+ All Categories
Home > Documents > Brachial Plexus Injuries

Brachial Plexus Injuries

Date post: 29-Oct-2014
Category:
Upload: prima-krishna-dharmawan
View: 127 times
Download: 0 times
Share this document with a friend
Popular Tags:
345
Transcript
Page 1: Brachial Plexus Injuries
Page 2: Brachial Plexus Injuries

Brachial Plexus Injuries

Edited by

Alain Gilbert MDInstitut de la MainParis, France

Published in association with theFederation of European Societiesfor Surgery of the Hand

M A RT I N � D U N I T Z

Page 3: Brachial Plexus Injuries

© 2001 Martin Dunitz Ltd, a member of the Taylor & Francis group

First published in the United Kingdom in 2001by Martin Dunitz Ltd, The Livery House, 7–9 Pratt Street, London NW1 OAE

Tel: +44 (0)20 7482 2202Fax: +44 (0)20 7267 0159Email: [email protected]: http://www.dunitz.co.uk

All rights reserved. No part of this publication may be reproduced, stored in aretrieval system, or transmitted, in any form or by any means, electronic,mechanical, photocopying, recording, or otherwise, without the prior permissionof the publisher or in accordance with the provisions of the Copyright Act 1988or under the terms of any licence permitting limited copying issued by theCopyright Licensing Agency, 90 Tottenham Court Road, London W1P OLP.

A CIP record for this book is available from the British Library.

ISBN 1-84184-015-7

Distributed in the USA by:Fulfilment CenterTaylor & Francis7625 Empire DriveFlorence, KY 41042, USAToll Free Tel: 1 800 634 7064Email: cserve@routledge_ny.com

Distributed in Canada by:Taylor & Francis74 Rolark DriveScarboroughOntario M1R G2, CanadaToll Free Tel: 1 877 226 2237Email: [email protected]

Distributed in the rest of the world by:ITPS LimitedCheriton HouseNorth Way, AndoverHampshire SP10 5BE, UKTel: +44 (0)1264 332424Email: [email protected]

Composition by Scribe Design, Gillingham, Kent, UK

This edition published in the Taylor & Francis e-Library, 2003.

ISBN 0-203-21640-7 Master e-book ISBN

ISBN 0-203-27262-5 (Adobe eReader Format) (Print Edition)

Page 4: Brachial Plexus Injuries

List of contributors v

I THE BRACHIAL PLEXUS

1 Anatomy of the brachial plexusAlexandre Muset i Lara, Carlos Dolz, andAlfonso Rodríguez-Baeza 3

2 Physical examinationTürker Özkan and Atakan Aydın 17

3 Radiological and related investigationsAlbert (Bart) CJ Slooff, Corneleus (Cees)WM Versteege, Gerhard Blaauw, andWillem JR van Ouwerkerk 31

4 Clinical neurophysiological investigationsJan W Vredeveld 39

II THE ADULT TRAUMATIC BRACHIALPLEXUS

5 EtiologyPanupan Songcharoen 47

6 Surgical techniques: neurolysis, sutures,grafts, neurotizationsMichel Merle and Aymeric Lim 51

7 Supraclavicular plexus injuriesJean Y Alnot 57

8 Complete palsyChantal Bonnard and Dimitri J Anastakis 67

9 Update on the treatment of adult brachialplexus injuriesHanno Millesi 77

10 Injuries of the terminal branches of thebrachial plexusRolfe Birch 91

11 The place of arthrodesisGiorgio A Brunelli 107

12 Palliative surgery: tendon transfers to theshoulder in adultsAydın Yücetürk 115

13 Palliative surgery: the elbow and forearmAlfred C Berger, Robert Hierner, and Lutz Kleinschmidt 123

14 Palliative surgery: the handJamal Gousheh 131

15 Palliative surgery: free muscle transfersKazuteru Doi 137

III OBSTETRICAL PARALYSIS

16 AetiologyJM Hans Ubachs and Albert (Bart) CJ Slooff 151

C O N T E N T S

Page 5: Brachial Plexus Injuries

17 Examination and prognosisHoward M Clarke and Christine G Curtis 159

18 Conservative treatment of obstetricalbrachial plexus palsy (OBPP) and rehabili-tationRobert S Muhlig, Gerhard Blaauw, Albert (Bart) CJ Slooff, Jan W Kortleve,and Alfons J Tonino 173

19 Surgical techniqueJose L Borrero 189

20 Indications and strategyAlain Gilbert 205

21 Results of repair to the obstetrical plexusAlain Gilbert 211

22 Results of surgery after breech deliveryGerhard Blaauw, Albert (Bart) CJ Slooff,and Robert S Muhlig 217

23 Palliative surgery: shoulder paralysisPiero L Raimondi, Alexandre Muset i Lara,and Elisabetta Saporiti 225

24 Palliative surgery: tendon transfers to theshoulder in childrenAydın Yücetürk 239

25 Medial rotation contracture and posteriordislocation of the shoulderRolfe Birch 249

26 Palliative surgery: elbow paralysisVincent R Hentz 261

27 Palliative surgery: pronosupination inobstetrical palsyEduardo A Zancolli (II) 275

28 Palliative surgery: forearm and handdeformitiesDavid C-C Chuang 293

29 Treatment of co-contractionRobert Hierner and Alfred C Berger 303

IV SPECIAL LESIONS

30 Traumatic brachial plexus injuries inchildrenAlain Gilbert and Christian Dumontier 315

31 War InjuriesJamal Gousheh 321

Index 326

iv CONTENTS

Page 6: Brachial Plexus Injuries

Jean Y AlnotService de Chirurgie Orthopédique etTraumatologiqueDépartement de Chirurgie de la Main et desNerfs PériphériquesCentre Urgences MainsHôpital Bichat-Claude Bernard46, rue Henri HuchardParis 75877Cedex 18France

Dimitri J AnastakisMinimally Invasive Surgery ProgramSurgical ServicesToronto Western Hospital399 Bathurst StreetFell Pavillion 4-140Toronto, ON M5T 2S8Canada

Atakan AydınDivision of Hand SurgeryPlastic and Reconstructive DepartmentIstanbul Medical FacultyIstanbul UniversityIstanbulTurkey

Alfred C BergerInternational Institute for NeuroscienceAlexis-Carrel-Strasse 4Hannover 30625Germany

Rolfe BirchRoyal National Orthopaedic HospitalPNI UnitBrockley HillStanmore, MiddlesexUK

Gerhard BlaauwDepartment of NeurosurgeryUniversity Hospital MaastrichtPO Box 58006202 AZMaastrichtThe Netherlands

Chantal BonnardService Universitaire de Chirurgie Plastique etReconstructivePermanence de LongeraieAvenue de la Gare 9CH-1003 LausanneSwitzerland

Jose L BorreroFlorida Hand Center610 Jasmine RoadAltamonte Springs, FL 32701USA

Giorgio A BrunelliVia Galvani 2625123 BresciaItaly

David C-C ChuangDepartment of Plastic and ReconstructiveSurgeryChang Gung Memorial Hospital199 Tung Hwa North RoadTaipei, Taiwan 105

Howard M ClarkeDivision of Plastic SurgeryDepartment of SurgeryUniversity of TorontoThe Hospital for Sick Children555 University Avenue, Suite 1524Toronto, ON M5G 1X8Canada

L I S T O F C O N T R I B U T O R S

Page 7: Brachial Plexus Injuries

Christine G CurtisDepartment of Rehabilitation MedicineUniversity of TorontoThe Hospital for Sick Children555 University Avenue, Suite 1524Toronto, ON M5G 1X8Canada

Kazuteru DoiDepartment of Orthopaedic SurgeryOgori Daaichi General HospitalOgoriYamaguchi-ken 754-0002Japan

Carlos DolzServicio de Cirugia Ortopedica y TraumatologiaHospital de ViladecansBarcelonaSpain

Christian DumontierInstitut de la MainClinique Jouvenet6, Place JouvenetParis 75016France

Alain GilbertInstitut de la MainClinique Jouvenet6, Place JouvenetParis 75016France

Jamal GoushehDepartment of Reconstructive and MicrosurgeryShahid Behesti University of Medical Sciences38 Keshavarz BoulevardTehran 14167Iran

Vincent R Hentz900 Welch RoadSuite 15Palo Alto, CA 94304USA

Robert HiernerClinic for Plastic, Hand and ReconstructiveSurgeryBurn CenterHannover Medical UniversitySchool of MedicinePodbielskistrasse 38030569 HannoverGermany

Lutz KleinschmidtClinic for Plastic, Hand and ReconstructiveSurgeryBurn CenterHannover Medical UniversitySchool of MedicinePodbielskistrasse 38030569 HannoverGermany

Jan W KortlevePlastic Surgery DepartmentAtrium Medical Center6401 CX HeerlenThe Netherlands

Aymeric LimHand and Reconstructive MicrosurgeryNational University Hospital5 Lower Kent Ridge RoadMain Building, Level 3Singapore 119074

Michel MerleInstitut Européen de la Main13, rue Blaise PascalF-54320 Maxéville-NancyFrance

Hanno MillesiUniversity of Vienna Medical SchoolLudwig-Boltzmann Institute of ExperimentalPlastic SurgeryLange Gasse 48A-1090 ViennaAustria

Robert S MuhligDepartment of RehabilitationAtrium Medical Center6401 CX HeerlenThe Netherlands

vi LIST OF CONTRIBUTORS

Page 8: Brachial Plexus Injuries

Alexandre Muset i LaraOrthopaedic Surgery DepartmentViladecans Hospital08840 Barcelona

Türker ÖzkanDivision of Hand SurgeryPlastic and Reconstructive DepartmentIstanbul Medical FacultyIstanbul UniversityIstanbulTurkey

Piero L RaimondiPlastic and Hand Surgery DepartmentLegnano Hospital20025 LegnanoMilanItaly

Alfonso Rodríguez-BaezaUnidad de Anatomia y EmbriologiaDepartamento de Ciencias MorfologicasFacultad de MedicinaUniversidad Autonoma de BarcelonaBarcelonaSpain

Elisabetta SaporitiPlastic and Hand Surgery DepartmentLegnano Hospital20025 LegnanoMilanItaly

Albert (Bart) CJ SlooffDepartment of NeurosurgeryUniversity Hospital MaastrichtPO Box 58006202 AZMaastricht,Department of NeurosurgeryFree UniversityPO Box 70571007 MBAmsterdamThe Netherlands,(contact address:)Rozenlaan 203620 LanakenBelgium

Panupan SongcharoenHand and Microsurgery UnitDepartment of Orthopaedic SurgeryFaculty of MedicineSiriraj HospitalMahidol UniversityBangkok 10700Thailand

Alfons J ToninoOrthopaedic Surgery DepartmentAtrium Medical Center6401 CX HeerlenThe Netherlands

JM Hans UbachsPijnsweg 336419 CJ HeerlenThe Netherlands

Willem JR van OuwerkerkDepartment of NeurosurgeryFree UniversityAmsterdamThe Netherlands

Corneleus (Cees) WM VersteegeDepartment of RadiologyAtrium Medical Center6401 CX HeerlenThe Netherlands

Jan W VredeveldDepartment of Clinical NeurophysiologyAtrium Medical CenterPostbox 44466401 CX HeerlenThe Netherlands

Aydın YücetürkClinic Plexus Tahran Cad. 3/3Kavaklıdere06700 AnkaraTurkey

Eduardo A Zancolli (II)Avenida Alvear 15351014 Buenos AiresArgentina

LIST OF CONTRIBUTORS vii

Page 9: Brachial Plexus Injuries
Page 10: Brachial Plexus Injuries

The Brachial Plexus

Page 11: Brachial Plexus Injuries
Page 12: Brachial Plexus Injuries

Introduction

The brachial plexus, on account of the progres-sive unions and divisions of its constituentnerves, is a more or less complex nerve forma-tion whose function is to innervate the muscles,articulations and tegument of the shoulder girdleand upper limb. In humans, the brachial plexusis formed from the anterior branches of the lastfour cervical nerves, and from the first thoracicnerve (Orts Llorca 1986). Additionally, it is irreg-ularly supplied by the C4 or T2 anteriorbranches. Such supply determines the so-calledplexus standards, pre- and post-fixed, respec-tively (Hovelacque 1927, Orts Llorca 1986,Williams 1998, Rouvière and Delmas 1999).Furthermore, it forms a union with the sympa-thetic cervical chain by means of communicatingbranches (Delmas and Laux 1933); it even formsa union with the paravertebral ganglia nodes ofthe second and third sympathetic thoracic chainby means of the Kuntz nerves (Orts Llorca 1986).

Topographically, the brachial plexus is locatedin the lower half of the neck’s lateral region,above the cervical pleural, projecting itself via aretro-infraclavicular path towards the axillarycavity (Fig. 1).

Taken as a whole, the brachial plexus presentsthe morphology of two triangles connected bytheir vertices (Hovelacque 1927). The upper trian-gle has a medial side oriented towards the spine,a base that coincides with the upper thoracicaperture, and an oblique lateral side orienteddownwards and outwards. The lower triangle,more irregular and mobile with arm movements(Lazorthes 1976), has a base coinciding with theemergence of the terminal branches of thebrachial plexus.

The most usual constitutional pattern for thebrachial plexus is through the formation oftrunks and cords (Feneis 2000). That is, the union

of the anterior branches of C5 and C6 forms thesuperior trunk. The union of the anteriorbranches of C8 and T1 forms the inferior trunk.The lower branch of C7, situated between thesetwo trunks, forms the middle trunk. Each of thetrunks subdivides into anterior and posteriorbranches. The posterior branches from the threetrunks unite to form the posterior cord, therebygiving place to the axillary (circumflex) and radialnerves. The lateral cord will provide the startingpoint to the musculocutaneous nerve and to theupper component of the median nerve. Themedial cord will provide the starting point to thelower component of the median nerve, the ulnarnerve and to the medial cutaneous nerves in thearm and forearm.

The suprascapular nerve, the posterior collat-eral branch of the superior trunk, is the mostlateral branch within the supraclavicular segmentof the brachial plexus, and its fibres have the

1Anatomy of the brachial plexusAlexandre Muset i Lara, Carlos Dolz, and Alfonso Rodríguez-Baeza

Figure 1

Ventral aspect of the brachial plexus.

Page 13: Brachial Plexus Injuries

function of innervating the supraspinatus andinfraspinatus muscles. It can be observed thatthe infraclavicular part of the brachial plexus isdivided into two planes between which theaxillary artery is located. The dorsal plane issimple, and is formed by the posterior cord. Theventral plane is more complex, and is made upof the lateral and medial cords.

Although the brachial plexus is essentiallydirected downwards and outwards, the directionof the different elements of which it is formedvaries significantly. Root C5 has a very obliquedirection downwards and outwards, whilst T1has an upward path. At the intervertebralforamen the C5 and C6 roots incline caudally onreaching the edge of the fissure of the spinalnerve made by the costo-transverse process ofthe corresponding cervical vertebrae. Root C7illustrates a direction coinciding with the plexusaxis. Roots C8 and T1 have an upward directionfrom the point of reflection realized in the pedicleof the vertebral arch and in the neck of the firstrib, respectively. The trunks have an oblique pathdownwards and outwards that causes them toconverge in the posterior edge of the clavicle.The angle of inclination is greater in the superiortrunk, and diminishes progressively in the medialand inferior trunks. In the infraclavicularsegment, their path is parallel, surrounding theaxillary artery. Nevertheless, they are verticallyinclined when the limb is in adduction andhorizontally inclined upon undergoing an abduc-tion of 90°.

Cervical supply to the brachialplexus

The brachial plexus’ cranial limit depends uponthe relationship established by roots C4 and C5in the constitution of the superior trunk. Kerr(1918) suggested a three-group classificationdepending upon the cervical supply to theplexus.

In the first group, a branch proceeding from C4anastomoses with C5, its size being highlyvariable, occasionally attaining diameters similarto the suprascapular nerve. Frequency for thishas been established at 63 per cent.

In the second group, the anterior C5 branch doesnot receive anastomotic branches, combining with

the anterior C6 branch in order to constitute thesuperior trunk. Frequency here is 30 per cent.

In the third group there is no C4 or C5 supply,but C5 contributes a nerve contingent to thecervical plexus. Frequency here is 7 per cent.

The supply of a significant nerve contingent byroot C4 to the brachial plexus defines a prefixedplexus. In such cases, part of the scapular girdle’sinnervating, which in classical patterns isattributed to the anterior C5 branch, may proceedfrom C4. This fact implies a cranial displacementof all the functions and innervations of the upperlimb, particularly when this supply coincides withthe scarcity of the T1 nerve contingent supply.Nevertheless, this aspect was neither defined norcorrelated in Kerr’s work (1918).

The cervical supply implies a cranial displace-ment of the brachial plexus axis, this being oneof the criteria used by certain researchers inorder to define a plexus as prefixed. However, nocompensation correlation has been establishedwith respect to the presence of cervical andthoracic supply, it being impossible to classifythe plexus as pre- or post-fixed in terms of thediameter of the nerves with which they areconstituted. Clinical work on quantifying nervecontingents supplied by each one of the roots(Slingluff et al 1996) defines a plexus as prefixedwhen C5 supply is greater than 15 per cent, andwhen that of T1 is less than 13 per cent; a plexusis defined as post-fixed when C5 supplies acontingent of between 6.8 and 12 per cent, andT1 from between 13.4 and 24.4 per cent.

With respect to the intra-plexus distribution,Slingluff et al (1996) consider that for prefixedplexus the superior trunk contributes to theformation of the posterior fasciculus in morethan 50 per cent, and to the innervating of thepectoral muscles in 75 per cent. The lateral fasci-culus receives no root C8 supply and less than 7per cent of the musculocutaneous nerve contin-gent comes from C7. These proportions areinverted in the post-fixed plexus, openingthereby a wide range of inter-individual possibil-ities and varieties in the plexus conformation.

Herzberg et al (1996) studied the radicularanastomoses between roots C4 and C5 on thebasis of 20 dissections. These researchersobserved that in five cases there was a branchfrom C4 to C5, in four cases a branch from C5 toC4, and in three of the cases there was noanastomosis.

4 THE BRACHIAL PLEXUS

Page 14: Brachial Plexus Injuries

Attention should also be focused on therelation between the phrenic nerve and the C4and C5 roots. The origin in C4 frequentlypresents anastomosis with C5, its neurolysisalways being possible in cases of very proximalresection for C5 as donor root in plexus injuries,without this causing any perceivable alteration indiaphragmatic function.

Anatomy of the foraminateregion

Knowledge of the topography, relationship anddistribution at a foraminate level of the spinalnerves as well as the path within the fissure fromthe transverse process of the cervical vertebraeis of fundamental practical interest to surgicalrepair of brachial plexus injuries. Access to thesupraclavicular–extrascalenus region of thebrachial plexus is undertaken via a lateral–cervi-cal approach. Nevertheless, it is the inter- andpre-scalenus dissection that allows us tohighlight the radicular segments that are usefulas donors, and to identify the posterior branchfor its intra-operational stimulation that willdefine for us, along with the remaining comple-mentary explorations, the condition of theanterior branch and its validity to the procedureof microsurgical reconstruction.

The intervertebral foramen is a space definedby the imposition of two adjacent vertebrae. Atthe cervical level, it is determined by the follow-ing anatomical elements: cranially and caudallyby the transverse process of the superior andinferior vertebrae, respectively; ventrally byunco-vertebral articulation and the inter-vertebraldisk; dorsally by the upper articular process(Testut and Latarjet 1979).

The transverse process of the cervical verte-brae is projected ventro-laterally, taking itsanterior starting point in the vertebral pedicle,and its posterior starting point in the osseouscolumn oriented vertically, culminating on thesuperior and inferior levels in articular surfacetracks. It presents two lateral bodies and acentral canal or fissure through which the spinalnerve runs. In its path proximal to the spinalnerve, with its anterior and posterior branches, itrelates posteriorly with articular processes andanteriorly with the vertebral vascular-nerve

parcel running through the transverse foramen.Upon reaching the spinal nerve, the externalmargin of the articular process gives rise to theposterior branch dorsally surrounding the articu-lar process in order to distribute itself in theposterior paravertebal musculature, in thetegument and in the articular capsule itself,providing a mixed sensory and motor innerva-tion. The intra-operational stimulation of thisbranch offers valuable information regarding thefunctional state of the spinal nerve (Fig. 2).

The anterior branch in the fissure is locatedbetween the anterior and posterior intra-transversal muscles. In this short path the nerve

ANATOMY OF THE BRACHIAL PLEXUS 5

Figure 2

Anatomy of the intervertebral foramen. (1) Spinal nerve; (2)vertebral pedicle; (3) anterior tubercle of transverseprocess.

21

3

Page 15: Brachial Plexus Injuries

receives the insertion of the transverse-radicularligament, which originates in the superior trans-verse process and, through an oblique out-to-in/upward–downward path terminates by fusingitself with the epineuro of the subjacent spinalnerve’s upper section (Fig. 3).

From the vascular point of view, the spinalnerves connect with the arteries whose functionis the arterial irrigation of the spinal cord(Rodríguez-Baeza and Doménech-Mateu 1993).The radicular and radiculo-medullar arteries ofthe inferior cervical region are branches of theascending cervical artery, of the costal–cervicaltrunk and of the vertebral artery. Supply intended

for medullar vascularization reaches the forami-nate space by means of an oblique upward andbackward path, connecting with the spinal nerveat the front and with the inter-transverse muscleat the back (Fig. 4).

In the external margin of the transverseprocesses, the anterior branches of the spinalnerves connect with the points of origin for thescalenus muscles, so as to subsequently enterthe inter-scalenus space (hiatus scalenicus),delimiting the anterior and middle scalenusmuscle.

The foraminal anatomy from C4 to C7 facili-tates the systemization of the radicular surgical

6 THE BRACHIAL PLEXUS

Figure 3

Foraminal anatomy of C5 and C6 roots (posterior view). (1)Radiculo-medular artery; (2) transverse-radicular ligament;(3) posterior tubercle of transverse process.

Figure 4

Arterial relationships of the brachial plexus. (1) Ascendingcervical artery; (2) vertebral artery; (3) transverse cervicalartery; (4) suprascapular artery.

Page 16: Brachial Plexus Injuries

approach, from distal to proximal, through thelocalization of the transverse process’s posteriortubercle, the dis-insertion of the middle andposterior scalenus muscles, and the section ofthe posterior inter-transverse muscle. This proce-dure highlights the nerve path that runs from theinter-vertebral foramen to the inter-scalenusspace without risk of injury to the arterial verte-bra. Additionally, we can expose the posteriorbranch, approximately 10 mm of the C5 and C6anterior branches and some 15 mm of the C7anterior branch path. These paths are generallyprotected at this level by the transverse-radicularligament. The relationship that the anterior C5branch maintains with the phrenic nerve servesto distinguish it in a certain manner from C6 andC7. The proximal surgical dissection of C5implies the dissection of anastomotic phrenicbranches in their distinct varieties (commentedon above). It is important, in this procedure, tobear in mind that the phrenic nerve receives itsprincipal nerve contingent from C4, and there-fore, when it requires a proximal resection of C5in order to obtain correct proximal stumpsegment quality, it can be sacrificed withoutdetriment to diaphragmatic function, on thecondition that a correct neurolysis and neuro-tomy, exclusive to the anastomotic branch, beundertaken. This surgical action will facilitateboth the radicular resection of C5 as well as itsproximal dissection without risk of injury to thephrenic nerve.

The foraminate anatomy of roots C8 and T1differ both in respect to their relationships andalso with regard to the means of radiculo-verte-bral union. At a vertebral level, the foramenpresents distinct limits due to the morphologicalmodification of the transverse process. In thethoracic vertebrae, the process is implantedwithin the vertical osseous column configuringthe articular process, orienting itself in a poste-rior–lateral direction. In this way, the foramen isdelimited cranially and caudally by the superiorand inferior pedicle respectively, dorsally by thearticular process and ventrally by the poste-rior–lateral margin of the superior vertebral bodyand by the inter-vertebral disk. Anterior relation-ships with the vertebral artery do not exist, andthe relationship that C8 and T1 maintain in theirimmediately extra-foraminate path are estab-lished with the neck of the first and second ribs.The markedly upward direction of the anterior T1

branch towards the inter-scalenus space bringsabout the relationship with the neck of the firstrib. Unlike what happens at higher levels, thereare no transverse–radicular ligaments here,thereby causing the considerable reduction ofresistance to traction; for this reason, radicularavulsions are more frequent. In the pre-scalenuspath, C8 and T1 are found in the Sébileauscalenus–vertebro-pleural space (Delmas andLaux 1933), this being an anatomical spacedelimited on the outside by the transverso-pleural ligament, on the inside by the vertebro-pleural ligament, on the underside by theposterior slope of the cervical pleura, and from

ANATOMY OF THE BRACHIAL PLEXUS 7

Figure 5

Waldeyer’s vertebral triangle. (1) Star-shaped node; (2)anterior scalenus muscle; (3) internal thoracic artery; (4)vertebral artery.

Page 17: Brachial Plexus Injuries

behind by the posterior extremity of the first tworibs and the spine. Upon surrounding the neckof the first rib, T1 connects with the star-shapednode, and is crossed by the superior intercostalartery. It moves outwards between the fasciculaof the costal–pleural ligament, becomingseparated from the subclavian artery by thefibres of the transverso-pleural ligament in itsinsertion into the cervical pleura.

The cervical–thoracic or star-shaped node isthe result of the union of the inferior cervicalnode with the first thoracic node. Its morphologyis levelled, being irregularly rounded, star-shaped or in the form of a half-moon (Testut andLatarjet, 1979). Its length is approximately 8 mmand it can extend itself from the transverseprocess of the seventh cervical vertebra to theneck of the second rib. The intimate relationshipthat it maintains with the lower part of thebrachial plexus justifies the appearance of aClaude–Bernard–Horner syndrome in proximalinjuries of the inferior plexus roots (Fig. 5).

Anatomy of the scalenus region

In the supra-clavicular region of the brachialplexus neck’s lateral region, there are connec-tions with the scalenus muscles. These muscles

form an irregularly triangular mass that extendsfrom the transverse cervical processes to the firsttwo ribs.

The anterior scalenus muscle originates in theanterior tubercles of the third to sixth cervicalvertebrae. The four portions, tendinous in origin,unite in a fleshy body that, orienting itselfdownwards and outwards, terminates by insert-ing itself within the first rib’s Lisfranc tubercle bymeans of a cone-shaped tendon. The middlescalenus muscle originates in the posteriortubercles of the last six cervical vertebrae, andterminates by inserting itself within the upperside of the first rib, behind the anterior scalenus.The posterior scalenus originates in the posteriortubercles of the fourth and sixth cervical verte-brae and terminates by inserting itself within theupper edge of the second rib.

The position of the scalenus muscles allowsfor delimiting a triangular space on the lowerbase, at the level of the first rib, known as thescalenus hiatus. The anterior margin is oblique,and the posterior is vertical, corresponding to theanterior and middle scalenus muscles respec-tively. Furthermore, the anterior scalenus musclehelps to delineate what is known as Waldeyer’svertebral triangle. The posterior scalenus muscleis separated from the middle muscle by an inter-stice in which we may locate the large thoracicnerve (Bell’s nerve) (Figs 6 and 7).

8 THE BRACHIAL PLEXUS

Figure 6

Scalenic anatomy. (1) Phrenicnerve; (2) intermediate node; (3)scalenus anterior muscle; (4)subclavian artery; (5) first rib.

Page 18: Brachial Plexus Injuries

There are multiple anatomical variations thatmay be observed in the scalenus muscles (Testutand Latarjet 1979), but, for our purposes, weshall only refer to those that directly affectrelationships with the brachial plexus.

The muscle referred to as the middle (or inter-mediate) scalenus, is a supernumerary muscularfasciculus that extends among the transverseprocesses of the sixth or seventh cervical verte-brae up to the first rib, interposing itself amongstthe brachial plexus and the subclavian artery inthe scalenus hiatus. The so-called Albinus andtransverso-pleural muscles may be consideredas variations of the middle scalenus. The Albinus

accessory muscle proceeds from the fourth, fifthand sixth cervical vertebrae, and reaches as faras the first rib, whilst the transverso-pleuralmuscle proceeds from the seventh cervical verte-bra, reaching the cervical pleural.

The low original points for the anteriorscalenus muscle leave the extra-foraminate C5path exposed, illustrating, in these cases, a pre-scalenus topography. In proximal radicularinjuries this consideration is important in ordernot to limit the proximal dissection to the inter-scalenus vertex, which may have an exclusiverelationship with C6. In other cases, we haveobserved C5 paths through the anterior scalenusmuscle.

Tendinous insertions in the first rib of theanterior and middle scalenus muscles may be incontinuity via a fasciculus referred to as ‘thescalenus’ sickle’. This formation closes thescalenus hiatus, being a cause of compressionfor the subclavian artery and the lower part ofthe plexus; this mechanism may be accentuatedwhen there are inter-scalenus muscular anoma-lies.

The anterior branches of the C3, C4, C5 and C6nerves give out direct branches for the anteriorscalenus muscle. The posterior and middlescalenus muscles receive branches from the C3,C4 and dorsal scapular nerves, this latter alsobeing known as the rhomboid nerve.

Through the anterior scalenus muscle, thebrachial plexus maintains relationships withanatomical structures that must be preserved inthe anterio-lateral approaches of the inter-scalenus space. These structures are, in a down-up description, the subclavian vein, thesubclavian muscle and the omohyoid muscle.The phrenic nerve and the ascending cervicalartery are located vertically in the ventral surfaceof the muscle, whilst the transverse cervical andsuperior scapular arteries cross this facetransversally. The inferior-medial part of theanterior scalenus muscle tendon connects withthe cervical pleural and is ligament supportsystem (a.k.a. Sébileau’s).

In the surgical dissection of the plexus’ inter-scalenus path, we need to bear in mind thepresence of the inter-scalenus artery. Its origingenerally lies in the subclavian artery, althoughon occasions it proceeds from the subscapular orcostocervical arteries. Its distribution is by meansof muscular branches for the scalenus muscles,

ANATOMY OF THE BRACHIAL PLEXUS 9

Figure 7

Intrascalenic anatomy. (1) Middle scalenus muscle; (2)Bell’s nerve; (3) anterior scalenus muscle (dis-inserted).

Page 19: Brachial Plexus Injuries

and by means of radicular branches for thebrachial plexus itself. Its muscular supplies arecomplemented by unnamed arterioles proceed-ing from the subclavian, dorsoscapular andcostocervical arteries.

Anatomical studies of NMR anatomy correla-tion for the pre- and inter-scalenus spaces haveallowed us to objectify the presence of fibro-muscular structures interposed between thesubclavian artery and the brachial plexus, as wellas the presence of pre-scalenus roots. Never-theless, regular clinical resolution does notdefine the ligament formations in the region ofthe thoracic inlet, obliging us therefore to reviewthis surgically in approaches for compressivesyndromes in the brachial plexus.

Anatomy of the extra-scalenusregion

In the lateral region of the neck, we find the poste-rior cervical triangle, delimited caudally by theclavicle, medially by the sternocleidomastoid andanterior scalenus muscles, and laterally by thetrapezius muscle. This triangular space, essen-tially clavicular, is subdivided by the presence ofthe omohyoidal muscle, the upper region beingomotrapezoidal and the lower being omoclavicu-lar or greater supraclavicular fossa (Fig. 8).

In order to accede to the plexus in this region,after incising the skin and the subcutaneouscellular tissue, the platysma colli muscle isexposed. This muscle is included in the divisionof the superficial cervical fascia, owing to whichits deep face rests on the fascia itself.

The superficial cervical fascia originates in theanterior middle raphe of the neck from where itmoves outwards in order to divide itself at thelevel of the sternocleidomastoid, and to form themuscle sheath. On its posterior edge, the twolayers unite and the fascia covers the greatersupraclavicular fossa only to divide once againon the medial edge of the trapezius muscle. Thisplane is separated from the medial cervicalfascia by the Meckel’s adipose mass, throughwhich runs the external jugular vein (Testut andLatarjet 1979).

The medial cervical fascia (the pre-tracheal layerof the cervical fascia) runs between the twoomohyoid muscles, reaching the semi-lunarnotches. In the mid-line it reaches the posterior lipof the sternal notch. At the clavicular level, itinserts into its posterior edge, surrounding thesubclavian muscle. The fascial expansion thatextends between the subclavian muscles and thecoronoid process continues with the fascia of theaxillary cavity. Therefore, this fascia reaches thesuperior orifice of the thorax, the sternum, theclavicles, first ribs, pericardium and subclavianfascia. It connects, via its deep face, with the

10 THE BRACHIAL PLEXUS

Figure 8

Anatomy of extrascalenus region.

Page 20: Brachial Plexus Injuries

brachial plexus and vascular structures of the neck,which runs superficially to the deep cervical fascia.

The cellular adipose layer extends cranially tothe omotrapezoidal triangle via a layer thatunites the superficial and deep cervical fasciaewith the medial cervical fascia. The externalbranch of the (accessory) spinal nerve runswithin this layer, as well as the transverse arteryof the neck, the suprascapular artery and thedorsal artery of the scapula. The path taken bythese arteries to the medial cervical fascia tendsto be deep, connecting directly with the brachialplexus. The superficial jugular vein remainssuperficial on this plane, whilst the sensorynerves in the cervical plexus perforate the cellu-lar adipose layer and that of the cervical fasciain order to situate themselves subcutaneously,and to distribute themselves within theanterior–lateral region of the neck and shoulder.

The suprascapular artery, a branch of thethyro-cervical trunk, crosses the anterior–medialsection of the tendon pertaining to the anteriorscalenus muscle, in order to subsequently locateitself deeply within the omohyoid muscle, and toreach the transverse scapula ligament, to whichthe artery takes an upper route.

The dorsal artery of the scapula, a branch ofthe inter-scalenus path of the subclavian artery,leaves the scalenus hiatus and locates itselfamong the middle and upper trunks of thebrachial plexus. It then crosses ventral and later-

ally to the middle and posterior scalenus musclesand reaches the muscular mass pertaining to thescapula lever, where it gives out the sub-trapezebranch and locates itself below the rhomboids.

The subclavian vein, when passing through thespace existing between the clavicle and the first rib,adheres to the fascia of the subclavian muscle inaddition to being united to the pre-tracheal layer.

The upper, middle and lower trunks areorganized and constituted in the extra-scalenusregion of the brachial plexus. The anatomicalvariations of major surgical relevance for thereconstruction of the plexus, or in canalicularsyndromes, correspond to the distribution of C7with respect to the anterior plane of the brachialplexus, upper and lower trunk. The complex andvariable distribution of the anterior C7 fibres hasallowed the establishment of a Gilbert’s classifi-cation of three types of plexus (A, B and C), whichexplain situations of apparent clinical paradox.

Anatomy of the (axillary)infraclavicular region

The brachial plexus reaches the vertex of theaxillary cavity, passing behind the clavicle. It isin this infraclavicular portion where the fasciclesand terminal branches of the plexus areorganized and structured (Fig. 9).

ANATOMY OF THE BRACHIAL PLEXUS 11

Figure 9

Anatomy of infraclavicular region.(1) Upper trunk; (2) middle trunk;(3) lateral cord; (4) medial cord; (5)posterior cord.

12 3

45

Page 21: Brachial Plexus Injuries

The axillary cavity is covered by a deep fascialevel that runs towards the coracobrachialis andto the axillary edge of the scapular from thepectoral muscle, subdividing itself into a superior(or semi-lunar) portion, and a lower (or scapular)portion (Testut and Latarjet 1979).

The semi-lunar portion is the part of Richet’sclavicular–coracoaxillary fascia, or Rouvière’sclavipectoral–coracoaxillary fascia (Paturet, 1951),which contributes to the Gerdy’s ligament supportsystem. This fibrous range has its vertex in thecoronoid process, its internal edge reaches thefascia of the pectoral minor, its lower edgereaches the skin of the axillary hollow, and itsexternal edge reaches the fascia of the armthrough the coracobrachialis and the short headof the biceps.

The scapular portion is the continuation ofGerdy’s ligament. It covers the anterior face ofthe trapezius muscle up to its scapular inser-tion, where it runs anteriorly to the subscapularmuscle, and inferiorly it covers the teres majorand the latissimus dorsi muscles. Its externaledge, close to the glenoid cavity, separatesfrom the scapula, freeing itself to fuse with thefibrous sheath of the coracobrachialis. This pathdetermines the axillary Langer’s arch, aninferior–external socket, through which a vascu-lar nerve structure runs from the axillary cavityof the arm (Paturet 1951). On occasions, anaccessory muscular fascicle (of a flat or trian-gular morphology) may be found between thelatissimus dorsi and the pectoral majormuscles, known as Langer’s muscle. On otheroccasions, there is a dense fibrous layer, or itmay be connected with the coracobrachialis orthe brachial biceps muscles, representing, inthese cases, an incomplete formation of thestructure in question.

The fascia of the axillary cavity’s internal wallcovers the anterior serratus muscle, being acellular adipose layer in which the large thoracicnerve (Bell’s nerve) is located.

The fascia of the axillary cavity’s anterior wallin direct relation to the brachial plexus is Richet’sclavicular–coracoaxillary fascia. Dense and resis-tant, it is perforated by the nerves and vesselsthat supply the pectoral major muscle. Itproceeds cranially from the subclavian musclesheath and from the coronoid processes. Itprojects itself towards the clavipectoral triangle,dividing itself with respect to the pectoral minor

muscle, subsequently reaching the axillarybase’s superficial fascia and the brachial fascia atthe level of the coracobrachialis. The expansionof the dermis constitutes the suspensoryligament of the axilla, triangular in form, with itsvertex in the coronoid process, its base at thelevel of its dermal insertion, an external edge incontinuity with the fascia of the coracobrachialismuscle and its internal edge in continuity withthat of the pectoral minor muscle.

The lateral cord of the brachial plexus is madeup of the union of the anterior branches from thesuperior and middle trunks. Many variationshave been described, but their frequency isscarce. On occasions the middle trunk is suppliedfrom the lower trunk before the point of originof its anterior branch; it may even unite with theanterior branch itself. On other occasions, themiddle trunk receives anastomosis from theposterior branch of the superior trunk before itsdivision (Fig. 9).

In certain cases, the lateral fascicle is directlyconstituted by the union of the C5, C6 and C7anterior nerve branches. The non-participation ofthe middle trunk in the formation of this fascicleimplies that, for such patients, the upper medianand the musculocutaneous nerves originate inC5 and C6, with supply from C4 in cases with apre-fixed plexus.

The medial fascicle is formed from the anteriorbranch of the lower trunk. There is, occasionally,union of the C8 anterior branch with the wholeof T1. This may also receive supply from C7. Afascicle making up the inferior median rarelydetaches itself from the nerve branch to movetowards the posterior fascicle.

The posterior cord is constituted by the unionof the posterior branches from the superior,middle and inferior trunks. On many occasions,it may be observed that the posterior branchesof the upper and middle trunks are joined, consti-tuting thereby a common fascicle to be subse-quently united with the posterior branch of thelower trunk. On other occasions, it is the poste-rior branches of the middle and lower trunk thatare first joined, being then followed by the poste-rior branch of the upper trunk. Only very rarelycan we observe the convergence of all threebranches simultaneously.

Other noteworthy variations (though infre-quent) are the following: additional supply fromthe upper and/or lower trunks via double or triple

12 THE BRACHIAL PLEXUS

Page 22: Brachial Plexus Injuries

branches; supply from the lower trunk proceed-ing from C8 without the participation of T1;branches proceeding from the lateral cord; andposterior branches proceeding directly from C5and C6 that join with the middle trunk in orderto subsequently anastomose with the inferiortrunk (Kerr 1918). Another interesting variation isthat in which the posterior cord only gives riseto the radial nerve.

The relationships maintained by the cords withthe vascular structures in the axillary cavitydetermine their topographical denomination. Theaxillary artery is located among the three fasci-cles, being entirely surrounded at the front bythe median nerve via supply from the lateral andmedial cords in the lower middle part of theaxillary cavity, in the lower retro-pectoral region.

Collateral branches of thebrachial plexus

These are topographically classified into supra-clavicular and infraclavicular, and have thefunction of innervating the muscles of thetronco-scapular apparatus (Orts Llorca 1986).They originate directly from the lower branchesof the medulla nerves forming the brachialplexus, or from its trunks or fascicles. The point

of origin may lie on the anterior or posterior face,depending upon the ventral or dorsal ontogenicsignificance, respectively (Fig. 10).

The supraclavicular branches are:Nerves for the deep muscles of the neck, that is,for the scalenus, longus colli and inter-transversemuscles. These proceed directly from theanterior branches of the lower cervical nerves atthe level of the intervertebral foramen.The dorsal nerve of the scapula. This originatesin the posterior face of the anterior C4 and C5nerve branches, usually via a single trunk. It runsbackwards, crossing the middle scalenus musclein order to reach the angular scapula muscle,which it innervates in its caudal fascicles. It thenconnects with the dorsal artery of the scapulaand innervates the rhomboid muscle.The long thoracic nerve. This is classicallyreferred to as Bell’s external respiratory nerve. Itoriginates in the posterior C5 to C7 faces,although a C7 component only exists in 40 percent of cases. The C5 component may originatewithin the dorsal nerve of the scapula. The twoupper branches cross the middle scalenusanastomosing at this level, or laterally to it. Theresulting branch descends behind the brachialplexus and the first portion of the axillary artery.It crosses the upper edge of the anterior serratusmuscle, descending via the lateral face of thethorax in the angle that is formed by the

ANATOMY OF THE BRACHIAL PLEXUS 13

Figure 10

Anatomy of the terminal branches.(1) Suprascapular nerve; (2) musculo-cutaneous nerve; (3) ‘V’ of mediannerve; (4) ulnar nerve; (5) radialnerve; (6) axillary nerve; (7) pectoralisnerves.

7

Page 23: Brachial Plexus Injuries

subscapular and anterior serratus muscles.When there is a C7 component, this emergesthrough the middle scalenus muscle. The longthoracic nerve gives off innervation branches toeach one of the digitations of the anterior serra-tus muscle, as the muscle’s upper part is inner-vated by C5 fibres, the middle part by C6 fibresand the lower part by C7 fibres (Lazorthes 1976).The subclavian nerve. This originates in theanterior C5 face or in the point of union betweenC5 and C6 (upper trunk). Descending obliquely infront of the plexus and the anterior scalenusmuscle and on the outside of the phrenic nerve.It has anastomosis with this latter nerve, givingrise to the accessory phrenic nerves (Hovelacque1927), and cranially to the subclavian vein, itmoves towards the subclavian muscle that itinnervates.The suprascapular nerve. This is one of the firstbranches leaving the brachial plexus. It proceedsfrom the upper trunk or directly from C5,although on certain rare occasions (particularlyin prefixed plexus) it may proceed from C4,following a C4–C5 union. It runs downwards andoutwards following the deep face of the omohy-oid muscle in order to reach the semilunar notch,passing the supraspinous fossa below the uppertransverse scapular ligament. It distributes itselfthroughout all the supra- and infraspinousmuscles.

The infraclavicular branches are:The pectoral nerve. This may originate in theanterior divisions of the upper and middle trunksor directly from the lateral fascicle via a singlebranch. It crosses in front of the axillary arteryand vein, passing through the clavipectoralfascia, distributing itself in the clavicular fascicleof the pectoral major muscle. It gives out ananastomotic branch that participates in theformation of the pectoral loop situated in frontof the first portion of the axillary artery, aroundthe point of origin for the acromio-thoracicartery. Fibres for the pectoral minor originatefrom the loop.The medial pectoral nerve. This proceeds fromC8 to T1 at the level of the medial fascicle, lyingbehind the axillary artery. It runs forwards by theinterstice between the axillary artery and vein,joining with the lateral pectoral nerve, under theacromio-thoracic artery, participating in thepectoral loop. It gives off innervation branches tothe pectoral minor muscle and to the sternal

fascicle of the pectoral major. The branchesleading to the pectoral major muscle reach theirdestination either by crossing the clavipectoralfascia or through the muscular fibres of thepectoral minor itself (Rouvière and Delmas 1999).The subscapular nerves. There are two or threebranches that proceed from the posterior cord ofthe brachial plexus, although on occasions theupper branch proceeds from the upper face ofthe upper trunk (Lazorthes 1976). Their functionis the innervating of the subscapular and teresmajor muscle.The thoraco-dorsal nerve. This belongs to thegroup of subscapular nerves, but is identified byits long pathway, parallel to the axillary edge ofthe scapula, accompanying the subscapularvessels. It innervates the latissimus dorsi andteres major muscles.

Terminal branches of thebrachial plexus in the axillaryregion

The terminal branches of the brachial plexus areclassified into ventral and dorsal groups, andproceed from the lateral, medial and posteriorfascicles, respectively. The posterior fasciclegives rise to the axillary (circumflex) and radialnerves. The axillary nerve is considered by someresearchers to be a collateral branch to theplexus because of its distribution in muscles ofthe shoulder girdle (Orts Llorca 1986). It carriesC5 and C6 fibres and runs downwards andoutwards, applied to the anterior face of thesubscapular muscle, to which it may provideinnervation, accompanying the posteriorhumeral circumflex artery. It leaves the axillarycavity by the Velpeau quadrilateral.

The radial nerve is the largest nerve in thebrachial plexus, and carries fibres from C5 to T1roots in most cases (Orts Llorca 1986, Feneis2000). It is the most posterior and internalelement in the axillary vascular nerve structures,lying behind the axillary artery and the mediannerve. It is located between the axillary vein andthe cubital nerve (which lie outside), and themusculocutaneous nerve (which lies inside). Itleaves the axillary cavity in connection with thelower edge of the latissimus dorsi tendon.

14 THE BRACHIAL PLEXUS

Page 24: Brachial Plexus Injuries

The ventral terminal branches are:The musculocutaneous nerve. This proceedsfrom the lateral fascicle and carries C5 fibres toC7. It runs downwards and outwards, lying later-ally with respect to the median nerve, andanterio-laterally with respect to the axillaryartery. In its path it crosses circumflex humeralvessels and perforates the coracobrachialismuscle upon reaching it, hence it is also referredto as Casserius’ perforating nerve.The median nerve. This is formed by the junctionof two roots, one lateral and one medial,proceeding from the lateral and medial fascicles,respectively. It carries C6 fibres to T1. The unionof the two roots gives rise to the V-shape of themedial nerve (Paturet 1951), located in front ofthe axillary artery, in the lower edge of the lesserpectoral muscle. The anterior humeral circumflexartery lies behind the nerve. It leaves the axillarycavity (Rouvière and Delmas 1999) in order tosituate itself within Cruveilhier’s brachial duct.The cubital nerve. This proceeds from the medialfascicle of the brachial plexus, and carries C8 andT1 fibres. It may occasionally receive C7 fibresproceeding from the lateral fascicle (Lazorthes1976). It is located in the anterior face of theinterstice separating the artery from the axillaryveins, amongst the median nerve and medialcutaneous nerves of the forearm. Behind this arethe subscapular and thoraco-dorsal vessels andnerves.

The medial cutaneous nerves of the arm andforearm originate from the medial fascicle, andhave been considered as sensory branches ofthe cubital nerve (Orts Llorca 1986). The arm’smedial cutaneous nerve is situated more deeplythan the forearm’s medial, and establishesanastomosis with the second intercostal nerve,giving rise to the so-called Hyrtl’s intercosto-brachial nerve (Lazorthes 1976). Both nerves areexclusively sensory and carry C8 and T1 fibres.

References

Bonnel F (1991) Les nerfs de la racine du membresupérieur. In: Bonnel F, Chevrel JP, Outrequin G, eds.

Anatomie Clinique. Les Membres. Springer-Verlag:Paris.

Delmas J, Laux G (1933) Anatomie médico-chirurgicaledu système nerveux végétatif (sympathique etparasympathique). Masson et Cie, éd: Paris.

Feneis H (2000) Nomenclatura Anatómica Ilustrada, 4thedn. Masson: Barcelona.

Herzberg and cols. (1996) Surgical approach of thebrachial plexus roots. In: Alnot JY, Narakas A, eds.Traumatic Brachial Plexus Injuries. Monographie GEM.Expansion Scientifique Française: Paris.

Hovelacque A (1927) Anatomie des Nerfs Craniens etRachidiens et du Système Grand Sympathique chezL’homme. Gaston Doin et Cie, éd: Paris.

Kerr A (1918) The brachial plexus of nerves in man, thevariations in its formation and branches, Am J Anat23(2).

Lazorthes G (1976) Sistema Nervioso Periférico. Toray-Masson: Barcelona.

Orts Llorca F (1986) Anatomia Humana, 6th edn, Vol 3.Científico Médica: Barcelona.

Paturet G (1951) Traité d’Anatomie Humaine, Vol II.Membres Supérieur et Inférieur. Masson et Cie, éd:Paris.

Rodríguez-Baeza A, Doménech-Mateu JM (1993)Anatomía de las arterias que irrigan la región cervicalde la Médula Espinal Humana. In: Bordas Sales JL, ed.Artrosis Cervical. Complicaciones Neurovasculares, 1stedn. Editorial Jims: Barcelona.

Rouvière H, Delmas A (1999) Anatomía Humana, 10thedn, Vol 3. Masson: Barcelona.

Singluff C. and cols. (1996) Surgical Anatomy of theHuman Brachial Plexus. In: Alnot JY, Narakas A, eds.Traumatic Brachial Plexus Injuries. Monographie GEM.Expansion Scientifique Française: Paris.

Testut L, Latarjet A (1979) Tratado de AnatomíaHumana, 9th edn, Vol 1. Salvat: Barcelona.

Williams PL (1998) Anatomía de Gray, 38th edn.Harcourt Brace: Madrid.

ANATOMY OF THE BRACHIAL PLEXUS 15

Page 25: Brachial Plexus Injuries
Page 26: Brachial Plexus Injuries

Introduction

Evaluation of brachial plexus injuries needs anunderstanding of many factors before a manage-ment plan can be tailored. The site of the lesion,nature and degree of the injury and expectedprognosis are part of this diagnostic process.

In the physical examination of the patient, thepurpose is to determine the type and the site ofthe brachial plexus injury. This is performed bya careful clinical examination including musclefunction (Table 1), (Kendall et al. 1993, Tubiana

et al. 1995, Clarkson 1999), sensorial examinationand specialized testing. At initial examination,the nature of the injuries (traction, penetrating,etc.), the entrance and exit wounds of penetrat-ing injuries, amount of bleeding at the time ofthe injury, and associated fractures are recorded.All the muscles of the upper extremity and shoul-der girdle innervated by the brachial plexus mustbe examined and graded on a scale from 0 to 5by the manual muscle tests according to theMedical Research Council scale on a brachialplexus chart (Alnot 1995, Boome 1997a) (Fig. 1).

2Physical examinationTürker Özkan and Atakan Aydın

Figure 1

Brachial plexus muscle test chart.

Page 27: Brachial Plexus Injuries

Sensibility tests are performed for eachdermatome, including cervical and brachialplexus. Pain and temperature appreciation, staticand moving two-point discrimination, constanttouch and vibration with a tuning fork at 30 and256 cycles per second, and the ninhydrin testdescribed by Moberg (Aschan and Moberg 1962)are tested and recorded (Fig. 2).

Physical examination is repeated, hours ordays later, and changes in the clinical findingsmust be recorded by a different colored ink onthe same chart or on a fresh chart. The functionalgrading of nerve compression compares themotor and sensory losses if they do not corre-late. Such a discrepancy may result if there isnerve compression rather than nerve division orrupture (Millesi 1984; Boome 1997b) (Table 1).

Severe neuropraxias may persist for up to 6–8weeks. Root avulsions, ruptures or neuromas incontinuity are possibilities when a specific rootfunction is completely absent, or each pathologymay be found at different root levels in the samepatient (Boome 1997b).

The paralysis of some muscles can givespecific information related to the level of theinjury (Table 2).

A serratus anterior muscle paralysis in associ-ation with total or upper trunk palsy suggests aC5 and C6 root avulsion, as its nerve supplyarises close to the vertebral foramen; whileparalysis of the levator scapulae, rhomboids anddeep muscles of the neck points to a proximal

injury or avulsion of C4 and C5 roots. Adiaphragmatic palsy suggests a C4 avulsion: ifassociated with a C5 and C6 palsy, it is likely thatC5 and C6 roots are ruptured close to theforamen within the vertebral canal.

If brachioradialis and teres major muscle paral-ysis is associated with paralysis of supraspinatus,infraspinatus, deltoid, teres minor and bicepsmuscles, then upper trunk injury is likely.However, if the brachioradialis and teres majormuscles are intact, a more peripheral injury of thenerves to the shoulder abductors and external

18 THE BRACHIAL PLEXUS

Figure 2

Brachial plexus sensibilityassessment chart.

Table 1 Functional grading of nerve compression

Page 28: Brachial Plexus Injuries

rotators and elbow flexors is likely. A normalsupraspinatus muscle function excludes a C5 rootavulsion or rupture in a normally fixed plexus(Bonnard and Narakas 1997).

Sensory evaluations also give some cluesabout the level and pattern of brachial plexusinjury.

Pain is usually a symptom correlating to anavulsion lesion of C7, C8 or T1 roots. The de-afferentation pain from root avulsion usuallybegins after a week or more, but if it appearsimmediately, more severe long-term de-afferentation pain can be expected. The presenceof any pain in an anesthetic hand or limb marksa root avulsion and severe pain syndrome pointsto C4–C5 root avulsion in 80 per cent of cases(Bonnard and Narakas 1997).

The presence of a proximal Tinel’s sign in theneck while testing in a disto-proximal fashion ofthe major peripheral nerves usually indicates aproximal neuroma and a sign of good prognosis.

However the absence of a Tinel’s sign in the neckis an important clinical finding pointing to a totalplexus avulsion.

Moisture of the skin can give useful informa-tion about the lesion: dry skin in an anestheticarea suggests a postganglionic lesion; on thecontrary, a normal moist skin suggests a pregan-glionic lesion. Sliding a plastic pen over the skinof the affected limb and comparing to the normalside can be used to test sweating function of theskin. The ninhydrin test is a more scientific testto detect sweating function.

The deep pressure sense (pinch) test is doneto determine whether continuity exits in a rootwith small nerve fibers which is least affected bycompression of a nerve trunk following injuryand swelling. To perform the pinch test, fullpinch pressure is applied to the patient’s finger-tips at the base of the nail and then the patient’sfinger is pulled sharply out from the examiner’sthumb and index finger, a maneuver that is

PHYSICAL EXAMINATION 19

Table 2 Major motor and sensory functions of the various parts of the brachial plexus

Anatomic Level Sensory Motor

Root C5Root C6

Root C7

Root C8Root T1(Upper trunk) Suprascapularnerve(Upper trunk) Posterior cordbranch (circumflex nerve)(Upper trunk) C7contribution to lateral cordLateral cord

Musculocutaneous nerveMiddle trunk

Posterior cord

Lower trunk and medialcord

Deltoid chevronCubital fossa, tip of the thumb

Thumb, index and middle fingers, dorsalradial handRing and little fingers, dorsal ulnar hand

Deltoid chevron

Thumb, index and middle fingers

Cubital fossaRadial forearmThumb, index and middle fingers (notback of the thumb)Cubital fossa, radial forearmThumb, index and middle fingers, radialforearm, radial dorsal hand

Deltoid chevron, back of thumb, index andmiddle fingersRing and little fingers, medial arm andforearm

Shoulder external rotation and abductionElbow flexion, extensor carpi radialislongusWrist and finger extensors, flexor carpiradialis, brachioradialis, pronator teresWrist and finger flexorsHand intrinsicsShoulder external rotation

Shoulder abduction

Pronator teresFlexor carpi radialisElbow flexionFlexor carpi radialisPronator teres

Elbow flexionNot external rotators of shoulder, elbowextension, brachioradialis, wrist and fingerextensorsShoulder abduction, elbow extension,brachioradialis, wrist and finger extensorsMost of wrist and finger flexors, medianand ulnar intrinsics

Page 29: Brachial Plexus Injuries

painful in a normal finger. In an apparentlyanesthetic finger, any burning sensation pointsto some continuity of the nerve supplying thatfinger. The tip of the thumb is used to test theC6 root with median nerve, the tip of the middlefinger for the C7 root through median nerve, andthe little finger for the C8 root with ulnar nerve,respectively. Neuropraxia can also affect trans-mission in these fibers, therefore the absence ofa deep pressure sense up to 6 weeks of the injuryis still not diagnostic of a rupture of that partic-ular nerve (Boome 1997b).

Root C8–T1 avulsion or a lesion close to thevertebral column of the corresponding spinalnerves compromises the sympathetic pregan-glionic fibers on the same side of the head andcauses vasodilatation, anhydrosis, miosis,enophthalmos and ptosis which is known asHorner’s syndrome. The absence of Horner’ssign is a good prognostic feature. If the avulsionof the rootlets accompanies a partial lesion of thespinal cord, Brown–Sequard syndrome occurs.Clinical signs of the patient show dissociatedchanges in the lower limbs, including spasticityand loss of tactile discrimination, position senseand vibration in the ipsilateral lower limb, whilethere is loss of pain and temperature discrimi-nation in the contralateral lower limb. Possibleparalysis of the intercostal nerves precludesneurotization of these nerves for reconstructionof the plexus (Boome 1997b).

Associated vascular injuries, bone and jointpathologies must also be taken into considera-tion and recorded during examination. Arterialrupture usually accompanies a infraclavicularplexus lesion but can be seen at the supraclav-icular level with C8–T1 root avulsions. Expandingswelling in the axillary area with or without abruit is a strong evidence of an arterial injuryeven in the presence of distal intact pulses.Progressive loss of function with increasingparalysis and sensory deficit suggests anexpanding hematoma or aneurysm compressingadjacent nerve trunks (Birch 1997).

Cervical transverse process fractures can beseen with C8–T1 root avulsions but also withC5–C6 root ruptures. Glenohumeral dissociationcan lead to peripheral plexus lesion and dislo-cated shoulder (scapulothoracic dissociation) is a

sign of complete avulsion with peripheral lesion(double level lesion). Upper humeral fracturessuggests infraclavicular nerve injuries andsevere abrasions on the tip of the shoulder andthe side of the head or helmet suggest supra-clavicular injuries (Millesi 1984; Bonnard andNarakas 1997).

References

Alnot JY (1995) Traumatic brachial plexus lesions inthe adult: indications and results. In: Grossman JAI, ed.Brachial Plexus Surgery. Hand Clinics (Nov). WBSaunders: Philadelphia:623–32.

Aschan W, Moberg E (1962) The Ninhydrin finger print-ing test used to map out partial lesions to hand nerves.Acta Chir Scand 132: 365–366.

Birch R (1997) Infraclavicular lesions. In: Boome RB, ed.The Hand and Upper Extremity: the Brachial Plexus.Churchill Livingstone: Edinburgh and New York: 79–88.

Bonnard C, Narakas AO (1997) Supraclavicular tractioninjuries in adults. In: Boome RB, ed. The Hand andUpper Extremity: the Brachial Plexus. ChurchillLivingstone: Edinburgh and New York: 71–78.

Boome RB (1997a) General discussion on the brachialplexus. In: Boome RB, ed. The Hand and UpperExtremity: the Brachial Plexus. Churchill Livingstone:Edinburgh and New York: 1–8.

Boome RB (1997b) Practical anatomy clinical assess-ment and surgical exposure. Boome RB, ed. The Handand Upper Extremity: the Brachial Plexus. ChurchillLivingstone: Edinburgh and New York: 9–18.

Clarkson HM (1999) Musculoskeletal Assessment, 2ndedn. Lippincott, Williams and Wilkins.

Kendall FP, McCreary EK, Provance PG (1993) Muscles:Testing and Function with Posture and Pain, 4th edn.Williams and Wilkins: Baltimore.

Millesi H (1984) Brachial plexus injuries, Clin Plast Surg11:115–121.

Tubiana R,Thomine JM, Mackin E (1995) Examinationof the Hand and Wrist, 2nd edn. Mosby-Yearbook: StLouis.

20 THE BRACHIAL PLEXUS

Page 30: Brachial Plexus Injuries

PHYSICAL EXAMINATION 21

Table 1 Clinical examination of brachial plexus injuries

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

SupraspinatusmuscleSuprascapularnerveC4, 5 and 6

Abducts theshoulder joint,stabilizes the headof humerus in theglenoid cavity

Initiation of abduction ofthe humerus whileapplying pressureagainst forearm in thedirection of adductionFixation is notnecessary

No effort is made todistinguish thesupraspinatus from thedeltoid in the strengthtest for grading, asthese muscles actsimultaneously inabducting the shoulder.To palpate thesupraspinatus, thetrapezius must berelaxed by extendingand laterally flexing thehead and neck

Deltoid muscleAxillary nerveC5 and 6

Shoulder abduction(chiefly by middlefibers)Shoulder flexionand mediallyrotation (anteriorfibers)Shoulder extentionand lateral rotation(posterior fibers)

Middle deltoid (sittingposition): shoulderabduction withoutrotation (A)Anterior deltoid (sittingposition): shoulderabduction in slightflexion with thehumerus in slight lateralrotation (B)Posterior deltoid (proneposition): shoulderabduction in slightextension with thehumerus in slight medialrotation (C)If the scapular fixationmuscles are weak theexaminer must stabilizethe scapula

In the presence ofparalysis of the entiredeltoid andsupraspinatus muscles,the humerus tends tosubluxate downwardsbecause the capsule ofthe shoulder jointpermits almost 2.5 cmof separation of thehead of the humerusfrom the glenoid cavity

continued

Pressure

A

B

C

Page 31: Brachial Plexus Injuries

22 THE BRACHIAL PLEXUS

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

InfraspinatusmuscleSuprascapularnerveC(4),5 and 6Teres minormuscleAxillary nerveC5 and 6

Laterally rotatesthe shoulder jointand stabilizes thehead of thehumerus in theglenoid cavity

(Prone position)Lateral rotation of thehumerus with the elbowheld at right anglesagainst pressure appliedin the direction ofmedial rotation(Supine position)Lateral rotation of thehumerus with the elbowheld at right angleagainst pressure appliedin the direction ofmedial rotationThis test requires strongfixation of trapezius

For the purpose ofobjectively grading aweak lateral rotatorgroup against gravityand for palpation of therotator muscles the testin prone position ispreferred for teresminor and in supineposition for infraspinatus

Latissimus dorsimuscleThoracodorsalnerveC6,7 and 8

Medially rotates,adducts andextends shoulderjoint. Alsodepresses theshoulder girdle andassists lateralflexion of the trunk

(Prone position)Adduction of the arm with extension in themedially rotated position against pressure on theforearm in the direction of abduction and slightflexion of the armCounter pressure is applied laterally on pelvis

Teres majormuscleLowersubscapular nerveC5,6 and 7

Medially rotates,adducts andextends shoulder

(Sitting position)Extension and adduction of the humerus in themedially rotated position against pressure on thearm above the elbow in the direction of abductionand flexion

Pectoralis majormuscle(Upper fibers)Lateral pectoralnerveC5,6 and 7(Lower fibers)Lateral andmedial pectoralnerveC6,7,8 and T1

Adduction andmedial rotation ofthe humerus;depression ofshoulder girdle.

Upper fibers (supine position). Starting with theelbow extended and the shoulder in 90° flexionand slight medial rotation, the humerus ishorizontally adducted toward the sternal end ofthe clavicle against pressure in the direction ofhorizontal abduction.Lower fibers (supine position). Starting with theelbow extended and the shoulder in flexion andslight medial rotation, adduction of the armobliquely toward the opposite iliac crest againstthe forearm obliquely in a lateral and cranialdirection

continued

Teres minor

Infraspinatus

(prone)

P

P

Page 32: Brachial Plexus Injuries

PHYSICAL EXAMINATION 23

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Pectoralis minormuscleMedial and lateralpectoral nerveC(6),7,8 and T1

Tilts the scapulaanteriorly androtates the scapulaso that thecoracoid processmoves anteriorlyand caudally

(Supine position)While the shoulder is inexternal rotation and 80°flexion and the elbow isflexed, the examinermoves the shouldergirdle cranially anddorsally along the shaftof the humerus to testmuscle strength

Weakness of thismuscle will increaserespiratory difficulty inpatients alreadysuffering frominvolvement ofrespiratory muscles

SubscapularismuscleUpper and lowersubscapular nerveC5,6 and 7

Medially rotatesthe shoulder jointand stabilizes thehead of thehumerus in theglenoid cavityduring movementsof this joint

(Supine/prone)Medial rotation of the humerus with arm at sideand elbow held at right angles against pressure inthe direction of laterally rotating the humerususing the forearm as a lever

RhomboidmusclesDorsal scapularnerveC4 and 5

Levator scapulaemuscleCervical 3 and 4,and dorsalscapular nerveC4 and 5

Trapezius muscleSpinal portion ofcranial nerve XIand ventral ramusC2,3 and 4

Adduct and elevatethe scapula androtate it so theglenoid cavity facescaudally

Elevates scapulaand assists inrotation so theglenoid cavity facescaudally

Adduction of thescapula performedchiefly by themiddle fibers withstabilization by theupper and lowerfibers

Rhomboid (prone)The patient raises the arm away from theback.The weight of the raised upper extremityprovides resistance to the scapular test motion.Rhomboid major can be palpated medial to thevertebral border of the scapula lateral to the lowerfibers of trapezius, near the inferior angle of thescapula. Note: inability to lift the hand off thebuttock may be due to shoulder muscleweakness, notably subscapularis not rhomboidmuscle weakness. Ensure that the hand ismaintained over the non-test side buttock andpatient adducts, medially rotates scapula (A).

Middle trapezius (prone)Adduction of the scapula with upward rotation(lateral rotation of the inferior angle) and withoutelevation of the shoulder girdle against pressureon the forearm in a downward direction towardthe table (B).

Upper trapezius (sitting)Elevation of the acromial end of the clavicle andscapula; postero-lateral extension of the neckbringing the occiput toward elevated shoulder withthe face turned in opposite direction (C)

continued

A

B

C

Page 33: Brachial Plexus Injuries

24 THE BRACHIAL PLEXUS

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Serratus anteriormuscleLong thoracicnerveC5,6,7 and 8

Abducts thescapula, rotates theinferior anglelaterally and theglenoid cavitycranially and holdsthe medial borderof the scapulafirmly against therib cage

(Standing position)Facing a wall with the elbows straight, the subjectplaces hands against the wall and pushes againstthe wall. This test is useful to differentiate onlybetween strong and weak for the purpose ofgrading (A).

A more objective test is to evaluate the ability ofthe serratus to stabilize the scapula in a positionof abduction and lateral rotation with the arm in aposition of approximately 120 to 130 flexionagainst pressure on dorsal surface of the armbetween shoulder and elbow downward in thedirection of extension and slight pressure againstthe lateral border of scapula in the direction ofrotating the inferior angle medially (B)

Adductor pollicismuscleUlnar nerveC8 and T1

Adducts thecarpometacarpal(CMC) joint andadducts and assistsin flexion of themetacarpophalangeal (MCP) joint sothat the thumbmoves toward theplane of the palm

Adduction of the thumbtoward the palm againstthe pressure on themedial surface of thethumb in the directionof abduction away frompalm. The hand may bestabilized by theexaminer or rest on thetable for support

A test that is frequentlyused to determine thestrength of the adductorpollicis is the ability tohold a piece of paperbetween the thumb andsecond metacarpalwhich can be difficult ina patient having musclebulk preventing closeapproximation of theseparts

Abductor pollicisbrevis muscleMedian nerveC6,7,8 and T1

Abducts the CMCand MCP joints ofthe thumb in aventral directionperpendicular tothe plane of thepalm

Abduction of the thumb ventralward from thepalm against pressure on the proximal phalanx inthe direction of adduction toward the palm.The examiner stabilizes the hand

Opponens pollicismuscleMedian nerveC6,7,8 and T1

Opposes the CMCjoint of the thumbin a position sothat by flexion ofthe MCP joint itcan oppose thefingers

Flexion, abduction and slight medial rotation of themetacarpal bone against pressure on metacarpalbone in the direction of extension and adductionso that the thumbnail shows in palmar view.The examiner stabilizes the hand

continued

A

B

Page 34: Brachial Plexus Injuries

PHYSICAL EXAMINATION 25

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Flexor pollicislongus muscleMedian nerveC(6),7,8 and T1

Flexing theinterphalangeal ( IP)joint of the thumb

Flexing the IP joint of the thumb against pressureon the palmar surface of the distal phalanx in thedirection of extension. The examiner stabilizes themetacarpal bone and proximal phalanx of thethumb in extension

Flexor pollicisbrevis muscle(Superficial head)Median nerveC6,7,8 and T1(Deep head)Ulnar nerve C8and T1

Flexes the MCPand CMC joint ofthe thumb

Flexing the MCP joint of the thumb withoutflexion of the IP joint against pressure on thepalmar surface of the proximal phalanx in thedirection of extension.The examiner stabilizes the hand

Extensor pollicislongus muscleRadial nerveC6,7 and 8

Extends IP jointand assists inextension of theMCP and CMCjoints of the thumb

Extension of the IP jointof the thumb againstpressure on the dorsalsurface of the IP joint ofthe thumb in thedirection of flexion.The examiner stabilizesthe hand and givescounterpressure againstthe palmar surface ofthe first metacarpal andproximal phalanx

In a radial nerve lesion,the IP joint of thethumb may be extendedby the action ofabductor pollicis brevis,flexor pollicis brevis,oblique fibers of theadductor pollicis or bythe first palmarinterosseus, by virtue oftheir insertions into theextensor expansion ofthe thumb

Extensor pollicisbrevis muscleRadial nerveC6,7 and 8

Extends the MCPjoint of the thumb,extends andabducts the CMCjoint

Extension of the MCP joint of the thumb againstpressure on the dorsal surface of the proximalphalanx in the direction of flexion.The examiner stabilizes the wrist

Abductor pollicislongus muscleRadial nerveC6,7 and 8

Abducts andextends the CMCjoint of the thumb

Abduction and slight extension of the firstmetacarpal bone against pressure on the lateralsurface of the distal end of the first metacarpaland the ability to abduct the wrist.The examiner stabilizes the wrist

Abductor digitiminimi muscleUlnar nerveC(7),8 and T1

Abducts, assists inopposition, andmay assist inflexion of the MCPjoint of the littlefinger

Abduction of the little finger against pressure onthe ulnar side of the little finger in the direction ofadduction toward the midline of the hand.The hand may be stabilized by the examiner orrest on the table for support

continued

Page 35: Brachial Plexus Injuries

26 THE BRACHIAL PLEXUS

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Opponens digitiminimi muscleUlnar nerveC(7),8 and T1

Opposes the CMCjoint of the littlefinger

Opposition of the fifth metacarpal toward the firstagainst pressure on the palmar surface along thefifth metacarpal in the direction of flattening thepalm of the hand.The hand can be stabilized by the examiner orrest on the table for support.The first metacarpalis held firmly by the examiner

Flexor digitiminimi muscleUlnar nerve C(7),8and T1

Flexes the MCPjoint of the littlefinger and assistsin opposition of thelittle finger towardthe thumb

Flexion of the MCP joint with IP joints extendedagainst pressure on the palmar surface of theproximal phalanx in the direction of extensionThe hand may rest on the table for support or bestabilized by the examiner

Dorsal interosseimusclesUlnar nerve C8and T1

Abducts the index,middle and ringfingers from theaxial line throughthe third digit.Assist in flexion ofMCP joints andextension of IPjoints of the samefingers

Abduction of the index, middle and ring fingersagainst pressure

Palmar interosseimusclesUlnar nerveC8 and T1

Adducts thethumb, index, ringand little fingertoward the axialline through thethird digit

Adduction of the corresponding fingers againstpressure

LumbricalesmusclesLumbricales I andIIMedian nerveC(6),7 and 8Lumbricales IIIand IVUlnar nerveC(7),8 and T1

Extends the IPjoints andsimultaneouslyflexes the MCPjoints of thesecond throughfifth digits

Extension of IP joints with simultaneous flexion ofMCP joints against pressure first on the dorsalsurface of the middle and distal phalanges in thedirection of flexion and then against the palmarsurface of the proximal phalanges in the directionof extension.The examiner stabilizes the wrist in slightextension if there is any weakness of wristmuscles

continued

Page 36: Brachial Plexus Injuries

PHYSICAL EXAMINATION 27

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Palmaris longusmuscleMedian nerveC(6),7,8 and T1

Tenses the palmarfascia, flexes thewrist. It may assistin flexion of theelbow

Tensing of the palmar fascia by strongly cuppingthe palm of the hand and flexion of the wristagainst pressure on thenar and hypothenareminences in the direction of the flattening thepalm of the hand and against the hand in thedirection of extending the wrist.The forearm rests on the table for support in aposition of supination

Extensor indicismuscleExtensor digitiminimi muscleExtensordigitorummusclesRadial nerveC6,7 and 8

Extends the MCPjoints and inconjunction withthe lumbricales andinterossei; extendsthe IP joints of thesecond throughfifth digits

Extension of the MCP joints of the secondthrough fifth digits with IP joints relaxed againstpressure on the dorsal surfaces of the proximalphalanges in the direction of flexion.The examiner stabilizes the wrist avoiding fullextension

Flexor digitorumsuperficialismusclesMedian nerveC7,8 and T1

Flexes the proximalIP joints of secondthrough fifth digits,assists in flexion ofthe MCP joints andin flexion of thewrist

Flexion of the proximal IP joint with the distal IPjoint extended of the second, third, fourth andfifth digits against pressure on the palmar surfaceof the middle phalanx in the direction ofextension.The examiner stabilizes the MCP joint with thewrist in neutral position or in slight extension.It appears to be the exception rather than the ruleto obtain isolated flexor superficialis action in thefifth digit

Flexor digitorumprofundusmusclesProfundus I, IIMedian nerveC7,8 and T1Profundus III, IVand ulnar nerveC(7), 8 and T1

Flexes distal IPjoints of index,middle, ring andlittle fingers andassists in flexion ofproximal IP andMCP joints

Flexion of the distal IP joint of the second, third,fourth and fifth digits against pressure on thepalmar surface of the distal phalanx in thedirection of extension.With the wrist in slight extension the examinerstabilizes the proximal and middle phalanges

Flexor carpiradialis muscleMedian nerveC6,7 and 8

Flexes and abductsthe wrist and mayassist in pronationof the forearm andflexion of theelbow

Flexion of the wrist toward the radial side againstpressure on the thenar eminence in the directionof extension toward the ulnar side.The forearm is in slightly less than full supinationand rests on the table for support.The palmaris longus can not be ruled out in thistest

Page 37: Brachial Plexus Injuries

28 THE BRACHIAL PLEXUS

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

Flexor carpiulnaris muscleUlnar nerveC7,8 and T1

Flexes and adductsthe wrist and mayassist in flexion ofthe elbow

Flexion of the wrist toward the ulnar side againstpressure on the hypothenar eminence in thedirection of extension toward the radial side.The forearm is in full supination and rests on thetable for support or is supported by the examiner.Normally fingers will be relaxed when the wrist isflexed.If the fingers actively flex as wrist flexion isinitiated, the finger flexors are tempting tosubstitute for the wrist flexors

Extensor carpiradialisLongus andbrevis musclesRadial nerveC6,7 and 8

Extends andabducts (longus)the wrist andassists in flexion ofthe elbow

Extension of the wrist toward radial side againstpressure on the dorsum of the hand along thesecond and third metacarpal bones while thefingers are allowed to flex.The forearm is in slightly less than full pronationand rests on the table for support

Extensor carpiulnaris muscleRadial nerveC6,7 and 8

Extends andadducts the wrist

Extension of the wrist toward the ulnar sideagainst pressure on the dorsum of the hand alongthe fifth metacarpal bone in the direction offlexion toward the radial side.The forearm is in full pronation and rests on thetable for support or supported by the examiner.Normally fingers will be in a position of passiveflexion when the wrist is extended.If the fingersactively extend as wrist extension is initiated, thefinger extensors are attempting to substitute forthe wrist extensors

Pronator teresmuscleMedian nerveC6 and 7

Pronates theforearm and assistsin flexion of theelbow joint

Pronation of the forearm with the elbow partiallyflexed against pressure at the lower forearmabove the wrist in the direction of supinating theforearm.The elbow should be held against the patient’sside or be stabilized by the examiner to avoid anyshoulder abduction movement

Pronatorquadratus muscleMedian nerveC7,8 and T1

Pronates theforearm

Pronation of the forearm with the elbowcompletely flexed in order to make the humeralhead of the pronator teres less effective by beingin a shortened position.The elbow should be held against the patient’sside to avoid shoulder abduction

Supinator muscleRadial nerveC5,6,(7)

Supinates theforearm

Supination of the forearm against pressure at thedistal end of the forearm above the wrist in thedirection of pronation.The examiner holds the shoulder and elbow inextension (tested with biceps elongated)

continued

Page 38: Brachial Plexus Injuries

PHYSICAL EXAMINATION 29

Table 1 Continued

Muscle and Function Muscle test Notes Test pictureinnervation and fixation

BrachioradialismuscleRadial nerveC5 and 6

Flexes the elbowjoint and assists inpronating andsupinating theforearm whenthese movementsare resisted

Flexion of the elbow with the forearm neutralbetween pronation and supination. The belly ofthe brachioradialis must be seen and felt duringthis test.The examiner places one hand under the elbow tocushion it from table pressure

CoracobrachialismuscleMusculocutaneousnerveC6 and 7

Flexes and adductsthe shoulder joint

Shoulder flexion in lateral rotation with the elbowcompletely flexed and forearm supinated againstpressure on the anteromedial surface of the lowerthird of the humerus in the direction of extensionand slight abduction. Assistance from the bicepsin shoulder flexion is decreased in this testposition because the complete elbow flexion andforearm supination place the muscle in too short aposition to be effective in shoulder flexion.Fixation is not necessary

Biceps brachiiand brachialismusclesMusculocutaneousnerveC5 and 6

Flex shoulder andelbow joint andsupinates theforearm (biceps)

Elbow flexion slightlyless than or at rightangles with forearm insupination againstpressure on the lowerforearm in the directionof extension.The examiner placesone hand under theelbow to cushion itfrom table pressure

If the biceps andbrachialis are weak as ina nervusmusculocutaneouslesion, the patient willpronate the forearmbefore flexing the elbowusing brachioradialis,extensor carpi radialislongus, pronator teresand wrist flexors

Triceps brachiimuscleRadial nerveC6,7,8 and T1

Extends the elbowjoint. In additionthe long headassists in adductionand extension ofthe shoulder joint

(Supine position)Extension of the elbowagainst pressure on theforearm in the directionof flexion.The shoulder is atapproximately 90°flexion with the armsupported in a positionperpendicular to thetable

While the triceps andanconeus act togetherin extending the elbowjoint, it may be usefulto differentiate thesetwo muscles. As thebelly of the anconeusmuscle is below theelbow joint, it can bedistinguished from thetriceps by palpation. It ispossible for a lesion toinvolve the branch ofradial nerve to anconeusleaving tricepsunaffected. The gradeof good elbowextension strength isactually the result of anormal triceps withoutanconeus

Page 39: Brachial Plexus Injuries
Page 40: Brachial Plexus Injuries

X-ray

Radiological investigations of brachial plexuslesions are both revealing and necessary.However, the value of the plain X-ray is under-rated: it may show the site and severity of theinjury.

X-rays of the cervical spine can reveal aluxation and/or fracture, pointing to the possibil-ity of associated lesions of the spinal cord and/orcervical spinal nerves or roots. Fracture of thetransverse process of the cervical spine often

implies a severe local (intraforaminal) lesion ofthe spinal nerve. Clavicular fracture or luxation israrely a cause of the brachial plexus lesion butreferes to the site of impact. A severefracture/dislocation will certainly have caused aninjury to the underlying plexus structures, whichwill require reconstruction (Fig. 1). In the sameway, X-ray of the shoulder can reveal the severecondition of a scapulo-thoracic dissociation(depression of the scapula and lateral displace-ment), in which the extensive neural injury isalways associated with a serious vascular lesion

3Radiological and relatedinvestigationsAlbert (Bart) CJ Slooff, Corneleus (Cees) WM Versteege, Gerhard Blaauw,and Willem JR van Ouwerkerk

Figure 1

Fracture/dislocation of the clavicle with pseudoarthrosis(arrow).

Figure 2

Scapulo-thoracic dissociation: lateral displacement of thescapula (arrows); fracture of the clavicle (white arrow), andtraumatic aneurysm of the subclavian artery (black arrow).

Page 41: Brachial Plexus Injuries

(Fig. 2). Shoulder dislocation or fractures can beaccompanied by neural injuries, sometimes incombination with a cuff lesion (Fig. 3). A scapulafracture running close or into the scapular notchmay cause an injury to the suprascapular nerve.It is interesting to know that a fracture of thecollum scapulae can be associated with a selec-tive injury of the nerve to the infrapinatus muscle(Fig. 4). A severe dislocated fracture of thehumerus can be followed by an injury to theradial nerve and thus present as one of thesecond level lesions in brachial plexus injuries inadults. The same problem occurs with theaxillary nerve – or other adjacent neural struc-tures, such as the lateral and posterior and/oreven medial cord – in fracture-dislocation of thehumeral head. Other injuries of the upperextremity do not have a diagnostic significancebut have to be considered in the context of thetotal treatment of the patient with brachial plexusinjury. X-ray of the chest is important in severalways, e.g. for diagnosing a diaphragmatic paral-ysis (C4) (Fig. 5), costal fractures, a pulmonary

lesion, but also in the overall treatment and morespecifically in relation to possible extraplexal(intercostal nerves) neurotizations.

In obstetric patients, a chest X-ray has to be astandard preoperative procedure, especially torule out a hemi-paralysis of the diaphragm.

32 THE BRACHIAL PLEXUS

Figure 3

Subcapital humerus fracture.

Figure 4

Fracture of the collum scapulae.

Figure 5

Left-side diaphragmatic paralysis.

Page 42: Brachial Plexus Injuries

Generally, this paralysis has to be treated firstby plication of the diaphragm, if possible by anendoscopic approach. A persisting phrenicparalysis is often associated with a homolateralsevere upper brachial plexus lesion, which hasto be treated surgically. Normally, fractures ofthe clavicle and humerus in obstetric cases haveno consequences, and certainly have noprognostic value in predicting spontaneousrecovery; they can be associated injuries of theobstetric plexus lesions and these fractures aresometimes wrongly supposed to be the cause ofthe malfunction of the extremity instead of theplexus injury itself.

CT-myelography

In the last decade myelography has been supple-mented and standardized as CT–myelography.Although MRI is the major tool in diagnosticneuroradiology for the assessment of rootletavulsions in brachial plexus lesions, CT-myelog-raphy still proves more reliable (Miller et al 1993,Francel et al 1995, Panasci et al 1995, Burge1997, Carvalho et al 1997, van Es 1997, Chow etal 2000).

Sunderland (1991) stated that traction of thebrachial plexus will normally result initially intearing of the arachnoidal and dural sheet of thenerve tract. But rootlet avulsions may existwithout a concomitant meningocele, probablybecause of a so-called central mechanism (seeChapter 16). A meningocele is a sign that the

patient has suffered from a traction force, but isnot proof of an avulsion. Also ‘complete’ recov-ery is possible in children with multiplemeningoceles. The existence of a meningocele isa complicating diagnostic factor because it isoften impossible to follow the course of rootlets,depending on the size of the meningocele.Sometimes the rootlet is displayed very clearlyin the middle of the meningocele but themeningocele can also push the rootlets away,even intact rootlets at adjacent levels, so thatthey cannot be judged to be secure.

Birch et al (1998) observe and remark on thedifferences in intradural injuries of the roots:intradural rupture and avulsion. The lesion may

RADIOLOGICAL AND RELATED INVESTIGATIONS 33

Figure 6

Absent (avulsed) right ventral root (C5).

Figure 7

Avulsed roots on the leftside, bilaterally and anteriorlylocated meningoceles.

Page 43: Brachial Plexus Injuries

be confined to either ventral (Fig. 6) or dorsalrootlets, or even be partial as we find moreoften in obstetric patients. These findings mayhave their operative consequences. In ourdepartments, CT-myelography is routinelyperformed, mainly in candidates for surgicaltreatment. The correlation between radiologicalinvestigations, the clinical picture and operativefindings is still higher in adults than in children,although important progress has been madeover recent years to improve delineation of therootlets using thin slices. Today we areconvinced about the value of CT-myelographyin obstetric patients (Fig. 7). Based on ourexperience with more than 600 surgicallytreated patients with brachial plexus lesions,including more than 300 children, we havereached the following conclusions:

• The presence of intact dorsal and ventralrootlets without a meningocele rules out anavulsion (Fig. 8);

• The absence of dorsal or ventral rootlets or bothwithout a concomitant meningocele is consid-ered to suggest a partial or complete avulsion;

• The presence of a meningocele is not proof ofan avulsion and can mask the existence ofintact rootlets, so that this level has to beexamined with special care;

• If the meningocele extends outside theforamen, which is less frequent in obstetricpatients, then an avulsion is very likely;

• Deformation and/or displacement of the spinalcord indicates a severe intradural injury (Figs9 and 10).

Technical notes for CT-myelography

For adults (16 years and older). Introduction of10 ml Omnipaque 300 by lateral C1–C2 puncture;this prevents post-puncture headache. The contrastis introduced under fluoroscopic guidance. Theparameters cited below define the quality of thepicture and the units quantify them.

Parameters: – slice thickness 1.5 mm– translation 3 mm– mAs 300– kV 120– FOV 80

Axial slices are taken from the level C2–T2.

For obstetric patients. CT-myelography isperformed under general anesthesia, normally atthe age of 3–4 months and we prefer to intro-duce 3–4 ml Omnipaque 240 by lumbar puncture.

Parameters: – slice thickness 1.5 mm– translation 2 mm– mAs 200– kV 120– FOV 80

Axial slices are taken from the level C2–T2.

34 THE BRACHIAL PLEXUS

Figure 8

A picture of a normal CT myelo-gram, revealing clearly the centraland dorsal roots (C6).

Page 44: Brachial Plexus Injuries

Magnetic resonance imaging

At present, in our opinion, MRI is not able todiscriminate the course of the rootlets betterthan CT-myelography quite apart from problemslike flow artifacts. Currently, variant MRtechniques for visualizing intradural lesions areunder evaluation. The 3D-CISS (three-dimen-sional constructive interference in steady state)procedure is especially promising for obstetric

lesions (Fig. 11). With this technique the courseof rootlets can be followed, even in the presenceof meningoceles. Advantages of the MRI are thelack of radiation and a less invasive type ofinvestigation, although general anesthesia,necessary in obstetric patients, may pose aproblem. MRI visualization of the plexus struc-tures beyond the vertebral foramen is the best ofall other imaging techniques. Traumatic injuriessuch as large neuromas may be followed along

RADIOLOGICAL AND RELATED INVESTIGATIONS 35

Figure 9

Deformation of the spinal cord.

Figure 10

Displacement of the spinal cordwith lateral and anterior meningo-celes; rootlets are difficult to distin-guish.

Page 45: Brachial Plexus Injuries

the plexus structures as well as entrapments andtumor formation and also the relation to vascu-lar structures. Hematomas in the vertebral canal,when avulsion is suspected, may be visualized aswell as hematomas in the paravertebral muscle,indicating the severity of trauma. Later, MRI canreveal the joint deformity, capsula tears andatrophy of muscles.

Angiography

In adult traumatic brachial plexus lesions,conventional angiography was frequentlyindicated but is at present mainly replaced byMRA. This is evident in an associated clinicalvascular deficiency (Fig. 12), but there willcertainly be ‘silent’ vascular lesions that couldpose a danger during the exploration and/orreconstruction (Fig. 13). Also this angiographic

36 THE BRACHIAL PLEXUS

Figure 11

3D-CISS picture with clearly visible intact rootlets on theright side (two white arrows), and absent rootlets on theleft side at the level of the meningocele (black arrow).

Figure 12

A traumatic occlusion of the axillary artery, with abundantcollateral circulation.

Figure 13

A ‘silent’ traumatic aneurysm of the axillary artery. (FromBlauuw 1999.)

Page 46: Brachial Plexus Injuries

information is necessary after a primary vascu-lar reconstruction. In our series we found a totalof 12 per cent associated vascular injuries inadult patients. Vascular injuries in obstetriccases are very rare (or exceptionally mentioned).

Future techniques

In the near future, endoscopic investigationswithin the vertebral canal may reveal avulsionsor ruptures of the roots more accurately, suchas partial avulsions or separate avulsion ofmotor and sensory roots. Apart from this aspectwe are not convinced of the necessity toperform a diagnostic cervical laminectomy.Moreover, especially in obstetric cases, the non-invasive technique of sonography is sometimescapable of detecting neuroma formation in thesupraclavicular region and may also be advan-tageous in exploring, for example, shoulderpathology.

References

Birch R, Bonney G, Wynn Parry CB (1988) SurgicalDisorders of the Peripheral Nerves. ChurchillLivingstone: Edinburgh: 157–60.

Blauuw G (1999) Letsels van de plexus brachialis.Tijdstroom: The Netherlands.

Burge P (1997) Diagnostic investigations. In: BoomeRB, ed. The Hand and Upper Extremity: the BrachialPlexus. Churchill Livingstone: New York: 19–29.

Carvalho GA, Nikkhah G, Matthies C et al (1997)Diagnosis of root avulsions in traumatic brachialplexus injuries: value of computerized tomography andmagnetic resonance imaging. J Neurosurg 86:69–76.

Chow BCL, Blaser S, Clarke HM (2000) Predictive valueof computer tomography in obstetrical plexus palsy.Plast Reconstr Surg 106: 971–7.

Van Es HW (1997) MR Imaging of the brachial plexus.Thesis, University of Utrecht: 67–9.

Francel PC, Koby M, Park TS et al (1995) Fastpin-echomagnetic resonance imaging for radiological assess-ment of neonatal brachial plexus injury. J Neurosurg83:461–6.

Miller SF, Glasier CM, Griebel ML, Boop FA (1993)Brachial plexopathy in infants after traumatic delivery:evaluation with MR imaging. Radiology 189:481–4.

Panasci DJ, Holliday RA, Shpizner B (1995) Advancedimaging techniques of the brachial plexus. In:Grossman JAI, ed. Brachial Plexus Surgery. HandClinics 11: (Nov). WB Saunders: Philadelphia: 545–53.

Sunderland S (1991) Nerve Injuries and Repair.Churchill Livingstone: Edinburgh: 147–58.

RADIOLOGICAL AND RELATED INVESTIGATIONS 37

Page 47: Brachial Plexus Injuries
Page 48: Brachial Plexus Injuries

Introduction

Neurophysiology is the study of the function ofthe central and peripheral nervous systems andof the muscles, and in clinical neurophysiologytechniques are employed to measure thesefunctions. It is important to realize that thesetechniques only measure electrical function.Anatomical information can be derived, but thisis secondary. Anatomical factors can also limitthe scope of measurements. However, becausefunction is measured at the level of single cells,it is sometimes possible to collect more informa-tion than with other methods, or with clinicalexamination. Clinical neurophysiologicaltechniques which are routinely used areelectroencephalography (EEG), electromyogra-phy (EMG), cortical, spinal or peripheral magneticstimulation (Magstim) and evoked potentialmeasurement (EP). This paper will discuss thevarious methods and their significance for theclinician in cases of brachial plexus lesions.

The EEG

The EEG is a measurement of the electricalfunction of the brain. Apart from its use inresearch, this technique offers no informationabout lesions in the peripheral nervous systemand will therefore not be discussed further.

The EMG

The EMG, in this chapter also including nerveconduction studies (NCS), measures the function

of the motor unit and the peripheral sensoryfibres and is thus the most important techniquefor analysing peripheral nerve lesions. Importantdefinitions are CMAP (compound muscle actionpotential) and SNAP (sensory nerve actionpotential). CMAP is measured by stimulating themotor nerve supramaximally and recording theelectrical activity from a muscle innervated bythis nerve by means of surface electrodes. It isthe summated potential of all muscle fibresbelow the electrodes and thereby indicative ofthe number of motor units that can still beactivated. The same is true for sensory fibres ina sensory or mixed nerve when measuring theSNAP. To understand the outcome, potential andlimitations of these tests, it is important to havesome knowledge of the pathophysiology of anerve lesion and to understand what can bemeasured.

Pathophysiology

As described by Sherrinton, the motor unitconsists of the anterior horn cell in the myelum,the axon, neuromuscular synapse and all musclefibres connected to it. A lesion of the motor unit,wherever it takes place, results in more or lessthe same findings on EMG: loss of function and,when the lesion is sufficiently severe, denerva-tion. Severance of a nerve can result in neuro-praxia, axonotmesis or neurotmesis.

In neuropraxia, the myelin sheath of the axonis damaged but the axon itself remains intact.The result of this is conduction loss over thedamaged segment so that the action potentialsfrom the anterior horn cell cannot reach themuscle fibres. Voluntary contraction of these

4Clinical neurophysiologicalinvestigationsJan W Vredeveld

Page 49: Brachial Plexus Injuries

motor units is not possible. However, as theaxon is still intact, denervation does not occur.Stimulating and measuring distal to the lesionresults in a normal CMAP or SNAP; stimulatingproximal to and measuring distal to the lesionresults in (partial) loss of the CMAP or SNAP.This is already the case immediately after theoccurrence of the lesion.

In the case of axonotmesis or neurotmesis,denervation occurs. Initially the nerve can still bestimulated distal to the lesion, but as Walleriandegeneration takes place, stimulation becomesless effective and when the degeneration ofnerve fibres is complete, stimulation no longerresults in a (normal) CMAP or SNAP. This meansthat the residual function of the nerve can bemeasured after about 1 week, following which nofurther deterioration will occur. However, inolder children and in adults it takes at leastanother 1–2 weeks before the muscle cellsdevelop spontaneous activity like fibrillationpotentials and positive sharp waves, also calleddenervation activity. In neonates these processesseem to follow a completely different timecourse, as we have found spontaneous activityalready 4–5 days after a well-documented lesion(unpublished data, Vredeveld).

As small changes in the CMAP or SNAP will beoverlooked, it is necessary to await the appear-ance of this spontaneous muscle activity (i.e. atleast 2–3 weeks after the accident) before theEMG is a reliable indicator of the extent of thelesion. The spontaneous muscle activity disap-pears as the denervated muscle fibres degener-ate or become reinnervated. It usually takes 1–2years, but sometimes much longer, for sponta-neous muscle activity to disappear completely.

Once Wallerian degeneration is complete, theproximal axonal stump starts to sprout and anew axon grows at a speed of roughly 0.5–1 mmper day, provided the correct path to the end-organ can be found.

Technique

To analyse a lesion in the peripheral nervoussystem, it is necessary to combine sensory andmotor measurements. As the anterior horn cell lieswithin the myelum and the dorsal root ganglionoutside it, the combination of a preserved SNAP

and denervation in the same myotome is indica-tive of a root or anterior horn cell lesion. In ourlaboratory, the EMG starts with measurement ofthe SNAPs. If necessary, e.g. in the case ofsuspected or underlying (poly)neuropathy, CMAPsare also measured. Both techniques measure atthe level of the nerve. Thereafter, needle EMG isperformed using concentric needles. Monopolarneedles can also be used. The needle EMG is ameasurement of the motor unit. Thus, findingspontaneous muscle activity indicates a lesion inthe anterior horn cell or in the muscle fibre itself,or somewhere between the two. It is impossible todifferentiate between these options using a singlemeasurement. Only the combination of severalmeasurements from several muscles, preferablyinnervated by the same root but by differentnerves, or by the same nerve but different roots,and the measurement of the SNAPs can give amore or less correct analysis of the lesion.Therefore, the following rules are important:

• Always sample a sufficient number ofmuscles, unless it is only necessary to checkreinnervation in a single muscle;

• Combine needle EMG with SNAPs; as statedabove, the dorsal ganglion is located outsideand the anterior horn cell inside the spinalcord;

• In the case of a lesion, all muscles innervatedby that root, trunk, cord or nerve must showspontaneous muscle activity in accordancewith the degree of innervation by this root,trunk, cord or nerve, unless there is a partiallesion. Thus, finding partial denervation insome muscles innervated by a nerve andcomplete denervation in one muscle inner-vated by the same nerve strongly suggests asecond lesion in that nerve somewhere proxi-mal to the latter muscle.

Intra-operative EMG

The advantage of intra-operative EMG is thatselective stimulation of roots, trunks, cords ornerves is possible. However, being able tomeasure something means that the neuromuscu-lar synapse is still functioning, otherwise it wouldonly be possible to measure nerve action poten-tials if there were functioning nerve fibres. Bystimulating a nerve and measuring at some

40 THE BRACHIAL PLEXUS

Page 50: Brachial Plexus Injuries

distance along that nerve, it is evident whetherthere are still functioning nerve fibres. However, itis impossible to tell if these are sensory or motorfibres unless it is possible to record activity fromthe muscle innervated by this nerve. Even then,the neuromuscular synapse must be functioning toallow some measurement, indicating that noneuromuscular blocking agents are effective. Thisseverely limits the use of intra-operative EMG.Furthermore, in our experience it proved possibleto obtain most of the necessary information beforethe operation by using an investigation protocolincluding nearly all the desired nerves andmuscles. Thus, only when selective information isrequired is there still a place for intra-operativemeasurement (see also SEP).

Protocol

Based on these rules, we have developed a proto-col for the analysis of brachial plexus lesions. Asthe findings in obstetric lesions differ from thosein older patients with the same lesion, the neona-tal EMG will be dealt with separately.

The standard EMG and NCS protocol used inour department is as follows:

• SNAPs from:Radial nerve to digit IMedian nerve to digits I and IIIUlnar nerve to digit Vand, if necessary, lateral cutaneous ante-brachial nerve;

• Concentric needle EMG from the followingmuscles:

Abductor pollicis brevisFirst dorsal interosseousFlexor carpi radialisFlexor carpi ulnarisExtensor digitorum communisBrachioradialisBiceps brachiiTriceps brachiiDeltoidPectoralis majorTrapeziusInfraspinatusLatissimus dorsiand Serratus anterior (not routinelysampled in neonates).

‘Adult’ lesions

This protocol has been verified in patients with atraumatic brachial plexus lesion (accepted: J ClinNeuromusc Dis). The patient group consisted of184 consecutive patients, 153 males (6–66 years,mean 25) and 31 females (11–74 years, mean 33),admitted to our hospital because of a traumaticbrachial plexus lesion. Nearly all patients wereadmitted between 6 weeks and 6 months afterthe accident, none after 1 year. In 155 patients (84per cent) our analysis was confirmed by imagingand operative findings. This leaves 29 patients (16per cent) in whom we missed or under-diagnosedthe severity or extent of the lesion. Why?

In 22 patients we missed at least one rootavulsion. As all 22 patients had a more or lesscomplete plexus lesion, the only way to find rootavulsions was by paraspinal EMG. However, inner-vation of the paraspinal muscles shows a consid-erable overlap, spanning up to six segments, and,when there is a limited number of root lesions,paraspinal denervation might disappear early.Therefore, paraspinal EMG will indicate the possi-bility of a root lesion, but it is impossible to be sureof exactly which root or of the exact number ofroots using paraspinal sampling alone. Partial rootlesions present an even greater problem. Correctlocation of a root lesion depends on the combina-tion with the findings in other muscles and theSNAPs. When there is a second more peripherallesion, the correct analysis of root lesionssometimes becomes impossible.

In one patient we missed a syrinx at the levelof C5 and C6, found by MRI. This patient had alesion of the upper and middle trunk, and thedenervation we found could well be explained bythe plexus lesion.

In five patients we found a normal SNAP in aseverely denervated area, and hence diagnosedroot avulsion. However, surgery revealed that thesepatients all had a peripheral lesion (neuroma) andno root avulsion. We have no explanation for thisfinding. So, why perform EMG?

As is apparent from our study, the EMG wasreliable in detecting and analysing brachial plexuslesions in 84 per cent of the patients. Furthermore,in 37 patients lesions outside the brachial plexuswere also found, most of them unexpected onclinical grounds but still important for the surgeonwhen operative treatment is being considered.What we found in these patients were additional,

CLINICAL NEUROPHYSIOLOGICAL INVESTIGATIONS 41

Page 51: Brachial Plexus Injuries

sometimes multiple, lesions outside the brachialplexus: 12 in the axillary nerve, eight in the radialnerve, seven in the musculocutaneous nerve, fivein the ulnar nerve, five in the median nerve includ-ing two carpal tunnel syndromes, and three in thelong thoracic nerve. Without knowledge of theselesions, a nerve graft to one of these nerves wouldhave resulted in a very long wait for an absentrecovery.

Another advantage is that reinnervation can befound and quantified even before this can bedone clinically.

‘Obstetric’ lesions

The EMG in obstetric lesions differs essentiallyfrom the EMG in older patients: first, denervationoccurs and disappears much earlier than in adults.We found that denervation can already be foundon the 4th or 5th day after a lesion and has usuallydisappeared by 4 months, unless the lesion issevere with multiple root avulsions or ruptures.Mostly, the EMG in obstetric brachial palsy is fartoo optimistic compared to the clinical picture. Tofind an explanation for this we have selectedinfants with a complete avulsion of roots C5 andC6 and compared them with patients with thesame lesion from the adult group. We found thatafter 4 months the infants showed a nearlynormal EMG on reflex activated contraction of thebiceps brachii, whereas the adults showed(almost) complete denervation. However, whenwe looked at infants with a complete avulsion ofC5 and C6 and with an additional lesion of C7 orthe middle trunk, the EMG showed the samepicture of almost complete denervation of thebiceps brachii as the adults with only avulsion ofC5 and C6. This means that the innervation of themuscles in neonates differs from that in olderchildren and adults, at least for the biceps brachii.Other studies also indicate a difference.Gramsbergen’s group, from Groningen University(Netherlands), found that in rats there waspolyneural innervation of the muscle fibres afterbirth, and they found the same in a 2-week-oldfull-term baby (personal communication). Thispolyneural innervation of the neuromuscularsynapse was considered not to exist after birth butto have disappeared around the 26th post-concep-tional week. Many still believe this. However, ithas been shown to exist after birth in term infants,

and is now believed to disappear within the firstfew months of life. This also explains the massivecontractions and movements seen in healthyneonates. Such a process of disappearance, alsocalled apoptosis or programmed cell death, playsan important role in the motor and sensory devel-opment. We think that, as we found in infants withan obstetric brachial plexus lesion, when the origi-nal innervation of a muscle is lost or severelydamaged, this extra innervation (or so-called‘luxury innervation’) stemming from other rootsbecomes the dominant innervation due to lack ofthe original dominant innervation and remainspresent probably during the rest of life. However,as the central motor pathways do not projectprimarily to the anterior horn cells of these roots,these muscles are clinically more or less paretic.They can be activated by reflex contraction, butthis happens at the spinal level. This explains thefar too optimistic EMG in these infants, andindicates that a second process of central devel-opment (learning) is necessary for normal motordevelopment. This learning process largelydepends on sensory input and is, therefore, alsohampered by the plexus lesion.

All this does not explain yet another differencein the EMG in neonates. As mentioned, thedenervation has largely disappeared by 4months. This is not true when there is completedenervation, as in the example of the brachialbiceps in avulsion of C5, C6 and C7. Providedthere is still some innervation of the muscle, wehave shown by single-fibre EMG that this veryearly reinnervation is due to massive collateralsprouting. This process is apparently muchquicker in neonates than in adults, althoughingrowth of new axons after a nerve graft seemsto occur at about the same speed as in adults,but the distances are much shorter.

The above has led us to the hypothesis that it isvery important to start trying to get sensory infor-mation into the central nervous system as early aspossible (physiotherapy, caressing and training,playing), and not to wait too long before repairmeasures are attempted, but also not to operatetoo early and thus prevent spontaneous recovery.Here the EMG can help to differentiate betweenthe infants with a good prognosis, i.e. sponta-neous recovery, and a bad prognosis, i.e. the needfor operative treatment. The above-mentionedfacts also indicate that the EMG in obstetric lesionshas to be viewed in a different way compared to

42 THE BRACHIAL PLEXUS

Page 52: Brachial Plexus Injuries

the EMG in ‘adult’ lesions; it is important that theelectromyographer assesses after reinnervation(i.e. high amplitude potentials), sometimes evenmore than after denervation, to reach a correctanalysis of the lesion. He or she also has to beaware of the fact that normal motor unit potentialsin neonates are much smaller than in adults. Thus,finding a normal ‘adult amplitude’ may alreadyindicate reinnervation.

The Magstim

The first, cortical, motor neurone can be stimu-lated electrically using high voltage stimulators,but this is very unpleasant and more or less thesame can be achieved by magnetic stimulation. Acoil is placed over the cortical motor area and,using a brief, high-powered current, a shortmagnetic pulse is produced. This gives a rapidlychanging magnetic field in the brain, resulting ina very small electric current, which can activatethe motor neurones in the cortex, probably byactivating some interneurones. These corticalmotor neurones then activate the anterior horncells, which in turn start firing and activating themuscle fibres. A contraction can then be regis-tered. Theoretically, this method is ideal fortesting the anterior roots. However, more periph-eral lesions will hamper the measurement andalso the activation is dependent on the patient’sco-operation. The best results are obtained with aslight pre-test contraction of the muscle. Otherdisadvantages are that it does not differentiate theroute by which the impulses reach the muscle,and cortical magnetic stimulation during narcosisis almost impossible. This method is thus lesssuitable for routine use and its value in analysingbrachial plexus lesions has yet to be established.

The somatosensory evokedpotentials (SEP)

The somatosensory evoked potentials are ameasurement of the somatosensory pathwayfrom a peripheral nerve up to the sensory cortex.Measurement is done by electrical stimulation ofa peripheral nerve, e.g. a digital nerve, and record-ing from electrodes overlying the somatosensorypathway up to the cortical representation area.

This pathway can be divided into smallersegments for more precise localization. When anormal cortical response is measured, the conclu-sion must be that at least a sufficient number ofsensory fibres, both peripheral and central, isfunctioning normally. This includes the plexusand the dorsal roots. When the response is lost,there must be a significant lesion somewherebetween the site of stimulation and the site ofmeasurement. Other lesions more central to thislesion can no longer be measured. This is also thelimitation of the use of the SEP in brachial plexuslesions; the response will be lost at the most distallesion in the somatosensory pathway.

However, it is still possible to register SEPduring narcosis. Hence, by stimulating the rootsduring surgery it can be established whether thedorsal roots (= somatosensory input!) are moreor less intact.

Conclusion

In lesions of the peripheral nervous system, theEMG and SEP are the most important techniquesfor analysing the severity and extent of the lesion,the EMG being the most important. In ‘adult’brachial plexus lesions the reliability of the EMGis high, the limitations are few and are known,and even clinically undetected but importantlesions can be found, provided the investigationhas been sufficiently thorough. Reinnervation canalso be detected even before it becomes clinicallyevident. In obstetric brachial plexus lesions, theEMG presents more problems due to the differ-ent anatomy and the early appearance and disap-pearance of denervation, and to the very earlyreinnervation. However, bearing this in mind, theEMG in neonates also offers a reliable analysis ofthe lesion.

Recommended reading

For EMG, most handbooks can be used. One ofthe most recent and detailed handbooks of EMGdealing with brachial plexus lesions is:

Dumitru D (1994) Electrodiagnostic medicine. In:Brachial Plexopathies and Proximal Mononeuropathies.Hanley & Belfus Inc: 585–642.

CLINICAL NEUROPHYSIOLOGICAL INVESTIGATIONS 43

Page 53: Brachial Plexus Injuries

For EMG prognosis of obstetric lesions:

Smith SJM (1996) The role of neurophysiologicalinvestigation in traumatic brachial plexus injuries inadults and children, J Hand Surg 21B:145–8.

For the difference in findings between obstetricand adult lesions:

Vredeveld JW, Blaauw G, Slooff ACJ et al (2000) Thefindings in paediatric obstetric brachial palsy differfrom those in older patients: a suggested explanation,Dev Med Child Neurol 42:158–61.

44 THE BRACHIAL PLEXUS

Page 54: Brachial Plexus Injuries

The Adult Traumatic Brachial Plexus

Page 55: Brachial Plexus Injuries
Page 56: Brachial Plexus Injuries

Historical introduction

Knowledge about the brachial plexus can betraced back to the second century when Galenaccurately described its anatomy (McHenry1969). Early on, the brachial plexus was viewedonly as a part of the peripheral nervous systemand has been treated accordingly for nearly twomillennia. At the beginning of the nineteenthcentury, brachial plexus injury became aseparate clinical entity from other peripheralnerve lesions (Malessy 1999). The etiologic factorin brachial plexus injury has gradually shiftedfrom one of ‘open injury’ in the American CivilWar and two World Wars to ‘closed injury’arising from road traffic accidents of the presentday, especially involving motorcycles.

Brachial plexus injury can be caused by a widevariety of circumstances. These etiologic factorscan be categorized according to their causativemechanisms as follows:

• Closed injuries– traction– compression– combined lesion

• Open injuries– sharp– gunshot

• Radiation

In closed brachial plexus injuries, the mostcommonly found causative mechanisms weretraction or contusion. In some circumstances, theinjuries were the result of a combination oftraction and contusion. At present, traction is themost frequently found mechanism of brachialplexus injury. In our series of adult brachial plexusinjuries, 95 per cent were caused by traction.

Traction lesions

Traction lesions result from forceful separationof the neck and shoulder or upper arm and trunk.The nerve pathology occurs between the twoanchoring points. The proximal anchoring pointis at the spinal cord and nerve root junction, andthe distal point is at the neuromuscular junction.The coracoid process is regarded as a temporarylever in forceful hyper-abduction of the shoulder.It is not only the direction of the applied force tobrachial plexus that determines the severity ofthe nerve damage, but also the speed of appli-cation of the traction force.

High velocity traction injury is the overallleading cause of brachial plexus injury in almostall reports (Alnot 1987, Songcharoen 1995). Themajority of traction injuries result from roadtraffic accidents. In our series of 1173 adultbrachial plexus patients, 82 per cent were causedby motorcycle accidents. The victim falls off aspeeding motorcycle and lands on the head andshoulder. On ground impact, the shoulder isdepressed and the head is forcefully flexed to theopposite side. The sudden widening of theneck–shoulder angle causes a severe tractioninjury to the clavicle and the underlying struc-tures including the brachial plexus and subcla-vian vessels. If the clavicle which is the strongestlink between the shoulder to the neck is broken,all the traction force is then transmitted to theneurovascular bundle (Fig. 1). This mechanism ofinjury causes greatest damage to the upperroots. While hyper-abduction of the shoulder orforceful widening of the scapulo-humeral anglemostly affects the C8 and T1 roots, the highvelocity traction injury can cause nerve rootavulsion from spinal cord (Fig. 2). The structuresprotecting the cervical nerve root from traction

5EtiologyPanupan Songcharoen

Page 57: Brachial Plexus Injuries

force are, firstly, the cone-shaped dural continu-ation in to the epineurium of the cervical spinalnerves, and secondly, fibrous attachmentsbetween the epineurium of spinal nerves C5, C6,and C7 and the cervical transverse process at theneural foramen. The absence of these ligaments

at C8 and T1 is the rationale behind the higherincidence of C8–T1 root avulsions compared withC5, C6, which sustain a higher incidence ofextraforaminal rupture. High velocity tractioninjuries are also incurred in speed-boat, car andski accidents.

48 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

(a, b) The sudden widening of the neck–shoulder angle causes a severe traction to the brachial plexus.

a b

Figure 2

Complete C5 to T1 root avulsionfrom the spinal cord from tractioninjury.

Page 58: Brachial Plexus Injuries

Low velocity traction injury has a much lowerincidence than the high velocity injury. Only 4 percent of patients in our series suffered this kind ofinjury. This injury is usually incurred by the victimfalling from a height and landing on the shoul-der, or may be caused by a heavy object fallingon an unprotected shoulder. In industrial situa-tions, the worker’s arm may be caught and pulledby a machine causing a stretch injury to theplexus. In sports, a rugby or American footballplayer tackling an opponent with head and shoul-der or a volleyball player practising heavyoverhead smashes may experience a transientparesthesia in the upper extremities. These lowervelocity injuries usually produce a lesser degreeof brachial plexus pathology.

Improper patient positioning in the operatingroom during general anesthesia can causetraction injuries to the brachial plexus (Cooper1991). The upper trunk can be injured due toprolonged extension and lateral bending of thehead to one side with the patient in the supine orlateral decubitus position. This posture willincrease the angle between the head and theaffected shoulder. The shoulder that is positionedon a sandbag or a roll can put the brachial plexusunder tension. Suspension of the arm from theoperating table screen in the lateral decubitusposition may stretch the brachial plexus,especially when the arm is in hyper-abduction.

Excessive abduction of the arms in either theprone or supine position (e.g. position for spinalprocedure) also causes stretching of the brachialplexus.

Compression lesions

The brachial plexus may be compressedbetween the clavicle and first rib. The compres-sion occurs when the traumatic force exerts itselfon the shoulder in the cephalocaudal direction.The bone fragments from a cervical transverseprocess fracture can compress the cervical nerveroots and a coracoid process fracture cancompress the lateral cord and musculocutaneousnerve. The fracture of the neck of scapula,humeral neck fracture and anteriorly dislocatedhumeral head compresses the posterior cord andaxillary nerve. The fracture of the scapular spinecan compress the suprascapular nerve.

Acute compression by seat belts may alsocause brachial plexus injuries in car accidents.Chronic compression by carrying heavy weightson the shoulders may cause a temporarybrachial plexus lesion.

Iatrogenic compression injuries to the brachialplexus in the operating room can occur byimproper placing of shoulder pads on anabducted arm while the patient is in a steepTrendelenburg position (Cooper 1991). If theshoulder pads are placed medial to the acromio-clavicular joint instead of indirectly over the joint,the pads directly press the brachial plexusagainst the first rib. Compression of brachialplexus can also occur as the arm and shoulderlie between the patient’s chest and the operatingtable when in the lateral decubitus position. Thebony structure of the shoulder and the armcompress the plexus against the rib cage.

Traction and compression

The complex trauma with multiple fracture ofcervical transverse process, clavicle, scapula, riband proximal humerus can cause both compres-sion and traction injury to the brachial plexus.The pathology on the plexus is usually diffusefrom nerve roots through terminal branches,disruption of brachial plexus can be found onmore than one site. This injury is usually associ-ate with vascular damage.

Open injury

Open injury of the brachial plexus is normallymuch less common than closed injury. In ourseries 4.3 per cent of patients sustained openinjury to the brachial plexus.

Sharp injury secondary to assault by knife inthe neck, chest or shoulder can directly injure thebrachial plexus. The open injury usually involvesonly part of the plexus. Associated vascular andintrathoracic injury can commonly be found. Thiskind of injury to the brachial plexus carries agood prognosis. The injured plexus can usuallybe treated by neurorrhaphy and nerve grafting.Iatrogenic sharp injury to the brachial plexusmay occur in brachial plexus block, tumor mass

ETIOLOGY 49

Page 59: Brachial Plexus Injuries

resection at the neck and supraclavicular area,occasionally an intact brachial plexus may bedamaged during attempts to stop severe bleed-ing from vascular injury or by insertion of subcla-vian lines.

Gunshot injuries to the brachial plexus may beencountered under several military and civiliancircumstances. This kind of injury may be accom-panied by life-threatening vascular or thoraciclesions. Gunshot injuries should be separatelyconsidered from sharp open injuries as there aresignificance differences in the extent and charac-ter of the neural and surrounding tissue damage.In gunshot injuries, the cause of tissue disruptionis the high velocity and the spin of the bullet. Thetissue is firstly crushed from direct contact withthe high velocity bullet and then stretched fromtemporary cavitation. The potential for tissuedisruption depends on the bullet mass, shape,construction and striking velocity (Omer 1991).

Low velocity gunshot injury is commonlyfound in civilian practice. This kind of injury wasfound in 2.7 per cent of patients in our series.This type of brachial plexus injury without anyassociated vascular injury carries a good progno-sis. More than half of the patients have a signif-icant functional recovery with non-operativetreatment.

High velocity gunshot injury has a greatertissue penetration. This kind of injury produces amore extensive and diffuse lesion where it isdifficult to define the extent of neural tissuedamage at the early stage. High velocity gunshotinjury is less common in civilian practice, andspontaneous functional recovery in less likely.

Shotgun injury usually produces extensivetissue damage and contamination. This kind ofinjury has less chance of spontaneous recoverythan low velocity gunshots.

Radiation injury

In general, peripheral nerves are considered tobe relatively radioresistant due to their deepsituation and low metabolic rate.

Radiation brachial plexus injury is sometimesfound in patients several years after radiationtherapy to the ipsilateral breast or axilla after treat-ment for breast cancer. Patients usually presentwith progressive motor and sensory deficit which

may be caused by radiation or compression ofrecurrent neoplasm. Investigations and meticulousclinical examination can clarify the diagnosis insome patients, but a number of patients willremain undiagnosed until a later stage. Surgicalexploration is difficult because of the fibrotic andischemic nature of the surrounding tissues.Despite the unrewarding results, surgical explo-ration is usually insisted upon by the patientbecause of the intractable pain.

Adult traumatic lesions of the brachial plexuscan also be classified according to the anatom-ical level within the plexus and their relation tothe clavicle. By using the clavicle as a reference,the lesion can be localized at three levels:supraclavicular, retroclavicular and infraclavicu-lar. To be more specific, the supraclavicularlesions are separated into two types in relationto the dorsal root ganglion: supraganglionicand infraganglionic. Although the clinical signif-icance of this classification is rather limited,because of the variation in other factors such asextent of the lesion, mechanism of injury, multi-level lesion, etc, this anatomical classificationmore or less helps the plexus surgeon to deter-mine the suitable type of surgical treatment andprognosis of the patient.

References

Alnot JY (1987) Traumatic brachial plexus palsy in theadult, Clin Orthop 237:9–16.

Cooper DE (1991) Nerve injury associated with patientpositioning in the operating room. In: Gelbermann RH,ed. Operative Nerve Repair and Reconstruction. JBLippincott: Philadelphia: 1231–42.

Malessy MJA (1999) Brachial Plexus Surgery. BVPasmans: The Hague.

McHenry LC (1969) Garrison’s History of Neurology. CCThomas: Springfield.

Omer GE (1991) Nerve Injury Associated with GunshotWounds of the Extremities. In: Gelbermann RH, ed.Operative Nerve Repair and Reconstruction. JBLippincott: Philadelphia: 655–70.

Songcharoen P (1995) Brachial plexus injury inThailand: a report of 520 cases, Microsurgery 16:35–9.

50 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 60: Brachial Plexus Injuries

Introduction

Direct surgery of the brachial plexus necessitatesthe level of technical equipment and team profi-ciency found in all vascular and nerve microsur-gical centers. Operating times are relatively long,from 3 to 10 hours depending on the nature ofthe lesions. The incision starts at the mastoid,follows the sternocleidomastoid muscle, then theanterior aspect of the clavicle and the deltopec-toral groove, and may extend, with a Z-shapedplasty, into the axillary region and even into thebrachial canal. When the preoperative diagnosisis precise, a shorter incision may suffice,centered on either the supraclavicular or thedeltopectoral area.

In the first 100 operations of our series, wewere led three times to divide the clavicle; thiswas done to gain better access to the C8 and D1roots that were adhering to the subclavian arteryor was made necessary by a significant retrac-tion of the fasciculi behind the clavicle.

Conversely, in the last 100 cases, the percent-age of clavicle divisions rose to 13 per cent. Wenow consider that this gives safety in controllingthe whole length of the subclavian artery andpermits better identification of lesions of the C8and D1 roots. Division of the clavicle has a lowcomplication rate when the osteosynthesis isperformed using screwed-on plates. To date, in23 clavicle divisions we have seen only oneseptic condition, and that had no adverse conse-quence on bone healing. Finally, this proceduredoes not lengthen the operative time; on thecontrary, it may shorten it by making dissectioneasier, especially when the most proximal rootsof the brachial plexus are retracted behind theclavicle.

Therapeutic options in the field of direct repairof brachial plexus lesions keep getting better.Only one matter remains controversial today,namely neurolysis. In our view, end-to-endsuture of brachial remains the most frequentlyused technique. Neurotizations are procedures oflast resort, reserved for the most severe plexuslesions.

Neurolysis

Microsurgical neurolysis of the brachial plexus isa controversial procedure with uncertain results,but could be indicated if a conduction block(neuropraxia) does not resolve itself sponta-neously. This condition is generally caused by adense perineural fibrosis. Fibrotic reaction isusually triggered by post-traumatic hematoma,but can also be found following stretch injuries.Operative palpation of roots and fasciculi revealsthe characteristic intraneural sclerous nodules.Only intraneural and interfascicular neurolysiscan free the fasciculi, but the method provokesdevascularization with a risk of a new fibrosis.

It is often the surgeon’s experience that deter-mines the type of neurolysis used. The danger inperforming a strictly extraneural neurolysis isthat it may leave undetected a complete rupturehidden under the apparent continuity of fibroticepineurium. Fibrosis and lack of flexibility of thenerve trunk must impel the surgeon to beaggressive and look for the fascicular rupture byintraneural dissection.

Our results demonstrate the difficulty of theprocedure. Objective amelioration after neuroly-sis was observed in less than 50 per cent of our

6Surgical techniques: neurolysis,sutures, grafts, neurotizationsMichel Merle and Aymeric Lim

Page 61: Brachial Plexus Injuries

patients; 10 per cent were made worse, and 40per cent had no amelioration following thismicrosurgical technique.

Experience has shown that when clinical andelectromyographic manifestations of nerveregeneration tend to stagnate, it is legitimate topropose a neurolysis before the sixth monthpost-injury.

Sutures

End-to-end suture

In our series of 386 cases of surgical procedureson the brachial plexus, we have done only fiveprimary repairs for lacerations, all of them bysuture.

In these cases of partial lesions the prognosisis good, because it is relatively easy to reconsti-tute the general fascicular organization of thenerve trunks. Such partial wounds must berepaired immediately. Neglecting them leads todangerous and complex secondary microsurgicalprocedures; during these, intraneural dissectionis required to separate the neuroma fromuninjured fasciculi, with a significant risk oflesion to these intact structures.

Early in the series we attempted immediaterepair of stretch injuries of the brachial plexus.The four cases operated on following thisscheme had to undergo secondary surgery, withfascicular grafting or neurotization. Fresh lesionsof the brachial plexus usually necessitate dissec-tion in the middle of a diffuse hematoma.Because we had underestimated the extent ofthe plexus lesions, the primary repairs by suture(or in a few cases by grafts) failed, andsecondary procedures were necessary.

End-to-side nerve suture

In 1992, Viterbo et al re-established the principleof neurotization by end-to-side nerve suture bedemonstrating its efficacy in the rat. Kennedy(1901), Ballance et al (1903), Harris and Low(1903) and Gatta (1938) were the first to applythis concept, but the results were far fromconvincing. Eventually, Lundborg et al (1994)

were able to show that it is possible for axono-lateral regeneration to produce function insecondarily innervated muscles. Mennen (1998)reported a series of 22 patients, all of whom hadundergone an end-to-side nerve suture. Anepineural window was made in that intact andreceiving nerve, taking care not to disrupt thefascicles. The proximal extremity of the otherwas then implanted with 8.0 prolene sutures.Four of the 22 cases involved neurotization of thebiceps, and of these, two patients recoveredpower of M4.

Grafts: free and vascularized

Most traumatic ruptures of the brachial plexuscan be repaired by fascicular grafts according toMillesi’s principles (Millesi et al 1973; Millesi1977). Appropriate restoration of the fascicularorganization of the brachial plexus is a ‘missionimpossible’. In the last 100 years or so, therehave been numerous attempts at clarifying theanatomy of the brachial plexus; these studiesstarted with Herringham (1886), Agostini (1887)and Adolfi (1898), followed by Kerr in 1918, KoHirasawa (1928) and Billet (1933). More recently,Sunderland (1977) and Seddon (1972a, 1972b)have contributed to the study of the fascicularorganization of the brachial plexus. Their studiesshow that there is no systematization of thebrachial plexus, that variations from root to rootamong individuals are great, and that both sidesof the same individual do not have the sameorganization.

The in-depth studies of Bonnel (Bonnel andRabischong 1976; Bonnel 1977) have shown thatfascicular organization in a given root varies witheach subject. Microsurgery helps to analyze thefascicular disposition and enhance the quality ofgraft/fascicle abutments. The quality of theseabutments and the rapidity with which surgerycan be performed have been further improved bythe recent availability of biologic glues such asBiocol and Tissucol (Merle et al 1987).

Microsurgical techniques also make it possibleto remove the connective tissue surrounding thefascicular groups. Bonnel has showed that 70 percent of a cross-section of the axillary nerve isoccupied by connective tissue; the ulnar nervecontains even more at 82 per cent (Bonnel and

52 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 62: Brachial Plexus Injuries

Rabischong 1976; Bonnel 1977). The repair strat-egy is greatly influenced by the type of lesionand by the number of roots or trunks that can berepaired.

First, the brachial plexus is dissected and thelesions are precisely identified, keeping in mindthat supraclavicular lesions are often in continu-ity; then plexus root trunks are sharply cut backto healthy neural tissue using Meyer’s guillotineknife. Nerve grafting is performed according toMillesi’s technique (Millesi et al 1973; Millesi1977). In the past few years, this technique hasbeen modified by the use of biologic glues suchas Tissucol (Merle et al 1987). Formerly we usedtwo or three stitches of 9.0 and 10.0 polypropy-lene monofilament to suture each fasciculargroup. This procedure was time-consuming andsomewhat damaging to perineural and neuraltissue.

Our experience of gluing nerve grafts hasshown that results obtained are significantlysuperior to those observed following conven-tional suturing; gluing facilitates abutment of thenerve substance and also minimizes handling ofthe nerve ends, thereby reducing unwantedmushrooming.

Short grafts (5–7 cm long) are used to repairthe upper trunk; surrounding tissues have afavorable influence on graft revascularization,which is obtained in a few days. Conversely,extensive lesions necessitate long grafts(10–15 cm), and the tissue environment is lessfavorable. In these cases, we prefer to use vascu-larized nerve grafts. The principle of vascularizedgrafts was developed by Taylor for the repair oflarge losses of substance in peripheral nerves(Taylor and Ham 1976). Later, Comtet put intouse the principle of vascularized transfer of thecutaneous antebrachii medialis nerve. Birchpresented a large series of plexus brachial graftsusing the vascularized ulnar nerve of the forearm(Bonney et al 1984). In our protocol, we use theulnar nerve vascularized by the collateral proxi-mal ulnar artery originating at the brachial artery.This donor site has the advantage of not inter-rupting the continuity of the principal vascularaxis, which are the cases when the forearm ulnarnerve is used. According to Lebreton, the collat-eral proximal ulnar artery exists in 94 per cent ofcases (Lebreton et al 1983). The vascularizedgraft can be used as a free graft as well as atransposed local graft. The vascularization of the

graft facilitates axonal progression, and withinthe first 6 months after vascularized grafting,progression of the Tinel sign averages 2 mm perday. Contractions of the biceps muscle appear atabout the ninth month. With conventional grafts,equivalent results are not seen before the twelfthmonth. We have observed, however, that resultsobtained with conventional grafts and withvascularized grafts tended to be similar in thelong run. Nevertheless, we believe that resultsare more consistent after vascularized grafts, aslong as there is no embolization of the anasto-mosis. In our series of 13 cases of vascularizedgrafts on the brachial plexus, seven were freegrafts with arteriovenous anastomosis and sixwere pediculized grafts. The length of the freegrafts varied from 12 to 18 cm; that of thepedicled grafts varied from 23 to 26 cm. We haveobserved three results at M4, six at M3+, two atM3–, and two at M2.

This technique, however, is of limited usebecause it implies the intradural avulsion of theC8 and T1 roots. Also, in the event of thrombo-sis of the vascular anastomosis, results will beinferior to those of conventional grafts becausethe central part of the grafts will undergo necro-sis.

Vascularized allografting is a way of gettingaround the problem of donor site scarcity, butmakes immunosuppressant treatment indispens-able. Studies by Bour et al (1986) have shownthat nerve regeneration is possible throughvascularized allografts as long as cyclosporine isadministered; interruption of the immunosup-pressive treatment brings about deterioration ofthe graft and, consequently, of the functionalcondition.

In the present state of experimental studies,use of these vascularized allografts is not justi-fied because continuous postoperative immuno-suppressive treatment is unacceptable inreconstructive surgery.

Neurotizations

Intradural avulsions can involve one, several oreven all the roots of the brachial plexus.Neurotization aims to salvage the functional ortrophic condition of the upper limb. The princi-ple of neurotization dates from the turn of the

SURGICAL TECHNIQUES 53

Page 63: Brachial Plexus Injuries

century. Harris and Low were the first to proposeit in 1903, followed by Tuttle who, in 1913,suggested the use of the spinal accessory nerve.Vulpius and Stoeffel (1920) suggested the use ofthe N. pectoralis. In 1929, Foerster used thethoracicus longus nerve. In 1948, Yeoman andSeddon (1961) and Fantis and Sîezak (1965)performed neurotization with the intercostalnerves, and the concept was taken up again byTsuyama in Japan in 1972 (Tsuyama and Hara1972). In 1980, Brunelli proposed using the wholecervical plexus (Brunelli and Monini 1984). As forus, we followed Kotani and Allieu and used thespinal accessory nerve, which comprises about1700 fibers, to reinnervate the biceps muscle(Kotani et al 1971; Allieu et al 1982).

Neurotization of the biceps usingthe spinal nerve

The spinal accessory nerve is used at its exitpoint from the sternocleidomastoid muscle; a5–10 cm long graft connects it to the fasciculargroup of the musculocutaneous nerve, which hasusually been dissected interneurally for a lengthof 4 cm in the lateral cord.

Out of 14 neurotizations of the biceps muscle,we obtained six good or very good results, twoaverage, and six poor or null results. Narakas, bypooling the results of all spinohumeral neuroti-zations performed by seven surgical teams, hasshown that 37 per cent of the cases had usefulresults as far as the biceps was concerned(Narakas 1989; Narakas et al 1989).

We continue to neurotize the biceps using thistechnique; it is advantageous because functionalresults are acceptable in nearly 40 per cent of thecases and because the inferior portion of thetrapezius muscle is preserved, being innervatedin most cases by rami from the cervical plexus(Bonnel and Rabischong 1976).

Neurotization using the intercostalnerves

There is an alternative to the use of bicepsneurotization; neurotization by intercostalnerves. Seddon’s idea was taken up again by

Tsuyama (Seddon 1972b; Tsuyama and Hara1972). Three or four intercostal nerves are used.

Respiratory potentials sent into the bicepsmuscle can restore true function to the elbowafter a period of time. The fatigue phenomenonsets in relatively quickly, but there is someadaptation in the long run: respiratory potentialsfade away and are replaced by voluntary muscleactivity. This method is presently pursued inJapan, where it was developed by Tsuyama(Tsuyama and Hara 1972). In Europe it is usedequally with neurotization by the spinal acces-sory nerve and complementary neurotization bythe cervical plexus as developed by Brunelli(Brunelli and Monini 1984). Narakas prefers touse the spinal accessory nerve to neurotize thesuprascapularis muscle, and to neurotize thebiceps muscle by intercostal nerve transfer(Narakas 1989).

Neurotization of the biceps usingthe intact ulnar nerve

In C5, C6, C7 root avulsions, Oberlin et al (1994)have suggested using 10 per cent of the fasciclesof the ulnar nerve for direct anastomosis with themotor nerve to the biceps. The suture is techni-cally easy, because the musculocutaneous andthe ulnar nerves are found side by side. In aseries of 18 cases reported by Loy et al (1997),seven out of eight patients who presented with aC5, C6 avulsion recovered elbow flexion, and fourout of nine patients presenting with C5, C6, C7root avulsions recovered flexion by this neuroti-zation technique only. Clinically, no sensorimotordeficits were created in the territory of the ulnarnerve. By intraoperative electrostimulation, theauthors prefer to use one or two of the fasciclesdestined for the extrinsic flexors, flexor digitorumprofundus or flexor carpi ulnaris.

Neurotization of the brachial plexusby the C7 contralateral nerve root

In 1989, Chuang (1999) reviewed the principle ofcontralateral C7 transfer, a technique firstproposed by Gu in 1986 (Gu 1989). Gu hadproceeded in two operative stages. The first

54 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 64: Brachial Plexus Injuries

involved anastomosis of the contralateral C7nerve root with a pedicled ulnar nerve graft takenfrom the side of the paralyzed limb. Then, 8–12months later, the distal part of the reinnervatedulnar nerve graft was sutured to the avulsedplexus, giving precedence to the muscolocuta-neous nerve, the median nerve, the axillarynerve and then other nerves.

Chuang prefers to transfer the ulnar nerve withthe ulnar artery and thus directly neurotize theplexus. In a series of 67 patients (Chuang et al1993), Chuang reported useful results in 67 percent of cases. He noted that sectioning of the C7root can create a temporary weakness of thetriceps of the donor upper limb and dyesthesiasover the skin of the dorsum of the hand suppliedby the radial nerve. The main difficulty for thepatient is producing movement on the contralat-eral side. In our experience, 18–24 months arerequired before the patient can integrate themovement of elbow flexion. This techniqueshould only be used when no homolateral neuro-tizations are possible.

Whichever technique is used, neurotization hastwo goals; to avoid amputation of the paralyzedarm and to provide trophic condition andcomfort compatible with everyday life. The resultis usually a live arm that performs as poorly asa prosthesis or even worse. The brachial plexuscontains at least 150 000 fibers, and the numberof fibers usable by the surgeon for arm neuroti-zation is no greater than 1300 in each intercostalnerve and 1700 in the spinal accessory nerve.

References

Adolphi H (1898) Uber das Verhalten der zweitenBrustnerven zum Plexus Brachialis beim Menschen,Anat Auz XV:25–36.

Agostini C (1887) Sulla composizione del plessobrachiale e sulle origini des susi rami verminali. PerginiVinceuzo: Santucci.

Allieu Y, Privat JM, Bounel F (1982) Les neuotisationspar le nerf spinal dans les avulsions radiculaires duplexus brachial, Neurochirurgie 28:115.

Ballance CA, Ballance HA, Steusard P (1903) Remarkson the operative treatment of chronic facial palsy ofperipheral origin, J Br Med 1:1009.

Billet H (1933) Les troncs primaire du plexus brachial,CR Assoc Anat Lisbonne 32.

Bonnel F (1977) Configuration interne histophysi-ologique du plexus brachial, Rev Chir Orthop 63:35.

Bonnel F, Rabischong P (1976) Contribution à l’étudede la Systématisation du Plexus Brachial. Actulatiés enReeducation et Readaptation Fonctionnelle. Massun:Paris.

Bonney G, Birch R, Jamieson AM, Eames RA (1984)Experience with vascularized nerve graft, Clin PlastSurg 11:137–42.

Bour C, Jupiter JB, Delamonte Schmitt RJ (1986)Experimental studies of vascularized nerve allografttransplantation. Rabbit model, Eur Surg Res Supp 1:82.

Brunelli G, Monini L (1984) Neurotization of avulsedroots of brachial plexus by means of anterior nerves ofcervical plexus, Clin Plast Surg 11:149.

Chuang DCC, Wei FC, Noordhoff MS (1993) Cross-chestC7 nerve grafting followed by free muscle transplanta-tions for the treatment of total avulsed brachial plexusinjuries: a preliminary report, Plast Reconstr Surg92:717–25.

Chuang DDC (1999) Controlateral C7 transfer (CC–MT)for avulsion injury of the brachial plexus. Techniquesin hand upper extremity, Surg 3:185–92.

Fantis A, Sîezak Z (1965) On the possibility of reinner-vation of total lesions of the brachial plexus by inter-costo-plexual anastomosis, Cesk Neurol 28:412.

Foerster O (1929) Die Therapie der Schussverletzungender Peripheren Nerve in Lewandoske – Handbuch derNeurologie. Springer: Berlin: 1677–91.

Gatta R (1938) Lateroterminal anastomosis of nervetrunks, Arch I Vol Chir 48:155.

Gu YD (1989) Microsurgical treatment for root avulsionof brachial plexus, Natl Med J Chir 69:653.

Harris W, Low VW (1903) On the importance ofaccurate muscular analysis in lesions of the brachialplexus, Br Med J 2:1035.

Herringham H (1886) The minute anatomy of thebrachial plexus, Proc R Soc XII 249:423.

Kennedy R (1901) On the restoration of co-ordinatedmovement after nerve-crossing, with interchange offunction of the cerebral cortical centers, Phil Trans RSoc London (Biol) 194B:127.

SURGICAL TECHNIQUES 55

Page 65: Brachial Plexus Injuries

Ko Hirasawa M (1928) Uber den Plexus Brachialismitterlung der Wurzeln des Plexus Brachial. Impressioseparato exactis Scholae Medianae. UniversitatisImperialis Kiotoenis XI, 98.

Kotani T, To Shima Y, Matsude H (1971) Post-operativeresults of nerve transposition in brachial plexus injury,J Orthop Surg 22:963.

Lebreton E, Bourgeon Y, Lascombes P et al (1983)Systématisation de la vascularisation du nerf ulnaire aubras en vue de son transfert microchirurgical, Ann ChirMain 2:211–18.

Loy S, Bhatia A, Asfazadourian H, Oberlin C (1997)Ulnar nerve fascicule transfer to the biceps musclenerve in C5–C6 or C5–C7 avulsions of the brachialplexus. Nineteen cases, Ann Chir Main Memb Super16(4):275–84.

Lundborg G, Zhao Q, Kanse M et al (1994) Can sensoryand motor collateral sponting be induced from intactperipheral nerve by end-to-end anastomosis? J HandSurg 19B:277–82.

Mennen U (1998) End-to-side nerve suture in thehuman patient, Hand Surg 3:7–15.

Merle M, Becker C, Pankovic C, Bagot D’Arc M (1987)La reparation microchirurgicale des nerfs periph-eriques et des vaisseaux par le Tissucol. Etude cliniqueet experimentale. Revue de laryngologie, Otol Rhinol108:13–14.

Merle M, Dautel G (1991) Vascularised nerve grafts, JHand Surg 16B:483–8.

Millesi H (1977) Surgical management of brachialplexus injuries, J Hand Surg 2:367.

Millesi H, Meissl C, Katzer H (1973) Zur Behandlung derVerîetzungen des Plexus Brachialis, Vorschlag einerintegrierten Therapie, Brun’s Beitr Klin Chir 220:4–429.

Narakas A (1989) Neurotisation par le nerf spinal dansles lesions du plexus brachial. In: Alnot JY, Narakas A,eds. Les Paralysies du Plexus Brachial. ExpansionScientifique Française: Paris: 162–72.

Narakas A, Allieu Y, Alnot JY et al (1989) In: Alnot JY,Narakas A, eds. Les Paralysies du Plexus Brachial.Expansion Scientifique Française: Paris: 191–9.

Oberlin C, Beal D, Leechavengvongs S et al (1994)Nerve transfer to biceps muscle using a part of ulnarnerve for C5–C6 avulsion of the brachial plexus:anatomical study and report of four cases, J Hand Surg19A:232–7.

Seddon HJ (1972a) Disorders of the Peripheral Nerves.Churchill Livingstone: Edinburgh.

Seddon HJ (1972b) Intercostal nerve transfer into themusculo-cutaneous nerve. In: Seddon HJ, ed. SurgicalDisorders of the Peripheral Nerves. Churchill Livingstone:Edinburgh.

Sunderland S (1951) A classification of peripheralnerve injuries producing loss of function, Brain 74:491.

Sunderland S (1977) Nerve and Nerve Injuries, 2ndedn. Churchill Livingstone: Edinburgh.

Taylor GI, Ham F (1976) The free vascularized nervegraft, Plast Reconstr Surg 57:413.

Tsuyama N, Hara T (1972) Intercostal nerve transfer inthe treatment of brachial plexus injury of root-avulsiontype. Experta Medica. In: 12th Congress of theInternational Society of Orthopaedic Surgery andTraumatology (SCIOT) Tel Aviv, 351–3.

Tuttle HK (1913) Exposure of the brachial pl;exus withnerve transplanatation, JAMA 61:15.

Viterbo F, Trindade JCS, Hoshino K, Mazzoni Neto A(1992) Latero-terminal neurorrhaphy without removalof the epineural sheath. Experimental study in rats, JSao Paulo Med 6:267–75.

Vulpius O, Stoeffel A (1920) Orthopaedische OperationsLehre, 2nd edn. Ruke: Stuttgart.

Yeoman PM, Seddon HJ (1961) Brachial plexusinjuries: treatment of the flail arm, J Bone Joint Surg43B:4933.

56 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 66: Brachial Plexus Injuries

Traumatic brachial plexus injuries are character-istic of young adults aged 18–20 who havesustained a motorcycle or car accident. Whateverthe clinical presentation, a patient showing norecovery at 4 or 6 weeks after a traumatic palsymust undergo investigation (CT-myelographyand EMG) in order to identify the surgical indica-tions early enough.

Our experience at Bichat Hospital includesmore than 1200 cases operated on between 1975and 1998, the results of which have beenpublished (Alnot 1993, Alnot et al 1993, Alnotand Oberlin 1993, Alnot 1995, Alnot et al 1996,Alnot et al 1998), as have the results of manyother authors (Narakas 1978, Brunelli and Monini1984, Millesi 1984, Sedel 1986, Terzis 1987, Allieuet al 1997). A better comprehension of patholog-ical lesions (Sunderland 1978, Millesi 1984, Alnotet al 1996) leads to a clearer classification ofinjuries, because classifications are essential ifwe want to evaluate the results, particularly afternerve repair, and the use of charts and diagramsdescribing the nerve injuries and the type ofrepair is a great help for further follow-up.

Lesions can be situated at any level from thebase of the nerve roots to the division of thebrachial plexus in the axillary regions, andseveral types of lesions can be differentiated:

• Supraclavicular lesions at root or primarytrunks level (75 per cent of cases);

• Infra- and retroclavicular lesions of secondarytrunks (10 per cent of cases);

• Terminal branches (15 per cent of cases).

Among the palsies due to supraclavicularlesions, the following can be distinguished:

• C5–C6 or C5–C7 palsies which occur in 20–25per cent of cases;

• C8–T1 palsies which occur in 2–3 per cent ofcases;

• C5–C8–T1 lesions which are the mostfrequent, occurring in 75–80 per cent of cases.

It is important to determine the exact site of thelesions as this will greatly bear on prognosis andon the future of the patient.

Careful clinical as well as paraclinical exami-nation should be rapidly undertaken in order toobtain an exact diagnosis and permit interven-tion within a period of 6 weeks to 3 months aftertrauma (Fig. 1).

Clinical examination, supported by paraclinicalinvestigations, permits evaluation of nervelesions according to the roots that have beenaffected, after which therapeutic indications anda prognosis can be made.

Some factors can be considered favorable and,for example, a patient with a brachial plexuspalsy secondary to dislocation of the shoulderdue to a minor trauma has a 90 per cent chanceof recovery (Sunderland grades 1 to 3) but insome cases there are ruptures at the level atterminal branches with indication for repair.

By contrast, the following factors carry a poorprognosis:

• Violent trauma involving the upper limb aswell as the plexus. Multiple bone fracturesand other traumatic lesions are frequentlyfound in the injured upper limb (21 per centof all cases); bone lesions are found in 58 percent of these cases and vascular lesions in 11per cent;

• Serratus anterior involvement and presence ofHorner’s syndrome, both of which indicate aproximal lesion;

• Presence of pain and medullary signssuggesting injury to the cord.

7Supraclavicular plexus injuriesJean Y Alnot

Page 67: Brachial Plexus Injuries

Repeated clinical examination shows the evolu-tion over time, and if there is no clinical recov-ery within a period of 30 days, paraclinicalinvestigations such as CT-myelography and EMGshould be done. The value of CT scan-myelogra-phy is to establish the existence of root lesionsand especially root avulsion with pseudo-meningocele (rupture of the dura sheath androotlets).

In peripheral mechanisms, depending on thedirection of traction, the forces have an action onanterior and posterior rootlets.

Avulsion is a very specific injury at the level ofspinal rootlets, and is beyond surgical possibilitiesfor repair. The rootlets are avulsed from the spinalcord, notably C7 and mostly C8 and T1 whichbecome horizontal with abduction of the arm.Superior C5 and C6 roots, because of their oblique

route, are often ruptured more distally in thescalenic region. Nerve injury can also be locatedimmediately next to the transverse canal withfrequently longitudinal disruption injuries andstaged ruptures of the nerve fascicular groups,from the transverse canal to the interscalenicspace. Nerve repair remains possible in somecases, but the proximal stump may, to someextent, have lost its possibility of axonal regener-ation. Besides, damage to the fascicular groups inthe transverse canal can cause retrograde degen-eration of nerve fibres, involving the motor cell inthe ventral horn or the sensory cell in the poste-rior ganglion, which is equivalent to avulsion.

Rupture can be more distal in the supraclavic-ular area, between the scalen muscles andbeyond, preserving a proximal stump of root orof trunks of variable length and quality, possibly

58 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

(a, b) Charts for recording clinical and paraclinical findings.

a b

Page 68: Brachial Plexus Injuries

available for nerve repair; EMG is also a part ofthe assessment and is particularly useful inC5–C6 and C5–C7 palsies.

The surgical indications depend on the clinicalevolution and decisions can be based on all thefactors discussed above. In our experience thereis no indication for immediate nerve repair; thepalsy is assessed by repeated motor and sensoryexamination and CT-myelography must be doneif there is no clinical and electrical recovery after4 to 8 weeks, depending on the general condi-tion of the patient. The prognosis depends on theanatomy and type of lesions and the early thera-peutic indications (first to third month) alsodepend on knowledge of the anatomical lesions.

Surgical procedure

The operation is carried out through a long zig-zag cervico-axillary incision and the wholeplexus must be explored. Two types of skinincision are described depending on the authors:

• A large ‘Z’ incision including a vertical cervi-cal incision at the posterior border of thesternocleidomastoid (SCM) muscle; a horizon-tal subclavicular incision, and a verticalincision in the deltopectoral groove;

• Multiple zig-zag incision at the level of thecervical area to avoid a retractile scar, andopen ‘V’ incisions in the subclavicular areaand in the deltopectoral groove.

The key of the cervical approach is the omohy-oid muscle. It must be located at the beginningof the dissection and must be transsected in itsmiddle and retracted laterally in order to exposethe supraclavicular plexus. The approach is doneat the posterior border of the SCM muscle; theexternal jugular vein must be preserved with theposterior SCM muscle belly. The lateral trans-verse branches must be ligated but the nervebranches of the superficial cervical plexus mustbe preserved.

At the upper part of the triangle made by theSCM and the trapezius muscle, the C4 loop mustbe preserved on to the SCM muscle belly and itrepresents an important topographic landmark.Then, the scalene outlet must be exposed andthe phrenic nerve, at the anterior aspect of the

anterior scalenus muscle, must be located andstimulated. The transverse cervical artery andvein must be ligated to complete the expositionof the plexus.

The suprascapular nerve is an essentiallandmark as there is no nerve element lateral to it.It is also important to locate the Charles Bell nervewhich must be preserved. Finally, the spinal acces-sory nerve must be dissected if a neurotization isscheduled. It is important not to dissect it proxi-mally in order not to destroy the branches for theupper and mid-trapezium. If neurotization must bedone with the distal accessory spinal nerve thesection can be performed after the departure of thebranches for the upper trapezius muscle.

The key of the subclavicular and axillaryapproach is the pectoralis minor muscle. Theapproach in the deltopectoral groove must bewide and must respect the cephalic vein. Dis-insertion of the lateral part of the clavicular inser-tion of the anterior pectoralis major can beconvenient.

The pectoralis minor must then be dissectedand exposed, and in certain cases it may benecessary to divide it.

A communication between the cervical areaand the axillary area is then established underthe clavicle using a sponge held with a clamp. Inthe majority of cases, it is not necessary to cutthe clavicle; this additional step which has beenrecommended systemically by some authors isnow only done in sub- and retroclavicularbrachial plexus palsies.

The musculocutaneous nerve is identified as itenters the coracobiceps muscle. This must besystematic in all explorations in order to avoid adouble level lesion.

The other nerves are identified and the dissec-tion is performed distal to proximal and proximalto distal.

The axillary artery can be also located.A complete evaluation of the lesions is made

after exploration of the whole plexus. However,if there are meningoceles on C8 and T1 it is notnecessary to explore these roots locating in adeep area with a dangerous dissection.

At the end of the procedure, depending on thetypes of exploration and on the repair, closure isdone plane by plane without drain, or only incertains cases with a superficial drain at thedistal part of the incision, at some distance fromthe nerve grafts.

SUPRACLAVICULAR PLEXUS INJURIES 59

Page 69: Brachial Plexus Injuries

In the immediate postoperative period, immo-bilization of the resected area with the elbowagainst the thorax is done for a period of 3weeks, associated with a cervical collar if nervegrafts have been performed at the level of theroots.

The type of the repair depends on the local-ization of the lesions and, in a majority of cases,there is a C7–C8 and T1 avulsion. C5 and C6roots are vertical and will have lost their obliquedirection and they will then be explored in thescalene outlet.

Nerve grafts (sural nerve or more rarely theulnar nerve as a free or vascularized graft) willbe performed depending on the root rupture inthe scalenic area. Neurotizations must also beperformed (spinal accessory nerve, intercostalnerves, etc).

For intercostal nerve neurotization, a skinincision is performed below the pectoralis majormuscle. Three intercostal nerves (D3, D4, D5) aretransected anteriorly to perform a direct suturewith the musculocutaneous nerve, for example.Several clinical pictures can be described.

Complete palsy withoutrecuperation

Total palsies with avulsions of the lower

roots (64 per cent of cases)

CT-myelography shows pseudomeningoceles atC8 and T1 and often at C7. Root C6 may have anabnormal aspect and C5 is usually normal in theCT scan. In these total palsies with avulsion ofthe lower roots (C7, C8, T1) when there is onlyone or two roots that are ruptured in the scalenicarea, it is not possible to graft all the plexus. Thesurgery must be performed early (6 weeks to 3months) and our approach is to aim for re-inner-vation of the proximal territories. The patientsmust be informed that they will have definitiveparalysis of the hand.

The results depend on the anatomo-patholog-ical lesions. When only one root (C5) can begrafted (Fig. 2a), our choice will be to repair theanterior part of the first trunk with the C5 rootand the suprascapular nerve by neurotization(direct suture) with the spinal accessory nervewhich is divided distal to the origin of the

60 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 2

Total palsies. (a) Avulsion C6–C7–C8–T1, one graftableroot. (b) Avulsion C7–C8–T1 two graftable roots. Ax:axillary; M: median; MC: musculocutaneous; SS: supras-capular; X1: spinal accessory.

a

b

Page 70: Brachial Plexus Injuries

branches for the superior and middle parts of thetrapezius. The goal is to obtain stabilization ofthe shoulder, an active pectoralis major toadduct the arm, flexion of the elbow and somepalmar sensibility in the forearm and palm.

When there are two roots (C5–C6) (Fig. 2b) thatcan be grafted, it is possible also to graft someparts of the posterior cord for radial or axillarynerve function. Every effort is made to connectthe anterior plane of the root grafts with theanterior plane of the plexus and the posteriorplane of the root with the posterior plane of theplexus in order to respect the cortical represen-tation and to avoid cocontractions betweenantagonist muscles. One must not try to grafteverything, and if the roots are thin thetechnique is similar to that used for one graftableroot. From the technical viewpoint, we mostfrequently use the sural nerve but when T1, C8and C7 are proved to be avulsed, it is possibleto use a vascularized ulnar nerve graft either freeor pedicled when the aspect and size of C5 orC5–C6 are good and when the length at the nervedefect is longer than 15 cm.

The results must be analyzed critically evalu-ating both motor and sensory function. They canbe evaluated only after sufficient time haselapsed because re-innervation after nerve graft-ing is always delayed. This requires 2 to 3 years,depending on the type of lesion and its location(roots, trunks, cords, and terminal branches), andthe results must be evaluated according to thefunction of the structures that have beenrepaired and, therefore, the therapeutic objec-tives achieved.

Finally, the pain syndrome and it is importantto stress that surgical interventions with nerverepair for any given region considerably modifythe afferents originating in the upper limb.

The final functions (Allieu et al 1997, Alnot etal 1993, Alnot et al 1996, Alnot and Narakas 1996,Alnot 1995, Brunelli and Monini 1984, Narakas1986, Rusch et al 1995, Sedel 1986, Terzis 1987)must be studied according to the nerve repairand depend on the number of grafted roots, anda useful result means at least that elbow flexionis possible. In our experience 75–80 per cent ofthe patients have had satisfactory results withgood elbow flexion M3 + M4. The pectoralismajor function is obtained in 60 per cent ofcases, allowing the possibility to hold objectsagainst the thorax.

The shoulder poses problems but it is possibleto obtain stabilization of the shoulder, some activeabduction and external rotation in 50 per cent ofcases by spinal accessory nerve neurotization;some authors perform shoulder arthrodesis.

It is rare to obtain function in the hand but inthe majority of the cases, a ‘shovel hand’ or‘paperweight hand’ is still useful to stabilize anobject on a table. Finally, some sensation in theforearm and hand is obtained and this in factmay explain why 80 per cent of the patients dosuffer little or no pain.

Total palsies with avulsions of all roots(24 per cent of cases)

CT-myelography shows meningocele or lacunaon all the roots, and therefore no root is avail-able for repair. In these cases (Fig. 3), neurotiza-tions are indicated using the spinal accessorynerve, the cervical plexus, the intercostal nerves,

SUPRACLAVICULAR PLEXUS INJURIES 61

Figure 3

Total palsies with avulsion of all the roots. Possibilities ofneurotizations. MC: musculocutaneous.

PlexusSpinalaccessorynerve

Page 71: Brachial Plexus Injuries

and more rarely the hypoglossus nerve or thecontralateral C7 root (Narakas 1996, Brunelli andMonini 1984, Alnot et al 1993, Allieu et al 1997).The goal is to provide elbow flexion by neuroti-zation of the musculocutaneous nerve; this canbe associated with shoulder arthrodesis.

When using the spinal accessory nerve associ-ated with the superficial cervical plexus, an inter-vening autograft is necessary with two strands ofsural nerve. The spinal motor fibres are connectedto the lateral part of the musculocutaneous nervetrunk and we add sensory fibres from the cervicalplexus connected to the medial part of the muscu-locutaneous nerve. When using intercostalnerves, the neurotization can be performed bydirect suture between three intercostal nerves 3–5divided in their anterior portion and the musculo-cutaneous nerve (Hara’s technique).

The results are good in 75 per cent of caseswith elbow flexion at M3 + M4 and the only issuethat remains is to know whether we can dobetter using some other neurotization.

Partial palsies C5–C6 or C5–C7 (25 per cent of cases)

The clinical pictures can be either C5–C6 orC5–C7 palsies or an initial total palsy with rapidrecovery in C8–T1.

The prognosis is dominated by the fact that thehand is normal or only partially involved butuseful. Surgery must be done early because thelesions are often in the scalenic area on the rootsor upper trunk with a good possibility for nerverepair with a satisfactory result.

Concerning the surgical approaches themusculocutaneous nerve is identified as it entersthe coracobiceps muscle and dissection fromdistal to proximal then allows to dissect thelateral cord and the anterior component of theupper trunk, as well as the posterior trunk andthen the posterior component of the upper trunk.

Then, the lesions must be located and repairedin the scalene space between the anterior andmiddle scalene muscles.

Lesions of C5–C6 and possibly C7 roots, areevaluated and if there are ruptures in the scalenicarea it is possible to perform grafts.

On the other hand, in C5–C6 avulsion weperform a medial approach of the upper arm,

approximately 120 mm distally below theacromial process in order to perform a neuroti-zation of the biceps nerve with a bundle of theulnar nerve (Oberlin 1994).

Nerve reconstruction and muscular transfer willbe studied according to a global scheme (Alnotand Oberlin 1993, Alnot et al 1998, Rostoucher etal 1998) and the indications are derived from theanatomo-pathological lesions (Fig. 4).

In C5–C6, palsies, when the two roots aredisrupted in the scalenic area (Fig. 4a) it is possi-ble, if they are a good size and aspect, to graftall the lesions. However if the size of the roots istoo small or when only one root is available (Fig.4b), one must not disperse the nerve fibres andthe graft must be performed to the anterior partof the first trunk.

A neurotization of the spinal accessory nerveadded to the suprascapular nerve gives betterresults than a graft from C5 with one strand ofthe sural nerve.

We can also perform a graft on the axillarynerve and use ulnar biceps neurotization. Finally,in C5–C6 palsies when no roots are available(Fig. 4c), we neurotize the spinal accessory nerveto the suprascapular nerve and perform at thesame operation a neurotization using a fascicu-lar group of the ulnar nerve with direct suture onthe biceps nerve (Oberlin 1994, Loy et al 1997,Leechavengvongs et al 1998) or perhaps afterend to side neurorraphy (Viterbo et al 1994,Franciosi et al 1998).

The present results of this ulnar biceps neuro-tization are good in C5–C6 palsies but the resultsare still uncertain in C5–C7 palsies and musculartransfer must be discussed and performed in thesame operating time.

Concerning C5–C7 palsies, the overall plan issimilar but is complicated by the severity of thelesions. When all the roots are avulsed and whenthere are no acceptable possibilities for muscu-lar transfer, we perform associated neurotiza-tions to restore elbow flexion and shoulderstability.

In C5–C6 palsies, active elbow flexion must beobtained in all cases by nerve surgery or muscu-lar transfer. The shoulder can be stabilized afteraccessory nerve neurotization with recoveryactive anteposition, and active external rotation isessential in order to allow more functional elbowflexion. However, the shoulder remains the mainproblem with 73 per cent of good results in C5–C6

62 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 72: Brachial Plexus Injuries

cases and only in 49 per cent of C5–C7 casesbecause of the greater severity of the lesions.

Rotation osteotomy of the humerus, ligamen-toplasty and arthrodesis can all be used.Concerning the axillary nerve an end-to-sideneurorraphy is a new possible way (Viterbo et al1994, Franciosi et al 1998). Finally, concerningthe wrist and the hand in C5–C7 palsies, there isalways the possibility of muscular transferswhich can be decided on early.

Partial palsies C8–T1

In these cases a decision regarding surgery willbe based upon results of the clinical examinationand other diagnostic studies. Here again, theprognosis is determined by the degree of handfunction and the severity of the nerve lesions.Myelography will reveal the presence of absenceof pseudomeningoceles, and based on thedegree of diagnostic certainty of the existence of

avulsion of different roots, the clinical statusshould be re-evaluated and a decision regardingsurgical exploration made. If pseudomeningoce-les involve the lower roots, exploration is notjustified and one proceeds to muscle transfers(Alnot 1993, Alnot and Oberlin 1993).

However, if the myelograms are normal butspontaneous regeneration has not occurred,surgery is appropriate for assessment and possi-bly nerve repair. It is important to remember ifC8 and T1 roots or even some more distallesions of the trunk and cords can be repaired bynerve grafts; the distance between the nervelesions and hand precludes re-innervation of theintrinsic muscles.

Conclusion

An update of this problem has been published inthe monograph of Alnot and Narakas (1996) withmany contributors.

SUPRACLAVICULAR PLEXUS INJURIES 63

Figure 4

C5–C6 palsies. (a) Two graftable roots with good aspect and size. It is possible to graft all the lesions. (b) One graftableroot C5 possibilities of repair associated with accessory nerve neurotization. A similar repair is performed if there are twograftable roots with a small size. (c) C5–C6 avulsion. Accessory nerve neurotization and ulnar biceps neurotization.Muscular transfers are also possible in C5–C6–C7 palsies. R: radial; M: median; SSsuprascapular; MC: musculocutaneous;U: Ulnar.

a b c

SpinalSpinal

Page 73: Brachial Plexus Injuries

Nerve repair and muscular transfers must beevaluated in a global plan. Preoperative diagno-sis is fundamental, and clinical evaluation andother investigations (CT-myelography and EMG)must be used to come to coherent therapeuticindications.

If surgery is indicated, the preoperative assess-ment must anticipate any pathological lesions thatmay be encountered in order to allow a specificsurgical plan. Early surgical exploration (from 6weeks to 3 months) will permit evaluation of theselesions and determination of the possibilities forneurolysis, nerve grafts or neurotization.

Results must be analyzed critically and anyregeneration should be credited to the nerverepair only if there is not other possible expla-nation for recovery. In any event, the results canbe evaluated only after sufficient time haselapsed because re-innervation after nerve graft-ing is always delayed. As mentioned earlier, thisrequires 2 to 3 years, depending on the type oflesion and its location (roots, trunks, etc) andmust be evaluated according to the function ofthe structures that have been repaired and theresulting therapeutic objectives achieved.

When evaluating motor function based on theinternational scale of 0 to 5, it is advisable to takeinto account the fatigability of the re-innervatedmuscles and the total function of the shoulder,elbow and arm. Return of sensory function isdifficult to evaluate, but recovery of function ofbridged structures confers considerable protec-tive sensitivity and definite improvement in thetrophic state of the extremity.

The pain syndrome must be treated. Themajority of our patients who underwent surgerydo not have pain, or have only some intermittentpain localized in the hand in the form of crampsor ‘electrical shocks’, which do not necessitateany medical treatment.

It is important to stress that surgical interven-tions with nerve repair for any given regionconsiderably modify the afferents originating inthe upper limb and that almost no patients havepain after surgery. This last observation is a veryimportant point that indicates the value of thesurgical treatment.

Finally, it is necessary to appreciate thepsychological aspect in young patients andprofessional consequences in the future.

The best results evidently occur in the upperroot partial palsies involving C5–C6 supraclavic-

ular lesions. However, in total root paralysis, theincreasing percentage of useful return offunction, depending upon the roots grafted andthe structures repaired, suggests that this type ofsurgery must be carried out with precise andearly indications.

References

Allieu Y, Chammas M, Picot MC (1997) Paralysie duplexus brachial par lesions supraclaviculaires chezl’adulte. Résultats comparatifs à long terme des greffesnerveuses et des trasferts nerveux, Rev Chir Orthop83:51–9.

Alnot JY (1993) La main plexique. Atteinte du poignetet de la main dans les paralysies traumatiques duplexus brachial de l’adulte. Presented at: Cahier desConferences d’Enseignement de la Société Françaisede Chirurgie de la Main (GEM), Expansion ScientifiqueFrançaise: Paris: 129–43.

Alnot JY (1995) Traumatic brachial plexus lesions inthe adult. Indications and results, Hand Clinics11(4):623–33.

Alnot JY, Narakas A (1996) Clinical examinations intraumatic brachial plexus injuries. ExpansionScientifique: Paris: 53–64.

Alnot JY, Oberlin C (1993) Tendon transfers in palsiesof flexion and extension of the elbow. In: Tubiana R,ed. The Hand, Vol IV. WB Saunders: Philadelphia:134–46.

Alnot JY, Narakas A et al (1989, 1995) Les paralysiesdu plexus brachial. Monographie du GEM, ExpansionScientifique, 1-er et 2-ème Ed. Expansion Scientifique:Paris: 1–297. [Traumatic brachial plexus injuries.Monographie du GEM, Expansion Scientifique, EnglishEdition, Paris, 1996: 1–279.]

Alnot JY, Daunois O, Oberlin C et al (1993) Total palsyof brachial plexus by supra-clavicular lesions, J OrthopSurg 7:58–66.

Alnot JY, Liverneaux PH, Silberman O (1996) Leslésions di nerf axillaire, Rev Chir Orthop 82(7):579–90.

Alnot JY, Rostoucher P, Oberlin C, Touam C (1998) Lesparalysies traumatiques C5–C6 et C5–C6–C7 du plexusbrachial de l’adulte par lésions supraclaviculaires, RevChir Orthop 84:113–23.

64 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 74: Brachial Plexus Injuries

Brunelli G, Monini L (1984) Neurotization of avulsedroots of brachial plexus by means of anterior nerves ofcervical plexus, Clin Plast Surg 11:144–53.

Franciosi LF, Modestti C, Mueller SF (1998) Neurotizationof the biceps muscle by end to side neurorraphybetween ulnar and musculocutaneous nerves. A seriesof five cases, Ann Chir Main 17(4):362–7.

Leechavenvongs S, Witoonchart K, Uepairojkit C (1998)Nerve transfer to biceps muscle using a part of theulnar nerve in brachial plexus injury (upper arm type):a report of 32 cases, J Hand Surg 23A:711–16.

Loy S, Bhatia A, Asfazadourian H, Oberlin C (1997)Ulnar nerve fascicle transfer on the biceps motor nervein C5–C6 or C5–C6–C7 avulsion of the brachial plexusbased on a series of 18 cases, Ann Hand Upper LimbSurg 16:275–84.

Millesi H (1984) Brachial plexus injuries. managementand results, Clin Plast Surg 11:115–21.

Narakas A (1978) Surgical treatment of traction injuriesof the brachial plexus, Clin Orthop 133:71–90.

Narakas A (1986) Les neurotisations ou transfertsnerveus dans le traitement des lesions traumatiques duplexus brachial. In: Tubiana R, ed. Traité de chirurgiede la main. Chirurgie des tendons, des nerfs et desvaisseaux, Vol 3. Masson: Paris: 542–68.

Oberlin C (1994) Nerve transfer to biceps muscle usinga part of ulnar nerve for C5–C6. Avulsion of the brachialplexus. Anatomical studies and report of four cases, JHand Surg 19A:232–7.

Rostoucher P, Alnot JY, Oberlin C, Touam C (1998)Tendon tranfers to restore elbow flexion aftertraumatic paralysis of the brachial plexus in adults, IntOrthop 22(4):255–63.

Rusch DS, Friedman A, Nunley JA (1995) The restora-tion of elbow flexion with intercostal nerve transfer,Clin Orthop 314:95–103.

Sedel L (1986) Resultats des reparations microchirurgi-cales du plexus brachial. A propos d’une serie de 170cas. In: Tubiana R, ed. Traité de chirurgie de la main.Chirurgie des tendons, des nerfs et des vaisseaux, Vol3. Masson: Paris: 568–71.

Sunderland S (1978) Nerves and Nerve Injuries, 2ndedn. Churchill Livingstone: Edinburgh and New York:854–900.

Terzis J (1987) Microreconstruction of Nerve Injuries.WB Saunders: Philadelphia.

Viterbo F, Trindade JCS, Hoschino K, Mazzoni Nito A(1994) End to side neurorraphy with removal of theepineural sheath. Experimental study in rat. PlastReconstr Surg 94:1038–47.

SUPRACLAVICULAR PLEXUS INJURIES 65

Page 75: Brachial Plexus Injuries
Page 76: Brachial Plexus Injuries

Introduction

Brachial plexus lesions result in complete palsyof the upper extremity in 50–64 per cent and anincomplete palsy in 50–36 per cent of cases(Songcharoen 1995, Birch et al 1998). Brachialplexus lesions are commonly classified as supra-clavicular or infraclavicular.

Supraclavicular lesions result from violentdistraction of the head away from the upperextremity. The roots are ruptured or avulsed fromthe spinal cord. The subclavian vessels are injuredin 15 per cent and the spinal cord is damaged in5 per cent of cases (Birch et al 1998).Infraclavicular lesions occur following shoulderhyperextension. With this mechanism of injury, anassociated axillary artery lesion is present in 30per cent of cases. In our experience, a combinedsupra- and infraclavicular lesion occurs in 15 percent of cases – the most frequent associationbeing an avulsion of C8T1 combined with arupture of the lateral and posterior cords orrupture of the axillary and musculocutaneousnerves. Over one half of patients presenting witha complete palsy may have single or multiplefractures (Songcharoen 1995, Terzis et al 1999).

In this chapter, we will describe the mostcommon reconstructions used for completebrachial plexus palsy following supraclavicularlesions, and their results. Our ability to comparefunctional results following brachial plexus recon-struction in complete palsy remains a challenge.The difficulty lies in the heterogeneity of eachseries, the lack of brachial plexus reconstructionstandardization, and because the evaluation offunctional results varies from one author to thenext. Surgical techniques such as the use ofcontralateral C7, rootlet reinsertion, and the role offree functioning muscle transfer will not be coveredin great depth, as they are discussed elsewhere.

Anatomy

In supraclavicular plexus injuries, the lesion islocated either above the spinal ganglion (pregan-glionic) or below it (postganglionic). Numerouscombinations of lesions exist (Narakas 1993).The most common combination is a rupture ofthe upper roots (C5C6 ± C7) and an avulsion ofthe lower roots (C8T1 ± C7) – this occurs in 48–54per cent of cases studied (Bonnard et al 1996,Bentolila et al 1999, Terzis et al 1999). Roots C5and C6 are less likely to be avulsed than thelower roots (C8T1) because of ligaments thatunite them to the osseous margin of the foram-inal outlet (Herzberg et al 1996).

The second most common combination is anavulsion of four roots with one ruptured root –this occurs in 23–27 per cent of cases (Bonnardet al 1996, Bentolila et al 1999, Terzis et al 1999).Typically, either root C5 or C6 is ruptured and theother roots are avulsed. Rarely, C5C6 and C8T1are avulsed and C7 is ruptured. Completeavulsion of C5 to T1 occurs in 13–25 per cent ofcases. In Birch’s series, complete avulsion wasmore frequent (Birch et al 1998). We have neverseen rupture of the lower trunk associated withan avulsion of the upper roots. If this patternexists, it must be extremely rare.

In preganglionic lesions, rootlets and rootsmay still be within or out of the foramen, andwith or without a tear of the dura. In the firstsituation, the peri-operative diagnosis is possiblewith intra-operative somatosensory evokedpotentials or with histological and histochemicalevaluation of the proximal stump during surgery(Terzis et al 1999). In postganglionic lesions thediagnosis of rupture is easier since the rootstumps may be dissected and their electricalstimulation results in contraction of the serratusand the deep neck muscles.

8Complete palsyChantal Bonnard and Dimitri J Anastakis

Page 77: Brachial Plexus Injuries

A double level lesion may exist in 15 per centof cases. Proximal avulsion and distal lesions areassociated with fractures of the clavicle (lateraland posterior cords), or of the scapula (supra-scapular nerve), with humeral fracture or gleno-humeral dislocation (axillary nerve), humerusfracture (musculocutaneous and radial nerve)and elbow dislocation (median and ulnar nerve).These peripheral nerves should be assessedintra-operatively. Double level injuries areassociated with a poor final outcome and maylimit proximal reconstruction. For example, it isof no value to repair the lower trunk when theulnar and median nerves have been stretched atthe elbow level. Up to 48 hours followingtrauma, a double level lesion is easier to detectbecause the injured root still responds to electri-cal stimulation intra-operatively. When stimula-tion of the avulsed spinal root fails to elicit acontraction of the corresponding muscle(s), itsuggests that a lesion exists between the avulsedroot and the target organ (Birch et al 1998).

Clinical finding

Complete paralysis of the serratus anterior andthe deep neck muscles suggest a C5C6C7avulsion. Horner’s syndrome suggests a C8T1avulsion. A complete palsy with sparing of thespinatii suggests an infraclavicular lesion(combined with C8T1 avulsion if Horner’ssyndrome is present). Sweat is present in theavulsed roots dermatomes. The skin is dry whenthe spinal roots are ruptured. Root avulsioncauses central pain. Patients describe a constantcrushing or burning pain in the anesthetic handwith episodes of excruciating shooting pain inthe whole limb. Pain typically appears within 2weeks post-trauma (Bonnard and Narakas 1985,Narakas 1992) and is common. Rarely, the painmay be associated with phantom-limb sensation.

Treatment

Nerve transfers/neurotizations

When a root is avulsed from the spinal cord, itsdistal stump must be reinnervated either by

another root (intraplexal neurotization) or by ahealthy nerve which resides outside the plexus(extraplexal neurotization).

Intraplexal neurotization: We abide by the‘proximity’ rule, which states that C6 should bereinnervated either with C5 or C7 but not withC8. This rule is valid in adults but not in children,where an avulsed C8T1 may be grafted to aruptured C6 (Birch et al 1998).

Extraplexal neurotization: The spinal accessorynerve, intercostal nerves, cervical plexus, phrenicnerve, hypoglossal nerve and the contralateralC7 root have been used as nerve transfers in thereconstruction of complete brachial plexus palsy.

Spinal accessory nerve

The spinal accessory nerve is commonly trans-ferred to the suprascapular or musculocutaneousnerves. The branches to the upper trapeziusmust be preserved to maintain scapular controland allow for future trapezius transfer. A C4injury is a contraindication to using the spinalaccessory nerve, due to the risk of completeshoulder instability post-operatively (Songcharoenet al 1996).

Intercostal nerves

The intercostal nerves can be harvested from T2to T6. T2 is commonly transferred directly toeither the lateral pectoral or the long thoracicnerve. Intercostal nerves T3 to T5 may be directlycoapted to the plexus at the upper arm level.Intercostal nerves from T6 and lower requirenerve grafts to reach the plexus. The intercostalnerves are commonly transferred to the motorportion or to the entire musculocutaneous nerve.They have also have been used to reinnervatethe radial, median, ulnar and axillary nerves, andfree functioning muscle transfer(s).

Intercostal nerves must not be used when thepatient suffers from Brown–Séquard syndromebecause the nerves are not functional. Brown–Séquard syndrome is present in 5 per cent ofcomplete brachial plexus palsy cases.

The harvesting of the intercostal nerves withan intact diaphragm does not change pulmonaryfunction (Allieu et al 1986, Giddins et al 1995).When there is a diaphragmatic paralysis, we donot use the intercostal nerves. We do not use acombined intercostal and phrenic nerve transfer.

68 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 78: Brachial Plexus Injuries

This last recommendation is contrary to theexperience of other surgeons who harvest thespinal accessory and phrenic nerves to reani-mate the shoulder and intercostal nerves toreanimate the elbow (Chuang 1999).

Three donor intercostal nerves provide similarresults to two intercostal nerves when trans-ferred to the musculocutaneous nerve (Naganoet al 1989). Recently, Berman et al (1996) havetransferred intercostal nerves to the lateral or themedial cord (depending of the location of thepain) for the sole purpose of pain relief.

Cervical plexus

Brunelli uses the cervical plexus to neurotize theaxillary and musculocutaneous nerves (Brunelli1980). Motor branches to the levator scapulae,rhomboids, upper trapezius, medial scalenus anddeep muscles of the neck may be transferred. Weuse only one or two motor branches to avoidcomplete denervation of the rhomboids andlevator scapulae muscles. Usually we transfermotor branches to the suprascapular nerve whenthe spinal accessory nerve is too small, or to thelong thoracic nerve to improve scapular mobil-ity. Sensory nerves of the cervical plexus havebeen used to reinnervate the median nerve. Theresults following this nerve transfer have beendisappointing.

Phrenic nerve

The phrenic nerve is commonly used as a nervetransfer by Asian surgeons (Gu et al 1987,Chuang et al 1995, Songcharoen et al 1996,Waikakul et al 1999a). Its use by Westernsurgeons is less common (Terzis et al 1999).Following phrenic nerve transfer, Gu found thatvital capacity dropped within the first post-opera-tive year but recovered to normal after 2 years(Gu 1989). We use the phrenic nerve only whenit is accompanied by an accessory phrenic nerve.

Hypoglossal nerve

Narakas first used this transfer for brachialplexus reconstruction in 1992. Results areencouraging when it is transferred to the muscu-locutaneous nerve. In our series, transferresulted in biceps motor strength of M3 or M4and a mean elbow flexion of 109° in five out of

eight cases. The results of hypoglossal nervetransfer to the median or ulnar nerve wereuniformly disappointing. This is most likely dueto the long distance between the proximal stumpand the distal end organ. Harvesting of thehypoglossal nerve is associated with no signifi-cant complications. Tongue hemiatrophy waswell tolerated in this group.

Contralateral C7 root

All or part of the contralateral C7 root may beused without severe sequelae. An entire chapteris devoted to the use of contralateral C7. Wehave not used this donor nerve to date.

Nerve root reimplantation

Since the early nineteen-nineties, Carlstedt hasshown that avulsed spinal roots can bereimplanted in the spinal cord under limitedconditions. A separate chapter is devoted to thisnew technique.

Reconstructive goals

In complete palsy, the reconstructive goals are: (1)reconstruction of brachiothoracic pinch; (2) recon-struction of elbow flexion; and (3) reconstructionof a basic hand function (i.e. sensation in thethumb and index finger, wrist extension, andfinger flexion). Plexus reconstruction may also beindicated to alleviate or improve pain (Berman etal 1996). These reconstructive goals are wellaccepted amongst most brachial plexus surgeons.However, a few surgeons prefer to restore elbowflexion first followed by wrist extension and fingerflexion, the last priority being the shoulder, whichmay be arthrodesed (Bentolila et al 1999).

Lesions, reconstruction andresults

Total avulsion C5–T1

When there are no associated orthopaedic orvascular lesions, we transfer the spinal accessory

COMPLETE PALSY 69

Page 79: Brachial Plexus Injuries

nerve to the suprascapular nerve and an inter-costal nerve to the long thoracic nerve to effectopening of the brachiothoracic pinch. When therhomboid muscles are functional, the serratusanterior needs to be reinnervated using inter-costal nerve transfers to avoid overaction of therhomboids. The T3 intercostal nerve is trans-ferred to the lateral pectoral nerve to improvebrachiothoracic pinch closure. Finally, intercostalnerves T4 and T5 are transferred to the muscu-locutaneous nerve to restore elbow flexion andforearm sensation. Intercostal nerves T6 and T7are transferred to the median nerve (Fig. 1).

When the scapula is fractured through the notchor when exploration confirms suprascapularnerve avulsion, we neurotize the long thoracicnerve to improve scapulothoracic joint function.Later, in a second stage, we perform gleno-humeral arthrodesis. The musculocutaneousnerve is neurotized either with the spinal acces-sory nerve (only when C4 is intact) or with inter-costal nerves. When neither transfer is available,we use the hypoglossal nerve. In our experience,these various transfers provide useful brachiotho-racic pinch, elbow flexion and minimal sensoryrecovery of the thumb and index finger. Handmotor function has never been useful (Table 1).

Alnot (1995) and Bentolila et al (1999) recom-mended transfer of the spinal accessory nerve to

the anterolateral portion of the musculocuta-neous nerve. The remainder of the musculocuta-neous nerve is covered by sensory branches ofthe cervical plexus (Alnot 1995) or by intercostalnerves (Bentolila et al 1999). The authorsdescribed biceps motor strength greater than M3in 75 per cent of their cases (Alnot 1995, Bentolilaet al 1999).

In 1999, Chuang recommended four neurotiza-tions in reconstruction of complete avulsion: thespinal accessory nerve transferred to the supra-scapular nerve, the phrenic nerve to the contri-bution of the upper trunk to the posterior cord,the intercostal nerves to the musculocutaneousnerve, and the contralateral C7 to the mediannerve. When the phrenic nerve is not available,Chuang uses the hypoglossal nerve (Chuang1999).

In 1999, Waikakul (1999b) presented a prospec-tive study of 205 patients with complete palsy.He compared elbow flexion recovery followingtransfer of the spinal accessory nerve or inter-costal nerves to the musculocutaneous nerve.The strength of elbow flexion was greatestfollowing spinal accessory nerve transfer.Superior sensory recovery and improved paincontrol were noted following intercostal inter-costal nerve transfers.

In another study by Waikakul (1999a), 96patients with complete root avulsion werereviewed. Waikakul transferred the phrenic nerveto the suprascapular nerve, the spinal accessorynerve to the musculocutaneous nerve and theanterior portion of the contralateral C7 root tothe median nerve, obtaining shoulder abduction(> 60°) in 85 per cent, elbow flexion (≥ M3) in 88per cent, forearm pronation (≥ M3) in 33 per cent,wrist flexion (≥ M3) in 29 per cent and fingerflexion (≥ M3) in 21 per cent of cases.

The importance of the patient’s age (Naganoet al 1989, Waikakul 1999a) and the intervalbetween trauma and plexus reconstruction(Nagano et al 1989, Akasaka et al 1990, Krakauerand Wood 1994, Chuang et al 1996, Songcharoenet al 1996, Bentolila et al 1999, Terzis et al 1999)have been well-studied recently. These twoimportant variables have had an impact on themethods of brachial plexus reconstruction used.

In 1989, Nagano described a series of 144 adultpatients undergoing neurotization of the muscu-locutaneous nerve with the intercostal nerves.Failure rates increased as the interval between

70 THE ADULT TRAUMATIC BRACHIAL PLEXUS

XI nerve SS nerve

Mot T2 long thoracic nerve

Mot T3 pectoral nerve

Mot T4 and T5 musculocutaneous nerve

Sens T4 to T6 median nerve

Figure 1

Plexus reconstruction following C5T1 avulsion.

Table 1 Results after reconstruction following C5–T1avulsion

No. of ≤ M2 M3 ≥ M3+patients

Thoracobrachial gripclosure 28 7 8 13opening 28 5 11 12Elbow flexion 32 9 6 17

Page 80: Brachial Plexus Injuries

trauma and plexus repair increased. Based onthis work, Nagano recommends that intercostalnerve transfers must be performed in patientsless than 40 years of age and within 6 monthspost-trauma.

Akasaka et al (1990) noted that intercostalnerve transfers resulted in poor results after adelay of more than 6 months post-trauma. Forlate cases they recommend using a free function-ing muscle transfer to restore elbow flexion.Krakauer and Wood (1994) suggest that 9months post-trauma should be the cut-off timefor intercostal nerve transfer to the musculocu-taneus nerve. After 9 months, they perform afree functioning muscle transfer innervated bythe intercostal nerves.

Songcharoen et al (1996) presented a series of216 patients – 158 cases with a total palsy and58 cases with a incomplete palsy. All patientshad an avulsion of C5C6 that was reconstructedwith a transfer of the spinal accessory nerve tothe musculocutaneous nerve. He studied theeffects of patient age and the interval betweentrauma and surgery. These factors were statisti-cally significant. Allieu again confirms that theresults of brachial plexus reconstruction are poorin patients over the age of 40, and notes that thebest results occur when reconstruction isperformed before 8 months post-trauma (Allieuet al 1997).

Chuang alters his reconstruction methodbased on the time post-trauma. If the patient isless than 6 months post-trauma, he transfersthree intercostal nerves to the musculocutaneousnerve. If the patient is between 6 and 12 monthspost-trauma, he transfers two intercostal nervesto the musculocutaneous nerve and uses twointercostal nerves to innervate a free functioningmuscle transfer. Finally, if the patient is greaterthan 12 months post-trauma, he does not neuro-tize the musculocutaneous nerve but rather uses

the three intercostal nerves to innervate a freefunctioning muscle transfer. The author prefersto innervate a free functioning muscle transferwith the intercostal nerves rather than the spinalaccessory nerve because direct nerve coaptationis possible (Chuang et al 1993, 1996). In 1999Terzis demonstrated that, in the restoration offinger flexion, a denervation time of less than 6months resulted in significantly better outcomesthan a denervation time of 6–12 months.

Doi describes a series of 26 patients withcomplete brachial plexus injury in whom upperlimb function was reconstructed using two freefunctioning muscle transfers (Doi et al 1995,2000). The reader is referred to the chapter onfree functioning muscle transfers for furtherinformation.

Rupture C5 and avulsionC6C7C8T1

Reconstruction depends on the quality of the C5root. When the stump of C5 is small or when itsquality is doubtful, we graft it to the posteriordivision of the upper trunk. This graft is done toobtain shoulder function. The rest of the recon-struction is the same as that in total avulsion.When the stump of C5 is of good quality, wegraft it with a vascularized ulnar graft to thelateral cord. The intercostal nerve T2 is trans-ferred to the lateral pectoral nerve and the inter-costal nerve T3 to the C7 contribution of the longthoracic nerve (Fig. 2).

To avoid co-contraction, Alnot grafts C5 to theanterior division of the upper trunk (lateral cord)only (Alnot and Narakas 1996) On the otherhand, Nagano usually grafts C5 to the posteriordivision of the upper trunk and transfers theintercostal nerve to the musculocutaneous nerve

COMPLETE PALSY 71

C5 C5C6 (post) C5 lateral cord

XI SS nerve (vascularized ulnar nerve graft)

T2 long thoracic XI SS nerve

T3 pectoral T2 long thoracic

Mot T4T5 MC T3 pectoral

Sens T4T6 median

a C5 is small b C5 is large Figure 2

Plexus reconstruction following C5rupture and C6T1 avulsion. (a) C5is small. (b) C5 is large.

Page 81: Brachial Plexus Injuries

(Nagano et al 1989). The spinal accessory nerveis transferred to the suprascapular nerve. Othersurgeons (Allieu, Merle; see Bonnard et al 1996)perform selective grafts from C5 to the muscu-locutaneous nerve, combining it with a graftfrom C5 to the radial nerve and even to themedian nerve if C5 is large enough (Terzis et al1999). When C5 is small, Terzis grafts it to themusculocutaneous nerve and neurotizes theaxillary nerve and the branch to the triceps withintercostal nerves.

Results following these classic reconstructionsare nearly the same as in total avulsion (Table2). In Bentolila’s study, results were better whengrafts from C5 were coapted distally to the lateralor posterior cord than when they were coaptedto the musculocutaneous or the radial nerves(Bentolila et al 1999). This finding was not statis-tically significant. The authors argued that whenthe length of graft is short there is more nervetissue available for grafting, which is a significantfactor in distal recovery.

Increasingly, surgeons are now combiningthese common reconstructions with a contralat-eral C7 transfer to the median nerve using anulnar vascularized nerve graft (Chuang 1999,Terzis et al 1999, Waikakul et al 1999a).

C5C6 rupture and C7C8T1 avulsion

We graft C5 and C6 to C5C6C7, respecting theanterior and posterior distribution of the fasci-cles, or C5 and C6 to the lateral and posteriorcord (Alnot et al 1992). The intercostal nerves aretransferred to the ulnar nerve, giving a positivetrophic effect and some wrist or finger flexion.The suprascapular nerve is reinnervated with thespinal accessory nerve (Fig. 3; Table 3). Severalauthors (Allieu, Sedel) prefer to leave the spinal

accessory nerve intact and use C5 to reinnervatethe suprascapular nerve (Bonnard et al 1996,Bentolila et al 1999). Terzis et al (1999) hasrecently shown that there is no statisticallysignificant difference following suprascapularnerve reconstruction with intraplexal neurotiza-tion or with spinal accessory nerve. Allieu et al(1997) studied elbow flexion recovery followinggrafting from a ruptured root or nerve transferwith the spinal accessory or intercostal nerves,and noted that the results were better followingnerve grafting (from a proximal stump of C6)(Allieu et al 1997).

Chuang prefers to transfer the intercostalnerves to the musculocutaneous nerve and usesC5 for the shoulder (i.e. suprascapular nerve andposterior part of the upper trunk), with or withoutthe phrenic nerve. He uses C6 for the hand (i.e.C8 or the median nerve) (Chuang et al 1995,Chuang 1999).

Terzis et al (1999) grafts C5 and C6 to themusculocutaneous nerve, radial and mediannerves distally using a vascularized ulnar graft.She uses intercostal nerves for transfer to thetriceps, the axillary nerve, the nerves to theserratus and to the latissimus dorsi. Later, sheperforms a free muscle transfer for the handreinnervated by a contralateral C7.

72 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 2 Results after reconstruction following C5rupture + C6T1 avulsion

No. of ≤ M2 M3 ≥ M3+patients

Brachiothoracic gripclosure 19 7 6 6opening 19 8 3 8Elbow flexion 26 7 7 12

Table 3 Results after reconstruction following C5C6rupture + C7T1 avulsion

No. of ≤ M2 M3 ≥ M3+patients

Brachiothoracic gripclosure 26 6 5 15opening 26 8 10 8Elbow flexion 28 12 6 10Wrist flexion 26 19 6 1Finger flexion 26 25 1

XI SS nerve

C5C6 C5C6C7

T3T4 ulnar nerve

Figure 3

Plexus reconstruction following C5C6 rupture and C7T1.

Page 82: Brachial Plexus Injuries

C5C6C7 rupture and C8T1 avulsion

Our preferred reconstruction includes graftingC5C6C7 to the upper and middle trunks and toC8 (for active finger extension), and transferringthe spinal accessory nerve to the suprascapularnerve (Fig. 4).

Chuang (1999) grafts C5 to the suprascapularnerve and the posterior part of the upper trunk,C6 to the median nerve, C7 to C7 itself, andtransfers three intercostal nerves to the muscu-locutaneous nerve.

The main difficulty is the availability of donornerve tissue for grafting – two sural nerve graftsare insufficient. The sensory branch of the radialnerve may be harvested, as well as a vascular-ized ulnar nerve nerve graft. Clavicle osteotomymay shorten the length of grafts needed(Bentolila et al 1999).

The results in this group are better than in thegroup presenting with three or more avulsions(Tables 4 and 5). This is probably due to thebetter quality of the proximal ruptured spinalroots C5C6 and C7. In Bentolila’s series (Bentolilaet al 1999), results were equivalent when five,four or three roots were avulsed (biceps ≥ M3+in 55.3 per cent of cases). Results are muchbetter (biceps ≥ M3+ in 80 per cent of cases)when only two roots are avulsed.

C5T1 rupture

All surgeons attempt an anatomical reconstruc-tion. The total amount of available nerve graft islimited; two sural nerves are usually not enough,

and sometimes the ulnar nerve must beharvested and used as a vascularized nerve graft(Table 6). Reconstruction of the lower trunk isworthwhile when the patient is young (less than20 years of age), when the interval betweentrauma and surgery does not exceed 2 months,and when the nerve graft is short.

Conclusions

Based on our experience and the reportedexperience of others, we reach the followingconclusions:

COMPLETE PALSY 73

XI SS nerve

C5C6 upper trunk

C7 middle trunk

C7 C8 (to post cord)

Figure 4

Plexus reconstruction following C5C7 rupture and C8T1avulsion.

Table 6 Results after reconstruction following C5T1rupture

No. of ≤ M2 M3 ≥ M3+patients

Brachiothoracic gripclosure 9 0 2 7opening 9 2 2 5Elbow flexion 9 1 3 5Wrist flexion 9 5 1 3Finger flexion (hand 9 7 1 1is useful as a hook)

Table 5 Results after reconstruction following C5C6C7rupture + C8T1 avulsion

No. of ≤ M2 M3 ≥ M3+patients

Brachiothoracic gripclosure 20 3 2 15opening 20 8 7 5Elbow flexion 20 4 2 14Wrist flexion 20 14 5 1Finger flexion (hand 20 18 2is useful as a hook)

Table 4 Biceps recovery according to the number ofavulsions

No. of Biceps ≥ M3+ % Biceps ≤ M2 %avulsions

5 17/32 patients 53 9/32 patients 284 12/26 patients 46 7/26 patients 273 10/28 patients 36 12/28 patients 43Total 39/86 patients 45 28/86 patients 332 14/20 patients 70 4/20 patients 20

Page 83: Brachial Plexus Injuries

1. Evaluating the quality of avulsed roots is veryimportant. In our experience, we overestimatethe quality of the roots, which may explainour poor results following C5 and C6 lesionsassociated with C7 to T1 avulsions.

2. Patient age and the interval between traumaand plexus repair are statistically significantvariables, and the type of reconstruction mustbe adapted accordingly. When the intervalbetween trauma and repair is greater than 9months, a free functioning muscle transfer isrecommended for biceps reconstruction.

3. The use of the phrenic nerve or the contralat-eral C7 root remains a controversial subject,despite studies demonstrating low post-operative morbidity.

4. Regarding the shoulder; we prefer to avoidglenohumeral arthrodesis. In extensive lesions(four or five root avulsions), reconstruction ofthe suprascapular, long thoracic and pectoralnerves improves scapular control andprovides the patient with brachiothoracicpinch. When C4 is avulsed, nerve surgery toreanimate the shoulder should be abandoned.A long thoracic nerve reconstruction is notrequired because of rhomboid paralysis. Thespinal accessory nerve should be preserved toavoid complete scapular instability.

5. Regarding the elbow, following reconstruc-tion of a complete plexus palsy, biceps motorstrength recovery was greater than M3+ inover 50 per cent of cases.

6. Regarding the hand, neurotization of thelower trunk using extraplexal transfers doesnot work. Neurotization of the ulnar or themedian nerves with intercostal nerves mayresult in useful wrist function and a protectivesensation. Functional results following freefunctioning muscle transfers are promising.

7. Plexus reconstruction alleviates central paincaused by root avulsion (Alnot et al 1992,Berman et al 1996, Bonnard et al 1996) and,if only for this purpose, plexus repair is worth-while.

References

Akasaka Y, Hara T, Takahashi M (1990) Restoration ofelbow flexion and wrist extension in brachial plexusparalysis by means of free muscle transplantation inner-vated by intercostal nerves, Ann Hand Surg 9:341–50.

Allieu Y, Clauzel AM, Mekhaldi A, Triki F (1986) Theeffects on respiratory function of traumatic plexusinjuries and some surgical treatment, Rev ChirurgOrthop 72:455–60.

Allieu Y, Chammas M, Picot MC (1997) Long termcomparative results between nerve grafts and nervetransfers for the treatment of supraclavicular brachialplexus injuries in adults, Rev Chirurg Orthop 83:51–9.

Alnot JY (1995) Traumatic brachial plexus lesions inthe adult, Hand Clin 11:623–31.

Alnot JY, Narakas A (1996) Traumatic brachial plexusinjuries. Expansion Scientifique Française: Paris.

Alnot JY, Daunois O, Oberlin Ch, Bleton R (1992) Totalpalsy of brachial plexus by supraclavicular lesions, RevChirurg Orthop 78:495–504.

Bentolila V, Nizard R, Bizot P, Sedel L (1999) Completetraumatic brachial plexus palsy, J Bone Joint Surg81A:20–8.

Berman J, Anand P, Chen L et al (1996) Pain relief frompreganglionic injury to the brachial plexus by late inter-costal transfer, J Bone Joint Surg 78B:759–60.

Birch R, Bonney G, Wynn Parry CB (1998) SurgicalDisorders of the Peripheral Nerves. ChurchillLivingstone: Edinburgh.

Bonnard C, Narakas A (1985) Syndromes douloureuxet lésions posttraumatiques du plexus brachial, HelvChir Acta 52:621–32.

Bonnard C, Allieu Y, Alnot JY et al (1996) Completepalsies through supraclavicular lesions. In: Alnot JY,Narakas A, eds. Traumatic Brachial Plexus Injuries.Expansion Scientifique Française: Paris: 126–55.

Brunelli G (1980) Neurotization of avulsed roots of thebrachial plexus by means of anterior nerves of thecervical plexus, Int J Microsurg 2:55–8.

Chuang D (1999) Management of traumatic brachialplexus injuries in adults, Hand Clin 15:737–55.

Chuang D, Epstein M, Yeh MC, Wei FC (1993) Functionalrestoration of elbow flexion in brachial plexus injuries:results in 167 patients (excluding obstetric brachialplexus injury), J Hand Surg 18A:285–91.

Chuang D, Lee G, Hashem F, Wei FC (1995) Restorationof shoulder abduction by nerve transfer in avulsedbrachial plexus injury: evaluation of 99 patients withvarious nerve transfers, Plast Reconstr Surg 96:122–8.

74 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 84: Brachial Plexus Injuries

Chuang D, Carver N, Wei FC (1996) Results of function-ing free muscle transplantation for elbow flexion, JHand Surg 21A:1071–7.

Doi K, Sakai K, Kuwata N et al (1995) Double free-muscle transfer to restore prehension followingcomplete brachial plexus avulsion, J Hand Surg20A:408–14.

Doi K, Muramatsu K, Hattori Y et al (2000) Restorationof prehension with the double free muscle techniquefollowing complete avulsion of the brachial plexus, JBone Joint Surg 82A:652–66.

Giddins GEB, Kakkar N, Alltree J, Birch R (1995) Theeffect of unilateral intercostal nerve transfer upon lungfunction, J Hand Surg 20B:675–6.

Gu YD (1989) Phrenic nerve transfer for brachial plexusmotor neurotization, Microsurg 10:287–9.

Gu YD, Wu MM, Zhen YL et al (1987) Microsurgicaltreatment for root avulsion of the brachial plexus,Chinese Med J 100:519–22.

Herzberg G, Narakas A, Comtet JJ (1996) Surgicalapproach of the brachial plexus roots. In: Alnot JY,Narakas A, eds. Traumatic Brachial Plexus Injuries.Expansion Scientifique Française: Paris:19–22.

Krakauer JD, Wood MB (1994) Intercostal nervetransfer for brachial plexopathy, J Hand Surg19A:829–35.

Nagano A, Tsuyama N, Ochiai N et al (1989) Directnerve crossing with the intercostal nerve to treatavulsion injuries of the brachial plexus, J Hand Surg14A: 980–5.

Narakas A (1992) Les syndromes douloureux dans lesarrachements du plexus brachial, Doul Anal 3:83–101.

Narakas A (1993) Lesions found when operatingtraction injuries of the brachial plexus, Clinical NeurolNeurosurg 95:56–64.

Songcharoen P (1995) Brachial plexus injury inThailand: a report of 520 cases, Microsurg 16:35–9.

Songcharoen P, Mahaisavariya B, Chotigavanich C(1996) Spinal accessory neurotization for restoration ofelbow flexion in avulsion injuries of the brachialplexus, J Hand Surg 21A:387–90.

Terzis J, Vekris M, Soucacos P (1999) Outcomes ofbrachial plexus reconstruction in 204 patients withdevastating paralysis, Plast Reconstr Surg 104:1221–40.

Waikakul S, Orapin S, Vanadurongwan V (1999a)Clinical results of contralateral C7 root neurotization tothe median nerve in brachial plexus injuries with totalroot avulsions, J Hand Surg 24B:556–60.

Waikakul S, Wongtragul S, Vanadurongwan V (1999b)Restoration of elbow flexion in brachial plexus avulsioninjury: comparing spinal accessory nerve transfer withintercostal nerve transfer, J Hand Surg 24A:571–7.

COMPLETE PALSY 75

Page 85: Brachial Plexus Injuries
Page 86: Brachial Plexus Injuries

Introduction

The brachial plexus is formed by the spinalnerves C5, C6, C7, C8 and T1, and it supplies thewhole upper extremity except for a small strip ofskin on the inner side of the upper arm whichreceives sensory fibres from T2.

At one extreme, a complete lesion of thebrachial plexus leads to a flail arm which is notonly useless but represents an obstacle: due to thebody asymmetry a scoliosis with consequent paindevelops, and the patient not only has no functionin the arm but also has additional problems.

The other extreme is a partial lesion, involvingonly one root. In this case the loss of function isminimal, because the vast majority of the musclesreceive nerve fibres from at least two roots. Thebest example is an isolated lesion of C7, whichunder normal anatomical fibre distributionproduces a weakness of the triceps and the exten-sor muscles of the forearm and a reduced sensi-bility of the thumb and the index finger, but allthese functional losses are temporary.

If, however, two or three neighbouring rootsare involved, a characteristic paralysis patterndevelops:

• A lesion of C5 and C6 is followed by reducedfunction of the shoulder muscles and lack ofelbow flexion;

• A lesion of C5, C6, and C7 produces, inaddition, a loss of wrist and finger extension;

• A lesion of C8 and T1 produces a loss of handfunction, with the shoulder and elbow function-ing normally.

The normally functioning uninvolved muscleswith paralysed antagonists become too strongand cause contractures. In lower brachial plexus

lesion (C8 and T1) a supination contracture maydevelop by the supinating action of theunopposed biceps muscle, and in the shoulderjoints much stronger inner rotators produce aninternal rotation contracture.

In such cases, muscle transfers can be plannedin order to restore the lost function. However,these partial brachial plexus lesions are differentfrom those lesions that originally produced acomplete loss of function but have partiallyrecovered. In the latter case some musclesremain permanently paralysed and othersrecover, but the recovered muscles are not asvalid for muscle transposition as those in theformer.

There are variations in the fibre distributionthat must be considered. The brachial plexusmay receive more fibres from C4 and less fromT1 (prefixation), or more from T2 and less fromC5 (postfixation).

The vast majority of brachial plexus lesions arecaused by motorcycle accidents, followed byother traffic accidents, sports injuries, workinjuries, and gunshot injuries.

The rootlets may be avulsed, i.e. no proximalstump is available. The spinal nerve, the trunksand the divisions of the cords may be inter-rupted, which means that a proximal stump isavailable or there might be a lesion in continu-ity. A lesion in continuity may be very severewith damage of the fascicular pattern (grade IVaccording to Sunderland 1951), or less severe(grades I, II or III). A grade IV lesion does notoffer the possibility of spontaneous recovery,and has to be treated like a complete loss ofcontinuity. The involved segment must beresected and the defect bridged by nerve grafts.

A grade I lesion is usually followed by acomplete spontaneous recovery. The same is

9Update on the treatment of adultbrachial plexus injuriesHanno Millesi

Page 87: Brachial Plexus Injuries

true for a grade II lesion, but this takes moretime. The spontaneous recovery can be impededor prevented by compression of the nerve tissueby fibrosis. A grade III lesion may show sponta-neous recovery, but this is always incomplete. Ingrade III lesions in particular, the chances ofspontaneous recovery may be impeded by exter-nal or internal compression. These grades maybe treated by neurolysis, and good functionalresults can often be achieved.

With increasing severity of the lesion thechances of a good result after surgerydecrease and they are, of course, the worst incases of avulsion of all five roots. In thesemost severe cases a high degree of recoverycannot be expected, and the patient may beunable to return to his original job. Even iffunction can be improved significantly, it willalways be impaired and the patient may needto be trained for a job that requires only onehand. However, this incomplete recovery isextremely important for daily living, and there-fore reconstructive efforts are justified even ifthe patient cannot resume his former occupa-tion.

A significant problem for many patients withbrachial plexus lesions is pain. The pain may beneuromal pain in cases with loss of continuity,pain produced by irritation of nerve fibres inlesions with preserved continuity, or pain causedat higher levels in cases of avulsion. The first twotypes of pain can be tackled by surgery, but thereal problem is central pain. Pain syndromesdevelop independently of attempts at surgeryand are also present in unoperated cases.However, there are theories that patients whohave undergone surgery develop painsyndromes to a lesser degree and less frequently(Narakas 1986).

Before the 1960s, successes in brachial plexussurgery were achieved in cases of clean transec-tions without defects, and in cases of preservedcontinuity after external neurolysis. Attempts toachieve restoration of continuity by resection ofsegments of the humerus failed, and the generalstrategy was to perform an amputation of theupper arm in cases of complete brachial plexuslesions with root avulsions.

The basis for current surgery of the brachialplexus was the introduction of new successfultechniques of nerve grafting (Millesi 1968, Millesiet al 1972, 1976).

Timing of surgery

If there is an open injury, primary surgical treat-ment is indicated. In a clean transection theinvolved segments of the brachial plexus arerepaired by neurorrhaphy. Fracture of the claviclehas to be osteosynthesized and the vascular lesiondealt with accordingly. The question ariseswhether the emergency vascular repair should becombined with elective surgery of the brachialplexus. Such a procedure should be consideredonly if an experienced surgeon is available. Aninexperienced surgeon does more harm thangood, and frequently even the diagnosis is unreli-able. If available the surgeon must considerwhether the small advantage of gaining some timeoutweighs the increased risk for the patient due tothe extended (10 or more hours) operation in theemergency situation. Experience suggests that inthis situation surgeons tend to reduce the time ofsurgery by performing as yet unproven proce-dures, e.g. end-to-side coaptations between trunks,and do not fully exploit the available possibilities.

Bonney (1983) preferred early surgery a fewdays after the accident, even in closed injuries,for two reasons:

1. The anatomical situation can be easily clari-fied because a few days after the accidentthere is no fibrosis or scar tissue formation;

2. Early primary surgery may save time.

However, the first argument is based on an errorin basic thinking. The fibrosis or scar tissue forma-tion is caused by the trauma. In cases of earlysurgery, it will develop after surgery and thesurgeon accepts this risk in order to deal withnerve stumps that can be easily seen but willbecome fibrotic in 2–3 weeks time. With earlysurgery the surgeon cannot allow for this. Inaddition, the eventual damage to neurons isdependent on the trauma and will happen anyway.There is good evidence that early secondary repairoffers a better chance for regeneration. Theneurons will be in an active phase, and thesecondary trauma creates greater axon sproutingthan the initial repair. It must also be rememberedthat with present diagnostic tools it is not possibleto establish an exact diagnosis immediately, andthere is therefore the possibility that a patient withthe chance of spontaneous recovery will beoperated on.

78 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 88: Brachial Plexus Injuries

Based on these considerations, the followingconclusions were reached regarding indicationsand timing:

1. If there is a clean transection, a primary repairwith neurorrhaphy is indicated;

2. If there is an open injury with blunt trauma tothe nerves and complicating factors such asvascular injury or fracture, it is advisable todeal primarily with the other injuries and toperform reconstruction of the brachial plexusas a secondary procedure;

3. If there is a closed injury in a brachial plexuslesion, an elective reconstruction of thebrachial plexus may be planned after clarify-ing the diagnosis and excluding the possibil-ity of spontaneous recovery.

Diagnosis

The diagnosis of brachial plexus lesion is estab-lished clinically, and the following differentiationcan be made:

• Complete brachial plexus lesion;• Upper brachial plexus lesion C5, C6;• Upper brachial plexus lesion C5, C6, C7;• Lower brachial plexus lesion C8 and T1;• Irregular brachial plexus lesion.

Information regarding the presence and absenceof root avulsions is desirable. In the early yearsthis was very important, because in cases ofavulsion of all five roots no surgery wasperformed. However, we have since learned todeal with these worst scenarios, and knowledgeof root avulsions does not play any role in estab-lishing the indication for operation.

The clinical examination provides importantinformation about the severity of the lesion. Ifthe serratus anterior muscle is not innervatedit can be assumed that an avulsion of the rootsC5, C6, and C7 has occurred, because theorigin of the long thoracic nerve is locatedvery proximal at these spinal nerves. If theserratus anterior is functioning, generally atleast one of these three roots is not avulsed;however, there is the rare situation when C4contributes to the innervation of the serratusanterior muscle.

The innervation of the deep muscles of theneck comes from the dorsal branch of the spinalnerves, and these muscles do not work in anextensive avulsion. Evaluation becomes moredifficult if there is a mixed situation, with someroots avulsed and others not.

The suprascapular nerve usually leaves thesuperior trunk at its distal end. If this nerve isconducting, the lesion should be distal to thislevel.

Two or three weeks after the trauma aTinel’s sign may become positive at the levelwhere axon sprouts are formed. In case ofavulsion a Tinel’s sign does not occur. In thissituation a positive Tinel’s sign will be presentonly if an element of the cervical plexus isinvolved. This can be differentiated by thediffering extent of the Tinel’s sign, which will,in case of involvement of the cervical plexus,spread towards the auricle or the anteriorneck.

If a present Tinel’s sign starts to advancetowards the periphery, the beginning of theneurotization of originally denervated distalsegments is indicated. Advancement andprogress of the Tinel’s sign does not,however, allow any prognosis as far as thequality of the regenerative process isconcerned, because it may depend on a few‘pioneer’ nerve fibres. It is the lack of occur-rence or progression of the Tinel’s sign thatallows a conclusion.

Ten days or so after the accident Walleriandegeneration has to be far enough advanced thatconductivity in the distal segments ceases. Thisfact – no conductivity in distal segments of motorfibres – allows exclusion of a grade I lesion ofthe nerve fibres.

The conductivity of sensory nerve fibres is lostif there is a lesion distal to the spinal ganglion(infraganglionic lesion). It remains positive incase of a pre-ganglionic lesion because thesensory nerve fibres are still in connection withthe neurons and do not develop Walleriandegeneration.

By clinical examination and simple electro-physiologic studies 3 weeks after the accident,the following conclusions can be reached:

• If motor conductivity remains positive, thereis grade I damage (neurapraxia). Spontaneousrecovery can be expected;

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 79

Page 89: Brachial Plexus Injuries

• If there is no Tinel’s sign, avulsion of all fiveroots may be suspected;

• If the Tinel’s sign is positive with radiationinto the arm and hand, at least one root is notavulsed.

During the following weeks further observationsallow the following conclusions:

• If the Tinel’s sign remains consistentlynegative, avulsion of all five roots may besuspected. This can be confirmed by CT-myelography or MRI. Surgery is stronglyindicated;

• If the Tinel’s sign becomes consistentlypositive but does not move, a grade IV or Vinfraganglionic lesion may be suspected.Surgery is indicated;

• If the Tinel’s sign becomes strongly positiveand moves in a distal direction, a grade II orIII infraganglionic lesion may be suspected.For the moment surgery is not indicated,because there is the chance of spontaneousrecovery;

• If further observation over the fourth and fifthmonths reveals satisfactory recovery, a gradeII lesion (axonotmesis) can be diagnosed. Nosurgery is indicated;

• If there is an initial recovery of the proximalmuscles, a grade III lesion is diagnosed.Surgery is indicated if recovery does notprogress;

• If the Tinel’s sign progresses without recov-ery, there is a grade III lesion with externalor internal compression. Surgery isindicated.

Summary

Surgery is indicated in all patients who do notshow full recovery within 6 months. Patients withroot avulsions and loss of continuity should beoperated upon as soon as the diagnosis is wellestablished and a lesion with preserved continu-ity can be excluded.

MRI gives a satisfactory indication that a rootavulsion is possible, but not safe enough onwhich to base a decision for the surgicalplanning. There is good reason to believe thatMRI techniques will improve and become 100

per cent reliable, and may also provide informa-tion about the status of a damaged nerve trunkwith preserved continuity.

Due to the complexity of lesions and the differ-ent possibilities of combinations, a classificationsystem is required.

Surgical options for treatment ofpatients with brachial plexuslesions

A variety of possibilities exists to improvefunction of patients with brachial plexus lesions.In partial brachial plexus lesions muscle trans-position may achieve this goal. The conductivityof non-conducting nerve segments withpreserved continuity may be restored by neurol-ysis. Lost continuity of original pathways isrestored by neurorrhaphy or nerve grafting, andnerve fibres may be transferred from one nerveto another if the original pathway cannot bereconstructed. New muscle tissue can be used ifthe original muscles are atrophied and beyondrepair. This section describes various surgicaloptions and demonstrates the enormous rangeof surgical possibilities.

Neurolysis

Lesions in continuity with external compressioncan be treated successfully by neurolysis. If thereis fibrosis of the paraneurium and the epifascic-ular or interfascicular epineurium, a para- orepineuriotomy, or an epi- or interfascicularepineuriectomy respectively can be performed.These procedures only have a good chance if thefascicular pattern and the endoneurial tissue ispreserved. If there is an interfascicular fibrosis ora loss of the fascicular pattern, neurolysis willnot be successful. In this situation, resection ofthe involved segment with restoration of conti-nuity by nerve grafts is indicated. Occasionally itis very difficult to make this differentiation, andif neurolysis ends in failure then an incorrectdecision was made. For the author, it is a rulethat, in case of doubt, it is better to resect andgraft than to attempt neurolysis.

80 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 90: Brachial Plexus Injuries

In some situations the surgeon may finish anoperation as a neurolysis at the level of thespinal nerve because the continuity seems to bepreserved into the intervertebral canal. However,in some of these cases the rootlets might beavulsed but the spinal nerve not extracted fromthe intervertebral canal and the situation notrecognized. To avoid such situations it wassuggested that hemilaminectomy be performedin order to clarify avulsion of the rootlets, andthe surgical procedure be performed in a secondstage. However, since it is possible to prove orexclude this situation by intraoperative stimula-tion (Turkof et al 1995, 1997) using stimulation ofthe gyrus praecentralis, the author believes thata hemilaminectomy is not indicated just fordiagnostic reasons.

Direct neurorrhaphy

Direct neurorrhaphy is a very successful treat-ment in clean transections, e.g. with a stabwound. In a typical brachial plexus lesion withloss of nerve substance, a direct neurorrhaphyhas no chance and should not be attempted.

Nerve grafting to bridge the defect

Seddon attempted nerve grafting in five caseswith upper brachial plexus lesion with someresult in three patients before 1948, but he gaveup these attempts after Clark (1946) publisheddetails of transfer of the major pectoralis muscleto replace the biceps function. With the devel-opment of a reliable grafting technique (Millesi1968, 1972, Millesi et al 1972, 1976), free autol-ogous nerve grafting became the standardtechnique in brachial plexus surgery. It isobvious that without grafting, modern brachialplexus surgery would not be possible. All avail-able proximal stumps are connected with theircorresponding distal stumps. For instanceneuromas at C5 and C6 are connected with thedistal stump of the superior trunk, or a proximalneuroma of a superior trunk is connected with adistal stump on the superior trunk. In otherwords, attempts are made to reconstruct the

anatomical situation as well as possible. It isobvious that such an anatomical reconstructionis possible only if the defects are limited. In longdefects, e.g. between the proximal end of atrunk and the distal stump in the correspondingcord, regenerating nerve fibres will not alwaysproceed to the target muscles, and there will bea mixed innervation of antagonistic muscleswith co-contractures. Anatomical reconstructionis also limited by the fact that not enough autol-ogous nerve grafts are available to restore thecontinuity of all structures. Therefore, prioritymust be established to neurotize distal stumpsserving very important muscles and muscleswith good chances of recovery. Less importantmuscles and muscles with poor chances ofrecovery are excluded. Chances of anatomicalreconstruction will increase when artificial nervegrafts with cultivated Schwann cells of thepatient become available; this will solve theproblem of allografting.

After the description of the use of vascularizednerve grafts (Taylor and Ham 1976) the ulnarnerve was utilized as a vascularized nerve graft,and there were great expectations of thistechnique. The many reports can be summarizedas follows:

Vascularized nerve grafts provide a quickerneurotization; the end result does, however, notdiffer from the use of free nerve grafts. There arecertain indications for the use of vascularizednerve grafts, especially in nerve transfers (seebelow). If one wants to benefit from the ulnarnerve as a graft donor without utilizing it as avascularized nerve graft, the ulnar nerve has to besplit into minor units, in order to be sure ofsurvival. A nerve trunk the size of the ulnar nervewill not survive free grafting well because therelation between surface and tissue masses isunfavourable. After the preparation of minor unitsfrom the nerve trunk of the size of the sural nerve,survival is assured (Eberhard and Millesi 1996).

Nerve transfer (nerve fibre transfer)

In a nerve transfer, nerve fibres from one nerve aretransferred to another denervated nerve in order toneurotize this nerve. This can be from a peripheralnerve, for example the ulnar nerve, in order tobring nerve fibres to the musculocutaneous nerve

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 81

Page 91: Brachial Plexus Injuries

(Oberlin et al 1994). The majority of nerve transfers,however, are performed with axon donors outsidethe brachial plexus, in order to achieve particularfunctions within the brachial plexus.

It is evident that motor nerves have to be usedas axon donors to achieve a motor function, andsensory nerves to achieve a sensory function.There are a few cases in which nerve transferscan be performed by direct neurrorhaphy. In themajority of cases, nerve grafts are necessary.

In a classical nerve transfer, the function of theaxon donor has to be sacrificed in order to be ableto transfer the nerve fibres to the recipient nerve.With the technique of end-to-side coaptation(Viterbo 1994), this is not necessary. By using thismethod, it is possible to achieve neurotization of adenervated nerve without loss of function of thedonor nerve. This works very well with smallmonofunctional nerves; however, whether it workswith major mixed nerves has still to be proven.

Transfer of normal muscles,arthrodesis, tenodesis

In cases of partial brachial plexus lesions the lostfunction can be replaced by the transfer offunctioning muscles. Examples include:

• The Steindler operation, consisting of trans-position of the origin of the common head ofthe forearm flexors to the humerus shaft inorder to give a better lever arm for flexion ofthe elbow joint (Steindler 1918);

• Transfer of a segment of the major pectoralismuscle to restore elbow flexion (Clark 1946);

• Transfer of the trapezius muscle for the shoul-der joint (Saha 1967).

It is obvious that these procedures are applica-ble only in cases with partial lesions. Without afunctioning muscle, a muscle transfer is notpossible.

The above-mentioned techniques mainly referto upper brachial plexus lesions of the type C5and C6.

A Steindler operation will not be successful ifthe wrist joint is not to be stabilized, as in casesof the upper brachial plexus lesion of C5, C5, andC7. Here, wrist arthrodesis must be performedbefore the Steindler operation.

The lost radial nerve function in partial lesionsof C5, C6, C7 frequently needs a tendon transfer,as for a radial nerve paralysis.

In a lower brachial plexus lesion, restoration ofgripping function can be attempted by transfer ofextensor tendons.

Arthrodesis or tenodesis of the wrist provideswrist flexors or extensors for transfer to thefingers.

If there is a very good serratus anterior andtrapezius muscle, arthrodesis of the shoulderjoint may provide sufficient active motility andmay be a good solution. If, however, there is aserratus anterior, arthrodesis of the shoulderjoint is less good than with the above-mentionedtrapezius transfer. Several patients have beenunhappy following shoulder joint arthrodesis.

Transfer of regenerated muscles

Experience shows that a regenerated muscle isless functional than an original one. Usually, fornerve transfer it is required that the muscle to betransferred has a force of M4. In the case of exten-sive brachial plexus lesions, transfers areperformed with weaker muscles with the hopethat muscle strength can be improved by exercise.

Adding new muscles bymicrovascular muscle transplantation

If the local muscles of the paralyzed extremityhave undergone atrophy and degeneration, theonly way to achieve a functional result is bytransplantation of a muscle from another place,using microvascular anastomosis.

Different basic approaches todealing with post-traumaticbrachial plexus lesions

Limited neurotization

After having given up an attempt to bridgedefects by nerve grafts, Seddon (1975)

82 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 92: Brachial Plexus Injuries

successfully performed a nerve transfer froman intercostal nerve via a nerve graft to themusculocutaneous nerve (Seddon 1963). Thisbecame a standard approach in Tokyo(Tsuyama et al 1968). He did not explore thebrachial plexus, but only performed the inter-costal nerve transfer to the musculocutaneousnerve if there was a suspicion of avulsion. Allrecently denervated muscles with potential ofregeneration are neglected.

Using two or three intercostal nerves (2, 3, 4)and dissecting in a longway the intervertebraldirection to have the nerves long enough for aneurorrhaphy, a high percentage of usefulrecoveries been achieved (Nagano et al 1992,Ogino and Naito 1995).

Primary free muscle grafting

The use of free muscle grafting in inveteratedcases or in connection with impaired motorbranches has been mentioned already. Doi(1995) suggested immediate application of twofree muscle grafts without exploring the brachialplexus, in order to re-establish elbow flexionand finger extension with one of these muscles,and elbow extension and finger flexion with theother. The use of one muscle for two differentfunctions is always problematic, but Doi demon-strated good results. However, he achieved nosatisfactory shoulder function; the muscles ofthe extremity, which could at least have partiallyresponded if the brachial plexus had beenoperated on, were wasted.

Combined treatment

Brachial lesions are such complex injuries withsuch severe consequences that all availabletechniques should be applied. This concept wassummarized in 1973 (Millesi et al 1973); at firstthe brachial plexus is operated, and all availabletechniques are then applied to achieve as muchneurotization as possible. After a period ofphysiotherapy, the result of nerve regenerationcan be evaluated. In a second phase, all suitablemuscle transfers are performed to maximize theresult.

Description and discussion ofMillesi’s personal strategy andtechnique

The principle is to use as many axon donors aspossible to apply as many nerve grafts as possi-ble, and to neurotize as many muscles withchances of recovery as possible. The problem isnot should we do this or should we do that; allreasonable chances must be exploited. The treat-ment starts with surgery on the brachial plexusand ends 5 years later, once all the possibilitiesof conservative treatment, reconstructive surgeryand mechanical improvement have been utilized.

Position

The patient lies supine with the scapular regionsupported; the arm, the left part of the upperthoracic wall and the neck are scrubbed. The armis completely mobile and can be lifted andpositioned in lateral and medial directions. Thisis especially important because during the surgi-cal dissection underneath the clavicle it is impor-tant to lift the elbow and the upper arm in orderto provide space between the clavicle and thefirst rib.

Incision

Three separate incisions are used instead of thesingle zig-zag used in the early years butabandoned in about 1995. The main incision isperformed in the flexion crease, which is visiblewhen the head is bent to the paralysed side andthe shoulder is lifted. This line has an acuteangle to the clavicle, is in an area where leasttraction is manifested during head and shouldermovements, and can be extended in dorsaldirection to the trapezius muscle. If necessarythe dissection can explore the brachial plexusfrom above and behind. With a zig-zag incisionthe supraclavicular fossa is exposed after liftinga dorsally-based flap, and the base of this flapprevents good access to the plexus from aboveand behind. The incision should not beextended too far ventrally because the final scarwill not be as good in the pectoral area as inthe supraclavicular fossa.

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 83

Page 93: Brachial Plexus Injuries

Dissection

By undermining in both directions, the clavicle isexposed. By entering across the fascia of thesupraclavicular fossa, the superior and mediustrunk are easily accessible between the anteriorand medius scalenous muscle.

The dissection enters the infraclavicular fossabetween the pectoralis major and the deltoidmuscle. To gain more access, the lateral originof the major pectoralis muscle can be disinsertedfrom the clavicle.

It is a basic principle to start dissection innormal tissue and to proceed to pathologictissues. If, therefore, a lot of fibrosis is encoun-tered in the supra- or infraclavicular fossa, theplexus in the deltopectoral sulcus is explored viathird incision, which extends to and follows theanterior axillary fold. After entering between thetwo muscles, the fascia between the minorpectoralis and the coracobrachial muscle istransected and the individual structures of thebrachial plexus including the axillary artery, aredefined. The dissection can now proceed to theinfraclavicular fossa.

If the area of the deltopectoral sulcus is alsoinvolved in scar tissue, which is frequently thecase if there was a vascular lesion, the incisionis extended to the midline of the medial aspectof the upper arm. The nerves and vessels areexplored in the upper arm and dissectionproceeds from here to the axilla and to thedeltopectoral sulcus.

The minor pectoralis muscle is also isolated. Ifthe dissection from the deltopectoral sulcus hasreached the infraclavicular fossa, the situationhas been cleared here. The clavicle is isolated, aswell as the subclavius muscle.

In the supraclavicular fossa, the A. transversacolli and the omohyoideus muscles are identi-fied, as well as the external jugular vein. The twooperative fields in the supra- and infraclavicularfossae are now united by elevating the clavicle.At this point the arm is lifted, with a flexed elbowjoint, in order to increase the space between theclavicle and the first rib.

The supra- and medius trunks have beenexplored already from the supraclavicular fossa.The inferior trunk, however, is best identified byfollowing the medial cord underneath the clavi-cle while lifting the clavicle. The formation of thespinal nerves C8 and T1 is reached at the level

of the Sibson’s fascia. Both spinal nerves canthen be followed to the exit of the respectiveintervertebral canal. The phrenic nerve is identi-fied on top of the anterior scalenous muscle. Atthe lateral side of the scalenus medius muscle,the dorsalis scapulae nerve and the long thoracicnerve are apparent.

The medial and lateral pectoral nerves, thethoracodorsal and the subscapular nerves mustalso be identified. The suprascapular nerve isidentified at its origin from the superior trunkand followed in a distal direction. If there issuspicion of a compression site at the incisurascapulae, the dissection is extended to this struc-ture and the incisura scapulae opened.

By the end of the dissection, the whole lengthof the brachial plexus is exposed and access isavailable through a series of ‘windows’ – acrossthe first incision at the neck; in the supraclavicu-lar fossa cranial and caudal to the omhyoideusmuscle; between the clavicle and the subclaviusmuscle; between the subclavius muscle and intothe pectoral sulcus medial and lateral to thepectoralis minor muscle; in the axilla from thecranial direction; and, if the incision has beenextended to the upper arm, to the axilla from thecaudal direction.

It is never necessary to perform an osteotomyof the clavicle bone. As a rule the transversingmuscles are saved; only if the major pectoralismuscle causes an internal rotation contractureis it disinserted and reinserted under reducedtension.

Dealing with the lesion

If there is a lesion in continuity, neurolysis isperformed as described above. If theparaneurium or the epifascicular epineuriumhave developed a fibrosis, a para- or epineuri-otomy is performed in order to achieve decom-pression. If this is not sufficient, a para- orepineuriectomy is performed.

If no fascicular pattern can be detected, agrade IV lesion is assumed and a resection of theinvolved segment is performed, with restorationof continuity by free nerve grafts.

The problem of involvement of the rootletswith the spinal nerve being still located in theintervertebral canal, has already been discussed.Even in cases of grade VI damage where it is

84 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 94: Brachial Plexus Injuries

necessary to resect and bridge the defect, it is alimited defect with well-defined proximal anddistal stumps.

Root avulsions

If the roots are avulsed there is no proximalstump, and all axons to be directed into thedenervated nerves come from another source.Under these conditions it is impossible to neuro-tize all distal structures and, therefore, a list ofpriorities must be established.

The most important function is active elbowflexion, and therefore neurotization of the bicepshas highest priority. The best available axondonor is reserved for the musculocutaneousnerve. This is very frequently the accessorynerve. Two or three intercostal nerves alsoprovide a good chance for the biceps.

Secondly, it is extremely important for thepatient to be able to stabilize the shoulder jointand to have at least a little abduction. This isprovided by neurotizing the suprascapular or theaxillary nerve from nerve fibres coming frommotor branches of the cervical plexus.

Thirdly, innervation of the serratus anteriormuscle to give a stable scapula is extremelyimportant. Neurotization of the long thoracicnerve is therefore necessary. This can beachieved by using the dorsalis scapulae nerve.For several years we have not transected thedorsalis scapulae nerve but have preserved itscontinuity and applied a distal stump of the longthoracic nerve as an end-to-side coaptation tothe trunk of the dorsalis scapular nerve. This hasgiven excellent results in all cases in which thistechnique was utilized.

Finally, even the best recovery of elbow flexiondoes not help very much if the patient is unable toperform external rotation. For this function themajor pectoralis muscle is earmarked. This muscleis reinnervated by neurotizing both the medial andthe lateral pectoralis nerves. In our series this hasbeen done via nerve grafts, which have beencoapted end-to-side to the phrenic nerve withexcellent results. Motor branches of the cervicalplexus can also be used for this. Following success-ful reinnervation of the major pectoralis muscle, itis later transferred to perform external rotation.

Nerve fibres from the phrenic nerve have alsobeen successfully applied via nerve grafts

coapted end-to-side to the nerves supplying themajor pectoralis muscle and going to the supras-capular nerve. Intercostal nerves are used toneurotize the radial nerve, with special referenceto the triceps branches.

Good recovery of the triceps is frequentlyobtained by using intercostal nerves. In cases inwhich several intercostal nerves have been usedto achieve biceps function, the triceps musclehas also been neurotized. Of course, the bicepsand triceps are antagonists, but this is with theintention to transfer the triceps to the bicepstendon in order to have both acting as elbowflexors. This has been useful in several cases inwhich the biceps was neurotized by intercostalnerves and did not respond well, and the tricepswas the second line of defence.

The second intercostal nerve is coapted byend-to-end neurorrhaphy to the thoracodorsalnerve, in order to achieve a latissimus dorsifunction. This can be used for external rotationif the major pectoralis does not recover.

It is, of course, also the aim to achieve at leasta primitive gripping function, even in patientswith avulsion of all five roots. For this reason aC7 transfer is performed. There are severaloptions. The C7 transfer can be performed as asecond stage. It can be performed using the ulnarnerve as a vascularized nerve graft, connecting itsoriginal distal end with the contralateral C7 andits original proximal end with the median nerve.This has, in a majority of cases, provided reinner-vation of the finger flexors, the flexor pollicislongus and the flexor carpi radialis. In othercases, especially in children, free saphenousnerve grafts have been used for the C7 transferand the median as well as the ulnar nerve havebeen neurotized. In these cases one graft wasconnected to the anterior and one graft to theposterior division of the medius trunk. In severalof these cases flexor and extensor functionreturned, which of course offers the best chancesfor reconstruction of gripping function.

If only the flexor pollicis longus, finger flexorsand the flexor carpi radialis recover, arthrodesisof the wrist joint is performed and the flexorcarpi radialis is transferred for finger extension.In this case, of course, only a key grip functioncan be achieved.

In some of these patients a considerablesource joins the brachial plexus from C4, and thisis used as an axon donor for shoulder muscles.

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 85

Page 95: Brachial Plexus Injuries

In cases of avulsion of the roots 6, 7, 8, and T1with C5 present as a stump after rupture, C5 (asthe best available axon donor) is used for themusculocutaneous, and the accessory nerve forthe suprascapular nerve. Otherwise, the sametransfers are applied.

With avulsion of three roots (C5, C6, C7), asimilar transfer is performed but no C7 transferis necessary.

With avulsion of C7, C8, and T1, the stumps ofC5 and C6 are utilized to neurotize the superiorand medius trunks. For the inferior trunk, a C7transfer is considered.

In cases of avulsion of two roots (C5 and C6),nerve transfers from the accessory nerve and thecervical plexus are performed for the subscapularand axillary nerve; the anterior division of C7 isconnected to the musculocutaneous nerve and theposterior division to the radial nerve. The continu-ity of C8 and T1 and the inferior trunk is restored.

In cases of avulsion of C6 and C7, C5 is utilizedto neurotize the musculocutaneous and themedian nerves, the accessory nerve is used for thesuprascapular, the intercostal nerve for the radialnerve, and the continuity of C8 and T1 is restored.

In cases of avulsion of C8 and T1, the stumps ofC5 and C6 are utilized to neurotize suprascapular,axillary and radial nerve (C5), respectively, muscu-locutaneous and lateral median nerve (C6).

In cases of avulsion of C8 and T1, the conti-nuity of 5, 6, and 7 with the peripheral structuresof the lateral and dorsal cord are restored, andfor the inferior trunk a C7 transfer may be consid-ered. If there is only an avulsion of one root, itcan be ignored.

Rupture at root level

It has already been mentioned that with shortdefects we prefer to bridge the defect with longgrafts, connecting the proximal stump directlywith the corresponding peripheral nerves. It isobvious that the continuity of as many structuresas is possible by the availability of graft donorswill be restored.

Lesions at the level of trunks and cords

These are treated according to the general rulesof peripheral nerve surgery.

Axon donors

Accessory nerve

The accessory nerve is a very good axon donor;it is very useful for the suprascapular nerve andalso for the median nerve. It is transected afterthe first branch has left it. The trapezius muscleis usually not completely denervated by this.

Motor branches of the cervical plexus areuseful donors, they are identified by electricstimulation and connected by grafts, preferablywith shoulder muscles.

Phrenic nerve

The phrenic nerve is a very potent axon donor,but there is the danger of the loss of diaphragmfunction. End-to-side coaptation offers a goodopportunity to use this nerve without transec-tion. It has proven its value in neurotizing andachieving recovery of the suprascapular nerveand supraspinatus muscle, and the pectoralisnerve and the major pectoralis muscle.

Hypoglossus nerve

We have no personal experience of using thisnerve.

Supraclavicular nerves

These purely sensory nerves have been success-fully connected via a nerve graft with the mediannerve, to improve sensation.

Second intercostal nerve

This is useful for neurotizing the thoracodorsalnerve for the latissimus dorsi muscle.

Intercostal nerves

Intercostal nerves III and IV can be mobilized andbrought into direct contact with the distal stumpof the musculocutaneous or the radial nerve toachieve direct neurotization. Intercostal nerves V,

86 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 96: Brachial Plexus Injuries

VI, VII, and VIII have been used for musculocu-taneous and/or radial nerve (triceps); a nervegraft is always necessary.

Intercostobrachialis nerve

In selected cases these nerves have beensuccessfully connected with fascicles of themedian nerve by long nerve graft.

Dorsalis scapulae nerve

This is a useful donor for the long thoracic nerve,preferable by end-to-side coaptation.

Contralateral C7

It is well known that the transection of C7 aloneunder normal anatomical conditions produces aweakness of the radialis nerve innervatedmuscles, especially the triceps and the wristand finger extensors; however, there is noparalysis. These muscles regain their originalstrength by internal sprouting. This relates tothe posterior division. The anterior divisioninnervates the pectoralis muscle and othermuscles to a much lesser degree without visibleloss of function. There is always a smallsensory loss in the thumb and the index finger.In some cases this loss is very important and isnot well tolerated by patients. In cases of pre-or post-fixation the composition of the spinalnerve C7 is different, and may contain moreimportant fibres. There was one case in ourseries in whom biceps and deltoid function wassupplied by C7, and this patient had a consid-erable loss of function.

For this reason, a C7 transfer is neverperformed immediately. Ligature of the anteriorand posterior divisions of C7 is alwaysperformed first and the patient is carefullyexamined the next day. The real loss of functioncan then be assessed. If there is a major loss ofsensory function, only the posterior division isused. If there is a more important loss of motorfunction, the transfer is not performed at all. Inthe above-mentioned patient, after cutting theligature, function recovered completely. In thisparticular case the nerve grafts were applied in

end-to-side fashion and some regenerationfrom C7 to the contralateral arm was achievedafter end-to-side coaptation, but without usefulrecovery.

End-to-side coaptation

Viterbo (1994) noted good neurotization in a dener-vated nerve after end-to-side coaptation to aninnervated nerve. Many further experimentalstudies have confirmed this observation.Apparently a denervated nerve poses such astimulus to the axons that they start to sprout evenif uninjured, and even across an intactperineurium. That a nerve graft can also stimulatethis sprouting is very helpful, because it enlargesthe possibilities of end-to-side coaptation. Theproblem, however, is that there is no controlwhatsoever over which fibres will start to sprout –sensory or motor and, if motor, which function.This might be the reason for failures of end-to-sidecoaptation in experiments with peripheral nerves.

Graft donors

The following nerves have been used as donorsfor free nerve grafts:

• Sural (bilateral);• Cutaneus antebrachii medialis;• Cutaneus antebrachii lateralis;• R. superficialis nervus radialis;• Cutaneus fem. lateralis nerve (bilateral);• Saphenous (bilateral);• Split ulnar.

The following nerves have been used as donorsfor vascularized nerve grafts:

• Ulnar nerve, based on the superior ulnarcollateral artery and vein;

• Saphenous nerve (bilateral).

Postoperative care

Patients who have undergone brachial plexussurgery with neurolysis only are immobilized for2 days and start active and passive exercises

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 87

Page 97: Brachial Plexus Injuries

from the third day. Patients having had nervegrafts are immobilized for 8 days and start activeexercises (using non-paralysed muscles) andpassive exercises from the third week. Electricstimulation with exponential currents is recom-mended. The use of splints and prostheses isencouraged.

Reconstructive surgery

It has already been mentioned that duringsurgery on the brachial plexus, muscles areearmarked for secondary transfer.

The best example is the major pectoralisnerve, which is neurotized as a priority and hasthe scope to be transferred for external rotationafter recovery. In cases of intercostal nerve trans-fer to the biceps and triceps, the triceps isearmarked for transfer, in order to support thebiceps which may be too weak alone. Thetriceps, of course, has to perform elbow flexionalone if the biceps does not recover.

A similar procedure (triceps transfer) isperformed in cases in which a simultaneousinnervation of biceps and triceps occurs withco-contractions. Biceps and triceps neutralizeeach other, and the procedure is sometimesregistered as a failure. In such cases there aretwo options. The first is to paralyse the tricepswith botulinum toxin. During paralysis theforce of the biceps muscle is increased as muchas possible by physiotherapy. If the tricepsrecovers, the biceps should be much strongerand perform elbow flexion in spite of co-contractions. Before botulinum toxin was avail-able, a lengthening operation of the tricepstendon was performed in order to diminish itsforce. If, however, co-contractions are sostrong that the force of biceps and triceps areequal, or that the triceps is even stronger thanthe biceps, then a transfer of the triceps on thebiceps tendon is performed; both muscles thenact as elbow flexors and extension is providedby gravity. The lack of active extension in thesecases is not as important as in other patientsbecause, due to the weakness of the shouldermuscles and the reduced function in the shoul-der joint, the patient will be unable to work atthe level where strong active extension isneeded.

Weak elbow flexion by a weak biceps can beimproved by transferring the biceps insertionmore distally to the radius, thereby increasingthe lever arm.

In cases of an upper brachial plexus lesion, aSteindler operation may be performed if theforce of the returning elbow flexion is insuffi-cient.

In cases of partial brachial plexus lesions, thepectoralis major (Clark 1946) or the latissimusdorsi (Zancolli and Mitre 1973) may be utilized torestore elbow flexion. The function of the shoul-der joint can be improved, as far as abduction isconcerned, by a trapezius transfer. Disinsertionof the trapezius muscle and reinsertion on thehumerus provides little active abduction butgood stability. A better option is to isolate thetrapezius muscle, mobilize the supraspinatus,and suture the trapezius to the supraspinatuswith its distal part across the canal to thehumerus, and to perform the insertion at therotator calf.

Active external rotation can be achieved bytransfer of the pectoralis major and, eventually,the pectoralis minor, the latissimus and the teresmajor; also by transferring the tendon of thesubscapularis.

In rare cases a supination contracture, asfrequently happens in obstetric brachial plexuslesions, may develop in a patient with an upperbrachial plexus lesion. It is corrected by transec-tion of the interosseus membrane and transfer ofthe biceps tendon or, to give the biceps tendonits pronating effect, by reinserting the biceps ina neutral position. In this case, of course, activesupination is lost, but the effect of the biceps inflexing the elbow joint is increased.

In cases of a C5, C6, and C7 lesion a tendontransfer is performed, as in radial nerve palsy.

The most challenging problem is the restora-tion of gripping function with the few musclesof the forearm that might have recovered. Inspite of the fact that arthrodesis is not always asuccessful procedure, this is performed in caseswhere only very few muscles are working andthe wrist flexors or extensors are needed. It isperformed using a plate between the thirdmetacarpal bone and the radius in the neutralposition, or even in slight flexion. Patients donot appreciate a dorsiflexion position (as wouldbe used in other cases) because they want to beable to put the hand in their pocket of their

88 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 98: Brachial Plexus Injuries

trousers. In many cases a tenodesis can beperformed to stabilize the wrist. For eachindividual patient a particular plan has to bedeveloped. If there are three muscles, onemuscle may be used to provide MP joint flexionand the second for finger flexion, beingconnected to the flexor digitorum profundustendons and the flexor pollicis longus tendon.The third muscle is used for finger and thumbextension. It is obvious that only a very primi-tive gripping function can be achieved. In orderto provide a sufficient key grip, arthrodesis ofthe IP joint in the neutral position is done. Thetendon of the flexor pollicis longus is disinsertedand transferred to the lateral dorsal aspect of thefirst phalanx, in order to give the thumb a goodadduction and a small amount of pronation. Ifthere are four muscles, one muscle is connectedto the flexor pollicis longus tendon to giveindependent thumb adduction and, if there arefive muscles, the fifth muscle is used for thumbextension.

Results

In almost 40 years of brachial plexus surgery,there has been a constant improvement inresults. Even in cases of five root avulsions, suffi-cient active elbow flexion can be expected inabout 80 per cent of cases, useful control of theshoulder joint and external rotation in about 60per cent of cases, and gripping function in theform of a key grip in at least 30 per cent of cases.

Critical remarks

The majority of cases achieve some sensoryinput, very often in the form of paraesthesiawithout being able to localize the touch. Patientswith intercostobrachial to distal radial transfersget protective sensibility.

The results of brachial plexus surgery dependon many individual factors, and, therefore, everysurgeon has a certain percentage of failure.However, cases are frequently seen which,although declared as final results by the originalsurgeon, still have the potential for furtherimprovement by muscle or tendon transfer,

tenodesis, arthrodesis or other reconstructiveprocedures. It is not unusual for a patient to beregarded as a failure because active elbowflexion was not achieved, due to co-contracturesbetween biceps and triceps.

It is also necessary to emphasize that, in theauthor’s view, it is a failure if all available optionsare not explored.

References

Bonney G (1983) The case for early surgery. Presentedat the Symposium on Brachial Plexus Injuries at StMary’s Hospital, London, January 1983.

Clark JPM (1946) Reconstruction of biceps brachii bypectoral muscle transplantation, Br J Surg 34:180.

Doi K (1995) Double free muscle transfer to reconstructprehension following complete avulsion of brachialplexus. Video presentation, 11th Congress of theInternational Society for Plastic Reconstructive &Aesthetic Surgery, Yokohama, April 16–22.

Eberhard D, Millesi H (1996) Split nerve graft, JReconstr Microsurg 12:71–6.

Millesi H (1968) Zum Problem der Überbrückung vonDefekten peripherer Nerven, Wien Med Wschr118(9/10): 182–7.

Millesi H (1972) Die Eingriffe an den Hand- undFingernerven. In: Wachsmuth W, Wilhelm A, eds.Allgemeine und Spezielle Chirurgische Operationslehre:Die Operationen an der Hand, Vol 10/3. Springer: Berlin:226–50.

Millesi H, Berger A, Meissl G (1972) The interfascicularnerve grafting of the median and ulnar nerves, J BoneJoint Surg 54A:727–50.

Millesi H, Meissl G, Katzer H. (1973) Zur Behandlungder Verletzungen des Plexus Brachialis. Vorschlag zurintegrierten Therapie, Bruns Beiträge Klin Chirurg220:429–46.

Millesi H, Berger A, Meissl G (1976) Further experi-ences with interfascicular grafting of the median,ulnar, and radial nerves, J Bone Joint Surg58A:227–30.

Nagano A, Ochiai N, Okinaga S (1992) Restoration ofelbow flexion in root lesions of brachial plexus injuries,J Hand Surg 17(A):815–21.

UPDATE ON THE TREATMENT OF ADULT BRACHIAL PLEXUS INJURIES 89

Page 99: Brachial Plexus Injuries

Narakas A (1986) Brachial plexus injury. Paper at the3rd Congress of the International Federation ofSocieties for Surgery of the Hand, Tokyo, November.

Oberlin C, Beal D, Leechavengvongs S et al (1994) Nervetransfer to biceps muscle using a part of ulnar nerve forC5–C6 avulsion of the brachial plexus: anatomical studyand report of four cases, J Hand Surg 19A:232–4.

Ogino T, Naito T (1995) Intercostal nerve crossing torestore elbow flexion and sensibility of the hand for aroot avulsion type of brachial plexus surgery,Microsurgery 16:571–7.

Saha AK (1967) Surgery of the paralyzed and flailshoulder, Acta Orthop Scand (Suppl) 97:1.

Seddon HJ (1963) Nerve grafting, J Bone Joint Surg45B:447.

Seddon HJ (1975) Surgical Disorders of the PeripheralNerves, 2nd edn. Churchill Livingstone: Edinburgh: 194.

Steindler A (1918) Reconstruction work on hand andforearm, NY Med J 108:117.

Sunderland S (1951) A classification of peripheralnerve injuries producing loss of function, Brain 74:491.

Taylor GI, Ham FJ (1976) The free vascularized nervegraft, Plast Reconstr Surg 57:413.

Tsuyama N, Sagakuchi R, Hara T, Kondo T et al (1968)Reconstructive surgery in brachial plexus injuries. In:Proceedings of the 11th Annual Meeting JapaneseSociety of the Hand, Hiroshima, p. 39–40.

Turkof E, Monsivais J, Dechtyar I et al (1995) Motorevoked potential as a reliable method to verify theconductivity of anterior spinal roots in brachial plexussurgery: an experimental study on goats, J ReconstrMicrosurg 11(5):357–62.

Turkof E, Millesi H, Turkof R et al (1997) Intra-operative electroneurodiagnostics (transcranialelectrical motor evoked potentials) to evaluate thefunctional status of anterior spinal roots and spinalnerves during brachial plexus surgery, Plast ReconstrSurg 99:1632–41.

Viterbo F (1994) Two end-to-side neurorrhaphies andnerve graft removal of the epineural sheath: experi-mental study in rats, Br J Plast Surg 47:75–80.

Zancolli E, Mitre H (1973) Latissimus dorsi transfer torestore elbow flexion. An appraisal of eight cases, JBone Joint Surg 55A:1265.

90 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 100: Brachial Plexus Injuries

Introduction

There are profound differences between supra-and infraclavicular lesions; chief amongst theseis the high incidence of intradural (pre-ganglionic) lesion in the former (Bonney 1954,Bonney and Gilliatt, 1958).

Injuries to the infraclavicular plexus and termi-nal branches of the brachial plexus are oftencaused by great violence and there is frequentlysevere damage to the scapula, clavicle andhumerus. Damage extends not only to the skele-ton but also to the nerves, where wide retractionof the stumps in traction injury or extensivedestruction of nerve substance is common.

Rupture or transection of the subclavian/axillary/common brachial arterial axis isfrequent; the author found this in 25 per cent ofclosed traction lesions, with a greater incidencein knife wounds or injuries caused by penetrat-ing missiles.

The neurological lesion is a complex one, withruptures of nerves at different levels and withavulsion of nerves such as the musculocuta-neous or circumflex directly from the relevantmuscle bellies. This complexity is increased bythe combination of lesions at two levels, whichoccurs in about 15 per cent of cases (Fig. 1).

The notion of ‘motor’ and ‘sensory’ nervesshould be abandoned, for no such distinctionexists within the peripheral nervous system. Thespinal accessory nerve has no cutaneous compo-nent and, as Bremner-Smith et al (1999) haveshown, the majority of its fibres are non-myeli-nated and are probably afferent nociceptors. Thesuprascapular nerve, another trunk with nocutaneous innervation, contains many myeli-nated afferents from the muscle spindles, tendonand capsule.

Diagnosis

A distinction must be made between degenera-tive and non-degenerative lesions. When anaxon is cut, Wallerian degeneration follows, theaxon degenerates, and conduction is lost atabout 3–4 days. All function is lost. A particularly

10Injuries of the terminal branches ofthe brachial plexusRolfe Birch

Figure 1

Diagram of brachial plexus.

Page 101: Brachial Plexus Injuries

important clinical sign in this situation is vaso-and sudo-motor paralysis owing to interruptionof the post-ganglionic sympathetic efferentfibres. The skin in the territory of the affectednerve becomes red and dry.

The progressive changes in the target organsof muscle, skin, blood vessels and the sensoryorganelles are well known. Muscle denervatedfor 2 years or more is lost to function, but ofequal significance are the changes that occurproximally after transection of the axons. Theproximal portion narrows, and in the parentneurone there is chromatolysis and retraction ofdendrites. According to Bonney (1998), ‘theseprocesses may continue to actual dissolution ofthe cell body’. Dyck et al (1984) were able tostudy the effect of permanent axonotomy in thespinal cord of two patients years after amputa-tion of a lower limb, and they found that ‘a lossof target tissue by axonotomy leads to atrophyand then loss of motor neurones’. These centralchanges are more extreme and more proximal inextensive and violent injuries, and they arefundamental factors in the prognosis after repair;they are the most important reasons for minimiz-ing the delay in effecting nerve repair.

Seddon (1943) classed nerve injuries into threegroups:

1. Neuropraxia (the nerve not working), whichimplies a physiological block to conductionbut no anatomical disturbance of the nerve.Distal conduction persists in this non-degen-erative lesion;

2. Axonotmesis (axon cutting), in which theaxon is severed; there is degeneration of thedistal axon, but the basal lamina of theSchwann cell remains intact;

3. Neurotmesis (nerve cutting), which is wherethere is interruption of continuity of allelements of the nerve.

In the latter two types, Wallerian degenerationoccurs and distal conduction is lost; the clinicalfeatures are very similar.

A favourable lesion may become a much lessfavourable one if the cause is not removed.Recovery of a nerve accidentally encircled by asuture or crushed under a plate is likely if treat-ment is immediate (Fig. 2). If the offending causeremains for days or weeks, then a far lesssuccessful outcome is inevitable.

Sunderland (1951) introduced a more elabo-rate system of classifying injuries, recognizingfive degrees of severity. Some clinicians mayfind Sunderland’s classification an improvementon Seddon’s, and of more practical use than the

92 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 2

Axillary artery and lateral cord wereencircled by sutures during anteriorstabilization of shoulder.

Page 102: Brachial Plexus Injuries

earlier method, but we tend towards a furthersimplification: to classification as ‘degenerative’or ‘non-degenerative’. This is how cliniciansshould regard nerve injuries; the first questionshould be, is this lesion degenerative or nondegenerative? (Table 1.)

Severe pain indicates substantial nerve injurythat cannot be consistent with a diagnosis of non-degenerative conduction block (neuropraxia).Tinel’s sign (1917) is an important clinical feature.A strongly positive Tinel’s sign over a lesion soonafter injury indicates rupture or severance, andwill not be found in a conduction block or a nondegenerative lesion. Failure of distal progressionof the Tinel’s sign in a closed lesion indicatesrupture or another injury not susceptible to recov-ery by natural process.

In closed injuries with associated fracture ordislocation, the history of the injury is all-impor-tant. The force expended upon the nerve trunkcan be estimated from knowledge of the veloc-ity and the force of impact of an object, or theheight of the fall. Local bruising or abrasions orlinear bruising along the course of the nerve isimportant evidence of this force. X-rays arevaluable in showing the extent of the displace-ment of bone fragments, and imperfect reductionor obstruction to reduction implies interpositionof muscle, nerve or artery (Fig. 3).

Classification of wounds

The single most important determinant ofoutcome is the violence of injury to the nerve

and the limb; the extent of destruction of thetissue is a reflection of this. Rank et al (1973)classified injuries of the hand as ‘tidy’ (amenableto primary repair) or ‘untidy’ (characterized byextensive destruction of tissue and by contami-nation). This important distinction can be appliedto nerve injuries.

In a tidy wound, caused by a knife, glass or thesurgeon’s scalpel, damage is confined to thewound itself and primary repair of all dividedstructures is desirable. Associated arterial injuryis common.

An untidy wound, commonly caused by openfractures or by penetrating missile injuries,

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 93

Table 1 Classification of focal mechanical nerve injury(from Thomas and Holdorff 1993, with permission)

I. Focal conduction block• Transient

IschaemicOther

• More persistent demyelinatingAxonal constriction

II. Axonal degeneration• With preservation of basal laminal sheaths of nerve

fibres• With partial section of nerve• With complete transection of nerve

Figure 3

Long segment of disruption of radial nerve from closedfracture of the shaft of humerus.

Page 103: Brachial Plexus Injuries

presents with extensive tissue damage and ahigh risk of sepsis. Arterial injury is common.The initial aims include restoration of the stabil-ity of the skeleton and of the circulation, withdelayed closure of the skin or early recourse toa pedicle or free cutaneous/myocutaneous flap.Nerves will almost always require grafting, andtiming is a matter of fine judgement. Woundscaused by industrial or agricultural machinery,burns or penetrating missiles do not, as a rule,permit primary repair. An amputated limbaffords the most uncompromising model of thistype of injury. Primary repair of the nerve givesthe best results in these severe cases.

The closed traction rupture –

nerves injured by fractures or

dislocations

Damage to the adjacent skeleton can cause nerveinjury by traction, laceration (by fragments ofbone), entrapment within the dislocated joint orin a fracture, and later, entrapment or compres-sion by callus. On the whole, dislocations aremore damaging. It seems to be widely assumedthat the prognosis for nerves injured in thismanner is good, but this is not generally the

case. Seigel and Gelberman (1991) reviewed thesubject thoroughly, finding 85 per cent of nervepalsies after closed fractures and 65–70 per centafter open fractures. Of those nerves that wenton to recover, 90 per cent had done so by 4months; these cannot have been wholly degen-erative lesions. Seigal and Gelberman’s indica-tions for intervention include: that the fractureneeds internal fixation, that there is associatedvascular injury; that wound exploration of anopen fracture is necessary; or that fracture ordislocation is irreducible. Two more indicationsmay be added; if the lesion worsens whilst underobservation; and if the lesion occurrs duringoperation for internal fixation of a fracture (Fig.4).

It is important to distinguish the true infra-clavicular lesion, in which multiple ruptures ofnerve trunks with associated fractures andarterial damage owing to lower energy transferinjuries is associated with fracture dislocation ofthe shoulder, in which damage to the rotatorcuff, mal- or non-union of the proximal humerus,or intra-articular damage to the gleno-humeraljoint are very unfavourable features. The infra-clavicular lesion is a distinct entity, generallycaused by violent hyperextension at the shoul-der; there is almost always a fracture of the shaftof the humerus or injury to the gleno-humeraljoint, the incidence of vascular injury is high, and

94 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 4

Closed fracture of the shaft ofhumerus in a 34-year-old woman.Rupture of radial nerve, with thedistal stump entrapped within thefracture. The fracture had beenpinned.

Page 104: Brachial Plexus Injuries

the level of proximal rupture is deep to thepectoralis minor, which acts as a guillotine onthe neurovascular bundle.

These injuries are immensely destructive ofnerves and axial vessels. They are characterizedby retraction of the ruptured nerves and vesselsand by considerable longitudinal damage withinthe ruptured trunks. When complicated byarterial lesions, which occur in about 25 per centof cases, the outcome following nerve repair isthe worst of all groups (Fig. 5).

Indications for operation andprinciples of nerve repair

The decision to intervene with a nerve injury isnever easy, except perhaps in the acute case ofan open wound, and in cases in which nerveinjuries are associated with major bones andblood vessels. The indications for such opera-tions are probably as follows:

1. Deep paralysis after wounding over thecourse of a main nerve or injection close tothe course of the main nerve;

2. Deep paralysis after a closed injury,especially high energy transfer injury associ-

ated with severe damage to soft tissues,bones or joints;

3. Deep paralysis after a closed traction injury;4. Association of a nerve lesion with evidence of

an arterial lesion;5. Association of a nerve lesion with fracture of

a related bone requiring early internalfixation;

6. Worsening of a nerve lesion whilst underobservation; failure to show evidence ofrecovery at a fixed time after a closed lesioninitially thought to be an axonotmesis;

7. Failure to show evidence of recovery inconduction block within 6 weeks of injury;

8. Persistent pain at almost any interval afterinjury.

Where direct suture of a nerve is not possible,then the mainstay technique remains autologousgrafting. The results of operating on peripheralnerves depend, to an extent hardly matched inany other branch of surgery, on the skill of thesurgeon and the quality of the technique. Nervesuture is performed when the gap after resectionis small, little mobilization of the nerve is neededto close it, and the repaired nerve lies withouttension and without excessive flexion of adjacentjoints. The experimental work of Clark et al(1992) clearly showed the ill effects of tension onthe repaired nerve.

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 95

Figure 5

25-year-old patient, high-speedroad traffic accident causing infra-clavicular traction lesion. Ruptureof axillary artery and radial nerve.

Page 105: Brachial Plexus Injuries

Vascular lesions

George Bonney introduced the policy of urgentprimary repair of nerves and vessels in thecombined neurovascular injury unit at St Mary’sHospital in the mid 1960s, and this policy hasbeen followed ever since. Of the cases describedhere, repair of nerves and vessels wasperformed within 48 hours in 98 cases, and laterthan this time in 66 cases. It was necessary torevise arterial repairs performed in the referring

hospital in 32 more cases. Diagnosis ofaneurysm or arteriovenous fistula was madebetween 10 days and 4 years from causal injury.Proper repair of a vascular lesion is the singlemost important factor governing prognosisamongst all of those within the surgeon’scontrol. Delay in the repair of nerves after arterialrepair presents technical difficulties, which aresometimes insuperable and invariably diminishthe level of functional return (Table 2).

Open injuries

The level of the arterial injury is not necessarilyrelated to the entry wound. Complete resectionof the artery does not inevitably lead toexanguination, as the vessel constricts in spasm(an observation first made by John Hunter). Theslowly expanding false aneurysm causesprofound loss of conduction in adjacent nervetrunks; the deep-seated aneurysm may not beclinically diagnosed unless the potential signifi-cance of an earlier wound is noted (Fig. 6).

Closed traction injuries

The first injury is fracture of the intima, which inmore violent injury progresses to rupture of all

96 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 2 Arterial injuries in association with nervelesions, 1975–1996 (from Surgical Disorders of thePeripheral Nerves (Bonney 1998))

Closed Open

Total Repaired Total Repaired

Vertebral 1 0 4 1Subclavian 82 40 9 9Axillary 63 54 31 27Brachial 7 6 35 31

Traumatic false aneurysms and arteriovenous fistulaewith associated nerve lesions

Region – artery Aneurysms Arteriovenous fistulae

Posterior triangle 3 2of neck

Axillary 14 3Brachial 4 1

Figure 6

Transection of axillary artery fromknife wound in a 33-year-oldwoman. There was little bleedingbecause of spasm within thedivided artery.

Page 106: Brachial Plexus Injuries

coats of the vessel. The intimal injury can easilybe seen as a pale crescent-shaped line, thedamaged segment of the artery being filled withthrombus (Fig. 7). Attempts to restore flow usingembolectomy catheters are invariable futile. Thelevel of the arterial injury is consistent in closedtraction injuries, the axillary artery being foundoccluded deep to pectoralis minor. Urgentamputation was performed in two cases of gasgangrene occurring with 24 hours of successfulrepair by prosthesis, in four cases with upperlimbs so badly damaged that there was no possi-bility of functional recovery, and in two cases(limbs amputated through the gleno-humeraljoint) where operation for replantation wasabandoned because of uncontrollable bleeding.Reasonable circulation to the skin is not theissue; what is important is whether muscle isperfused, and failure to restore the flow througha damaged axillary or common brachial arterywithin 6 hours of injury is almost alwaysfollowed by a degree of post-ischaemic fibrosis.The exposure of choice is that developed byFiolle and Delmas (1921); the reader is referredto Surgical Disorders of the Peripheral Nerves(Birch, 1998) for further discussion of this andother technical issues.

Distal circulation was successfully restored in148 cases, with strong peripheral pulses. In threecases, failure to restore bloodflow necessitatedamputation.

Results

Measurement of outcome

Seddon (1975) described a system for measure-ment of outcome which was drawn from theMedical Research Council system, and classifiedresults of nerve repair as fair, poor or bad. Wehave simplified this further into good, fair or poor,using the grade ‘excellent’ rarely and in excep-tional cases where function is indistinguishablefrom normal. For some nerves muscular functionis a good deal more important than recovery ofsensation, and for such nerves (e.g. the spinalaccessory, suprascapular, circumflex, musculocu-taneous and radial) little significance is attached tothe extent of recovery of cutaneous sensationexcept when recovery is complicated by severepain, when the result is considered poor. Recoveryfor sensation has been given equal importance tomuscle function in the description of results inmedian and ulnar nerves, and it could be arguedthat sensibility is the most important function ofthe median nerve. There are of course defects andlimitations of the MRC system, but it has stood thetest of time. Kline and Hudson (1995) described avaluable grading system for motor and sensoryfunction, and for the whole nerve. The particulardifficulties of measurement of limb function incases of injury to the adult brachial plexus were

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 97

Figure 7

63-year-old man. Rupture of axillaryartery after fracture dislocation ofshoulder.

Page 107: Brachial Plexus Injuries

addressed by Narakas (1989), and his system iscertainly useful in the measurement of outcome ofthe more severe cases of multiple ruptures of trunknerves (Tables 3, 4).

Spinal accessory nerve injury

The XIth cranial nerve innervates the sternoclei-domastoid and trapezius muscles, and has no

cutaneous distribution. The nerve is particularlyat risk from surgeons where it emerges fromdeep to the sternomastoid into the posteriortriangle of the neck. Damage may be inevitableduring radical neck surgery for cancer. In mostcases injuries were inflicted during the course oflymph node biopsy or other operations for abenign condition within the posterior triangle ofthe neck. Williams et al (1996) have described thecharacteristic syndrome of symptoms and clini-cal signs as pain, drooping of the shoulder, lossof abduction, winging of the scapula, andnumbness of the face and ear caused by damageto the adjacent transverse cervical and greaterauricular nerves (Fig. 8; Tables 5, 6).

98 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 3 Classification of outcome of nerve recovery(from Surgical Disorders of the Peripheral Nerves(Bonney 1998))

Motor recoveryM0 No contractionM1 Return of perceptible contraction in the proximal

musclesM2 Return of perceptible contraction in both proximal

and distal musclesM3 Return of perceptible contraction in both proximal

and distal muscles of such degree that allimportant muscles are sufficiently powerful to actagainst resistance

M4 Return of function as in Stage 3, with the additionthat all synergic and independent movements arepossible

M5 Complete recovery

Sensory recovery

S0 Absence of sensibility in the autonomous areaS1 Recovery of deep cutaneous pain sensibility within

the autonomous area of the nerveS2 Return of some degree of superficial cutaneous pain

and tactile sensibility within the autonomous areawith disappearance of any previous over-reaction

S3+ Return of sensibility as in Stage 3, with theaddition that there is some recovery of two-pointdiscrimination within the autonomous area

S4 Complete recovery

Table 4 Grading of results (from Surgical Disorders ofthe Peripheral Nerves (Bonney 1998))

Motor recovery Sensory recovery

M4 or better GoodM3 FairM2 PoorM1 & O Bad

S4 (normal) or S3+ GoodS3 FairS2 PoorS1 & 0 Bad

Figure 8

Lesion of the spinal accessory nerve. Note the apparentwinging of the scapula.

Table 5 Accessory nerve lesions: classification ofrecovery (from Williams et al 1996)

Outcome Grade Number

A No change Poor 7B Pain improved Fair 10

Movement improvedC Almost normal (difficulty Good 15

with overhead work)D Normal (from patient’s point Excellent 4

of view)

Page 108: Brachial Plexus Injuries

Nerve to serratus anterior

This nerve is most commonly damaged inlesions of the brachial plexus, but it is particu-larly vulnerable to accidental damage where itcrosses the first and second ribs. When the serra-tus anterior is paralysed, the inferior pole of thescapular does not move forwards but slidesmedially and cranially. Confusion with spinalaccessory palsy is a common error. Deep, achingpain is usual, and is sometimes severe. When thenerve was repaired after a stab wound (sevencases) or intra-operative damage (11 cases),results were very good, perhaps better than inalmost any other peripheral nerve. In cases

where the proximal stump was unavailable inter-costal transfer to the distal stump of the nerveregularly restored useful function within themuscle (Fig. 9).

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 99

Table 6 Results of repairs of the spinal accessory nerve

Cause Number of repairs Outcome

Poor Fair Good Excellent

Stab 7 2 1 2 2PMW 5 1 2 2 0Traction 3 1 1 1 0Iatropathic 63 7 12 36 8TOTAL 78 11 16 41 10

Figure 9

Serratus anterior palsy. The scapula moved cranially andtowards the midline.

Figure 10

Rupture of circumflex nerve. Full elevation of shoulderthrough the suprascapular nerve and rotator cuff.

Page 109: Brachial Plexus Injuries

The circumflex and suprascapularnerves

Ochiai and colleagues (Ochiai et al 1997, Mikamiet al 1977) showed that the suprascapular nervemight be severely damaged in several places,and the recommended exposure is described indetail in his valuable paper. We reportedfindings in 129 nerve injuries in 98 patients(Birch and Spilsbury 1996). Spilsbury’s myomet-ric study suggested that the deltoid muscle wasresponsible for over 50 per cent of the power ofabduction, about 30 per cent of forward flexionpower and 80 per cent of extension power of theshoulder. Improvements in strength andstamina were not impressive, even in thosecases where the outcome was considered good.Results were particularly bad when there wasassociated rupture of the rotator cuff. Intra-artic-ular fracture provoked post-traumatic arthritis ina number of cases. All circumflex nerve repairsperformed more than 1 year after injury failed;most of those repaired after an interval of 6months fared better. Bonnard et al (1999)presented a detailed analysis of 146 cases, andamongst the many notable findings was the factthat: ‘the dramatic decrease in the rate ofsuccess seen with longer delays suggests thatsurgery should be undertaken within threemonths of injury’.

It is extremely unwise for the clinician to makea diagnosis of isolated circumflex nerve palsy ina patient who is unable to abduct the arm.Patients with an intact suprascapular nerve androtator cuff have a good range of abduction, andin many cases that range is full. Patients whopresent with paralysis of the deltoid and whocannot abduct the arm must have rupture of therotator cuff and/or rupture of the suprascapular

nerve unless proven otherwise (Fig. 10). Ourprefered investigation in these situations is MRscanning.

Another remarkable feature about injuries tothe circumflex nerve, which is not widely appre-ciated, is the high incidence of damage to theoffsets of the axillary artery, most especially thecircumflex-humeral and subscapular vessels.

100 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 7 Grading of results for the circumflex and suprascapular nerves (from Surgical Disorders of the PeripheralNerves (Bonney 1998))

Circumflex nerveGood Deltoid MRC 4 or better Abduction Elevation at least 120°Fair Deltoid MRC 3+ or better Abduction Elevation 90–120°Poor or bad Deltoid Less Less

Suprascapular nerveGood Abduction 120° or more Lateral rotation 3° or moreFair Abduction 90–120° Lateral rotation 0–30°Poor Less Less

Table 8 Results of repairs to the circumflex andsuprascapular nerves

Good Fair Poor Total

Circumflex nerve35 31 15 81

Suprascapular nerve24 8 6 38

Figure 11

Delayed rupture of false aneurysm of subscapular artery in83-year-old woman 8 weeks after dislocation of shoulderand with rupture of circumflex nerve.

Page 110: Brachial Plexus Injuries

These injuries lead to considerable fibrosis oreven to false aneurysm formation (Tables 7, 8;Fig. 11).

The Radial nerve

This is the largest terminal branch of the brachialplexus, and was the most commonly injured nervetrunk in 16 500 cases of war wounds (Sunderland1978). In civilian life, the radial nerve seems to beinjured less commonly than the median or ulnarnerves. Zachary (1954) studied 113 cases of repairby direct suture within 6 months of injury, themaximum amount of nerve resected being 5 cm.There was a good or fair outcome in 61.5 per centof cases. In Seddon’s series (Seddon 1975) of 63nerve sutures, 77.8 per cent of results were gradedgood or fair. Kline and Hudson (1995) described171 operated cases of radial nerve injuries, andfound that the results were best in more distallesions. The outcome after repair of laceratednerves was better than that following repair ofnerves damaged by fracture or gun shot. Results

were best after primary repair, followed by thosewith secondary repair; the worst results were inthose cases requiring grafting. Birch et al (2001)described the outcome of 242 repairs of radialnerves, finding only 30 per cent good results and28 per cent fair. The violence of injury was themost important factor in prognosis. Of the open,tidy repairs, 79 per cent achieved a good or fairresult; 36 per cent of cases with arterial injuryreached this level. Most repairs failed when thedefect in the nerve trunk exceeded 10 cm. Ofrepairs performed within 14 days of injury, 49 percent achieved a good result; only 28 per cent oflater repairs did so. All repairs undertaken beyond12 months failed. It is perhaps significant that inthis series no less than 16 of 18 repairs of theposterior interosseous nerve achieved a goodresult (Table 9).

The musculocutaneous nerve

Lesions of this nerve accounted for less than 2per cent of a series of 14 000 nerve injuries

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 101

Table 10 Results of repair of musculocutaneous nerve (85 nerves)

Type of injury Good (elbow flexion Fair (elbow flexion Poor (elbow flexion TotalM4 or better) M3, M3+) M2 or less)

Open ‘tidy’ 12 0 1 13Open ‘untidy’ 15 7 2 24Closed traction 30 10 8 48

Effect of rupture of axillary or brachial arteryArterial injury 25 10 8 43None 32 7 3 42

Effect of skeletal injuryAssociated skeletal injury 23 10 10 43None 34 7 1 42

The Chi-square test was used to analyse these data.The P-value for type of injury and outcome was 0.1779, for association with arterial injury was 0.1617, and for association with skeletalinjury was 0.0067.

Table 9 Results of repair of radial nerve by wound type

Wound No. of nerves repaired Good Fair Poor

Open ‘tidy’ 73 38% (28) 41% (30) 21% (15)Closed traction 62 31% (19) 27% (17) 42% (26)Open ‘untidy’ 52 25% (13) 25% (13) 50% (26)Associated vascular injury 55 22% (12) 14% (8) 64% (35)Total 242 30% (72) 28% (68) 42% (102)

Page 111: Brachial Plexus Injuries

incurred in World War II, and it was rarelydamaged in isolation (Sunderland 1978). Resultsof repair are rather better than for other nervesin the upper limb. Seddon (1975) reported satis-factory results in all 10 of his cases, and Klineand Hudson (1995) obtained a useful outcome in29 cases of rupture in ‘infraclavicular/stretchinjuries’.

Osborne et al (2000) recently published a studyof results of 85 repairs, in which there were 57good results. The type of injury was the mostimportant factor in determining the result; twelveof 13 open, tidy lesions gave good results,compared with 30 of 48 closed traction lesions.Results were better when the nerve was repairedwithin 14 days of injury and when grafts wereless than 120 cm long. They were worse in thepresence of an associated arterial or bony injury.Four of the 16 repairs performed after 180 daysor more failed (Table 10).

Median and ulnar nerves

The effects of age and level of injury are partic-ularly evident for these two nerves. Of equalimportance, if not even more so, is the effect ofthe cause of injury and the delay between injuryand repair. Cavanagh et al (1987) described theoutcome in 74 adults with complex infraclavicu-lar injury in which several nerve trunks were tornor ruptured by severe traction or by missiles.There was rupture of the axillary artery in nearlyone-half of cases. Results were significantlybetter in the ‘open’ wounds, and were very poorindeed in those cases where there was a ‘doublelevel’ lesion above and below the clavicle.

Results for the median and ulnar nerves weredistinctly worse than for the other nerve trunksstudied. This was particularly evident whenresults were related to the interval betweeninjury and repair. The results after repair of ahigh median and ulnar nerve lesion in the adult

102 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 11 Results related to interval between injury and repair (from Surgical Disorders of the Peripheral Nerves(Bonney 1998))

Lesion Good Fair Poor Total

Median and ulnar 14 13 6 33Radial and musculocutaneous 19 10 1 30Suprascapular and circumflex 7 2 2 11Total 40 25 9 74

Late: after 14 days (22 weeks on average)Median and ulnar 1 8 13 22Radial and musculocutaneous 10 10 7 27Suprascapular and circumflex 5 1 1 7Total 16 19 21 56

Table 12 Grading of results in high median and ulnarnerve repair (from Surgical Disorders of the PeripheralNerves (Bonney 1998))

MedianGood Long flexor muscles MRC 4 or better

Localization to digit, withouthypersensitivityReturn of sweating

Fair Long flexor muscles MRC 3 or 3+‘Protective sensation’, moderate or nohypersensitivitySweating diminished or absent

Poor or bad Long flexor muscles MRC 2 or less‘Protective sensation’ but severehypersensitivity or no sensation

UlnarGood FCU and FDP little and ring MRC 4 or

betterIntrinsic muscles MRC 2 or betterLocalization to little and ring fingers, nohypersensitivityReturn of sweating

Fair FCU and FDP little and ring fingers MRC 3or 3+No intrinsic muscle function‘Protective’ sensation little and ring fingersModerate hypersensitivityLittle or no sweating

Poor or bad FCU and FDP little and ring fingers MRC 2No intrinsic muscle function‘Protective’ sensation with severehypersensitivity or no sensationNo sweating

Page 112: Brachial Plexus Injuries

are, on the whole, much more modest than thosefollowing more distal repairs, and so a lessdemanding a system of assessment is used(Tables 11, 12).

Nonetheless, we were able to follow thecourse of recovery in 145 repairs of the medianor ulnar nerves in the axilla or arm, and foundthat useful recovery was seen in more than one-half of cases of the tidy and of the untidywounds. Results were poor or bad in well overone-half of repairs in the closed traction group(Tables 13, 14).

Penetrating missile injuries

Delorme (1915), then Inspector General of theMedical Services of the Armies of France,outlined a method of treatment for shell andbullet wounds based on three principles: resec-tion of the scar until a healthy bed is secured;

excision of damaged nerves until healthy stumpsare reached; and tension-free suturing for graft-ing by adequate immobilization and flexion ofadjacent joints. His paper was heavily criticizedby the great and good, even before the days ofso-called evidence-based medicine, but historyhas proved him right.

The largest published series regarding civiliangunshot injuries comes from Kline (1989), whodescribed 141 wounds to the brachial plexus (90operated cases), and from Kline and Hudson(1995), who described over 150 cases of woundsto the lumbo-sacral plexus, femoral nerve andthe sciatic nerve and its divisions. In nearly one-third of the brachial plexus injuries there wasmajor vascular injury. A total of 125 nerveelements were repaired, and recovery was usefulfor C5, C6 and C7 lesions, for lesions to the upperand middle trunks, and for the posterior andlateral cords.

Stewart and Birch (2001) studied a series of 58patients with penetrating missile injuries.Patients were operated for known or suspectedvascular injury (16 cases), severe persistent pain(35 cases) or complete loss of function in thedistribution of one or more plexus elements.False aneurysms or arteriovenous fistulae wererepaired in 13 cases. Repair of the nerve andvascular lesions abolished or significantlyreduced severe pain in 31 (94 per cent) cases.Fifty-six nerve trunks (36 patients) were grafted;useful results were seen in 35. External neuroly-sis of lesions in continuity produced good oruseful results in 21 (91 per cent) of cases. Avigorous approach is justified in the treatment ofpenetrating missile injury to the brachial plexus,and primary intervention is mandatory wherethere is evidence of a vascular lesion.Worthwhile results can be achieved with earlysecondary intervention in cases with debilitatingpain, failure to progress and deterioration of thelesion under observation. There is cause foroptimism in nerve repair, particularly for the C5,C6 and C7 roots and for the lateral and posteriorcords, but the prognosis for complete lesions ofthe plexus associated with damage to the cervi-cal spinal cord is particularly poor.

One striking finding from this study is thedevastating effect of close-range shotgun injury,in which there is massive energy transfer leadingto gross destruction of the skeleton and of thesoft tissues (Figs 12, 13).

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 103

Table 13 Results in 85 repairs of median nerves in axillaor arm: adults and children (from Surgical Disorders ofthe Peripheral Nerves (Bonney 1998))

Tidy Untidy (incl. Traction Totalpenetrating missile injuries)

Good 6 4 3 13Fair 8 8 15 31Poor or bad 3 12 26 41TOTAL 17 24 44 85

This includes 28 repairs of either lateral or medial root of thenerve in the axilla

Table 14 Results in 60 repairs of medial cord or ulnarnerve in axilla and arm: children and adults (from SurgicalDisorders of the Peripheral Nerves (Bonney 1998))

Tidy Untidy (incl. Traction Totalpenetrating missile injuries)

Good 3 2 0 5Fair 5 13 7 25Poor or bad 2 10 18 30Total 10 25 25 60

All five good results followed repair within 48 hours of injury –two in children. The results of repair 3 months after injury werealways poor in the untidy and traction lesions.

Page 113: Brachial Plexus Injuries

Acknowledgements

We are grateful to Mr George Bonney for permis-sion to refer to his material and also for permis-sion to use a number of the Figures and Tables,which originally appeared in Surgical Disordersof the Peripheral Nerves (Bonney, 1998).

References

Birch R (1998) Compound nerve injuries: traumaticlesions of the brachial plexus. In: Birch R, Bonney G,Wynn Parry CB, eds. Surgical Disorders of thePeripheral Nerves. Churchill Livingstone: Edinburgh:pp. 123–156, 157–208.

104 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 12

The nerve lesion in causalgia. A 23-year-old, shot in the axilla with ahandgun (low energy transfer). Thebullet partially severed the mediannerve; the ulnar nerve wasunscathed. Relief by sympathec-tomy and repair of the nerve.

Figure 13

37-year-old man with a low energy transfer bullet injury. At 2 months he presented with causalgia, cardiac failure andmedian, musculocutaneous and radial nerve palsies. Digital subtraction angiogram confirmed a broad-based arteriovenousfistula.

Page 114: Brachial Plexus Injuries

Birch R, Bonney G, Munshi P, Shergill G (2001) Theradial and posterior interosseous nerves: results in 260repairs, J Bone Joint Surg (in press).

Birch R, Spilsbury J (1996) Some lesions of the circum-flex and suprascapular nerves (abstract), J Bone JointSurg 73B:Suppl. 1, 59.

Bonnard C, Anastakis DJ, van Melle G, Narakas AO(1999) Isolated and combined lesions of the axillarynerve, J Bone Joint Surg 81B:212–18.

Bonney G (1954) The value of axon responses in deter-mining the site of lesion in traction lesions of thebrachial plexus, Brain 77:588–609.

Bonney G (1998) reaction to injury: iatropathic injuryIn: Birch R, Bonney G, Wynn Parry CB, eds. SurgicalDisorders of the Peripheral Nerves. ChurchillLivingstone: Edinburgh: 37–55, 293–334.

Bonney G, Gilliatt RW (1958) Sensory nerve conductionafter traction lesion of the brachial plexus, Proc CollMed 51:365–7.

Bremner-Smith AS, Unwin AJ, Williams WW (1999)Sensory pathways in the spinal accessory nerve, JBone Joint Surg 81B:226–8.

Cavanagh SP, Bonney G, Birch R (1987) The infraclav-icular brachial plexus: the case for primary repair, JBone Joint Surg 69B:489.

Clark WL, Trumble TE, Swiontowski M et al (1992)Nerve tension and blood flow in a model of immedi-ate and delayed repairs, J Hand Surg 17A: 677–87.

Delorme E (1915) The treatment of gunshot wounds ofnerves, Br Med J 1: 853–5.

Dyck PJ, Nukada H, Lais CA, Karnes J (1984)Permanent axotomy: a model of chronic neuronaldegeneration produced by axonal atrophy, myelinremodelling and regeneration. In: Dyck PJ, Thomas PK,Lambert EH, Bunge R, eds. Peripheral Neuropathy W BSaunders: Philadelphia: 660–90.

Fiolle J, Delmas J (1921) The Surgical Exposure of theDeep Seated Blood Vessels (Translated and edited byCG Cumston). W Heinemann: London: 61–7.

Kline DG (1989) Civilian gun shot wounds to thebrachial plexus, J Neurosurg 70:166–74.

Kline DG, Hudson AR (1995) Nerve Injuries. W BSaunders: Philadelphia.

Mikami Y, Nagano A, Ochiai N, Yamamoto S (1997)Results of nerve grafting for injuries of the axillary andsuprascapular nerves, J Bone Joint Surg 79B:527–31.

Narakas AO (1989) Critères d’évaluation des résultatsIn: Alnot J-Y, Narakas AO, eds. Les Paralysies duPlexus Brachial. Expansion Scientifique Française:Paris: 160.

Ochiai N, Nagano A, Mikami Y, Yamamoto S (1997) Fullexposure of the axillary and suprascapular nerves, JBone Joint Surg 79B:532–3.

Osborne A, Birch R, Bonney G, Munshi P (2000) Themusculocutaneous nerve: results of 85 repairs, J BoneJoint Surg 82B:1140–42.

Rank BK, Wakefield AR, Hueston JT (1973) Surgery ofRepair as applied to Hand Injuries, 4th edn. ChurchillLivingstone: Edinburgh.

Seddon HJ (1943) Three types of nerve injury, Brain66:237–88.

Seddon HJ (1975) Surgical Disorders of PeripheralNerves, 2nd edn. Churchill Livingstone: Edinburgh.

Seigel DB, Gelberman RH (1991) Peripheral nerveinjuries associated with fractures and dislocations. In:Gelberman RH, ed. Operative Nerve Repair andReconstruction. J B Lippincott: Philadelphia: 619–33.

Stewart MPM, Birch R (2001) Peneterating missileinjuries of the brachial plexus, J Bone Joint Surg (inpress).

Sunderland S (1951) A classification of peripheralnerve injuries producing loss of function, Brain 74:491–516.

Sunderland S (1978) Nerves and Nerve Injuries, 2ndedn. E&S Livingstone: Edinburgh: 164.

Thomas PK, Holdorff B (1993) Neuropathy due tophysical agents. In: Dyck PJ, Thomas PK, Griffin JW etal, eds. Peripheral Neuropathy, 3rd edn. W B Saunders:Philadelphia.

Tinel J (1917) Nerve Wounds (revised and edited byJoll CA). Ballière Tindall and Cox: London.

Williams WW, Donell ST, Twyman RS, Birch R (1996)The posterior triangle and the painful shoulder: spinalaccessory nerve injury, Ann R Coll Surg Eng 78:521–5.

Zachary RB (1954) Results of nerve sutures, MedicalResearch Council Special Report Series No 282:363.

INJURIES OF THE TERMINAL BRANCHES OF THE BRACHIAL PLEXUS 105

Page 115: Brachial Plexus Injuries

Birch R (1998) Compound nerve injuries: traumaticlesions of the brachial plexus. In: Birch R, Bonney G,Wynn Parry CB, eds. Surgical Disorders of thePeripheral Nerves. Churchill Livingstone: Edinburgh:pp. 123–156, 157–208.

Birch R, Bonney G, Munshi P, Shergill G (2001) Theradial and posterior interosseous nerves: results in 260repairs, J Bone Joint Surg (in press).

Birch R, Spilsbury J (1996) Some lesions of the circum-flex and suprascapular nerves (abstract), J Bone JointSurg 73B:Suppl. 1, 59.

Bonnard C, Anastakis DJ, van Melle G, Narakas AO(1999) Isolated and combined lesions of the axillarynerve, J Bone Joint Surg 81B:212–18.

Bonney G (1954) The value of axon responses in deter-mining the site of lesion in traction lesions of thebrachial plexus, Brain 77:588–609.

Bonney G (1998) reaction to injury: iatropathic injuryIn: Birch R, Bonney G, Wynn Parry CB, eds. SurgicalDisorders of the Peripheral Nerves. ChurchillLivingstone: Edinburgh: 37–55, 293–334.

Bonney G, Gilliatt RW (1958) Sensory nerve conductionafter traction lesion of the brachial plexus, Proc CollMed 51:365–7.

Bremner-Smith AS, Unwin AJ, Williams WW (1999)Sensory pathways in the spinal accessory nerve, JBone Joint Surg 81B:226–8.

Cavanagh SP, Bonney G, Birch R (1987) The infraclav-icular brachial plexus: the case for primary repair, JBone Joint Surg 69B:489.

Clark WL, Trumble TE, Swiontowski M et al (1992)Nerve tension and blood flow in a model of immedi-ate and delayed repairs, J Hand Surg 17A: 677–87.

Delorme E (1915) The treatment of gunshot wounds ofnerves, Br Med J 1: 853–5.

Dyck PJ, Nukada H, Lais CA, Karnes J (1984)Permanent axotomy: a model of chronic neuronaldegeneration produced by axonal atrophy, myelinremodelling and regeneration. In: Dyck PJ, Thomas PK,Lambert EH, Bunge R, eds. Peripheral Neuropathy W BSaunders: Philadelphia: 660–90.

Fiolle J, Delmas J (1921) The Surgical Exposure of theDeep Seated Blood Vessels (Translated and edited byCG Cumston). W Heinemann: London: 61–7.

Kline DG (1989) Civilian gun shot wounds to thebrachial plexus, J Neurosurg 70:166–74.

Kline DG, Hudson AR (1995) Nerve Injuries W BSaunders: Philadelphia.

Mikami Y, Nagano A, Ochiai N, Yamamoto S (1997)Results of nerve grafting for injuries of the axillary andsuprascapular nerves, J Bone Joint Surg 79B:527–31.

Narakas AO (1989) Critères d’évaluation des résultatsIn: Alnot J-Y, Narakas AO, eds. Les Paralysies duPlexus Brachial. Expansion Scientifique Française:Paris: 160.

Ochiai N, Nagano A, Mikami Y, Yamamoto S (1997) Fullexposure of the axillary and suprascapular nerves, JBone Joint Surg 79B:532–3.

Osborne A, Birch R, Bonney G, Munshi P (2000) Themusculocutaneous nerve: results of 85 repairs, J BoneJoint Surg 82B:1140–42.

Rank BK, Wakefield AR, Hueston JT (1973) Surgery ofRepair as applied to Hand Injuries, 4th edn. ChurchillLivingstone: Edinburgh.

Seddon HJ (1943) Three types of nerve injury, Brain66:237–88.

Seddon HJ (1975) Surgical Disorders of PeripheralNerves, 2nd edn. Churchill Livingstone: Edinburgh.

Seigel DB, Gelberman RH (1991) Peripheral nerveinjuries associated with fractures and dislocations. In:Gelberman RH, ed. Operative Nerve Repair andReconstruction. J B Lippincott: Philadelphia: 619–33.

Stewart MPM, Birch R (2001) Peneterating missileinjuries of the brachial plexus, J Bone Joint Surg (inpress).

Sunderland S (1951) A classification of peripheralnerve injuries producing loss of function, Brain 74:491–516.

Sunderland S (1978) Nerves and Nerve Injuries, 2ndedn. E&S Livingstone: Edinburgh: 164.

Thomas PK, Holdorff B (1993) Neuropathy due tophysical agents. In: Dyck PJ, Thomas PK, Griffin JW etal, eds. Peripheral Neuropathy, 3rd edn. W B Saunders:Philadelphia.

Tinel J (1917) Nerve Wounds (revised and edited byJoll CA). Ballière Tindall and Cox: London.

Williams WW, Donell ST, Twyman RS, Birch R (1996)The posterior triangle and the painful shoulder: spinalaccessory nerve injury, Ann R Coll Surg Eng 78:521–5.

Zachary RB (1954) Results of nerve sutures, MedicalResearch Council Special Report Series 282:363.

106 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 116: Brachial Plexus Injuries

Introduction

The treatment of brachial plexus palsies hasimproved greatly over the last quarter of a centurydue to better knowledge of the lesions themselvesand of the pathophysiology of healing of theperipheral nerves, as well as improvements insurgical techniques and the new possibilities ofneurotizing the paralysed muscles. However, in alarge number of cases results are still unsatisfac-tory, especially regarding the shoulder.

According to the most recent theories, this ispartially due to the fact that, when the surgeondecides how to distribute the limited nervous fibressurviving after trauma, the hand is the prioritybecause its function is certainly more important. Asa consequence, in many cases the shoulderremains paralysed in spite of the result of surgeryat the elbow and hand because of the palsy of themuscles from the scapula to the humerus.

Complete shoulder palsy is a severe functionalhandicap. However, the muscles moving theshoulder upon the thorax are often preservedand the scapula can be displaced in all direc-tions. An operation blocking the humerus to thescapula therefore allows movement of the armby the action of the thoracoscapular muscles.This operation is called arthrodesis, andprevents any movement of the joint, transform-ing the useless joint into a rigid lever (AOARC1942, Botteri 1960, Brittain 1952, Charnley 1966,Cofield 1979, Delitala 1940, Hawkins 1987,Johnson 1986, Nagano 1989, Richards 1993, Ron1991, Rowe 1974, Scaglietti 1933, Vidal 1987).

Indications

Scapulo-humeral arthrodesis can give extremelysatisfactory results provided that hand function

and elbow flexion are good. These are prerequi-sites. That said, in few cases the author hasperformed contemporary arthrodesis at theelbow and wrist or opposition osteodesis of thefirst metacarpal when the function of the shoul-der arthrodesis alone would have been useless.In these cases, multiple arthrodeses suppliedrudimentary but useful movements.

Results of arthrodesis are more consistent andefficient than any muscular recovery due tomuscular tendinous transfer; the result is perma-nent whereas transferred muscles may deterio-rate with the time.

Of course, shoulder fusion is an admission of thesurgeon’s incapacity to restore normal movement.Sacrificing a joint is also repellent to the surgeon,but it is a matter of ‘propitiatory sacrifice’ to obtainactive movement of the paralysed glenohumeraljoint. Palagi, quoted by Logroscino (1960), cleverlysaid that arthrodesis at the shoulder ‘creates a newmovement of the arm, whilst also satisfying somecosmetic requirements (Figs. 1 and 2).

Shoulder arthrodesis was performed first byAlbert in 1879 and then by Vulpius in 1898(Mezzari 1960). The operation was later taken upby many other authors, including Albee, Putti,Delitala and Straub. Even so this operation wasinitially harshly criticized by many other authorsbecause of the poor results due to inaccurateindications, and lack of detailed advice regardingthe position of the humerus and the type oftechniques to be performed.

At the present time shoulder fusion in upperarm palsies receives general support providedthat indications are correct. Moreover, resultsimprove over the years due to a series of tricksand functional adaptations by the patient. Eventhe results obtained by means of other opera-tions on the hand improve, because they can bebetter exploited thanks to the improved dynamicpositioning of the hand itself.

11The place of arthrodesisGiorgio A Brunelli

Page 117: Brachial Plexus Injuries

Indications for shoulder arthrodesis are funda-mentally the palsies of abduction of the upperlimb (trapezium and supraspinatus).

Prerequisites are:

• Good function of the thoraco-scapularmuscles; the trapezium, levator scapulae,rhomboids and, above all, the serratusanterior (partial impairment of one of thesemuscles can be tolerated, but it must beremembered that strength is necessary toraise a whole upper limb). The function of thepectoralis major muscle must also be takeninto consideration, not because its strengthmay prevent abduction (in fact the pectoralismajor muscle is an antagonist), but becauseits presence is essential for the thoraco-brachial pliers.

• Good distal function of the arm. It is uselessto perform a shoulder arthrodesis if the handis completely paralysed.

• Integrity of the joints of the scapular belt; theacromioclavicular, sternoclavicular and scapu-

lothoracic joints. Stiffness of these jointsfollowing joint fracture, or post-fractureadhesions, reduce or annul the results. It isalso important to avoid any involvement ofthe acromioclavicular joint during surgery.

It must be remembered that those joints whichtake the place of the glenohumeral joint, mustsustain extra work and will therefore undergodeterioration and arthritis with the years. Also,the muscles suffer extra stress, which may leadto hypertrophy or exhaustion and atrophy.

Lack of distal function is not an absolutecontraindication to shoulder fusion. As previ-ously stated, in very severe cases (mainly bilat-eral and especially in polio) the author hasassociated shoulder fusion with arthrodesis ofthe elbow or wrist in a functional position, andsometimes with opposition of the firstmetacarpal. This serial arthrodesis gaverudimentary function.

Arthrodeses can be classified as intra-articular,extra-articular or mixed. They can be simple or

108 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

Various types of arthrode-sis:A: VulpiusB: AlbeeC: Watson JonesD: ZancolliE: LangeF: First type of personal

arthrodesisG: Present type of

arthrodesis with decor-tication, plate bent at100° and addition ofspongious bone (longarrows).

Page 118: Brachial Plexus Injuries

modelling, i.e. performed in such a way as togive a particularly favourable position. They canbe done by decortication alone, or with theaddition of bone grafts. Numerous types ofarthrodesis have been suggested, some of whichare shown in Figure 1. The position for arthrode-sis has been the subject of much discussion. Inshoulder fusion, we must obtain an abductionsufficient to position the hand in the workingplane but not so great as to prevent efficientactive adduction for the thoraco-brachial pliers,which is very important. There must be enoughstrength to hold an object under the arm. Toogreat an abduction would cause muscle fatiqueand prevent the thoraco-brachial pliers.

The author therefore considers that the armshould be placed in 30° of abduction from thethorax (Fig. 3). In this position, movement of thescapula would allow active abduction of about60° whilst still allowing the contact with thethorax (Fig. 4). After surgery, X-rays should showan angle of about 65° between the lateral borderof the scapula and the medial cortex of thehumerus.

The arm should be placed in 45° of anteposi-tion. Movement of the scapula will in time allowan anteposition of 65–75° as well as a restingposition almost vertical. The rotation should begraduated so that, by the residual movements ofthe scapula, the hand can touch the abdomen,reach the mouth with a fork and be put in apocket (intra-rotation of 40°).

THE PLACE OF ARTHRODESIS 109

Figure 2

X-rays of an ancient shoulder arthrodesis fixed by twoscrews.

Figure 3

By positioning the arm at 30° of abduction, the activemovement of the scapula allows an abduction of 60°.

Figure 4

The limited abduction allows a strong grasp which is veryimportant, especially if the hand is not valid.

Page 119: Brachial Plexus Injuries

In reality, the ideal position for fusion dependsupon the individual functional goals, the globalneurological deficit, and possible elbow or wristdeformities (Ducloyer et al 1991).

Surgical techniques

Surgical techniques are numerous and vary intheir approach, in the type of arthrodesis itself,the type of immobilization and the type of bonegraft (autologous) etc. Here, the authorspreferred technique is described. This consists ofa reversed V-shaped approach on the posterioraspect of the shoulder, following the spine of thescapula throughout its length and then continu-ing on the lateral aspect of the deltoid region for10 cm. This approach is not considered too largeas it has to allow the positioning of a plate bentat 100° and of sufficient length for the number ofscrews required for solid and stable fixation.

After retraction of the skin edges, the paral-ysed deltoid muscle is severed (as if a militaryepaulette), detaching it totally from the spine, theacromion and the clavicle. This allows a clearview of the joint. This region is very rich invessels, and therefore the procedure should beperformed very carefully with progressivehaemostasis. The capsule is then severed andthe head of the humerus subluxed. Decorticationof both the humeral head and the glenoid isperformed by means of a large hollow chisel,until bleeding spongious bone is achieved. Greatcare must be taken not to alter the congruity ofthe bony heads. The inferior aspect of theacromion is also decorticated, taking care tospare the acromioclavicular joint.

At this point, the head of the humerus ispushed into the cavity of the glenoid and againstthe acromion. The arm is placed in 45° of antepo-sition, 30° of abduction and 40° of internalrotation. This position is temporarily fixed bymeans of two screws, one from the humerus tothe glenoid and one from the acromion to thehumeral head. All authors have stressed the diffi-culty of estimating the correct position of thearthrodesis during the operation.

A large pedicled bonegraft may now beelevated from the acromion or from the spine ofthe scapula, by means of progressive osteotomestrikes from medial to lateral. This chip is

inserted into the humeral head, leaving a ‘bridge’in continuity with the scapula.

While an assistant maintains the arm inposition (held by the two screws), a 10- or 12-hole plate is modelled to fit above the spine ofthe scapula, the acromion and the lateral aspectof the humerus, with an approximate angle of100°. The plate is then screwed according to theroutine technique.

In more recent years, to avoid possible non-consolidation, the author performed spongiousautologous bone grafts (Fig. 5). The bone istaken (by means of a large curette) from theproximal meta-epiphysis of the tibia, and is thencrushed using a pestle and mortar. Using thethumb, the resultant pulp is pushed into all thegaps of the arthrodesis like a ‘stucco’, accordingto the technique described 35 years ago (Brunelli1963, 1972). Careful haemostasis follows. A drainis introduced for 2 days and the tissues arereconstructed. Skin sutures and an immobilizingcast follow (the cast should be prepared 48 hoursin advance and opened for the operation).

Although techniques vary, it is important tofollow five fundamental rules:

1. Thorough decortication;2. Careful alignment of the bone surfaces;3. Addition of cancellous bone;

110 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 5

X-rays of a recent arthrodesis fixed by means of plate,screws and cancellous bone addition (arrow).

Page 120: Brachial Plexus Injuries

4. Solid fixation;5. Immobilization until there is complete fusion.

Regarding timing of the operation, it is impor-tant:

1. To wait until nerve regrowth up to themuscles has occurred;

2. To first perform all the other operationsrequired for the hand and elbow, since itwould be very difficult to perform them aftershoulder fusion and might give poor results.

In general, patients prefer the safe stiffness ofarthrodesis to the pre-operative instability. If thefunction of the hand and elbow are good, theresult of shoulder fusion may be excellent.However, even with poor function of the handthe result greatly improves daily activitiesbecause the paralysed arm becomes an aid tothe contralateral one, it may serve as a paper-weight, achieve a thoraco-brachial pliers and beused raising light objects.

Evaluation of results is shown in Table 1.Since 1963 the author has performed 58 shoul-

der arthrodeses, 20 of which have been lost tofollow-up. Most of these 20 were cases of polioand were operated in the 1960s (Fig. 6).

• Thirty-eight cases have been followed up; ofthese, 23 male and eight female, average age23 years (range 8–65) were reviewed. Twenty-one were scored ‘very good’, 13 ‘good’, three‘fair’ and one ‘poor’;

• Twenty-five cases had obtained a pretty goodrecovery of both the function of the hand andflexion of the elbow following nerve surgery;

• Three cases were sequelae of polio. In threecases (elderly people), arthrodesis had beenperformed primarily;

• In the first nine cases, fixation was obtainedby screw only. In nine more cases, a pedicled

bone graft from the spine of the scapula wasadded;

• In the last 20 cases, arthrodesis wasperformed by decortication, addition ofpulped cancellous bone and fixation by a longplate and screws;

• External fixation was never used;• Healing was obtained on average in 41⁄2

months (range 3–7 months).

Regarding complications:

• There was one case of non-consolidation at 7months due to abandoning immobilization tooearly (45 days). It was cured by a secondaryoperation where a cortical spongious wedgegraft was inserted into the anterior opening ofthe joint;

• There was no pseudoarthrosis (which is oneof the commonest complications in otherauthors’ series). This was probably due to theprimary addition of cancellous bone;

• There were no fractures of the humerus,another common complication reported inliterature (Chamas et al 1995). This wasprobably due to the modest degree of abduc-tion of the humerus;

• There was one delayed consolidation, whichhealed at 7 months;

• In two cases the plate gave decubitus of theskin and further surgery was required tomobilize the surrounding skin for a bettercoveraged;

• In three cases, the plate had to be removedfollowing bone consolidation because itprotruded under the skin at the acromionregion;

• In two more cases loosening screws had to beremoved.

All the patients were able to reach their mouths;very easily if elbow flexion was good, or by

THE PLACE OF ARTHRODESIS 111

Table 1 Evaluation of the results of shoulder fusion

Abduction (active) Fork to the mouth Thoraco-brachial pliers Manual work Daily activity

Very good � 60° Yes Yes Yes YesGood 45–60° Long fork Yes ± YesFair 30–45° Very long fork Weak No ±Poor ≤30° No No No No

Page 121: Brachial Plexus Injuries

means of long spoons or forks if flexion waspoor (Fig. 6). The position in internal rotationobviously gives a lack of external rotation,making it difficult or impossible for patients toreach the back of their neck, but it is very usefulto reach the abdomen and the side pocket. Therigid position of the shoulder may cause sometrouble sleeping. Almost all the patients wereunable to sleep on their fused shoulder. Youngpeople were able to resume sporting activitiesthat did not involve the upper arm.

From the subjective point of view, all thepatients were satisfied.

Conclusion

Shoulder arthrodesis gives a predictable result.This result is certainly partial but improves thefunction of the upper limb greatly. The strengthof movement of the arthrodesed shoulder iscertainly better than that obtained by musculartransfers. The range of movement of the armdepends on the pre-operative condition of thescapulothoracic muscles, of which the serratusanterior is the most important. Of course, thefunction of the limb after shoulder arthrodesisdepends upon the function of the elbow and the

hand. However, even in an almost flail limbmultiple arthrodeses can give a very rudimentaryfunction, which is much appreciated by patients.In the general planning of surgery, foreseeing ashoulder arthrodesis for a paralysed upper limballows the surgeon to distribute the nervouselements to the nerves for the elbow and hand,so obtaining a much better result.

References

American Orthopedic Association Research Committee(1942) Survey and results on stabilization of paralyticshoulder, J Bone Joint Surg 24:699.

Botteri G (1960) Arthrodesis. In: Mezzari A ed. LaPoliomielite. Idelson: Napoli: 197–9.

Brittain H (1952) Architectural principles in Arthrodesis.Livingston: London.

Brunelli GA (1963) L’innesto d’osso spongiosomalleabile nelle pseudartrosi di tibia. CongressoCROMA, San Benedetto.

Brunelli GA (1972) Soft cancellous bone grafts for non-union and joint fusion. International Congress Series291 Orthopaedic Surgery and Traumatology. ExcerptaMedica: Amsterdam: 952–4.

Brunelli GA (1995) L’arthrodèse de l’épaule.Symposium on Extensive Paralysis of the UpperExtremity. Soc Française de la Main: Paris.

Chamas M, Allieu Y, Meyer G (1995) L’arthrodèse del’épaule: indications – resultats. In: Alnot JY, NarakasA, eds. Les Paralysies du Plexus Brachial, 2nd edn.Expansion Scientifique Française: Paris: 231–8.

Charnley J, Houston JK (1966) Compression arthrode-sis of the shoulder, J Bone Joint Surg 46B:614–20.

Cofield RH, Briggs BT (1979) Glenohumeral arthrode-sis. Operative and long-term functional results, J BoneJoint Surg 61A: 668–77.

Delitala F (1940) I movimenti del braccio nelle anchilosidell’articolazione scapolo omerale e nelle paralisi deldeltoide. Contributo alla fisio-patologia articolare, ChirOrg Mov 26:230–43.

Ducloyer Ph, Nizard R, Witvoet J (1991) Arthrodèsesd’épaule dans les paralysies du plexus brachial, RevChir Orthop 77:396–405.

112 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 6

Demonstration of the movement allowed by arthrodesis.This patient also had a clavicle non-union, which had beenpreviously cured by cancellous bone graft and plate.

Page 122: Brachial Plexus Injuries

Hawkins RJ, Neer CS (1987) A functional analysis ofshoulder fusions, Clin Orthop 223:65–76.

Johnson CA, Healy WL, Brooker AFJ et al (1986)External fixation shoulder arthrodesis, Clin Orthop211:219–23.

Logroscino D (1960) La spalla paralitica. In: Mezzari A,ed. La Poliomielite. Idelson: Napoli: 334–53.

Nagano A, Okinaga S, Ochiai N et al (1989) Shoulderarthrodesis by external fixation, Clin Orthop247:97–100.

Richards RR, Beaton D, Hudson AR (1993) Shoulderarthrodesis plate fixation: functional outcome analysis,J Shoulder Elbow Surg 2:225–39.

Ron A, Birch R (1991) Traitements chirurgicaux palliat-ifs de l’ épaule paralysée après lésion du plexusbrachial. In: Tubiana A, ed. Traité de chirurgie de laMain. Masson: Paris: 155–61.

Rowe CR (1974) Re-evaluation of the position of thearm in arthrodesis of the shoulder in the adult, J BoneJoint Surg 56A:913–22.

Scaglietti O (1933) Dell’artrodesi mista, intra-articolareed extra-articolare di spalla negli esiti di poliomielitedell’arto superiore, Chir Org Mov XVII:609–15.

Vidal J, Nekach G, Rabishong P et al (1987) Arthrodèsescapulo-humérale par fixation combinée externe etinterne, Rev Chir Orthop 73:171–7.

THE PLACE OF ARTHRODESIS 113

Page 123: Brachial Plexus Injuries
Page 124: Brachial Plexus Injuries

Introduction

An estimated 25–30 per cent of the nerve fibreswithin the brachial plexus pass to the shouldergirdle, with its large number of muscles andcomplex shoulder movements (Birch 1995, Allieu1999). Movement about the shoulder requiresthe coordinated moment of four joints and over20 muscles, and the shoulder and scapularmuscles and their innervations are shown inTables 1 and 2 (Kendall et al 1993).

Thirty-three per cent of shoulder elevation isfrom the scapulothoracic joint. Saha believedthat good elevation requires prime movers(deltoid or clavicular head of the pectoralismajor), a steering group (supraspinatus,infraspinatus and subscapularis) and a depressorgroup (pectoral head of pectoralis major, latis-simus dorsi, teres major and teres minor) (Karev1986, Price and Grossman 1995, Kotwall et al1998). Saha confirmed that when any two of thesteering group of muscles were paralysed, asingle muscle transfer to replace the deltoidwould not provide abduction beyond 90°(Kotwall et al 1998). Therefore, in paralysis of theprime movers the importance of the steeringgroup of muscles must be considered whenperforming a tendon transfer to restore shoulderabduction: with paralysis of all the steeringgroup, transfer of one muscle (such as thetrapezius into the deltoid attachment) permitsarm lifting only to 90° and results in the loss ofsynchronous scapulohumeral rhythm. If thepatient has one of the steering group muscles,the subscapularis, this avoids the necessity forone additional tendon transfer to restore theirfunction (Karev 1986). Abduction of the shoulderto 90° is provided by the supraspinatus anddeltoid; for abduction greater than this at theglenohumeral joint, the main external rotators ofthe scapula are serratus anterior and trapezius,while other muscles are involved in rotating andstabilizing the humeral head.

Paralytic shoulder may be caused by:

1. Supraclavicular and infraclavicular injuries;2. Nerve injuries of the terminal branches of the

brachial plexus (axillary, suprascapular andmusculocutaneous nerves);

12Palliative surgery: tendon transfers tothe shoulder in adultsAydın Yücetürk

Table 2 Scapular muscles (Kendall et al 1993)

Abductor – full flexion: serratus anterior (C5-6-7-8)Lateral rotators: serratus anterior (C5-6-7-8), trapezius(nerve XI – accessory) and ventral ramus (C2-3-4)Adductor – full abduction: trapezius (nerve XI –accessory) and ventral ramus (C2-3-4)Lateral rotators: trapezius (nerve XI – accessory) andventral ramus C2-3-4), serratus anterior (C5-6-7-8)Adductors, medial rotators and elevators – full extension:rhomboids (C4-5), levator scapula (C3-4-5)Ant. tilt of scapula by pectoralis minor (C6-7-8-T1)Adductors – to side against resistance: rhomboids (C4-5);trapezius (nerve XI – accessory) and ventral ramus (C2-3-4)

Table 1 Shoulder muscles (Kendall et al 1993)

Flexors: anterior deltoid (C5-6), biceps (C5-6), pectoralismajor upper (C5-6-7), coracobrachialis (C6-7)Abductors: deltoid (C5-6), supraspinatus (C5-6), bicepslong head (C5-6)Lateral rotators: infraspinatus (C5-6), teres minor (C5-6),posterior deltoid (C5-6)Extensors: posterior deltoid (C5-6), teres major (C5-6-7),latissimus dorsi (C6-7-8), triceps long head (C6-7-8-T1)Adductors: pectoralis major (C5-6-7), teres major (C5-6-7),latissimus dorsi (C6-7-8), triceps long head (C6-7-8-T1)

Page 125: Brachial Plexus Injuries

3. Lesions of the nerves to the shoulder – theaccessory nerve (trapezius) and long thoracicnerve (serratus anterior) (Alnot 1996).

For the best functional recovery, both the axillaryand the suprascapular nerves must be repairedsurgically (Alnot 1996). It is important to remem-ber that these two nerves can be crushed twice intrauma; besides damage at Erb’s point, the supras-capular nerve can also be injured in the scapularnotch and the axillary nerve in the axillary fossa(Fig. 1). Reinnervation of the pectoralis major isalso important. If just the suprascapular nerve isrepaired, no more than 30° of shoulder abductionand flexion can be expected.

In complete shoulder palsy (active trapeziusand serratus anterior muscles) shoulder instabil-ity is obvious, with inferior subluxation and theweight of the upper limb causing pain anddisability (Alnot 1996) (Fig. 2).

Axillary ± suprascapular nervepalsy

Over the past 30 years, the efficacy of brachialplexus surgery has been established for thetreatment of certain injuries. Although some

muscle function returns spontaneously or follow-ing surgery, certain muscles (most notably thedeltoid) may remain paralysed. In tractioninjuries the axillary nerve is that most commonlydamaged, followed in order of frequency by themusculocutaneus nerve and the suprascapularnerve. Because the axillary nerve is composedmostly of motor fibers and travels only a shortdistance from its origin to its muscle insertion,there is a good prognosis for full functionalrecovery after surgical repair. This nerve is rarelycompletely severed but is usually stretched overseveral centimetres. Stretch injuries of the infra-clavicular brachial plexus have a better progno-sis for spontaneous recovery than suprascapularinjuries (Fiedman et al 1990). Suprascapularnerve repair is difficult because of the route ofthe nerve, and because it may also be avulsedfrom the muscle (Alnot 1996).

Secondary shoulder surgery followingtraumatic brachial plexus injuries includes:

1. Ligamentous plasty – acromial insertion of thecoracoacromial ligament transfer onto thelesser tubercule;

2. Arthrodesis;3. Rotational osteotomy;4. Muscle transfer, which may include any

combination of the following:

116 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

Brachial plexus injury and fractured scapula. Axillary andsuprascapular nerve possible double crush injury.

Figure 2

Flail shoulder and multidirectional instability.

Page 126: Brachial Plexus Injuries

• Transfers for deltoid and supraspinatusparalysis – trapezius (if spinal accessorynerve was not done before) (Alnot 1996);pectoralis major and teres major; latis-simus dorsi; gracilis; combined biceps andtriceps;

• Other combinations of teres major, latis-simus dorsi, levator scapula, biceps,coracobrachialis and pectoralis major(Ruhmann et al 1998).

Alnot’s surgical aproach according to shouldermuscle palsies is shown in Table 3.

Ligamentous plasty

Ligamentous plasty is not effective for shoulderfunction, but can be used for shoulder subluxa-tion.

Arthrodesis

According to Allieu, neurotization of the muscu-locutaneous nerve by the intercostal nervewithout interpositional graft and shoulderarthrodesis gives the best and most constantfunctional results for shoulder and elbow (Allieu1999). According to Kline, substitutive proce-dures for loss of shoulder abduction with proxi-mal brachial plexus injuries remain controversial(Kline and Hudson 1995).

Fusion of the glenohumeral joint is probablythe procedure of choice when the plexus injuryspares hand function. The advantage of a shoul-der arthrodesis is the greater increase in power-ful active function, and this must be selected forthose patients who have physically demandingwork and still retain good function in the elbowand hand (Ruhmann et al 1999). The transfer ofa portion of the lateral trapezius to provide activeabduction of the shoulder is a less certain proce-dure, and it can be difficult to achieve a balancedand therefore usable transfer with this operation(Kline and Hudson 1995).

Disadvantages of shoulder fusion includetechnical difficulties, long rehabilitation, compli-cations, the loss of passive mobility, and theirreversibility of the procedure (Aziz et al 1990).

Saha argues against arthrodesis after poliomyeli-tis and points out that the fulcrum is moved tothe scapulothoracic joint, which gives a muchlonger lever arm for the thoracic muscles(Kotwell et al 1998). Cofield and Briggs point outthat the disadvantages of arthrodesis include ahigh incidence of fracture, worsening of pain,and relative reduction of passive movements(Kotwell et al 1998).

In summary, arthrodesis permits only a limitedrange of abduction and has a high incidence ofcomplications. It inhibits the passive mobility ofthe joint making some daily activities such asdressing, putting the hand in the pocket andpulling a zipper difficult (Mir-Bullo et al 1998). It isirreversible, and gives less active and passivemovement than trapezius transfer (Aziz et al 1990).

Rotational osteotomy

Derotation osteotomy of the humeral shaft canbe performed, and enhances recovery of elbowflexion. Alnot advises rotational osteotomy whenboth the deltoid and the infraspinatus are palsied(Alnot 1996) (Table 3).

Muscle transfer

Muscular transfers of the latissimus dorsi andteres major are not satisfactory in adults,

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN ADULTS 117

Table 3 Alnot’s surgical aproach according to shouldermuscle palsies

Only deltoid muscle palsy:Trapezius transfer with acromial portion to deltoidinsertion (Bateman’s procedure) or transfer of the longhead of the triceps to acromion (Sloaman)

Both deltoid and infraspinatus palsy:Derotation osteotomy of the humeral shaft for externalrotation

Deltoid, infraspinatus and supraspinatus palsy:Stabilization by long head of biceps muscleDerotation osteotomyShoulder arthrodesisTrapezius transfer

Page 127: Brachial Plexus Injuries

although they are very effective in children(Alnot 1996). Different muscle transfers havebeen used (Table 4); trapezius transfer was usedfor deltoid paralysis by Hoffa in 1891, Lewis in1910, Lange in 1911 and Mayer in 1927(Ruhmann et al 1999). Today, trapezius transferis commonly used for deltoid palsy (Karev 1986,Aziz et al 1990, Kotwell et al 1998, Mir-Bullo et al1998, Ruhmann et al 1998, 1999).

Indications for trapezius transfer include:

• Failure of nerve repair;• Late brachial plexus injuries;• Paralysis of deltoid and supraspinatus clini-

cally and by EMG;• Trapezius strength of at least M4, ideally M5;• A normal glenohumeral joint;• At least 80° of passive abduction;• An intact rotator cuff.

Contraindications include:

• Trapezius strength of less than M4;• Glenohumeral joint problems, including

degenerative arthritis, septic arthritis seque-lae, old unreduced shoulder dislocation, andmalunited humeral head or glenoid fractures;

• Workers with physically demanding jobs.

Relative contraindications are: complete rotatorcuff tears, spinal accessory nerve neurotizationto the musculocutaneous nerve (even partial),and no elbow flexion.

Difficulties may include:

• A preoperative stiff shoulder that needsphysiotherapy;

• Scars from previous nerve repair;

• Humeral fractures;• Humerus osteomyelitis following an open

fracture;• Acromioclavicular dislocation;• Severe osteoporosis.

Complications may include: infection, looseningof screws, non-union of the acromial bonefragment, and nerve injuries.

Trapezius muscle transfer has several advan-tages (Aziz et al 1990, Ruhmann et al 1999):

1. It reduces shoulder subluxation and disloca-tion;

2. It gives a functional improvement;3. It eliminates pain;4. Muscle transfer operations are more success-

ful in terms of passive function and havelower complication rates;

5. Surgery is relatively short.

According to Aziz et al, the absence of clearindications for trapezius transfer and expectingtoo much from this transfer alone have led to itsinfrequent use, and the only contraindication isadvanced degeneration of the shoulder.

Saha’s modification of the trapezius transfer,described by Bateman, provides a more distalfixation of the transfer after a more proximalrelease. Distal fixation has the following advan-tages: it gives a longer lever arm; the bony inser-tion transfered from the acromium allows betterfixation; it allows transfer for paralysed musclesof the rotator cuff; it improves control of thehumeral head; and it prevents subluxation(Kotwell et al 1998).

Even when functional recovery is incomplete,trapezius transfer is strong enough to keep theshoulder stable and allow some active abduc-tion, while allowing a full passive range (Aziz etal 1990).

Trapezius transfer can be combined withother transfers to achieve maximum use of thearm, e.g. latissimus dorsi transfer to the elbow,Steindler’s elbow flexorplasty, and tendontransfer to the wrist or wrist fusion (Aziz et al1990).

The acromion should be transferred to belowthe greater tuberosity of the humerus and fixedwith screws (Figs 3 and 4). The point of fixationto the humerus is a decisive factor regarding theextent of postoperative function, especially

118 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 4 Muscle transfers to shoulder

Pectoralis major transfer Hildebrandt (1906)Trapezius muscle transfer Hoffa (1902)

Bateman (with acromium andspina scapula bone en bloc)

Long head of triceps to Sloamanacromium

Short head of biceps Ober, Hass, Davidson,Harmon

Latissimus dorsi Shultze-BergeL’Episcopo

Page 128: Brachial Plexus Injuries

abduction and forward flexion. Sometimestrapezius transfer distalization is limited by therestricted degree of mobilization. Multidirectionalshoulder instability is reduced by the procedure(Ruhmann et al 1999) (Fig. 2).

For successful trapezius transfer, Karev (1986)points that the muscle must be at least M4,

there must be excellent passive shouldermotion, the direction of the muscle must bestraight, and tension must be adequate. Aftertrapezius transfer with the acromion intraumatic patients, Ruhmann et al (1998)achieved a satisfactory improvement in stabilityand function (Fig. 5).

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN ADULTS 119

Figure 3

Trapezius transfer with acromium bone block.

Figure 4

Trapezius transfer and acromium bone block fixation tohumerus.

Figure 5

(a) Preoperative shoulder abduction, (b) Postoperative 45° of abduction after trapezius transfer.

a b

Page 129: Brachial Plexus Injuries

Long thoracic nerve palsy

Seventeen muscles are attached to or take theirorigin from the scapula, the most important ofwhich are the serratus anterior, trapezius,rhomboids and levator scapula. These muscleshelp the scapula to remain in position during thefull range of motion of the shoulder.

The normal function of the serratus anterior isto maintain the scapula in apposition to thethorax when the arm is elevated forward at theshoulder. The trapezius and rhomboid musclescontrol scapular rotation and provide a mobileyet stable base for moment of the upper extrem-ity. Paralysis of the serratus anterior musclecauses the scapula to rotate posteriorly on itsvertical axis, producing the characteristic appear-ance of ‘winging’. Spontaneous recovery of thelong thoracic nerve occurs within 1 year, onaverage, in 70–80 per cent of cases (Alnot 1996).

Paralysis of the serratus anterior muscle canbe functionally disabling (Perlmutter and Leffert1999) and may result from a variety of causes,including:

• Acute traumatic and traction injuries;• Viral illnesses;• Brachial neuritis;

• Iatrogenic injuries (transaxillary resection ofthe first rib, cervical dicectomy and arthrode-sis, mastectomy and axillary nodes dissection,Bankart procedure, acromioclavicular jointreconstruction etc.) (Alnot 1996, Perlmutterand Leffert 1999).

Indications for pectoralis major transfer include:

• Marked scapular winging;• Loss of shoulder strength;• Difficulty with the activities of daily living;• Electrodiagnostic evidence of chronic dener-

vation of the serratus anterior present for12–18 months (with the exception of theParsonage–Turner syndrome – brachial neuri-tis) (Perlmutte and Leffert 1999).

Electrical evidence of long thoracic nerve injuryis required to confirm that scapular winging isbeing caused by serratus anterior dysfunction(Warner and Navarro 1998).

Long thoracic nerve recovery may take asmuch as 2 years, but patients with severesymptoms who have undergone 12 months ofconservative treatment may benefit from surgerytreatment (Winter and Flatow 1999). Severalprocedures have been reported in the literature,

120 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 6

(a) Scapular winging, (b) After pectoralis major transfer to the medial border of scapula.

a b

Page 130: Brachial Plexus Injuries

including transplantation of the pectoralis majormuscle into the serratus itself, fascial slingsuspension, pectoralis minor transfer, rhomboidtransfers, scapulo-thoracic articulation arthrode-sis, and pectoralis major transfer with fascia lataautogenous graft to the scapula medial border.Pectoralis major tendon transfer is an effectivetreatment for scapular winging (Fig. 6). Thissurgical approach sometimes gives unacceptablecosmosis, and there may be local morbidity tothe donor site of the iliotibial band graft that isused to augment the tendon transfer (Warnerand Navarro 1998).

Povacz and Resch (2000) stabilized the scapulaby split pectoralis major tendon transfer directlyattached to the inferior angle of the scapula, andautogenous graft complications were removed.Medial and lateral pectoral nerve injury resultingin recurrent scapular winging after pectoralismajor transfer was reported by Litts et al (2000).Scapulo-thoracic arthrodesis reduces thesymptoms, but decreases the motion of theshoulder (Perlmutter and Leffert 1999).

Spinal accessory nerve palsy

Trapezius paralysis may be seen after biopsy ofthe cervical nodes, trauma, or radical dissectionin the neck. Spinal accessory nerve paralysissymptoms include pain, visible deformity anddysfunction of the shoulder girdle. The functionaldisability is compounded by discomfort, whichmay be secondary to traction on the brachialplexus, pericapsular muscle spasm, frozen shoul-der, subacromial impingement, or acromio-calvicular synovitis. Thoracic outlet syndromemay also develop (Bigliani et al 1996).

Physical examination findings includeasymmetry of the neckline, dropping of theshoulder girdle with lateral displacement of thescapula, and weakness of active elevation.

The spinal accessory nerve provides the solemotor innervation to the trapezius muscle, whilebranches of the third and fourth cervical nervesprovide proprioceptive function. Reconstructiveprocedures to substitute for a paralysed trapez-ius include stabilization of the scapula to thespinal processes of the thoracic vertebra with thefascia lata and/or transfer of the levator scapula,and transfer of the levator scapula and rhomboid

major and minor muscles together (Eden–Langeprocedure). This combined transfer is a reason-able salvage procedure for a patient who haspain, deformity and diminished function of theshoulder girdle caused by irreparable injury ofthe spinal accessory nerve (Bigliani et al 1996).

References

Allieu Y (1999) Evolution of our indications for neuriti-zation. Our concept of functional restoration of theupper limb after brachial plexus injuries, Ann Chir Main166–7.

Alnot JY (1996) Brachial plexus palsies: palliative surgery.In: Alnot, Anarakas, eds. Traumatic Brachial PlexusInjuries. Expansion Scientifique Française: Paris: 218–20.

Aziz W, Singer RM, Wolfe TW (1990) Transfer of thetrapezius for flail shoulder after brachial plexus injury,J Bone Joint Surg 72B(4): 701–4.

Bigliani LU, Compito CA, Duralde XA, Wolfe IN (1996)Transfer of the levator scapula, rhomboid major,andrhomboid minor for paralysis of the trapezius, J BoneJoint Surg 78A:1534–40.

Birch R (1995) Lesions of the upper brachial plexus:C5/6 and C5/6/7 injury. In: Wastamaki M, Jalovaara P,eds. Surgery of the Shoulder. Elsevier: Oxford: 373–8.

Fiedman AH, Nunley JA, Urbaniak JR, Goldner RD(1990) Results of isolated axillary nerve lesions afterinfraclavicular brachial plexus injuries: case reports,Neurosurgery 27(3):403–7.

Karev A (1986) Trapezius transfer for paralysis of thedeltoid, J Bone Joint Surg 11B(1):81–3.

Kendall FP, McCreary EK, Provance PG (1993) MusclesTesting and Function. Upper Extremity and ShoulderGirdle Strength Tests, 4th edn. Wiliams & Wilkins:235–98.

Kline DG, Hudson AR (1995) Nerve Injuries.Reconstructive Procedures. WB Saunders Co:Philadelphia: 505–12.

Kotwall PP, Mittal R, Malhotra (1998) Trapezius trans-fer for deltoid paralysis, J Bone Joint Surg80B(1):114–16.

Litts CS, Hennigan SP, Williams GR (2000) Medial andlateral pectoral nerve injury resulting in recuurent

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN ADULTS 121

Page 131: Brachial Plexus Injuries

scapular winging after pectoralis major transfer: a casereport, J Shoulder Elbow Surg 9(4):347–9.

Mir-Bullo X, Hinarejos P, Mir-Batlle P et al (1998)Trapezius transfer for shoulder paralysis, Acta OrthopScand 69(1):69–72.

Perlmutter GS, Leffert RD (1999) Results of transfer ofthe pectoralis major tendon to treat paralysis of theserratus anterior muscle, J Bone Joint Surg81A(3):377–84.

Povacz P, Resch H (2000) Dynamic stabilization ofwinging scapula by direct split pectoralis major trans-fer: a technical note, J Shoulder Elbow Surg 9(1):76–8.

Price AE, Grossman AI (1995) A management approachfor secondary shoulder and forearm deformities

following obstetrical brachial plexus injury, Hand Clin11(4):607–17.

Ruhmann O, Wirth CJ, Gosse F, Schmolke S (1998)Trapezius transfer after brachial plexus palsy.Indications, difficulties and complications, J Bone JointSurg 80B(1):109–113.

Ruhmann O, Gosse F, Wirth CJ (1999) Trapezius trans-fer for shoulder paralysis, Acta Orthop Scand70(4):407–8.

Warner JJ, Navarro RA (1998) Serratus anteriordysfunction. Recognition and treatment, Clin Orthop349:139–48.

Wiater JM, Flatow EL (1999) Long thoracic nerve injury,Clin Orthop 368:17–27.

122 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 132: Brachial Plexus Injuries

Introduction

The reconstruction of elbow flexion in patientswith a complete paralysis of the upper extremityleads to highly improved arm function.Regarding amplitude of motion and power of theelbow flexion, grades of function can be differ-entiated. For high quality restoration anintegrated concept is necessary, which includes:

• a precise description of the brachial plexuslesion and/or ventral muscles of the upper arm;

• correct timing of the operation;• preconditions for successful tendon transfer;• specialist knowledge of brachial plexus lesions;• awareness of all therapeutic options;• knowledge of patient-related factors.

Classification of the lesion

A precise preoperative analysis of the lesion isextremely important, and the following clinicalsituations should be differentiated:

• complete paralysis of the upper extremity;• partial paralysis after spontaneous regenera-

tion or operative revision without any regen-eration of the muscles for elbow flexion (M0);

• partial paralysis after spontaneous regenera-tion or operative revision with insufficientregeneration of the muscles for elbow flexion(M1 or M2).

For partial paralysis, the residual function has tobe balanced. Special attention should be paid tothe important muscles of the forearm for elbow

flexion. In a few cases the forearm muscles areso powerful that active elbow flexion is possible(M2+ or M3). This is called the Steindler effect.

In partial paralysis after spontaneous regener-ation or operative revision with insufficientregeneration of the muscles for elbow flexion(M1 or M2), better results can be reached byaugmenting the function by using certain opera-tive techniques when compared to the resultsthat can be achieved by patients with completeparalysis (M0). This may be a result of the bettersensory-motor regulation of joint function byregenerated muscles.

Timing

Reconstructive operations should be performedafter 2–3 years in brachial plexus patients withinsufficient regeneration, spontaneous regenera-tion, or after operative revision. In patients withoutthe option of neurotization and patients withadditional destruction of muscles, an immediatemuscle and/or tendon transfer is indicated.

Preconditions for successfultendon transfer

Very important basic preconditions for success-ful muscle and tendon transfer are glidingtissues and sufficient skin coverage. Sometimesit is necessary to provide a bed for the musclesand tendons by local or distant flap transfers.There has to be free passive motion of the jointsbecause contracture of the tissue can lead to a

13Palliative surgery: the elbow and forearmAlfred C Berger, Robert Hierner, and Lutz Kleinschmidt

Page 133: Brachial Plexus Injuries

significant loss of motion and power. Themuscles chosen for the transfer should have asimilar function or should be at least activated inthe same movement phase. The muscle must betransposed completely. It is not possible to usea muscle both as an agonist and as an antago-nist. There must be sufficient amplitude ofmotion and a sufficient power available for theoperation to succeed.

To avoid unnecessary loss of power a straightrun of the tendon must be chosen. If it is neces-sary to change the direction, a pulley is required.Appropriate tension of the muscle and tendontransfer is important for later function. Lesions ofthe paratendineon should be avoided, and atten-tion must be paid to the donor site defect. Thegain in function must be greater than thefunctional loss at the donor site.

An optimal result for the patient can only begained when all those involved in the treatment(surgeon, anesthesiologist, nursing staff, physi-cal therapist, general practitioner and almoner)are co-operating. Pre- and postoperative physio-therapy in particular are very important. Toachieve a good result, it is beneficial if thepatient has preoperatively learned to contractand isolate the muscle to be transferred.

Specialist knowledge of brachialplexus lesions

In patients with a brachial plexus lesion, the useof regenerated muscles for a reconstruction is

often necessary. Therefore, it is important tofollow some basic rules.

Only muscles graded > M3 are useful for atransposition. Owing to a higher rate of fibrosis,a smaller mass of muscle tissue and a disturbedinnervation, regenerated muscles are normallyless resistant. Excessive tension during theoperation and premature exertion after themobilization should be avoided.

Postoperatively the elbow is fixed in 100°flexion in a whole-arm splint (wrist and fingersfixed in intrinsic plus). Six weeks after the opera-tion active and passive exercises can normallybe started, although still protected against exten-sion by the splint. Each week the splint ismodified so that the elbow is extended by afurther 10°. After 3 months, there should be adeficit of extension of 30–40°.

Therapeutic options

Different techniques are possible for secondaryreconstruction of elbow flexion. These techniquesdiffer regarding the innervation, amplitude ofmotion, power, donor defect and influence on thejoints. Besides reconstructing bimanual use ofthe extremity, there is the possibility of achievingstabilization of the glenohumeral joint andsupination in the forearm. A summary of resultsin the literature is given in Table 1.

The latissimus dorsi is innervated by thethoracodorsalis nerve (C6/C8), which is a branchof the fasciculus posterior. lt is a very powerful

124 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Table 1 Results after muscle/tendon transfer for elbow flexion restoration

Treatment option TAM (°) Extension Flexion Rom (°) Power (kg)lag (°) (ex/flex) (°)

Nerve reconstruction 100 0–60 – 0/30/130 6–16(Narakas 1989)

Latissimus dorsi transfer 65–140 0–20 65–140 0/0/115 0.5–4(Zancolli and Mitre 1973)

Pectoralis major transfer 20–150 0–60 80–150 0/20/150 1–4.5(Schottstaedt et al 1955)

Triceps transfer – – – – 1–2(Narakas 1989)

Steindler transfer 20–140 0–90 20–40 0/22/115 1–2(Literature review)

Modified Steindler transfer 20–140 20–40 95–140 0/33/113 3–4(personal results)

Page 134: Brachial Plexus Injuries

muscle with a constant anatomic position and along neurovascular pedicle. By bipolar transpo-sition of the latissimus muscle, an averagerange of motion of 0–0–115° and power of0.5–4 kg can be achieved. In our series (n = 10),one muscle showed a significant loss of powerpostoperatively although there was very goodperfusion of the skin island. Nine of the 10patients achieved an average active range ofmotion of 0–30–130° and power of 4 kg (in 90°flexion of the elbow). With this technique it is

possible to gain stabilization of the gleno-humeral joint and moderate supination of thelower arm (Fig. 1a–c).

The pectoralis major is innervated by the ansapectoralis (C5–T1). The clavicular part is inner-vated from the fasciculus lateralis (C5/C6/C7) andthe sternocostal part from the fasciculus medialis(C8/T1). Because of the broad innervation, thismuscle can often be used in partial lesions.Preparation of the neurovascular pedicle is moredifficult than in the latissimus dorsi because of

PALLIATIVE SURGERY: THE ELBOW AND FOREARM 125

Figure 1

Bipolar latissimus dorsitransfer for elbow flexionreconstruction: (a) intraop-erative aspect (proximalfixation of the tendon atthe coracoid); (b) intraoper-ative aspect after flap inset-ting; (c,d) postoperativeclinical aspect (activeelbow flexion and lifting of8 kg).

a c

b d

Page 135: Brachial Plexus Injuries

anatomical variations. Owing to the unaestheticscar, this transfer should not be used for women.The pectoralis major is a strong motor musclefor thoracohumeral function, and therefore itshould only be used in a transfer when the latis-simus dorsi or the teres major facilitate activeadduction of the upper arm against the thoracicwall. Patients with a pectoralis major muscletransfer achieve an average range of motion of0–20–150° and power of 1–4.5 kg. We do nothave any experience of this transfer.

The pectoralis minor is innervated by the ansapectoralis (C5–T1) and, because of the broadinnervation, this muscle can also be used inpartial lesions. The pectoralis minor is a compar-atively weak and short muscle, and thereforeshould only be transferred in combination with aSteindler transfer or as an augmentation transferof the elbow (M2). Some recentralization of thecaput humeri can be reached by this transfer.

The triceps is innervated by the radial nerve(C7/T1). Because the triceps is an antagonist ofthe biceps, there are few difficulties in the postop-erative learning period. In our series (n = 15), onemuscle lost its function. The other 14 patientsreached an average active range of motion of0–40–100° and power of 2 kg. This muscle trans-position has one disadvantage in that the patientloses the possibility of active elbow extension. Ifabduction in the shoulder or flexion greater than90° is not possible, the donor site morbidity is notvery important; however, this transfer should notbe used in patients who need elbow extension fordaily living – for example when they have to usecrutches or a wheelchair. Because it is only amonopolar transfer, there is no further stabiliza-tion of the glenohumeral joint. In co-contractionof the biceps and triceps, the triceps transferprovides a very good technique and excellentresults (Fig. 2).

The muscles for the flexion and pronation of theforearm are innervated by the median nerve(C6–T1). These muscles are proximalized by aSteindler transfer and an average range of motionof 0–22–115° and power of 0–2 kg can beachieved. In our series (n = 6), we used a modifi-cation of the Steindler operation with a moreproximal position of the flexor muscles, 8–10 cmproximal to the elbow joint. We lost one casebecause of deep infection; five of the six patientsreached an average active range of motion of0–20–101° and power of 3.2 kg. This transfer

should be avoided if active extension from thewrist and fingers is not possible or cannot bereconstructed by transfer of the flexor carpiulnaris tendon, owing to the pronation and flexioncontracture that follows this technique (Fig. 3).

In patients with a complete paralysis, elbowflexion can only be achieved by microsurgicalmethods. The first stage is a nerve grafting proce-dure. The accessory nerve, intercostal nerves orparts of the contralateral C7 radicula are used asaxon donors. Twelve to 18 months after the firstoperation a biopsy of the distal end of the graftsis performed to assess the quality of the nerveregeneration. The number and quality of theaxons will be quantified. If this shows a usefulnumber of motor functions, a free microvascularmuscle transfer with the latissimus dorsi or rectusfemoris or gracilis is performed. To reconstructelbow flexion and flexion or extension in thewrist, the muscle can be used bifunctionally andplaced under a pulley in the elbow area. Thismeans an insertion in two joint areas, but ofcourse the power in both areas will be dimin-ished. In 50 per cent of these patients, usefulfunctional motion can be achieved. In our series(n = 4), the range of motion was on average0–45–95° and the power 0.7 kg (Fig. 4a,b).

126 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 2

Triceps transferfor elbow flexionreconstruction.

Page 136: Brachial Plexus Injuries

Patient-related factors

These factors influence the choice of method ofreconstruction and the outcome of the proce-dure, and include age, sex, state of health,profession and hobbies, intelligence, expecta-tions of the patient, compliance, social environ-ment and motivation.

PALLIATIVE SURGERY: THE ELBOW AND FOREARM 127

a

b

c

Figure 3

Modified Steindler transfer for elbow flexion reconstruction:(a) intraoperative aspect; (b) postoperative clinical aspect(elbow extension); (c) postoperative clinical aspect (elbowflexion).

Figure 4

Free functional muscle transfer for reconstruction ofelbow flexion: (a) schematic drawing of transfer using theFCU as pulley; (b) postoperative clinical aspect (elbowextension); (c) postoperative clinical aspect (elbowflexion).

a

b

c

Page 137: Brachial Plexus Injuries

Discussion

For the treatment of brachial plexus lesions, anintegrated therapy concept is used. Nerve recon-struction has the highest priority because withthe different techniques of neurolysis, directsuture of the nerve (rare) and nerve transplanta-tion the best functional results can usually begained.

Specific movements can be reconstituted oraugmented by single or multiple tendon trans-fers. Monopolar or bipolar tendon transpositioncan be performed. One joint (monoarticular) ormultiple joints (polyarticular) can be moved. If nolocal muscle is available, a free microvascularfunctional muscle transfer can be performed. Ifthere is no sufficient innervation available, amultiple stage procedure is necessary. The firststep is nerve grafting, and after histologicalassessment of the grafts a free microvascularfunctional muscle transfer can be performed.

All these techniques for reconstruction can beimproved further by using adjuvant operationsor techniques such as tenodesis or capsulodesis.

In our concept, reconstruction of the elbowflexion has highest priority. Next, active functionin the wrist and finger area should be achieved.When possible, the function of the thumb shouldbe reconstructed. Finally, motion in the shoulderarea should be improved.

Selection of the replacement muscle dependson the degree of the defect, the necessity offurther transfers and the age of the patient.

In patients with a complete paralysis, elbowflexion can only be gained by microsurgicalmethods with primary nerve grafting andsecondary free microvascular muscle transferwith the latissimus dorsi. Junctions of the nervegrafts are the ipsilateral accessory nerve, theipsilateral intercostal nerves or parts of thecontralateral C7 radicula. Because of the disap-pointing results of the bifunctional transfers,elbow flexion only should be reconstituted. Ifthere is complete loss of elbow flexion, latis-simus dorsi transfer is the method of choicebecause of the motion amplitude, the power, thelimited donor defect and the additional stabiliza-tion of the glenohumeral joint. Of course, if thismuscle is necessary for the shoulder area itshould not be used for elbow reconstitution. Ifthere is the possibility of re-establishing externalrotation, the shoulder must be able to abduct

more than 90°. When this is possible, the teresmajor or pectoralis major can be considered asa second choice for reconstructing the externalrotation of the shoulder.

Triceps to biceps transfer can also be used, butit should be remembered that when a shoulderabduction of more than 90° is possible, the handcan sometimes hit the patient’s face becausethere is no active extension in the elbow. Thismethod should not be used in children becauseof the possibility of dysarthrosis.

Owing to the donor defect, pectoralis majortransfer is used only as a third choice, especiallyin women.

Modified transfer of the pronator and flexormuscle mass (Steindler) is a good indication inlesions from the C5/C6 area. Absent active wristand finger extension is a relative contraindica-tion. It is possible to reduce this functionalimpairment in these cases by doing a primarytransfer of the flexor carpi ulnaris tendon forwrist and finger extension.

In partial lesions and partial regenerations ofelbow function (M1 or M2), all these methodscan be used as augmentation transfers.

One particular case should be noted; if there isa co-contraction of the triceps and biceps, thetriceps to biceps transfer is the therapy of firstchoice.

In partial paralysis after spontaneous regener-ation, after operative revision with regenerationin the elbow flexion muscles of grade M3 or if aSteindler effect is present, further treatmentshould be discussed with the patient. If thepatient does not need to lift heavy weights indaily life or at work, the indication to operate isquestionable. In these patients, muscle or tendontransfer should only be performed after preciseevaluation and when the patient understandswhat may be lost or achieved.

Summary

Elbow flexion plays a key role in the globalfunction of the upper extremity. In the case ofunilateral complete brachial plexus lesion, restora-tion of elbow flexion will dramatically increase thepatient’s chance of regaining bimanual prehen-sion. Furthermore, depending on the type ofreconstruction, stability of the glenohumeral joint

128 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 138: Brachial Plexus Injuries

as well as some supination function of theforearm can be restored to a varying degree at thesame time. Depending on the level of brachialplexus lesion and/or reinnervation, differentreconstructive procedures are available. In orderto select the best treatment option for the patient,it is necessary to know the extent of the lesion ofthe brachial plexus and/or ventral upper armmuscles, to time the operation appropriately, tobe aware of all treatment possibilities and to havespecialist knowledge of tendon transfer forbrachial plexus patients. Our concept is based onour experience with more than 1100 patientspresenting a brachial plexus lesion between 1981and 1996 and treated in our institution. Therewere 528 operative revisions of the brachialplexus. Some 225 patients underwent secondarymuscle/tendon transfers. In 35 patients elbowflexion was reconstructed by bipolar latissimusdorsi transfer (n = 10), triceps-to-biceps transfer (n= 15), modified flexor/pronator muscle massproximalization (n = 6) and multiple-stage freefunctional muscle transfer after intercostal nervetransfer (n = 4).

References

Akasaka Y, Hara T, Takahashi M (1990) Restoration ofelbow flexion and wrist extension in brachial plexusparalyses by means of free muscle transplantationinnervated by intercostal nerve, Ann Hand Surg9:341–50.

Allieu Y, Triki F, de Godebout J (1987) Les paralysiestotales du plexus brachial. Valeur de la conservation dumembre et de la restauration de la flexion du coude,Rev Chir Orthop 73:665–73.

Alnot JY, Abols Y (1984) Réanimation de la flexion ducoude par transferts tendineux dans les paralysiestraumatiques du plexus brachial de l’adulte, Rev ChirOrthop 70:313–23

Alnot JY, Oberlin C (1991) Transferts musculaires dansles paralysies de la flexion du coude et de l’extensiondu coude. Techniques chirurgical. In: Tubiana R (ed).Traité de Chirurgie de la Main, Vol. 4. Masson: Paris:162–75.

Alnot JY, Rostoucher P, Houvet P (1995) Paralysie dela flexion du coude. In: Alnot JY, Narakas A (eds). LesParalysies du Plexus Brachial, 2nd edn. ExpansionScientifique Française: Paris: 239–42.

Berger A (1985) Ersatzoperationen nach Plexusbrachialis-Verletzungen. In: Hase U, Reulen H-J (eds)Läsionen des Plexus Brachialis. Walter de Gruyter:Berlin: 107–20.

Berger A, Flory P-J, Schaller E (1990) Muscle transferin brachial plexus lesions, J Reconstr Microsurg 6:113–15.

Berger A, Schaller E, Mailänder P (1991) Brachialplexus injuries: an integrated treatment concept, AnnPlast Surg 26 70–6.

Berger A, Brenner P (1995) Secondary surgery follow-ing brachial plexus injuries, Microsurgery 16: 43–7.

Berger A, Schaller E, Becker MH-J (1995) Pulley zurVerstärkung einer Muskelersatzoperation über zweiGelenke bei Plexus Brachialis-Läsionen: Beschreibungder Operationstechnik, Handchir Mikrochir Plast Chir26:51–4.

Berger A, Hierner R, Becker M (1997a) SekundäreErsatzoperationen zur Wiederherstellung derEllenbogenbeugefunktion nach Läsionen des Plexusbrachialis, Orthopäde 26:643–50.

Berger A, Hierner R, Becker M (1997b) Die frühzeitigemikrochirurgische Revision des Plexus brachialis beigeburtstraumatischen Läsionen – Patientenauswahlund Ergebnisse Orthopäde 26:710–18.

Brüser P, Noever G (1988) Welches Ergebnis könnenwir nach einer Bizeps-Ersatzoperation erwarten?Handchir Mikrochir Plast Chir 20:211–17.

Bunnell S (1951) Restoring elbow flexion to theparalytic elbow, J Bone Joint Surg 33A:566–9.

Burkhalter WE (1974) Tendon transfers in brachialplexus injuries, Orthop Clin North Am 5:259–70.

Caroll RE, Hill RA (1972) Triceps transfer to restoreelbow flexion. A study of 15 patients with paralyticlesions and arthrogryposis, J Bone Joint Surg52A:239–44.

Caroll RE, Kleinmann WB (1971) Pectoralis major trans-plantation to restore elbow flexion to the paralyticlimb, J Hand Surg 4:501–7.

Clark JMP (1946) Reconstruction of biceps brachii bypectoral muscle transplantation, Br J Surg 34:180–1.

Hierner R, Berger A (in press) The Steindler flexor-plasty – personal experience and current conceptreview, J Hand Surg (Asian Issue).

PALLIATIVE SURGERY: THE ELBOW AND FOREARM 129

Page 139: Brachial Plexus Injuries

Hierner R, Noah EM, Becker M, Berger A (1996)Ergebnisse und Besonderheiten von motorischenErsatzoperationen bei Verletzungen des PlexusBrachialis, Handchir Mikrochir Plast Chir28:K16–K17.

Le Coeur P (1953) Procédés de restauration de laflexion du coude paralytique par transplantation dupetit pectoral, Rev Chir Orthop 39: 655–6.

Mantkelow RT (1993) Functioning free muscle transfer.In: Green DP (ed). Operative Hand Surgery, 3rd edn.,Vol I. Churchill Livingstone: New York: 1159–77.

Masse P, Morchoisine P (1967) Procédés de restau-ration de la flexion du coude paralytique par trans-plantation du petit pectoral, Rev Chir Orthop53:357–72.

Millesi H (1992) Chirurgie der Peripheren Nerven.Urban & Schwarzenberg: Vienna

Narakas A (1989) Examen du patient et de la fonctiondes divers groupes musculaires du membre supérieur.Critères d’évaluation des résultats. In: Alnot JY,Narakas A (eds). Les Paralysies du Plexus Brachial.Expansion Scientifique Française: Paris: 49–64.

Richards RR (1992) Operative treatment of irreparablelesions of the brachial plexus. In: Gelbermann RH (ed).Operative Nerve Repair and Reconstruction. LippincottCo: Philadelphia: 1303–27.

Rüdigier J (1991) Motorische Ersatzoperationen deroberen Extremität, Band 1: Schultergürtel, Oberarm,Ellenbogen. In: Buck-Gramcko D, Nigst H (eds).Bibliothek für Handchirurgie. Hippokrates: Stuttgart:

Saha AK (1967) Surgery of the paralysed and flailshoulder, Acta Orthop Scand Suppl 97:5–90.

Schottstaedt ER, Larsen LJ, Bost FC (1955) Completemuscle transposition, J Bone Joint Surg 37A:897–918.

Segal A, Seddon HJ, Brooks DM (1959) Treatment ofparalysis of the flexors of the elbow, J Bone Joint Surg41B:44–50.

Smith RJ (1987) Tendon Transfers of the Hand andForearm. Little, Brown and Co: Boston.

Steindler A (1919) Operative treatment of paralyticconditions of the upper extremity, J Orthop Surg 1:608.

von Lanz T, Wachsmuth W (1959) Praktische Anatomie,Vol 11 Part 3, Arm 2. Aufl. Springer: Berlin.

Wood MB (1992) Restoration of elbow flexion bytendon or muscle transfer. In: Conférences d’enseigne-ment 1992, Cahier d’enseignement de la SOFCOTno. 42. Expansion Scientifique Française: Paris: 51–60.

Zancolli E, Mitre H (1973) Latissimus dorsi transfer torestore elbow flexion. An appraisal of eight cases, JBone Joint Surg 55A:1265.

130 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 140: Brachial Plexus Injuries

Reanimation of digital flexion orextension by latissimus dorsimuscle island flap

The latissimus dorsi muscle has been used forthe restoration of elbow flexion (Zancolli andMitre 1973), but it has not often been used torestore digital flexion. This restriction in the useof the latissimus dorsi muscle seems to stemfrom the fact that the thoracodorsal artery hadbeen designated as the main blood supply forthe upper two-thirds of this muscle, and theperforating branches of the intercostal andother collateral arteries as the main supply forthe rest of the muscle (Tobin et al 1981).However, our experiences revealed that thethoracodorsal artery is sufficient for thenourishment of the entire length of the muscle.If the entire length of the muscle from theaxillary fossa to the iliac crest is dissected andfreed, it will usually exceed 40 cm, and this islong enough to reach the wrist and be suturedto the flexor or extensor tendons of the fingers.Moreover, the range of contraction of thismuscle is quite sufficient for generating thecomplete range of digital movement. Therefore,the use of this method and some re-educatingof the patient can restore the digital flexion orextension (Gousheh et al 2000).

Operative technique

Under general anesthesia with the patient in thelateral position and the affected side uppermost,the entire upper extremity is prepared along with

half the chest down to the iliac crest. During theoperation any prolonged or sudden pulling orexcessive abduction of the limb is avoided.

The skin incision, from the axillary fossa to theposterior iliac crest, is made in the shape of a large‘S’. For cases in which overlying skin is required,the length and width of the skin is marked on theanterior part of the muscle and the connectionbetween the muscle and the skin is entirelypreserved. Since the perforating arteries from themuscle supply the skin, the edges of this skin aresutured to the muscle by a few separate stitchesto prevent possible sliding of the skin on themuscle and damage to the nutrient arteries. Theskin flap can be used only for covering the upperpart of the forearm, because the arteries in thedistal part of the muscle are incapable of nourish-ing the skin. The width of the skin flap should alsobe small enough that the edges of the donor sidecan be approximated and sutured directly.

Dissection of the muscle starts from its anterioredge, where it is gradually separated from theserratus anterior muscle, and is continueddownwards. At the lower part, the muscle isseparated from the last three ribs and finally fromthe iliac crest. We prefer to harvest a part of thegluteal aponeurosis along with the latissimusdorsi muscle, and this aponeurosis is used forcovering the tendon suture line to prevent subse-quent adhesion. At this point the entire musclecan be lifted and the neurovascular pedicle on thesuperior part of this muscle is revealed. Arelatively large and important branch enters theserratus anterior muscle, which is ligated anddisconnected. All the small hemorrhagic vesselsare carefully coagulated. The entrance of thevascular pedicle into the muscle is approximately9 cm below the axillary artery and 1.5 cm away

14Palliative surgery:the handJamal Gousheh

Page 141: Brachial Plexus Injuries

from the anterior edge of the muscle (Manktelow1986). Dissection of this pedicle must beperformed carefully and continued up the axillaryartery. Transection of the circumflex scapularartery is not usually necessary. Finally the actionof the nerve is checked by an electrical stimulator,and with this stimulation the entire length of themuscle must contract. At this point the entiremuscle up to its highest point is released and onlyits connection to the humerus and the neurovas-cular pedicle are left intact. It is better to transectthe proximal attachment of the muscle, which islocated at the proximal part of the humerus bone,and transfer and fix it to the coracoid in caseflexion of the elbow is also required.

The anterior and posterior edges of the muscleare sutured with a few separate stitches, and themuscle is then shaped like a tube with a length,usually, of more than 40 cm in adults. It is thenprepared for transfer into the arm (Fig. 1).

Restoration of digital flexion

In the lower third of the forearm the tendons,arteries and nerves are identified, and theprofundus flexor tendons are prepared for sutureto the muscle. A few horizontal incisions aremade along the axillary fossa, the medial aspect

of the arm, and the anterior aspect of the elbowand the forearm. A tunnel is then made underthe skin and the muscular tube guided throughthe tunnel and sutured to the profundus flexortendons of the fingers and the thumb. Theharvested part of the gluteal aponeurosis is usedfor covering the suture area. This significantlyreduces the possibility of subsequent adhesionsin the region of the suture. Attention must bepaid to the following:

• The neurovascular pedicle must be free oftension and torsion;

• The tension of the muscle must be appropri-ate. We keep the elbow at 90° flexion whilesuturing the muscle to the tendons. With thistension, there will be automatic digital flexionwhen the elbow is extended;

• Haemostasis should be done carefully and agood drain placed appropriately.

Postoperatively the elbow is kept in 90° flexionby the use of a cast. The drain can usually beremoved on the second or the third day after theoperation, and the cast after 3 weeks for childrenand 4–5 weeks for adults. During this period thepatient is trained to contract this muscle. Afterremoval of the cast, extensive physiotherapy andexercises are recommended.

132 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

The latissimus dorsi istransferred to the flexoror the extensor tendons.

Page 142: Brachial Plexus Injuries

Restoration of digital extension

An analogous procedure can be used to restoredigital extension. For this, a few horizontalincisions are made along the axillary fossa, themedial aspect of the arm, and the posterior aspectof the elbow and the forearm. A tunnel is thenmade under the skin and the muscular tubeguided through the tunnel and sutured to theextensor tendons of the fingers and the thumb.Again the harvested part of the gluteal aponeuro-sis is used for covering the suture area. Thetension of the muscle must be appropriate. Wekeep the elbow and the fingers completelyextended while suturing the muscle to thetendons. With this tension, there will be automaticdigital extension when the elbow is extended.

Additional procedures

To obtain optimal results it is necessary that theantagonist muscles also function properly. Forcases with complete paralysis of flexion andextension of the fingers, one of these deficien-cies can be remedied by transfer of the latis-simus dorsi muscle and the other by transfer ofthe gracilis muscle. We prefer to use the latis-simus dorsi muscle for generation of digitalflexion and the gracilis muscle for digital exten-sion. This is particularly applicable in cases withsevere Volkmann’s ischemic contracture, inwhich a suitable nerve exists for the reinnerva-tion of the transferred muscle. It should bementioned that in some patients with completeflexor and extensor paralysis the effect of a pre-existing extensor tenodesis is sufficient forrelatively useful function, and we feel that it isunnecessary to transfer a second muscle forthese cases. In some cases we can increase theefficiency and improve the function of the trans-ferred muscle by tenodesis of the antagonisttendons. For good function, it is necessary forthe thumb to have sufficient opposition. In somecases, use of tenodesis for proper adjustment ofthe thumb tendons will improve the function.For good results the wrist joint must have arelatively good function. Although in general wedo not agree with arthrodesis of the wrist, insome rare cases, for example when the wrist iscompletely unstable, it can help to generateeffective flexion of the fingers due to contractionof the latissimus dorsi muscle.

Discussion

We believe that this procedure is a good practi-cal solution for the following types of cases:

• Old and permanent paralysis in the territoryof the lower elements of the brachial plexus,due to war or traction injuries;

• Permanent paralysis of the hand due toirreparable lesions to the median, radial andulnar nerves above the elbow;

• Cases in which there has been a loss of theflexor and/or extensor muscles of the fingersdue to extensive debridement (e.g. in somewar injury cases (Gousheh 1995)).

The transferred muscle is usually strong, andafter the operation patients are capable ofcarrying out daily activities such as writing;some can even carry objects weighing up to10 kg.

One problem that we have not yet been ableto correct is the bowstring deformation of thelimb in the area of the elbow during contractionof the muscle. However, experience has shownthat after a few years this is reduced due toformation of new fascia, and in some cases evenbecomes unnoticeable.

Free muscle transfer for oldcomplete paralysis of brachialplexus

For cases in which all of the brachial plexuselements have been damaged, for example inthe case of complete root avulsion, or whensurgical repairs of brachial plexus elements havefailed, free muscle transfer can be used. Thegracilis muscle is the most suitable one for thispurpose. For the innervation of this muscle wecan use the local nerves of the paralyzed side forexample the spinal accessory or intercostalnerves (Samardzic et al 1992). The use of thenerves from the contralateral (healthy) side hasalso been suggested:

1. Some authors have recommended the use ofthe seventh cervical root of the healthy side(Chwei-Chin Chuang et al 1993). They believe

PALLIATIVE SURGERY: THE HAND 133

Page 143: Brachial Plexus Injuries

that transferring this nerve does not induceany major problem or side effect in the upperextremity of the healthy (donor) side.However, our experience indicates thatrelative atrophy of the triceps and the latis-simus dorsi will develop. Moreover, mildsensory disruption of the index and themiddle fingers will develop;

2. The use of medial pectoralis nerve has noside effects and will be explained in thefollowing section.

Operative technique

In the first stage, a 6-cm horizontal incision of theskin is made below the clavicle of the healthyside. The pectoralis major muscle is dissectedfrom about 1 cm below the lower border of theclavicle. The pectoralis nerve can be observedentering the deep aspect of the muscle. It can bechecked with an electrical stimulator, and thentransected and prepared for grafting. It is best tomark its end with thin nylon, so that it remainseasily accessible.

The entire length of the sural nerve isharvested, and its distal end is sutured to themedial pectoralis nerve with a few stitches of10–0 nylon. The entire nerve is guided under theskin to the other side, until it reaches the proxi-mal part of the arm region. There it is markedwith colored nylon and is left under the skin.Usually after 1 year a small neuroma is observedat this end of the nerve, which indicates nervegrowth.

At the second stage, which is usually 1 yearlater, the gracilis muscle is transferred to thearm. The gracilis muscle is located in the medialaspect of the leg, and extends from the pubisbone to the tibial tuberosity. It is about 42 cmlong in adults, and the distal 10 cm is tendon. Itsnerve is the anterior branch of the obturatornerve, and is 8–10 cm in length. Its dominantvessel pedicle, a branch of the deep femoralvessel, is about 6 cm long. This pedicle, togetherwith the nerve, usually enters the muscle 9 cmbelow the proximal end of the muscle, and iscapable of nourishing the entire muscle(Manktelow 1986).

For harvesting of the gracilis muscle, the skinincision starts from the insertion of the adductor

longus tendon into the adductor tubercle andends at the tibial tubercle. After the skin incisionthe adductor longus muscle can be recognized,and the vessel pedicle is between this muscleand the gracilis muscle. The entire vessel pediclecan be easily observed on lifting of the adductorlongus muscle. There are two veins accompany-ing the artery. These vessels are dissected to theend and the artery and the veins are markeddifferently and harvested along with the muscleand its nerve for the transfer to the arm (Figures2–4).

Simultaneously, a second surgical team pre-pares a suitable vein and artery in the proximalarm region for anastomosis. The end of the suralnerve, which has been transferred in the firststage, is also found and prepared for anastomo-sis. The proximal end of the gracilis muscle isfixed to the coracoid, and its distal end to thebiceps muscle tendon just where the tendonjoins the radius bone. The tension of the muscleis adjusted so as to keep the elbow in 45° flexion.Anastomosis of the vessels and nerve is thenperformed using 10–0 nylon and a microscope.The arm is fixed at 90° of elbow flexion for 5weeks.

The following points should be noted:

1. It is better to harvest part of the skin alongwith the muscle; this can then be used as amonitoring device for checking the blood flowto the muscle postoperatively;

2. After recovery of elbow flexion with sufficientstrength, finger flexion can be regenerated bytransferring the distal end of the muscle to theprofondus digital flexor tendons. This transferis accomplished using the fasciae of thetensor fasciae latae muscle;

3. It is possible to try to generate elbow anddigital flexion in one procedure. Since themuscle is about 42 cm long, it might bepossible to fix the proximal end of the muscleto the coracoid and its distal part directly tothe profondus digital flexor tendons. Whenthis is not possible, the position of the proxi-mal end of the muscle can be adjusted sothat its distal end is sutured directly to thefinger flexor tendons. In order to do this, theproximal end of the muscle is fixed to theproximal part of the humerus bone at theregion of the insertion of the pectoralis majormuscle tendon.

134 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 144: Brachial Plexus Injuries

PALLIATIVE SURGERY: THE HAND 135

Figure 2

The long incision onthe forearm allowsthe placement of themuscle

1 Median nerve;2 Anterior

interosseousmuscle.

Figure 3

The muscle is placedin the forearm. Theproximal tendon isfixed on the medialepicondyle. Thelength of the muscleis restored by usingthe 5-cm intervalsbetween the sutures.

Figure 4

In this case reconstruction of elbow and finger flexion is performed with the samemuscle transplant.

Page 145: Brachial Plexus Injuries

References

Chwei-Chin Chuang D, Wei FC, Noordhoff MS (1993)Cross-chest C7 nerve grafting followed by free muscletransplantations for the treatment of total avulsedbrachial plexus injuries: a preliminary report, PlastReconstr Surg 92:717–27.

Gousheh J (1995) The treatment of war injuries of thebrachial plexus, J Hand Surg 20A:S68–76.

Gousheh J, Arab H, Gilbert A (2000) The extended latis-simus dorsi muscle island flap for flexion or extensionof the fingers, J Hand Surg 25B:160–5.

Manktelow RT (1986) Microvascular Reconstruction.Springer Verlag: 46–52.

Samardzic M, Grujicic D, Antunovic V (1992) Nervetransfer in brachial plexus traction injuries, JNeurosurg 76:191–7.

Tobin GR, Schusterman M, Peterson GH et al (1981)The intramuscular neurovascular anatomy of the latis-simus dorsi muscle: the basis for splitting the flap,Plast Reconstr Surg 67:637–41.

Zancolli E, Mitre H (1973) Latissimus dorsi transfer torestore elbow flexion, J Bone Joint Surg 55A:1265–75.

136 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 146: Brachial Plexus Injuries

The restoration of prehension after completeavulsion of the brachial plexus has been thefocus of recent interest in reconstruction of theupper limb following brachial plexus injuries.The nerve cross-over technique, similar to trans-fer of the intercostal nerves to the median nerveto restore finger function (Millesi 1987, Narakas1987), has failed for several reasons. Thedistance is too great between the site of nerveanastomosis and the neuromuscular junction ofthe forearm musculature. More than 1.5–2 yearsis needed for the regenerating axon to reach thetarget muscle and by this time atrophy of themuscle has ensued. Regenerating axons are alsomisdirected frequently and further contribute tothe compromised result. Hence, simple nervetransfer should not be attempted to restorefinger function following brachial plexus injuries.It can be used to achieve shoulder stability andactive elbow flexion (Nagano et al 1989).

Re-innervated free-muscle transplantation hasallowed recovery of motor function in cases ofseverely injured extremities, such as Volkmann’scontracture and traumatic muscle loss(Manktelow et al 1978, 1984, Doi et al 1993). Itcan also be used to provide reliable and power-ful recovery of motor function following brachialplexus injury. Classically, free muscle transferhas been used for reconstruction of elbowflexion in non-acute cases with flail upper limbsecondary to brachial plexus injury. It has alsobecome possible to regain finger and wristfunction following complete avulsion of C5 to T1roots using free muscle transfer. Inadequaterecovery of power did not allow use of the upperextremity for activities of daily living. Recently,attempts have been made to restore morefunction than simple elbow flexion in cases oftotal avulsion injury of the brachial plexus.

Free muscle transfer can provide reliable andpowerful motor recovery for finger function. Theneuromotor units of the free muscle are in theupper arm nearer to the donor nerve, and thenerve to the muscle is purely motor. Followingbrachial plexus injury, free muscle transfercombined with multiple nerve transfers of thespinal accessory nerve and intercostal nervescan be used to restore prehensile function.

Prehension is a basic function of the humanhand. A patient with complete avulsion of thebrachial plexus has a normal contralateral upperlimb. He or she can perform most of the activi-ties of daily living with the unaffected upperlimb. Few activities require the use of bothhands, such as lifting a heavy box or holding abottle while opening its cap. These actions needa powerful grip, independent of the contralaterallimb. Hence, direct activation of finger flexion isimperative for a powerful grip (Doi et al 1991).

Several surgical approaches have been devel-oped to restore prehension following completebrachial plexus avulsion (Berger et al 1990,Akasaka et al 1991, Doi et al 1995). To restorepinch (key grip), the Moberg type of simplehandgrip reconstruction by activating radial wristextensors has been attempted in the past(Moberg 1976). This technique along with simplenerve crossing failed to achieve independentactivation of forearm muscles. Some investiga-tors incorporated free muscle transfer toovercome these difficulties (Berger et al 1990,Akasaka et al 1991). Although the transferredmuscle functioned well, finger flexion was weak,as flexion was achieved indirectly by synergisticaction. Patients with brachial plexus injury avoidusing the hand reconstructed by this synergisticaction, as the contralateral normal hand caneasily perform simple pinch.

15Palliative surgery: free muscle transfersKazuteru Doi

Page 147: Brachial Plexus Injuries

Grasp release is also essential for prehension.This can be achieved by secondary tenodesis ofthe finger extensor tendons in a Moberg-typereconstruction. With this type of reconstruction,finger extension is assisted by gravity-aidedwrist flexion and release can only be achievedwith the elbow in flexed position. This cannot beaccomplished with the elbow in extendedposition, making it difficult for the patient to usethe reconstructed hand for everyday activities.Hence voluntary finger extension independent ofelbow position is another prerequisite for recon-struction.

Stability of the shoulder and elbow joints isperemptory to use the transferred muscleefficiently. Re-innervated, transferred freemuscles and triceps and shoulder girdle muscleprovide stability to shoulder joint. If instabilitypersists glenohumeral arthrodesis should beconsidered. Most authors (Berger et al 1990,Akasaka et al 1991) denied the significance ofelbow stabilization, maybe because of its techni-cal difficulty, and failed to provide some form ofstabilization. Subsequently, even if their patientsregained powerful wrist extension or fingerflexion, they were not able to use the fingersoptimally in daily activities because the elbowwas unstable. All transferred muscles act simul-taneously to cause elbow flexion and fingerextension or finger flexion. Action is similar tothe transferred brachioradialis muscle in cases ofspinal cord injury. In such situations, the patientsstabilize the unstable elbow by the contralateralhand because of non-functioning elbow exten-sors. Patients find it too inconvenient and avoidusing the reconstructed upper extremity.Ultimately atrophy of transferred muscles ensuesand the limb becomes useless for everyday activ-ities. Reconstruction of the elbow extension isimperative whenever prehension is being recon-structed by the transfer of one muscle thatmoves multiple joints simultaneously (Doi et al1997).

Basic sensory functions such as protectivesensation and position sense should be restoredwhen the motor function is reconstructed for theseverely paralysed limb (Ihara et al 1996).

Limited numbers of motor nerves are availablefor reconstruction in brachial plexus injury. Wehave preferred to use the spinal accessory andthe third to sixth intercostal nerves. Other avail-able donor nerves are the phrenic nerve and

contralateral C7 nerve root (Chung et al 1993, Guet al 1998). However, we chose not to use thesenerves because of possible adverse risk. Thelimited number of the available motor nervesforced us to explore the possibility of one muscletransfer for a two-function concept and to doreconstruction with double free muscle transferto achieve prehension in a flail limb in brachialplexus injury.

The most useful functions in everyday activi-ties for these patients are powerful grip, even ifit is hook grip, flexion and extension of theelbow, and stability and rotation of the shoulderjoint. While planning reconstruction, priorityshould be given to these factors.

The operative technique and long-term resultsof the double free-muscle transfer procedure forreconstruction of prehension following completeavulsion of the brachial plexus are described inthis chapter (Doi et al 1995, Doi et al 1999, Doiet al 2000).

Indications

Patient selection

Based on the authors’ experiences, the mostimportant prerequisites for this procedure arethe patient’s motivation and financial support tocontinue the postoperative rehabilitation for atleast one year, the interval between injury andoperation, and the patient’s age. A motivatedpatient without financial worries about the treat-ment will be able to concentrate more on therehabilitation programme, which will greatlyassist a better recovery. Elbow stability is imper-ative to obtain a satisfactory result and is usuallyachieved by the re-innervated triceps brachimuscle. When this procedure was done laterthan one year after injury, severe atrophy of themuscle resulted. No useful prehension for thepatient was restored, because of incompleterecovery of the triceps brachi which producedloss of elbow stability. This procedure should beperformed at least within 6 months following theinjury. Poor re-innervation of the transferredmuscles and other problems such as jointcontracture and causalgia may result in inade-quate recovery in older patients.

138 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 148: Brachial Plexus Injuries

This procedure should not be performed inthe presence of subclavian artery injury. Therecipient vessels for anastomoses to the nutri-ent vessels of the transferred muscle mighthave been injured and development of throm-bosis in the anastomosed vessels is more likelyto occur. Other important prerequisites for thistransfer are finger and elbow joint mobility, thepresence of undamaged tendons in the handand forearm, and good skin coverage in the armand forearm.

Donor muscle selection

During the initial period in this series, the latis-simus dorsi, gracilis and rectus femoris muscleswere used as the donor muscle. However, long-term results with the latissimus dorsi and rectusfemoris were not satisfactory. The latissimusdorsi failed to provide satisfactory finger functionbecause of adhesion of the muscle to the recon-structed pulley system and rupture of its tendondue to ischaemic necrosis of the portion distal tothe pulley. Because of the shorter fascicle length,the rectus femoris muscle had less excursion,resulting in poor finger function. The gracilis isthe choice of donor muscle.

Donor nerves selection

Only a limited number of motor nerves are avail-able for reconstruction in brachial plexus injury.Simultaneous reconstruction of two functionscan be achieved by transfer of a single muscle.This at times may resolve the discrepancybetween the number of available motor nervesand the number of basic functions that neededrestoring. The spinal accessory and the third tosixth intercostal nerves are commonly used fortransfer for motor and sensory reconstruction.Elbow flexion and finger extension are restoredby using a single free muscle transfer innervatedby the spinal accessory nerve. Finger flexion isrestored by a second free muscle neurotized bythe fifth and sixth intercostal nerves. The thirdand fourth intercostal nerves are anastomosed tothe motor branch of the triceps brachi muscle torestore elbow extension and to stabilize the

elbow, for negating the tendency of elbowflexion that may occur when the fingers areextended.

For sensory reconstruction of the hand, theintercostal nerves showed better recovery thanthe supraclavicular nerves (Ihara et al 1996).

Technique

This consists of six established reconstructiveprocedures:

1. Exploration of the brachial plexus, intraoper-ative monitoring of spinal evoked potentialand sensory action potential and repair of theruptured motor nerves if possible;

2. The first free muscle transfer for elbow flexionand finger extension – neurotization of trans-ferred muscle by the spinal accessory nerve;

3. The second free muscle transfer for fingerflexion – neurotization of the transferredmuscle by the fifth and sixth intercostalnerves;

4. A nerve-crossing procedure to achieve elbowextension – the motor branch of the tricepsbrachi muscle is neurotized by using the thirdand fourth intercostal nerves. This is donesimultaneously with the second muscle trans-fer;

5. The intercostal sensory rami is coapted to themedial cord of the brachial plexus to restoresensibility of the hand;

6. Secondary reconstruction – arthrodesis of theglenohumeral joint, carpometacarpal joint ofthe thumb and wrist joint, to increase stabil-ity; if indicated, tendon transfer of re-inner-vated infraspinatus muscle to stabilize theelbow joint if triceps brachi muscle recoveryis not adequate; and tenolysis of the trans-ferred muscle and tendons.

Timing of the various reconstructive proceduresis important and is guided by several criteria.Procedures (1) and (2) are performed at the firststage operation. Procedures (3), (4) and (5) aredone at the second stage operation, usually twoor three months after the first. Operationsmentioned in procedure (6) are done dependingon the condition of recovery, around one and ahalf years after the first stage operation.

PALLIATIVE SURGERY: FREE MUSCLE TRANSFERS 139

Page 149: Brachial Plexus Injuries

Brachial plexus exploration

Surgical exploration of the brachial plexus wascarried out in all the patients. Spinal-evokedpotentials and sensory nerve action potentialswere recorded intraoperatively to confirm thediagnosis (Fuchigami et al 1994) and to definethe pattern and level of injury. Reconstructionwas undertaken only after confirming completebrachial plexus palsy secondary to avulsion ofthe C5 to T1 nerve roots. If, during the explo-ration, rupture (post-ganglionic injury) of the C5nerve root was observed, nerve grafting betweenthe proximal remnant of the C5 root and thesuprascapular nerve was done to restore shoul-der function.

First muscle transfer

The first stage was to reconstruct simultaneouselbow flexion and finger extension with a freemuscle transfer using the spinal accessory nerveas the motor nerve (Fig. 1).

Gracilis muscle harvesting

The donor gracilis muscle from the contralateralthigh was selected, as the orientation of itsneurovascular bundle matched the donorvessels. The length of the gracilis muscle, fromits pubic bone origin to its tibial insertion wasmeasured. (The length of muscle should besufficient to span the distance between theacromion and midforearm without tension.) Thedonor muscle was then harvested from its proxi-mal origin to the distal attachment. Gracilismuscle harvesting by the conventionaltechnique leaves a long and unacceptable scarat the donor site. To minimize the length of scar,endoscopic harvesting of the gracilis muscle isrecommended (Doi et al 1997). Correct muscletension is critical for good postoperativefunction. Prior to detaching the muscle, theresting length of the muscle was recorded byplacing black silk ligatures on the surface of themuscle at every 5 cm, as described byManktelow et al (1984).

Selection of donor nerve and recipientvessels

The distal portion of the spinal accessory nervewas used for neurotizing the transferred muscle.Sparing the upper branch preserved the motorfibres to the superior fibres of the trapezius

140 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 1

The first free muscle to restore finger extension and elbowflexion simultaneously is transferred in the anterior surfaceof the upper arm. The nutrient vessels of the muscle areanastomosed to the thoracoacromial artery and cephalicvein, and the motor nerve is connected to the spinal acces-sory nerve. The muscle origin was sutured to the acromionand latera clavicle proximally, passed underneath thepulley of the brachioradialis and wrist extensors in theanterior elbow and the distal tendinous portion of themuscle was sutured to the extensor comminus tendonsdistally in the forearm. (a) Accessory nerve; (b) motorbranch of the muscle transplant; (c) thoracoacromial arteryand branches of the cephalic vein; (d) nutrient artery andveins of the muscle transplant; (e) muscle transplant; (f)the brachioradialis and wrist extensors serving as a pulley;(g) extensor digitorum communis tendon. Reproducedwith permission from Doi et al 2000.

Page 150: Brachial Plexus Injuries

muscle. The distal portion of the spinal accessorynerve was dissected, divided, and transferred.The middle and distal branches of the spinalaccessory nerve were coapted to the motornerve of the transferred muscle, which waspassed underneath the clavicle. The nutrientvessels of the muscle transplant were thenanastomosed to the thoracoacromial artery andvein, or the cephalic vein.

Position of the transferred muscle

A simple and straight route for the free musclewas created from its new origin, the acromionto its final insertion in the forearm to maximizethe force of contraction. It was placed on theanterior portion of the deltoid muscle, thelateral aspect of the upper arm and dorsalsurface of the forearm; deep to the brachiora-dialis and radial wrist extensor muscles, justdistal to the elbow to prevent bowstringing.This position is optimal for elbow flexion andfinger extension, although the grip may weakenwhen the elbow is flexed. The origin of thetransferred muscle was sutured to theacromion. The distal tendon of the transferredmuscle was coapted to the extensor digitorumcommunis tendons.

Adjustment of muscle tension

For optimal function, it is essential that thecorrect muscle tension be reproduced in theupper limb before final suturing of the muscle tothe finger extensors. The muscle was stretchedto restore the original length, until the distancebetween markers was once again 5 cm. Whileadjusting the tension, the elbow was kept inminus 30° of extension; wrist in neutral and thefingers were in fully extended position. Theposition of the stumps of the extensor digitorumcommunis tendons on the donor muscle tendonwas noted and marked and the extensor digito-rum communis tendons were coapted to themuscle transplant. The elbow was then flexed to90° and with the fingers fully flexed and the wristin a neutral position, tenorrhaphy was done atthe previously marked sites. The biomechanicswere examined.

Second free muscle transfer

The second free muscle transfer for reconstruc-tion of finger flexion was done 2–3 months afterthe first operation (Fig. 2). By this time postop-erative contracture of the elbow and finger jointshad improved. The third to sixth intercostalnerves were dissected up to the mid-clavicularline and transferred to the axillary region. Thesecond gracilis muscle was harvested from theipsilateral thigh, making sure that it would spanthe distance from the second rib to the mid-forearm. The proximal end of the gracilis musclewas sutured to the second and third ribs near themid-axillary line. For fixation, holes were drilledin the ribs and the muscle was anchored withnon-absorbable sutures passed through theseholes. The muscle was placed on the medialaspect of the upper arm and forearm so as notto be a secondary elbow flexor.

PALLIATIVE SURGERY: FREE MUSCLE TRANSFERS 141

Figure 2

The second free muscle to restore finger flexion is trans-ferred in the medial surface of the upper arm, and thenutrient vessels of the muscle are anastomosed to thethoracodorsal artery and vein individually. The motor nerveof the muscle is connected to the fifth and sixth intercostalnerves. The muscle is sutured to the second and third ribproximally, and the distal tendinous portion of the tendonis sutured to the flexor digitorum profundus tendonsdistally in the forearm, following passing the muscle under-neath the pulley of the pronator teres and wrist flexors. (a)muscle transplant; (b) long finger flexor tendons. (c) prona-tor teres and wrist flexors serving as a pulley; (d) thora-codorsal artery and vein; (e) nutrient artery and veins ofthe muscle transplant; (f) the second and third ribs; (g) thefifith and sixth intercostal nerves; (h) motor branch of themuscle transplant. Reproduced with permission from Doiet al 2000.

Page 151: Brachial Plexus Injuries

In the forearm, just distal to elbow, the tendi-nous part of gracilis was passed under thepronator teres and long wrist flexors. The distalportion of the gracilis muscle being tendinousand thin passed easily through the small hiatusdeep to the muscles and was coapted to theflexor digitorum profundus. Muscle tension wasdetermined using principles as described above.The nutrient vessels were anastomosed to thethoracodorsal artery and vein. In the axilla, thefifth and sixth intercostal nerves were anasto-mosed without tension to the motor nerve of thesecond muscle transplant.

Nerve-crossing to the motor branchof the triceps brachi muscle

The third and fourth intercostal nerves wereconnected to the motor branch of the tricepsbrachi muscle, to activate the elbow extensors.This was done prior to the detachment of thesecond muscle from the thigh (Fig. 3).

Sensory reconstruction

The sensory rami of intercostal nerves weresutured to the medial cord of the brachial plexus atthe second operation to restore hand sensibility.

Postoperative management

The upper limb was immobilized without tensionon the transferred muscles, motor nerves andnutrients vessels for 4 weeks postoperatively andthen the rehabilitation programmes were startedas described below.

Rehabilitation

Initial stage: before electromyographicre-innervation of transferred muscles

The use of electrical stimulation for the trans-ferred muscles and nerve-repaired muscles is

still controversial. However, the authors preferthe electrical stimulation of the paralysed targetmuscles, such as the two transferred gracilismuscles, the triceps brachi, and the supraspina-tus, and infraspinatus if the suprascapular nervewas repaired. The low-intensity electrical stimu-lation of the muscle was started from the thirdpostoperative week and continued untilelectromyographic re-innervation was detectedin the muscles.

At this stage, functional orthosis was used toimmobilize the reconstructed upper limb. Theauthors preferred the air-bag type orthosis(Nakamura brace, Shimane, Japan) to immobilizethe shoulder and elbow joints and a plaster-of-Paris long-arm cast was used for immobilization

142 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 3

Nerve-crossing of the third and fourth intercostal nerves tothe motor branch of the triceps brachi muscle to restoreelbow extension and stabilization following completebrachial plexus avulsion is also done at the second freemuscle transfer. (a) the third and fourth intercostal nerves;(b) motor branch of the triceps brachi muscle; (c) tricepsbrachi muscle. Reproduced with permission from Doi et al2000.

Page 152: Brachial Plexus Injuries

of the wrist and finger joints. Four weeks follow-ing operation, the long-arm cast was removedand only passive flexion of the elbow joint wascommenced. During the early postoperativeperiod, a plastic static splint was used to maintainthe wrist in a neutral position and the proximaland distal interphalangeal joints in extension toallow these joints to contract in this position. Atthe sixth postoperative week, while protecting themuscle–tendon suture site of the transferredmuscle from over-tension by keeping the wristjoint in extension following the first free muscletransfer or in flexion following the second freemuscle transfer, passive extension of the elbowwas started.

Only the metacarpophalangeal joints weremoved passively, since the transferred musclesintended to move the single joint action todecrease the effect of claw finger deformity. Atthe ninth postoperative week, the air-bag ortho-sis was discontinued and the elbow sling-typeorthosis was applied to prevent subluxation ofthe shoulder.

Late stage: after re-innervation

Following electromyographic documentation ofre-innervation of the transferred muscle, usuallybetween three and eight months postoperative,electromyographic biofeedback techniques werestarted to train the transferred muscles to movethe elbow and fingers. Muscular facilitation or re-education is indicated when patients displayminimal active contraction with an identifiedmuscle or muscle group. The initial goal of re-education is for patients to reactivate voluntarycontrol of the muscle. When the patient isworking with a weak muscle, initially the inten-sity of the motor unit activity and the frequencyof the muscle contraction are emphasized.Treatment sessions should be short and endwhen fatigue is noted by a decreasing ability ofthe patient to achieve the set goal level.

Independent finger flexion and extension train-ing using electromyographic biofeedbacktechniques commenced following recovery ofactive motion of the elbow and fingers. Patientsalso practised skilled activities, such as lifting,holding, carrying, and pinching. All patientsshould follow the rehabilitation programme every

day for 6 months postoperatively. And after thepatients have mastered activation of the trans-ferred muscle, the home programmes consistingof power-up exercise, individual activation of thetransferred muscles, and daily-use practiceshould be detailed and recommended to thepatients to be performed by themselves. Theauthors recommend the patients to continue therehabilitation programme at least for 3 yearspostoperatively, as functional recovery will beexpected to occur even after this period.

Secondary reconstruction

Elbow extension (dynamic stability)

Inadequate recovery of triceps muscle followingintercostal nerve neurotization resulted in elbowinstability. This ruined prehension function, evenif the power strength of re-innervated transferredmuscles was enough to move the fingers. Elbowstability could be provided by supplementalreinforcement of elbow extension by transferringthe re-innervated infraspinatus to the triceps.Tenodesis of the triceps brachi when theinfraspinatus recovery is not enough for transfer,is another optional procedure to provide stabil-ity of the elbow (Doi et al 1997).

Tenolysis

One-third of our 34 cases with double freemuscle transfer resulted in sliding insufficiencyof the transferred muscle in spite of satisfactorycontraction of the muscle, which needed tenoly-sis. Under local anesthesia, tenolysis of thetransferred gracilis and distal tendons wasperformed to allow full evaluation from theproximal musculotendinous junction of thegracilis to the distal insertion of the tendon. (Careshould be taken when releasing adhesionsunderneath the pulley system.)

Carpometacarpal arthrodesis of the thumb

After significant recovery of the active fingermotion, the carpometacarpal fusion of the thumbwith intentional contracture of the metacar-

PALLIATIVE SURGERY: FREE MUSCLE TRANSFERS 143

Page 153: Brachial Plexus Injuries

144 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Figure 4

A 25-year-old man sustainedcomplete avulsion of his leftbrachial plexus and underwentthe double free muscle proce-dure. Finger flexion (a) andextension (b) with the elbowin extension; finger flexion (c)and extension (d) with theelbow in flexion; unscrewing abottle cap with both hands (e);and lifting a 5-kg box withboth hands (f) 36 monthspostoperatively. Reproducedwith permission from Doi et al2000.

a b

fe

c

d

Page 154: Brachial Plexus Injuries

pophalangeal joint and interphalangeal joint ofthe thumb and proximal and distal interpha-langeal joints of the fingers provides stable pinchfunction.

Glenohumeral arthrodesis

If the instability of shoulder joint persists, arthro-desis of the glenohumeral joint may be performedat a later stage, following re-innervation of thetransferred muscle. Arthrodesis of the gleno-humeral joint will allow the scapulothoracic jointto be moved by the remnant of the trapeziusmuscle for patients with an unstable shoulderjoint. This permits control of the upper limb andprevents dispersion of power of the transferredmuscles. Glenohumeral arthrodesis should bedone as a final stage in the reconstructiveprogramme, since it makes the reconstructiveprocedures difficult to perform in the position ofshoulder adduction.

Complications

Vascular compromise

The accompanying skin flap helps to monitorcirculation of the transferred muscle, which maydevelop vascular insufficiency due to thrombosisin the anastomosed vessels. In case of vascularproblems, prompt exploration of the anasto-mosed vessels is imperative for survival of thetransferred muscle. After revision of the anasto-mosed vessel, not only fresh bleeding from themuscle, but also muscle contraction to electricalstimulation should be present. Otherwise, themuscle will develop ischemic necrosis andanother free muscle transfer may be needed.

Delayed or failed re-innervation ofthe transferred muscles, andtriceps brachi muscle

Electromyographic evidence of re-innervation ofthe transferred muscle was detected between 3

and 10 months after surgery, depending on thedonor motor nerve used. Re-innervation wasdetected much earlier in muscles where thespinal accessory nerve was used (mean 3.9months) than those re-innervated by the inter-costal nerve (mean 4.8 months). Voluntarycontraction was detected about 2 months afterelectromyographic documentation of re-innerva-tion. Re-innervation of the triceps muscleoccurred even much later than those of the trans-ferred muscles re-innervated by the intercostalnerve (mean 8.2 months).

Delayed re-innervation later than the timedescribed above or no re-innervation at allindicates a poor functional prognosis. Delayedand failed re-innervation will result in muscleatrophy. Atrophied muscle will not be able toactivate joint motion. In such situations, furtherreconstructive plans should be abandoned.

Adhesion of the transferred muscle

Adhesion of the transferred muscle to thesurrounding tissue occurred more or less in allthe cases. One-third of the cases underwenttenolysis. Tenolysis was indicated when activefinger function was not achieved despite strongcontraction of the transferred muscle. All thetransfers done with the gracilis muscle resultedin improved range of finger motions postopera-tively; however, the latissimus dorsi had recur-rence of adhesions.

Instability of the proximal joints

The transferred double free muscle moved multi-ple joints simultaneously. The first free musclewas for both elbow flexion and finger extension.The second free muscle was for finger flexionand for this reason it was not placed in theflexion–extension plane of the elbow. Thismuscle was not intended for elbow flexion.However, this also caused finger flexion andelbow flexion simultaneously. Hence, simultane-ous elbow flexion occurred with fingermovement as long as the antagonist of elbowflexion did not recover. The third and fourthintercostal nerves were anastomosed to the

PALLIATIVE SURGERY: FREE MUSCLE TRANSFERS 145

Page 155: Brachial Plexus Injuries

motor branch of the triceps brachi muscle torestore elbow extension and to stabilize theelbow, negating the tendency of elbow flexionwith finger movement. Even if the power of thetriceps brachi was weak, it could contribute tostability of the elbow with the aid of gravity. Ifre-innervation of the triceps brachi has failed, re-innervated infraspinatus transfer to the tricepsbrachi or tenodesis of the triceps brachi isrecommended to restore elbow stability.

The re-innervated free muscle, triceps brachiand shoulder girdle muscles without arthrodesiscan achieve stability of the glenohumeral joint.During exploration of the brachial plexus, if theC5 nerve root is available, it should be crossedto the suprascapular nerve using nerve graft, notonly to improve shoulder function, but also to re-innervate paralysed muscles for use as possibledonor muscles for transfer, if the triceps brachidid not recover. If the glenohumeral jointremains unstable even after recovery of thesemuscles, glenohumeral arthrodesis can be done,although this will limit several activities, forexample, turning over during sleep may becomedifficult. Care must be taken to prevent fractureof the proximal humerus.

Sensibility of the hand

Restoration of basic modalities, e.g. protectivesensation and position sense, is imperative whenprehensile function is reconstructed for irrepara-ble brachial plexus injury. Half of the patientsachieved sensitivity of the palm better than S2(Highet’s grading system) and adequate positionsense. None of the patients achieved sensitivityover the finger tips. Even so, protective sensa-tion did not recover over the ulnar side of thehand and finger. Patients tended to sustainminor injury and burned their hands over theulnar side of the hand.

Pain syndrome

Unlike reports in the literature, none of ourpatients had severe causalgia that could not becontrolled using the usual analgesics.

General functional outcome

Of patients reconstructed by the double freemuscle procedure and followed-up longer than24 months after the second free muscle trans-fer (mean follow-up, 40 months) 26 out of 32were assessed for long-term outcome of univer-sal prehension, including motion and stabilityof the shoulder and elbow, voluntary andindependent motion of the fingers, sensibilityand ADL functions. Functional outcome ofprehension according to the authors’ classifica-tion (Doi et al 1995) was excellent in 4 patients(which implies restoration of more than 90°elbow flexion, dynamic stability of elbow, whilemoving fingers and more than 60° of totalactive motion of fingers; TAM) (Fig. 4), good in10 (same as excellent, except TAM 30–60°) fairin 2 (TAM < 30°) poor in 9 and bad in 1.Satisfactory results, better than good, wereobtained in 14 out of 26 patients (54 per cent)who obtained more than 90° elbow flexion,dynamic stability of elbow, voluntary fingermotion at any position of elbow, more than 30°of total active motion of fingers, and daily useof their reconstructed hand for both-handsactivities, such as holding a bottle whileopening a cap or lifting a heavy object. Thesesatisfactory results were obtained only frompatients:

• of age younger than 32 years;• with an interval between injury and surgery

shorter than 8 months;• with a longer follow-up than 55 months;• without serious accompanied injuries of

subclavicular artery, spinal accessory nerveand spinal cord;

• for whom the bilateral gracilis muscles wereselected as donor muscle;

• with successful restoration of elbow stabilitywith nerve-crossing or secondary tendontransfer.

Satisfying these prerequisites, the double freemuscle procedure should provide reliable anduseful prehensile function to the patient withcomplete avulsion of the brachial plexus toenable the use of their otherwise useless limb(Doi et al 2000).

146 THE ADULT TRAUMATIC BRACHIAL PLEXUS

Page 156: Brachial Plexus Injuries

References

Akasaka Y, Hara T, Takahashi M (1991) Free muscletransplantation combined with intercostal nerve cross-ing for reconstruction of elbow flexion and wrist exten-sion in brachial plexus injuries, Microsurgery 12:346–51.

Berger A, Flory PJ, Schaller E (1990) Muscle transfersin brachial plexus lesions, J Reconstr Microsurg6:113–16.

Chung DC, Wei FC, Noordhoff MS (1993) Cross-chestC7 nerve grafting followed by free muscle transplanta-tions for the treatment of total avulsed brachial plexusinjuries: a preliminary report, Plast Reconstr Surg92:717–25.

Doi K, Hattori Y, Tan S-H et al (1997) Endoscopicharvesting of the gracilis muscle for re-innervated free-muscle transfer, Plast Reconstr Surg 100:1817–23.

Doi K, Kuwata N, Muramatsu K et al (1999) Doublemuscle transfer for upper extremity reconstructionfollowing complete avulsion of the brachial plexus,Hand Clin 15:757–67.

Doi K, Muramatsu K, Hattori Y et al (2000) Restorationof prehension with the double free muscle techniquefollowing complete avulsion of the brachial plexus, JBone Joint Surg 82A:652–66.

Doi K, Sakai K, Ihara K et al (1993) Reinnervated freemuscle transplantation for extremity reconstruction,Plast Reconstr Surg 91:872–83.

Doi K, Sakai K, Kuwata N et al (1995) Double-muscletechnique for reconstruction of prehension aftercomplete avulsion of brachial plexus, J Hand Surg20A:408–14.

Doi K, Sakai K, Kuwata N et al (1991) Reconstruction offinger and elbow function after complete avulsion ofthe brachial plexus, J Hand Surg 16A:796–803.

Doi K, Shigetomi M, Kaneko K et al (1997) Significanceof elbow extension in reconstruction of prehensionwith re-innervated free muscle transfer followingcomplete brachial plexus avulsion, Plast Reconstr Surg100:364–72.

Fuchigami Y, Doi K, Kawai S et al (1994) Intraoperativeelectrodiagnosis for brachial plexus injury, J Jpn SocSurg Hand 11:559–62.

Gu YD, Chen DS, Zhang GM et al (1998) Long-termfunctional results of contralateral C7 transfer, JReconstr Microsurg 14:57–9.

Ihara K, Doi K, Sakai K et al (1996) Restoration of sensi-bility in the hand after complete brachial plexus injury,J Hand Surg 21A:381–6.

Manktelow RT, McKee NH (1978) Free muscle trans-plantation to provide active finger flexion, J Hand Surg3:416–26.

Manktelow RT, Zuker RM, McKee NH (1984)Functioning free muscle transplantation, J Hand Surg9A:32–9.

Millesi H (1987). Brachial plexus injuries: managementand results. In: Terzis J, ed. Microreconstruction ofNerve Injuries. WB Saunders: Philadelphia: 347–60.

Moberg E (1976) Reconstruction hand surgery intetraplegia, stroke and cerebral palsy: some basicconcepts in physiology and neurology, J Hand Surg1:29–34.

Nagano A, Tsuyama N, Ochiai N et al (1989) Directnerve crossing with the intercostal nerve to treatavulsion injuries of the brachial plexus, J Hand Surg14A:980–5.

Narakas AO (1987) Thoughts on neurotization or nervetransfers in irreparable nerve lesions. In: Terzis J, ed.Microreconstruction of nerve injuries. WB Saunders:Philadelphia: 447–54.

PALLIATIVE SURGERY: FREE MUSCLE TRANSFERS 147

Page 157: Brachial Plexus Injuries
Page 158: Brachial Plexus Injuries

Obstetrical Paralysis

Page 159: Brachial Plexus Injuries
Page 160: Brachial Plexus Injuries

History

The aetiology of the obstetric brachial plexusinjuries has an interesting history. As early as1764, Smellie suggested the obstetric origin of aparalysis of the arm in children. But only in 1872,in the third edition of his book De l’électrisationlocalisée et de son application à la pathologie età la thérapeutique, Duchenne de Boulognedescribed four children with an upper brachialplexus lesion as a result of an effort to deliverthe shoulder. The classical description by Erb in1874 concerned the upper brachial plexus paral-ysis in adults, with the same characteristics asthose described by Duchenne de Boulogne.Using electric stimulation, he found in healthypersons a distinct point on the skin in the supra-scapular region, just anterior to the trapeziusmuscle, where the same muscle groups could becontracted as those affected in his patients. It isthe spot where the fifth and sixth cervical rootsunite, and where they are optimally accessible toelectric current by virtue of their superficialposition. Pressure on this ‘point of Erb’, causedeither by fingers by traction on the armpits, byforceps applied too deep, or by a haematomawere for Erb, and many obstetricians after him,the only possible cause of the lesion.

But not everybody accepted the compressiontheory. Poliomyelitis and toxic causes werementioned. Some even pointed to the possibilityof an epiphysiolysis of the humerus, caused bycongenital lues, and consequently a paralysis ofthe arm. Doubts about the pressure theory,however, were raised as a result of observation ofHorner’s syndrome, indicating damage of thesympathical nerve, together with an injury of thelower plexus. Augusta Klumpke, the first femaleintern in Paris, explained in 1885 Horner’s sign inthe brachial plexus lesion by avulsions of theroots C8–T1 and involvement of the homolateral

cervical sympathic nervous system (Klumpke1885). Klumpke later married Dejerine, and there-fore the lower plexus palsy is sometimes calledthe Dejerine–Klumpke paralysis, as opposed tothe upper plexus palsy, which is named theErb–Duchenne paralysis. Thornburn (1903) wasone of the first to assume that the injury was the

16AetiologyJM Hans Ubachs and Albert (Bart) CJ Slooff

Figure 1

Engelhard’s photograph demonstrating the result of exces-sive stretching during the delivery (Engelhard 1906).

Page 161: Brachial Plexus Injuries

result of rupture or excessive stretching of thebrachial plexus during the delivery.

Pathogenesis

To test Thornburn’s assumption, Engelhardinvestigated the influence of different positionsand assisted deliveries on a dead fetus, in whichthe brachial plexus was dissected. In his doctoralthesis he demonstrated in 1906, with for thatperiod excellent photographs, that the pressuretheory was highly improbable (Fig. 1). Obstetricinjury of the brachial plexus could only be theresult of excessive stretching of that plexusduring the delivery. In particular, he warnedagainst strong downward traction of the fetalhead developing the anterior shoulder incephalic deliveries, and extensive lateralmovement of the body in breech extractions.And his words still have their validity. Morerecently, Metaizeau et al (1979) repeated thesestudies and explained the differences in injury.The results of these investigations have beenconfirmed by our clinical and surgical observa-tions (Ubachs et al 1995, Slooff 1997). Shoulderdystocia occurs mostly unexpected, and it is oneof the more serious obstetric emergencies. Theshoulder is impacted behind the symphysispubis, and although there is a long list ofmanoeuvres to disimpact the shoulder, not oneis perfect. Excessive dorsal traction, the firstreaction in that situation, bears the danger ofoverstretching with consequent damage of thebrachial plexus (Fig. 2). In breech presentation,even of small infants, the injury is caused bydifficulties in delivering the extended andentrapped arm and therefore a combination offorceful traction with too much lateral movementof the body.

Reconstructive neurosurgery of the obstetricbrachial plexus lesion, together with neurophys-iological and radiological investigation, gives theopportunity to gain a clear understanding of therelationship between the anatomical findingsduring operation and the obstetric trauma. Theinjury may be localized in the upper or lower partof the brachial plexus, resulting in differentphenotypes. Erb’s palsy results from an injury ofthe spinal nerves C5–C6 and sometimes C7. Itconsists of a paralysis of the shoulder muscles,

resulting in a hanging upper arm in endorota-tion, a paralysis of the elbow flexors and conse-quently an extended elbow in pronating position,caused by the paralysis of the supinators.Combination with a lesion of C7 results in aparalysis of the wrist and finger extensors andthe hand assumes the so-called waiter’s tipposition. The total palsy, often incorrectly calledKlumpke’s palsy, is caused by a severe lesion ofthe lower spinal nerves (C7–T1) but is alwaysassociated with an upper spinal nerve lesion ofvarying severity. The impairment mainlyincludes a paralysis of the muscles in forearmand hand, sometimes causing a characteristicclawhand deformity, and sensory loss of thehand and the adjacent forearm. Involvement ofT1 is frequently paralleled by cervical sympa-thetic nerve damage, an injury that will give riseto Horner’s syndrome.

Furthermore, stretching of the brachial plexusmay result in two anatomically different lesionswith different morbidities. The lesions are easilydistinguished during surgery. Either the nerve ispartially or totally ruptured beyond the vertebralforamen, causing a neuroma from expandingaxons and Schwann’s cells at the damaged site,or the rootlets of the spinal nerve are torn fromthe spinal cord, a phenomenon called anavulsion.

152 OBSTETRICAL PARALYSIS

Figure 2

Excessive dorsal traction in shoulder dystocia with conse-quent damage of the brachial plexus. (From Ubachs et al1995.)

Page 162: Brachial Plexus Injuries

Patients

Study of the first 130 patients, operated on fromApril 1986 to January 1994 in De Wever Hospital(today the Atrium Medical Centre) in Heerlen, TheNetherlands, offered the opportunity to proveEngelhard’s assersion in 1906. Moreover, it wasinteresting to determine whether the presentationof the fetus during the preceding delivery –breech or cephalic – contributed to the localiza-tion and anatomical severity of the lesion. Theresults of that study, the first where the anatom-ical site of the damage was compared with thepreceding obstetric events, were published in1995. The indication for neurosurgical interven-tion was based on the criteria from Gilbert et al(1987). The obstetric history was traced by analy-sis of the obstetric records made at the deliveryand compared much later with the anatomicalfindings at surgery. Demographic and obstetricdata regarding a large proportion (146 533) of the196 700 deliveries in The Netherlands in 1992were obtained from The Foundation of PerinatalEpidemiology in The Netherlands (PEN) and theDutch Health Care Information Centre (SIG).These data were used to identify specific featuresin the study population (Table 1).

Of the operated infants with obstetrics brachialplexus lesions (OBPLs), 102 were born incephalic and 28 in breech position. Patients whohad been delivered in cephalic presentation wereborn more frequently from a multiparous

mother, were more frequently macrosomic,experienced intrapartum asphyxia more oftenand required instrumental delivery more often.Patients born in breech differed from the refer-ence population by a higher incidence of intra-partum asphyxia. The gestational age at birth didnot differ significantly.

In one-third (40/130) of the OBPL population,the preceding pregnancy had been complicatedby treated gestational diabetes, the suspicion ofidiopathic macrosomia (percentile of birth weightfor gestation ≥ 90), obesity and even the explicitwish to give birth in a standing position, a strat-egy which tends to aggravate mechanicalproblems encountered during the second stage.Two-thirds (87/130) of the infants with OBPLswere delivered by multiparous mothers and, inalmost half of them (39/87) macrosomia, instru-mental delivery and/or other potentiallytraumatic manipulations had complicated thesecond stage of labour. Whereas the cephalicgroup was characterized by a disproportionatenumber of macrosomic infants, the distributionof the percentile of birth weight for gestation inthe breech group did not differ significantly(Table 1 and Fig. 3). The mean neonatal weightof the children born in the cephalic position was4334 g with a range from 2550 to 6000 g. Infantsborn by breech weighed a mean 3050 g with arange from 1230 to 4000 g. In spite of thismarked weight difference, the incidence ofmechanical problems during passage of the birth

AETIOLOGY 153

Table 1 Demographic and obstetric characteristics of the two obstetric brachial plexus lesion (OBPL) populations inrelation to their respective reference populations. Values are given as percentages (From Ubachs et al 1995)

Cephalic delivery Breech delivery

Characteristics OBPL Control OBPL Control(n = 102) (n = 138 702) P (n = 102) (n = 7926) P

ProportionMultipara 75 56 < 0.05 39 44 NSMales 50 52 NS 46 46 NS

IncidencePre-term birth* 7 14 NS 21 33 NSPost-term birth 7 5 NS 7 3 NSSmall for dates (≤ 10%)** 0 10 < 0.001 14 18 NSLarge for dates (≥ 90%)** 71 10 < 0.0001 4 6 NSBirth asphyxia (Apgar score ≤ 6) 65 1 < 0.0001 86 4 < 0.0001Caesarean birth* 0 7 < 0.01 0 38 < 0.001Forceps/vacuum birth 49 11 < 0.001 0 1 NS

*In the breech reference group the incidence of preterm deliveries and that of Caesarean sections was higher than in the cephalicreference group (P < 0.05, �2 test). **According to Dutch intrauterine growth curves (Kloosterman 1970). NS, not significant.

Page 163: Brachial Plexus Injuries

canal and that of intrapartum asphyxia (1 minApgar score ≤ 6) was similar in the two groups(Table 2). It is uncertain whether the asphyxiawas caused by the difficulty in delivery, or if itwas one of the factors in the nerve damage bycausing muscular hypotonia. Obviously, excessmacrosomia in the cephalic group explains thehigh incidence of shoulder dystocia. It is inter-esting that twice as many right- than left-sidedinjuries were observed in the children deliveredin vertex presentation. This is most likely to be adirect consequence of fetal preference for aposition with the back to the left side, and hencea vertex descent in a left occipital anteriorpresentation (Hoogland and de Haan 1980). Thepreference for the right side was also noted forthe breech group. However, this was not signifi-cant, possibly because of the smaller group size(Table 3).

An unexpected finding was the difference inclinical and anatomical type of lesion betweenthe children born in breech and cephalic presen-tations (Table 4 and Fig. 4). Mechanically, a diffi-cult breech delivery with often brusquemanipulation to deliver the first arm, togetherwith excessive traction on the entire neck wasexpected to predispose towards more extensivedamage reflected in the Erb’s type C5–C7 or thetotal C5–T1 lesions. Similarly, overstretching bytraction and abduction in an attempt to deliverthe first shoulder was expected to predispose forC5–C6 damage. To our surprise, two-thirds(19/28) of the injuries after breech deliveryconsisted of pure Erb palsies (C5–C6) caused, inthe majority of cases (16/19), by a partial orcomplete avulsion of one or both spinal nerves.Total lesions were rare in the breech group.Conversely, the most common lesion after

154 OBSTETRICAL PARALYSIS

Table 2 Traumatic birth and intrapartum asphyxia in thetwo birth groups. Values are given as n (%). Differences(P) not significant

Cephalic Breech(n = 102) (n = 28)

Complicated 2nd stage* 92 (90) 22 (79)Intrapartum asphyxia 66 (65) 24 (86)

*Shoulder dystocia or difficult breech extraction.

Table 3 Incidence of the left- and right-side lesions:cephalic birth (n = 102) and breech (n = 28). Values aregiven as n (%)

Birth group Left side Right side P

Cephalic 37 (36) 65 (64) < 0.01Breech* 10 (36) 18 (64) NS

*Several of these infants had a bilateral OBPL. The operatedlesion is mentioned. NS: not significant.

Figure 3

The weight at birth of 130 childrenwith OBPLs.

Number of patients

<10 10–25 25–50 50–75 75–90 90–95 >95Percentile of birth weight for gestation

Breech Cephalic

50

40

30

20

10

0

Page 164: Brachial Plexus Injuries

cephalic birth was the more extensive Erb’s palsy(C5–C7) usually resulting from an extraforaminalpartial or complete nerve rupture, closelyfollowed by the total palsy. In fact, a total palsywas an almost exclusive complication (43/45) ofcephalic delivery, with nerve rupture and nerveavulsion seen equally frequently. Interestingly, if

in this group the lesion was not total (C5–T1), thedamage was always more severe as indicated bythe incidence of nerve rupture. Apparently,unilateral overstretching of the angle of neck andshoulder in the cephalic group led to a moreextensive damage, including the lower spinalnerves of the plexus.

An explanation of this phenomenon might besought in tight attachment of the spinal nerves C5and C6 to the transverse processes of the cervi-cal vertebrae (Sunderland, 1991). As a result ofthat, unilateral overstretching in shoulder dysto-cia preferentially leads to an extraforaminallesion of the upper spinal nerves and often to anavulsion of the lower spinal nerves C8–T1 fromthe spinal cord. A different causal mechanism,however, should be considered in difficult breechdeliveries (Slooff and Blaauw, 1996). Hyper-extension of the cervical spine and consequentlya forced hyperextensive moment or elongation ofthe spinal cord in such a delivery, combined withthe relatively strong attachment of the spinalnerves C5 and C6 to their transverse processes,might cause an avulsion by acting directly on thenerve roots between their attachment to the cordand their fixed entry in the intervertebralforamen. Sunderland calls this the ‘centralmechanism’ of an avulsion (Sunderland 1991,Fig. 18.7, p. 157).

Associated lesions were frequent. Fractures ofthe clavicle or the humerus were evenly

AETIOLOGY 155

Table 4 Effect of presentation at birth on type andseverity of the OBPL birth groups. Values are given aspercentages (From Ubachs et al 1995)

Cephalic BreechType of lesion (n = 102) (n = 28) P

Erb C5–C6Avulsion* 2 57 < 0.0005Rupture 5 11 NSTotal: 7 68 < 0.0005

Erb C5–C7Avulsion* 8 18 NSRupture 43 7 < 0.0005Total: 51 25 < 0.05

Total lesion C5–T1Avulsion* 20 4 < 0.05Rupture 22 3 < 0.05Total: 42 7 < 0.005

Any lesionAvulsion* 29 79 < 0.0005Rupture 71 21 < 0.0005

*At least one spinal nerve.NS: not significant.

Figure 4

Presentation at birth, morbidity andtype of lesion in 130 children.(From Ubachs et al 1995)

Breech

Avulsion Rupture

Erb C5–C6

Erb C5–C7

Total lesion C5–T1

Cephalic

40 30 20 10 0 10 20 30 40 50 60

Page 165: Brachial Plexus Injuries

distributed over the two groups, whereas persis-tent paralysis of the phrenic nerve was notedmore frequently in infants born by breech andbilateral OBPL was seen exclusively after abreech delivery (Table 5).

Intrauterine maladaptation was neversuspected, as no infant in these series was bornby Caesarean section and all vaginal deliverieswere either operative or were complicated byother potentially traumatic manipulations. ACaesarean section, for that matter, is not alwayssafe and atraumatic: especially in malpositions, aCaesarean delivery can be extremely difficult. Asearly as 1980, Koenigsberger found in neonateswith plexus injuries whose deliveries wereuncomplicated, in the first days of lifeelectromyographic changes characteristic ofmuscle denervation, which, in adults, take at least

10 days to develop. In neonates denervationactivity is found much earlier, in our experiencealready after 4–5 days (see Chapter 4). It is there-fore dificult to prove intrauterine maladaptationas a cause of nerve injury. This would demandelectromyographic investigation within the firstdays after the delivery. Study of the aetiology,and the anatomic injury as its consequence,should teach a lesson. As already said, shoulderdystocia is not always predictable. Estimation ofthe child’s birth weight is inaccurate. The averagedifference between the estimated weight beforedelivery and the birth weight is, independent ofthe method used, about 15–20 per cent. But evenassuming a 100 per cent precision in predicting abirth weight of > 4500 g estimations are that from58 to 1026 Caesarean deliveries would be neces-sary to prevent a single, permanent brachialplexus injury (Sacks and Chen 2000). There aremany obstetric measures and manoeuvresdescribed to overcome a shoulder dystocia.However, the crucial factor is that every midwifeor obstetrician should have a well-conceived planof action, which can be executed rapidly.Computer techniques to measure the forces usedin shoulder dystocia have been developed (Allenet al 1994). In future, they might be used as amodel for obstetricians in training to teach thehandling of such a difficult and frequentlyunexpected problem.

The realization of the risk of birth trauma inbreech presentation (and its legal consequences)

156 OBSTETRICAL PARALYSIS

Table 5 Incidence of associated lesions in the two birthgroups. None of the children had a spinal cord or facialnerve lesion. Values are given as n (%) (From Ubachs etal 1995)

Cephalic BreechAssociated lesions (n = 102) (n = 28) P

Sternocleidomastoid 9 (8) 5 (18) NSFracture

Clavicle 9 (9) 3 (11) NSHumerus 6 (6) 2 (7) NS

Phrenic nerve lesion 3 (3) 10 (36) < 0.0005Bilateral OBPL 0 (0) 7 (25) < 0.0005

NS: not significant.

Figure 5

The ‘central mechanism’ of anavulsion (Sunderland 1991).

Page 166: Brachial Plexus Injuries

has made the number of Caesarean sections forthat position in the Netherlands rise from 28.4per cent in 1990 to 46.2 per cent in 1997. Thisnumber undoubtedly will increase inversely tothe consequential lack of experience of theobstetrician.

The recent international study by Hannah et al(2000), involving 2083 women in 26 countries,confirmed that planned Caesarean section for theterm fetus in breech presentation is better thanplanned vaginal birth, with similar maternalcomplications between the two groups.

Conclusion

The high number of abnormal precedingpregnancies or deliveries in the group of multi-parous women suggests the risk of recurrence.Consequently, a multiparous woman with ahistory of mechanical problems during a previ-ous delivery and with her current pregnancycomplicated by even the suggestion of fetalmacrosomia should alert the obstetrician torecurrent mechanical complications during deliv-ery. If the fetus is in a cephalic presentation, avaginal birth can be anticipated, althoughabdominal delivery should be considered if anydelay develops in the first stage. On the otherhand, if the fetus is in breech presentation, aprimary Caesarean section seems recommend-able to circumvent the markedly elevated risk formechanical injury during vaginal birth.

References

Allen RH, Bankoski BR, Butzin CA, Nagey DA (1994)Comparing clinician-applied loads for routine, difficult,and shoulder dystocia deliveries, Am J Obstet Gynecol1971:1621–7.

Duchenne G (1872) De l’électrisation localisée et de sonapplication à la pathologie et à la thérapeutique. JBBaillière et fils: Paris: 357–62.

Engelhard JLB (1906) Verlammingen van den plexusbrachialis en n. facialis bij het pasgeboren kind.(Doctoral thesis) P. Den Boer: Utrecht.

Erb W (1874) Uber eine eigentümliche Lokalisation vonLähmungen im Plexusbrachialis, Verhandl Naturhist

Med Vereins. Carl Winters’ Universitats Buchhandlung:Heidelberg: Vol 2:130–6.

Gilbert A, Hentz VR, Tassin FL (1987) Brachial plexusreconstruction in obstetric palsy: operative indicationsand postoperative results. In: JR Urbaniak, ed.Microsurgery for Major Limb Reconstruction. CVMosby: St. Louis: 348–64.

Hannah ME, Hannah WJ, Hewson SA et al (2000)Planned caesarean section versus planned vaginalbirth for breech presentation at term: a randomisedmulticentre trial, Lancet 356:1375–83.

Hoogland HJ, de Haan J (1980) Ultrasonographicplacental localization with respect to foetal position inutero, Eur J Obstet Gynecol Reprod Biol 11:9–15.

Kloosterman GJ (1970) On intrauterine growth, Int JGynaecol Obstet 6:895–912.

Klumpke A (1885) Contribution à l’étude des paralysiesradiculaires du plexus brachial, Rev Méd (Paris)5:591–616, 738–90.

Koenigsberger MR (1980) Brachial plexus palsy at birth:intrauterine or due to delivery trauma?, Ann Neurol8:228.

Metaizeau JP, Gayet C, Pleriat F (1979) Les lésionsobstétricales du plexus brachial, Chir Pediatr 20:159–63.

Sacks DA, Chen W (2000) Estimating fetal weight in themanagement of macrosomia, Obstet Gynecol Surv55:229–39.

Slooff ACJ (1997) Obstetric brachial plexus lesions. In:Boome RB, ed. The Brachial Plexus ChurchillLivingstone: New York: 89–106.

Slooff ACJ, Blaauw G (1996) Some aspects of obstet-ric brachial plexus lesions. In: Alnot JY, Narakas A, eds.Traumatic Brachial Plexus Injuries. ExpansionScientifique Française: Paris: 265–7.

Smellie W (1764) A Collection of Preternatural Casesand Observations in Midwifery. Vol III. Wilson andDurham: London: 504–5.

Sunderland S (1991) Nerve Injuries and their RepairChurchill Livingstone: Edinburgh: 151–8.

Thorburn W (1903) Obstetrical paralysis, J ObstetGynaecol Br Emp 3:454–8.

Ubachs JMH, Slooff ACJ, Peeters LLH (1995) Obstetricantecedents of surgically treated obstetric brachialplexus injuries, Brit J Obstet Gynaecol 102:513–17.

AETIOLOGY 157

Page 167: Brachial Plexus Injuries
Page 168: Brachial Plexus Injuries

Introduction

It is self-evident that the child with a suspectedbrachial plexus lesion should be examined asearly as possible in order to make a definitivediagnosis, to begin recording the naturalprogression of recovery and to initiate educationand support for the family. For example,problems such as positional torticollis can betreated effectively if early intervention, includingappropriate positioning, is undertaken.

In order to illustrate our approach to theseinfants, the methods for and timing of examina-tion of brachial plexus lesions in newborns, theprognosis regarding primary surgical interven-tion and the assessment of surgical outcomeswill be discussed.

Initial evaluation

History

A careful obstetrical history should be obtained.Parents are routinely questioned about previouspregnancies and deliveries, the history of thecurrent pregnancy including diabetes and toxaemia,the duration of labour and method of delivery.Further enquiries outline the early postnatal periodincluding respiratory difficulties, evidence offractures or Horner’s syndrome and the extent ofthe paralysis seen in the first few days of life. Oftenthe most difficult data to retrieve concern themechanism of the delivery itself. This information issometimes sketchy and may not be reliable inattempting to reconstruct the birth history.

The parents can, in some cases, give anextremely detailed account of the early recoveryof movement in the limb. This information

provides the introduction to a detailed examina-tion of the infant.

Physical examination

Physical examination of the newborn should bethorough in order to rule out other diagnosesand determine the full extent of possible birthtrauma. Observation of the position of the head,neck and arms gives useful clues to underlyingpathology (Fig. 1). The sternocleidomastoid

17Examination and prognosisHoward M Clarke and Christine G Curtis

Figure 1

The typical posture of a 6-week-old infant with a right uppertrunk (Erb’s) palsy. The extremity is held adducted at the sidewith the elbow straight. The wrist, fingers and thumb areflexed, and the infant often looks away from the affectedside. (From Clarke and Curtis 1995, with permission.)

Page 169: Brachial Plexus Injuries

muscles are palpated to determine if a pseudo-tumour is present or if a muscle is shortened.Many have noted the tendency for infants withbrachial plexopathy to turn the head away fromthe involved arm. If left unchecked this can leadto a contracture of the shortened sternocleido-mastoid muscle and a true torticollis candevelop.

Careful attention should be given to theposition of the affected arm of the child. Theclassic position of Erb’s palsy resulting frominvolvement of the upper roots is adduction andinternal rotation of the shoulder, extension of theelbow, pronation of the forearm and flexion ofthe wrist and fingers. This typical posture mayalso occur in the absence of elbow extensionsince gravity holds the arm at the side of thesupine infant. Total palsy is characterized bycomplete atonia of the extremity (Fig. 2). Thefingers may rest in a flexed posture, which is theresult of the tenodesis effect at the wrist ratherthan true power in the long flexors. Sensationmay be absent, although this is difficult to test inan infant. Some arm movement may occur as aresult of shoulder elevation, and this should notbe confused with true shoulder joint movement.Klumpke’s paralysis is extremely rare in obstet-rical injuries (Al-Qattan et al 1995), but would bediagnosed when paralysis of the hand is

observed in the presence of normal shoulder andelbow movement.

Palpation of the clavicles, humeri and ribs forfractures is part of a thorough examination.These fractures can produce a pseudoparalysissimilar in initial presentation to a true brachialplexus lesion. Pseudoparalysis is caused bycompression of the brachial plexus by thefractured bone, by swelling around the plexus, orby involuntary splinting of the arm in thepresence of pain but in the absence of directinjury to the plexus itself. Characteristically,pseudoparalysis resolves more rapidly than atrue obstetrical lesion of the plexus. Plain X-raysmay be indicated to rule out fractures.Dislocation of the shoulder has also been associ-ated with true obstetrical brachial plexus palsy(Stojčević-Polovina 1986, Eng 1971).

Observation of the abdomen for symmetricaldiaphragmatic movement may help to indicatewhether phrenic nerve paralysis has occurred.Fluoroscopy is probably the best single test toassess diaphragmatic function. Formal investiga-tion of the position of the diaphragm shouldalways be undertaken prior to surgery in case thepatient develops respiratory difficulties followingsurgery, typically an increased frequency andseverity of upper respiratory tract infections thenext winter. A paralysed hemidiaphragm pre-dating surgery may require plication to improvefunction. If the diaphragm was of normal excur-sion before surgery and is paralysed postopera-tively, it may recover by the next winter season,sparing the child the need for plication.

The eyes are inspected for the signs ofHorner’s syndrome, especially in the presence oftotal paralysis. The four signs seen in Horner’ssyndrome are ptosis, myosis, enophthalmos andanhydrosis on the ipsilateral face. These findingsare taken as indications of proximal injury(usually avulsion) of the lower trunk, as originallydescribed by Klumpke in adult injuries (Klumpke1885). She found that the Horner’s resulted fromavulsion of T1, which disrupts the communicat-ing branch supplying sympathetics to the stellateganglion.

Assessment of motor function

The most challenging aspect of the assessmentof the newborn infant with paralysis of the

160 OBSTETRICAL PARALYSIS

Figure 2

In a 6-month-old patient with a total plexus lesion frombirth, the signs of denervation of the hand are seen withan intrinsic minus claw hand. No active extension of thefingers or thumb was seen, but flexion was full. (FromClarke and Curtis 1995, with permission.)

Page 170: Brachial Plexus Injuries

upper extremity is to determine a practical andreliable method for quantitating motorfunction. The infant cannot cooperate, therange of motor movement normally seen inyoung infants does not match that of the adult,and the power of even a normal infant limb isdwarfed by that of the adult examiner. Inaddition, we have need of an assessment toolthat readily discriminates between scores thatindicate the possibility for useful function andthose which suggest that the function achievedby spontaneous recovery will be of little valueto the child.

In 1943, the British Medical Research Council(MRC) suggested a system of recording power inpatients with peripheral nerve lesions (Aids tothe Investigation of Peripheral Nerve Injuries,1943) (Table 1). The administration of this testwas dependent on the patient understanding thenature and object of the examination. Thesystem as originally described failed to distin-guish whether active movement was through afull or partial range of motion. In current usagethis test is often modified to require that fullrange of movement be obtained to score Grades2 through 5.

Although some authors (Boome and Kaye1988, Laurent and Lee 1994) utilize the MRCscale for assessment of motor power in infantswith brachial plexus lesions, others have recog-nized the limitations of evaluating youngpatients with this system. Infants will only rarelyuse full power when being examined. Gilbertand Tassin suggested a modified British MedicalResearch Council classification, shown in Table2, simplifying it to account for the difficulties ofexamining infants (Gilbert and Tassin 1987). M2in this scale covers a wide range of active

movements, beginning with slight movementwith gravity eliminated and progressing to nearfull range of motion against gravity. This makesthe scale difficult to use in assessing outcomessince most results typically fall in the M2category and substantial improvements may notbe documented.

Like Gilbert and Tassin, we have found it diffi-cult to administer the MRC scale in infants, whocannot be expected to cooperate in demon-strating full voluntary power of individualmuscles. In our experience, the M0–M3 scaledoes not accurately reflect the improvements inmotor recovery seen in these children. Forthese reasons we have developed our ownscale for assessing active movement in theupper extremities of infants and young childrenwith brachial plexus lesions. The ActiveMovement Scale (Table 3) is an eight-pointscale designed to capture subtle and significant

EXAMINATION AND PROGNOSIS 161

Table 1 Medical Research Council Muscle Grading System

Observation Muscle grade

No contraction 0Flicker or trace of contraction 1Active movement, with gravity eliminated 2Active movement against gravity 3Active movement against gravity and resistance 4Normal power 5

Data from Aids to the Investigation of Peripheral Nerve Injuries(British Medical Research Council 1943).

Table 2 Gilbert and Tassin Muscle Grading System

Observation Muscle grade

No contraction M0Contraction without movement M1Slight or complete movement with M2

weight eliminatedComplete movement against the M3

weight of the corresponding segment of extremity

Data from Gilbert and Tassin (1987).

Table 3 Hospital for Sick Children Muscle Grading System

Observation Muscle grade

Gravity eliminatedNo contraction 0Contraction, no motion 1Motion � 1⁄2 range 2Motion > 1⁄2 range 3Full motion 4

Against gravityMotion � 1⁄2 range 5Motion > 1⁄2 range 6Full motion 7

Full active range of motion with gravity eliminated (Muscle Grade4) must be achieved before active range against gravity is scored(Muscle Grades 5–7). (From Clarke and Curtis 1995, withpermission.)

Page 171: Brachial Plexus Injuries

changes in movement in the arm. A full scoreof 7 does not necessarily reflect full musclestrength, as the scale represents activemovement only. To our knowledge, no reliablemethod of testing true muscle power or resis-tance in infants exists.

There are a number of advantages in using theActive Movement Scale. It can be used to grademovement in infants and young children, anddoes not require the child to perform tasks oncommand. Overall joint movements are evalu-ated in contrast to individual muscle testing,which may be difficult to perform in infants.Smaller changes in movement can be detected,and it can be used as a preoperative as well aspostoperative evaluation tool.

We have developed the following guidelines inan effort to standardize the use of the ActiveMovement Scale:

1. A score of 4 must be achieved (full range ofmotion with gravity eliminated) before ahigher score can be assigned. This clarifiesscoring when limited movement is presentboth with gravity eliminated and againstgravity;

2. Movement grades are assigned within theavailable range of passive motion. Forexample, if a flexion contracture is present atthe elbow, full range of extension is scored ifthe elbow can be extended to the limits of thecontracture;

3. Movement grades are assessed within theage-appropriate range of motion as assessedin the contralateral limb. For example,newborn infants normally do not flex theshoulder a full 90° above the horizontal. Theuninvolved limb should be used as a controlto estimate the extent of available normalrange (Fig. 3);

4. Extension of the digits is assessed at themetacarpophalangeal joints. Flexion of thedigits is evaluated by observing the distanceat rest between the finger tips and the palmand then observing the active motion as afraction of that distance both without andagainst gravity;

5. Digital flexion or extension is given a singlegrade by using the movement score of the bestdigit. For example, if the index finger scores agrade of 7 for flexion and the other digits score2, then the finger flexion score is 7.

Assessment using the Active Movement Scale isperformed with the upper body and arms of theinfant exposed. Ideally, the child is placed on aflat, firm surface where he can move or roll. Avariety of toys to stimulate movement should beavailable (Fig. 4). Gravity-eliminated movementsare assessed first to determine if higher scorescan then be assigned. For example, to gradeshoulder flexion the child is placed in thegravity-eliminated position of side-lying with theaffected arm uppermost. A toy is placed withinthe child’s view and moved in a way to attractattention. Tactile stimulation of the arm usingthe toy followed by movement of the toy in aforward direction draws attention to the arm andencourages flexion of the shoulder. The anteriordeltoid region of the shoulder is palpated todetect flickers of movement if minimal activemovement is seen. If less than full range ofavailable passive movement is obtainedcompared to the normal side, then a score of 3or lower is given. If full range of forward flexionis obtained (giving a score of 4), the child isplaced in a supported sitting position to viewmovement against gravity. Again the child isencouraged to reach forward for an object. Anagainst-gravity score of 5 or more is assigned

162 OBSTETRICAL PARALYSIS

Figure 3

By presenting the same stimulus to both the normal andabnormal sides (though not of necessity simultaneously asshown here), a direct comparison can be made of therange of motion obtained. Here supination to neutral isseen on the affected right side and no finger or thumbextension. (From Clarke and Curtis 1995, with permission.)

Page 172: Brachial Plexus Injuries

depending on the greatest range of motionobserved. In this way, all joint movements arescored after observation in gravity-eliminatedand against-gravity positions. Parents may alsoparticipate in encouraging movement if a childis especially anxious with strangers. Withpractice, all joint movements can be graded byobservation of play in three positions: supine,side-lying and sitting.

Scores are given for the following jointmovements: shoulder flexion, abduction,

adduction, internal rotation and externalrotation; elbow flexion and extension; pronationand supination of the forearm; wrist flexion andextension; finger flexion and extension; andthumb flexion and extension. These scores arerecorded at the initial assessment and at 3-monthly intervals in the first year of life or untilsurgery intervenes. They are also used postop-eratively to evaluate the results of surgery. Theadvantage of this system is that a small amountof movement against gravity is not sufficient toyield a high score in situations where it may beof limited functional value. The disadvantagesare the time and practice required to carry outthis technique successfully and the difficulty indetermining the effect of gravity on suchmotions as finger flexion.

Curtis has demonstrated the reliability of theActive Movement Scale in a two-part study(Curtis 2000). Part A was an inter-rater reliabil-ity study in which two physiotherapists, experi-enced in using the scale, separately assessed 63infants with obstetrical brachial plexus palsy.Part B examined the dispersion of ActiveMovement Scale scores of infants with obstet-rical brachial plexus palsy as evaluated byphysiotherapists with varying levels of priorexperience after a single training session.Overall quadratic weighted kappa analysis inPart A demonstrated that the raters’ scoreswere at the highest level of agreement(Kquad = 0.89). Part B established that thevariability of scores due to rater factors, waslow compared with patient factors, and that thevariation in scores due to rater experience wasminimal. The Active Movement Scale is areliable tool for the evaluation of infants up to1 year of age with obstetrical brachial plexuspalsy when raters are trained in the use of thescale.

Another approach to the evaluation of childrenwith brachial plexus lesions is to assess globalmovement of the extremity and look at patternsof movement that may be either functional ormaladaptive. Such a grading scale has beenestablished by Mallet (Mallet 1972) (Fig. 5), andis commonly used. The disadvantage of thissystem is that it is practicable only with childrenof 3–4 years of age, who can reliably performvoluntary movements on command. Recordingthe natural history of recovery in infant patientswith this system is difficult.

EXAMINATION AND PROGNOSIS 163

Figure 4

Bright toys with rattles and bells were used to attract theattention of this 5-month-old infant to the affected side.Stroking the forearm or hand with the toy will often elicita motor response. (From Clarke and Curtis 1995, withpermission.)

Page 173: Brachial Plexus Injuries

Assessment of sensory function

The assessment of sensation in infants isextremely difficult. In many cases it is only possi-ble to determine if the child responds to painfulstimuli and to examine for the signs of self-mutilation, which in children can indicatedecreased sensory awareness. Narakas hasclassified the sensory response in infants into

four grades which can be used to collect descrip-tive data (Narakas 1987) (Table 4). Narakas quali-fies the scale by stating that the recovery ofsensation is capricious and that the sensory scalemay not consistently indicate the clinical progressof the lesion. Distinguishing between S1 and S2can be difficult. In a completely paralysed limb,only the reaction to painful stimuli (S1) can beusefully evaluated.

164 OBSTETRICAL PARALYSIS

II

Inferior to 30°

Inferior to 20°

30° to 90° Superior to 90°

Superior to 20°

Impossible Difficult Easy

Clarion Small clarion

HA

ND

TO

MO

UT

HH

AN

DT

OB

AC

KE

XT

ER

NA

LR

OT

AT

ION

AC

TIV

E

AB

DU

CT

ION

HA

ND

TO

NA

PE

OF

N

EC

K

S1 T12Impossible

III IVFigure 5

Mallet’s classification offunction in obstetricalbrachial plexus palsy.Grade 0 (not shown) is nomovement in the desiredplane and Grade V (notshown) is full movement.(From Gilbert 1993, withpermission.)

Page 174: Brachial Plexus Injuries

Classification

The most complete anatomical classification forbrachial plexus injuries includes the followingcategories: upper plexus palsy (Erb’s) involvingC5, C6 ± C7 (Erb 1874); intermediate plexuspalsy involving C7 ± C8, T1 (Al-Qattan andClarke 1994); lower plexus palsy (Klumpke’s)involving C8, T1 (Klumpke 1885) and total plexuspalsy involving C5, C6, C7, C8 ± T1 (Terzis et al1986). In infants with obstetrical injuries, Gilbertfound that two clinical appearances predomi-nated in 1000 babies examined 48 hours afterbirth; paralysis of the upper roots and completeparalysis. Klumpke’s paralysis with isolated

involvement of the distal roots was not seen(Gilbert et al 1991).

Narakas has graded infants with obstetricalbrachial plexus palsy based on the clinicalcourse during the first 8 weeks of life (Narakas1986) (Table 5). The classification is not anatom-ical but grades the overall severity of the lesionand implies a progressive degree of injury withincreasing force applied at the time of delivery.Clinical types are assigned as follows. Type I ismild and heals in a few weeks. Type II shows anunpredictable prognosis in the first few weeks.Usually the shoulder does not recover but theelbow functions satisfactorily. Some of thesepatients do not recover wrist and finger extensionand require tendon transfers. Type III involves theupper trunk, has avulsion of C7 and a stretchinjury of the lower trunk. These may appearcomplete at birth with temporary Horner’ssyndrome. Type IV includes avulsion of C8 andT1 and persisting Horner’s syndrome. Significantrecovery of C5 and C6 function may occur,however. Type V shows severe injury involvingall nerve roots. The Horner’s sign is permanent,which, along with paralysis of the rhomboids,levator scapulae and serratus anterior, is a signof a poor prognosis. Narakas classification is

EXAMINATION AND PROGNOSIS 165

Table 4 Narakas Sensory Grading System

Observation Sensory grade

No reaction to painful or other stimuli S0Reaction to painful stimuli, none to touch S1Reaction to touch, not to light touch S2Apparently normal sensation S3

(Adapted from Narakas 1987, with permission.)

Table 5 Narakas Classification of Obstetrical Brachial Plexus Palsy

Clinical picture Pathology grades Recovery(Sunderland 1951)

Type I C5–C6 1 & 2 Complete or almost in 1–8 weeksType II C5–C6 Mixed 2 & 3 Elbow flexion: 1–4 weeks

Elbow extension: 1–8 weeksC7 Mixed 1 & 2 Limited shoulder: 6–30 weeks

Type III C5–C6 4 or 5 Poor shoulder: 10–40 weeksElbow flexion: 16–40 weeks

C7 2 or 3 Elbow extension: 16–20 weeksWrist: 40–60 weeks

C8–T1 1 Hand complete: 1–3 weeks(No Horner’s sign)

Type IV C5–C7 4 and/or 5 Poor shoulder: 10–40 weeksElbow flexion: 16–40 weeks

C8 Mixed 2–3 Elbow extension incomplete, poor: 20–60 weeks or nilT1 1 and 2 Wrist: 40–60 weeks(Temporary Horner’s sign) Hand complete: 20–60 weeks

Type V C5–C7 5 Shoulder and elbow as aboveC7 or avulsedC8 3 or avulsed Wrist poor or only extension: poor flexion or noneT1 2 and 3C8–T1 Avulsed Very poor hand with no or weak(Horner’s sign usually present) flexors and extensors; no intrinsics

(From Narakas 1986, with permission.)

Page 175: Brachial Plexus Injuries

extremely valuable in providing clues to progno-sis in the first 2 months of life. A further studyusing statistical methods to verify these prognos-tic factors would be highly informative.

Prognosis for recovery

Although many infants with plexopathy recoverwith minor or no residual functional deficits, anumber of children do not regain sufficient limbfunction and subsequently develop functionallimitations, bony deformities and joint contrac-tures. In a thorough study by Bager, half of 52consecutive patients had normalized handfunction on clinical assessment at 6 months ofage but half had identifiable residual impairmentsat 15 months of age (Bager 1997). Furthermore,Bellew et al have found that children withbrachial plexus palsy, regardless of severity,showed more behavioural problems than norma-tive data would suggest (Bellew et al 2000). Thechildren with more severe palsies had even morebehavioural problems and scored less well ondevelopmental assessment. Determining whichinfants may develop such sequelae is of obviousimportance in planning therapy.

Opinion varies widely on the spontaneousrecovery of children with obstetrical brachialplexus palsy. Nonetheless, the majority ofpatients do well and do not require primarysurgical intervention (Greenwald et al 1984,Jackson et al 1988, Piatt 1991, Michelow et al1994). The lack of a uniform system for compar-ing outcome makes comparison of publishedstudies difficult.

While all of the factors discussed above provideuseful insights, our real need is to understand thenatural history of this condition sufficiently topredict, at a few months of age, the probableoutcome and the need for surgery. Ultimately, alarge series of patients studied in a statisticallysound manner will be necessary to provide securepoints of reference. In our own attempt to under-stand these factors we have reviewed the recordsof the Brachial Plexus Clinic at the Hospital forSick Children (Michelow et al 1994). Included were28 patients (42 per cent) with upper plexusinvolvement and 38 (58 per cent) with totalplexopathy. Sixty-one patients (92 per cent) recov-ered spontaneously and five patients (8 per cent)

required primary brachial plexus exploration andreconstruction. Observations of shoulder abduc-tion and adduction, as well as flexion and exten-sion at the elbow, wrist, thumb and fingers, wererecorded at or close to 3, 6, 9 and 12 months ofage. A record of the natural history of obstetricalbrachial plexus palsy from birth to 12 months ofage was generated (Fig. 6). Inspection of thegraph demonstrated an early improvement inlimb movement in the patients who recoveredspontaneously in contrast to patients with severeplexopathy requiring surgery.

Adapting the classification of Narakas, poorrecovery was defined as elbow flexion of half orless than half the normal range and shoulderabduction of less than half the normal range(Narakas 1985). Recovery was otherwise consid-ered to be good. Each patient was then classifiedinto either a good recovery group or poor recov-ery group, based on their scores at 12 months ofage. The assignment was made based onspontaneous recovery alone and not on whethersurgery was undertaken.

Stepwise discriminant analysis (SAS: PROCSTEPDISC with stepwise option (SAS User’sGuide: Statistics 1985)) was used to study which

166 OBSTETRICAL PARALYSIS

Lim

b m

otio

n sc

ore

2018

161412

10864

20

Operated group

Age (months)0 3 6 9 12

Mean

Figure 6

The natural history of obstetrical brachial plexus palsy asevaluated by Limb Motion Scores is depicted (solid line) forpatients recovering spontaneously in their first year of life(n = 66). The shaded area represents ± 1 SD of variationfrom the mean score. Early improvement in limb movementis seen in patients who recovered spontaneously. The meanscore for the operated patients has been plotted separately(dashed line) and shows a slower, less remarkable improve-ment. (From Michelow et al 1994, with permission.)

Page 176: Brachial Plexus Injuries

parameters at birth and 3 months were usefulpredictors of the two recovery groups at 12months. The significant parameters were thenanalyzed using discriminant analysis (SAS: PROCDISCRIM (SAS User’s Guide: Statistics 1985)).

The analysis demonstrated that a number ofparameters were highly significant in their abilityat 3 months to predict subsequent recovery at 12months (Table 6). Elbow flexion at 3 monthsincorrectly predicted recovery in 12.8 per cent ofcases (Table 7, Fig. 7). When appropriate param-eters were combined, particularly elbow flexionwith elbow, wrist, thumb and finger extension;recovery was incorrectly predicted in only 5.2 percent of cases (Table 7, Fig. 8).

Indications for surgery

Many authors agree that attempts to avoid perma-nent sequelae necessitate intervention early in thefirst year of life in appropriate cases (Terzis et al1986, Kawabata et al 1987, Alanen et al 1990). Isit possible, therefore, to predict by 3 months ofage whether or not a child will spontaneouslyrecover sufficiently to avoid unnecessary primaryplexus surgery? While clinical examination is the

best single method for determining the need forsurgery (Yılmaz et al 1999), what does theexaminer evaluate?

EXAMINATION AND PROGNOSIS 167

Table 6 Individual discriminants of recovery

Parameter p

Elbow flexion 0.0004*Elbow extension 0.018*Wrist extension 0.0042*Thumb extension 0.023*Finger extension 0.0069†

*n = 39, †n = 38. (From Michelow et al 1994, with permission.)

Table 7 Discriminants of recovery

Parameter Rate of incorrect prediction (%)

Elbow flexion (3 months) 12.8*Elbow flexion (3 months) 7.1*

+ finger flexion (birth)Elbow flexion + finger extension 5.2†

(3 months)Elbow flexion + elbow, wrist, thumb 5.2†

and finger extension (3 months)

*n = 39, †n = 38. (From Michelow et al 1994, with permission.)N

umbe

r of

pat

ient

s

Poor recoveryGood recovery

Numerical score0 0.3 0.6 1 1.3 1.6 2

109876543210

Figure 7

All patients, irrespective of whether primary plexus surgerywas or was not performed, were classified into good andpoor recovery groups based on their elbow flexion andshoulder abduction at 12 months of age (n = 39). Thegroups were then evaluated retrospectively with respect toelbow flexion at 3 months. A score of 0 at 3 months wasseen with almost equal frequency in both groups indicat-ing the poor discriminating ability of elbow flexion as apredictor. (From Michelow et al 1994, with permission.)

Figure 8

Based on elbow flexion and shoulder abduction scores at12 months of age, patients were classified into good andpoor recovery groups (n = 38). The Test Score of elbowflexion plus elbow, wrist, thumb and finger extension at 3months is shown. A score of 3.5 out of 10 was the water-shed between the groups. All patients with scores below3.5 were in the poor recovery group and all patients in thegood recovery group scored 3.5 or better. (From Michelowet al 1994, with permission.)

Poor recoveryGood recovery

Elbow flexion + (elbow + wrist + thumb + finger) extension

0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5

876

54

32

10

Num

ber

of p

atie

nts

Page 177: Brachial Plexus Injuries

Gilbert and Tassin relied on spontaneousrecovery of the biceps as the indication forsurgery (Gilbert and Tassin 1984). If the recoveryof the biceps had not begun at 3 months of age,the functional prognosis was poor and surgicalrepair of the plexus was warranted. More specif-ically, they suggested that surgery was indicatedwhen there was a total palsy with a flail arm after1 month and Horner’s syndrome, when infantswith complete C5–C6 palsy after breech deliveryshowed no signs of recovery by the third month,and when biceps was completely absent by thethird month in infants with C5–C6 palsies.Because of the necessities of scheduling and forsafety of anesthesia, surgery is performed in thethird month (Gilbert et al 1988). These guidelinesare widely used in many centres and are proba-bly the most common indications in current use.

Narakas divided patients into three groups(Narakas 1985). Those patients who startedrecovering within 3 weeks would recovercompletely and would not require surgicalmeasures. Patients who started to recover afterthe third week and continued to improve wouldoften require secondary surgical procedures.Finally, those patients who did not start torecover after the second month of life would dopoorly and were explored as soon as possible.

In the Waters’ series, 66 patients followed fromless than 3 months of age were divided intogroups depending on the month of life in whichbiceps strength recovered (Waters 1999). In thiscarefully performed study, analysis of variancewas used to demonstrate statistically that theearlier the biceps recovers, the better the finalresult for the patient. He concluded that patientswith total lesions at 3 months of age (flail armplus Horner’s syndrome) and patients who hadno recovery of the biceps muscle by 5 months ofage should be offered surgery.

Some authors (Berger et al 1997, Grossman etal 1997, Chuang et al 1998, McGuiness and Kay1999, Yılmaz et al 1999, Basheer et al 2000) feel,however, that the evaluation of elbow flexionalone is not sufficient to distinguish all patientswho are suitable candidates for surgery. It hasbeen our experience that a number of patientswith absent elbow flexion at 3 months of ageimproved sufficiently by 9 months of age toobtain functionally useful elbow flexion ofgreater than half range against gravity (Michelowet al 1994), Grade 6 or 7 on the Active Movement

Scale. Indeed, almost half of the patients in ournatural history study with no elbow flexion at 3months of age went on to have good extremityfunction according to Narakas’ criteria (Narakas1985) (Fig. 7).

Using the data from the natural history studyoutlined above (Michelow et al 1994), a TestScore was developed to determine the likelihoodof a good outcome without surgery. The TestScore developed was based on a grading systemthat has since been supplanted in our clinic bythe Active Movement Scale. In order to convertcurrent scores on the Active Movement Scale(Table 3) to former numerical scores for testingpurposes a conversion system is used (Table 8).A Test Score (x, range 0–10) can then beassigned to any patient at 3 months of age bysumming the former numerical score (range 0–2)for the clinical grade for the following jointmotions:

x = elbow flexion + elbow extension + wristextension + thumb extension + finger extension

The linear discriminant function for this TestScore was:

y = 3.3 – 0.94x

If y < 0, good recovery is predicted. If y ≥ 0, apoor outcome is expected. Solving the equationfor y = 0 suggests a good outcome for cases withx > 3.5.

In practice, at 3 months of age the MuscleGrade (Table 3) of five selected joint movements(elbow flexion and elbow, wrist, thumb and

168 OBSTETRICAL PARALYSIS

Table 8 Conversion from Current Muscle GradingSystem to Former Numerical Scores*

Current Muscle Grade Former Numerical Score

0 01 0.32 0.33 0.64 0.65 0.66 1.37 2.0

*These conversions were required to utilize the previouslypublished Test Score. (From Clarke and Curtis 1995, withpermission.)

Page 178: Brachial Plexus Injuries

finger extension) are converted into numericalscores (Table 8). The five numerical scores areadded to give a Test Score out of 10. Infants witha Test Score of ≤ 3.5 are booked for surgicalexploration of the brachial plexus. If the TestScore is > 3.5, the infant continues to be followedin the clinic. Clearly the conversion of scalesmakes this evaluation method cumbersome. Anew analysis of data obtained using the currentActive Movement Scale is underway.

The above system is useful in identifyingpatients with total palsy who require earlysurgery. Supporting evidence indicating surgeryin some total palsy patients can be deduced fromthe fact that none of the patients with Horner’ssyndrome in our series of 48 total plexus palsypatients went on to satisfactory spontaneousrecovery (Al-Qattan et al 2000).

Some patients with upper trunk lesions whoshow good early recovery and have Test Scores> 3.5 may still not develop adequate elbowflexion by the end of the first year of life and mayhave poor shoulder function. Our presenttechnique for selecting these patients for surgeryis to continue to monitor the Active MovementScores and if, at the age of 9 months, elbowflexion is less than Grade 6 (less than half rangeof motion against gravity), surgical exploration isoffered.

To assess elbow flexion at 9 months of age weuse what we have called the ‘Cookie Test’. Thistest is performed with the child in a comfortable,sitting play situation. With the child’s uninvolvedhand occupied with a toy, the tester gentlyrestrains the involved arm in a position of adduc-tion against the child’s trunk. The arm isrestrained in this way to limit the compensatoryshoulder abduction and internal rotation thatchildren with upper root lesions characteristicallyuse to bring the hand to the mouth (the trumpetsign). One half of an Arrowroot cookie is thenoffered to the involved hand and the child isencouraged to put it to the mouth. The cookieshould be small to encourage full flexion of theelbow. The child passes the test, and is rejectedas a surgical candidate, if the cookie is taken tothe mouth by elbow flexion against gravity andwith less than 45° of neck flexion (Fig. 9). If thecookie does not reach the mouth, or if markedflexion of the neck is required to reach thecookie, the child fails the test and surgery isconsidered (Fig. 10).

Many of the concepts presented above aresummarized in the treatment algorithm proposedby Berger et al. (1997). Total plexus lesions arerecognized and operated upon early, but moder-ate lesions may be followed at multiple visitsbefore a final evaluation of ongoing recovery ismade.

In our clinic at the Hospital for Sick Childrenwe use the Test Score at 3 months of age to

EXAMINATION AND PROGNOSIS 169

Figure 9

The ‘Cookie Test’ is administered with the child sittingcomfortably. The elbow of the affected side is held gentlyat the side of the body and the child is encouraged to putthe cookie to the mouth. If less than 45° of neck flexionis required in addition to active elbow flexion to get thecookie to the mouth, the test is passed. In this case the9-month-old child passes the test and primary surgery tothe plexus is not necessary. (From Clarke and Curtis 1995,with permission.)

Page 179: Brachial Plexus Injuries

select patients with severe and usually totalplexus lesions for primary surgery. The unequiv-ocal presence of Horner’s syndrome is anabsolute indication for surgery. The Cookie Testof elbow flexion is used at 9 months of age todistinguish additional patients, usually withupper trunk lesions, who have not recoveredsufficiently and require surgery. The selectedpatients are offered surgical intervention at theearliest available opportunity once the decisionto operate has been made.

Assessment of surgical results

Narakas originally believed there was noadequate classification to demonstrate theresults of brachial plexus reconstructionbecause of the complexity of the lesions and ofthe repair (Narakas 1985). Nonetheless Narakasdid provide us with some practical categories(Narakas 1985) (Table 9) as follows: Good resultsdemonstrated abduction and flexion of theshoulder to 90°; external rotation to at leastneutral; elbow flexion of 120° with MRC Grade4 or better; elbow extension lag of not morethan 20° with MRC Grade 3 or better; extensionof the wrist to at least neutral; flexion of thewrist with MRC Grade 3 or better and a handthat could grasp an object the size of an egg andappreciate at least light touch. Fair resultsshowed abduction of the shoulder to 50°–85°;external rotation with the elbow flexed and theforearm against the chest to at least 30°; elbowflexion to 90°–115° with a MRC Grade 3 or

170 OBSTETRICAL PARALYSIS

Table 9 Narakas’ Grading System for outcome in obstetrical brachial plexus palsy

Good Fair

Range MRC Range MRC

Shoulder abduction and flexion 90° 50°–85°Shoulder external rotation ≥ Neutral ≥ 30°Elbow flexion 120° ≥ 4 90°–115° ≥ 3Elbow extension Lag ≤ 20° ≥ 3 Lag 35°–50°Wrist flexion ≥ 3Wrist extension NeutralHand Grasp egg, light touch Weak grasp,

protective sensation

MRC = Medical Research Council Muscle Grade (Table 1). Poor and nil were considered self-explanatory. (Data from Narakas 1985, withpermission.)

Figure 10

In this 9-month-old infant elbow flexion against gravity waslimited although full flexion with gravity eliminated waspresent. The cookie did not even approach the mouthleaving the child visibly upset. The Cookie Test was failedand surgery was recommended. (From Clarke and Curtis1995, with permission.)

Page 180: Brachial Plexus Injuries

better; a passive or active extension lag at theelbow of 35°–50°; a hand with a weak grip withsome fingers capable of holding a light objectand protective sensation at least in the mediannerve territory. Poor results failed to achieve theabove criteria. Nil results were self-evident.

A detailed, updated classification system forthe assessment of hand function in operatedpatients with obstetrical lesions has beenproposed by Raimondi (Clarke and Curtis 1995)(Table 10). This scale incorporates evaluationof both sensation and movement and attemptsto distinguish useful from functionless results.The grading system has been simplified tomake it easier to use. This system addressesthe functional deficits seen in these patients.Grade III and above are considered usefuloutcomes.

We feel that two considerations are importantin evaluating patients following primary surgeryto the brachial plexus. The first is that the samesystem of recording should be applied bothbefore and after surgery to allow direct compar-ison of paired data and facilitate statistical analy-sis. Secondly, it would be extremely valuable ifall centers engaged in this surgery could form aconsensus grading scale to be validated in amulti-center inter-rater reliability study. Ourcurrent approach to the evaluation of surgicalresults at the Hospital for Sick Children is to usethe Active Movement Scale (Table 3) because itis reliable, readily used both in infants and youngchildren, and allows statistical analysis of boththe natural history and the results of surgicalintervention.

References

Alanen M, Ryöppy S, Varho T (1990) Twenty-six earlyoperations in brachial birth palsy, Z Kinderchir 45:136–9.

Al-Qattan MM, Clarke HM (1994) A historical note onthe intermediate type of obstetrical brachial plexuspalsy, J Hand Surg 19B:673.

Al-Qattan MM, Clarke HM, Curtis CG (1995) Klumpke’sbirth palsy: Does it really exist?, J Hand Surg20B:19–23.

Al-Qattan MM, Clarke HM, Curtis CG (2000) Theprognostic value of concurrent Horner’s syndrome intotal obstetric brachial plexus injury, J Hand Surg25B:166–7.

Bager B (1997) Perinatally acquired brachial plexus palsy– a persisting challenge, Acta Paediatr 86:1214–19.

Basheer H, Zelic V, Rabia F (2000) Functional scoringsystem for obstetric brachial plexus palsy, J Hand Surg25B:41–5.

Bellew M, Kay SPJ, Webb F, Ward A (2000)Developmental and behavioural outcome in obstetricbrachial plexus palsy, J Hand Surg 25B:49–51.

Berger AC, Hierner R, Becker MH-J (1997) Die frühzeit-ige mikrochirurgische Revision des Plexus brachialisbei geburtstraumatischen Läsionen. Patientenauswahlund Ergebnisse, Orthopäde 26:710–18.

Boome RS, Kaye JC (1988) Obstetric traction injuries ofthe brachial plexus. Natural history, indications forsurgical repair and results, J Bone Joint Surg70B:571–6.

EXAMINATION AND PROGNOSIS 171

Table 10 Raimondi’s Grading System for outcome of hand function in obstetrical brachial plexus palsy

Observation Hand Grade

Complete paralysis or slight finger flexion of no use; useless thumb with no pinch – 0no or some sensation

Limited active flexion of fingers; no extension of wrist or fingers; possible lateral pinch Iof thumb; supinated forearm

Active extension of wrist gives passive flexion of fingers (by tenodesis); passive lateral IIpinch of thumb; pronated forearm

Complete active flexion of wrist and fingers; mobile thumb with partial abduction and IIIopposition; some intrinsic balance; no active supination – good sensation – good possibilities for secondary surgery

Complete active flexion of wrist and fingers; active wrist extension; weak or absent finger IVextension; good thumb opposition with active intrinsics; partial active pronation and supination

Grade IV plus active finger extension and near complete pronation and supination V

Hand Grades of III or better are considered to be useful functional outcomes. (From Clarke and Curtis 1995, with permission.)

Page 181: Brachial Plexus Injuries

British Medical Research Council (1943) Aids to theInvestigation of Peripheral Nerve Injuries. His Majesty’sStationery Office: London.

Chuang DC-C, Ma H-S, Wei F-C (1998) A new evalua-tion system to predict the sequelae of late obstetricbrachial plexus palsy, Plast Reconstr Surg101:673–85.

Clarke HM, Curtis CG (1995) An approach to obstetri-cal brachial plexus injuries, Hand Clin 11:563–80.

Curtis CG (2000) The Active Movement Scale: AnEvaluative Tool for Infants with Obstetrical BrachialPlexus Palsy. Master of Science Thesis, Institute ofMedical Science, University of Toronto.

Eng GD (1971) Brachial plexus palsy in newborninfants, Pediatrics 48:18–28.

Erb W (1874) Über eine eigenthümliche Localisationvon Lahmungen im plexus brachialis, Naturhist-MedVer Heidelberg Verh 2:130.

Gilbert A, Tassin JL (1984) Réparation chirurgicale duplexus brachial dans la paralysie obstétricale, Chirurgie110:70–5.

Gilbert A, Tassin J-L (1987) Obstetrical palsy: A clinical,pathologic, and surgical review. In: Terzis JK (ed).Microreconstruction of Nerve Injuries. WB SaundersCompany: Philadelphia: 529–53.

Gilbert A, Razaboni R, Amar-Khodja S (1988)Indications and results of brachial plexus surgery inobstetrical palsy, Orthop Clin North Am 19:91–105.

Gilbert A, Brockman R, Carlioz H (1991) Surgical treat-ment of brachial plexus birth palsy, Clin Orthop264:39–47.

Greenwald AG, Schute PC, Shiveley JL (1984) Brachialplexus birth palsy: a 10-year report on the incidenceand prognosis, J Pediatr Orthop 4:689–92.

Grossman JAI, Ramos LE, Shumway S, Alfonso I (1997)Management strategies for children with obstetricalbrachial plexus injuries, Int Pediatr 12:82–6.

Jackson ST, Hoffer MM, Parrish N (1988) Brachial-plexus palsy in the newborn, J Bone Joint Surg70A:1217–20.

Kawabata H, Masada K, Tsuyuguchi Y et al (1987) Earlymicrosurgical reconstruction in birth palsy, Clin Orthop215:233–42.

Klumpke A (1885) Contribution à l’étude des paralysiesradiculaires du plexus brachial, Rev Méd 5:739–90.

Laurent JP, Lee RT (1994) Birth-related upper brachialplexus injuries in infants: operative and non-operativeapproaches, J Child Neurol 9:111–17.

Mallet J (1972) Paralysie obstétricale du plexusbrachial. Traitement des séquelles. Primauté du traite-ment de l’épaule – méthode d’expression des résultats,Rev Chir Orthop 58, suppl 1:166–8.

McGuiness C, Kay S (1999) Mini-symposium: surgicalneurology of the upper limb: (iii) obstetrical brachialpalsy, Curr Orthop 13:20–6.

Michelow BJ, Clarke HM, Curtis CG et al (1994) Thenatural history of obstetrical brachial plexus palsy,Plast Reconstr Surg 93:675–80.

Narakas AO (1985) The treatment of brachial plexusinjuries, Int Orthop 9:29–36.

Narakas AO (1986) Injuries to the brachial plexus. In:Bora FW Jr (ed). The Pediatric Upper Extremity:Diagnosis and Management. WB Saunders Company:Philadelphia: 247–58.

Narakas AO (1987) Obstetrical brachial plexus injuries.In: Lamb DW (ed). The Paralysed Hand, Vol 2. ChurchillLivingstone: Edinburgh: 116–35.

Piatt JH Jr (1991) Neurosurgical management of birthinjuries of the brachial plexus, Neurosurg Clin NorthAm 2:175–85.

SAS User’s Guide: Statistics (1985) SAS Institute Inc.:Cary, North Carolina.

Stojčević-Polovina M (1986) Risk factors in infants withparesis of the brachial plexus, Acta Med Iugo 40:3–14.

Sunderland S (1951) A classification of peripheral nerveinjuries producing loss of function, Brain 74:491–516.

Terzis JK, Liberson WT, Levine R (1986) Obstetricbrachial plexus palsy, Hand Clin 2:773–86.

Waters PM (1999) Comparison of the natural history,the outcome of microsurgical repair, and the outcomeof operative reconstruction in brachial plexus birthpalsy, J Bone Joint Surg 81A:649–59.

Yılmaz K, Çalișkan M, Öge E et al (1999) Clinical assess-ment, MRI, and EMG in congenital brachial plexuspalsy, Pediatr Neurol 21:705–10.

172 OBSTETRICAL PARALYSIS

Page 182: Brachial Plexus Injuries

Introduction

It is essential to be aware of the natural historyof OBPP and the possible sequelae of this birthinjury in order to be able to consider which kindof treatment is most opportune. Conservativetreatment and surgery, whether a primary neuro-surgical reconstruction or secondary surgery,should not be regarded as alternatives, butrather as complementary. Everyone involved inthe conservative treatment of OBPP should,therefore, also be aware of the surgical indica-tions. Knowledge of the natural history andpossibilities of conservative treatment of OBPPcan help with selecting those patients who willbenefit from primary neurosurgical reconstruc-tion or secondary surgery. It is not realistic to talkabout conservative treatment of OBPP withoutalso considering when neurosurgery andsecondary surgery may be required.

Natural history

There is a real need to understand the naturalhistory of OBPP in order to be able to predict theprobable outcome and the need for surgery at anearly stage. Naturally the parents of a baby withOBPP, having been confronted with an unexpectedcomplication during the delivery of their child, arelonging for information regarding the prognosis. Itis, however, not possible to predict with completecertainty the ultimate consequences of this injuryimmediately after diagnosis.

A large number of children with OBPP experi-ence a degree of paralysis in the affected arm foronly a few days. Some have complete paralysisof the whole arm, but show rapid recovery of thedistal muscles. If there is persistent completeparalysis 6 weeks after birth, the prognosis willbe poor. External rotation of the shoulder andsupination in the lower arm usually recoverrelatively late. Wrist and finger extension areoften more troublesome than flexion. Eventually,some degree of biceps function will alwaysdevelop. It is remarkable that despite poor handfunction, good recovery of the sensation in thehand can occur. Return of motor function cancontinue until 21⁄2 years of age, and sensoryfunction beyond 3 years.

Eng et al (1996) performed electrodiagnosticstudies which showed that reinnervation of thebiceps occurs by 4–6 months of age, but activeelbow flexion may not be apparent until 3–4weeks later; forearm muscle reinnervation occursat 7–8 months of age, and reinnervation of thehand muscles by 12–14 months. The value ofEMG findings in predicting the recovery of OBPPcan be considered dubious.

Gilbert et al (1988) noted that, throughout thelast century, a question frequently posed byneurologists and surgeons was: ‘does the recov-ery of an Obstetrical Brachial Plexus Palsy(OBPP), which always exists but may well beincomplete, justify additional treatment, surgicalor otherwise?’ Specht (1975) performed anextensive literature search concerning theprognosis of brachial plexus palsy in thenewborn. He found that opinions varied from: ‘in

18Conservative treatment of obstetricalbrachial plexus palsy (OBPP) andrehabilitationRobert S Muhlig, Gerhard Blaauw, Albert (Bart) CJ Slooff, Jan W Kortleve,and Alfons J Tonino

Page 183: Brachial Plexus Injuries

the vast majority of infants the recovery offunction begins within several days and paraly-sis clears promptly’, to ‘spontaneous recoveryoccurs to some degree in more than half thepatients with paresis of the upper part of theplexus and is complete in only about 10 percent’. Furthermore, he noted that, in general,neurologists, neurosurgeons and paediatricianswere more optimistic than orthopaedic surgeons,who indicated a significantly higher incidence ofincomplete recovery, possibly because the lattersee these patients later because of the sequelaeof OBPP. Another reason for this variation inopinions concerning the prognosis of OBPPcould be that the term ‘complete recovery’ ispoorly defined (Hoeksma et al 2000).

If complete recovery is defined as a child withOBPP who regains normal muscle strengthtogether with normal sensation, Hoeksma founda complete recovery rate of 72.6 per cent.Different percentages may be found if completerecovery is defined in more functional terms,since remaining paresis may be accompanied bysatisfactory function of the upper extremity. Itmight be expected that, if the classification ofNarakas (1993) were followed, a much higherrecovery rate score would be achieved: a ‘pooroutcome’ by Narakas was defined as elbowflexion of 50 per cent or less of the normal range,and shoulder abduction of less than 50 per centof the normal range. Otherwise the recovery,according to Narakas, was considered to be‘good’.

The conclusion must be that, due to theabsence of a uniform assessment method andthe fact that the groups of children with OBPPthat have been followed-up are far from uniform,the results between the different outcomescannot really be compared. This reinforces theview that the prediction of outcome is very diffi-cult.

Neurosurgery

There have always been disputes about theeffectiveness of neurosurgery on OBPP because,in the natural history of the condition, even themost severe lesions will show some degree ofrecovery. It also seems that recovery of activeelbow flexion and of a certain degree of sensa-

tion is the rule. Kennedy (1903) reported the firstsurgical procedure for treating a brachial plexusinjury. At that time, the results of this surgerywere poor. Contrary to the general opinionnowadays, in 1955 Wickstrom et al stated that aneurosurgical reconstruction did not providebetter results than non-operative treatment.

In his thesis, Tassin (1983) reviewed therecords of 44 children who were not operatedupon. His findings included the following:

1. When biceps and deltoid muscles began theirrecovery before 2 months of age, the resultwas a normal or nearly normal shoulder;

2. When biceps or deltoid muscle began torecover before the third month, the end resultwas good;

3. When biceps and deltoid muscles began theirrecovery after 3 months of age, the end resultwas average or poor.

Because of Tassin’s findings, Gilbert et al (1988)started to explore the brachial plexus in everychild with OBPP who did not have any bicepsfunction by the age of 3 months. The biceps wasconsidered to be the ‘key muscle’ because of therelation found between the time that the bicepsshowed clinical signs of reinnervation and theexpected degree of recovery of shoulder functionin particular.

Clarke and Curtis (1995) followed up 66patients with OBPP, and found that 61 patients(92 per cent) recovered spontaneously and fivepatients (8 per cent) required primary explorationand reconstruction. It appeared that elbowflexion at 3 months incorrectly predicted recov-ery in 12.8 per cent of cases. When appropriateparameters (biceps together with triceps, wristextension, finger extension) were combined,recovery was predicted incorrectly in only 5.2 percent of cases.

Strömbeck et al (2000) studied functionaloutcome at the age of 5 years in 247 patientswith OBPP. They compared the outcome inchildren with an upper lesion (C5–C6 and C5–C7)who had no early recovery, i.e. exhibited nomuscle activity in their biceps or deltoid musclesat 3 months of age. It was found that shoulderfunction in C5–C6 palsies was significantly betterin the operated group, but as far as other param-eters were concerned there were no differencesbetween the operated and non-operated group.

174 OBSTETRICAL PARALYSIS

Page 184: Brachial Plexus Injuries

This difference was not statistically significant inthe C5–C7 group.

These findings are in accordance with earlierreports by Gilbert and Tassin. In this series asubstantial number of the non-operated patientsshowed good late recovery, and thus the writerssuggest postponing the decision to operate inthe C5–C7 lesion. It is remarkable that in theC5–C6 lesion group, decrease in grip strengthand bimanual function was found. In this series,total lesions were generally treated early.

Basheer et al (2000) concluded in their studyof 52 patients that the upper limb function gradu-ally improved until the sixth month, after whichthere was no significant recovery. Although 90per cent of the patients achieved their final scoreat the sixth month, there was no significantimprovement between the third and sixthmonths. This result supports Gilbert’s policy ofperforming early surgery at 3 months, ratherthan that of Zancolli and Zancolli (1988) whofound that 75 per cent of their patients startedrecovering after the fifth month.

In his study of 66 patients with OBPP, Waters(1999) found that the results in the six patientswith neurosurgical reconstruction were signifi-cantly better than of those in the 15 patients whowere not operated on and had a biceps functionrecovery at the age of 5 months. The results inthese six patients, however, were no better ifthey were compared with the results of the 11patients in whom the biceps recovered at the ageof 4 months.

This study confirms Gilbert’s observation thatit is very rare for children to have completerecovery if the biceps function returns after 3months. There is, despite all these findings, noconsensus concerning the indications for neuro-surgery in OBPP. If the indication for neuro-surgery, according to Gilbert’s criteria, wasreached on the basis of no biceps function at theage of 3 months, 39 of the 66 patients in Waters’study would have been operated, not just six.

Although there are no comparative outcomedata concerning children who were treated neuro-surgically and children who underwent conserva-tive treatment, we agree with Birch et al (1998)that, in general, a child with severe damage to thebrachial plexus will not be worse off if neurosur-gical reconstruction is performed. There are somestrong arguments in favour of neurosurgical treat-ment. Birch et al (1998) stated that: the palliative

procedures of deformities secondary to the neuro-lesions are deeply unsatisfactory and the resultsof musculotendinous transfers in OBPP are farinferior to those following good nerve regenera-tion and on the whole they are inferior to thoseobtained for the treatment of poliomyelitis orsimple peripheral nerve injuries. Reasons include:widespread weakness of muscle, which is at timesnot fully appreciated, inadequate cutaneoussensation and proprioception and later skeletaldeformities’. Gilbert et al (1988) stated that: ‘It isan advantage of operated cases that their muscu-lotendon transfers are made possible by thelarger recovery of shoulder muscles and theabsence in most cases of the typical cocontrac-tions that occur with spontaneous recovery. Inmost cases, therefore, the end result after surgi-cal treatment will be better than spontaneousrecovery. This surgical treatment includes theinitial plexus repair and the subsequent opera-tions: joint release and muscle tendon transfers,which need to be done at an early age’.

There is no doubt about the positive results ofneurosurgery in selected cases of children withOBPP. There is, however, still no consensus,especially about when neurosurgical reconstruc-tion should be performed.

The Brachial Plexus Work Group in Heerlen,The Netherlands, developed a flowchart as ageneral guideline on how and when to act if achild is born with a brachial plexus lesion (seeAppendix 1). Of course not everyone will agreewith the policy as described in this flowchart, butuntil there is a consensus it is important to haveconsistency in indications for intervention, tohave an accepted and standardized method ofassessment for evaluation and, if necessary, toadapt the indication parameters.

Conservative management

Range of motion (ROM) exercises should beginimmediately to prevent the otherwise rapiddevelopment of contractures at the shoulder,elbow and wrist while waiting for the brachialplexus to recover. It should be remembered thatthe extent of the paralysis will sometimesregress, especially in the distal muscles of theaffected arm. This does not exclude the possibil-ity of a severe lesion in the upper roots.

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 175

Page 185: Brachial Plexus Injuries

Exercises must therefore be performedfrequently, and should be aimed at maintainingfull passive external rotation of the shoulder withthe arms in adduction. To achieve this, bothupper limbs must be exercised together.Attention should also be paid to maintaining theinferior and posterior scapulo-humeral angle.We, together with Birch et al (1998), believe thatparents must be involved in the treatment of thechild from the outset. The best physiotherapist isthe mother or father. There is no scientific proofof the effectiveness of physical therapy in theprevention of joint contractures; however, wenevertheless believe that many fixed deformitiescan be prevented if the parents perform theappropriate exercises regularly during the day(for example, after every meal, after changingnappies etc) instead of waiting for the physio-therapist to come three times per week. On theother hand, severe fixed endorotation–adductioncontractures have been seen in the shoulders ofchildren whose parents faithfully performed theexercises as instructed.

It is believed that the development of shouldercontracture is due to motor imbalance and isdependent on the degree of neurological injury.Manipulation of the shoulder joint is probablynot enough to prevent contractures completely ifmotor recovery in the deltoid and externalrotators is insufficient to balance the internalrotators.

Posterior dislocation of the radial head with agradual curving of the ulna can develop from thefourth month of age (Eng et al 1996). Forcedsupination of the pronated forearm may aggra-vate radial dislocation, and this radial head dislo-cation may be responsible for a progressiveflexion contracture of the elbow. Since the bicepsand brachialis at that age are often still paral-ysed, these muscles cannot be the cause of thiscontracture. Whenever there is a progressiveflexion contracture of the elbow, splinting isrecommended. Excellent results have beenachieved with serial casting, provided this treat-ment is started in time. Splinting has otherwisegenerally been condemned, although somerecommend the use of functional bracing inchildren, stating that it may be helpful in encour-aging early hand use.

Although splinting is generally regarded asobsolete, Eng et al (1996) reported that all exceptmild cases were treated with a wrist/hand cock-

up splint with the thumb in opposition. Later, astatic elbow extension or dynamic elbow flexionor extension and supination splints were used asindicated. We recommend that redressing splintsonly be used at night. Usually this involves cock-up wrist splints, that also correct ulnar deviationin the wrist. If there is a weak extension of thewrist causing a ‘dropping hand’ and preventingthe child from grasping, it is recommended thata cock-up splint only be used for a few hoursduring the day; we believe that this type of splintinterferes with the sensation of the hand, andthis could discourage the child from using theaffected hand. Moreover, a cock-up splint mayprevent the strength of the wrist extensors devel-oping.

There is frequently a lack of active supination,and it seems that the biceps must reach normalstrength before supination can be successfullyaccomplished. To perform supination of theelbow in extension, the supinator muscle itselfmust be strong or at least of normal strength(Eng et al 1996).

The thumb tends to become tight in flexionand adduction, and if this occurs the child canbenefit from a nightly redressing splint, becausea ‘thumb in palm deformation’ will limit handfunction considerably.

The rate of recovery in OBPP can be very slow.It is therefore imperative to keep monitoring andif necessary treating the child during this period.Depending on the age of the child and the seque-lae of the OBPP, the therapy should be adapted.As soon as the child is able to participate activelyduring therapy this should be encouraged.Passive exercises should only be performed ifsome functions, such as lateral/medial rotation ofthe shoulder or supination/pronation in the lowerarm, are absent; however, even in this case thechild should be stimulated to use the affectedarm/hand for that particular function. The healthyunaffected arm can be used to support theaffected side in performing active exercises: thechild holds a stick with both hands supinated andthe arms adducted, then the unaffected armrotates externally and pulls the affected arm inmedial rotation. When the unaffected arm is putin medial rotation, the affected arm is pushedinto external rotation. Rotational exercises forthe lower arm can be done in the same way.Special care should be taken when training thechild to be independent as far as activities of

176 OBSTETRICAL PARALYSIS

Page 186: Brachial Plexus Injuries

daily life are concerned. Even if, for example,medial rotation is limited and doing up buttonsis troublesome, the child should be encouragedto do it without too much support from theparents. Sporting activities such as swimmingshould be encouraged at a young age. It is,however, our experience that even with optimalconservative management some sequelae of theOBPP cannot be prevented. There is almostalways some degree of scapular winging, later-alization of function to the unaffected other side,and some shortening of the affected side – thepercentage of the shortening being dependenton the severity of the lesion.

Chuang et al (1998a) states that cross-innerva-tion in the neuroma formation of the damagednerves, muscular imbalance and growth are thethree main causes of shoulder deformity inOBPP. Cross-innervation causes co-contractionsof synergistic and antagonistic muscle groups.Muscular imbalance and co-contraction will leadto muscular contractures, which again are themain cause of development of shoulder andelbow deformities. Investigations are currently inprogress regarding the use of botulinum toxin inchildren with cerebral palsy, and perhaps it willbe possible in the future to use botulinum inchildren with OBPP and troublesome co-contrac-tions. Berger and his colleagues in Hannoverhave started a trial with botulinum and haveconcluded in a very limited series that it mightbe an effective tool (Rollnick et al 2000).

Children with an OBPP are sometimes seenwho hardly involve their affected hand inperforming activities, despite the fact thatdoctors are convinced that there are no longerany serious neurological deficits. This kind of‘agnosia’ should be related to initial disturbancesin sensation and to a lack of creation of ‘cerebralcircuits’ at a very early stage of life.Investigations are under way to see whetherEMG-triggered myofeedback has an influence onthese problems.

Sequelae of OBPP

Shoulder

The shoulder is by far the most frequentlyaffected joint in OBPP. The medial rotation

contracture with all its complications for theglenohumeral joint, is most common. Stimsonfirst described a posterior dislocation of theshoulder in OBPP in 1888.

Fairbank (1913) expected that all children withOBPP with incomplete neurological recoverywould develop a shoulder deformity, andsuggested that this is in fact the only hindranceto complete recovery. Fairbank consideredmuscular imbalance to be the main cause of theposterior subluxation in the shoulder because hefound it remarkable that no luxation was seen incomplete paralysis.

L’Episcopo (1934) found that the disabilityresulting from obstetrical paralysis of the upperarm type is essentially due to internal rotationdeformity of the humerus, which is a very poorfunctional position. The patient is unable toperform the necessary movements requiredwhen eating, combing the hair, putting on acollar and tie, dressing etc.

Green and Tachdjian (1963) reviewed all theirOBPP cases between 1943 and 1961, and foundthe following to be the most common residualdeformities requiring surgery:

• Fixed internal rotation deformity of the shoul-der;

• Limited abduction of the shoulder andcontractures of the adductors;

• An elongated coracoid;• An elongated and downwards hooked

acromion;• Dislocation of the head of the radius.

Dunkerton (1989) noticed a loss of passive exoro-tation beyond neutral to be the main clinicalsymptom of a posterior dislocation. Troum et al(1993) considered that in young children (under6 months of age) with a fixed medial adductioncontracture, the birth trauma that caused theOBPP also causes the dislocation. It is indeedhard to imagine that a muscular imbalance atthis age can lead to dislocation of the shoulder.

Birch and Chen (1996) reported that of 108children requiring surgery because of severelimitation in passive external rotation of theshoulder, 45 had an uncomplicated medialrotation contracture, 56 had a posterior subluxa-tion or dislocation of the head of the humerus,and seven had a dislocation complicated byovergrowth of the corocoid and acromion. In

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 177

Page 187: Brachial Plexus Injuries

three cases, dislocation occurred at birth. In fivebiopsies of the subscapularis muscle, changeswere seen that were consistent with post-ischaemic fibrosis. In all other cases progressivedeformity was caused by muscular imbalance –the subscapularis and other medial rotatorsinnervated by C7, C8 and T1 being unopposedby the lateral rotators innervated by C5.

Torode and Donnan (1998) calculated that theincidence of posterior luxation in OBPP is about40 per cent.

Pearl and Edgerton (1998) found three consis-tent patterns of shoulder deformity in childrenwith an apparently incomplete recovery of anOBPP:

1. Flattening of the glenoid;2. A biconcave glenoid;3. A pseudoglenoid with a posterior concavity

distinct and separated from the remainingoriginal articular surface.

All patients with a glenoid deformity had apassive lateral rotation of less than –10°.However, not every glenoid appeared deformed,even if there was a lateral rotation limitation.Both Zancolli and Zancolli (1988) and Pearl andEdgerton (1998) found that 28 per cent ofpatients had a normal glenoid despite having afixed internal rotation contracture. Nonetheless,it is obvious that an internal rotation contracturesecondary to OBPP has a high likelihood of beingassociated with glenoid deformity.

Birch et al (1998) produced a very useful classi-fication of shoulder deformities in OBPP, describ-ing four types with increasing severity caused bya medial rotation contracture:

1. The medial rotation contracture: the onlyabnormality is restriction of passive lateralrotation, which is diminished by 30–40°.Overgrowth of the coracoid was alsodescribed as an interesting cause of medialrotation contracture;

2. Posterior subluxation (simple): passive lateralrotation is restricted to about 10°. There is asyet no secondary deformity of the acromion,the coracoid or the glenoid;

3. Posterior dislocation (simple): the head of thehumerus can be seen and palpated behind theglenoid. It may be possible to click thehumeral head in and out of the glenoid. X-

rays confirm displacement, and there is acharacteristic windswept or curved appear-ance of the proximal humerus;

4. Complex subluxation/dislocation: in this finalstage, marked skeletal abnormalities areapparent on clinical and radiological exami-nation. These patients have pain, there isfixed flexion at the elbow with pronation ofthe forearm, and the compensatory thora-coscapular movement seen in many youngchildren has disappeared.

Three skeletal abnormalities are significant:

1. An elongated coracoid;2. An elongated and downwards hooked

acromion;3. A bifacetal appearance in the glenoid. The

true glenoid lies above and anterior; the falseglenoid lies below and posterior. In complexsubluxation the articulation is between thehead of the humerus and the false glenoid. Incomplex dislocation the articulation isbetween the lesser tuberosity and the falseglenoid.

Hoeksma et al (2000) found that shouldercontracture occurred in at least one-third of thechildren with delayed recovery and at least two-thirds of the children with incomplete recovery.Delayed recovery was defined as recovery thattook more than 3 weeks; complete recovery wasdefined as complete neurological recovery withnormal muscle strength in all muscle groupstogether with normal sensibility. Even in thegroup with complete neurological recovery but adelayed recovery of more than 3 weeks, about 30per cent developed a shoulder contracture. In thegroup of incomplete neurological recovery, thefrequency of shoulder contracture was as high as65 per cent.

Shoulder surgery

Sever stated as early as 1918 that if there is afixed medial rotation contracture, conservativetherapy is useless. He suggested a tenotomy ofthe pectoralis major and subscapularis. Gilbertet al (1991) advised that if external rotationdrops progressively to below 20°, therapy is nolonger efficacious. It is therefore appropriate to

178 OBSTETRICAL PARALYSIS

Page 188: Brachial Plexus Injuries

deliberate for some time on the surgical aspectsof a fixed medial contracture when it appearsthat conservative management is unable toprevent this deformity, and we will concentratehere on such surgery. For all other shouldersurgery in OBPP, the reader is referred to theappropriate chapter of this book.

It is sometimes difficult to persuade physiother-apists and doctors who are treating children withan OBPP to allow them to undergo surgery, andthe argument is often used that there is no needfor an operation because the child does not showor complain of any disabilities and is still improv-ing. This argument of the absence of disabilitiesand therefore the absence of indications forsecondary surgery is misleading for two reasons:

1. A child born with a completely paralysed armthat persists will adapt and become entirelyindependent as far as activities of daily lifeare concerned; thus disabilities will not beobvious. Although in the case of adultsconsideration may be given to withdrawingtreatment if there are no disabilities, thisshould not be done in the case of a child withan OBPP. A child will not miss a function heor she never had. It is, however, an obligationfor everyone involved in the treatment ofchildren with OBPP to do their utmost tocreate function in the affected arm that mightbe of invaluable benefit to the child.

2. Even if there are no disabilities at the time theoperation is suggested, they might arise inthe future if the suggested surgery is notperformed. A fixed medial adduction contrac-ture of the shoulder will generally not causedisability or pain to the child, but maybecome an insoluble problem because ofirreversible shoulder deformities in the futureadult. Gilbert et al (1991) formulates it veryclearly: ‘Contrary to traditional thinking, thesurgeon should not wait to treat an internalrotation contracture. In the absence of surgi-cal treatment recovery is limited, abduction isimpossible, the extremity is dysfunctional,and, most important, osseous and articulardeformity will occur. Posterior subluxationand deformity of the humeral head perma-nently worsens the prognoses. These anoma-lies, which have long been considered theresult of obstetrical palsy, are in fact simply aconsequence of untreated contractures’.

Subscapularis operations

Gilbert favours a posterior release of thesubscapularis if the external rotation drops to20°, but at the same time warns that soft tissuerelease of the internal rotation contractureshould only be performed if the joint is congru-ent and the humeral head is round. Gilbertfollowed up 66 patients who had undergoneposterior subscapularis release over more than 5years. He found excellent results if the childrenwere operated before the age of 2 years, butthere was an 18 per cent failure rate if they wereover 2 years, if there was an incorrect preopera-tive evaluation of articular deformities or if therewas no postoperative physical therapy.

In Gilbert’s opinion, posterior dislocation hasdisappeared with this aggressive treatment formedial rotation contracture of the shoulder.However, Birch found that children who weretreated with a subscapularis slide/release devel-oped bone deformities, even though the shoul-der had been concentrically reduced. Hence,treatment with a subscapularis slide failed in thelong term in one-third of the patients. He advisesthat in simple dislocation and subluxation thesubscapularis tendon should be exposed andlengthened.

In Heerlen we follow Birch’s policy and do notperform subscapularis releases because webelieve that in most cases of a fixed medialadduction contracture there is already somedegree of posterior subluxation.

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 179

Table 1 Comparing results of Gilbert and the Heerlengroup on subscapularis operations

Gilbert Heerlen (release) (lengthening)

No. of patients 66 84Prior neurosurgical 28 40

reconstructionAge at operation < 2 years 44 27Age at operation 2–4 years 14 30Age at operation > 4 years 8 27Active exorotation after 31 (47%) 27 (32%)

operationFailures 18% 23%Relapse requiring second 7

lengtheningExorotation contracture 12

Page 189: Brachial Plexus Injuries

Children have been known to lose the abilityto actively medially rotate following a subscapu-laris tenotomy. It is therefore very important notto perform a subscapular tenotomy, but a length-ening of the subscapular tendon.

In Heerlen in the period 1995–2000, 84 childrenunderwent an anterior subscapularis lengthen-ing, 26 in combination with a resection of thecoracoid. Twenty-seven (32 per cent) developedactive external rotation within 4 months oflengthening, and therefore this procedure is nolonger combined with a muscle transfer forexorotation. Our results with subscapular length-ening are comparable with the results achievedin the 66 children on whom Gilbert performed asubscapular release (see Table 1).

Gilbert does not specify what he considers tobe a failure. Of course this is arbitrary andrequires discussion. At Heerlen, a recurrence ofthe fixed medial adduction contracture wasconsidered to be a failure. Zancolli and Zancolli(1988) warned against a soft tissue procedure ona medial adduction contracture with joint defor-mity because of the risk of creating an exorota-tion contracture, which in fact produces muchmore disability than a medial rotation contrac-ture. Therefore, an exorotation contracture wasalso scored as a failure.

In fact, every time progression of shoulderdeformity is not prevented, despite a subscapu-lar release or lengthening, this should be consid-ered a failure. To our knowledge, however, theseresults have not yet been reported.

It is noteworthy that after release or lengthen-ing of the subscapularis without any kind ofadditional treatment, active external rotationdeveloped in about of 30 per cent of the operatedchildren.

Chuang et al (1998b) measure active externalrotation with the arm in abduction. The arm ofthe patient is held in 90° abduction and 90°flexion of the elbow, and the patient is thenasked to perform external rotation of the shoul-der. If the hand cannot be raised above the chest(exorotation less then 60°), a poor result isscored. If the hand can reach the ear (exorota-tion 60–90°), the score is good. If the hand canreach the occiput (shoulder exorotation morethen 90°), the result is excellent.

We assess active exorotation according to theMallet score with the arm in adduction and thearm in 90° of flexion. It is striking that some

children, despite good recovery of the infraspina-tus on EMG, are not able to actively rotate thearm if it is held in adduction, but can activelyexternally rotate the arm if in abduction. There isno apparent explanation for this phenomenon,and this again shows the need for consensus onthe method of assessment in order to compareresults.

Following the Sixth International Workshop onOBPP held in Heerlen in November 2000, wedecided to apply the following rules if a fixedmedial contracture is found in a patient:

1. A fixed medial rotation contracture should betreated surgically if there is a persisting limita-tion of the passive external rotation of < 30;

2. If there is no posterior displacement of thehead of the humerus, a subscapular slide willbe performed;

3. If there is posterior displacement of the headof the humerus, an anterior approach withsubscapular lengthening will be carried out; ifthere is an elongated coracoid, this will beshortened;

4. If there is a relapse of a fixed medial rotationcontracture, subscapular lengthening will becarried out together with a muscle transfer tocreate active external rotation;

5. If the infraspinous muscle does not showsigns of reinnervation by the age of 2 yearsand there is a fixed medial rotation contrac-ture, subscapular lengthening will beperformed together with a muscle transfer tocreate active external rotation;

6. If there is a fixed medial rotation contractureand a posterior luxation of the head of thehumerus with deformities of the glenoid(bifacetal, retroversion), subscapular length-ening together with a derotational osteotomyof the humerus should be considered,because of the risk of creating a fixed exter-nal rotation contracture if repositioning of thehead of the humerus is achieved bysubscapular lengthening.

In Heerlen CT scan arthrography of the shoulderis performed preoperatively to obtain more infor-mation about the glenohumeral joint. In childrenolder than 5 years, a CT scan without arthrogra-phy will suffice.

Following subscapular slide, the operated armis immobilized in maximal lateral rotation for 3

180 OBSTETRICAL PARALYSIS

Page 190: Brachial Plexus Injuries

weeks. After subscapular lengthening, theimmobilization is for a period of 6 weeks.

Osteotomy of the humerus

As early as 1955, Wickstrom et al stated that allsurgery around the shoulder should aim at anincrease in the lateral rotation and abduction. Hispolicy was as follows:

• If there is no shoulder deformity, this goal canbe reached by simple tenotomy/release ofcontracted medial rotators. This should onlybe performed in combination with muscletransfers to strengthen the lateral rotators;

• If there is a dislocation or incongruency of theglenohumeral joint, no release or muscletransfer should be carried out. An open reduc-tion should be performed in combination withan osteotomy of the humerus.

Most authors nowadays agree with Wickstrom,although not everyone advocates the combina-tion of release with muscle transfers. Zancolliand Zancolli (1988) consider it a serious error tothink that reduction of a posterior subluxationcan be established if there is a deformity of thejoint. An exorotation osteotomy is indicated inthese cases, even if there is only a slight defor-mity of the head of the humerus, and usuallyafter the age of 3–4 years.

Kirkos and Papadopoulos’ (1998) indicationsfor rotation osteotomy are:

• Patients at least 4 years of age, with a fixedmedial rotation contracture and decreasedstrength of the teres major and latissimusdorsi, dislocation of a deformed humerushead, and relapse of deformity/medial rotationcontracture after a soft tissue procedure;

• Patient over 8 years of age with a fixed medialrotation contracture or a limitation of theactive rotation of the arm.

Following up a series of 22 cases, they found anincrease in active abduction and in the arc ofrotation. Intensive physiotherapy was notneeded. This made it easier to manage youngerpatients because a high level of co-operation andcompliance is not necessary, as it is after softtissue surgery, especially tendon transfers.

Soft tissue procedures are recommended foryoung children who are under 6 years old andwho have a severe internal rotation contracturewithout osseous changes in the humeral head.Release of the soft tissue contracture improvesthe cosmetic appearance but produces onlyslight functional improvement. There is anincrease in external rotation without an increasein abduction, and there is also a risk of anteriordislocation of the shoulder. In addition, the rangeof rotational movement that is achieveddecreases with time and with recurrence of thefixed internal rotation deformity.

Elbow/wrist/hand

If elbow flexion remained insufficient a Steindlerprocedure was used at Heerlen, transposing thepronator/flexor group from its insertion at themedial epicondyl to a more proximal site on thehumerus, provided that the wrist extensors werestrong enough to stabilize the wrist when thepatient was using the wrist flexors to produceflexion of the elbow. Unfortunately, on severaloccasions there was a serious deformity of theelbow a few years after the Steindler procedure,and this became very unstable because transfer-ring the flexor–pronator group had caused agrowth disturbance of the medial epicondyl. It istherefore understandable that a latissimus dorsitransfer for active elbow flexion might bepreferred to a Steindler procedure.

Zancolli and Zancolli (1988) mentioned theproblem of a supination contracture, and thisshould be corrected before a ‘fixed deformity’develops. The deformity causes an inner rotationof the lower arm bones with a volar subluxationof the distal ulna, and sometimes this is accom-panied by a luxation of the head of the radius. Asupination contracture causes serious disabilitiesin daily life. Activities of daily life request elbowflexion and pronation. Zancolli and Zancolliperformed a re-routing of the biceps tendon torestore active pronation. The results at Heerlenwith this procedure, especially in fixed deformi-ties with a contracted interosseous membrane,are disappointing. Recently, functional improve-ments have been observed in children with asupination contracture who had undergone apronation osteotomy of the radius. The hand is

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 181

Page 191: Brachial Plexus Injuries

preferably placed in a position of 30° pronation.Birch et al (1998) warn that this operation mayhave to be repeated as the child grows.

Another disabling deformity is that in whichthe thumb lies in the ulnar deviated hand. In thissituation, the extensor carpi ulnaris (ECU) can betransferred to the abductor pollicis longus if wristextension is adequate. If wrist extension is poor,it might be better to transfer the ECU to theextensor carpi radialis brevis. Generally,however, the results of transfers in the hand aredisappointing. As Birch et al (1998) mentioned:‘The results of tendon transfers in the hand areunpredictable and on the whole they are worsethan those for any other neurological disorder’.

Developmental and behaviouraloutcome

There is always a danger of becoming too preoc-cupied with one aspect of the very complicatedproblems of OBPP and forgeting that ‘life ismore’ for children with OBPP than just theiraffected arm. This may be the reason whyrelatively little research has been carried out intogeneral developmental aspects in OBPP. In 1984,Greenwald et al suggested that psychologicaltesting has not revealed any differences betweennormal individuals and those who sustainedbrachial plexus birth palsy. However, Bellew etal (2000) assessed children with OBPP withregard to both developmental attainment andbehavioural problems, and found a high level ofthe latter. The children whose initial plexus injurywas so severe that it required nerve surgerywere found to have significantly poorer develop-ment than those whose injury did not requiresurgery, and the developmental delay wasglobal. This study suggests that children withOBPP, particularly those with more severeinjuries, may be at risk of developmentalproblems previously not identified. Becausethese developmental and behavioural problemshave not previously been identified, the childrenhave not had appropriate recognition or supportfor them. Whereas psychosocial risk factorsbecame more prominent with increasing age andwere related to poorer outcomes in children inall areas of functioning (motor, cognitive and

social emotional development), organic riskdecreased in influence.

Assessment

One of the biggest problems is to compare theresults of the different treatment policiesbecause of lack of consensus about the methodof assessment and how to use the variousscoring systems. This of course makes it verycomplicated to compare, for example, the resultsof a more conservative attitude in treating OBPPwith a more aggressive surgical approach. TheBrachial Plexus Work Group in Heerlen usesseveral assessment methods (see Appendix 2).

The Mallet scoring system is not really suitablefor children under the age of 3 years. The childrenare asked to perform some movements to assessthe function of the shoulder and the elbow.Determining the Mallet score is, however, not acomplete assessment. It provides no informationabout the hand function or the passive ROM.Another problem is that some score ‘active exoro-tation’ with the arm in abduction, while Malletscored active exorotation with the arm adductedto the trunk. The results differ in different positionsof the arm. In Heerlen, grades I to V are used. Birchuses only three grades, and adds the five differentscores to obtain a maximal score of 15. Thereshould be a consensus about what is considereda good, fair or poor result: 15 will of course be agood result, but what should be considered a fairresult? Any score between 8 and 13?

The passive ROM should be examined. Theexorotation, measured with the arm adducted tothe trunk, is particularly important in settingpolicy. A passive exorotation of less than 30° thatdoes not improve with exercise should betreated surgically. The choice of surgery dependson the patient’s age, degree of shoulder defor-mity and activity of lateral rotators (infraspinousmuscle).

The shoulder joint should be examinedcarefully. The coracoid and acromion should bepalpated and compared with the unaffectedhealthy side. The presence of a posterior oranterior displacement of the head of thehumerus should be established.

The Gilbert/Raimondi score for shoulder,elbow and hand function is a very useful addition

182 OBSTETRICAL PARALYSIS

Page 192: Brachial Plexus Injuries

in assessment of the function of the arm andhand in children with OBPP. However, determin-ing the correct scoring grade is frequently diffi-cult. For example, if a child shows full activeexorotation in the shoulder but less than 90°abduction, what should the stage be accordingto the scoring system of Gilbert and Raimondi?It is not possible to score a stage V or a stage II.In Heerlen, in this particular case a score of stageII plus would be allocated, but others might scorea stage V minus. There are comparable problemsin scoring hand function with this system.

It is essential to reach agreement about whichassessment methods should be used and how.Another problem is agreement on interpretingthe scores: which scores should be interpreted asgood, fair or poor? Also, what exactly is meantby a good, fair or poor result – does this applyto the neurological state of recovery, or to thelevel of functional abilities? The latter could verywell be good, while at the same time the neuro-logical state of recovery might be poor.

Conclusions

OBPP is a very complicated, multi-faceted disor-der. Conservative and surgical treatments cannotbe separated, but should be used in conjunction.A multidisciplinary approach involving theparents, physiotherapists, occupational thera-pists, neurosurgeons, plastic surgeons, ortho-paedic surgeons, rehabilitation specialists andpsychologists is probably the way to achieveoptimal results in treating children with OBPP.However, ‘prevention’ is better than ‘cure’, andtherefore the final word in this chapter goes toNarakas (1987): ‘Prevention may solve theproblems of obstetrical palsy, as it has done foracute poliomyelitis.’

Appendix 1 Flow diagram –guidelines for action in OBPP(see next page)

Notes

What to do if you encounter a child with a neona-tal palsy of the brachial plexus?

1. Diagnosis:• History of pregnancy and delivery –

number of weeks, child no., presentation,cephalic/breech;

• Difficulties, shoulder dystocia;• Reanimation, Apgar score;• Birth weight;• Help of nurse/specialist.

2. Neurological examination:• Look at the posture of the limb, the

spontaneous movements;• Stimulate, sensory and motor testing, take

time and be patient.3. Differential diagnosis:

• Clavicle fracture;• An obstetric lesion after vaginal delivery,

cephalic/breech;• An obstetric lesion after Caesarean

section;• Intra-uterine compression, maladaptation;• Congenital anomaly of the plexus;• Hereditary plexopathy;• Intra-uterine infection.

4. Tests:• Electromyography within the first days if

an intra-uterine lesion is suspected;• Radiological examinations of thorax, clavi-

cle, humerus if a phrenic paralysis issuspected, and/or a fracture;

• Look for associated lesions: haematoma,fractures, phrenic-, spinal cord-, bilaterallesion, tracheal lesion or lesions of nervesVII, XI and XII.

5. Therapy: 3 weeks in a rest position, arm infront of the chest; no splints!

Encourage early presentation and control visitsto augment the knowledge on the natural evolu-tion of the lesion and to be sure that theexercises/physiotherapy proceed correctly.

Physical therapy is a continuing activity, inwhich parents play an important role, in order tostimulate muscle activity, and sensory function,and to mobilize the joints to prevent contrac-tures.

If after 6 weeks a (nearly) totally flail limb stillpersists, further investigations such aselectromyography and myelography will benecessary and surgical intervention will beunavoidable.

For other cases, with a more limited lesion tothe upper plexus, a decision to intervene surgi-

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 183

Page 193: Brachial Plexus Injuries

184 OBSTETRICAL PARALYSIS

0 weeks birth

radiological investigations

arm in rest position

3 weeks

no improvement improvement

physical therapy (physical therapy)

6 weeks

no improvement improvement

(physical therapy)

total lesion upper plexus lesion

neurophysiological physical therapy

investigations

CT-myelography

neuro-surgery

physical therapy

3-4 months

(attention to joint contractures; subscapularis release)

no improvement improvement

(physical therapy)

neurophysiological

investigations

CT-myelography

check criteria surgery

positive dubious

neuro-surgery physical therapy

physical therapy

no improvement improvement

neuro-surgery physical therapy

physical therapy

6-10 months insufficient improvement

shoulder function

neuro-surgery

physical therapy

24 months

secondary surgery, muscle/tendon transfer, osteotomy

Neurological investigations

Neurological investigations

Neurological investigations

Neurological investigations

special attention to joint contractures, possible radiological investigations of joints

insufficient improvement of specific muscles / functions

Figure 1

Guidelines for action inOBPP.

Page 194: Brachial Plexus Injuries

cally has to be made in the third or fourthmonth; supplementary investigations are sched-uled. Check the criteria for neurosurgical treat-ment.

It is far more difficult to make a decision later,when some, but insufficient, shoulder functionrecovers.

In premature children and in children withoutor with poor physical therapeutic help and/orwith severe contractures, the recovery offunctions will be slower and incomplete.Contractures need special attention.

Criteria for neurosurgical treatment• Biceps function M0 after 3 months, eventually

combined with insufficient recovery of exten-sor muscles of elbow, wrist and fingers;

• Evidence of a severe lesion: Horner’ssyndrome, persisting hypotonic paralysis,persisting phrenic paralysis, severe sensorydisturbances;

• EMG: persisting denervation with no actionpotentials;

• CT-myelography: meningocele formationoutside the vertebral foramen.

Timing of neurosurgical intervention• Generally at the age of 3–4 months;• In severe (sub)total lesions and/or avulsions,

as soon as possible;• In doubtful cases of upper plexus lesions, not

later than the age of 1.5–2 years(?);• In late (sub)total lesions, until the age of 2

years or even later.

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 185

Appendix 2 Assessment ofOBPP

Mallet Score

Active exorotation with arm abducted:Hand–stomach reach:Lat. dorsi:Passive ROM:Exorotation with arm adducted:Endorotation:Inf. GH angle:Post. GH angle:Elbow extension:Supination:Pronation:Shoulder examination:Elongated coracoid:Aromion enlarged:Posterior displacement:Anterior displacement:

Figure 2

Mallet score (Gilbert modification) (from Gilbert 1993 withpermission).

Page 195: Brachial Plexus Injuries

References

Basheer H, Zelic V, Rabia F (2000) Functional scoringsystem for obstetric brachial plexus palsy, J Hand Surg[Br] 25(1):41–5.

Bellew M, Kay SP, Webb F, Ward A (2000)Developmental and behavioural outcome in obstetricbrachial plexus palsy, J Hand Surg (Br) 25(1):49–51.

Birch R, Bonney G, Wynn-Parry CB (1998) Birth lesionsof the brachial plexus. In: Surgical Disorders of thePeripheral Nerves. Churchill Livingstone: Edinburgh:209–33.

Birch R, Chen L (1996) The medial rotation contractureof the shoulder in obstetric brachial plexus palsy, JBone Joint Surg 73B(Suppl):68.

Chuang DC, Ma HS, Wei FC (1998a) A new strategy ofmuscle transposition for treatment of shoulder defor-mity caused by obstetric brachial plexus palsy, PlastReconstr Surg 101(3):686–94.

Chuang DC, Ma HS, Wei FC (1998b) A new evaluationsystem to predict the sequelae of late obstetric brachialplexus palsy, Plast Reconstr Surg 101(3):673–85.

Clarke HM, Curtis CG (1995) An approach to obstetri-cal brachial plexus injuries, Hand Clin 11(4):563–81.

186 OBSTETRICAL PARALYSIS

Gilbert/Raimondi assessment

Stage Shoulder assessment Gilbert/Raimondi: Elbow assessment Gilbert/Raimondi: Stage0: Complete flail shoulder Flexion: nil, some contraction 1I: Abduction = 45°, no active exorotation Incomplete flexion 2

Complete flexion 3II: Abduction < 90°, no exorotation No extension 0III: Abduction 90°, weak exorotation Weak extension 1IV: Abduction < 120°, incomplete exorotation Good extension 2

Ext. deficit: 0°–30° 0V: Abduction > 120°, active exorotation Deficit 30°–50° –1

> 50° –2

Hand-assessment Gilbert/RaimondiComplete paralysis or slight finger flexion of no use; useless thumb – no pinch; some or nosensation 0Limited active flexion of fingers; no extension of wrist or fingers; possibility of thumb lateral pinch IActive extension of wrist with passive flexion of fingers (tenodesis), passive lateral pinch of thumb (pronation) IIActive complete flexion of wrist and fingers – mobile thumb with partial abduction–opposition;intrinsic balance – no active supination; no wrist extension; possibilities for palliative surgery IIIActive complete flexion of wrist and fingers; active wrist extension, weak or absent fingerextension. Good thumb opposition with active intrinsic. Partial pro- and supination IVHand IV with finger extension and almost complete pro- and supination V

Rotation lower arm (Brachial Plexus Work Group Heerlen)

Pronation SupinationAbsent 0 Absent 00°–45° 1 0°–45° 1> 45° 2 > 45° 2Complete 3 Complete 3

Page 196: Brachial Plexus Injuries

Eng GD, Binder H, Getson P, O’Donnell R (1996)Obstetrical brachial plexus palsy (OBPP) outcome withconservative management, Muscle Nerve 19(7):884–91.

L’Episcopo JB (1934) Tendon transplantation in obster-trical paralysis, Am J Surg 25:122.

Fairbank HAT (1913) Birth palsy: subluxation of theshoulder joint in infants and young children, Lancet1:1217–23.

Gilbert A (1993) Obstetrical plexus palsy. In: Tubiana R,ed. The Hand, WB Saunders Company: Philadelphia:579.

Gilbert A, Brockman R, Carlioz H (1991) Surgical treat-ment of brachial plexus birth palsy, Clin Orthop264:39–47.

Gilbert A, Razaboni R, Amar-Khodja S (1988)Indications and results of brachial plexus surgery inobstetrical palsy, Orthop Clin North Am 19(1):91–105.

Green WT, Tachdjian MO (1963) Correction of residualdeformity of the shoulder from obstetrical palsy, JBone Joint Surg 45A:1544–5.

Greenwald AG, Schute PC, Shiveley JL (1984) Brachialplexus birth palsy: a 10–year report on the incidenceand prognosis, J Pediatr Orthop 4(6):689–92.

Hoeksma AF, Wolf H, Oei SL (2000) Obstetrical brachialplexus injuries: incidence, natural course and shouldercontracture, Clin Rehabil 14(5):523–6.

Kennedy MA (1903) Suture of the brachial plexus inbirth paralysis of the upper extremity, BMJ 7:298–301.

Kirkos JM, Papadopoulos IA (1998) Late treatment ofbrachial plexus palsy secondary to birth injuries:rotational osteotomy of the proximal part of thehumerus, J Bone Joint Surg 80A(10):1477–83.

Narakas AO (1987) Obstetrical brachial plexus injuries.In: Lamb DW, ed. The Paralysed Hand. The Hand andUpper Limb, Churchill Livingstone: Edinburgh: 116–35.

Pearl ML, Edgerton BW (1998) Glenoid deformitysecondary to brachial plexus birth palsy (publishederratum appears in J Bone Joint Surg 80A(10):1555–9),J Bone Joint Surg 80A(5):659–67.

Rollnik JD, Hierner R, Schubert M et al (2000)Botulinum toxin treatment of cocontractions after birth-related brachial plexus lesions, Neurology 55(1):112–4.

Sever JW (1918) The results of a new operation forobstetrical paralysis, Am J Orthop Surg 16:248–57.

Specht EE (1975) Brachial plexus palsy in the newborn.Incidence and prognosis, Clin Orthop 110:32–4.

Strombeck C, Krumlinde-Sundholm L, Forssberg H(2000) Functional outcome at 5 years in children withobstetrical brachial plexus palsy with and withoutmicrosurgical reconstruction, Dev Med Child Neurol42(3):148–57.

Torode I, Donnan L (1998) Posterior dislocation of thehumeral head in association with obstetric paralysis, JPediatr Orthop 18(5):611–15.

Troum S, Floyd WED, Waters PM (1993) Posterior dislo-cation of the humeral head in infancy associated withobstetrical paralysis. A case report, J Bone Joint Surg75(9):1370–75.

Waters PM (1999) Comparison of the natural history,the outcome of microsurgical repair, and the outcomeof operative reconstruction in brachial plexus birthpalsy, J Bone Joint Surg 81A(5):649–59.

Wickstrom J, Haslam ET, Hutchinson ET (1955) Thesurgical management of residual deformities of theshoulder following birth injuries of the brachial plexus,J Bone Joint Surg 27A:27–36.

Zancolli EA, Zancolli ER (1988) Palliative surgical proce-dures in sequelae of obstetric palsy, Hand Clin4(4):643–69.

CONSERVATIVE TREATMENT OF OBSTETRICAL BRACHIAL PLEXUS PALSY AND REHABILITATION 187

Page 197: Brachial Plexus Injuries
Page 198: Brachial Plexus Injuries

The neural lesion in obstetrical brachial plexuspalsy is one of traction or stretch. It should beclear that when a nerve is stretched the resultinginjury is very variable. For the most simple ofsituations, such as when a single peripheralnerve is stretched, opinions differ as to theproper treatment (surgery or wait?). Differencesin opinions continue even after deciding that thetreatment is to be surgical (neurolysis orneuroma resection and grafts?).

Controversies are enormous when instead of asingle nerve the highly variable stretch injuryinvolves multiple nerves, located in a poorly acces-sible or understood part of the body, in proximityto vital structures, and do not follow a precise,constant pattern. Such is the situation with tractioninjuries to the brachial plexus. The situation isfurther complicated when the patient is a baby.Objective, precise evaluation is difficult if not impos-sible in the young pediatric patient. In addition,these patients create unavoidable emotions, whichlead to unrealistic hopes of superior regenerativepowers or exaggerated fears of patient brittleness.

It should be no surprise to recognize thesubject of surgical treatment for obstetricalbrachial plexus palsy as one sensitive, difficult,poorly understood and even feared matter. It isa subject full of controversies. This chaptersummarizes the author’s personal opinions andapproach based on over 15 years of experience,which includes visiting and participating insurgery with surgeons in Europe and America aswell as being the primary surgeon in 82 cases.

Anesthesia

Risk

Patients undergoing brachial plexus explorationand reconstruction are usually 3–8 months of

age. Statistics show that the risk of generalanesthesia is highest during the first month oflife (Cohen et al 1990). Neonates therefore havethe highest incidence of intraoperative adverseevents. After the first month the risk of anesthe-sia decreases. In fact, adverse events duringanesthesia are less frequent for ages 1–12months than for ages 1–5 years (Cohen et al1990). Accurate assessment of age and relation-ship to risk calls for calculation of both gesta-tional age and postconceptual age. Both ages aredefined as the time from fertilization to birth.

For our purposes the most common andimportant risk factor relates to the respiratorysystem and the occurrence of postoperativeapnea. Apnea (absence of breathing for 20seconds or longer) can be central (absence ofrespiratory effort) or obstructive. Central apneais associated with prematurity. Preterm infants(having a gestational age of less than 37 weeks)have a higher risk of postoperative apnea up to50 or 60 weeks postconceptual age (Cote 1993).Obstructive apnea is associated with upper respi-ratory tract infections (URI). During the first yearof life a baby has an average of six URIs or aboutone every two months. Each of these episodesmay take longer than two weeks to completelyclear up and not be a factor leading to obstruc-tive apnea. (Tate and Knight 1987). Thus, optimalconditions allowing elective surgery withoutadditional risk are not always found during thesecond or third months of life. If optimal condi-tions are met (40 weeks gestational age,completely cleared URI, appropriate hospital,experienced anaesthesiologist), surgery duringthe second or third month should not representan added risk. Yet no one knows with completecertainty which patients have less anesthesia riskand at what postconceptual or gestational age(Cote 1993). Delaying surgery on a healthy,normal baby after the second or third month is

19Surgical techniqueJose L Borrero

Page 199: Brachial Plexus Injuries

based on the ‘level of comfort’ of the anesthesi-ologist as well as convenience for the surgeonand not on measurable anesthesia risk for thepatient.

Temperature

Hypothermia is a factor to consider in this patientgroup. Besides decreased ability to generateheat, the larger ratio of body surface to bodyweight in babies allows for faster heat loss thanin adults. Hypothermia causes hypoxia, hyper-capnea, acidosis and hypoglycemia. Babies alsoincrease their oxygen consumption with lowertemperatures. Effort must be made to preventheat loss. The room temperature should beraised to maintain the infant’s temperature. Heatlamps and warming blankets are also useful inpreventing thermal stress. The core temperatureshould be monitored during surgery.

Fluids

Given that the operation involves one upperextremity and as both lower extremities are tobe used as graft sources, access to peripheralveins is limited to only one arm. With antici-pation the anesthesiologist should know thatonly the non-injured upper extremity is avail-able for intravenous fluid administration.Experience in placing a reliable intravenousline is important. The inexperienced may findthis difficult, cause unnecessary delays, mayrely on inadequate scalp veins, or need torequest additional help for jugular or subcla-vian catheter placement.

Infants, having a faster metabolic rate and alarger body surface to weight ratio than adults,become dehydrated more easily. An average 6hours fasting period is generally accepted,although several studies have shown no differ-ence in gastric residual volume or pH in childrenallowed to ingest unlimited clear liquids up until2 or 3 hours prior to anesthetic induction(Schreiner et al 1990, Splinter and Schaefer1990, Cote 1993). The deficit created by fastingfor a patient under 10 kg is estimated as 4 ml/kgtimes the number of hours fasting (Savarese

1991). Fifty per cent of this total is to be admin-istered in the first hour. Insensible losses result-ing from evaporation depend on the incisionsize and site, the viscera exposed and respira-tion. These losses are small during brachialplexus surgery.

Maintenance fluid requirements for a babyunder 10 kg are estimated as 4 ml/kg/h. Becausethese operations may take a long time, unrecog-nized hypoglycemia is of particular concern, thusdextrose is included in the maintenance solution(5 per cent dextrose in 0.45 per cent normalsaline). A balanced salt solution (lactatedRinger’s) is used to replace deficits.

Monitoring

For patients between 3 and 8 months old, theaverage weight is between 6 and 8 kg. For thisgroup, the average respiratory rate, heart rate,blood pressure and blood volume are 20–40breaths per minute, 110–180 bpm 60–110 mmHgsystolic pressure, and about 80 ml/kg, respec-tively (Savarese 1991). In this group, urine outputis expected as 0.5–1.5 ml/kg/h. Inhalationanesthesia, hypoperfusion and hyperthermia arecauses of decreased urine output (McGowan andChlebowski 1991).

Surgery

Position

A small folded towel along the posterior midlinebrings the chest forward and allows the shoul-ders to fall back in retraction. The head isslightly turned to the non-injured side with mildneck extension. Localized areas of pressureagainst the scalp should be prevented. The prepincludes both lower legs to the groin, the injuredarm and hand, and the axilla, anterior chest wall,supraclavicular fossa, and neck on the affectedside. Heat loss during this preparation should beminimized. Recording electrodes are not placedbecause in my experience intraoperative record-ing has not been fruitful for key decisionmaking. I recognize opinions differ in this regard(Fig. 1).

190 OBSTETRICAL PARALYSIS

Page 200: Brachial Plexus Injuries

Magnification

Although naked eye dissection of the plexus isboth possible and practical, magnification usingloupes offers undisputed advantages, leading toa more delicate and precise dissection. At least�2.5 magnification is recommended for thegeneral dissection of the brachial plexus. Forprecise dissection of the roots at the foramen,often the operating microscope is best suited.

Incision

For all upper plexus cases and for most totalplexus cases a transverse supraclavicularincision has been my preference for the past 10years. This incision, 2 cm above the clavicle,extends from the lateral border of the sterno-cleidomastoid muscle to the anterior border ofthe trapezius muscle (Fig. 1). This incision allowsdissection of the spinal accessory nerve andaccess to the vertebral foramen of plexus rootsC5, C6, C7, and C8. It also permits exposure ofthe clavicle and all structures within the supra-clavicular fossa. If further exposure is needed,the superior border of the clavicle is exposed. Alongitudinal incision is made along the clavicular

periosteum exposing a small central area of bareclavicle. The clavicle is cut transversely at thispoint. Inferior retraction produces inferiorangulation of the clavicle and provides anadditional centimeter of access to the distalportions of upper and middle trunks retroclavic-ularly (Fig. 2). Realignment of the clavicle with a

SURGICAL TECHNIQUE 191

Figure 1

Position and incision. A small towel roll is placed betweenshoulder blades. The head is slightly turned with mild neckextension. The incision (a) is 2 cm above the clavicle fromthe lateral or posterior border of the sternocleidomastoidmuscle (b) to the anterior border of the trapezius muscle(c). It follows skin lines.

Figure 2

Exposure of distal upper and middle trunks. (A) Skin andtissue over clavicle is reflected inferiorly. Longitudinalincision on clavicular periosteum allows exposure of bareclavicle (1), which is cut transversely. (B) Retracting splitclavicle inferiorly allows additional exposure of distal upperand middle trunks. (a) Neuroma in upper trunk, (b) middletrunk, (c) suprascapular nerve, (d) posterior division uppertrunk, (e) anterior division upper trunk.

A

B

a b

1

ab

cd

e

Page 201: Brachial Plexus Injuries

simple but tight suture through the bone andrepair of the periosteum are sufficient closure forcomplete bone healing.

In total plexus cases a complete clavicularosteotomy and mobilization is necessary when:(a) the subclavian artetry is found ‘high’ (almostsuperior to the superior border of the clavicle),or covering C8: (b) roots C8 ,T1 need completeexposure, particularly if one is not avulsed: (c)the infraclavicular plexus and branches need tobe clearly defined.

In these cases, the posterior end of the skinincision is extended obliquely to below the clavi-cle. The pectoralis major is detached from itsclavicular border exposing the inferior border ofthe clavicle allowing circumferential dissection.All soft tissue over the superior border of theclavicle, and the subclavius muscle inferiorly isdivided, remembering that the suprascapularartery travels in this vicinity. Opposing periosteal

flaps are created and elevated and the bare clavi-cle is exposed at its mid-portion (Fig. 3). Theclavicle is cut transversely and reflected asnecessary for complete exposure of the plexusand access to the subclavian artery.

When closing, clavicular approximation andalignment are maintained with a tight figure-of-eight suture passing through small drill holesnear the divided ends of the clavicle. Periostalflaps are sutured back over the clavicle restoringnormal anatomy.

Surface anatomy

The posterior triangle is bordered by the clavicle,the lateral (or posterior) border of the sternoclei-domastoid muscle and the anterior border of thetrapezius muscle. This triangle is further divided

192 OBSTETRICAL PARALYSIS

Figure 3

Splitting of the clavicle for complete exposure of theplexus. After exposing the clavicle, two opposing, u-shaped periosteal flaps (d) are elevated revealing the bareclavicle which is cut transversely. At closure, the clavicleis approximated with a tight absorbable suture (c) and theperiosteal flaps are sutured back anatomically. (a)Neuroma, upper trunk, (b) suprascapular tissue containingsuprascapular artery (ligated).

Figure 4

Surface anatomy. The posterior triangle is formed by:sternocleidomastoid (8), trapezius (2), and clavicle. Theomohyoid (4) divides the posterior triangle into two smallertriangles. The accessory nerve (3) travels obliquely accom-panied by lymph nodes. Branches from the cervical plexusare superficial and contain mostly sensory fibers: lesseroccipital (1), greater auricular (5), cervical cutaneous (6) andsuprascapular (7) which further divides into anterior middleand posterior branches.

Page 202: Brachial Plexus Injuries

into two triangles by the omohyoid muscle. Thelarger superior triangle contains nerves of thecervical plexus and the spinal accessory nerve.These nerves enter the floor of the triangle at themid-portion of the lateral border of the sternoclei-domastoid muscle. Abundant lymph nodessurround the spinal accessory nerve at this point.That is why injury to this nerve can occur duringnode biopsy. The much smaller inferior trianglecontains the upper plexus trunk and transversecervical vessels. In this triangle the external jugularvein which travels superficially, parallel to thelateral border of the sternocleidomastoid, divesunder the corner where the sternocleidomastoidmeets the clavicle to join the internal jugular vein.On the left side, the thoracic duct ends at the anglebetween the internal jugular and subclavian veins.The sternocleidomastoid muscle may fan into abroad, flat attachment to the clavicle (clavicularportion) (Fig. 4).

Procedure

The transverse incision is deepened through theplatysma. An effort is made to undermine theskin at the subplatysmal plane, both superiorlyand to reach the level of the clavicle inferiorly.This facilitates exposure of the clavicle and lateron closure by allowing identification and approx-imation of the platysmal layer.

Bleeding is controlled with the bipolar cautery.Instead of subcutaneous injection of epinephrinesolution, which only helps while making theincision, gauze moistened with epinephrinesolution is used throughout the procedure. Thisis particularly helpful in controlling bothersomeoozing when operating through the microscope.The solution is made by diluting 1 ml ofepinephrine 1:1000 in 100 ml of normal saline.

The lateral border of the sternocleidomastoid isidentified and mobilized from its clavicular attach-ment. This insertion is divided at the clavicle usingthe cautery, yet leaving a periosteal cuff for repairwhen closing. External and internal jugular veins(and thoracic duct on the left side) lie underneaththis insertion. The lateral border of the sternoclei-domastoid is dissected exposing nerves of thecervical plexus (superficial or cutaneous branches)as they drape over this muscle at its mid-portion.These nerves correspond to the cervical

cutaneous, traveling anteriorly, towards themidline, the greater auricular going superiorly infront of the ear and the lesser occipital, goingsuperiorly behind the ear. At this same level buttraveling towards the clavicle the superficialsensory supraclavicular nerves are seen (anterior,middle, posterior). These nerves should bepreserved for they can also serve as nerve grafts.The spinal accessory nerve can be found at thissame level traveling obliquely in a posterior direc-tion. It lies slightly deeper and posterior, in thevicinity of lymphatic nodes that bleed with ease.Its presence is verified with the nerve stimulator.A long suture loop is used as a marker anticipat-ing it being needed for neurotization later on. Ifind that for this nerve, at this early stage insurgery, small silicone rubber loops (vessel loops)are too big and get in the way.

SURGICAL TECHNIQUE 193

Figure 5

Dissection, upper trunk neuroma. Underneath the omohy-oid, fatty tissue rich in lymphatics (12) is mobilized as aflap hinging posteriorly exposing the plexus.1 – C6 plexus root2 – Omohyoid, cut and reflected3 – Phrenic over anterior scalene4 – Anterior scalene5 – Transverse cervical vessels, ligated6 – Clavicular attachment of sternocleidomastoid, cut and

reflected7 – Internal jugular vein8 – Middle trunk9 – Anterior division, upper trunk

10 – Posterior division, upper trunk11 – Suprascapular nerve12 – Fat pad and lymphatics, retracted by hook retractor13 – Supraclavicular nerve14 – C5 plexus root

Page 203: Brachial Plexus Injuries

Next, the omohyoid muscle is identified,dissected and either divided or pushed to theside. Underneath this muscle, fatty tissue rich inlymphatics covering the supraclavicular fossa isfirst separated from the lateral border of thesternocleidomastoid and anterior scalenemuscle. This tissue is mobilized as a flap,hinging posteriorly. The plexus lies directlyunderneath. During closure this tissue is laidback over the reconstructed plexus serving asuseful protective cover. The transverse cervicalartery and vein are contained in this fatty tissue.To facilitate dissection they can be divided andligated (Fig. 5).

Fibrous portions of the anterior scalene aredissected from the usual neuroma comprisingthe upper trunk. The phrenic nerve may beadhered to this neuroma. Normally this nerve isattached to the upper trunk or to C5 plexus rootby a small branch and travels over the anteriorscalene muscle. Stimulation verifies its presence(Fig. 5).

The C5 plexus root is dissected as far proxi-mal into the foramen as possible. Small vesselswithin the foramen cause bothersome oozingwhen disrupted. They can be controlled withepinephrine soaked sponges and bipolar electro-cautery. The operating microscope may beneeded to obtain a clear, precise dissection ofthe root as it exits the foramen. Reflecting theanterior scalene muscle medially while protect-ing the phrenic nerve allows anterior or medialdissection of the upper trunk and exposes C6and C7 plexus roots. These roots are alsodissected to their foramina and circled withsilicone loops.

The middle scalene is found when dissectingthe upper trunk posteriorly. It may also befibrotic and scarred. Important branches formingthe long thoracic nerve usually run through themiddle scalene at this level. A normal serratusmuscle response, noted when stimulating thesenerves, indicates proximal root continuity withthe spinal cord and that the injury is distal tothese branches. These nerves should bedissected to their origin from C5, C6 and C7plexus roots and protected.

The middle trunk may be matted against theupper trunk as one large neuroma. No attempt ismade to separate these trunks proximally.Instead attention is given distally to the branch-ing of the upper trunk. Posteriorly and just above

the clavicle these branches are the suprascapu-lar nerve, the posterior division and the anteriordivision. The posterior division is usually inclosest proximity to the suprascapular nerve. Thedivisions continue underneath the clavicle.Therefore, if an inferior or retroclavicularexposure is necessary to visualize the divisionsof the upper trunk and the terminal portion of themiddle trunk, the clavicle should be exposed andif necessary divided as described earlier under‘incision’. The middle trunk branching into itsanterior and posterior divisions occurs retroclav-icularly. The middle trunk may have two anteriordivisions, one merging with the anterior divisionof the upper trunk to form the lateral cord andone merging with the anterior division of thelower trunk to form the medial cord.

When separating the middle trunk from C8plexus root, one must consider the dorsal scapu-lar artery. This vessel can be a branch of thethyrocervical trunk or can originate directly fromthe subclavian artery. This dorsal scapular arterymay lie over C8 and the middle trunk but underthe upper trunk (Huelke 1990) (Fig. 6). It is best toexpose, divide and ligate this vessel.

If C8 plexus root lies completely behind thesubclavian artery, safe exposure calls for split-ting the clavicle with inferior retraction of thesoft tissues. When dissecting soft tissuessurrounding the clavicle and dividing thesubclavius muscle, it is necessary to rememberthat the suprascapular artery (branch of thethyrocervical trunk) is in close proximity. It isbest to ligate and divide it (Fig. 3). After gentlemobilization of the subclavian artery, C8 plexusroot can be clearly defined. T1 plexus root isoften found inferior and anterior to C8. It is moretransverse in orientation and maybe muchsmaller than C8. As a rule, T1 and C8 merge toform the lower trunk. This merging is usuallyretroclavicular, thus the lower trunk as such maynot be seen in supraclavicular dissections orunless the clavicle is divided. The lower trunkthen splits into anterior and posterior divisions.The posterior division joins to form the poste-rior cord. This merge can be quite distal andvariable. Sometimes this posterior divisionoriginates directly from C8 (still retroclavicu-larly). In this situation the posterior cord doesnot receive contributing fibers from T1 (Bonnelet al 1980, Bonnel and Rabischong 1981, Bonneland Canovas 1996).

194 OBSTETRICAL PARALYSIS

Page 204: Brachial Plexus Injuries

Reconstruction and decisionmaking

Meticulous exploration allows for definition ofthe anatomical location of the injury. Once theexact location of the injury is determined, precisedefinition of the severity or extent of this stretchinjury is difficult if not impossible. (The physio-logical deficit may not correlate with the anatom-ical distortion). Even under the most favorable ofconditions and availability of resources, predic-tion as to expected regeneration and recovery is

only an educated guess. Reconstruction shouldbe individualized, for no two cases are identical.That is why reports analysing results aftersurgery are often inaccurate unless they includespecific comparable cases.

Neuroma appraisal: neurolysis

Most commonly, a neuroma with some physicalcontinuity is encountered. The surgeon now facesthe first major dilemma. Should the neuroma beresected or a neurolysis conducted?

Neurolysis is, of course, appealing but the sadreality is that it is usually not the best choice.Precise prediction of the regenerative capacity ofthe nerve fibers within a neuroma in continuityrequires exact quantitative and qualitative analy-sis of those nerve fibers within the neuroma.This is not possible today in supraclavicularplexus injuries in babies.

Most surgeons with experience agree that theresults after neurolysis are discouraging.Neuroma resection and grafting is preferred forit gives the best results (Kawabata et al 1985,Sloof 1995, 1997, Gilbert 1996, Birch 2000). Mypersonal experience has been the same. Afterencouraging intraoperative studies, neurolysisalone (no neuroma resection) was carried out intwo cases of C5, C6 rupture. Shoulder functionin these two patients was inferior to the 17cases of C5, C6 rupture treated by neuromaresection and grafting. I believe, however, thatin cases of C5, C6 injury one should inspect andconsider neurolysis of C7. Likewise, microneu-rolysis of C8 should be considered in cases ofupper and middle trunk neuroma (C5, C6, andC7 injury).

Proximal root stump appraisal:histology

After the neuroma is resected, the surgeon facesthe second major dilemma. Is the proximalplexus root stump a satisfactory source of nervefibers suitable for grafting?

Correct appraisal of the proximal stump duringplexus surgery is of critical importance.Connecting nerve grafts to an unsuitable proximal

SURGICAL TECHNIQUE 195

Figure 6

Arteries in the supraclavicular fossa involved in the dissec-tion of the brachial plexus. The dorsal scapular artery (1)crosses over the C8 plexus root. It may continue underthe middle trunk or as illustrated over the middle trunk andunder the upper trunk. Its origin may be from the subcla-vian artery directly (illustrated) or from the thyrocervicaltrunk. Transverse cervical (2), and suprascapular arteries(3) usually arise from the thyrocervical trunk (Huelke 1990).a – Phrenic nerveb – Dorsal scapular nervec – Long thoracic nerve (Bell)d – Upper trunke – Middle trunkf – Suprascapular nerveg – Anterior division of upper trunkh – Lower trunk

Page 205: Brachial Plexus Injuries

root will lead to a poor result. In appraising theproximal root stump the surgeon considers thesecriteria:

1. The combined preoperative evaluation (radio-logic findings, electrodiagnostic studies,physical examination);

2. The gross examination of the root involved(appearance, palpation, how it feels on dissec-tion, how it cuts);

3. The intraoperative muscle response to directnerve stimulation;

4. Special nerve studies during surgery (evokedpotentials, intraoperative nerve conductionstudies);

5. Histological appraisal of the root.

Intraoperative frozen sections usually meandelays and added strain for the surgeon. Ifappropriate rapport and communication areestablished with members of the pathologydepartment, frozen histological sampling can beobtained in a reliable, efficient manner, in lessthan 10 minutes. For the last 7 years histologicalappraisal of the extent of fibrosis (H & E stain),and myelin content (trichrome stain) has beenobtained as a routine in our hospital (Floridahospital Medical Center, Orlando Florida) (Meyerand Claussen 1995, Malessy et al 1999). This, hasallowed certain conclusions:

1. Histological sampling is not needed in everycase, but it is helpful when there is doubt asto the quality of the proximal stump;

2. Histological sampling is a direct and objectivemethod of defining proximal root suitability;

3. Presence of ganglion cells defines avulsion.The root distal to this avulsion may appearhistologically normal. This root is unsuitableand should not be used;

4. It is best to dissect and transect the plexusroot as far proximal into the foramen aspossible.

Grafting

The gap created after neuroma resection isbridged with nerve grafts, which establish nervecontinuity. Segments of sural nerve are mostcommonly used. Portions of the cervical plexus

(supraclavicular nerves) and cutaneous nerves ofthe arm and forearm can also serve as grafts.These are autologous nerve grafts. Sadly, as ofyet, a reliable substitute is not available.

Sural nerve grafts are obtained with a longitu-dinal, posterior, midline incision from thepopliteal fossa to the lateral malleolus. A tourni-quet is not used. Epinephrine soaked spongesand bipolar cautery provides minimal blood loss.Usually, 10–12 cm of sural nerve is obtained perleg in a 6–8 kg baby. On the average, graft lengthfor supraclavicular plexus reconstruction is2–3 cm each. Graft orientation is maintained byplacing a dot in the distal end of the sural nervewith the marking pencil. (The dot is distal.)

Grafts can be coapted individually or as acable made by grouping several nervesegments held together with fibrin glue. Thesecables can be further organized to match theresected portion of the plexus. This is easierdone in a separate surface and then transferredas a unit to the patient where distal and proxi-mal ends are glued or sutured. This methodcalls for great precision in orientation and inestimating graft length and relative diameter ofboth proximal and distal ends so that a nearperfect fit is obtained. Experts like Gilbert canperform this maneuver with great precision andefficiency.

When grafts are sutured individually, I preferto lay the entire sural nerve within sight in theneck wound. This maintains the graft moist andvery accessible. Segments of nerve aremeasured and cut as needed without referring tothe back table. Suction should be used withextreme caution or avoided completely. Theawful empty feeling that accompanies watchingnerve grafts being swallowed into the suctiontubing will stay with you forever. Coaptation iscompleted with a single microsuture of 9–0 or10–0 nylon. It makes sense to anchor deepergrafts first and more superficial grafts last. It ispossible to maintain root fascicular orientationand to coapt grafts to specific fascicular groupswithin the plexus root. The merit gained by thiseffort is speculative.

Conventional suturing calls for passing theneedle through both root and graft epineuron inindependent throws. I find it efficient instead tobring the graft into the needle that is held in itsposition after it passes half-way through theproximal plexus root epineuron.

196 OBSTETRICAL PARALYSIS

Page 206: Brachial Plexus Injuries

Fibrin glueIt is known that the combination of cryoprecipi-tate thrombin and calcium results in a substancethat can be used as a sealant or adhesive. Theusual proportions call for one unit of cryoprecip-itate, 10 000 units of thrombin and 5 ml of 10 percent calcium chloride.

Cryoprecipitate does not contain red or whitecells. It contains antibodies and some coagulat-ing factors. One unit of cryoprecipitate containsa minimum of 150 units of fibrinogen and 80units of factor VIII (Florida Hospital 1999).Cryoprecipitate is obtained from pooled bloodand needs to be cross-matched with the recipi-ent. The sealant or adhesive properties resultwhen fibrinogen is converted to an insolublefibrin matrix. Thrombin, calcium and factor VIIIcatalyze this reaction (Malessy et al 1999).

Commercially available adhesive and sealantcombine the same elements (fibrinogen, throm-bin and calcium) in higher concentrations andmore precise proportions. This results inimproved rheological properties (elasticity,tensile strength, and adhesiveness). In addition,some products contain a substance calledaprotinin, which retards the process of fibrinoly-sis, prolonging the adhesive property.

During plexus surgery, the tissue adhesive isused when bonding several grafts together alongside to form a polyfascicular nerve trunk, whichis then used to bridge the plexus gap created byresecting the neuroma. Some surgeons relysolely on this adhesive to maintain end-to-endcoaptation. Others prefer a combination of oneanchoring suture and fibrin adhesive. It has beenshown that the presence of this fibrin matrix isnot a barrier and does not interfere with thepassage of regenerating axons (Romano et al1991, Palazzi et al 1995).

NeurotizationIn its broadest sense neurotization is defined asnerve regeneration after its division. For thebrachial plexus surgeon, it is synonymous withnerve transfer or nerve crossing. It is to be usedwhen the proximal end of a divided nerve is notavailable.

Nerve transferring requires a donor or proximalnerve source and a recipient or distal nerve. The

proximal source can be from nerves outside theplexus (extraplexual), i.e. the spinal accessoryserving the suprascapular nerve, or within theplexus itself (intraplexual), i.e. the C5 plexus rootserving the lower trunk. A direct nerve transfer isnot always possible. In many instances a nervegraft is needed to bridge the gap between thedonor and recipient nerve. It should be obviousthat these variables defining the nerve transferwill affect results. Anatomical and physiologicaldisparities between donor and recipient nervesalso create variables inherent to the specific trans-fer. The number and type of fibers in the donorand recipient nerves must have some similarity(Bonnel et al 1979, Bonnel and Canovas 1996).

The most commonly used extraplexual transferis the spinal accessory to the suprascapular(Kawabata et al 1994). This transfer can be madedirect, without a graft and without tension whilepreserving the proximal branches to the trapez-ius. Both nerves are about equal in caliber andfiber content. Useful infra- and supraspinatusmuscle function can be obtained with this trans-fer. Other extraplexual sources for transferringinclude the phrenic nerve, the hypoglossus nerve,the cervical plexus, intercostal nerves and thecontralateral C7 root. The deficit created as wellas the expected useful muscle function gained isto be carefully considered prior to carrying outany of these transfers (Brunelli 1984, Kawabataand Shimada 1995, Narakas 1987, Sloof 1995,1997).

Intraplexual transfer dictates using nerve grafts.This is the most accepted method of reconstruct-ing the lower plexus with favorable results. Ofcourse transferring one proximal donor source,like C5 plexus root to multiple distal elementssuch as the lower trunk and the upper trunk intro-duces the problem of fiber dilution or dispersion.It should be obvious that one proximal root aloneis not sufficient or adequate to fulfill all the needsof upper and lower trunks. Available proximalroots used during reconstruction must be knownin order to properly analyse and compare resultsafter reconstruction.

Reconstruction strategy

Reconstruction is dictated by findings at the timeof surgery. Each case needs to be individualized

SURGICAL TECHNIQUE 197

Page 207: Brachial Plexus Injuries

recognizing that variations in the number ofroots affected, as well as, the location and extentof their injury are enormous. Ruptures or extraforaminal injuries can be repaired by resectionand grafting assuming both ends are suitable.Avulsion or intraforaminal injuries are trulyspinal cord injuries and cannot be repaired. Inthese cases, reconstruction is accomplished bydirecting nerve fibers from another source intothe injured distal plexus (neurotization).

Based on injury location, two major categoriesexist: Injuries affecting the upper plexus; andinjuries affecting the plexus totally. The upperplexus category is further divided into two types:those affecting roots C5, C6, and those affectingroots C5, C6 and C7.

According to Gilbert, the most common injuryis one involving two upper roots (C5, C6). Hereports that out of 436 operated cases, 48 percent involved C5, C6 roots, 29 per cent involvedC5, C6, C7 roots and 23 per cent were consideredcomplete (Gilbert and Tassin 1987, Gilbert 1995,1996). My personal experience differs from thatobservation. From a total of 82 cases, 40 (48 percent) involved C5, C6 and C7, 19 (24 per cent)involved C5 and C6, and 23 (28 per cent) wereconsidered total (involving more than the threeupper roots).

Fortunately, the more common upper plexusinjuries are also more commonly extraforaminalruptures, thus repairable. Avulsions are morecommon in central roots (C6, C7 and C8).Isolated upper roots (C5, C6) injuries arefrequent. I have not seen an isolated lower root(C8, T1) injury.

Recognizing the hand as most important for auseful limb, the order of preference duringreconstruction should be (1) hand, (2) elbowflexion, and (3) shoulder. Guidelines for repair ofspecific injuries are included.

1. C5, C6 injury: ruptures. Dissect and transectroots as far proximally as possible into theforamen. Protect and check response afterstimulating root branches to the serratusmuscle (long thoracic nerve) and rhomboidmuscles (dorsal scapular nerve). Expect nearnormal functional results (Mallet IV or better)when grafts bridge suitable healthy proximalstump and a neuroma free distal upper trunk.If there is doubt as to the quality of the distalupper trunk one must not hesitate to transect

further distally and coapt grafts individually tothe suprascapular nerve and to upper trunkdivisions. Fibrosis in the distal upper trunkand its branches should alert the surgeon todissect and inspect C7 plexus root (Fig. 7a).

2. C5, C6 injury: one root ruptured the otheravulsed. Since C5 commonly contains lessfibers than C6, avulsion of C6 with rupture ofC5 leaves the smaller of the two roots as proxi-mal source of axons. This is the less favorablesituation. C6 is also considered to contributemore to biceps than to deltoid. The contraryholds for C5 (Birch 2000). Reconstruction callsfor direct neurotization with distal spinal acces-sory to the suprascapular nerve and grafts fromthe ruptured proximal root to anterior andposterior divisions of the upper trunk. It seemsreasonable to direct the best proximal fascicleto the anterior division of the upper trunk, soproviding the most favorable situation forelbow flexion. If the proximal root stump is asuitable source of fibers, results will be quitegood. I believe that C7 and the middle trunkshould be inspected and neurolysis carried outunless noted perfectly normal (Fig. 7b).

3. C5, C6 injury: both roots avulsed. This uncom-mon plexus injury is seen after breech delivery,where it can occur bilaterally affecting babiesthat are not large (Geutjens et al 1996). Atsurgery if the avulsed roots have remainedintraforaminally the diagnosis of avulsion is notevident as one sees root continuity into theforamen. In addition there may be littleneuroma or fibrosis present. The operativefindings, therefore, do not correlate with theprofound loss in elbow flexion and shoulderfunction noted preoperatively. To make mattersworse, if one decides to cut the root that doesnot look so bad, histology of an avulsed plexusroot will reveal minimal fibrosis and nearnormal myelinization, an observation for whichI personally have not yet found an explanation.

In C5, C6 root avulsion injuries there are noavailable proximal nerve fibers for plexusreconstruction. Neurotization is the only hope.Since the hand is normal in a C5, C6 injury,priorities should be elbow flexion, then shoul-der function, in that order. For elbow flexion,I favor spinal accessory extended with a graftto the musculocutaneous nerve. I use aseparate infraclavicular incision whichseparates the clavicular attachment of the

198 OBSTETRICAL PARALYSIS

Page 208: Brachial Plexus Injuries

pectoralis major allowing its inferior andmedial mobilization and exposes thepectoralis minor which is reflected to eitherside or detached from its insertion at thecoracoid process. Other surgeons prefer

extending the original incision, detaching thepectoralis major from the clavicle and expos-ing the pectoralis minor. Without splitting theclavicle, tunneling is necessary to pass thegraft in both circumstances.

SURGICAL TECHNIQUE 199

Figure 7

(a) C5, C6 injury: rupture. Both proximal root stumps are suitable for grafting. Distal upper trunk with little fibrosis andgood fascicular organization. 7 grafts, 2–3 cm in length. All supraclavicular. Expect an excellent, near normal result. SS,suprascapular; PD, posterior division upper trunk; AD, anterior division upper trunk; UT, upper trunk. (b) C5, C6 injury, oneroot ruptured, the other one avulsed. Theoretically, prognosis is best if the smaller C5 root is avulsed and if C6 root stumpis healthy and normal. Four nerve grafts to upper trunk are shown. Consider neurotization, distal spinal accessory tosuprascapular. Results should be very good. (c) C5, C6 Injury, both roots avulsed. Assuming C7 is perfect and hand isnormal, priorities should be elbow flexion and shoulder function. The problem is not having a proximal source of nervefibers. Neurotization, spinal accessory to musculocutaneous (with a graft under clavicle), and cervical plexus to uppertrunk. Other options described in text. Poor shoulder function is expected. XI, Spinal accessory; CP, cervical plexus; MC,musculocutaneous nerve.

a b

c

Page 209: Brachial Plexus Injuries

Shoulder function will be poor buthopefully the intact medial pectoral nerveprovides for useful pectoral muscle function,resulting in shoulder stability although withexcessive unopposed internal rotation.Branches of the cervical plexus (sensorysupraclavicular) are transferred to the uppertrunk (mostly the anterior division) withoutneeding a graft (Fig. 7c). I do not have enoughexperience to prove this as worthwhile,although a reasonably good result (able toreach mouth with supinated palm of hand,Mallet III) has been obtained in one case.

In an effort to obtain useful elbow flexion,others have considered transferring the spinalaccessory to the anterior division, or themedial pectoral or intercostal to the musculo-cutaneous nerve (Gilbert and Tassin 1987,Gilbert et al 1991, Slooff 1997). Transferring aportion of the ulnar nerve to the musculocu-taneous nerve at the upper arm level is alsoan alternative (Oberlin et al 1994, Sloof 1997,Leechavengvongs et al 1998). Precise datacomparing these alternatives and their resultsdoes not exist. Gains from these efforts maybe modest, yet important, considering theenormous incapacity that results from zeroshoulder abduction and zero elbow flexion ifleft untreated.

Another alternative is to utilize C7 as asource of fibers with end to side graftsleading into the upper trunk. This concept isstill experimental.

In one case I have transferred a large fasci-cle from the uninjured middle trunk directlyinto the upper trunk. A deficit resulting fromdividing portions of a normal middle trunkwas not observed post operatively. Resultsare expected to be favorable for this surgicalprocedure.

4. C5, C6, C7 combined rupture and avulsion. Inthese cases as in all upper plexus problems,hand function is quite good with normal fingerand thumb flexion and normal or weak exten-sion. Priorities should be elbow flexion, shoul-der control and wrist extension. In agreementwith Gilbert, I have observed that the mostcommonly avulsed root noted in this, threeroot, upper plexus injury is C7, particularly ifthe neuroma involves the divisions of theupper trunk (Gilbert and Tassin 1987). Oftenthis root remains partially within the foramen.

With gentle dissection and tugging, the rootseparates from the foramen and the ganglionbecomes evident thus defining the injury as aC7 avulsion.

If three proximal roots are available, multi-ple grafts are used for plexus to plexus repair.It is common not to have enough graftmaterial even after using both surals. Portionsof the cervical plexus and the medianantebrachial cutaneous nerve serve asadditional grafts. The connection with thedistal middle trunk may be retroclavicular andthe clavicle may need to be divided andretracted inferiorly as described under‘incision’. Inspect C8 and the lower trunk andconsider neurolysis if there is any fibrosis(Fig. 8a).

If two or one proximal root is available asa source of fibers, my preference is transfer-ring the spinal accessory to the suprascapu-lar. nerve. Grafts from the healthiest lookingproximal fascicle are directed to the anteriordivision of the upper trunk. Other grafts arethen directed to the posterior division and tothe middle trunk. I do not know if being soselective is worthwhile.

If only one proximal root is available, theproblem is not having enough proximalsource of nerve fibers. It is not possible toadequately supply the upper and middletrunks with one plexus root, especially thesmaller C5. In theory, directing grafts fromone proximal root to all distal recipientsresults in dilution or dispersion of the motorfibers. Muscles may be partially re-innervatedwith limited useful function.

I have directed all grafts from one healthyroot into the upper trunk ignoring the middletrunk completely. The resulting strong elbowflexion with poor shoulder control, weaktriceps and virtually no radial nerve function(zero wrist, thumb or finger extension) wasdisappointing. But, why was the shoulder andradial nerve function so poor? For unknownreasons, regenerating fibers traveled to theanterior division of the upper trunk andignored the suprascapular nerve and poste-rior division of the upper trunk. Anatomicalvariations of the posterior cord is anotherexplanation. In other words, the ignoredmiddle trunk may have been the predominantcontributor to the posterior cord and the

200 OBSTETRICAL PARALYSIS

Page 210: Brachial Plexus Injuries

radial nerve, which becomes even moreimportant when C5 and C6 are missing. Athird explanation is unsuspected fibrosis atthe distal upper trunk. The neuroma resectionwas not carried out distally enough. Or,maybe the proximal root was truly not that‘healthy’. Finally, maybe nerve regenerationthrough a graft simply did not progress asexpected. The point is that sometimes results(good or bad) cannot be explained.

When only one proximal root is available,my preference today is to directly transfer thespinal accessory to the suprascapular and tograft the healthiest proximal group fascicle tothe anterior division of the upper trunk. Allremaining grafts to the remaining upper trunkand to the middle trunk. There is no choicebut to accept dilution or dispersion of fibers.This seems better than directing long graftsinto specific target nerves (Fig. 8b).

5. Total plexus injury: C5, C6, C7 rupture, andavulsion of C8 and/or T1. Priorities should behand (wrist and fingers), elbow flexion andshoulder function. The spinal accessory istransferred to the suprascapular nerve. Graftsfrom the healthiest proximal fascicle aredirected to the lower trunk and next to theanterior division of the upper trunk. Othergrafts if available are directed to the posteriordivision of the upper trunk and to the middletrunk. Portions of the cervical plexus arecoapted without a graft to the middle trunk. Itis difficult to explain to parents and non-surgeon clinicians that results will be poor forsome muscle groups, yet overall superior to notreatment at all (spontaneous recovery) (Fig. 9).

A similar strategy is followed when two oreven one proximal root is available as a sourceof nerve fibers. Results of course, will beinferior in these severe total plexus injuries.

SURGICAL TECHNIQUE 201

Figure 8

(a) C5, C6, C7 ruptured, all proximal stumps suitable, distal upper and middle trunks with little fibrosis. Multiple graftsused for plexus to plexus repair. In this example 10 grafts (2–3 cm each) are shown. The problem is not having enoughgrafts. Expect good results if all proximal stumps are healthy and suitable and distal upper and middle trunk show minimalfibrosis supraclavicularly. UT, upper trunk; MT, middle trunk. (b) C5, C6 C7 injury: two roots avulsed, one root ruptured.Neurotization spinal accessory to suprascapular (extraplexual), grafts from available root to divisions of upper trunk andto middle trunk. Dilution of proximal nerve fibers is significant. Distal upper or middle trunk fibrosis calls for a more distaltransection, thus longer grafts to anterior and posterior division of upper trunk and to distal middle trunk will be needed.This distal connection may be retroclavicular (may need to split the clavicle). The problem may be not having enoughproximal fibers. In this example, two grafts are directed to the anterior division to illustrate priority for elbow flexion. XI,spinal accessory; SS, suprascapular; PD, posterior division upper trunk; AD, anterior division upper trunk; MT, middle trunk.

a b

Page 211: Brachial Plexus Injuries

Postoperative

The baby is immobilized to prevent disruption ofthe grafts. This calls for maintaining the affectedarm in a sling and avoiding the angle betweenhead and shoulder to exceed 90° on the affectedside. It is best to have a splint or custom ‘babyholder’ brace made prior to surgery. This alsohelps parents understand what to expect andhow to care for the baby after surgery. Aminimum of 3 weeks immobilization is necessary.

Babies older than 5 months do not tolerate thisperiod as well as younger ones. This intoleranceand excessive movement can adversely affectresults in those operated on late.

Routine postoperative care includes anobservation period of at least 24 hours. Duringthis time IV fluids are reduced as dietprogresses to normal. Pain and restlessnessare addressed. Monitoring of O2 saturationsmay be necessary.

Complications

In modern day, complications are few. Morbidityis rather low for such a delicate and complexprocedure. Presumed injury to the phrenic nerveresults in transient paralysis of the diaphragmand low O2 saturations post operatively. I havehad a baby with respiratory insufficiency,secondary to presumed aspiration, necessitatingventilatory support for 4 days. Another child wasnoted to have localized hair loss, presumably theresult of localized pressure to the scalp eitherduring surgery or postoperatively while in the‘baby holder splint’.

Conclusions

Brachial plexus surgery in the newborn reopensand expands the frontier of peripheral nervemicroneurosurgery. Surgery calls for precisedelicate technique with a solid knowledge of theanatomy. The surgeon must have a preconceivedoperative plan as well as understand reconstruc-tive priorities and expected results. These arestressful, delicate operations that call forenormous sense of integrity and discipline inorder not to yield to the easiest, fastest surgicalsolution including unjustified delays. In spite ofsuch efforts, important factors affecting resultsare not known or controlled by the operatingsurgeon.

Acknowledgement

All artwork by Jose L Borrero.

202 OBSTETRICAL PARALYSIS

Figure 9

Total plexus injury: C5, C6 rupture, C7, C8, T1 avulsion.In this example two proximal roots allowing 8 grafts areavailable for reconstructing the entire plexus. Prioritiesshould be hand, elbow flexion and shoulder function.Best proximal sources provide grafts to lower trunk andto anterior division of upper trunk. Neurotization of spinalaccessory to suprascapular. Consider neurotization:Cervical plexus to middle trunk (without a graft). In totalplexus injuries, as in all plexus injuries, reconstructionstrategy is variable depending on what is available. Graftshortage and not enough proximal source of fibers areserious problems. Results will vary and major deficit willremain. Still these patients are better off than if nottreated (spontaneous recovery). XI, Spinal accessory; SS,suprascapular; PD, posterior division, upper trunk; AD,anterior division, upper trunk; MT, middle trunk; LT, lowertrunk.

Page 212: Brachial Plexus Injuries

References

Alnot JY, Oberlin C (1996) Nerves available for neuro-tization, 1. The spinal accessory nerve. In: Alnot J,Narakas A, eds. Traumatic brachial plexus injuries.Expansion Scientifique Française: Paris: 33–8.

Birch R (2000) Obstetric brachial plexus palsy. In: GuptaA, Kay SPJ, Scheker LR, eds. The Growing Hands.Mosby: London: 461–74.

Bonnel F, Allieu Y, Bruner P et al (1980) Anatomical andsurgical principles of the brachial plexus in newbornchildren, Int J Microsurg 2(1):12–15.

Bonnel F, Rabischong P (1981) Anatomy and system-atization of the brachial plexus in the adult, Anat Clin2:289–98.

Bonnel F, Allieu Y, Sugata Y, Rabischong P (1979)Anatomico-surgical bases of neurotization for root-avulsion of the brachial plexus, Anat Clin 1:291–6.

Bonnel F, Canovas F (1996) Anatomy of the brachialplexus of the newborn and the adult, macroscopic andhistologic structures. In: Alnot JY, Narakas A, eds.Traumatic Brachial Plexus Injuries. ExpansionScientifique Française: Paris: 3–13.

Brunelli G (1984) Neurotization of avulsed roots ofbrachial plexus by means of anterior nerves of cervicalplexus, Clin Plast Surg 11(1):149–52.

Cohen MM, Cameron CB, Duncan PG (1990) Pediatricanesthesia, mobility and mortality in the peri-operativeperiod, Anesth Analg 70:160–7.

Cote CJ (1993) Pediatric anesthesia. In: Cote CJ, RyanJF, Todres ID, Goudsouzian N, eds. A Practice ofAnesthesia for Infants and Children, 2nd edn. WBSaunders: Philadelphia: 2088–117.

Florida Hospital (1999) Cryoprecipitate, thawed,pooled: Laboratory-Blood Bank Dept., Policy andProcedure Manual, dept/class 711.05, index no. 189.372 AL-1–AL-4.

Geutjens G, Gilbert A, Helsen K (1996) Obstetricbrachial plexus palsy associated with breech delivery,a different pattern of injury, J Bone Joint Surg78B:303–6.

Gilbert A, Brockman R, Carliuz H (1991) Surgical treat-ment of brachial plexus birth palsy, Clin Orthop264:39–47.

Gilbert A (1995) Long term evaluation of brachialplexus injuries, Hand Clin 11(4): 563–81.

Gilbert A, Tassin JL (1987) Obstetrical palsy: a clinical,pathological and surgical review. In: Terzis JK, ed.Microreconstruction of Nerve Injuries. WB Saunders:Philadelphia: 529–53.

Gilbert A (1996) The newborn palsies. In: Alnot J,Narakas A, eds. Traumatic Brachial Plexus Injuries.Expansion Scientifique Française: Paris: 254–64.

Huelke D (1990) A study of the transverse cervical anddorsal scapular arteries, Anat Rec 132:233–5.

Kawabata H, Kwai H, Masatoni T, Yasui N (1994)Accessory nerve neurotization in infants with brachialplexus birth palsy, Microsurgery 15:768–72.

Kawabata H, Masada K, Tsuyuguchi Y et al (1985) Earlymicrosurgical reconstruction in birth palsy, Clin OrthopRel Res 215:233–42.

Kawabata H, Shimada K (1995) Intercostal neurotiza-tion in birth palsy, Microsurgery 16:232–5.

Leechavengvongs S, Witoonchart K, Uerpairojkit C etal (1998) Nerve transfer to biceps muscle using a partof the ulnar nerve in brachial plexus injury (upper armtype): a report of 32 cases, J Hand Surg 23A: 711–16.

Malessy M, Van Duinen S, Feirabend H, Thomeer RT(1999) Correlation between histopathology of C5 andC6 nerve stumps and motor recovery following nervegrafting in brachial plexus repair. In: Malessy M, ed.Brachial Plexus Surgery: factors affecting functionalrecovery. Pasmans BV: Den Haag and Leiden: 61–77.

McGowan FX, Chlebowski SM (1991) Anesthetic manage-ment of children with renal and endocrine dysfunction. In:Bell C, Hughes CW, Oh TH, eds. The Pediatric AnesthesiaHandbook. Mosby-Year Book: St Louis: 320.

Meyer RD, Claussen G, Oh S (1995) Modified trichromestaining technique of nerve to determine proximalnerve viability, Microsurgery 16:129–32.

Narakas A (1987) Obstetrical brachial plexus injuries.In: Lamb DW, ed. The Paralyzed Hand. the Hand andUpper Limb. Churchill Livingstone: Edinburgh: Vol 2:116–35.

Oberlin C, Beal D, Leechavengvongs S et al (1994)Nerve transfer to biceps muscle using a part of ulnarnerve for C5 C6 avulsion of the brachial plexus:anatomical study and report of four cases, J Hand Surg19A:232–7.

SURGICAL TECHNIQUE 203

Page 213: Brachial Plexus Injuries

Palazzi S, Vila-Torres J, Lorenzo JC (1995) Fibrin glueas a sealant and not a nerve barrier, J ReconstrMicrosurg 11:135–9.

Romano VM, Blair SJ, Kerns JM, Wurster RD (1991)Comparison of fibrin glue, bioresorbable tubing andsutures in peripheral nerve repair, Restor NeurologNeurosci 3:75–8.

Savarese A (1991) The preoperative evaluation of thepediatric patient. In: Bell C, Hughes CW, Oh TH, eds.The Pediatric Anesthesia Handbook. Mosby-Year Book:St Louis: 64.

Schreiner MS, Treibwasser A, Keon TP (1990) Ingestionof liquids compared with preoperative fasting pediatricoutpatients, Anesthesiology 72:593–7.

Slooff ACJ (1997) Obstetric brachial plexus lesions. In:Boome RS, ed. The Hand and Upper Extremity: theBrachial Plexus. Churchill Livingstone: New York:14:89–106.

Slooff ACJ (1995) Obstetric brachial plexus lesionsand their neurosurgical treatment, Microsurgery16:30–4.

Splinter WM, Schaefer JD (1990) Unlimited clear fluidingestion two hours before surgery in children doesnot affect volume or pH of stomach contents, AnaesthIntensive Care 18:522–5.

Tait AR, Knight PR (1987) The effects of generalanesthesia on upper respiratory tract infections inchildren, Anesthesiology 67:930–5.

204 OBSTETRICAL PARALYSIS

Page 214: Brachial Plexus Injuries

The basis of the indications for brachial plexusrepair is correct assessment. There is no validdecision without a precise and detailed examina-tion of the patient. Although EMG, CT scan andMRI may be of use in the exact determination ofthe lesion, I strongly believe that the surgicaldecision should be taken on clinical grounds.

The interpretation of the EMG is not easy andvery different from that of an adult, where signsof recovery will usually mean a favorableoutcome. In the baby, these signs mean only thata few fibers have crossed the neuroma andreached the muscle, but not that this muscle willhave a controlled activity.

Examination of the baby andspontaneous recovery

The examination should be done with the babylying down and then sitting, if he or she is oldenough. The individual assessment of muscles israrely possible; it is usually more correct tospeak of functions: abduction, forward elevation,external rotation etc. (Figs 1 and 2). Some ofthese movements are difficult to provoke, and toeffect them requires a very good knowledge ofthe child’s reflexes. However, these tests shouldbe simple enough for a pediatrician, physiother-apist or surgeon to do, and need not be limitedto specialized centers.

In his thesis, Tassin (1984) followed the devel-opment of 44 patients from birth to the end ofrecovery. In these non-operated patients he wasable to trace the relationships between recoveryof individual muscles or functions and the endresult. He demonstrated that if a patient hadsome sort of contraction of the biceps before theend of the third month, the end result would begood enough to avoid direct surgery (Mallet

Grade IV). On the contrary all the patients whohad not demonstrated any biceps recovery at 3months were likely to have an incomplete result(Grade III or less). Based on this study we have

20Indications and strategyAlain Gilbert

Figure 1

The typical presentation of an upper root lesion.

Figure 2

A complete lesion with flail arm and hand.

Page 215: Brachial Plexus Injuries

used these criteria for more than 20 years todecide on the need for surgical exploration of theplexus and we have since proved that operatedpatients are largely improved by surgical repair(Gilbert 1984, 1995, 1998). Several other testmuscles were studied by Tassin but only thebiceps and deltoid were found to be useful. Thedeltoid, however, cannot be used because itsthree components often recover separately andtheir examination is difficult as other musclesmay provoke the movements (pectoralis major,external rotators, etc.). This was the reason forour exclusive choice of the biceps.

In the baby, the muscles cannot be studied onthe same scale as the adult (M0 to M5). Thevoluntary movements are not reproducible in theyoung child and we have therefore been using amuch simpler scale:

M0: no movement, no contractionM1: contracting, no or very limited movementM2: weak movement, incomplete rangeM3: complete range of motion

Indications for surgicalexploration

In 1903, Kennedy thought that the absence ofrecovery at 3 months after birth justified anexploration and a surgical repair. Two decadesago we found, based on the studies of Tassin(1980) that this indication was justified and wehave been using it satisfactorily for the past 20years. However, discussions have arisen fromdifferent surgical teams on the timing of theoperation and the criteria used for the decision.

Many authors (Kawabata et al 1987, Terzis etal 1986) have adopted the 3 month deadline;others prefer to wait longer (4–6 months afterbirth), either for practical reasons if the child isreferred late (Petrolati et al 1994), or for philo-sophical reasons (Clarke, 1995, Grossman 2000).Although recovery of the biceps is the main testadopted by most authors, some, after Clarke(1995), prefer to use a more complex test, givingnumbers to each function and, in particular,giving great importance to the wrist extension.Clarke (1995) claims that by waiting and usinghis grading system, predictability for outcome is7 per cent better than by the simple biceps test.

However, there are drawbacks to the utilizationof this system;

• It is complex and time consuming, and needsa specialized physiotherapist;

• Some patients will start a biceps recovery at4 or 5 months, leading to a poor shoulderresult if not operated on;

• At 6 months, a family may not accept the lossof already-acquired function of the upper limband therefore refuse operation.

In Mediterranean or Middle-eastern environ-ments this latter point can lead to many childrenescaping treatment. The precision for indicationsthat is gained by waiting is balanced by thepsychological difficulties with the decisionwhether or not to operate.

In patients with complete paralysis the situa-tion is different and the aim is no longer biceps

206 OBSTETRICAL PARALYSIS

Figure 3

Horner’s syndrome.

Page 216: Brachial Plexus Injuries

recovery but hand recovery. There are caseswhere the shoulder and elbow recover rapidlyand extensively without hand function. In thesecases one must be prepared to sacrifice some ofthe already recovered upper limb functions inorder to obtain a better hand.

The presence of Horner’s syndrome at the firstexamination will almost certainly lead to surgicalrepair (Fig. 3). Horner’s syndrome is pathog-nomonic and there are very few exceptions. AlQattan has shown that in his experience nopatient with Horner’s syndrome had any satis-factory spontaneous recovery (Al Qattan et al1998). Chuang et al, reviewing older, non-operated patients, states that an overweightbaby (more than 4.5 kg) with Horner’s syndromewill show a poor recovery. On the contrary,Horner’s syndrome is not so indicative in a babyof less than 4 kg at birth (Chuang et al 1998). Theindications of surgical approach are summarizedin Fig. 4.

Special cases

Late exploration

In some cases the child is seen much later, up toseveral years, having a flail arm or a poor recov-ery with or without co-contractions (Fig. 5).

The most common feature is a child seen atthe age of 3 or 4 years, who had a completepalsy: the shoulder and elbow have recoveredsufficiently but the hand is flail or extremelyweak.

The situation is very difficult as, in order toobtain a good chance of recovery in the hand,it would be necessary to sacrifice some of theupper nerve roots. The use of extraplexualneurotization will never allow a good recoveryin the hand. The decision is totally dependenton the family. The surgeon must assess theparents and judge whether they will accept thetemporary (or even definitive) loss of somefunction in order to give the child a chance ofrecovery in the hand. This is one of the mostdifficult surgical decisions. If the whole arm isflail, the problem is not so acute. Then the timelimit will be determined by the survival of themuscles (checked by EMG) and the brain adapt-ability. In obstetrical palsy, if there is noavulsion, the muscles may stay alive for manyyears as some fibers manage to cross theneuroma. However, even if the muscle can bereinnervated after several years, the child maynever use it, as the brain cannot adapt to thisnew situation.

Of course, in any of these late cases, the choicewill be weighed against the possibilities ofmuscle and tendon transfers, which might give asufficient function without the risks of nerve

INDICATIONS AND STRATEGY 207

Obstetrical birth palsy

Incomplete Complete

1–2 months Check for bicepswrist extensors 3 months

3 months

Some biceps No biceps Hand recovery No handrecovery (Horner)

No operation Operation No biceps Operation

Figure 4

Indications for surgical repair. Apatient who has not a recov-ered biceps (even a contrac-tion) at 3 months is a candidatefor surgical exploration andrepair. With complete paralysis,repair will be necessary, evenif the biceps is recovering, aslong as the hand has not recov-ered and/or Horner’s syndromeis present.

Page 217: Brachial Plexus Injuries

surgery. This is true for shoulder and elbow butrarely of the hand.

Breech presentation

In breech presentation, the average birth weight ismuch lower than usual (Geutjens et al 1996, Slooff1997). The specificity of the lesion is due to thehigh percentage of avulsion injuries of the upperroot. When a baby of small weight, born by breechpresentation, is seen with a severe upper palsy,extensive lesions should be expected. The shoul-der is usually flail and the EMG shows absence ofresponse in the C5–C6 muscles.

A myelograph CT scan may show meningoce-les in the upper roots (Fig. 6). The difficulty is therepair. Often in these localizations, explorationshows the roots in situ. They are soft, whitishand do not respond to stimulation. The decisionis not easy as we know that these partialavulsions recover in 50 per cent of cases(Gilbert). It is then best to leave the roots in place

and close the wound. After 6 months a decisionwill be taken according to the recovery. If thereis no or weak recovery, another exploration willallow repair by either neurotization or end-to-side suture (or both) (Fig. 7).

208 OBSTETRICAL PARALYSIS

Figure 5

(a) A 9-year-old non-operated child with no elbow function. (b) A 25 cm graft is done between C5 and the musculocutaneousnerve. (c) Flexion of the elbow after 18 months.

a b c

Figure 6

A CT-myelogram which can show a meningocele.

Page 218: Brachial Plexus Injuries

Conclusions

There is no strategy applicable in all cases ofobstetrical palsy. In some clear-cut cases ofcomplete C5–C6 or C5–C7 lesions with no signsof recovery at 3 months, or in a case of completeparalysis with no hand function and Horner’ssyndrome at 3 months, the decision for surgicalexploration is easy and well understood by thechild’s family. However, sometimes it will not beso easy: at 3 months the surgeon may find nobiceps function but there is some elbow flexiondue to the brachioradialis; the hand has veryweak movements but the elbow flexes. In suchcases it is difficult to convince the family of thedesirability for surgical intervention and nooperation should be done on the child underthese circumstances. It is even worse when thechild is older and a surgical decision and repairwill necessarily mean the loss of some recoveredfunction.

This situation is often made more difficult bycomments made by doctors, ignorant of theproblem and usually giving negative advice.

However, patient support groups often counter-balance this information.

References

Al Qattan NM, Clarke HM, Curtis CG (1998) Theprognostic value of concurrent phrenic nerve palsy innewborn children with Erb’s palsy, J Hand Surg25B:166–7.

Chuang DC, Ma HS, Wei FC (1998) A new evaluationsystem to predict the sequelae of late obstetric brachialplexus palsy, Plast Surg 101(3):673–85.

Geutjens G, Gilbert A, Helsen K (1996) Obstetric brachialplexus palsy associated with breech delivery. A differ-ent pattern of injury, J Bone Joint Surg 78B(2):303–6.

Gilbert A (1995) Long-term evaluation of brachial plexussurgery in obstetrical palsy, Hand Clinics 11:583–95.

Gilbert A, Tassin JL (1984) Réparation chirurgicale duplexus brachial dans la paralysie obstétricale, Chirurgie110:70–5.

INDICATIONS AND STRATEGY 209

a

b

Figure 7

(a) Breech presentation. A myelogram shows two upper meningo-celes. (b) Intraoperative view showing the roots in continents andno neuroma.

Page 219: Brachial Plexus Injuries

Grossman JA (2000) Early operative intervention forbirth injuries to the brachial plexus, Semin PediatrNeurol 7(1):36–43.

Kawabata H, Masada K, Tsuyuguchi Y et al (1987) Earlymicrosurgical reconstruction in birth palsy, Clin Orthop215:233–42.

Kennedy B (1903) Suture of the brachial plexus in birthparalysis of the upper extremity, BMJ I:298.

Petrolati M, Raimond PL, Cavallazzi RM, Saporiti E (1994)Il trattamento microchirurgico delle paralisi ostetriche:15 anni di esperienza, Riv Ital Chir Plastica 26:299–306.

Slooff ACJ (1997) Obstetric brachial plexus lesions. In:Boome RS, ed. The Hand and Upper Extremity, Vol 14,the Brachial Plexus. Churchill Livingston: New York:89–106.

Tassin JL (1984) Paralysies obstétricales du plexusbrachial. Thesis, Université Paris VII.

Terzis JK, Liberson WT, Levine R (1986) Obstetricbrachial plexus palsy, Hand Clin 2:77.

210 OBSTETRICAL PARALYSIS

Page 220: Brachial Plexus Injuries

Evaluation and grading ofresults

To evaluate the results of obstetrical plexusrepair, the use of a clear system of grading isessential. This is the reason why, with PieroRaimondi, we have worked for several years onthe definition of methods of assessment ofoutcome after repair of the plexus.

The evaluation of brachial plexus repair hasalways been very difficult as it encompassesmany muscles and joints and several functions.Some authors (Narakas 1993) have proposedthe use of a purely functional scale. Two diffi-culties arise: the need for voluntary, wellcontrolled, active movement which is not possi-ble in the young child; and the establishment ofa hierarchy of movements. Others (Clarke andCurtis 1995) have been using functions to whichnumerical values have been allotted. Thisallows one to give with one number the overallvalue of the upper limb. However, we find againthe difficulty of comparing different functions:does an upper extremity with an excellentshoulder abduction and rotation, a perfectflexion and extension of the elbow have anyvalue if the hand is flail? In this case, the sumof all the numbers will give an overall resultwhich seems acceptable. Individual testing isnot possible in the baby for every muscle,owing to absence of voluntary movements. Theuse of reflex movements is limited to some raremuscles.

After looking into various systems, evaluatingthem, we found that the only possibility was toevaluate synergic groups of muscles, assessthem in detail but give independent results for

each joint. This does not solve completely theoverall evaluation of the arm but allows precisecomparison of each joint before and after opera-tion.

We are, then, studying the shoulder, the elbowand the hand independently. The follow-up isanother crucial factor. Not only must the child beexamined at regular intervals (every 3 months,then 6, then 12), but check-ups should continueuntil the end of growth. We know too well thedeleterious effect of growth on some results (sizeof the limb, joint abnormality, loss of abduction)to accept results which have not withstood thelong-term evaluation. This puts us in a difficultposition: can we publish results of evaluationbefore the end of growth? That means aminimum of 13 to 14 years follow-up. It wouldbe the ideal situation but, as it is a relativelyrecent technique, it would not enable us toexchange information. The compromise solutionis to publish follow-up details as late as possible,letting the reader know that our results may stillbe modified later.

In order to facilitate comparison, we havedivided the cases in these categories:

• upper root lesions (C5–C6);• upper and middle root lesions (C5–C6–C7);• complete lesions.

Results of upper root lesions

We had reviewed 103 patients operated on forC5–C6 lesions before 1996 (Gilbert 1995). Thesepatients have been reviewed at 2 and 4 years

21Results of repair to the obstetricalplexusAlain Gilbert

Page 221: Brachial Plexus Injuries

post-operation (Figs 1, 2). We have used ourgrading system (see Chapter ••). At 2 years theresults were:

• Grade IV–V (good or excellent) 52 per cent;• Grade III (average) 40 per cent;• Grade II or less (poor) 8 per cent.

212 OBSTETRICAL PARALYSIS

Figure 1

Left C5–C6 repair at 7 years postoperative. (a) Results on externalrotation. (b) Results on abduction. In this case, the shortening is verylimited.

Figure 2

Left C5–C6 repair at 11 years. (a) Good abduction but marked shorten-ing of the left arm. (b) Excellent external rotation.

a

b

a

b

Page 222: Brachial Plexus Injuries

Most of the patients with poor results hadavulsion injury of the upper roots after breechdelivery.

One-third of these C5–C6 patients hadsecondary surgery of the shoulder, either forrelease (13 subscapularis releases) or tendontransfers (33 latissimus dorsi and 6 trapeziustransfers). At 4 years, the new evaluation showed:

• Grade IV–V 80 per cent;• Grade III 20 per cent;• Grade II or less 0 per cent.

Results of upper and middleroot lesions

In C5–C6–C7 defects, the anatomical lesions areusually more severe with many associatedruptures and avulsions (Fig. 3). At 2 years, theresults of 61 patients were:

• Grade IV–V 36 per cent;• Grade III 46 per cent;• Grade II or less 18 per cent.

During the following years, 7 patients had sub-scapularis release and 25 had tendon transfers(24 latissimus dorsi and 1 trapezius). The patientswere evaluated again at 4 years:

• Grade IV–V 61 per cent;• Grade III 29 per cent;• Grade II 10 per cent.

Although these results are not as good as thosefor as the C5–C6 patients, they were largelyimproved by surgery.

Results of complete lesions

A series of 73 patients with complete paralysishas been reviewed with a minimum of 8 yearsfollow-up (Gilbert and Haerle 2000). The patientswere reviewed at 2-year intervals.

At 2 years, the results of recovery in the shoul-der were very limited, with only 25 per cent withGrade III and IV recovery. The upper roots areusually ruptured but a large part is used to re-innervate the avulsed lower roots, thus weakening

RESULTS OF REPAIR TO THE OBSTETRICAL PLEXUS 213

Figure 3

Right C5–C6–C7 repair at 9 years. (a) Slight limitation of medial rotation.(b) Shortening and limitation of external rotation.

b

a

Page 223: Brachial Plexus Injuries

the scapular belt. During the following years someimprovement occurs but the changes are duemostly to tendon transfers (Fig. 4). A review at 8years resulted in 77 per cent of patients of GradesIII, IV and V with 44 per cent of Grades IV and V.

During these years 46 secondary operationswere done, either releases for joint stiffness orlatissimus dorsi transfers.

The results for elbow flexion recovery wereexcellent: 68 per cent of good or excellent results

at 2 years; 81 per cent at 8 years, after only 9secondary operations.

Forearm supination contracture is verycommon after complete palsy. This contracturewas important and needed treatment in 13 cases (6 biceps re-routing and 7 rotationalosteotomies).

In the hand and wrist the early evaluation at 2years was very poor (only 35 per cent of usefulresults 3–4–5) (Fig. 5).

214 OBSTETRICAL PARALYSIS

Figure 4

Complete right paralysis. Avulsion ofC8–T1 repaired with upper roots. (a)Limited results on abduction. (b,c)Excellent function of the hand.

b c

a

Page 224: Brachial Plexus Injuries

During the following years, 22 transfers weredone for wrist extension with only 14 valuableresults. Very few transfers were done for thefingers.

At 8 years hand recovery had improvedmarkedly, with 76 per cent of useful results. Withavulsion injuries of the lower roots, intraplexusrepairs using the upper roots give an acceptableresult, justifying the surgical strategy of repair-ing the lower roots at any cost.

Conclusion

The long-term results of obstetrical brachialplexus repair confirm the advantages of surgicalrepair over spontaneous recovery. In upper rootlesions, the results of shoulder recovery aftersurgical repair were far better than the resultsafter spontaneous recovery in the same grade ofpatients.

In complete lesions, the intraplexual nervetransfers allow three-quarters of patients to

recover a useful hand in cases of lower rootavulsions who otherwise would have had a verypoor and useless hand.

References

Clarke HM, Curtis CG (1995) An approach to obstet-rical brachial plexus injuries, Hand Clin11(4):563–80.

Gilbert A (1988) Tendon transfers for shoulder paraly-sis in children, Hand Clin 4(4):633–42.

Gilbert A (1995) Long-term evaluation of brachialplexus surgery in obstetrical palsy, Hand Clin11(4):583–94.

Gilbert A, Haerle M (in press) The result of treatmentafter complete obstetrical paralysis of the plexusbrachial.

Narakas A (1993) Muscle transposition in the shoulderand upper arm for sequelae of brachial plexus palsy,Clin Neurol Neurosurg 95, Suppl:S89–91.

RESULTS OF REPAIR TO THE OBSTETRICAL PLEXUS 215

Figure 5

Avulsion injury ofC7–C8–T1 (a,b) GradeII hand – poor function.

a b

Page 225: Brachial Plexus Injuries
Page 226: Brachial Plexus Injuries

Patients

During the period January 1986 to January 2000,we examined 896 children with an obstetricalbrachial plexus palsy in our outpatient depart-ment. We performed primary surgery 302 timesin 299 patients. In 239 of these delivery followedcephalic presentation and in 60 the palsyoccurred after breech delivery; one of thesefollowed Caesarean section delivery of a breechpresentation. Breech delivery is complicated,brachial plexus lesions occurring at a higher ratethan in patients who are born after cephalicpresentation (see Chapter 19 in this volume). In57 patients, unilateral primary (nerve) surgerywas performed. Although a bilateral brachialplexus palsy had occurred in 13 patients, bilat-eral primary surgery was only necessary in threeof these. Thus in 60 patients, we performedprimary surgery on 63 sides. Two-year follow-upwas possible for 56 operations carried out on 54patients (Table 1). We report the results of surgi-cal treatment of 53 cases of upper brachialplexus lesions; complete lesions were notincluded because of the small number. In thelatter cases the results of gain in hand functionhave until now been disappointing, and in onechild the follow-up period was just over twoyears, thus distal results could yet be evaluatedas recovery takes many years to develop.

Clinical presentation

The birth weight of the neonates in this series(Table 1) contrasts with our cases of brachialplexus palsy after cephalic presentation (Ubachs

et al 1995). The latter cases are almost invariablymacrosomic neonates.

In all children, it was noted directly after birththat there was a functional disturbance of the arm,although in some at first a fracture was suspectedto be the cause. In fact a local haematoma pointedto the site of the trauma in eight cases (14 percent). In five cases the clavicle was fractured (9per cent), in three cases the humerus (5 per cent),and one child had a fracture of the femur.

22Results of surgery after breechdeliveryGerhard Blaauw, Albert (Bart) CJ Slooff, and Robert S Muhlig

Table 1 Data of 56 cases of surgery performed on 54patients with obstetrical brachial plexus palsy afterbreech delivery

Gender:Female 22Male 32

Birth weight (grams) Range: 1230–4250Mean: 3018SD ± 684

Operated side:Right 31Left 21Bilateral 2

Concomitant lesions:local haematoma 8phrenic paralysis 13fracture of clavicle 5fracture of humerus 3bilateral brachial plexus palsy 13

Distribution:Upper palsy

C5, C6 37C5, C6, C7 16

(Sub)-total palsy 3

Age at operation (months) Range: 3–21Mean: 6.3SD ± 3.9

Page 227: Brachial Plexus Injuries

In 95 per cent of the cases (n = 53) there wasparalysis of the upper roots: in 37 cases (66 percent) of C5 and C6; in 16 (29 per cent) of C5, C6and C7. In these babies the arm is internallyrotated and pronated, and there is no activeabduction or elbow flexion. If C7 is alsodamaged, the extensor muscles are paralysed,the elbow may lie slightly flexed and waiter’s tipposition of the hand is pronounced (Gilbert andWhitaker 1991). In 13 of these babies hemiphrenicnerve palsy was present (see below). Pectoralismajor, finger and thumb flexors are usuallyactive, but muscle atrophy often develops early(Fig. 1). In 13 babies, bilateral lesions werepresent (Fig. 2). Distal sensation and vasomotorcontrol are usually unaffected. A complete palsyof upper and lower roots was present in onlythree cases. In one of these three cases, isolatedearly recovery of suprascapular function hadoccurred. Horner’s syndrome was present; thephrenic nerve was normal.

Phrenic hemiparalysis

Phrenic hemiparalysis was present in 13neonates. Most paediatric surgeons onlyconsider surgical treatment (plication of thediaphragm) to be necessary in those neonateswho have respiratory difficulties. This proved tobe the case in seven patients; in six patientsplication was performed before brachial plexussurgery. One patient who was refused plicationprior to plexus surgery developed serious respi-ratory difficulties directly following plexussurgery. We regarded the cause to be the impactof prolonged surgery and immobilization in thecast. We are convinced that the indication forplication must not be too strict. In some children,hemiphrenic plication may prove to be a disad-vantage; during thoracotomy required for plica-tion, section of the latissimus dorsi may render

218 OBSTETRICAL PARALYSIS

Figure 1

Baby with palsy due to paralysis of C5, C6 and C7. Notewaiter’s tip position of hand and early muscle atrophy.(From Blauuw 1999)

Figure 2

Boy with a bilateral lesion. On the left side, the total lesionshows early proximal recovery. On that side there is alsoa phrenic palsy, which required plication of the diaphragm.

Page 228: Brachial Plexus Injuries

this muscle unsuitable for later latissimus dorsitransfer. At present the indication for phrenicplication has been broadened because of the lessinvasive endoscopic technique now available.The added advantage is that no harm is done tothe latissimus dorsi muscle.

Radiological findings

Radiologically, involvement of cervical dural sacand root sleeves was a frequent finding in thisseries. The myelogram was normal in only sixpatients, showing normal roots and root sleeves.In one patient we refrained from myelographybecause of the child’s clinical condition. Hencethe myelogram was abnormal in 47 patients.Traumatic arachnoid cysts (meningoceles),derived from torn root sleeves, often bilateral,seemed almost to be the rule in obstetricbrachial plexus lesions after breech delivery.Rarely, the cyst reached outside the foraminalcanal. In most cases arachnoid cysts werepresent on the upper roots only. Althoughmyelography was performed in all but one case,the results in our early cases did not allow us todecide on the presence or absence of intraduralrootlets, thus radiologically an accurate diagno-sis of nerve root avulsion was usually not possi-ble. The improvement in the quality ofCT-myelography makes current assessments atpresent more reliable. The reader is referred toChapter 3 in this volume.

Neurophysiology

The interpretation of the neurophysiologicalfindings in brachial plexus lesions differs inneonates and adults, possibly due to anatomicaldifferences. As mentioned previously, brachialplexus lesions following breech delivery have asignificant degree of intradural root involvement;avulsions may be total or partial. In adult cases,the presence or absence of sensory nerve actionpotentials (SNAPs) differentiates between rootlesions and extraforaminal lesions. In neonates,SNAPs are often difficult to measure owing tothe small size of the fingers and hands and tothe lack of cooperation. Also, special electrodes

have to be used. For further information thereader is referred to Chapter 4 in this volume.

Indications for surgery

We follow Gilbert and Tassin’s (Gilbert 1995)criteria, supplemented by those of Clarke andCurtis (1995), to operate when biceps functionhas not returned at the age of 3 months. Gilbertconcluded that babies without a normal deltoidor biceps by the age of three months could notexpect a good outcome, and so considered thisto be the indication for surgical intervention. Weconsider this to also be the case when there isevidence of a severe lesion, such as a Horner’ssyndrome, persisting hypotonic paralysis,persisting phrenic paralysis, and severe sensorydisturbances. The neurophysiological and radio-logical findings may add to the indication tooperate but the decision is usually made onclinical grounds, because the radiologicalfindings are sometimes difficult to interpret andneurophysiological changes may give a rosierpicture than justified by the patient’s finaloutcome (Gilbert and Whitaker 1991, Clarke andCurtis 1995, Slooff and Blaauw 1996a, Slooff1997).

Most patients were operated before they hadreached 6 months of age (Table 1), although 14patients were older than 6 months at the time ofoperation: five were 7 months old, one was 8months, and two were 9 months. Four under-went surgery at 10, 12, 14 and 16 months,respectively, and two at the age of 21 months. Intwo patients who had undergone bilateralsurgery, the second operation was performedwithin one month of the first, well before the ageof 6 months.

Surgical intervention

The surgical findings were grouped as shown inTable 2. We scored certain avulsion when therewere one or several exteriorized root ganglia. Inthe extreme case the ganglion was without anycontact with the foramen, but sometimes themotor and sensory roots had some anatomicalconnection with the foramen, although the

RESULTS OF SURGERY AFTER BREECH DELIVERY 219

Page 229: Brachial Plexus Injuries

ganglion was clearly visible. When we found oneor several soft pale roots which could be tracedup to the foramen, without a neuroma andwithout presentation of the ganglion, and whichdid not respond to stimulation, we scored proba-ble avulsion (avulsion in situ), especially whenthis was supported by the radiological findings.When we found a neuroma, and when the rootsresponded to stimulation we scored: no avulsion;in eight patients there was no avulsion of anyroot. In eight other patients we found certain orprobable avulsion of one or more roots incombination with a neuroma of another root ortrunk (Fig. 3).

Table 3 presents the types of primary interven-tion. Often more than one type of interventionwas performed in the same patient. Intraplexalreconstructions required resection of theneuroma and suralis nerve interposition andsometimes this was added with free transplants

from supraclavicular nerves. Accessory nervetransfer was performed in 44 surgical cases,mainly to the suprascapular nerve (n = 39). In fourcases the nerve was connected to the site of entryof spinal nerve C6 to the upper trunk, and in onecase of complete avulsion of C6, the accessorynerve was connected to its motor root, andsensory nerves from the cervical plexus to itssensory root. Accessory nerve transfer wascombined with transfer of medial pectoral nervesto the musculocutaneous nerve in 17 cases; inone patient motor nerves from the cervical plexuswere added to the neurotization; in one, pectoralnerves were also sutured to the axillary nerve. Inone patient a secondary nerve operation wasperformed which required the suturing of inter-costal nerves to the musculocutaneous nerve,

220 OBSTETRICAL PARALYSIS

Table 2 Surgical findings

Certain avulsion 5Probable avulsion 41No avulsion 8Neuroma spinal nerve C5 12Neuroma spinal nerve C6 1Neuroma upper trunck 10Neuroma upper and middle trunk 1

Figure 3

Intraoperative pictures. On the leftside the figure shows a neuromaof C5 and a possible avulsion(avulsion in situ) of C6. The pictureon the right shows the surgicalresult following resection of theneuroma, and grafting and suturingof C5 to the upper trunk and to thedistal stump of C6. (From Blauuw1999)

Table 3 Surgical interventions

Intraplexal reconstruction 20Accessory nerve to:

Suprascapular nerve 39Other nerve 5

Medial pectoral nerve to:Musculocutaneous nerve 19Anterior division of upper trunk 2Other nerve 3

Hypoglossal nerve 2Intercostal nerves to musculocutaneous nerve 2Other nerve transfer 1

Page 230: Brachial Plexus Injuries

because of failure of the primary transfer to themusculocutaneous nerve. Also, pectoral nerveswere connected to the anterior division of theupper trunk in two cases, to the posterior divisionof the upper trunk in one case and to the entryof C5 into the upper trunk in another case. Thehypoglossal nerve was used to neurotize (Slooffand Blaauw 1996b) the anterior division of theupper trunk twice in order to obtain elbow flexionand to enhance median nerve function.Intercostal nerve transfer was performed twice asstated above, in one case as a secondary proce-dure. We have added transfer of the cutaneousbranch of the musculocutaneous nerve to neuro-tize the brachioradialis muscle in two patients,who had had a pectoral–musculocutaneous trans-fer, and this was successful at follow-up.

When a root was regarded as a probableavulsion, our usual attitude was to leave the rootin place, so that spontaneous recovery couldoccur, because of the possibility of avulsion beingpartial. In these instances we still often performednerve transfers for shoulder control and elbowflexion at the time of the first exploration. Gilbert(1995) found some kind of spontaneous recoveryin 50 per cent of the cases with probable avulsionafter breech delivery, ranging from recovery ofone muscle to full recovery, supporting thepresence of partial avulsion in these cases. Whenrecovery did not take place, he performed asecondary intraplexal neurotization after 1 year.

Table 4 depicts the number and type ofsecondary surgical procedures. In many patientssecondary surgery was not necessary (n = 39)due to a good result; in others more than one procedure was performed successively.Although subscapular release and shortening ofthe coracoid was noted in this table, these proce-dures are not strictly speaking secondary inter-ventions. Also, re-Steindler should be omitted forsimilar reasons, so in fact 18 secondary surgicalinterventions were performed in 11 patients, andin 42 cases no secondary operations as suchwere necessary. The majority of secondarysurgery was performed earlier in this series,possibly because neurotizations for shouldercontrol and elbow flexion have become standardin the last 7 years. In 10 cases one procedurewas performed, in six cases, two procedures andin two cases, three or more. Remarkably in thisseries, subscapular release was only required infour cases and it is notable that joint contractures

rarely occur in obstetrical palsy after breechdelivery, especially in the presence of completeor partial avulsion.

Results

For the assessment of postoperative results weused the Mallet scale (Gilbert 1993) (Fig. 4). This

RESULTS OF SURGERY AFTER BREECH DELIVERY 221

Table 4 Secondary surgical intervention

Anterior subscapular release 4Latissimus dorsi transfer 3Double transfer 4Steindler 9Re-Steindler 3Surgery for wrist extension 2Shortening of coracoid 1

Figure 4

The Mallet scale.

Page 231: Brachial Plexus Injuries

is an assessment of global movement of theextremity, looking at patterns of movement thatmay be either functional or maladaptive. Thisscale can be applied in children aged 3–4 years,who can perform voluntary movements reliablyon command, scoring shoulder functions as wellas elbow flexion. It is particularly useful forassessing results of obstetrical brachial plexuspalsy after breech delivery, as in these casestotal lesions rarely occur and upper lesions arepredominant: in the present series this was 95per cent (n = 53). We used the original five-pointscale with score 1 for no detectable function andscore 5 when the function was normal, althougha three-point scale (as used by Birch) may bemore practical taking the original score 2 asscore 1, and 4 as 3. In our experience, recoveryof shoulder and other proximal muscle re-inner-vation is largely completed 2 years after nervesurgery. Thus, we report only on cases who havea minimum follow-up of 2 years. Only in the caseof secondary surgery will the functions changeat later follow-up.

The results of scoring following the Malletscale are presented in Table 5, which shows thatgrade 3 was scored 70 times, and grade 4 wasscored 155 times. A completely normal function(grade 5) was not scored in any instance and sograde 4 was the maximum score. If all cases hadhad a grade 4 recovery for the respective Malletscale functions, the total number of grade 4scores would have been 265 (5 � 53), hence wemay conclude that only in slightly more than halfof the cases were grade 4 scores given (58 percent). In the earlier cases secondary procedureswere performed more frequently; the Malletscores of these are shown in Table 6. The resultsof the cases where only primary surgery wasperformed (Table 7) contrasts favourably withthese cases.

Accessory nerves were used in 44 cases (83per cent); in 39 the nerve was connected to thesuprascapular nerve. In these patients shoulderabduction of 30–90° (Mallet 3) was found in 17cases, and abduction of more than 90° (Mallet 4)in 19. Active shoulder external rotation 0–20°(Mallet 3) was present in four of these cases andexternal rotation of more than 20° (Mallet 4) in26. This illustrates the effectiveness of acces-sory–suprascapular nerve transfer.

In the present series, where upper lesions werepredominant, the medial pectoral nerve, a

branch of the lower trunk, should be regarded asa good alternative for neurotizing the musculo-cutaneous nerve, and so intercostals were rarelyused. According to Gilbert, we score good recov-ery (M3) when there is full range of activemovement. Due to lack of cooperation in babiesthe MRC five-point scale cannot be used. Goodrecovery of biceps developed in 17 out of 19cases having such neurotization. Good voluntarybiceps function also occurred in two patients inwhom the hypoglossal nerve was used forneurotization.

Table 8 shows the distribution of Mallet sumscores. Scores of 18 or more were regarded as

222 OBSTETRICAL PARALYSIS

Table 5 Results of operations in 53 cases of upperbrachial plexus palsy scored using the Mallet scale

Grade 1 2 3 4

Abduction 3 2 23 25External rotation 11 7 35Hand to nape of neck 8 15 30Hand to back 2 7 11 33Hand to mouth 2 5 14 32Total: 7 33 70 155

Table 6 Results of operations in 11 cases of upperbrachial plexus palsy scored following the Mallet scale,measuring results of both primary and secondarysurgery.

Grade 1 2 3 4

Abduction 3 1 6 1External rotation 3 2 6Hand to nape of neck 7 4Hand to back 2 3 3 3Hand to mouth 2 1 4 4Total: 7 15 19 14

Table 7 Results of operations in 42 patients of upperbrachial plexus palsy scored using the Mallet scale,measuring results of primary surgery only

Grade 1 2 3 4

Abduction 1 17 24External rotation 8 5 29Hand to nape of neck 1 11 30Hand to back 4 8 30Hand to mouth 4 10 28Total: 18 51 141

Page 232: Brachial Plexus Injuries

a good result, as was the case in 29 cases (54per cent), and a fair result, of between 14 and 18was present in 12 cases (23 per cent). Whenpatients are not able to position their good handin a good working position, this must beregarded as a poor result. Generally, this is thecase when the score is less than 15, as found in12 cases (23 per cent). The inability to externallyrotate forms a major component in this group.

Discussion

Generally, the results of both primary and ifnecessary secondary operations in this serieswere rewarding with 54 per cent good results and23 per cent fair results. In both these groups, mostpatients are able to position and use their moreor less normal hand such that they can functionnormally, without serious impairment, accordingto the international classification. In the group thatwe classified as a poor result, many children werestill able to function fairly well by compensating,making the number of cases which were judgedas satisfying by parents still greater.

Our finding of the predominance of an upperbrachial plexus lesion with a great number ofmyelographic abnormalities after breech deliverycan be regarded as typical for this group. Thisfinding is confirmed by earlier reports (Geutjenset al 1996).

In the present series at primary surgery, weaimed at recovery of externally rotate by neuro-tization of the suprascapular nerve in a largenumber of the cases. The figures show that wewere successful in a great number of the cases.Also, the rate of recovery of shoulder abductionwas significant. There is variability in the subjec-tive appreciation of the functional result betweenthe medical profession on the one hand and theparents and the patients on the other. Patientsand parents are entheusiastically positive aboutthe functional result to a greater extent than themedical profession might conclude from the

Mallet scale. A reason for this variability may bethat although the Mallet scale may show thatsome children actually do have impairments,these will often not cause children to suffer fromdisabilities.

The need for intercostal nerve transfer was notgreat in the present series, because of thesuccessful use of the almost invariably well-functioning medial pectoral nerves, althoughintercostals can offer a good alternative. In thisseries intercostal nerves were only used twice.Also, due to the large number of cases of phrenicnerve palsy, we refrained from the use of inter-costals, as in these cases intercostal nerveharvesting may increase the danger of respira-tory difficulties.

The good results after accessory nerve trans-fer and pectoral nerve transfer support ourattitude that this technique is a particularly usefulmethod in cases of brachial plexus palsy afterbreech delivery. This is in accordance with anearlier study concerning an evaluation of 119cases of accessory–suprascapular nerve transferin obstetrical birth palsy after cephalic or breechpresentation (Blaauw et al 1997). Because ofthese good results, we do not think it justified touse other, often more complicated, techniquesfor neurotization in obstetrical birth palsy afterbreech delivery.

Conclusion

Brachial plexus injuries after breech deliveriesare associated with a different pattern of injuryin comparison with birth palsies after cephalicpresentation. The data from this series and fromearlier publications (Geutjens et al 1996) showthat obstetrical birth palsy after breech deliveryis characterized by a high proportion of pregan-glionic injury (partial or complete avulsion) of theupper roots supplying the brachial plexus, illus-trated by traumatic meningoceles in the preop-erative myelogram. In our series, in the majority

RESULTS OF SURGERY AFTER BREECH DELIVERY 223

Table 8 Mallet sum scores in 53 cases of upper brachial plexus palsy, measuring results of both primary andsecondary surgery

Scores 9 10 12 13 14 15 16 17 18 19 20Number 1 1 1 2 7 6 1 5 7 11 11

Page 233: Brachial Plexus Injuries

of these lesions, the damaged root had remainedin the foramen (avulsion in situ). Following thispattern of injury, phrenic nerve lesions and bilat-eral brachial plexus lesions are almost charac-teristic complications of a breech delivery.

In this type of injury we found that the earlyapplication of selective nerve transfers is impor-tant and effective. The combination of accessory–suprascapular and pectoral–musculocutaneousnerve transfers proved particularly rewarding. Ina minority of cases in this series, ruptures andneuromas were present, and classical intraplexalreconstruction with sural nerve transplants wasnecessary and effective, usually added by acces-sory–suprascapular nerve transfer. Remarkably,in this pattern of injury, development of jointcontractures is rare.

References

Blauuw G (1999) Letsels van de plexus brachialis.Tijdstroom: The Netherlands.

Blaauw G, Slooff ACJ, Slooff WB (1997) Evaluation ofaccessory nerve to suprascapular nerve transfer usingfibrin glue in obstetric brachial plexus palsy, ClinNeurol Neurosurg Suppl 1:181–2.

Clarke HM, Curtis CG (1995) An approach to obstetri-cal brachial plexus injuries, Hand Clin 11:563–80.

Geutjens G, Gilbert A, Helsen K (1996) Obstetricbrachial plexus palsy associated with breech delivery.A different pattern of injury, J Bone Joint Surg78B:303–6.

Gilbert A (1993) Obstetrical brachial plexus palsy. In:Tubiana R, ed. The Hand, Vol 4. Philadelphia: WBSaunders: 579.

Gilbert A (1995) Long-term evaluation of brachialplexus surgery in obstetrical palsy, Hand Clin11:583–94.

Gilbert A, Whitaker I (1991) Obstetrical brachial plexuslesions, J Hand Surg 16B:489–91.

Slooff ACJ (1997) Obstetric brachial plexus lesions. In:Boome RS, ed. The Hand and Upper Extremity, Vol 14,the Brachial Plexus. Churchill Livingstone: New York:89–106.

Slooff ACJ, Blaauw G (1996a) Some aspects in obste-tric brachial plexus lesions. In: Alnot J-Y, Narakas A,eds. Traumatic Brachial Plexus Injuries, Monographiede la Societé Française de Chirurgie de la Main (GAM).Expansion Scientifique Française: Paris: 265–7.

Slooff ACJ, Blaauw G (1996b) The hypoglossal nerve.In: Alnot J-Y, Narakas A, eds. Traumatic BrachialPlexus Injuries, Monographie de la Societé Françaisede Chirurgie de la Main (GAM), Expansion ScientifiqueFrançaise: Paris: 50–2.

Ubachs JMH, Slooff ACJ, Peeters LLH (1995) Obstetricantecedents of surgically treated obstetric brachialplexus injuries, Br J Obstet Gynaecol 102:813–17.

224 OBSTETRICAL PARALYSIS

Page 234: Brachial Plexus Injuries

Introduction

Treatment of sequelae of shoulder paralysisvaries depending on the degree of root involve-ment and, consequently, the severity of functionalimpairment.

First, as in all palliative surgery, the main goalis to treat a joint with minimal or no contracturesor deformities. Therefore the first step is to treatany such capsular contracture or muscular retrac-tion to make further muscular transfer easy.

In the pre-microsurgical era, the task was totreat spontaneously-recovered obstetric paralysisat an increased age where there was a significantamount of capsular contracture and muscularretraction and imbalance. The results were veryseldom satisfactory due to the reduced range ofpassive movement of the shoulder joint andsevere bone deformities. The more recent earlymicrosurgical repairs, as shown by Gilbert(1995), Petrolati et al (1994) and Slooff (1995),allowed better functional results, especiallyregarding shoulder function, substantially reduc-ing the need for palliative surgery. Nevertheless,the severity of plexus lesions in many cases(upper roots avulsion in breech presentation andtotal paralysis, in which the upper roots aremainly devoted to reconstructing the lowerplexus (Raimondi et al 1998), with consequentpartial reconstruction of the upper plexus) leadsto the need for secondary surgery. In order toobtain better results, Gilbert and Dumontier(1991) suggested performing very early palliativesurgery, either passive as in capsulotomy,muscle sliding or lengthening, or active as inmuscle transfers.

Different clinical pictures can be describeddepending on different situations:

• Complete flail shoulder in breech presentationwith C5–C6 avulsion in obstetric brachialplexus palsy (grade 0 shoulder followingGilbert’s classification). In these cases there issignificant articular instability of the thoraco-scapular joint with insufficiency of the muscu-lar couples (flexion–extension and elevation–descent); these facts, together with a completelack of intrinsic (rotator cuff) and extrinsicmuscles (deltoid system), lead to lack of eleva-tion of the arm. These cases do not developjoint or bone deformities;

• Incomplete reinnervation (either spontaneousor following microsurgical repair) but acomplete deficit due to extrinsic and intrinsicmuscular imbalance and articular deformities.This category can be assumed as grade 0(Gilbert’s classification);

• Incomplete spontaneous recovery with thetypical intrarotated arm and different degreesof shoulder abduction, with more or lesscomplete lack of external rotation. The degreeof functional impairment varies depending onthe degree of recovery of the different muscu-lar groups; thus clinical pictures also vary andit is not always easy to classify them;

• Incomplete recovery following early microsur-gical repair (with different degree of deficit;the most frequent deficit is external rotationrather than shoulder abduction).

Although it is not our task here to study thebiomechanics of the shoulder joint we mustremember the prominent role of the rotator cuffsystem, which allows the glenohumeral joint toachieve intrinsic stability, complete elevation andexternal rotation (Celli et al 1985, Comtet et al1989). This muscular complex, and particularly the

23Palliative surgery:shoulder paralysisPiero L Raimondi, Alexandre Muset i Lara, and Elisabetta Saporiti

Page 235: Brachial Plexus Injuries

supraspinatus muscle, constitutes a functionalsystem that must be reconstructed (Howell et al1986). This explains the interest in levator scapu-lae transfer to the supraspinatus in shoulder grade0 or functionally grade 0 in order to allow thetrapezius to deltoid transfer to work better andachieve good abduction in the scapular plane(Saha 1967, 1971).

We know that the latissimus dorsi muscle repre-sents the first-choice transfer as a glenohumeralintrinsic stabilizer. The problem arises in theabsence of a strong latissimus dorsi, and in thesecases we utilize the levator scapulae muscle as anintrinsic stabilizer.

Joint contracture release

Before undertaking any surgery it is fundamen-tal to treat the internal rotation contracture of theshoulder (Birch et al 1998) that very oftenappears due to the imbalance in muscular recov-ery (external rotators generally recover partiallyor do not recover at all, while internal rotatorsare usually stronger).

The predominance of internal rotator musclescombined with the absence of deltoid and exter-nal rotator muscles produces, with time, poste-rior subluxation and articular deformity of thehumeral head, hypoplastic glenoid fossa retro-version, posterior and inferior traction of theacromion (especially in cases with partial recov-ery of the deltoid) and lateral and inferior tractionof the coracoid process; this will constitute ablock to the repositioning of the humeral headinto the glenoid fossa and limit the externalrotation due to contracture of the coracohumeralligament. This retracted ligament together withthe predominant function of the subscapular,pectoral major and teres major muscles will limitthe external glenohumeral rotation.

To avoid all these severe joint deformities,which generally develop after 3–4 years of age, itis essential to treat them early. Gilbert andDumontier (1991) recommended not waiting forosseous and articular deformities to occur, butrather treating the lack of passive external rotationwhen it becomes less than 20°. MRI or CT arthro-grams are very useful for evaluating the conditionof the joint and of the articular surface of thehumeral head. It is obvious that if we expect to

release the contracture later (at 7–10 years of ageor later), when bone deformities have developed,we will not have a good chance of success. In thepast Sever’s operation (1927) was very popular,but with long-term follow-up we realized that therewas a constant and significant lack of internalrotation. Even in lengthening the subscapulartendon, which we performed in the past togetherwith the latissimus dorsi transfer, we could notconstantly obtain sufficient external rotation due tothe frequent capsular adhesions at the level of thetendon lengthening. Moreover, the reduction orcomplete lack of internal rotation occurfed veryfrequently. In the last 8 years we have performedthe subscapular release as described by Carliozand Brahimi (1971) and popularized by Gilbert etal (1988), and (Gilbert and Dumontier 1991), whorecommend a very early operation when indicated– that is when the joint is normal and the humeralhead is not flattened.

Subscapular release: surgical technique

The baby is positioned in a lateral position on theoperating table, with the shoulder elevated by apillow and the scapula completely free to bemoved during the operation. The skin incisionfollows the external border of the scapula from theposterior pillar of the axilla to the tip of the scapula,while the shoulder is maintained in abduction byan assistant. The latissimus dorsi is then movedposteriorly and the border of the scapula exposed.Gilbert suggests transfixing the tip of the scapulaby a strong suture, which can easily allow tensionof the bone during the release. The subscapularmuscle is then detached from the scapula, startingfrom the lateral border and by blunt dissectionfrom all the anterior surface of the scapula (Fig. 1).If the muscle is not detached entirely from thescapula, the desired external rotation release willnot be obtained; for this reason special care mustbe taken in detaching it both from the inner borderand the angle of the scapula, remembering that atthe superior border injury to the suprascapularnerve at the notch must be avoided. We then tryto externally rotate the shoulder; sometimes thiscan be achieved only after progressive andrelatively forced maneuvers, which finally releasethe possible associated capsular contractures. Thisis generally the case in older patients with morefixed capsular contractures.

226 OBSTETRICAL PARALYSIS

Page 236: Brachial Plexus Injuries

One or two suction drains are recommendedas a certain degree of bleeding often occurs fromthe periosteal vessels, which cannot be electro-coagulated, especially in the inner or upper partof the muscular detachment.

The arm is then immobilized in externalrotation with the shoulder adducted for 3 weeks.

Once complete passive external rotation hasbeen obtained, a spontaneous recovery of activeexternal rotation due to the reinforcement ofweak external rotator muscles will be observedover a period of a few months. This occurs in atleast 30–40 per cent of cases, and is the reasonwhy we never perform subscapular release andtendon transfer in the early cases. In late cases(at 4 years or more), where recovery of the exter-nal rotators cannot be expected, the two opera-tions can be performed at the same time.

Regarding muscle transfer, two main differentsituations can occur as already mentioned: atransfer in a complete flail shoulder where nocapsular contractures are present, or transfer inan incomplete shoulder recovery; in the formera double muscular transfer will be necessary,while in the latter often a subscapular releasefirst and, in a second stage, a latissimus dorsitransfer will be required.

Transfers in grade 0 shoulders

The role of the scapulothoracic jointAs already mentioned, this type of reconstructionis performed either following breech presentation

outcomes (total flail shoulder) or in shoulders inwhich, despite a certain degree of reinnervation,the functional results correspond to grade 0 inGilbert’s classification. The only difference is thatin flail shoulder neither bone deformity norcapsular contracture are present while in thesecond group they are always present to somedegree.

Correct function of the thoracoscapular joint isfundamental, as it allows correct orientation andthus greater mechanical advantage in the glenoidfossa. The combined movements of scapularbascule are based on a muscular couple consti-tuted by the serratus inferior and rhomboid;scapular suspension is based on the couplelevator scapulae and lower trapezius; thoraco-scapular cohesion in protraction–retractionmovements controlled by the couple constitutedby the middle trapezius and middle serratusallows adequate positioning of the glenoid fossaduring the movement of elevation of the arm,projecting the superior external angle of thescapula forward, upward and laterally. This reori-entation places the intrinsically unstable articularscapulohumeral surfaces in a position of maximalcongruency and mechanical advantage, allowingthe glenohumeral stabilizing action of the rotatorcuff.

The normal function of the thoracoscapularjoint depends on the integrity of the cervicalplexus responsible for innervation of the levatorscapulae, the dorsal scapular nerve for innerva-tion of the rhomboids, and Bell’s nerve for inner-vation of the serratus anterior muscle.

A full clinical examination of the thoraco-scapular muscles helps us to assess the possibleoutcome of secondary reconstructive surgery ofthe glenohumeral complex in cases withoutdeltoid and rotator cuff function.

In sequelae of obstetric brachial plexus palsy,either with partial or absent spontaneous recov-ery and in early operated cases with poor recov-ery, a scapulohumeral rhythm at a cortical levelhas never been integrated by the patient. If weassume this, the transfer of the upper trapeziusto the deltoid will never manage to supply thiscomplicated functional system: it only pretendsto put this primitive elevation mechanismacquired by the patient himself in a morefavourable distal insertion. Transferring theacromial insertion of the upper trapezius as nearas possible to the original insertion of the deltoid

PALLIATIVE SURGERY: SHOULDER PARALYSIS 227

Figure 1

Under the retractor the lateral border of the scapula is visible,from where the subscapular muscle has been detached.

Page 237: Brachial Plexus Injuries

to the humerus converts this muscle from athoracoscapular muscle to a thoracohumeralone, thus jumping two joints.

Consequently in cases in which the inferiorscapular angle and the inner border of the scapulaare not under the control of the serratus-rhomboidsystem (which happens in complete flail shoulderfor upper roots avulsion), the action of the trans-fer will have repercussions on the scapuloglenoidsystem in such a way that, during the superiortranslation of the humeral head, it will project thehead against the acromion, elevating the superiorexternal scapular angle. As a consequence of thisconcept of functional anatomy of the shoulder, wecan understand why the possibilities offered bysecondary reconstructive surgery are limited. Thepresence of structural deformities which stillpreserve a passive range between 110° and 120°of elevation may not influence the results of pallia-tive surgery. Limitations of passive range of exter-nal rotation may be an influence in those patientsin which the transfer allows an elevation of 90°.When they reach 90° of abduction, the greattuberosity is in conflict with the acromion, block-ing the progression of the elevation movement.

Levator scapulae transfer

The levator scapulae muscle arises from thetransverse processes of the atlas and the axisand the posterior tubercles of the transverseprocesses of the third and fourth cervical verte-brae by means of four independent tendons. Thefirst tendon is usually the biggest, and covers theothers. These insertions are usually fused withthe longissimus capitis and splenius cervicistendons posteriorly, and with the middle scalenemuscle anteriorly. The distal insertion of thelevator scapulae is at the border of the upperangle of the scapula. This attachment is made upof short tendinous fibers, which overlap thecostal surface of the serratus anterior fascia.

The blood supply of the muscle is multisegmen-tary. The upper part is vascularized by the ascen-dent cervical artery (terminal branch of thethyreo-cervical trunk), and sometimes by branchesof the vertebral artery. The middle part of themuscle is supplied by the transverse cervical artery,posterior scapular or deep transverse cervical arter-ies (a collateral of the subclavian artery). The lower

part of the levator scapulae is vascularized by theinferior scapular artery, and the anastomoticsystem with the posterior scapular artery.

The nerve supply is also multisegmentary. Thecervical plexus provides from one to four motorbranches and two to four anastomotic loops. Thecervical branches arise from beneath the poste-rior border of the sternocleidomastoid musclefollowing a craniocaudal direction (downward).The dorsal scapular nerve had branches to themuscle in less than 50 per cent of the specimensin our cadaveric study.

Due to its very proximal innervation, the levatorscapulae muscle always functions, even in verysevere total plexus lesions. This fact allows us toutilize this transfer as a glenohumeral intrinsicstabilizer when no latissimus dorsi is available.

Surgical technique

With the patient in lateral decubitus, the surgicalapproach is through a vertical incision parallel to theinner border of the scapula, continuing with anincision following the scapular spine to the deltoidregion. Detaching the trapezius to allow its simulta-neous transfer to the deltoid allows exposure of theupper part of the vertebral border of the scapulaand consequently, the scapular insertion of thelevator scapulae muscle. The levator scapulae isdetached from the scapula with a long strip ofperiosteum along the inner border of the scapula(until the inferior angle) in order to obtain a longtendon for its distal reattachment to the greatertuberosity. It is important to be careful of the nervesand vessels for the rhomboid, which lie in a deeperand more medial plane compared to the levatorscapulae; the surgical plane between the levatorscapulae and the rhomboid is a very safe plane asnerves and vessels for the levator scapulae reachthe muscle anteriorly. Special care must be takenwhen detaching the muscle at its distal insertion tothe upper medial angle of the scapula; it is impor-tant to try to maintain a sufficiently strong continu-ity between the muscle and its prolongationthrough the periosteal strip. After detaching thelevator scapulae from the scapular angle with itsperiosteal prolongation, it is sufficient to free itproximally no more than 2–3 cm until we can seethat the direction of the transplant is completelystraight and in line with the greater tuberosity.

228 OBSTETRICAL PARALYSIS

Page 238: Brachial Plexus Injuries

The distal insertion to the greater tuberosity inchildren is done using 3-0 or 4-0 nylon, while inadults a bone-to-bone fixation is used for a strongerinsertion, harvesting a piece of bone from thescapula at the end of a periosteal strip so that it canbe fixed with a screw to the greater tuberosity.

After detaching the periosteal strip from thescapula, the rhomboid must be reinserted to thescapula in order to preserve its function (remem-ber that rhomboid, together with serratus, isresponsible for bascule movements of thescapulo-thoracic joint).

Trapezius transfer: surgical technique

As the trapezius transfer is normally performed inconjunction with levator scapulae transfer, thepreviously described incision following thescapular spine is used, continuing in a straightline following the middle deltoid to the region ofits insertion to the humerus. In this way the proxi-mal insertion of the deltoid to the acromion andthe distal insertion of the trapezius are exposed.

The trapezius is detached from the overlying skinusing electrocautery, as there may be bleedingfrom the numerous cutaneous vascular branches.

The trapezius is detached from the spine of thescapula, taking care to spare the spinal nerve forthe middle and lower trapezius, which lies 2 cmmedially to the scapula border in the muscularmass of the trapezius; this nerve can easily bedamaged by traction and coagulation. Differentopinions have been expressed regarding the distalbone detaching from the acromion. If theacromion is to be preserved, complete detachmentof the superior trapezius is necessary in order toproject the distal insertion as distally as possible.

Saha has described transfer of the trapezius inpoliomyelitis patients, in which the acromion wassectioned with the external extremity of the clavi-cle, sparing the insertion of the coracoclavicularligaments. A piece from the bone of scapular spineand acromion sufficiently wide to allow a strongand safe distal insertion are thus obtained. Afterdetaching it from the acromion and from thescapular spine, the trapezius is submitted to gentletraction in order to detach it from the underlyingsupraspinatus muscle and free it completely. Thedeltoid, previously detached from its proximalinsertion, is retracted, and the trochiter is exposed.At approximately 2 cm distal to the trochiter the

bone cortex is rasped for 2 mm in depth in orderto prepare for reinsertion of the piece of acromionincluded in the trapezius transfer. This piece ofacromion has to be decorticated on its lowersurface (articular surface). The distal insertion maybe performed with two screws (in adults) or withnon-absorbable suture, e.g. nylon, in children. Thepoint of insertion may vary depending on circum-stances and need; if anteversion (external rotation)is required the distal insertion will be located in amore anterolateral position; if more internalrotation is required (a very rare situation) the inser-tion will be more posterolateral. Throughout thistime the arm must be maintained in abduction toavoid stretching of the trapezius muscle fibers.When the trapezius transfer has been completedover the previous transfer of the levator scapulae,the deltoid is reinserted over the trapezius transferto re-establish a better profile of the shoulder,improving the cosmetic aspect.

Two drainage tubes are positioned, one in thelower angle of the scapula and the other underthe trapezius, and the subcutaneous and skinsuturing is completed.

Immobilization in a plaster cast with shoulderabduction of 90–100° and in external rotation of50–60° is maintained for five weeks.

Comments

Despite different techniques of tunneling andreinforced insertion, the insertion of the trapez-ius to the humerus in some cases undergoesdehiscence requiring re-operation. Moreover, thesparing of the acromion together with the longperiod of postoperative immobilization favorsthe development of adhesions, which in someinstances lead to the need for reintervention. Onthe other hand, the preservation of the acromionprevents cranial translation and subluxation ofthe humeral head, favoring a pulley whichimproves the axis of traction of the levatorscapulae after its transposition to the supraspina-tus. Nowadays we detach only a piece ofacromial bone, sparing the acromial vault.

In some cases early results showed 110–120°of elevation of the arm, which later stabilized at80–90°; in other cases after growth, augmenta-tion of weight and lack of a correct and ongoingrehabilitation program, there was a significantloss of abduction to 60°.

PALLIATIVE SURGERY: SHOULDER PARALYSIS 229

Page 239: Brachial Plexus Injuries

In the situation of performing the transfer of theupper trapezius following Saha’s technique withno stable scapulothoracic joint, transposition ofthe levator scapulae will not produce a centeringaction of the humeral head, but it will producetraction of the humeral head with a conjointmovement of rotation of the glenohumeral jointand elevation of the upper external scapular angle.

The synergy of function between the uppertrapezius and the levator scapulae makes theirconjoint action easy, strengthening the action ofelevation of the shoulder and contralateral spinelateralization, which is normally corticallyintegrated very well by these patients.

In obstetric brachial plexus palsy, we couldnever obtain the same results as Saha (1967,1971) showed in his polio patients in terms ofabduction and external rotation. Nevertheless, agrade 0 shoulder that can reach an average eleva-tion of 80–90° with a tenodesic external rotationof 20–30° has been considered satisfactory.

In a second operation, we can adjoin the latis-simus dorsi transfer to the infraspinatus in orderto obtain a degree of active external rotation. Thistransfer is done through a posterior approach tothe rotator cuff, as will be described later.

Palliative surgery in a partiallyfunctioning shoulder (shouldergrade II–III)

The following transfers are indicated in patientswith grade II or III shoulders (Gilbert’s classifica-tion); that is, in patients with partial recovery ofabduction and partial or absent recovery of exter-nal rotation. Schulze-Berge (1917) was the first tosuggest a transfer of the latissimus dorsi as anexternal rotator, but it is thanks to L’Episcopo(1934) that the technique has been popularized,associated with a section of both subscapular andpectoralis major muscles as suggested by Sever.Hoffer et al (1978) modified the distal insertion ofthe transfer of the conjoint tendon of both latis-simus dorsi and teres major to the rotator cuff atthe point of insertion of the teres minor muscle.

The ideal indication is in cases with a more orless good abduction having a deltoid of strengthM4, but with a weak or absent external rotation.Of course the latissimus dorsi that is to be trans-ferred must also be of sufficient strength.

It is evident that a capsular contracture ormuscular retraction must be treated in order toavoid fixed deformity of the shoulder joint.Today it is easier to establish a correct algorithmwith early treatment of articular deformity and toperform the transfer as a second stage.Frequently after an early subscapular releasethere is a substantial improvement in activeexternal rotation thanks to the re-establishmentof correct muscular balance, which allowsreinforcement of a weak external rotator.

In some instances the patients present verylate, after a spontaneous partial recovery andwith a significant degree of fixed shoulder defor-mity. In these cases the treatment would besimultaneous tendon transfer and correction ofthe deformity by means of tendon lengtheningor muscular release. This was our experienceuntil the late 1980s; at that time the technique weutilized was transfer of the latissimus dorsi(through a unique anterior approach) withassociated coracoid osteotomy and subscapularrelease if necessary (Morelli and Saporiti 1985).

Latissimus dorsi transfer

Anterior approach

The skin incision follows the deltopectoralgroove, running in the axilla up to the lateralborder of the scapula. Osteotomy of the coracoidprocess is performed and the three musclesinserted onto it are reflected. The subscapularistendon is completely exposed. At the scapularborder the subscapular muscle is detachedsubperiosteally from the anterior plane of thescapula, as already described. If this is insuffi-cient and resistance to the passive externalrotation continues to exist, the subscapulartendon is lengthened, taking care to avoidopening the glenohumeral joint and causinganterior luxation of the humeral head.

Once the problem of contractures is solved, wemove on to the transfer. In the past we usuallyutilized the transfer with the conjoint tendon oflatissimus dorsi and teres major; some authorsprefer to separate the tendon of the teres majorfrom that of the latissimus dorsi (Gilbert et al1988, Gilbert and Dumontier 1991). Until the endof the 1980s we utilized the conjoint tendon,which was sectioned as near as possible to its

230 OBSTETRICAL PARALYSIS

Page 240: Brachial Plexus Injuries

bone insertion. The deltoid muscle is elevated,creating a passage for the tendon from behindup to the rotator cuff. The arm is internallyrotated in order to reach the insertion point ofthe teres minor. The conjoint tendon of teresmajor and latissimus dorsi is then passed frombehind under the deltoid, and is fixed with twostitches of 2-0 or 3-0 nylon to the point of inser-tion of the teres minor muscle to the rotator cuff.The stitches are secured with the arm in maximalinternal rotation. If the coracoid has beendetached, it is wired in its original position afterhaving shortened 1–1.5 cm. A plaster castimmobilizes the arm in 100° of abduction andmaximal external rotation with the elbow flexedat 90° for five weeks.

These associated static and dynamic opera-tions in the past gave relatively good results,considering the associated longstanding defor-mity. Today this reconstructive strategy has

changed thanks to a better algorithm with theearly treatment of postural deformities.

The operation we prefer, when passive mobil-ity of the shoulder allows it, is the isolated trans-fer of the latissimus dorsi to the external rotators.

Posterior approach

Nowadays we utilize only the posterior approach;the patient is positioned in lateral decubitus onthe opposite, healthy side. The skin incision runsfrom the posterior border of the deltoid throughthe posterior pillar of the axilla and follows theinner border of the scapula. The latissimus dorsitendon is detached from the humeral shaft andthe muscles isolated for a few centimetersdownwards, which allows better movementupwards to the new insertion (Figs 2 and 3). Weprefer to isolate the neurovascular bundle in

PALLIATIVE SURGERY: SHOULDER PARALYSIS 231

Figure 2

The common tendon of the latissimus dorsi and the teresmajor has been divided and the difference in length of thetwo tendons is easily visible: the latissimus dorsi tendoncan now reach a higher point than the teres major.

Figure 3

In this case the teres major tendon has been transferredto the infraspinatus and the latissimus dorsi passedbeneath the posterior deltoid and transferred to thesupraspinatus.

Page 241: Brachial Plexus Injuries

order to avoid possible tension or lesions. Thehumeral head and the rotator cuff are exposedthrough a transdeltoid incision following themuscle fiber direction. The tendon is passedunder the posterior deltoid and then reinserted ashigh as possible onto the rotator cuff. The tensionof the transplant has to be sufficient, while thearm is maintained in a 120° of abduction andmaximal external rotation; the distal insertion isgenerally performed with a non-absorbablesuture (generally a 3-0 nylon). As usual, a plastercast is used for five weeks.

Regarding the utilization the isolated tendonof latissimus dorsi or the conjoint tendon oflatissimus dorsi and teres major, we agree thatthe course of the two tendons is different(Fig. 2), and moreover the teres major isinserted onto the scapula and limits the ascentof the latissimus dorsi. This fact suggestsseparating the two tendons and utilizing thetendon of latissimus dorsi alone, sectioning theunion with the teres major tendon. In someinstances, when the latissimus dorsi does notseem too strong, it can be supported by theteres major: in this case the tendon is dividedand, after gliding to re-establish the correctrelationship lengthwise, the teres major tendonis fixed once again to the latissimus dorsi butat a lower point. This trick (suggested byGilbert) allows joining of the two forces, avoid-ing elevation of the scapular angle.

Results

The results differ depending on the preoperativefindings (Figs 4, 5, 6). In paralytic shoulder thereal difficulty is not the surgical technique, butrather the correct clinical preoperative evalua-tion, and therefore the best surgical indication.We want to emphasize that there is no indicationin severe bone and joint fixed deformities. Inthese cases the indication may be for a staticoperation such as an external rotation correctivehumerus osteotomy.

To evaluate the results, we must differentiatethe transfer for grade 0 shoulder or functionallygrade 0 from the transfers for the shoulder withsome function spared or recovered. We utilizeGilbert’s scale rather than Mallet’s scale becauseit appears that the functional evaluation is more

strictly related to the shoulder function (abduc-tion and external rotation).

Personal experience with the transfer of thelatissimus dorsi in 43 patients with differentdegrees of deficit ranging from grades I to III(Morelli and Saporiti 1985) gave the resultsshown in Table 1.

On the whole we obtained two grade shoul-ders II (4.6 per cent), 17 grade III shoulders (39.5per cent) and 24 grade IV shoulders (55.8 percent), which represents an important functionalimprovement as compared to the preoperativesituation: nine cases of grade I shoulders (20.9per cent), 31 grade II shoulders (72.1 per cent),and three grade III shoulders (7 per cent).

The first conclusion when looking at theseresults is that all the patients improved theirfunction with surgery. Of course the results arerelated to the severity of the preoperative situa-tion. Moreover, in this series published in 1985,35 per cent of cases had significant joint defor-mities, mainly due to late surgical treatment.This can explain the relatively poor resultsobtained in the group shoulder grade I.

Another conclusion, made retrospectively, isthat we operated only a few patients with agrade III shoulder: this means that, at least in thepast, a grade III shoulder was considered suffi-cient from the functional point of view, and thatparents were not so inclined to undergo a surgi-cal operation with the possibility of only a limitedfunctional upgrading. Nowadays of course wesuggest transfers more often in these cases, aswe are aware of the good results obtainable.

In a series of 44 patients operated using trans-fer of the latissimus dorsi alone and by a poste-rior approach, Gilbert et al (1988) achieved, inmost cases, a considerable improvement inactive external rotation; even grade IV shoulders(where we generally do not advise surgery) could

232 OBSTETRICAL PARALYSIS

Table 1 Results of transfer of the latissimus dorsi

Preoperative grade Postoperative grade (patients) (patients)

I (9) II (2)III (3)IV (4)

II (31) III (14)IV (17)

III (3) IV (3)

Page 242: Brachial Plexus Injuries

PALLIATIVE SURGERY: SHOULDER PARALYSIS 233

Figure 4

(a) An 18-month-old childwith a spontaneouslyrecovered OBPP with95° of shoulder abduc-tion and no externalrotation, neither activenor passive due to amarked retraction ofsubscapular muscle andanterior capsular contrac-ture (Gilbert grade IIIshoulder). A subscapularrelease and latissimusdorsi transfer to therotator cuff have beenperformed with a singleantero-lateral surgicalapproach. (b–d) Thepatient 2 years aftersurgery, with goodrecovery of active exter-nal rotation and a signifi-cant improvement inshoulder abduction. Thescar is hardly visible atthe lateral border of thescapula, the axilla andthe deltopectoral groove.

a b

c d

Page 243: Brachial Plexus Injuries

234 OBSTETRICAL PARALYSIS

Figure 5

(a,b) An 8-year-old boy with aspontaneous recovered OBPPwith 90° of shoulder abduction,no external rotation and severecapsular contracture andsubscapular muscle retraction(Gilbert shoulder III). Throughan antero-lateral approach asubscapular release from thescapula, a coracoid partialresection and latissimus dorsi+ teres major muscle transfersto the rotator cuff have beenperformed. (c–e) Very goodactive external rotation andshoulder abduction 2 yearsafter surgery.

c d e

a b

Page 244: Brachial Plexus Injuries

PALLIATIVE SURGERY: SHOULDER PARALYSIS 235

a b c

d e f

Figure 6

(a,b) A 12-year-old boy with spontaneous evolution of a severe total OBPP with a shoulder grade 0. Shoulder palliativesurgery had been indicated due to the presence of some function at the hand. A triple transfer has been performed: (c)The trapezius transfer is lifted up with a piece of acromion, while underneath the levator scapula has been harvestedwith a periosteal strip to reach the supraspinatus insertion; laterally, the latissimus dorsi has been prepared for transferto the infraspinatus. (d–f) Functional result 1 year later. The patient can achieve 95° of abduction in the scapular plane, asatisfactory result if we consider that it has given an upgrade from a grade 0 shoulder to a grade III shoulder. The lackof elbow extension is now the main problem.

Page 245: Brachial Plexus Injuries

reach an average of 35° of external rotation. Of23 cases of grade III shoulders, 13 could obtainan average of 45° improvement in shoulder exter-nal rotation (upgrading to grade IV shoulder) andfour obtained an average improvement of 36°. Offive cases of grade II shoulders only twoupgraded to grade III. In a more recent series,Gilbert and Haerle (in publication) have followed122 cases with a minimum of 3.8 years follow-upand an average of 5.9 years. Of these 122 cases,52 had been operated after C5C6 paralysis, 47after C5C6C7 paralysis, 3 after C5C6C7C8 lesions,and 20 following complete paralysis. On average,the transfer had been done 3.2 years after theplexus repair. The average improvement inabduction was 29° after 1 year, 31° after 3 years,but only 15° after 5 years, showing the late deteri-oration of the result, probably due to disuse. Theresults regarding external rotation were almostconstant, with an 83 per cent improvement at fiveyears.

Results for grade 0 shoulders

A different situation is found in terms of resultsin the cases of transfer in grade 0 shoulders.

We studied a series of 27 patients with a grade0 shoulder in which we performed the followingtype of secondary reconstruction:

Levator scapulae transfer 1 caseTrapezius + latissimus dorsi + levator

scapulae 16 casesTrapezius + levator scapulae 6 casesTrapezius + teres major + latissimus

dorsi + levator scapulae 4 cases

The following results were achieved.

Latissimus dorsi + trapezius + levatorscapulae (16 cases)

Spontaneous Cases Results Complicationsrecovery

Good scapulo- 12 Gilbert Myolisis thoracic grade III 1 case

11 casesDeficit of 0scapulo-thoracic

Post- Cases Results Complicationsmicrosurgery

Good scapulo- 2 Gilbert Nothoracic grade III

2 casesDeficit of 2 Gilbert Noscapulo-thoracic grade II

2 cases

Trapezius + levator scapulae (six cases)

Spontaneous Cases Results Complicationsrecovery

Deficit of 6 Gilbert Noscapulo-thoracic grade II

6 cases

Trapezius + latissimus dorsi + teres major+ levator scapulae (four cases)

Spontaneous Cases Results Complicationsrecovery

Good scapulo- 3 Gilbert Nothoracic grade III

3 cases

Post- Cases Results Complicationsmicrosurgery

Deficit of 1 Gilbert Noscapulo-thoracic grade II

1 case

Levator scapulae (1 case)

Post- Cases Results Complicationsmicrosurgery

Good scapulo- 1 Gilbert Nothoracic grade III

1 case

It is evident that in this category of patients theresults are less good than in the previous group;in the cases with the best results we managed toreach grade III of Gilbert’s classification. Theexternal rotation obtained is characteristic of adynamic tenodesis effect and is approximatelyabout 20°, and 30° as an integrated movement in

236 OBSTETRICAL PARALYSIS

Page 246: Brachial Plexus Injuries

the elevation of the arm in the scapular planewith a tendency to flexion in the sagittal plane.In the poorest cases the shoulder achieved goodstability, which could help elbow flexion andshoulder abduction, with projection of the handbetween 30° and 60° allowing a totally uselessarm to be transformed into a useful support tothe contralateral healthy arm.

The main problem is the progressive impair-ment of function with time that was observed ina certain number of cases. After surgery wenormally immobilize the shoulder with plaster forfive weeks, and the plaster or a splint is thenmaintained for at least three months andremoved temporarily only for the rehabilitativeactive and passive exercises. Later on, followingcomplete removal of the splint, the patients aresubmitted to a rehabilitative program, withspecial care not to force adduction and internalrotation, for one or two more months. We try inthis way to avoid functional overcharge of thisnew muscular system, which can gradually returnto its preoperative situation. During the immedi-ate postoperative period and later on during thefirst months of the rehabilitation program there isa high risk of myolisis. In some cases we havebeen forced to reoperate the patient and start therehabilitation program again.

The key to good, long lasting results is thepresence of correct scapulo-thoracic motoriza-tion, which allows stabilization of the scapularsuspension and bascule. This facilitates thecombined action of these muscles, avoiding theovercharge provoked by the conjoined elevationof the humerus and of the superior externalangle of the scapula due to the action of activemobilization of the inferior scapular vertex.

Conclusions

Palliative surgery in obstetric brachial plexusparalysis must not be considered as an obsoletetreatment. We are aware that early microsurgicalrepair of the brachial plexus leads to substantialimprovement in terms of functional results ascompared to a spontaneous late recovery.Nevertheless, in many instances it is still neces-sary, especially for the shoulder area, to performmuscular transfers to improve the results further.This is especially true in avulsion lesions of

upper roots or in total paralysis, in which theupper roots are mainly devoted to reconstructionof the lower plexus.

Generally the results are much better thanthose obtained in the past using palliativesurgery alone.

What has changed is the global strategy inobstetric paralysis treatment: early microsurgicalrepair plays the most important role, as it givesthe child the unique opportunity to recover to thebest degree and especially in total paralysis, itgives the opportunity of recovering useful handfunction. However, palliative surgery plays suchan important role in the general reconstructiveplanning that the surgeon who performs micro-surgery must already have in mind all the possi-ble surgical steps and the correct timing torealize them.

References

Bateman JA (1954) The Shoulder and Environs. CVMosby: St Louis.

Birch R, Chen L (1996) The medial rotation contractureof the shoulder in obstetric brachial plexus palsy, JBone Joint Surg 73B(suppl.1):68.

Birch R, Bonney G, Wynn Parry CB (1998) Birth lesionsof the brachial plexus. In: Surgical Disorders of thePeripheral Nerves. Churchill Livingstone: Edinburgh:209–33.

Carlioz H, Brahimi L (1971) La place de la désinsertioninterne du sous-scapulaire dans le traitement de laparalysie obstétricale du membre superieur chezl’enfant, Ann Chir Infantile 12:159–68.

Catalano F, Fanfani F, Mazzone V (1985) The dynamicoperations: tendon transfer of latissimus dorsi andteres major, Chirurgia della Mano 22(1):59–65.

Celli L, Balli A, De Luise G, Rovesta C (1985) Some newaspects of functional anatomy of the shoulder, Ital JOrthop Traumatol 11:83–91.

Clarke HM, Curtis GC (1995) An approach to obstet-rical brachial plexus injuries, Hand Clin11(4):563–80.

Comtet JJ, Herzberg G, Al Naasan I (1989)Biomechanical basis of transfers for shoulder paralysis,Hand Clin 5:1–14.

PALLIATIVE SURGERY: SHOULDER PARALYSIS 237

Page 247: Brachial Plexus Injuries

Gilbert A, Romana C, Ayatti R (1988) Tendon transfersfor shoulder paralysis in children, Hand Clin4(4):633–42.

Gilbert A, Dumontier C (1991) Etude clinique et évolu-tion spontanée des paralysies obstétricales du plexusbrachial. In: Traité de Chirurgie de la Main, Tubiana R.Ed. Vol. 4. Masson Paris: 610–34.

Gilbert A (1995) Long-term evaluation of brachialplexus surgery in obstetric palsy, Hand Clin11(4):583–94.

Harmon PH (1950) Surgical reconstruction of theparalytic shoulder by multiple muscle transplantations,J Bone Joint Surg 32A:583–95.

Hoffer MM, Wickenden R, Ropper B (1978) Brachialplexus birth palsies: results of tendon transfer torotator cuff, J Bone Joint Surg 60A:691–5.

Howell SM, Imobersteg AM, Seger DH et al (1986)Clarification of the role of supraspinatus muscle inshoulder function, J Bone Joint Surg 68A:398–404.

Itoh Y, Sasaki T, Ishiguro T et al (1987) Transfer of latis-simus dorsi to replace a paralysed anterior deltoid. Anew technique using an inverted pedicle graft, J BoneJoint Surg 69B:647–51.

Jacchia GE, D’Arienzo M, Pavolini B (1985) The staticoperations for the sequelae of obstetric paralysis in theshoulder, Chirurgia della Mano 22(1):45–57.

Johnston TB (1937) The movements of the shoulderjoint. A plea for the use of the ‘plane of the scapula’ asa reference for movements occurring at thehumeroscapular joint, Br J Surg 25:252.

L’Episcopo JB (1934) Tendon transplantation in obstet-rical paralysis, Am J Surg 25:122–5.

L’Episcopo JB (1939) Restoration of muscle balance inthe treatment of obstetric paralysis, NY State J Med39:357–63.

Mayer L (1927) Transplantation of the trapezius forparalysis of abductors of the arm, J Bone Joint Surg9:412–20.

Morelli E, Saporiti E (1985) Mixed static–dynamic inter-ventions: surgical treatment of shoulders in outcomesof obstetrical palsy, Chirurgia della Mano 22(1):67–74.

Narakas AO (1993) Muscle transpositions in the shoul-der and upper arm for sequelae of brachial plexuspalsy, Clin Neurol Neurosurg 95(Suppl.):89–91.

Petrolati M, Raimondi P, Cavallazzi RM et al (1994)Microsurgical treatment of obstetrical palsies: ourexperience in 15 years of surgery, Riv Ital Chir Plastica26(3):299–306.

Price AE, Grossman JAI (1995) A managementapproach for secondary shoulder and forearm defor-mities following obstetrical brachial plexus injuries,Hand Clin 11(4):607–17.

Raimondi P, Petrolati M, Cavallazzi RM et al (1998)Obstetric palsy: microreconstruction of lower roots. In:Roth JH, Richards RS (eds). 7th Congress of the IFSSHVancouver (Monduzzi International ProceedingsDiv):69–75.

Saha AK (1967) Surgery of paralysed and flail shoul-der, Acta Orthop Scand 97(Suppl.): 9–90.

Saha AK (1971) Dynamic stability of the glenohumeraljoint, Acta Orthop Scand 42:491–505.

Saloff-Coste J (1966) A propos du traitement desparalysies obstétricales du plexus brachial: désinser-tion du sous-scapulaire sans capsulotomie, Rev ChirOrthop Reparatrice Appar Mot 52:395–400.

Scaglietti O (1938) The obstetrical shoulder trauma,Surg Gynecol Obstet 65:867–77.

Scaglietti O (1941) Le Lesioni Ostetriche Della Spalla.Licinio Cappelli Ed: Bologna: 1–147.

Schulze-Berge A (1917) Propfung des Nervus axillarisin den Ulnaris und Ersatz der Heber des Oberarmes(Deltoideus und Coracobrachialis) durch denLatissimus dorsi, Zentralbl Chir 25:551–4.

Sever JW (1916) Obstetric paralysis: its etiology,pathology, clinical aspects and treatment, with a reportof 470 cases, Am J Dis Child 12:541.

Sever JW (1927) Obstetrical paralysis, Surg GynecolObstet 44:547–9.

Slooff ACJ (1995) Obstetric brachial plexus lesions andtheir neurosurgical repair, Microsurgery 16:30–34.

Wickstrom J (1962) Birth injuries of the brachial plexus:treatment of defects in the shoulder, Clin Orthop23:187–96.

Zachary RB (1947) Transplantation of teres major andlatissimus dorsi, Lancet 253(11):757–8.

Zancolli EA, Zancolli ER (1988) Palliative surgical proce-dures in sequelae of obstetrical palsy, Hand Clin 4:643–70.

238 OBSTETRICAL PARALYSIS

Page 248: Brachial Plexus Injuries

Introduction

The first description of obstetrical birth palsy wasgiven by Smellie in 1764. Duchenne in 1872 andErb in 1874 described upper brachial plexuspalsy, and in 1885 Klumpke gave accounts ofpalsy, its lower part which includes C8 and T1,and of complete palsy. Twenty years later, Fieuxquestioned Erb’s theory, believing that the lesionis caused by overstretching (Egloff et al 1995).

It is generally accepted that obstetrical brachialplexus palsy occurs as a result of extreme lateraltraction on the head of the infant away from theshoulder during the last phase of the delivery.According to Metaizeau et al, a force of between20 and 40 kg is necessary to produce a lesion(Clarke and Curtis 1995). Most of the theseinjuries resolve without operative intervention,and patients recover with minor or no residualfunctional deficits. Twenty-five per cent ofestimated nerve fibres within the brachial plexuspass to the shoulder girdle (Birch 1995).Shoulder and scapular muscle innervations areshown in Tables 1 and 2.

Most infants who show signs of recovery inthe first month generally have normal function,but if they do not recover in the first month oflife they have a considerable risk of long-termlimited strength and range of motion. As thedelay in recovery extends, the risk increasesproportionately (Waters 1997). A number needplexus repair. The clinical examination is impor-tant and EMG predictions frequently have noclinical correlation in infants. If recovery of thebiceps has not begun at three months, thefunctional prognosis is poor and plexus repair isindicated (Gilbert et al 1991, Gilbert 1995,Yücetürk 1996, Waters 1997, Chen et al 1999).

The characteristic posture of the shoulderfollowing upper type obstetrical birth palsy isone of internal rotation and limited abduction.The primary abductors, the supraspinatus anddeltoid muscles are affected.The powerfulexternal rotator infraspinatus, is also palsied orparalysed. The shoulder external rotator may be

24Palliative surgery: tendon transfers tothe shoulder in childrenAydın Yücetürk

Table 1 Shoulder muscles and innervations (Kendall et al1993)

Flexors: anterior deltoid (C5–6), biceps (C5–6), pectoralismajor upper (C5–6–7), coracobrachialis (C6–7)Abductors: deltoid (C5–6), supraspinatus (C5–6) bicepslong head (C5–6)Lateral rotators: infraspinatus (C5–6), teres minor (C5–6),posterior deltoid (C5–6)Extensors: posterior deltoid (C5–6), teres major (C5–6–7),latissimus dorsi (C6–7–8), triceps long head (C6–7–8–T1)Adductors: pectoralis major (C5–6–7), teres major(C5–6–7), latissimus dorsi (C6–7–8), triceps long head(C6–7–8–T1)

Table 2 Scapular muscles and innervations (Kendall et al1993)

Abductor: full flexion: serratus anterior (C5–6–7–8)Lateral rotators: serratus anterior (C5–6–7–8), trapezius(nerve XI – accessory) and ventral ramus (C2–3–4)Adductor: full abduction: trapezius (nerve XI – accessory)and ventral ramus (C2–3–4)Lateral rotators: trapezius (nerve XI – accessory) andventral ramus (C2–3–4), serratus anterior (C5–6–7–8)Adductors: medial rotators and elevators – full extension:rhomboids (C4–5), levator scapula (3–4–5)Ant. tilt of scapula by: pectoralis minor (C6–7–8–T1)Adductors: to side against resistance: rhomboids (C4–5),trapezius (nerve XI – accessory) and ventral ramus(C2–3–4)

Page 249: Brachial Plexus Injuries

functioning but its force is not enough againstthe strong internal rotators and shoulder adduc-tors such as the pectoralis major, subscapularis,teres major and latissimus dorsi (Hentz 1999,Waters and Peljovich 1999).

Motion about the shoulder requires the coordi-nated moment of four joints and over 20muscles. Thirty-three per cent of shoulder eleva-tion is from the scapulothoracic joint. Sahabelieved that good elevation requires primemover (deltoid or pectoralis major), steeringgroup (supraspinatus, infraspinatus andsubscapularis) and depressor group (pectoralhead of pectoralis major, latissimus dorsi, teresmajor and teres minor muscles) (Price andGrossman 1995).

Biceps recovery at later than three monthsgenerally needs secondary surgery (Gilbert et al1991, Gilbert 1995). Waters documented thenatural history of brachial plexus birth palsy inrelation to the recovery of biceps function. Infantswho had recovery of biceps function during thefourth, fifth or sixth months of life later hadsignificantly worse function, according to thecriteria described by Mallet, than those who hadrecovery in the first three months (Waters andPeljovich 1999, Waters 1999). According toGilbert, after plexus repair one third of Erb’s palsypatients need secondary surgery (Gilbert 1995).

Shoulder problems in obstetricalbirth palsy

In 1913 Fairbank focused on the frequency ofsubluxation of the shoulder in obstetrical birthpalsy. Moore, in 1939, noted that internalrotation contractures with external rotation palsyare often associated with a posterior dislocation(Egloff et al 1995; Fig. 1). Babbitt and Cassidydescribed anterior and inferior dislocation of theshoulder related to a dynamic phenomenon ofmuscular imbalance, and according to theirobservations dislocation took place during thefirst six months (Egloff et al 1995).

240 OBSTETRICAL PARALYSIS

Figure 1

(a) 3D CT of the shoulder dislocation; (b) CT of the shoulder subluxation.

Table 3 Historical review of different techniques forshoulder tendon transfers

Pectoralis major transfer Hildebrandt (1906)

Trapezius muscle Hoffa (1902)transfer Bateman (with acromium and

spina scapula bone bloc)

Short head of biceps Ober (1944)

Latissimus dorsi ’Episcopo (1934), Hoffer (1978)

Multiple muscle Harman (1950)transfers

ba

Page 250: Brachial Plexus Injuries

Since the main problem of the shoulder inobstetrical birth palsy is muscle imbalance, agreat number of techniques and modificationshave been described (Table 3).

Shoulder abnormalities

Abnormalities in the shoulder after obstetricalbirth palsy include: poorly formed and hypoplas-tic humeral head; a short, abnormally formedclavicle and a hypoplastic elevated scapula witha shallow glenoid fossa; an inferiorly directedcoracoid process; an abnormally taperedacromium, and subluxated shoulders (Pollockand Reid 1989, Waters et al 1998, Ogawa et al1999; Fig. 2). As in Figure 2 sometimes forcefulexternal rotation results with air arthrogram.

Contractures and shoulder dislocation

Internal rotation and adduction contracturesdevelop at the glenohumeral joint because ofmuscle imbalance. Shoulder paralysis (Gilbertgrading system) and evaluation of external

rotation are important for the treatment andfollow-up (Tables 4 and 5). As time passes,progressive deformity occurs in the gleno-humeral joint and sometimes dislocation of thejoint can be seen (Egloff et al 1995, Waters et al1998, Hentz 1999, Waters and Peljovich 1999). In1905, Whitman described three origins of shoul-der dislocations in infants: true congenital dislo-cation, traumatic dislocation during delivery, anddislocation secondary to the obstetrical birthpalsy. Fairbank in 1913, focused on thefrequency of shoulder subluxation in obstetricalpalsy and described the pathophysiology (Egloffet al 1995, Pearl and Edgerton 1998) (Fig. 1).

Initially, paralysed subscapularis recovers morequickly than the external rotators and abductors.The anterior capsule is short because of themalposition of the humeral head. Posteriorly, theinfraspinatus, teres minor and deltoid musclescannot provide sufficient support. The musclesthat are not paralysed (teres major, latissimusdorsi and subscapularis) pull the head down andbackward (Egloff et al 1995).

The age for operation for contractures isgenerally 12–24 months (Bennett and Allan1999). Adduction and internal rotation contrac-tures are common, and the Carliotz operation(release of subscapularis origin from scapulaanterior) can be performed once the child is over8 months of age (Gilbert 1997). In late cases(over 2 years), of age subscapularis lengthening

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN CHILDREN 241

Table 4 Shoulder paralysis (Gilbert grading system)

Stage 0 Complete flail shoulderStage 1 Abduction 45°, no external rotationStage 2 Abduction <90°, no external rotationStage 3 Abduction =90°, weak external rotationStage 4 Abduction <120°, incomplete external

rotationStage 5 Abduction >120°, active external rotation

Table 5 Evaluation of external rotation

Negative: Absent external rotation –45°Insufficient: Weak active external rotation –30°,–20°Fair: External rotation until neutral 0°Good: External rotation beyond neutral +20°,+30°Excellent: Complete external rotation +45°Figure 2

Glenohumeral deformity black arrow: lengthened coracoid;white arrow: sometimes forceful external rotation resultswith air arthrogram.

Page 251: Brachial Plexus Injuries

is preferable without opening the anterior shoul-der capsule (Bennett and Allan 1999). Shoulderdislocations are treated by open reduction andmuscle transfers and humeral osteotomy(Dunkerton 1989, Yücetürk 1999). Isolated abduc-tion contracture is rare, and release of the deltoidmuscle and sometimes lengthening of thesupraspinatus is the treatment of choice (Bennettand Allan 1999). Abduction and external rotationcontractures are another form of contracture;transfer of infraspinatus tendon to teres minortendon and release or recession of infraspinatusand supraspinatus tendons with or withoutrelease of deltoid muscle is the treatment ofchoice (Bennett and Allan 1999).

Glenohumeral deformity

Progressive glenohumeral deformity occurs inobstetrical birth palsy due to muscle imbalanceas the patient’s age increases. According toChuang et al, three main causes of shoulderdeformity due to obstetrical birth palsy are:cross-innervation of the muscles (caused bymisdirection of regenerated axons); muscularimbalance (caused by paresis or earlier recov-ery); and growth (Chuang et al 1998a, 1998b).

As time passes, incongruity of the gleno-humeral joint, deformity of the humeral headand hypoplasia of the glenoid is seen. In the X-ray, ossification of the humeral head is limitedcompared to the unaffected side. Hernandez andDias found CT useful in evaluation of the shoul-der for proper placement of the humeral head inplaster or orthotic devices (Hernandez and Dias1988) Waters et al evaluated 42 brachial plexusbirth palsy patients’ shoulders, using computer-ized tomography or magnetic resonanceimaging. They found that the degree of retrover-sion of the glenoid on the affected side was–25.7° compared with –5.5° on the unaffected

side. Sixty-two per cent of the 42 shoulders hadevidence of posterior subluxation of the head(Waters et al 1988). Gudinchet et al studied fiveshoulders, using MRI, and blunt anterior andposterior labrum was seen (Gudinchet et al1995).

Intraoperative arthrograms were performed byPearl and Edgerton and showed that 72 per centof the patients had a deformity of the posterioraspect of the glenoid. Of 18 patients, five hadflattening of the posterior aspect of the glenoid,seven had biconcave glenoid with the humeralhead articulating with the posterior of the twoconcavities, and six had a so-called pseudo-glenoid. This deformity, which is the result ofmuscle balance and internal rotation contracture,is severely advanced by the time that the child is2 years old (Pearl and Edgerton 1998).

Surgical procedures

The timing of surgery is important and de-pends upon the pathology and treatmentchoice (Table 6).

Tendon transfers

In obstetrical birth palsy, surgical procedures areperformed to improve shoulder function(increase of shoulder abduction and externalrotation). Various transfers are described in theliterature (Table 1). Most of the authors selectedone technique or combined or modified them.The L’Episcopo techique was, for example,modified by Covey et al, who transferred thelatissimus dorsi and re-routed it from the axillaryapproach (Covey et al 1992). In the literature theresults generally appear to be satisfactory

242 OBSTETRICAL PARALYSIS

Table 6 Timing of the surgery

Algorithm of secondary procedures in obstetrical palsy

Contractures Shoulder dislocation Tendon transfers Bony procedures8–24 months immediately after diagnosis after 24 months after 5 years of age

Page 252: Brachial Plexus Injuries

whichever technique is used. Preoperative physi-cal examination and the power of the musclesare very important, and also techniques must notbe standardized; and combined procedures mustbe used according to the preoperative planningand surgical findings.

Chuang et al described perioperative studiesdemonstrating the existence of muscle recoveryby cross innervation, and a new strategy ofmuscle transposition to minimize the influence ofcross innervation by releasing the antagonisticpectoralis major and teres major muscles andaugmenting the paretic muscles. Augmentationwas done by transferring teres major to theinfraspinatus muscle and reinserting both endsof the clavicular part of the pectoralis majormuscle laterally (Chuang et al 1998b).

Price and Grossman transferred the teres majorand latissimus dorsi as one conjoined tendon tothe posterior aspect of the greater tuberosity(Price and Grossman 1995). Gilbert believes thatthe excursion of the teres major and latissimusdorsi is different, and the latissimus dorsi mustbe transfered alone (Gilbert 1997). After trapeziustransfer with the acromium in traumatic patients,Ruhmann et al were satisfied with the improve-ment in stability and function (Ruhmann et al1998). In obstetrical cases, trapezius transfer isgenerally combined with other muscle transfers,especially in patient with weak deltoid andsupraspinatus muscle functions.

L’Episcopo muscle tendon transfer alsoimproves the functional outcome in adulttraumatic brachial plexus lesions (Beauchamp etal 1998).

After Sever–L’Episcopo transfers transient andpermanant axillary nerve palsies have beenreported (Strecker et al 1990).

Humeral osteotomy

Restoration of lateral rotation at the shoulder canimprove its function (Goddard and Fixsen 1984).Vulpius and Stoffel performed external humeralosteotomy to correct internal rotation contrac-ture in 1913. Zancolli used the same osteotomy,and noted that osteotomy improves abduction aswell (Egloff et al 1995).

Twenty-two patients between the ages of 4and 17 who were unable to perform self-care

activities were treated by Kirkos et al withrotational osteotomy of the proximal part of thehumerus. All patients had decreased strength ofthe lateral rotator and abductor muscles andnormal strength of the subscapularis andpectoralis major muscles. Osteotomy wasperformed between the insertions of thesubscapularis and pectoralis major muscles. Theaverage increase in active abduction was 27°,and the average increase in the arc of rotationwas 25° (Kirkos and Papadapoulos 1998). Waters’studies indicate that both tendon transfer andhumeral osteotomy can uniformly improvefunction in children with chronic brachialplexopathy according to their age and gleno-humeral deformity (Waters and Peljovich 1999;Fig. 3).

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN CHILDREN 243

Figure 3

Humeral osteotomy.

Page 253: Brachial Plexus Injuries

Scapula

Seventeen muscles are attached or take originfrom the scapula. The most important musclesare the serratus anterior, trapezius, rhomboidsand levator scapula. The muscles help thescapula to remain in position during the fullrange of motion of the shoulder.

The effects of scapular instability are functional(dynamic) and cosmetic. Functional effectsinclude diminished arm abduction and externalrotation. Cosmetic effects include shoulderasymmetry and impaired scapular growth.

In 1984, Hertzmark et al reattached therhomboid muscle to the scapula. In 1989,Whitman et al used fasciat lata graft between thescapula and spinal processes, and in 1980Ketencian et al used fascia lata graft between thescapula and the ribs for stabilization of thescapula. Tensor fascia lata graft was also used forthe transfer of the pectoralis major to the inferiormedial part of the scapula for scapular winging.

Surgical results

Between January 1994 and September 2000, theauthor operated on 116 patients for obstetricalbirth palsy shoulder sequelae. Sixty-seven (57.8per cent) were male and 49 (42.2 per cent) werefemale. In 62 patients (53.4 per cent) the rightside was affected and in 54 patients (46.6 percent) the left side was affected. The age atsurgery was between 5 months and 22 years(average 7.2 years). Five families had twochildren with the same problem. Seventy-threepatients (63 per cent) had combined elbow,forearm and hand problems. Generally all theproblems are operated in one stage if the patientwas over the age of 5 years.

Shoulder posterior dislocation

Five cases were operated after CT or MRI diagno-sis; three were under the age of 1 year. A poste-rior approach was used for the latissimus dorsiand teres major transfer to the rotator cuff, andan anterior approach was used for subscapularislengthening and pectoralis major transfer to the

rotator cuff. In two cases, shoulder stiffness wasseen. Shoulder stiffness was not seen in twocases which were operated after 1.5 years of age.Shoulder subluxation was seen in 18 cases andthe same tendon transfers were performed.

Contractures

After 8 months of age, the Carliotz operation wasperformed (six cases). All had combined tendontransfer (latissimus dorsi and teres major trans-fer to rotator cuff, subscapularis lengthening andpectoralis major transfer to rotator cuff) after 2years of age. The author did not achieve goodresults with the Carliotz operation.

For the abduction contracture, two patientshad subscapularis lengthening. Eleven patientshad infraspinatus and teres minor lengtheningbecause of an internal rotation deficit and exter-nal rotation contracture.

Ninety-eight per cent of patients were operatedfor internal rotation and adduction contracture.Only two had humeral osteotomy alone. Theposterior incision was made first, from 1 cminferior to the posterior acromium to the axillaryfold. The latissimus dorsi and teres major weretransferred to the rotator cuff and sutured withnon-absorbable sutures with the patient’s armresting on the trunk. If the patient had an inter-nal rotation deficit, infraspinatus lengtheningcould be done at the same time. If the teresmajor was stiff and not mobile enough to reachthe rotator cuff, tenotomy only was performed.According to Gilbert, the excursion of the twomuscles is different and they must therefore notbe transferred together.

In 25 cases, tenotomy of the teres major asperformed. In this approach, it is important toprotect the axillary nerve, posterior circumflexhumeral artery, radial nerve just under the teresmajor and axillary nerve entrance to the deltoidmuscle when suturing the tendon to the rotator.

A deltopectoral incision was made for theanterior approach. The cephalic vein wasprotected. The pectoralis major was cut at theinsertion site and the pectoralis minor tendonwas protected. The subscapularis tendon wasisolated and lengthened. The anterior circumflexhumeral artery, axillary nerve and anteriorcapsule were protected. External rotation was

244 OBSTETRICAL PARALYSIS

Page 254: Brachial Plexus Injuries

examined. In Klumpke patients, if the pectoralisminor was contracted then lengthening wasperformed. When 20° of external rotation wasmaintained, the pectoralis major was transferredto the rotator cuff and sutured with non-absorbable sutures. After release, if there wasstill an external rotation deficit and if the patientwas over 5 years of age, humeral osteotomy wasperformed just over the pectoralis major inser-tion to maintain 20° of external rotation and plateand screw fixation was used.

When the deltoid and suprascapularis muscleswere weak, trapezius transfer with acromial boneblock was performed in patients over 5 years ofage. If the patient was under 5 years of age, only

tendinous transfer was done. The levator scapulawas not used in this series.

Incisions were closed with two layers ofabsorbable sutures. The skin was closed withabsorbable subcuticular sutures. In the last 3years, 3M Soft-Cast was used for immobilizationin 90° abduction and neural rotation. The axillaryregion was augmented with six layers of Soft-Cast splint. Immobilization was for 6 weeks, andwounds were not opened during this time. Fivedays of antibiotic cover were given. After 6weeks, the cast was removed with scissors; acast-cutter was not necessary. Fifteen days afterremoval of the cast, physiotherapy began ifpossible.

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN CHILDREN 245

a b

c

Figure 4

(a) preoperative(b) postoperative abduction(c) postoperative external rotation.

Page 255: Brachial Plexus Injuries

Following the operation, if the shoulder eleva-tion increases to more than 90° and the tricepsmuscle power is under 3 (antigravity weakness),the patient has a disability problem. The sameproblem can be seen after transferring triceps tobiceps for elbow flexion, combined with shoul-der tendon transfer. Almost all the patients’families were satisfied with the post-operativeresult (Fig. 4).

Glenohumeral deformity

If the patient had glenohumeral deformity, theapproach was the same as above. The authordoes not immediately perform humeralosteotomy; for this, the patient must haveenough internal rotation before surgery.

Shoulder arthrodesis

Only one case (15-year-old patient), which hadweak shoulder muscles and bony atrophy, hadshoulder arthrodesis and plate screws fixation.

Coracoid osteotomy

In some cases a lengthened coracoid was seen,especially in direct X-ray. During this series noneof them blocked the external rotation, andnothing was done to the coracoid process.

Scapular winging

In one case the pectoralis major, lengthened withtensor fascia lata, was transferred to the medialborder of the scapula and the result was good.

References

Beauchamp M, Beaton DE, Barnhill TA et al (1998)Functional outcome after the L’Espiscopo procedure, JShoulder Elbow Surg 7(2):90–6.

Bennett JB, Allan C.H (1999) Tendon transfers aboutthe shoulder and elbow in obstetrical brachial plexuspalsy, J Bone Joint Surg 81A:1612–27.

Birch R (1995) Lesions of the upper brachial plexus:C5/6 and C5/6/7 injury. Surgery of the shoulder. In:Wastamaki M, Jalovaara P, eds. Elsevier Proceedingsof the 6th International Congress on Surgery of theShoulder (ICSS), 27 June–1 July, Helsinki. Elsevier:Amsterdam: 373–8.

Chen L, Gu Yu-dong, Xu J (1999) Operative treatmentfor the medial rotation contracture of the shouldercaused by obstetric brachial plexus palsy. 1999International Symposium on Upper Limb Problems inChildhood, PR China, Oct 24–28, 27.

Chuang DC, Ma HS, Wei FC (1998a) A new evaluationsystem to predict the sequelae of late obstetric brachialplexus palsy, Plast Reconstr Surg 101(3):673–85.

Chuang DC, Ma HS, Wei FC (1998b) A new strategy ofmuscle transposition for treatment of shoulder defor-mity caused by obstetric brachial plexus palsy, PlastReconstr Surg 101(3):686–94.

Clarke HM, Curtis CG (1995) An approach to obstetri-cal brachial plexus injuries, Hand Clin 11(4):563–81.

Covey DC, Riordan DC, Milstead ME, Albright JA (1992)Modification of the L’Episcopo procedure for brachialplexus birth palsies, J Bone Joint Surg 74B(6):897–901.

Dunkerton MC (1989) Posterior dislocation of the shoul-der associated with obstetric brachial plexus palsy, JBone Joint Surg 71B(5):764–6.

Egloff DV, Raffoul W, Bonnard C, Stalder J (1995)Palliative surgical procedures to restore shoulderfunction in obstetrical brachial palsy, Hand Clin11(4):597–606.

Gilbert A (1995) Long-term evaluation of brachial plexussurgery in obstetrical palsy, Hand Clin 11(4):583–5.

Gilbert A, Brockman R, Carlioz H (1991) Surgical treat-ment of brachial plexus birth palsy, Clin Orthop264:39–47.

Goddard NJ, Fixsen JA (1984) Rotation osteotomy ofthe humerus for birth injuries of the brachial plexus, JBone Joint Surg 66B(2):257–9.

Gudinchet F, Maeder P, Oberson JC, Schnyder P (1995)Magnetic resonance imaging of the shoulder inchildren with brachial plexus birth palsy, Pediatr Radiol25:S125–8.

246 OBSTETRICAL PARALYSIS

Page 256: Brachial Plexus Injuries

Hentz VR (1999) Brachial plexus palsy – internalrotation contracture of the shoulder. Restoration ofexternal shoulder rotation in Erb’s brachial plexuspalsy. 1999 International Symposium on Upper LimbProblems in Childhood, PR China, Oct 24–28, 20–25.

Hernandez RJ, Dias L (1988) CT evaluation of the shoulderin children with Erb’s palsy, Pediatr Radiol 18(4):333–6.

Kendall FP, McCreary EK, Provance PG (1993) MuscleTesting and Function. Upper Extremity and ShoulderGirdle Strength Tests. Williams & Wilkins: Baltimore:235–98.

Kirkos JM, Papadopoulos IA (1998) Late treatment ofbrachial plexus palsy secondary to birth injuries:rotational osteotomy of the proximal part of thehumerus, J Bone Joint Surg 80A:1477–83.

Ogawa K, Yoshida A, Inokuchi W (1999) Deltoidcontracture: a radiographic survey of bone and jointabnormalities, J Shoulder Elbow Surg 8(1):22–5.

Pearl ML, Edgerton BW (1998) Glenoid deformitysecondary to brachial plexus birth palsy, J Bone JointSurg 80A(5):659–67. Published erratum appears in JBone Joint Surg 80A(10):1555–9.

Pollock AN, Reed MH (1989) Shoulder deformities fromobstetrical brachial plexus paralysis, Skeletal Radiol18(4):295–7.

Price AE, Grossman AI (1995) A management approachfor secondary shoulder and forearm deformitiesfollowing obstetrical brachial plexus injury, Hand Clin11(4):607–17.

Ruhmann O, Writh CJ, Gosse F, Schmolke S (1998)Trapezius transfer after brachial plexus palsy.Indications, difficulties and complications, J Bone JointSurg 80(1):109–13.

Strecker WB, McAllister JW, Manske PR et al (1990)Sever–L’Episcopo transfers in obstetrical palsy: a retro-spective review of twenty cases, J Pediatr Orthop10(4):442–4.

Waters PM, Smith GR, Jaramillo D (1998)Glenohumeral deformity secondary to brachial plexusbirth palsy, J Bone Joint Surg 80A(5):668–77.

Waters PM (1999) Comparison of the natural history,the outcome of microsurgical repair, and the outcomeof operative reconstruction in brachial plexus birthpalsy, J Bone Joint Surg 81A(5):649–59.

Waters PM, Peljovich AE (1999) Shoulder reconstruc-tion in patients with chronic brachial plexus birth palsy.A case–control study, Clin Orthop 364:144–52.

Waters PM (1997) Obstetric brachial plexus injuries:evaluation and management, J Am Acad Orthop Surg5(4):205–14.

Yücetürk A (1996) EMG problems in the preoperativeevaluation of obstetrical brachial plexus, Turkish JHand Surg Microsurg 4(5):21–4.

Yücetürk A (1999) Tendon transfers to the shoulder,elbow and hand in OBP. 1999 International Symposiumon Upper Limb Problems in Childhood, Oct, PR China,28–29.

PALLIATIVE SURGERY: TENDON TRANSFERS TO THE SHOULDER IN CHILDREN 247

Page 257: Brachial Plexus Injuries
Page 258: Brachial Plexus Injuries

Introduction

Medial rotation contracture and posterior dislo-cation of the shoulder is the most common andmost significant secondary deformity in obstetri-cal brachial plexus palsy (Birch et al 1998, Birch2000). Surgery to correct the deformity provednecessary in more than one third of the 1200children seen in our unit from 1982. Diagnosis isoften delayed, and the untreated deformity hassevere consequences for function within theupper limb as a whole. By late adolescence orearly adult life, movements of the shoulder girdleare greatly restricted. The upper limb lies in fixedmedial rotation; there is pain from the disorga-nized gleno-humeral joint; and there is flexion

pronation posture of the elbow, often compli-cated by subluxation of the head of the radius(Fig. 1).

Most children with obstetrical brachial plexuspalsy (OBPP) progress to good or at least usefulneurological recovery, and it is the duty oforthopaedic surgeons to encourage this potentialfunction by earlier detection and more appropri-ate treatment of this deformity (Fig. 2).

Historical background

The causation, course, skeletal consequences andmethods of treatment were set out by workers

25Medial rotation contracture andposterior dislocation of the shoulderRolfe Birch

Figure 1

Posterior dislocation in a 21-year-old man. Movements ofthe gleno-humeral joint are restricted and painful. There isnow a fixed flexion pronation posture at the elbow withsecondary dislocation of the head of radius.

Figure 2

The typical posture of posterior subluxation in a 2-year-oldboy. Note the flexion pronation posture of the forearm.

Page 259: Brachial Plexus Injuries

during the last century. It seems that their workhas been neglected or misunderstood, in spite ofthe recent resurgence of interest in the neuro-logical lesion itself. Whitman (1905) distinguishedbetween congenital (rare) and acquired(common) subluxation, stating that acquiredsubluxation was caused by fibrosis and contrac-tures during the period of paralysis: it was aconsequence of the neurological lesion. Fairbank(1913) wrote: ‘the muscle which is most affectedand offers the strongest bar to outward rotationis the subscapularis’. He described the results ofsurgery in 18 cases where a technique was usedin which the subscapularis tendon and capsule,and the coraco-humeral ligament, were divided.In three cases the coracoid was sectioned. Theradiological features of the deformity were fullydescribed by Sever in his analysis of a series of1100 children with OBPP (Sever 1925). He recog-nized delayed ossification of the head of thehumerus, progressive deformation of the glenoidand, in later stages, overgrowth of the acromionand ‘marked elongation of the coracoid process,due probably to the pull of the contracted coraco-brachialis muscle’. Sever was unable to demon-strate any case of epiphysis separation or ofdislocation caused during delivery, but Putti,whose work was reported by Scaglietti (1938),thought that the deformity was caused by directinjury to the proximal humerus: ‘the mostconstant and characteristic change is the defor-mation of the angle of declination’ (i.e. of retro-version of the head of humerus upon the shaft).Controversy continues. Zancolli and Zancolli(1993) thought that damage to the growth plate(epiphysiolysis) was the major factor in causationof the deformity, but in the same volume Gilbert(1993) wrote that: ‘posterior subluxation defor-mity of the humeral head permanently worsensthe prognosis. These anomalies that have longbeen considered the result of obstetrical palsyare, in fact, in consequence of untreated contrac-tures’. Goddard (1993) studied over 200 casestreated by Gilbert, and uncovered only twoinstances of injury to the growth plate. Tubiana(1993) commented on the reasons for thesedifferences in interpretation, ‘which actuallycorrespond to two different patient groups’. Ourexperience suggests that both views are correctand represent important contributions.

We described our experience with 86 opera-tions of subscapularis recession and 59 cases of

subluxation or dislocation treated by the anteriorapproach in 1996 and in 1998 (Birch and Chen1998, Birch et al 1998), noting a high incidenceof recurrence of deformity in the former (29cases) and substantial loss of medial rotation inthe latter. These findings led to a significantchange in our approach owing to better under-standing of the deformity of the glenoid inadvanced cases and the high incidence of retro-version of the head upon the shaft.

The present paper is based on analysis of 166children operated for the deformity between1992 and 1996. All these had at least usefulneurological recovery, either spontaneously orafter repair of the brachial plexus. The functionalgrades were at least Raimondi 4 for the hand andGilbert and Raimondi 4 for the elbow, and themean range of elevation at the shoulder was150° at final review. In other words, all of thesechildren had substantial recovery through allelements of the brachial plexus. We excludedfrom discussion 44 other children who weretreated for the shoulder deformity, who hadmore severe neurological injuries and in whomthe problem was one of paralysis. It is essentialto distinguish between these two groups; thedifficulties and methods of solution are quitedifferent. In the ‘paralytic’ group with severeneurological injury, the priority is to improveinnervation of the upper limb as far as possible.Muscle transfers are only of palliative value, buttransfer of the latissimus dorsi to the lateralrotators may be necessary to improve lateralrotation and so restore the muscular balance atthe shoulder. Such muscle transfers were notneeded in the 166 children described, all ofwhom regained functional lateral rotation. Insome the difficulty was restoration of adequatemedial rotation. Of the 166 children, 28 had hadprevious subscapularis recession and 21 opera-tions of anterior release, and the deformity hadrecurred in these 49 children. All of these 166children fell into Groups I, II or III of the Narakasclassification: none was delivered by breech.

Causation

In 40 children the deformity was detected at birthor shortly afterwards, and in 30 more it wasrecognized within the first year of life. One fathergave a clear description of the mechanism of

250 OBSTETRICAL PARALYSIS

Page 260: Brachial Plexus Injuries

injury. During a difficult vertex delivery theafflicted arm was pulled into abduction and thenacross the chest, into forced flexion with medialrotation. In 13 cases the subscapularis musclewas densely fibrosed, which may represent apost ischaemic or compartment syndrome lesionanalogous to that described in the adult case byLandi et al (1992). However, in 57 children thedeformity developed or progressed whilst underobservation, and in 20 the deformity occurredafter repair of the upper trunk and progressed inspite of continuing observation and assiduousexercises. In 11 children recovery was so goodthat they were discharged from the clinic withnormal or near-normal function, only to returnlater with established deformity! Birch and Chen(1996), in their analysis of the earlier 120 consec-utive cases, made the alarming observation thatin no less than 12 children, the shoulder await-ing treatment for uncomplicated medial rotationcontracture had progressed to dislocation whilstawaiting admission. This suggests that the

primary cause is, in most cases, the neurologicallesion, which invariably afflicts C5 and C6irrespective of whether the rest of the plexus isdamaged or not. There is paralysis of the lateralrotator muscles, of infraspinatus and teresminor, innervated by the fifth cervical nerve. Themedial rotators (most especially the subscapu-laris muscle) innervated by the seventh andeighth cervical nerves are never paralysed or areonly weakened for a short time so that theiraction is unopposed. Muscular imbalance ispotent cause of deformity in the growing limb,and the deformity of the shoulder is a reflectionof this general principle. We have not encoun-tered a single instance of anterior dislocation ofthe gleno-humeral joint.

Classification of the deformity

The rate of progression of the deformity variesfrom child to child, and is not necessarily related

MEDIAL ROTATION CONTRACTURE AND POSTERIOR DISLOCATION OF THE SHOULDER 251

Table 1 A clinical classification of shoulder deformity (based on observations by Dr Liang Chen, in Birch et al 1998)

Type Relation of head of Clinical evidence Radiological evidence Supplementary humerus to glenoid investigations

Medial rotationcontracture

Simplesubluxation

Simpledislocation

Complexsubluxation

Complexdislocation

Loss of passive lateralrotation of 30° or more

Lateral rotation toneutral. Head palpableposteriorly

Fixed medial rotationcontracture at about30°. Head evidentlylying behind glenoid

Lateral rotation toneutral or less.Overgrowth of coracoidand acromion palpable.

Fixed medial rotationcontracture of 30° ormore, obvioussecondary bonechanges

Normal: coracoid maybe elongated

Incongruent. No otherskeletal abnormality

Head of humerusbehind glenoid. Noother skeletal deformity

Extent of coracoid andacromion abnormalityseen: ‘double facet’ ofglenoid.

Head of humerusbehind glenoid:overgrowth of coracoidand acromion;abnormality of glenoid

Congruent

Head of humerus infalse glenoid

Head of humerusposterior to glenoid

Head of humerus infalse glenoid. Secondarybone deformity

Head of humerusbehind glenoid.Secondary bonedeformity

Ultrasound – congruent.MR scan may showretroversion of headupon shaft of humerus

Ultrasound, CT and MRscans confirmincongruency:retroversion, and‘double facet’ glenoidmay be seen

Ultrasound, CT, MRscans confirm.Retroversion may beseen

Confirm incongruencyand skeletal abnormalitybut may mislead aboutglenoid shape

Confirms dislocation andextent of skeletalabnormality

In all cases, a flexion pronation posture of elbow and forearm is seen. In advanced cases this deformity becomes fixed, and may beassociated with dislocation of the head of radiusThe extent of retroversion of head upon shaft of humerus cannot be measured accurately by any ancillary investigation, and it is bestdetermined at operation of open reduction.

Page 261: Brachial Plexus Injuries

to age. Advanced secondary bone changes havebeen seen in children aged 3 years or less, whilstdislocation in the presence of only minor defor-mity has been seen in children aged 11 or 12years. The deformity is progressive, and there isa spectrum from medial rotation contracture tocomplex dislocation of the shoulder (Table 1).

Diagnosis

The diagnosis is made by physical examination.For babies, the infant is placed supine and theexaminer holds both upper limbs with theelbows flexed to 90°. The arms are heldadducted against the chest and the upper limbsare gently rotated into lateral rotation. Anydiminution in the range of passive lateralrotation in the afflicted upper limb is significant.Serious errors in diagnosis are caused by incor-rect examination, and the practice of examiningone limb in isolation or of estimating passiverange for rotation within the arm in abduction iscondemned. The examination is done gently,and if the infant protests then the examiner mustsuspect incongruity at the gleno-humeral joint. Inolder children, the posture of the upper limb ischaracteristic, lying in medial rotation withflexion and pronation at the elbow. The contourof the shoulder is abnormal, with head seenprominently behind the glenoid. Palpationreveals abnormalities of the coracoid andacromion. In nearly every case a clear impres-sion of the extent of bone deformity can beformed from the physical examination supple-mented by plain anteroposterior and axial radio-graphs. Arthrograms, CT and MR scans havebeen used in diagnosis (Waters et al 1998, Pearlet al 1998). Ultrasound examination is a suitableaid to diagnosis before the first 12 months of lifein those cases where the clinician is suspiciousabout the congruity of the shoulder.

Classification of the secondary bone deformi-ties is set out in Table 2. The changes in coracoidand acromion are best detected by clinicalexamination; the extent of retroversion at opera-tion; and the changes of the head and of theglenoid by plain anteroposterior and axillaryradiographs supplemented by inspection atoperation. MR scans provide valuable informa-tion in the advanced case, but the findings from

these, and from CT scans, must be interpretedwith care; they must not be used in place ofcareful clinical examination (Figs. 3, 4).

Recording shoulder function

Careful recording of clinical data is essential.Without such a record, comparison betweendifferent series is impossible and much clinicalwork is rendered valueless. We have followedthe proposals of Gilbert and of Raimondi, andrecommend that other clinicians use thesesystems, which are summarized below.

Three systems are used in our unit. Nearly allchildren aged 1 year or more willingly demon-strate these exercises using coloured crayons orsmall toys as encouragement. Observation of theinfant gives useful information about the rangeof active movements, the passive range can bemeasured. Records are made at each clinic atten-dance, and the three systems, taken together,provide useful information about shoulderfunction.

The first method (Table 3) is derived from thesystem proposed by Alain Gilbert, and has beenmodified to record the presence of functional

252 OBSTETRICAL PARALYSIS

Table 2 Summary of bone deformities

Coracoid Normal 30Moderate 98Severe 38

Acromion Normal 121Anterior spur 36Overgrowth of whole 9

Glenoid Normal 38Double facet 99Planar 22Severe posterior defect 16

Head of Humerus Normal 125Conoid or oval 25Flattened 9Trench or bifid defect 7

Retroversion Less than 30° 8330–50° 2050–70° 36Over 70° 27

In the advanced case, the glenoid exhibits both planar andposterior defect.Some of the deformities of head and glenoid were caused byearlier inappropriate treatment.

Page 262: Brachial Plexus Injuries

medial rotation. The child with fixed medialrotation contracture such that the passive rangeof lateral rotation is restricted to the neutralposition or less is, by convention, given a gradeof Stage 1 and no more. Mallet’s system (Table4) records five shoulder functions, according amaximum of three points to each. A child with15 points has a good shoulder but by no meansa normal one. Our system of recording the activeand passive range of movements of the shoulderis set out in Table 5, and this includes measure-ment of the range of active prono-supination,which usually improves after successful reloca-tion of a dislocated joint. This system recognizes

MEDIAL ROTATION CONTRACTURE AND POSTERIOR DISLOCATION OF THE SHOULDER 253

Figure 3

Complex posterior dislocation in a 7-year-old boy. Althoughthe true glenoid appeared to be reasonably well developedat operation, posterior bone block was necessary to securethe stabilization. This had to be done as a second opera-tion. Final scores in this case: Gilbert 5+, Mallett 15.

Figure 4

AP radiograph of complex subluxation in a 7-year-old girl.There is overgrowth of the anterior acromion. Note theelongation of the coracoid.

Table 3 A method of staging shoulder function in OBPP(drawn from Gilbert 1993)

Stage 0 Flail shoulderStage I Abduction or flexion to 45°

No active lateral rotationStage II Abduction < 90°, lateral rotation to neutralStage III Abduction = 90°, weak lateral rotationStage IV Abduction < 120° – incomplete lateral rotationStage V Abduction > 120° – active lateral rotationStage VI Normal

The suffix + is added to indicate sufficient medial rotationpermitting the hand to come against the opposite shoulder.Our convention restricts children with no lateral rotation beyondneutral to Stage I, usually I+, because adequate medial rotationis maintained.

Page 263: Brachial Plexus Injuries

that loss of active range may arise from paraly-sis or weakness of muscles, but also from fixedcontracture. Two specific measures meritdescription.

The inferior gleno-humeral angle

This is the angle between the axes of thehumerus and the lateral border of the scapulawith the arm in adduction. In the normal shoul-der, this is at least 150°. In some children,contracture of the inferior capsule and of thelatissimus dorsi and teres major is so tight thatthe active and the passive angle is diminished to30–40°. Weakness of the abductor muscles of theshoulder is shown by substantial discrepancybetween the passive and active ranges.

The posterior gleno-humeralcontracture

This is measured as follows. The affected handis placed onto the opposite shoulder, with the

254 OBSTETRICAL PARALYSIS

Table 4 The Mallet system of measuring function at theshoulder

Value of active = 1 2 3shoulder function

Global abduction

Global external rotation

Hand to neck

Hand to spine

Hand to mouth

Range of Movement

DATE Forward flexions Lat rotation Inferior GH angle Post GH angle Abduction Medical rotation Rotation forearm

Active Passive Active Passive Active Passive Active Passive Active Passive Active Passive Active Passive

Table 5

Page 2 (OBPP shoulder)

Page 264: Brachial Plexus Injuries

axis of the humerus parallel to the ground. Theangle between the axes of the humerus and ofthe blade of the scapula is measured. In a normalshoulder this should be 70°, whilst in a severecontracture it may be reduced to 0°. Some of thiscontracture arises from capsular tightness, and itcan be overcome by firm depression of thescapula onto the chest wall whilst holding thearm in the position described. However, most ofthe contracture is caused by bone deformity(retroversion of the neck of the humerus), and itis commonly seen after successful relocation ofdislocation. Treatment, which is straightforwardenough, is outlined below (Fig. 5).

Treatment

Exercises

Obviously, the best treatment is prevention. It ispossible to overcome an uncomplicated medialrotation contracture by assiduous but gentlestretching of both limbs into lateral rotation withthe arms adducted against the side. It is for theclinician to teach the parents this exercise and toinsist that physiotherapists perform the stretchcorrectly. Close monitoring is essential, andthese children are reviewed at intervals notexceeding 6 weeks. The exercises are performedfour or five times before every feed in the caseof infants, and the arms are held in the positionof lateral rotation for 4–5 seconds. If the exerciseprovokes pain, then the clinician should assumethat the shoulder is incongruent. Forceful stretch-ing, or persisting with these exercises when theshoulder is plainly incongruent, is damaging.

Surgery

Subscapularis recession was described byCarlioz and Brahimi in 1986. Gilbert (1993)emphasized that the operation should only beperformed if the shoulder is congruent. Theincidence of recurrence has already beenmentioned. It is clear that those surgeonsinclined to follow this operation must perform,meticulously, the indications and techniques ofthe originators: above all the operation must bedone only if the shoulder is congruent and if thelateral rotator muscles have recovered suffi-ciently to restore balance. As in all work inobstetric brachial plexus palsy prolonged andcareful clinical follow up is essential.

Lateral rotation osteotomy of the shaft ofhumerus has no place in the treatment of thedeformity. It does nothing to secure congruentreduction of the head of the humerus into thetrue glenoid, it increases retroversion of the headupon the shaft of the humerus. Results havebeen singularly unimpressive in those caseswhen it was used as palliation for the irreducibleshoulder.

Correction of deformity by the anteriorapproach is intended to secure congruent reloca-tion of the head of the humerus into the true

MEDIAL ROTATION CONTRACTURE AND POSTERIOR DISLOCATION OF THE SHOULDER 255

Figure 5

Loss of active medial rotation in a child with a posteriorgleno-humeral angle of no more than 10°. This wascorrected by de-rotation osteotomy of the shaft ofhumerus.

Page 265: Brachial Plexus Injuries

glenoid. The obstacles to this reduction includeovergrowth of the coracoid process, contractureof the coraco-humeral ligament, and contractureof the subscapularis muscles. Significant retro-version of the neck of the humerus must also becorrected. The technique of operation is fullydescribed elsewhere (Birch 2001a, 2001b).

The record of the abnormalities displayedshould include:

1. The condition of the deltoid and pectoralismuscles;

2. The length, breadth and inclination of thecoracoid, and whether it abuts against thehead of the humerus in attempted lateralrotation;

3. The contribution of the coraco-humeralligament to the contracture;

4. The state of the subscapularis muscle;5. The condition of the labrum;6. The depth and breadth of the true glenoid;7. The location of the false glenoid and the

presence or absence of an intervening carti-laginous ridge between this and the truesocket;

8. Overgrowth of the acromion;9. Any abnormality of the anterior part of the

head of the humerus;10.The stability of reduction, between full lateral

rotation and medial rotation. Re-dislocation of

256 OBSTETRICAL PARALYSIS

Figure 6

AP radiographs of complex sublux-ation in an 8-year-old girl.

Figure 7

Axial radiograph of complex subluxation in an 8-year-oldgirl.

Page 266: Brachial Plexus Injuries

the head of the humerus into the false glenoidor behind it at neutral rotation or even in theinner range of lateral rotation is one indica-tion of retroversion.

We estimate retroversion of the head upon theshaft of humerus by grasping the head of thehumerus between the index finger and thumb ofone hand, to define the coronal plane of thehead, and grasping the epicondyles of thehumerus with the finger and thumb of the otherhand. The coronal plane of the head of humerusand the distal humerus are parallel in the normalinfant shoulder. Retroversion of the head inexcess of 30° is significant; retroversion inexcess of 50° demands de rotation osteotomy.This was done in 55 of these children; wherepossible it should be done at the same operation,but the arm is too small in children under the ageof 18 months. The shaft of the humerus isexposed between the pectoralis major anddeltoid tendons.

MEDIAL ROTATION CONTRACTURE AND POSTERIOR DISLOCATION OF THE SHOULDER 257

Figure 8

Function in 3-year-old boy, 18 months after relocation ofsimple dislocation. Gilbert 5+, Mallett 15.

Figure 9

Axial radiograph of the case shown in Fig. 4 8 years later,showing a degree of remodelling of the head, although itsenlargement suggests that there may have been a degreeof avascular necrosis.

Figure 10

The clinical outcome in this girl, now aged 15. Gilbert score5+, Mallett 15.

Page 267: Brachial Plexus Injuries

A posterior bone block is indicated if there isa severe defect of the postero-inferior wall of theglenoid. It has now been done in 12 cases. Theprecise indication for this technique has yet to beclarified, but it is probably necessary inlongstanding cases with marked abnormality ofthe posterior lip of the glenoid. In these cases,MR scanning has a place in defining glenoidmorphology. Our experience with osteotomy ofthe glenoid has been very unsatisfactory.

Results

The children were followed for a minimum of 3years, and the changes in the Gilbert and Malletfunctional scores are summarized in Table 6. Thechanges in the mean ranges of active

movements at the shoulder and of prono-supina-tion before and after operation are summarizedin Table 7. Some of the improvement recordedmust inevitably reflect continuing neurologicalrecovery. We point out that the low Gilbert scorerecorded before operation is a reflection of ourmodification of his system of measurement (seeTable 2).

The long-term outcome from this interventionwill not be known for a number of years. It islikely that later reconstructive operations willprove necessary in some of these children whenthey reach adult life, and it is hoped that, byimproving the anatomical relation of the gleno-humeral joint, subsequent arthroplasty or evenarthrodesis will be a practical proposition.

The improvement in function at the elbow andin the forearm is, at times, remarkable. In somechildren, improvement of active extension of thewrist was seen. The observations of parents andour functional assessments in these childrensuggest that successful relocation of the shoul-der brings about marked improvement not onlyat the shoulder but also in function of the limbas a whole.

Acknowledgements

Mr George Bonney kindly gave permission foruse of the photographs in this chapter, drawnfrom ‘Surgical Disorders of the PeripheralNerves’. Mrs Margaret Taggart was responsiblefor collating all clinical records and for prepara-tion of the manuscript.

258 OBSTETRICAL PARALYSIS

Table 6 Mean Gilbert and Mallet scores before and afteroperation

Gilbert Mallet

Deformity Cases Pre- Post- Pre- Post-

Medial rotation 12 1.85 5.1 10.2 14.0Simple subluxation 42 1.50 4.3 9.2 13.3and dislocationComplex subluxation 112 1.70 4.3 9.3 12.0and dislocation

The low Gilbert scores recorded before operation are a reflectionof our modification of this system (see Table 2).

Table 7 Means of ranges of active movements at shoulder before and after operation

Deformity Cases Forward flexion Abduction Lateral rotation Medial rotation Prono-supinationPre- Post- Pre- Post- Pre- Post- Pre- Post- Pre- Post-

Medial rotation 12 124 153 136 156 – 6 65 94 86 104 152Simple subluxation 42 105 127 111 128 –17 55 90 81 105 156and simple dislocationComplex subluxation 112 96 122 103 128 –17 42 90 76 87 142and simple dislocation

There is no doubt that some of the improvement in flexion and abduction is a reflection of continuing neurological recovery.

Page 268: Brachial Plexus Injuries

References

Birch R (2000) Obstetric brachial plexus palsy. In: Kay SJP,ed. The Growing Hand. Mosby: Philadelphia: 461–74.

Birch R (2001a) The shoulder in OBPP. In: Dupard J,Carlioz H, eds. Surgical Techniques in Orthopaedicsand Traumatology. Elsevier (in press).

Birch R (2001b) Shoulder dislocation in OBPP. In:Benson M, Fixen J, MacNicol M, Parsch K, eds.Children’s Orthopaedics and Fractures. ChurchillLivingstone: Edinburgh: in press.

Birch R, Bonney G, Wynn Parry CB (1998) Birth lesionsof the brachial plexus. In: Birch R, Bonney G, Wynn ParryCB, eds. Surgical Disorders of the Peripheral Nerves, 1stedn. Churchill Livingstone: Edinburgh: Chapter 1.

Birch R, Chen L (1996) The medial rotation contractureof the shoulder in obstetric brachial plexus palsy, JBone Joint Surg 73B(Suppl.):68.

Carlioz H, Brahimi L (1986) La place de la désinsertioninterne du sous-scapulaire dans le traitement de laparalysie obstétricale du membre supérieur chezl’enfant, Ann Chir Infant 12:159.

Fairbank HAT (1913) Subluxation of shoulder joint ininfants and young children, Lancet 1:1217–23.

Gilbert A (1993) Obstetrical brachial plexus palsy. In:Tubiana R, ed. The Hand. W B Saunders: Philadelphia:576–601.

Goddard N (1993) The development of the proximalhumerus in the neonate. In: Tubiana R, ed. The Hand.WB Saunders: Philadelphia: 624–31.

Landi A, Schoenhuber R, Funicello R et al (1992)Compartment syndrome of the scapula, Ann HandSurg 11:383–8.

Pearl ML, Edgerton BW (1998) Glenoid deformitysecondary to brachial plexus birth palsy, J Bone JointSurg 80A:659–67.

Scaglietti O (1938) The obstetrical shoulder trauma,Surg Gyn Obstet 66:868–77.

Sever JW (1925) Obstetrical paralysis. Report of elevenhundred cases, J Am Med Assoc 85:1862–5.

Tubiana R (1993) Editorial note. In: Tubiana R, ed. TheHand. W B Saunders: Philadelphia: 575.

Waters PM, Smith GR, Jaramillo D (1998) Gleno-humeral deformity secondary to brachial plexus birthpalsy, J Bone Joint Surg 80A:668–77.

Whitman R (1905) Treatment of congential andacquired luxation at the shoulder in childhood, J AmMed Assoc 42:110–15.

Zancolli EA, Zancolli ER (1993) Palliative surgicalprocedures in sequelae of obstetrical palsy. In:Tubiana R, ed. The Hand. W B Saunders: Philadelphia:602–3.

MEDIAL ROTATION CONTRACTURE AND POSTERIOR DISLOCATION OF THE SHOULDER 259

Page 269: Brachial Plexus Injuries
Page 270: Brachial Plexus Injuries

Introduction

Elbow deformity and dysfunction are frequentconsequences of obstetrical brachial plexuspalsy. In fact, lack of elbow flexion, along withpoor shoulder movement may be the firstrecognized sign of a perinatal palsy immedi-ately after birth. A variety of problems arerecognized including paralysis or paresis ofeither elbow flexion or extension or both, orstiffness of the elbow. Stiffness is typicallymanifest as a flexion contracture of the elbow,and/or a supination or pronation contracture ofthe forearm. These sequelae of obstetricalbrachial plexus palsy may be seen in infantsand children who have recovered sponta-neously, or in those who have undergone earlyneural reconstructive procedures.

In terms of incidence, the most commonpresentation is weakness of elbow flexion,followed by flexion contracture. Less commonare supination contractures of the forearm.Pronation contractures are somewhat lesscommon than supination contractures. Unlikethe adult circumstances, it is exceeding rare tosee a child with total paralysis of thebiceps–brachialis muscle groups and total inabil-ity to flex the elbow, either after spontaneousrecovery or surgery. This seems true whether theinitial insult involved only the upper elements ofthe brachial plexus, as in Erb’s palsy, or whetherthe injury was more global. In the last 10 years,we have seen only one child with total absenceof elbow flexion among more than 200 childrenexamined in our clinic (Fig. 1). Total absence ofany elbow extension is also exceedinglyunusual.

26Palliative surgery:elbow paralysisVincent R Hentz

Figure 1

This child, born via breech delivery, is the only child in ourseries who regained absolutely no elbow flexion. Hisplexus was explored at 6 months of age. C5 and C6 rootsappeared normal to observation but only small corticalresponses could be measured. The roots were not recon-structed. By 18 months of age, no biceps activity could beobserved, although EMG studies showed some electricalactivity. He subsequently underwent ulnar motor to bicepsbranch crossover.

Page 271: Brachial Plexus Injuries

Anatomy and pathology

Paralysis and paresis

Flexion of the elbow is typically a product ofactive contraction of the prime elbow flexors, thebiceps, and brachialis muscle groups. These areinnervated in most instances by axons emanat-ing from the anterior parts of the C5 and C6nerve roots. These motor axons join at thesuperior trunk and then travel via the anteriordivision of this trunk as it contributes to forma-tion of the lateral cord. They leave the lateralcord as the musculocutaneous nerve.

The elbow can be partially flexed by accessoryelbow flexors, such as the post-axial brachiora-dialis and extensor carpi radialis longus muscles,and by pre-axial muscles that have origin abovethe elbow joint, such as the pronator teres andthe flexor carpi ulnaris. All of these accessorymuscles lie too close to the axis of rotation to domore than bring the forearm to slightly less than90° and lack power at this range to performreally useful work. Moreover, in the case of thepre-axial muscles, the individual must recruitnearly all of the fibers of these muscles in orderto initiate elbow flexion, and as they do so, theforearm pronates, the wrist flexes and the fingersclench. This posture and resultant initiation ofweak elbow flexion is referred to as the‘Steindler effect’ (Steindler 1946). The brachiora-dialis and extensor carpi radialis longus musclesare typically innervated by the C5 and C6 roots.

The pronator teres and flexor carpi radialis areinnervated by the C6 and C7 roots, and the flexorcarpi ulnaris by the C7 and C8 roots.

It has been my experience that children withobstetrical brachial plexus palsy injuries affectingthe upper roots are much less adept at utilizingthis Steindler effect than adults who suffer upperroot injuries. This may have more to do with thelack of children who persist in exhibitingcomplete paralysis of the biceps and brachialismuscles after obstetrical brachial plexus palsy,compared to adults who suffer traumatic obstet-rical brachial plexus palsy.

The prime elbow extensor is the tricepsmuscle, innervated by motor axons from the C6,C7, and C8 roots. These axons travel via theposterior divisions of the upper, middle, andlower trunks, as part of the posterior cord andfinally, the radial nerve. Therefore, the tricepsremains innervated in the majority of childrenwho suffer upper root injuries. Paralysis of elbowextension is thus seen only in more globalobstetrical brachial plexus palsy injuries.

Elbow stiffness

The relatively high incidence of residual elbowflexion contracture is a subject of some conjec-ture. These children exhibit a flexion contractureof varying severity (Fig. 2). The development ofan elbow flexion contracture in the context of a

262 OBSTETRICAL PARALYSIS

Figure 2

Example of an elbow flexioncontracture in a previously unoper-ated child who suffered a C5–C6(Erb’s) perinatal brachial plexuspalsy. (From Ballinger and Hoffer,1994 with permission.)

Page 272: Brachial Plexus Injuries

weak biceps is somewhat paradoxical. The factthat this can appear relatively soon after birthhas led to speculation that the same birth traumathat led to injury to the brachial plexus maycause a direct injury to upper limb muscles,including the deltoid and the biceps. This birthtrauma may lead to unrecognized consequences,such as an unrecognized post-partum compart-ment-like state associated with death of musclefibers. These necrotic muscle fascicles will berapidly replaced by scar that may then undergocontraction, leading to elbow flexion contracture.This theory is somewhat reinforced by the greatdifficulty experienced in overcoming an elbowcontracture by means of splinting. This resis-tance to conservative splinting mirrors thatexperienced in trying to open the contractedhand and wrist following Volkmann’s contrac-ture.

This seeming paradox of this frequentlyprogressive deformity has been investigated byAitken (1952) who recognized a common triad ofdeformity including proximal ulnar curvature,radial neck clubbing, and posterior subluxationof the radial head. The pathologic mechanismsresponsible for this clinical presentation have notbeen clearly determined. Their consequences arewell recognized however and treatment is verydifficult, if not impossible.

Ballinger and Hoffer (1994) have documentedthe incidence of such contractures in theirpopulation. They studied 121 patients over theage of 3 years with Erb’s C5–C6 palsies, elimi-nating other causes of elbow pathology such asradial head dislocation, and eliminating thosewith more global palsies. In 38 children, therewas an average of 19° of contracture with arange from 0 to 40°. Only four of the 38 patientshad no elbow contracture. Of the remaining 34children, 29 had extension strength thataveraged one grade higher than flexion strength.Seven patients had what was termed ‘clinicallysignificant’ contracture, meaning greater than30°. All seven had greater flexor than extensorpower.

Ballinger and Hoffer (1994) offered threepossible explanations including persistence ofthe fetal flexion posture into the early postnatalperiod. They posited that the presence of aflexion contracture might be beneficial to theinfant with weak flexor strength, and thereforethe flexed position might be maintained during

the day. They also offered the possibility that theflexor may recover before the extensor and, bythe time the extensor recovers, the contracture isalready present.

Associated stiffness about the elbow may alsobe manifest as a supination or pronation contrac-ture of the forearm. Enlargement of the bicipitaltuberosity of the radius may occur as a conse-quence of biceps muscle pathology. Asmentioned above, the proximal ulna maybecome curved. The consequence may be proxi-mal radio-ulnar impingement.

Eberhard (1997) identified nine children whohad suffered global obstetrical brachial plexuspalsy injuries presenting with fixed supinationdeformity. Impingement of the bicipital tuberos-ity on the ulna was the main cause of the supina-tion deformity in all nine children. All wereimproved by disinsertion and reinsertion of thebiceps tendon on the bicipital tuberosity, amodification of Zancolli’s (Zancolli and Mitre1973) biceps re-routing procedure.

Functional pathology

Elbow flexion

Most children with weakness of elbow flexionafter obstetrical brachial plexus palsy injury willhave biceps–brachialis strength at the M2–M3level. They may be able to lift the hand of theaffected limb to their mouth but are unable tosustain this against resistance. As the oppositearm is typically of normal strength the child willhave used this limb since infancy for tasks thatrequire great strength. Unless these children arefrequently tested, the weakness of elbow flexionmay go unrecognized for some time because thechild will perform most activities of daily livingsomewhat effectively even with weak unilateralelbow flexion.

As the elbow functions in concert with moreproximal joints in positioning the hand, defor-mity and weakness of the shoulder willcompound the functional consequences of elbowflexor weakness. The shoulder pathology maydraw more attention than the elbow weakness,because shoulder deformity much more oftenresults in an abnormal posture of the entire limb.For example, as the child begins to walk and

PALLIATIVE SURGERY: ELBOW PARALYSIS 263

Page 273: Brachial Plexus Injuries

then to run, his internally rotated limb begins toimpinge against his body. Without sufficientexternal rotation at the shoulder, as the childflexes the elbow, the arm impinges against thebody. The child learns quickly to abduct theshoulder so that the flexing arm can clear thebody and be brought toward the mouth. This co-contracture of elbow flexion and shoulder abduc-tion, termed the ‘trumpet’ sign, may be aconsequence of misguided axonal regeneration,and thus co-innervation or a learned activity orsome element of both may exist. Regardless ofetiology, if the child cannot lift the arm easilybecause of the internal rotated humerus, thechild is more likely to avoid flexing the elbow,and the biceps and brachialis muscles maybecome somewhat ignored, underutilized andunderexercised. The consequence is persistentweakness. Correction of the pathologic shouldermay be necessary before one can adequatelyassess elbow function, particularly in the youngchild.

An additional functional problem occurs whenre-innervation has been misguided and the childco-contracts elbow flexors and extensors.Whether the etiology is a consequence of misdi-rected regenerated motor axons, or central orperipheral ephaptic transmission, or aberrantmuscle recruitment is not clearly determined.However, the functional consequences areclearly visible. The result is weakened elbowflexor power, typically seen when the child triesharder to flex the elbow.

Recently, Rollnik et al (2000) reported theapplication of botulinum toxin type A in sixchildren with severe biceps–triceps co-contrac-tion after spontaneous nerve regenerationfollowing obstetrical brachial plexus palsy injury.The children ranged in age from 2 to 4 years. Allwere treated two to three times over a period ofeight to 12 months with 40 units of botulinumtoxin at two sites along their triceps muscle. Inthis small series, elbow range of motionincreased from 0 to 25–50° to 0 to 25–100° andflexion strength from MRC classification of 1.7 to3.7. Most interestingly, there was no recurrenceafter a one-year follow-up. These findings arepreliminary but provide a degree of optimism.They strengthen the concept that neural plastic-ity plays an important role in the ultimatefunctional outcome. Weakening the overactivetriceps may allow the child to begin using the

arm more normally, thus reinforcing the creationof more appropriate neural pathways andpermitting the regression of inappropriate ones.

Elbow extension

Weakness of elbow extension may be conse-quential as the infant and later the child strivesto achieve developmental milestones. However,weakness in extending the elbow is usually feltto be less functionally disabling than weaknessin flexion at the elbow. None the less, weaknessof elbow extension interferes with the infantturning him or herself from side to side, andinterferes with attempts to push up from a sittingto a standing position. Later in life, weakness inextending the elbow interferes with the efficiencyof performing many tasks, such as straighteningthe arm to make it easy to put the arm throughthe shirt sleeve. Jones et al (1985) described theindications for surgically strengthening elbowextension in four children who had sufferedobstetrical brachial plexus palsy injury. Theaverage age at surgery was 10 years in thesefour patients.

Elbow stiffness

The consequences of an elbow with a fixedflexion contracture may be both esthetic andfunctional, especially when one considers that allthese children exhibit some limb length discrep-ancy. The fixed flexed elbow exaggerates theappearance of a short limb and further decreasesthe reach and sphere of movement of theaffected limb. As mentioned above, a contractureof greater than 30° is said to be clinically signif-icant (Morrey et al 1981).

Treatment

Paralysis/paresis of elbow flexion

The functional goals for the child with absent orweak elbow flexion after obstetrical brachialplexus palsy injury are to provide increased

264 OBSTETRICAL PARALYSIS

Page 274: Brachial Plexus Injuries

strength of flexion and elbow control. Given thatthe function of the shoulder influences elbowfunction so significantly, problems of shouldercontrol strength and range of motion aretypically addressed as well, either at an initialstage or simultaneously.

The conservative and surgical management ofthe brachial plexus injury itself is adequatelydiscussed in preceding chapters of this volume.I will focus on the management of weak orabsent elbow flexion in the child who has eitherbeen allowed to undergo spontaneous evolutionof the injury or who has already undergoneneurosurgical reconstruction.

Timing and indications

There are no clear guidelines regarding eithertiming or indications available in the medicalliterature. Most clinicians will elect to observeboth the operated babies and those allowed tospontaneously evolve over many months, oreven years, before considering introducingpalliative measures. However, one should notrestrict one’s thinking to only classical palliativetreatments such as muscle–tendon transfers,contracture release, or osteotomy. As referred toabove, aggressive early treatment of recogniz-able functional problems, such as dysfunctionalbiceps–triceps co-contracture, may play animportant role.

Early intervention

There is a clear role for secondary neurosurgicalreconstruction with the goal of restoring elbowflexion. The most common circumstanceinvolves an infant who has undergone primaryplexus exploration and reconstruction whichincluded procedures specifically designed torestore elbow flexion, but who has demonstratedno biceps recovery within the expected interval.One example is a child, now between six to ninemonths following nerve grafts from the C5 andC6 root to the anterior division of the superiortrunk with no evidence of voluntary bicepscontractions. One may consider obtaining anEMG in hopes of this providing some direction

for further reconstructive decisions. However, itis clear that the EMG provides overly optimisticassessment and is not particularly indicative offunctional outcome in obstetrical brachial plexuspalsy injuries (Benaim et al 1999). Strong consid-eration should be given to going forward withsurgical procedures that add to the motor axonpopulation potentially able to re-innervate thebiceps and brachialis. This might involve re-exploration of the original reconstructive site butmore likely involves finding fresh sources ofmotor axons. Some clinicians have termed thisas ‘supercharging’ the muscle.

Axonal supercharging

Several different methods have been described.All have as their goal, the achievement in a rapidfashion of growth of motor, as opposed to mixedmotor and sensory axons into the target muscle.Both plexal and extraplexal axonal sources havebeen described.

The most popular today is some variation ofthe procedure introduced by Oberlin (Loy et al1997), who described adjoining nearby expend-able motor fascicles directly to the motorbranches of the biceps and brachialis muscles.Oberlin suggested using motor elements of theulnar nerve and recommended sacrificing thoseinnervating the extrinsic forearm muscles ratherthan intrinsic muscles. Oberlin’s cases wereprimarily adults who had suffered traumaticplexus injuries, but this philosophy has been andcan be applied equally successfully to childrenwho have suffered obstetrical brachial plexuspalsy injury to the upper roots as opposed tothose with a global palsy. For these globallyinjured babies, other extraplexal sources ofmotor axons must be sought.

Other clinicians have looked to other sourcesof motor axons, again describing the proceduresprimarily in adults. Gilbert (Dumontier andGilbert 1990) utilized ipsilateral and evencontralateral medial and lateral pectoral nerves.Motor elements of the median nerve, or thethoraco-dorsal nerve may be used to directlysupply the biceps and brachialis to restore elbowflexion. Indeed, recovery of biceps function issuch a high priority that one could even justifyusing a less expendable motor nerve to achieve

PALLIATIVE SURGERY: ELBOW PARALYSIS 265

Page 275: Brachial Plexus Injuries

this priority. One might argue that sacrificing anerve such as the thoraco-dorsal nerve, and bydoing so accepting sacrifice of the latissimusdorsi muscle, a muscle that is often used torestore elbow flexion through a secondarymuscle–tendon transfer, may be cavalier.

The indications for such muscle ‘supercharg-ing’ are not well defined as mentioned above.Neither is the duration of the window of oppor-tunity to achieve some success. Anecdotal infor-mation indicates that the window of opportunityis of much greater duration if the muscle hasbeen at least partially re-innervated by the origi-nal surgical procedure or through spontaneousevolution. A muscle that is completely dener-vated for more than one year, even in the child,will probably not respond in a systematic fashionto this supercharging procedure.

Some argue that this window of opportunity istoo short and that children’s muscles responddifferently to nerve injuries. There is onlyanecdotal evidence to support this concept, butthat is enough for some clinicians to proposeeven primary plexus repair or reconstruction onchildren even four years or more from birth.

The outcome of such procedures for thisobstetrical brachial plexus palsy populationcannot be determined from a review of the liter-ature. Too few cases have been performed in thiscohort.

Late intervention

Timing and indications

Late intervention relies primarily on classicalmuscle–tendon transfers and rarely on suchprocedures as microvascular functional freemuscle transfers. Timing and indications formuscle transfers to restore elbow flexion aresimilar to those for most related procedures.They must typically await some level of psycho-logical maturity on the part of the child. Mosttransfers benefit from a period of formal therapyas part of the re-education process. Childrenbetween ages 2 and 4 rarely exhibit the neces-sary maturity.

Once sufficient maturity has been achieved, itis possible to assess in a meaningful and reliablemanner the strength of remaining muscles. Prior

to this age, this assessment is difficult if notimpossible.

The indications for muscle–tendon transfersto strengthen or restore elbow flexion in thispopulation must be carefully individualized.Even weak flexion will be effective flexion formost activities. Reconstruction of elbow flexionrequires surgery of far greater technicalcomplexity than most other standard muscle–tendon transfers. Re-education of the mostcommonly used muscle resources, such as thetriceps, latissimus dorsi, and pectoralis major isfar more complex than that associated withother transfer procedures. One must carefullyconsider the risks and benefits in performingrelative complex reconstructive surgery requir-ing sophisticated re-education maneuvers in avery young patient. Resources to restore elbowfunction are too scarce to waste through surgi-cal misadventure.

Specific procedures

Steindler flexor-plasty (Steindler 1946)

There is insufficient information in the literatureto allow one to determine which of the severalcommon transfers is best for the pediatric post-obstetrical brachial plexus palsy injury popula-tion. From the standpoint of technical andre-education simplicity, the Steindler proceduremust prevail over latissimus dorsi or pectoralismajor transfer. One might consider transferringthe pectoralis minor either in isolation orcombined with the Steindler procedure, if oneseeks a more modest boost to elbow flexionpower. I have preferred this procedure in thosepediatric patients who exhibit good wrist stabil-ity and active elbow that is flexion sufficient toallow them to initiate elbow flexion. In this case,one can attach the transfer under a little lesstension and usually avoid extension loss whileanticipating better power in the mid-range ofelbow flexion. These patients need power in thisrange and not necessarily at the end range.

The Steindler procedure is best performed inpatients who can exhibit the Steindler effectproduced by pronating the forearm and flexingthe wrist and at the same time swinging the armat the elbow to overcome gravity. With this

266 OBSTETRICAL PARALYSIS

Page 276: Brachial Plexus Injuries

maneuver, they should be able to then maintaintheir arm somewhat flexed at the elbow againstthe force of gravity. Those patients lacking goodwrist extension are poor candidates for thisprocedure, as it accentuates wrist flexor power.

Author’s preferred technique

A curved incision is planned that is orientedlongitudinally across the medial epicondylar areaat the elbow and extends proximally and distallyover sufficient length to expose the junction ofthe distal one-third and middle one-third of thehumerus and the proximal one-half of the

forearm flexor pronator muscle mass (Fig. 3, A).The incision is made through subcutaneoustissues with care being taken to preserve majorcutaneous nerves. The medial epicondyle isexposed, and the ulnar nerve is located andcleared above the cubital tunnel behind themedial epicondyle. The nerve is gently elevatedout of the cubital tunnel, preserving motornerves coursing to the flexor carpi ulnaris (Fig.3, B). The flexor carpi ulnaris is split distally fromthe point where the ulnar nerve penetrates it toallow free mobilization of the nerve and muscle.The subcutaneous dissection is carried distallyover the pronator teres, which forms the radialborder of the flexor pronator muscle mass.

PALLIATIVE SURGERY: ELBOW PARALYSIS 267

Figure 3

The operative steps of theSteindler flexor-plasty are illus-trated. The flexor/pronator origin isdetached along with a piece of themedial epicondyle (A,C). The ulnarnerve is protected and the mediannerve is dissected to allow proxi-mal migration of the branches tothe pronator teres (B). The bonyattachment of the muscle origins ismoved proximally on the humerusand is placed on the mid-anteriorsurface of the humerus at a pointcomfortably reached by the muscleorigin/epicondylar block with theelbow flexed 45 to 60° (D). Theepicondyle may be fixed to thehumerus with a wire or screw (E).

Page 277: Brachial Plexus Injuries

The median nerve is best located at elbowlevel just ulnar to the biceps tendon and beneaththe bicipital fascia (lacertus fibrosus) andadjacent to the brachial artery. The median nerveis elevated and dissected distally, special carebeing exercised to preserve motor branches tothe pronator teres (Fig. 3, B).

The origin of the muscle mass from the medialepicondyle is identified, and one may judgewhere an osteotomy should be placed to elevateand move the entire conjoined origin of themuscles. With the median and ulnar nervescarefully protected, the epicondyle may beseparated from the humerus, together with themuscle origins, by controlled osteotome osteo-tomy (Fig. 3, C). The muscle origins, along withthe freed epicondyle, may then be reflected up,and the flexor pronator muscle mass may bedissected from the underlying joint capsule in anatural separation plane. The muscle group isstripped from the ulna distally, to the extentallowable by motor nerve branches from both themedian and the ulnar sides. The flexor carpiulnaris is split along the plane started by thepenetration of the muscle by the ulnar nerve.With the elbow flexed to about 45°, one mayreflect the freed epicondyle and muscles acrossthe elbow to get an impression of how far proxi-mally on the humerus the transferred mass willreach (Fig. 3, D). The anterior surface of thehumerus is exposed at this level by splitting whatremains of the atrophic biceps and brachialismuscles. With the median nerve reflectedradially, the proper site for bone-to-boneepicondyle-to-humerus juncture is determined. Itis placed on the mid-anterior surface of thehumerus at a point comfortably reached by themuscle origin/epicondylar block with the elbowflexed 45 to 60°. The anterior cortex of thehumerus is removed to receive the epicondyle.The epicondyle may be fixed in place with a largewire suture or by screws (Fig. 3, E). After initialfixation, the extent of elbow extension is gentlytested with the forearm supinated and pronated.It should extend to 60 or 70° without unduetension. The lack of traction on and freedom ofthe median and ulnar nerves are checked, and thetourniquet is removed to achieve satisfactoryhemostasis before closure of the skin wound.

The elbow is immobilized at about 60° offlexion in a dressing with an overlying splint.After 3 weeks of healing, gentle passive exten-

sion may be initiated, with progression to activeand passive range-of-motion exercises over aperiod of 3 more weeks. Night splinting inprogressive maximum extension is important.Some prefer a turnbuckle extension device forprogressive night-time extension splinting. Manymonths of instructed and self-administeredphysical therapy prove rewarding to patientswho have undergone this operation.

Latissimus dorsi transfer (Zancolliand Mitre 1973)

I prefer this technique for pediatric patients whohave a strong latissimus and little elbow flexionpower. Rehabilitation is difficult and biofeedbackand functional muscle stimulation have provedhelpful.

Author’s preferred technique – bipolartransfer

The three critical elements of the procedure areas follows:

1. Careful dissection of the neurovascular struc-tures to allow proper re-positioning of thelatissimus dorsi from posteriorly to anteriorlywithout the risk of kinking or torsioning theneurovascular pedicle;

2. Transfer of the normal tendinous insertion ofthe latissimus dorsi on the humerus to thecoracoid process of the scapula to align themuscle properly as an elbow flexor;

3. Weaving of the cut end of the biceps insertionthrough the fibrous origin of the transferredlatissimus dorsi so that the origin of the latis-simus dorsi became its new insertion.

Function of the latissimus dorsi is first checkedby palpating the posterior axillary fold andwatching the muscle area on the back while thepatient is asked to adduct and dorsally displacethe arm (as in elbowing one’s way through acrowd). Tensing of the muscle is readilydetected, and often the outline of the musclemay be seen on the back.

Knowing that the arc of rotation is close to thespine of the scapula along the axillary border,

268 OBSTETRICAL PARALYSIS

Page 278: Brachial Plexus Injuries

one may plan a skin island, if it is needed, so thatit will be carried to an area of deficit upon thearc of rotation. An incision is made along theaxillary surface of the muscle in order not toviolate the island of skin but to allow exposureof the thoraco-dorsal neurovascular bundle.Once the bundle is noted, the island of skin, ifneeded, is circumscribed by a continuation of theincision, and the skin edges are tacked down tothe epimysial fascia with temporary sutures. Therest of the superficial surface of the latissimusdorsi is exposed by dissecting the areolar planesuperficial to the muscle fascia. A marking sutureis placed proximally, close to the muscle origin,and another is placed distally, near the insertionsite. The distance between these sutures is notedwith the arm in adduction and then in abduction.The whole muscle may be dissected up along itsdeep surface, carefully avoiding inadvertentelevation of the serratus anterior muscle with thelatissimus. The origin from fascia along thelumbosacral vertebrae and posterior iliac crest iscut, preserving enough fascia on the musclefrom which to fashion a new tendon of insertion.Segmental vessels to the muscle along themedial origin constitute the minor blood supplyto the muscle. They may be sectioned withoutconcern for flap survival when the majorthoraco-dorsal pedicle is preserved. The muscleis dissected to its humeral insertion, carefullypreserving the pedicle, and the insertion itself issectioned in preparation for the bipolar transfer.

The muscle may now swing through a widearc, with the neurovascular pedicle as a pivot,and passes across the axilla. Final tension will beadjusted so that the original resting length of thelatissimus dorsi is re-established. This is judgedby placing marking sutures regularly along themuscle before its release from its original site.The muscle is tubed and tunneled betweenincisions made at the shoulder level and on thefront of the distal arm, just proximal to the elbow.The distal end of the latissimus dorsi is attachedto the biceps tendon by a weave technique (Fig.4, A). The proximal end is attached to thecoracoid process (Fig. 4, B). When the elbow isput through its range of flexion and extensionwith the arm abducted and adducted, the rangeshould duplicate as closely as possible that notedwhen the muscle was in its original position.

After closure of the wound, the limb is splintedwith the elbow flexed at 90°. After 6 weeks of

immobilization, active exercises are initiated, andthe elbow is allowed to move gradually to itsfully extended position in the following 3 weeks.

Re-training for elbow flexion can be difficult, asit often is with the latissimus dorsi transfer.Careful physical and occupational therapy andbiofeedback exercises are helpful. The patientmay be able to accelerate rehabilitation bycoughing to stimulate the latissimus dorsi. Sixmonths or more are usually required to achievefull strength and excursion of the transferredmuscle.

Pectoralis major transfer

I have little experience with this procedure in thepediatric population. The scarring in this agegroup is significant and a source of parentalconcern.

Author’s preferred technique (Carroll andKleinman 1979)

Integrity of the pectoralis major muscle isassessed by grasping the anterior axillary foldand observing the anterior chest wall while the

PALLIATIVE SURGERY: ELBOW PARALYSIS 269

Figure 4

Operative steps in bipolar transfer (A, B) of the latissimusdorsi to restore elbow flexion. See text for details.

A

B

Page 279: Brachial Plexus Injuries

patient adducts and internally routes the armagainst resistance.

An incision is planned that will course alongthe muscle 2 cm below the superior insertion onthe clavicle and 2 cm lateral to the sternal origin.The incision is carried down to the xyphoid areato expose the superior extent of the anteriorrectus sheath (Fig. 5, A).

Laterally, the incision is extended along thedeltipectoral groove marked by the cephalic vein.If skin is to be carried with the pectoralis muscle,the incision is designed to circumscribe an islandof skin in a position appropriate to meet the needin the arm. The position is determined by design-ing it in relation to the pivotal pedicle in therecipient and then the donor position.

A second incision is planned in the antecubitalfossa to expose the biceps tendon and its inser-tion into the biceps tuberosity of the radius.

The incision is initiated superiorly to exposethe deltopectoral groove and cephalic vein. Thegroove is deepened, retracting the cephalic veinsuperiorly. The pectoralis minor insertion on thecoracoid process is seen and cleared. The clavic-ular origin of the pectoralis is detached to exposethe pectoral vessels and the medial and lateralpectoral nerves. Electrical stimulation helpsidentify the nerves. Once the neurovascularstructures are identified, they are carefullyprotected as the dissection proceeds.

270 OBSTETRICAL PARALYSIS

Figure 5

Operative steps for transfer of thepectoralis major to restore elbowflexion. See text for details.

Page 280: Brachial Plexus Injuries

If an island of skin is to be carried, a circum-scribing incision is made around it, and the skinis tacked down to the pectoral fascia with tempo-rary sutures. The rest of the muscle is cleared inthe extrafascial plane over its entire superficialsurface.

Marking sutures may be placed, one at thejunction of the muscle with the rectus sheath anda second close to the muscle’s insertion on thehumerus. Measurements of the distancebetween these two marking sutures with the armabducted and externally rotated and then withthe arm adducted and medially rotated arecorded for later use.

A tongue of rectus fascia for use as a distalmuscle attachment is outlined and elevated,leaving it attached to the pectoralis muscle (Fig.5, B). The superior, medial, and inferior muscleattachments are severed, and the whole muscleis elevated, dissecting in the areolar subpectoralplane superficial to the pectoralis minor. Greatcare is exercised to avoid inadvertent injury tothe primary neurovascular bundle. If the muscleis to be transferred as a bipolar transfer, theinsertion of the pectoralis major on the humerusis sectioned close to bone (Fig. 5, C, large arrow).

A second incision is made across the antecu-bital fossa and is extended proximally on themedial side and distally on the lateral side toexpose the tendon of the paralyzed biceps closeto its insertion on the radial tuberosity (Fig. 5, D).

If the skin and subcutaneous tissue coveringthe arm are uninjured and sufficient, a tunnel iscreated by blunt and sharp dissection just super-ficial to the biceps fascia. The tunnel connectsthe proximal and distal incisions along theanterior (biceps) side of the arm. If skin is insuf-ficient by virtue of loss or scarring, it is resectedto good skin on all sides, and enough skin willbe brought from the chest wall with the muscleto cover the transferred muscle without skintension.

The pectoralis muscle is shifted to check foradequacy of the pedicle dissection. The medialinferior corner of the muscle with its rectussheath tongue is swung distally to the bicepsinsertion, and the fascia end, previously insertedon the humerus, is swung upward to theacromion process beneath the retracted deltoid(Fig. 5E). The muscle is shifted proximally anddistally to a position that will allow it to contractand relax with elbow flexion and extension and

with the arm abducted, creating the least possi-ble tension on the neurovascular structures. Themuscle is rolled into a fusiform shape, and thejunctures are made distally and proximally (Figs5, D, E). The tension of the muscle is judged bysetting the two insertion sites such that when theelbow is flexed and extended, the measurementsbetween the marking sutures come closest toduplicating the range of contraction noted whenthe muscle was in its original bed.

Wounds are closed with or without the skinisland according to indication, and the arm isimmobilized on the anterior trunk with the elbowflexed during healing. Immobilization for 4–6weeks is followed by progressive passive andthen active range-of-motion exercises.

Other muscle–tendon transfers torestore elbow flexion

Triceps to biceps transfer: This transfer (Hoanget al 1989) has a long history and has proven tobe effective in restoring a good range of motion,but at the expense of all active elbow extension.This transfer is probably best performed incircumstances where dysfunctional biceps–triceps co-contraction fails to respond to conser-vative treatment such as botulinum toxin injec-tion, or where no other suitable local muscle isavailable.

Sternocleidomastoid to biceps transfer: Thistransfer is rarely performed today, primarilybecause of the significant cosmetic deformitythat results. Kumar et al (1992) have described amodification of the standard technique that mayreduce what is otherwise a strikingly alteredappearance.

Advancement of the insertion of the biceps:Nemoto et al (1996) recommends moving theattachment of the biceps about 2 cm distal to itsoriginal insertion. He reported improved flexionpower in six patients.

Functional free muscle transfers: The indicationsfor the application of these procedures in thepediatric post-obstetrical brachial plexus palsypopulation are similar to those for the adultpopulation. The sources available to innervate

PALLIATIVE SURGERY: ELBOW PARALYSIS 271

Page 281: Brachial Plexus Injuries

the transfer are similar to those discussed for theadult traumatic brachial plexus patient. This isdiscussed more completely in other chapters.One must be concerned about using severalintercostal nerves to re-innervate the free muscletransfer in a young child who has a coexistingipsilateral phrenic nerve palsy with evidence ofreduced pulmonary function.

Results after muscle–tendontransfers to restore elbowflexion

Most of the literature dealing with reconstructionof elbow flexion describes global results and thepatient population is primarily adult males whohave suffered traumatic brachial plexus palsy.No series defines results in only pediatric cases.In these reports, there is no consensus on whichdonor muscle performs the best. Botte andWood (1989) reviewed the Mayo Clinic seriesand found that latissimus and pectoralis transferprovided about the same arc of antigravitymotion. Moneim and Omer (1986) found disap-pointing results on standardized daily livingevaluations in his series of patients who hadundergone latissimus transfer. Marshall et al(1988) reviewed London’s St. Mary’s Hospitalexperience and determined that latissimus andtriceps performed best; Steindler procedureswere adequate while the results of pectoralistransfers were disappointing. Liu et al (1993)reported 71 consecutive cases of Steindler flexor-plasty including 13 patients who suffered obstet-rical brachial plexus palsy. Seventy-nine per centwere judged to have good or excellent results.Dutton and Dawson (1981) reviewed 25 consec-utive Steindler procedures. The mean arc offlexion was 95° with an average loss of exten-sion of 36°. Brunelli et al (1995) reported goodresults with this procedure. Matory et al (1991)described modifications in the incisions forpectoralis transfer that reduce the scar effects.Hoang (1989) reviewed seven patients followingtriceps to biceps transfer, five of whom couldbring their hand to their mouth.

It is difficult to extrapolate results obtained inadults to the pediatric population. Many adultswho need these operations have flail shouldersand in many arthrodesis has been performed.

The salutary effects of shoulder arthrodesis onelbow flexion strength has long been recognized.Few obstetrical brachial plexus palsy patientsundergo gleno-humeral arthrodesis.

Elbow extension

The methods to achieve improved elbow exten-sion include such traditional muscle–tendontransfers as latissimus to triceps and posteriordeltoid to triceps transfers (Moberg 1975). Jones(1985) reported on the outcome of latissimus totriceps transfers in four children after obstetricalbrachial plexus palsy. The surgical technique issimilar to that described for reconstruction ofelbow flexion, differing only in that the transferworks well as a unipolar transfer (Fig. 6).Postoperatively, all demonstrated increasedstrength of elbow extension and improvementsin activities of daily living.

Elbow stiffness

The treatment of the stiff elbow in the post-obstetrical brachial plexus palsy population has

272 OBSTETRICAL PARALYSIS

Figure 6

Use of the latissimus dorsi for elbow extension. A simplermonopolar transfer can be performed.

Page 282: Brachial Plexus Injuries

proved difficult. Once a contracture has devel-oped it is very resistant to conservative treat-ment. Dynamic extension splinting has beenminimally successful in restoring lost extensionrange but is effective in preventing progressionof the deformity.

For significant deformity (greater than 30°)surgical release may be appropriate. Soft tissuecontracture will account for some of the contrac-ture and thus soft tissue release will be partlyeffective. This includes anterior capsular releaseand lengthening of the biceps, either as an intra-muscular release or by formal Z-lengthening.Some fractional lengthening of the humeralorigins of the pronator teres and flexor carpiradialis may be necessary.

Very severe contractures are rare and areprobably best managed by some combination ofsoft tissue release and a dome osteotomy of thehumerus at the supracondylar level.

References

Aitken J (1952) Deformity of the elbow joint as a sequelto Erb’s obstetrical paralysis, J Bone Joint Surg34B:352–58.

Ballinger S, Hoffer M (1994) Elbow flexion contracturein Erb’s palsy, J Child Neurol 9(2):209–10.

Benaim J, Jouve J, Bardot J et al (1999) Pseudo-paral-ysis of the brachial biceps in obstetrical brachial plexuslesions (OBPL): concerning the ‘overly optimistic’ EMGin OBPL, Neurophysiol Clin 29(6):490–4.

Botte M, Wood M (1989) Flexorplasty of the elbow, ClinOrthop 245:110–16.

Brunelli G, Vigasio A, Brunelli G (1995) ModifiedSteindler procedure for elbow flexion restoration, JHand Surg 20A(5):743–6.

Carroll R, Kleinman W (1979) Pectoralis major transferto restore elbow flexion to the paralytic limb, J HandSurg 4:501–7.

Dumontier C, Gilbert A (1990) Traumatic brachialplexus palsy in children, Ann Chir Main Memb Super9(5):351–7.

Dutton R, Dawson E (1981) Flexorplasty. An analysis oflong-term results, J Bone Joint Surg 63A(7):1064–9.

Eberhard D (1997) Transposition of the bicipitaltuberosity for treatment of fixed supination contracturein obstetric brachial plexus lesions, J Hand Surg22B(2):261–3.

Hoang P, Mills C, Burke F (1989) Triceps to bicepstransfer for established brachial plexus palsy, J BoneJoint Surg 71B(2):268–71.

Jones B, Manske P, Schoenecker P, Dailey L (1985)Latissimus dorsi transfer to restore elbow extension inobstetrical palsy, J Pediatr Orthop 5(3):287–9.

Kumar V, Satku K, Pho R (1992) Modified technique ofsternomastoid transfer for elbow flexion, J Hand Surg17A(5):812–13.

Liu T, Yang R, Sun J (1993) Long-term results of theSteindler flexorplasty, Clin Orthop 296:104–8.

Loy S, Bhatia A, Asfazadourian H, Oberlin C (1997)Ulnar nerve fascicle transfer onto to the biceps musclenerve in C5–C6 or C5–C6–7 avulsions of the brachialplexus. Eighteen cases, Ann Chir Main Memb Super16(4):275–84.

Marshall R, Williams D, Birch R, Bonney G (1988)Operations to restore elbow flexion after brachialplexus injuries, J Bone Joint Surg 70B(4):577–82.

Matory W Jr, Morgan W, Breen T (1991) Technicalconsiderations in pectoralis major transfer for treat-ment of the paralytic elbow, J Hand Surg16A(1):12–18.

Moberg E (1975) Surgical treatment for absent single-hand grip and elbow extension in quadriplegia: princi-ples and preliminary treatment, J Bone Joint Surg57:196–206.

Moneim MS, Omer GE (1986) Latissimus dorsi muscletransfer for restoration of elbow flexion after brachialplexus disruption, J Hand Surg 11A(1):135–9.

Morrey B, Askew L, An K, Chao, E (1981) A biome-chanical study of normal elbow motion, J Bone JointSurg 63:872–879.

Nemoto K, Itoh Y, Horiuchi Y, Sasaki T (1996)Advancement of the insertion of the biceps brachiimuscle: a technique for increasing elbow flexion force,J Shoulder Elbow Surg 5(6):433–6.

Rollnik J, Hierner R, Schubert M et al (2000) Botulinumtoxin treatment of cocontractions after birth-relatedbrachial plexus lesions, Neurology 55:5–6.

PALLIATIVE SURGERY: ELBOW PARALYSIS 273

Page 283: Brachial Plexus Injuries

Steindler A (1946) Orthopaedic Operations, Indications,Technique and End Results. Charles C Thomas:Springfield IL.

Zancolli E, Mitre H (1973) Latissimus dorsi transfer torestore elbow flexion. An appraisal of eight cases, JBone Joint Surg 55A(6):1265–7.

274 OBSTETRICAL PARALYSIS

Page 284: Brachial Plexus Injuries

Obstetrical palsy of the upper limb represents asevere traumatic complication during delivery,which basically involves the brachial plexus andoccasionally the osteoarticular structures of theshoulder and its deep periarticular muscles.

Our aim is the description and treatment of therelatively frequent obstetrical palsy sequela –consequent to extensive lesions of the brachialplexus – represented by forearm supinationdeformity, usually associated with ulnar deviation

27Palliative surgery: pronosupination inobstetrical palsyEduardo A Zancolli (II)

Figure 1

(A) Early flexible supination deformity in a 2-year-old girl after a whole-limb with partial recovery type of sequela. Theseinitial flexible forearm deformities offer the opportunity to prevent the retraction of the interosseous membrane throughthe indication of early tendon transfers to rebalance the pronosupination function. (B) Very severe (180°) supination defor-mity of the forearm in a 15-year-old girl after a severe sequela of the distal type. A postoperative X-ray is seen in Fig. 11.

a b

Page 285: Brachial Plexus Injuries

of the wrist and variable deficits of hand function.Forearm supination deformity tends to beprogressive, culminating in severe retraction ofthe interosseous membrane and subluxation ordislocation of the distal radio-ulnar (DRU) jointand in a few advanced cases to dislocation of theradio-proximal radius head (Figs. 1b, 3, 4). Earlyrecognition of this deformity is of paramountimportance to prevent fixed deformities, whichincrease the hand deficit (Fig. 1a).

General considerations

Before dealing with the forearm supinationdeformity, we consider it of importance tomention some general concepts related toobstetrical palsy and its sequelae.

Sequelae of obstetrical palsy depend on threemain initial lesions: (1) the brachial plexus, (2)the osteoarticular system, particularly at theproximal humeral level, and (3) direct damage tothe deep periarticular muscles of the shoulder.

Brachial plexus injury manifests by differentlesions of the motor, sensitive and sympatheticfibers: neuropraxia, axonotmesis and rootavulsion. In axonotmesis the magnitude and timefor recovery will depend on its degree of severity.Three main topographical initial brachial plexuslesions have been classically recognized: (1) uppertype (Duchenne–Erb paralysis) (Duchenne 1872,Erb 1874), where the cervical nerves C5 and C6 areinvolved; (2) upper and middle type, with lesion ofC5, C6 and C7 (C7 lesion can be total or partial),and (3) total type (Duchenne–Erb–Klumpke paraly-sis) with lesion of C5–T1. In our series we have notfound isolated lesions of the C8–T1 (Klumpke’sparalysis, Klumpke 1885). Occasionally the patternof nerve lesions is irregular.

This classification of initial brachial plexusinjuries differs from the classification we haveadopted (Zancolli III et al 1979a,b,c, Zancolli II 1981,Zancolli II and Zancolli 1988, 2000), to group thediverse sequelae of the upper limb (see below).

The primary osteoarticular lesion of the proxi-mal humerus (epiphysiolysis) was initiallydescribed by Kuestner (1885) and thoroughlystudied by Putti (1932) and Scaglietti (1938).Primary epiphysiolysis was present in 32.5 percent of our series (1979). We consider that thedeep periarticular muscles of the shoulder joint

can be affected by the initial obstetrical trauma.This explains early muscular fibrosis, such as thecontracture of the subscapularis muscle a fewmonths after birth. A similar interpretation forcongenital torticolis by direct birth injury hasbeen made by Suzuki et al (1984).

Other factors that contribute to increase theinitial lesions are: incorrect or delayed initialtreatment; forced manipulations with the inten-tion to elongate severely retracted muscles,ligaments and capsules; plaster casts or bracesfor the shoulder in forced ‘Statue of Liberty’position, and surgical pitfalls.

A delay in the rebalance of paralytic muscles orthe prevention and correction of muscular contrac-tures or ligament retractions usually allowsprogressiveness of deformities and joint subluxa-tions or dislocations. A typical example is repre-sented by the classical and common supinationdeformity of the forearm caused by the muscularimbalance created in distal brachial plexus lesions.In these cases the interosseous membrane beginsto retract within a few years after birth (Fig. 1) andwith time the DRU joint subluxates or dislocates.Another example is represented by early andsevere internal contractures of the shoulder thatfrequently demand surgical release early on –usually after 6–8 months of life – to prevent defor-mity of the humeral epiphysis.

These examples indicate the importance ofcreating new muscular balances at differentlevels of the upper limb, particularly in thosedeformities that tend to be progressive, produc-ing bone deformities and joint subluxation ordislocations.

Clinical material

Our clinical material to obtain a classification ofsequelae of the upper limb in obstetrical palsywas represented by an initial study of 368patients evaluated at the National RehabilitationCenter of Buenos Aires, Argentina (1979)(Zancolli et al 1979a,b,c). Of this series, 203patients were operated on for sequelae of theshoulder and 93 patients received reconstructivesurgery of the elbow, forearm and hand. Ananalysis of another series of 148 cases (ZancolliII and Zancolli III 1988) showed that 63 per centof patients were female of whom 54 per cent

276 OBSTETRICAL PARALYSIS

Page 286: Brachial Plexus Injuries

were affected on their right upper limb and 4 percent were affected on both sides. In this series ahigh birth weight (> 4 kg) was recorded in 92 percent of cases. Breech presentation occurred in 12per cent. Other pathologies observed at birthwere: congenital dislocation of the hip (1.3 percent), torticolis (2.7 per cent), and partial paraly-sis of the lower limb caused by a combinedspinal cord lesion (1.2 per cent). All these percent-ages relate to a sample group of 148 patients.

Recovery of muscle paralysis was variable. Thesooner after birth that the recovery begins, thebetter is the final result. Particularly goodfunctional results are obtained from those caseswhere recovery begins before week 4–6; fair if itbegins after week 6 and poor, with severe seque-lae, if recovery does not begin before 6 months ofage. In favorable cases spontaneous recovery cancontinue improving until 18–24 months. Our recentobservations (Zancolli II and Zancolli III 1988, 2000)showed that biceps function begins to recover in75 per cent of the cases after the fifth month. Thisindicated us that direct surgery of the upper plexusshould not be undertaken before the fifth month,approximately. Some 18 per cent of our cases didnot recover elbow flexion and this should be thecase for direct reconstructive surgery of the upperbrachial plexus (C5–C6 nerve tissue replacement inpostganglionar lesions). Of our patients withmiddle-plexus injuries, 46 per cent had a definitivetricep paralysis. In total plexus lesions, 79 per centrecovered some finger function but 21 per centhad no recovery at all (complete hand paralysis).When the hand begins to recover before 3 monthsof age the final functional result will generally beacceptable. If hand active movements have notappeared at 6 months, the final function will bevery poor or non-existent. Horner’s syndrome wasinitially present in the cases with severe paralysisof the hand (21 per cent).

Classification of sequelae

Very little has been written relating to classificationof the residual deformities of obstetrical paralysisof the upper limb. Typically, classifications haveusually referred to the initial lesions of the brachialplexus, but not to its sequelae after the recoveryperiod. Only the shoulder has received specialattention (Steindler 1923, Gilbert et al 1988).

Based on our clinical material we initially(Zancolli II et al 1979a,b,c) recognized three maintypes of sequelae: (1) proximal upper-limb, (2)predominant distal and (3) predominant whole-limb paralysis. Through subsequent evaluations(Zancolli II and Zancolli III 1988, 2000), we nowrecognize five types of sequelae that give us abetter correlation with the selection of theperipheral reconstructive procedures (Table 1).

In the proximal limb (type 1) (53 per cent) theprincipal pathology is located in the shoulderand elbow and occasionally a moderate prona-tion contracture of the forearm can be demon-strated. The shoulder sequelae have beendivided by us into two main groups: (1) jointcontracture and (2) flaccid paralysis. Shouldercontractures are divided into five subgroups inrelation to the local pathology and surgical treat-ment (Zancolli II 1984).

In the proximal middle-limb (type 2) (10 percent), the affected proximal limb muscles aresimilar to type 1, with the addition of muscularparalysis depending on the C7 cervical nervelesion, such as: triceps, pronator teres, extensorsof the wrist and fingers and flexor carpi radialis.The latissimus dorsi is usually very weak and notsuitable for transfer. If the triceps is very weak aparalytic flexion contracture of the elbow isfrequently present. This group of patients maypresent a forearm supination deformity due tothe muscle unbalance created between thepronosupination muscles.

The posterior cord (type 3) (10 per cent) ischaracterized by severe functional deficit presentin shoulder abduction and elbow, wrist, fingersand thumb extension, due to the distal musclesbeing innervated by the nerves depending on theposterior cord of the brachial plexus.

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 277

Table 1 Classification of obstetrical palsy sequelae

Type Clinical picture Percentagea

1 Proximal limb 532 Proximal middle-limb 103 Posterior cord 104 Whole-limb with partial recovery

(forearm supination or pronation deformities) 25

5 Whole-limb with complete flaccid paralysis 2

aPercentages referred to 148 patients evaluated (Zancolli II andZancolli III, 1988, 2000).

Page 287: Brachial Plexus Injuries

In the whole-limb with partial recovery (type 4)(25 per cent) the entire upper limb is affected butpartial recovery has been achieved. The principaldeficit is at the distal part of the limb (forearm,wrist and hand). These are the patients who

most frequently present with supination defor-mity of the forearm. In our series we have notfound a single case of exclusive sequelae of thedistal part of the upper limb, presenting normalshoulder and elbow. In our series of the limb of

278 OBSTETRICAL PARALYSIS

Figure 2

(a) Anatomy of the interosseous membrane. (Dorsal view of the forearm.) r, radius; u, ulna; 1, strong descending radio-ulnar fibers (DRUF); 2, weak descending ulno-radial fibers (DURF). (b) This drawing show the two types of fibers (dorsalview of the forearm). The DRUF (1) run between the anterior aspect of the interosseous border of the radius to theinterosseous border and posterior aspect of the ulna. These fibers represent the principal part of the membrane. Theirfibers are very thick and strong and retract very easily in the forearm supination deformity in obstetrical palsy. Dorsallyto the DRUF are thin fibers, the DURF (2), that run distally, crossing the DRUF. The DRUF tense in supination and theDURF tense in pronation. (C) We have found in cadaveric investigations that the interosseous space widens at itsmaximum in a supination of 20° (20 mm) and that reduces to 14 mm in maximum supination (D) and to 10 mm inmaximum pronation (E). (Investigation in an adult cadaveric specimen.)

a b

Page 288: Brachial Plexus Injuries

type 4 sequelae has the following characteristics:(1) moderate internal rotation contracture of theshoulder and severe functional deficit to abduc-tion; (2) flexion contracture of the elbow, (3)supination deformity of the forearm, frequentlyassociated with ulnar deviation of the wrist (Fig.8); (4) pronation contracture of the forearm (Fig.12), and (5) different types of digital paralysis(Figs 1–3, 8, 12). At the end of the growing periodthe upper limb usually presents hypotrophy, witha considerable reduction of its length.Occasionally Horner’s syndrome may persist.

Shoulder abduction was limited in our cases toan average of 56° – range between 10 and 90°.The Putti’s sign was present in 27 per cent of thecases. Elbow flexion paralysis was present in 18per cent and elbow extension paralysis in 46 percent. Flexion contracture of the elbow was

present in 62 per cent of the patients (average30°). There were three cases with dislocation ofthe proximal end of the ulna and pronationdeformities of the forearm (Fig. 12) and one casewith osteonecrosis of the humeral trochlea.

In the whole-limb with complete flaccid paral-ysis (type 5) the entire upper limb is paralyzed,without joint deformities. The only active musclecontrolling the limb is the trapezius. This is avery uncommon type of obstetrical sequela (2per cent).

Forearm supination deformityassociated with ulnar deviationof the wrist

Clinical picture

Very frequently, supination deformity of theforearm results from extensive paralysis of thelower part of the upper limb after obstetricalpalsy. Very few papers have been written regard-ing its treatment (Grilli 1959, Owings et al 1971,Putti 1932, Steindler 1923, Zancolli II and ZancolliIII 2000). According to our series the supinationdeformity of the forearm was particularlyobserved in the types 2 (proximal middle-limb)and 4 (whole-limb with partial recovery) of resid-ual deformities. In both types of sequelae animbalance is produced between the active supina-tor muscles (biceps and supinator) and the paral-ysed pronator muscles (pronator teres andpronator quadratus). Considering both types ofsequelae, totalling 35 per cent of all the upperlimb sequelae, the supination deformity waspresent in 69 per cent of the cases with persistentand extensive distal paralysis. Pronation contrac-ture was present in 28 per cent of our series.

Initially, forearm supination deformity may bereduced passively (Fig. 3, 7). With time andgrowth – usually after 2 years of age – theinterosseous membrane begins to retract and thedeformity cannot be passively reduced. In thesupination position of the forearm theinterosseous membrane space of the forearmreduces and the interosseous becomes retractedin its strong descending radio-ulnar fibers (Fig.1b, 2). Under these circumstances these fibersretract and pronation is blocked. The deformity

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 279

Figure 3

Progressiveness of the forearm supination deformity inobstetrical palsy. (A) Initial deformity with or withoutretraction of the interosseous membrane. The DRU jointis congruent. (B) Fixed deformity with retraction of theinterosseous membrane and the dorsal structures of theDRU joint and volar dislocation of the distal ulna. (C) Volardislocation of the distal ulna and the radius head (arrows).

a

b

c

Page 289: Brachial Plexus Injuries

becomes fixed very quickly and with time thedeformity produces a curvature of the forearmbones, especially the radius (Figs 3, 7, 8), andvolar subluxation and dislocation of the distalend of the ulna (Figs 3, 4, 8). In severe deformi-ties the radial head dislocates volarly (Figs 3, 4b,8). The dislocation of the distal end of the ulnais palpated over the volar aspect of the wrist withthe forearm in supination (Fig. 8b).

Associated with forearm supination, the wristhyperextends due to the paralysis of its volarflexors and the partial activity of the dorsi flexors

tendons (Figs. 4b, 8a, b). With forearm in supina-tion gravity increases wrist hyperextension andulnar deviation. The frequent presence of anactive extensor carpi ulnaris (ECU) muscle tendsto accentuate the ulnar deviation of the dorsi-flexed wrist. The power of the ECU is usuallygreater than that of the radial extensors of thewrist and flexor carpi ulnaris.

Supination contracture of the forearm is a verydisabling deformity. Owing to its presence, manycommon activities in daily life, such as dressing,eating and writing, require elbow flexion andabduction plus internal rotation of the shoulder.With this deformity the patient is not motivatedto use the hand and has a functional deficit outof proportion to the real muscular and handsensory conditions. This indicates the impor-tance of early correction of the deformity toimprove hand function before bone deformitiesand joint dislocations occur (Fig. 1a).

Associated with forearm supination and defor-mity of the wrist (dorsiflexion and ulnar devia-tion) the fingers and thumb often show greatweakness or paralysis, especially of their intrin-sic muscles. Usually, the metacarpophalangealjoints of the fingers are stiff in extension due tocontracture of their collateral ligaments. The

280 OBSTETRICAL PARALYSIS

Table 2 Classification (stages) of obstetrical supinationdeformities of the forearm

Stage I Flexible deformity (congruent DRU joint)(a) active triceps(b) paralytic triceps

Stage II Fixed deformity (contracture of theinterosseous membrane)1. Congruent DRU joint

(a) active triceps(b) paralytic triceps

2. Incongruent DRU joint (subluxation ordislocation)

3. Volar subluxation or dislocation of the distalulna and radial head

Figure 4

(A) X-ray in a supination deformitywith incurvation of the forearmbones and volar dislocation of thedistal ulnar head (arrow). Inter-osseous membrane retracted. (B)Very severe supination contracturewith volar dislocation of the radialhead and distal ulnar head (arrows).

b

a

Page 290: Brachial Plexus Injuries

thumb is usually paralysed and drawn in adduc-tion. A flexion contracture of the elbow isfrequently present due to the imbalance betweena paralytic triceps and an active biceps.

The forearm supination deformity presentsinto two stages: (1) flexible (posture in supina-tion) and (2) fixed (contracture). The latter can bemild (20°), moderate (20–60°) or severe (over60°). In severe long-standing cases, the rotationof the forearm can reach up to 180°, as in two ofour cases (Fig. 1B). Table 2 shows our clinicalclassification, which is closely related to theoperative procedures (Table 3).

Surgical treatment

Indications and timing

Surgical indications basically depend on thepathology of the interosseous membrane –flexible or retracted – and the condition of thedistal radio-ulnar (DRU) joint and the tricepsfunction. Other factors to be considered are: thedeformities of the shoulder and elbow and theremnant function of the hand. In general, asevere shoulder deformity should be correctedfirst. A mild or moderate flexion contracture ofthe elbow is not a surgical contraindication toforearm surgery. The supination deformity andthe dorsiflexed and ulnarly deviated wrist arecorrected next. The reconstructive surgery ofthe hand should be the final step. Surgicaltechniques depend on the pathology of theinvolved soft tissues, joints and muscular condi-tion. Surgery should be ideally indicated assoon as the supination deformity begins toincrease (stage I) (Fig. 1,A). Conservative treat-ment (manipulations and splint) cannot preventthe contracture of the interosseous membraneand the progressiveness of the deformity.Correction of the supination deformity usuallyincreases digital function.

Operative procedures

The author shall refer to the principal distaldeformities of the upper limit after obstetricalpalsy and to his prefered surgical procedures.

1. Forearm supination/wrist ulnardevication deformity

This is a complex deformity where two compo-nents are usually combined: (i) forearm supina-tion and (ii) wrist dorsiflexion with ulnardeviation.

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 281

Figure 5

(A) Re-routing of the biceps tendon. The tendon is split ina Z fashion (1). It is exposed through a volar elbow incision.In a fixed supination deformity the interosseousmembrane is initially released through a long longitudinalposterior incision. After this, a forceful manipulation inpronation is performed (Fig. 8f). This technique is indicatedin a flexible deformity with a congruent DRU joint with afunctioning triceps (stage I,a) and in a fix deformity with acongruent DRU joint and good triceps (stage II-1,a). (B) Thepart of the tendon attached to the radius is passed dorsallyaround the radius, from the ulnar to the radial side, deepto the supinator (1). With the forearm in 30° of pronationand with the elbow in 80° of flexion the two ends of thetendons are sutured side to side. A Kirschner wire canmaintain the forearm in pronation during 5 weeks. A longposterior splint with the elbow in 90° is applied during thepostoperative period.

A

B

Page 291: Brachial Plexus Injuries

(i) Forearm supination componentDifferent techniques to correct forearm supina-tion deformity have been published. Blount(1940) introduced closed osteoclasis of themiddle third of both bones of the forearm.Zaoussis (1963) indicated osteotomy of the proxi-mal end of the radius. In 1967 we presented theresults of treatment of 14 patients by release ofthe interosseous membrane and re-routing of thebiceps tendon to obtain pronation (Zancolli II1967). Release of the interosseous membranewas initially described by Putti (1940) and the re-routing of biceps by Grilli (1959). We havecombined both procedures (Zancolli II 1967).Owings et al (1971) have presented their experi-ence with biceps re-routing. Our surgical indica-tions in recent years have been based in theclassification presented in the Table 2. In ourexperience osteoclasis and osteotomies of theforearm bones will usually recur in time due tothe persistence of the pronosupination muscularimbalance.

In the case of a flexible supination deformity(congruent DRU joint without contracture of theinterosseous membrane) and with a good triceps

(stage I,a) we re-route the biceps tendon toproduce pronation (Figs 5, 7). The transfer isfixed with the forearm in 30° of pronation andwith the elbow in 80° of flexion. The positionobtained can be provisionally secured with onKirschner wire between both forearm bonesduring 5 weeks (Table 3). Our results have beensatisfactory: patients obtain some degree ofpronation (Fig. 7).

Biceps tendon re-routing is contraindicated inflexible deformities with triceps paralysis (stageI,b). Under this condition the procedure willproduce a postoperative flexion contracture ofthe elbow. In these cases, the flexor carpi ulnaristendon – when active – is transferred deep to theflexor tendons of the forearm and the mediannerve and fixed to the distal radius at thebrachioradialis tendon insertion. This techniquecan produce partial active pronation. If the exten-sor carpi ulnaris (ECU) tendon is paralysed thedeformity is corrected by a distal radio-ulnarfusion.

In fixed supination deformities (stage II) thesurgical indication vary in relation to the localpathology. In a fixed deformity with a congruentDRU joint and good triceps (stage II–1,a) the

282 OBSTETRICAL PARALYSIS

Figure 6

(A) Release of the interosseousmembrane (1) and reduction of the defor-mity, with distal excision of the ulnarhead and distal radio-ulnar fusion, fixedwith two screws (2). Bone grafts areadded (3). This technique can beindicated in the following conditions:flexible supination deformity without afunctioning triceps (stage I,b); fixeddeformity with a congruent DRU joint buta paralytic triceps (stage II-1,b); fixeddeformity with an incongruent and DRUjoint, with or without a functioningtriceps (stage II-2), and in associateddislocation of the radius head (stage II-3).(B) Transfer of the ECU tendon throughthe forearm interosseous space. Thetendon surrounds the radius volarly andis fixed to the extensor carpi radialislongus (ECRL) and brevis (ECRB)tendons. It is indicated in wrist dorsiflex-ion and ulnar deviation flexible deformi-ties and usually is associated with bicepsre-routing or distal radio-ulnar fusion.

a

b

Page 292: Brachial Plexus Injuries

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 283

Figure 7

(A) A case of a flexible supination deformity with a congruent DRU joint in a 11-year-old boy. Through a passive maneu-ver it is possible to pronate the forearm. Triceps was active. A transfer of the biceps to pronate the forearm was indicated.(B) Reinsertion of the entire biceps tendon into the dorsal side of the radius neck to obtain pronation (arrow). Thistechnique differs from that shown in Fig. 5. (C) A good position in pronation was obtained. (D) The patient reaches theface very easily.

a b

c d

Table 3 Surgical techniques most frequently indicated in forearm supination deformity. In all the cases a transfer of theECU tendon may be indicated to correct ulnar deviation of the wrist and to maintain active wrist extension (Fig. 6b)

Stage Indication

Flexible deformityStage I-a (good triceps) Biceps tendon re-routing to pronationStage I-b (bad triceps) Transfer of the FCU or distal radio-ulnar fusion

Fixed deformityStage II-1,a (good triceps) Excision of the interosseous membrane and re-routing of biceps to pronationStage II-1,b (bad triceps) Excision of the interosseous membrane with distal radio-ulnar fusionStage II-3 Excision of the radial head, release of the interosseous membrane and distal radio-ulnar

fusion

Page 293: Brachial Plexus Injuries

284 OBSTETRICAL PARALYSIS

Figure 8

(A) A case of a fixed supination deformity (1) without functioning triceps in a 12-year-old girl. Ulnar deviation of the wrist (2).(B) The distal ulnar was palpated (dislocated) on the volar aspect of the wrist (arrow). (C) Volar dislocation of the distal ulnain lateral view (arrow) (1). In anteroposterior view a separation is seen at the DRU joint (2). Note incurvation of the forearmbones. (D) Long dorsal incision to expose the ECU tendon – usually active – and the interosseous membrane (IM). (E) Ampleexcision of all the retracted fibers of the interosseous membrane. (F) Intraoperative passive maneuver to obtain maximumpronation after the release of the interosseous membrane. (G) Radio-ulnar fusion in the desired angle of pronation (20°).Bone grafts are added. Fixation with two screws (arrow). The ECU tendon is retracted radially. (H) Radio-ulnar fusion proxi-mally to the radius growing cartilage (arrow). The ulnar head has not been excised. (I) The ECU tendon is passed volarly tothe radius (arrow). R, distal radius. (J) The ECU tendon has been passed through the interosseous space (1) and emergesclose to the distal radius epiphysis (2). The ECRL and ECRB are exposed (3). (K) Final position of the forearm (20° prona-tion) and the wrist after the distal radio-ulnar fusion and transfer of the ECU tendon have been performed. (L) Fixation ofthe ECU tendon to the ECRL and ECRB tendons (arrows). (LL) Final result after five years from the operation. Active wristextension was possible up to 60°. (M) Wrist flexion (30°) through the effect of gravity (arrow). The flexor tendons of thewrist were paralyzed. (N) Wrist position and alignment were maintained in all functions of the upper limb.

a b

c d

e f

Page 294: Brachial Plexus Injuries

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 285

i

j

h

k

l

n

m

ll

g

Page 295: Brachial Plexus Injuries

deformity is corrected by an ample excision ofthe retracted interosseous membrane, followedby manipulation of the forearm towards prona-tion and re-routing of the biceps tendon toproduce pronation. The interosseous membraneis particularly released at the middle third of theforearm where strong retracted fibers arepresent (descending radio-ulnar fibers) (ZancolliII and Cozzi 1992) (Fig. 2). Poitevin and Valente(1999) has described in detail the different fibersof the interosseous membrane.

Eighteen biceps re-routing operations wereperformed in obstetrical palsy patients. Thepatients’ age at time of surgery ranged from 4 to12 years – mean 10.5 years. In 12 patients a releaseof the interosseous membrane was indicated inthe same surgical stage. None of the cases failedto correct the supination deformity. In three casesa flexion contracture of the elbow was produced.In these cases the triceps was paralyzed.

In a fixed deformity with congruent DRU jointand paralytic triceps) (stage II-1,b), and in a

286 OBSTETRICAL PARALYSIS

Figure 9

(A) Distal radio-ulnar fusion in a fixed supination deformity in a 4-year-old girl. Synostosis is seen performed proximally tothe radius growing cartilage (arrow). (B) Nine years after the operation the fusion moves proximally. The initial correctionof the forearm was maintained. The growing cartilage was not affected.

a b

Page 296: Brachial Plexus Injuries

flexible deformity I,b, biceps transfer to obtainmuscular rebalance should not be attempted,otherwise a postoperative flexion contracture ofthe elbow is likely to result.

In our experience, the cases of fixed deformitywith congruent DRU joint and paralytic triceps(stage II-1, b), or with subluxated or dislocatedDRU joint, with or without triceps (stage II-2), thebest indication consists of:

1. ample excision of the interosseous membranethrough long dorso-ulnar incision, followingthe ECU tendon;

2. release of the dorsal capsule and ligaments ofthe DRU joint;

3. manipulation of the forearm to obtainmaximum pronation;

4. excision of the distal end of the ulna;5. radio-ulnar fusion proximal to the growth

plate of the radius (Table 3) (Figs 6, 8).

The synostosis produced between the ulna andradius is fixed with two screws or crossedKirschner wires. The forearm is fixed in 20° ofpronation to compensate through gravity thelack of wrist flexors tendons when the arm is atthe side of the body. Bone graft is added fromthe resected distal ulna.

We have indicated this technique in 32patients between 1972 and 1998. Forearm

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 287

Figure 10

In the case of this 10-year-old girl, the growing cartilage ofthe radius was injured during a distal radio-ulnar surgicalsynostosis. Two years later the cartilage was seen blockedinto its ulnar side.

Figure 11

Distal radio-ulnar fusion (1) is combined to wrist fusion (2)in a 16-year-old girl with a severe supination deformity. Theinterosseous membrane was excised. Incurvation of bothbones of the forearm is seen. (This postoperative X-ray isof the patient in Fig. 1B.)

Page 297: Brachial Plexus Injuries

supination deformity was corrected in all thecases. We have not seen late deformation of thedistal radius or its growing cartilage (Figs 8,H;9), except in one case where this cartilage wasviolated during surgery (Fig. 10). This case

involved a fall, 10 years after the surgical proce-dure, where a fracture at the bone fusion wasproduced (at the screws level). Associated witha re-routing of biceps or to a radio-ulnar fusion,a transfer of the ECU tendon can be indicated

288 OBSTETRICAL PARALYSIS

Figure 12

(A) Pronation deformity in a whole-type with partial recovery of an obstetrical palsy sequela, in a 20-year-old patient. Thefingers were in a claw deformity. (b) Incongruent DRU joint and radius in curvation (anteroposterior view). (C) Dorsal dislo-cation of the distal ulna (lateral view). (D) Dislocation of the proximal end of the ulna. In this case a distal radio-ulnarfusion was performed after an ample excision of the interosseous membrane.

a b

c d

Page 298: Brachial Plexus Injuries

to correct ulnar deviation of the wrist (Fig 6, 8)(see below).

In very severe hand paralysis, after thegrowing period, the distal synovitis of theforearm can be combined with the wrist fusion(Fig. 11).

In a supination contracture with volar disloca-tion of the distal ulna and radial head (stage II-3) the supination contracture is corrected by theexcision of the interosseous membrane andexcision of the radial head followed by a distalradio-ulnar fusion, as described above.

(ii). Wrist dorsiflexion with ulnardeviation component

As mentioned, is usually orsiflexion position ofthe wrist with ulnar deviation can be associatedwith forearm supination (27 per cent) (Zancolli IIand Zancolli III 1988). In the wrist deformity thereexists usually a muscular imbalance where thegreat majority of the muscles that pass throughthe radial aspect of the wrist are paralysed(thumb extensors, extensor carpi radialis brevisand longus aNd flexor carpi radialis). Frequentlythe only active wrist muscle is the extensor carpiulnaris.

The ulnar deviation is initially reducible andbasically produced by the effect of gravity andthe traction of an unopposed active ECU tendon.There are three possible treatments for thisdeformity: (1) a brace, (2) wrist arthrodesis afterthe growing period, or (3) tendon transfers. Weprefer tendon transfers when treating a childduring the growing period. As we havementioned before this technique is ideallyperformed simultaneously with the correction ofthe supination contracture (biceps transfer ofradio-ulnar fusion).

The ECU tendon is passed through the excisedinterosseous membrane towards the volarcompartment of the forearm and then re-routedaround the volar aspect of the radius (passingdeep to the radial vascular bundle) and suturedto the extensor carpi radialis longus and breviswith a tension to maintain the wrist in 20° ofdorsiflexion (Fig. 8). The route given to thetendon transfer is very effective to maintain thewrist in the axis of the radius.

This technique gives excellent results incorrecting ulnar deviation and restores a goodwrist dorsiflexion. It prevents a future fixed ulnardeviation of the wrist (Fig. 8) and eliminates theuse of wrist splints in a growing child. Thepreservation of wrist extension through the activ-ity of the transferred ECU tendon and wristflexion by the gravity effect can be used to createtenodesic effects over the fingers and a thumb invery severe digital paralysis.

2. Forearm pronation deformity

Pronation contracture deformity of the forearm, asmentioned before, can be present in distal seque-lae of obstetrical palsy. In this deformity the distalulna dislocates opposite to the supination defor-mity, that is dorsally. The interosseous membraneretracts very easily. Our proposed surgical treat-ment when the DRU joint is incongruent, sublux-ated or dislocated, consists of an ample excisionof the interosseous membrane, tenotomy of theretracted pronator muscles and distal radio-ulnarfusion in position (Fig. 12). Tendon transfers orwrist fusion can be associated.

3. Hand sequelae

The clinical presentation of hand function isvariable depending on the degree of spontaneousrecovery. Hand sequelae are difficult to classify,and consequently it is difficult to standardizesurgical procedures. If tendons are available fortransfers, hand function may be reconstructed.Generally muscles to transfer are weak andreconstruction of hand function is a challenge.

In most of the cases the surgeon has only oneor two wrist motors available for transfers andthese are used only after a wrist arthrodesis isperformed (after the growing period).

An attempt was made to classify the differenttypes of hand sequelae (Zancolli II and ZancolliIII 1988, 2000).

The usual patterns of paralysis we have see are:

• radial-type (complete or dissociated) (30 percent),

• median and ulnar-type (8 per cent),

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 289

Page 299: Brachial Plexus Injuries

• diffuse paresis (13 per cent),• total paralysis (21 per cent),• irregular distribution of paralysis (28 per cent).

The dissociated radial paralysis presents whenthere is a good wrist extensor and bad fingerextensors, or vice versa. The diffuse paretic handhas only a few degrees of digital motion inflexion and extension (none of these muscles canbe transferred). Tendon transfers employ onlymuscles of value 4 or 5. Total paralysis (21 percent) is compatible with root avulsion of C7, C8,and T1. In these cases the combination of a distalradio-ulnar fusion to a wrist fusion can producea good assistant limb (Fig. 11).

Conclusion

We have presented our preferred palliative surgi-cal procedures to correct residual distal deformi-ties of the upper limb after obstetrical palsy. Wehave studied, in particular, the supination defor-mity of the forearm usually associated with wristdorsiflexion and ulnar deviation. The distal resid-ual sequelae have been classified. This classifica-tion permits selection of the most appropriatereconstructive procedures. The surgical proce-dures here proposed have passed the test of time.Our aim has been to communicate our experienceof this challenging and severe pathology.

References

Blount WP (1940) Osteoclasis for supination deformi-ties in children, J Bone Joint Surg 22:300–14.

Duchenne GBA (1872) L’electrization localisé et de sonapplication à la pathologie therapeutique, 3rd edn.Baillière: Paris: 337–82.

Erb WH (1874) Ueber eine Eigenthumliche Localizationvon Lahmungen in Plexus Brachialis. Verhandl DNaturhist, Medisch Verlag: Heidelberg: NF, 2:130–61.

Gilbert A, Romana C, Ayatti R (1988) Tendon transfersfor shoulder paralysis in children, Hand Clinics4(4):633–42.

Gwilli FP (1959) Il trapianto del bicipite brachali infunzione pronatoria

Archivo ‘Putti’ 12:359–71.

Kuestner O (1885) Vaber epiphys are Diaphysen frakturam Humerusende des Neugeborenen, Arch F Klin Chir31:310–15.

Klumpke A (1885) Contribution a L’étude des paralysisradiculaires du plexus brachial. Paralysis radiculairestotales. Paralysis radiculaires inferieures. De la partici-pation des lilets sympathiques ocul-pupillaires dansces paralysis, Rev Med 5:591–616.

Owings R, Wickstrom J, Perry J, Nickel V (1971) Bicepsbrachii rerouting in treatment of paralytic supinationcontracture of the forearm, J Bone Joint Surg53A(1):137–42.

Poitevin L, Valente S (1999) La membrana interosea yla estabilidad longitudinal del antebrazo, Revista de laAsociación Argentina de Ortopedia y Traumatología64(3):220–8.

Putti V (1932) Análisis della triada radiosintomaticadegli stati di prelussazione, Chir Org di Mov 17:453–9.

Scaglietti O (1938) The obstetrical shoulder trauma,Surg Gynecol Obstetr 65:868–77.

Steindler A (1923) Congenital deformities of the shoul-der. Birth Palsy. In: Reconstructive Surgery of theUpper Extremity, Ch II. 12–21. Appleton Century: NewYork.

Suzuki SH, Yamamuro I, Fujita A (1984) The aetiologi-cal relationship between congenital torticolis andobstetrical parálisis, Int Orthop 8:175–81.

Zancolli EII (1967) Paralytic supination contracture ofthe forearm, J Bone Joint Surg 49A:1275.

Zancolli EII, Aponte F, Zancolli EIII (1979a) Parálisisobstétrica. Clasificación de las secuelas, Bol TrabajosSoc Arg Ort Traum. 44(3), 320–35.

Zancolli EII, Aponte F, Zancolli EIII (1979b) Parálisisobstétrica. de tipo braquial superior. Clasificación desus secuelas y su corrección quirúrgica, Bol TrabajosSoc Arg Ort Traum. 44(5), 288–306.

Zancolli EII, Aponte F, Zancolli EIII (1979c) Parálisisobstétrica. del miembro superior. Secuelas de tipoparalítico total y global a predominio inferior, BolTrabajos Soc Arg Ort Traum. 44(6), 417–430.

Zancolli EII (1981) Classification and management ofthe shoulder in birth palsy, Orthop Clin North Am12(2):433–57.

290 OBSTETRICAL PARALYSIS

Page 300: Brachial Plexus Injuries

Zancolli EII, Zancolli EIII (1988) Palliative surgical proce-dures in sequelae of obstetrical palsy, Hand Clinics4(4):643–69.

Zancolli EII, Cozzi E (1992) Atlas of Surgical Anatomyof the Hand. Churchill Livingstone: Edinburgh.

Zancolli EII, Zancolli EIII (2000) Reconstructive Surgeryin Brachial Plexus Sequelae. In the Growing Hand.Mosby: St Louis: 805–23.

Zaoussis AL (1963) Osteotomy of the proximal end ofthe radius for paralytic supination deformity inchildren, J Bone Joint Surg 45B(3):523–7.

PALLIATIVE SURGERY: PROSUPINATION IN OBSTETRICAL PALSY 291

Page 301: Brachial Plexus Injuries
Page 302: Brachial Plexus Injuries

Introduction

Clarke and Curtis (1995) classify birth palsy intofour categories according to anatomy: (1) Erb’spalsy, a palsy involving C5, C6 and sometimesC7, also called upper plexus palsy, which causesdeformities mostly of the shoulder and elbow; (2)Intermediate plexus palsy involving C7 andsometimes C8, T1; (3) Klumpke’s palsy, a palsyinvolving C8, T1, and (4) total plexus palsy,involving C5–C8 and sometimes T1. The lastthree types cause deformities mostly of forearmand hand. In their 3508 cases of obstetricalbrachial plexus palsy (OBPP) evaluation, only 20cases (0.6 per cent) were Klumpke’s palsy.

Gilbert (Gilbert et al 1991), evaluating 1000babies with birth palsy at 48 hours after birth,found that two types predominated in thispopulation: paralysis in the upper roots; andcomplete paralysis. Isolated involvement of thedistal roots, Klumpke’s palsy, was not seen.Chuang et al (1998a) have classified OBPP intotwo types: (1) initial OBPP (or Infant OBPP, I-OBPP), some of whom required early nervesurgery; and (2) late OBPP with deformity, calledsequelae OBPP (S-OBPP, or child OBPP), manyof whom require surgical correction. As a resultof evaluation of 121 S-OBPP children, Chuangcreated a ‘score of 10’ system for functionalevaluation. ‘Erb’s score of 10’ gives 10 points forupper plexus functions including shoulderabduction, shoulder external rotation, elbowflexion, elbow extension, forearm supination andpronation, and trumpet’s sign. ‘Klumpke’s scoreof 10’ gives 10 points for lower plexus functionsincluding wrist extension, wrist flexion, metacar-pophalangeal (MP) joint extension, interpha-langeal (IP) joint extension, finger flexion, thumb

adduction and thumb abduction. The author hasfound that hands in patients with S-OBPP fortu-nately tend to be functional. No one was foundwith complete hand palsy; some C8 or T1function still remained (ie incomplete avulsion);or some C8–T1 function from the aberrant re-innervation from the upper plexus was found (ieavulsed C8–T1 is possibly connected to thedisrupted upper trunk).

Forearm and hand reconstruction in S-OBPPwith a poor Klumpke’s score presents more diffi-culty in reconstruction than the shoulder orelbow. Few papers in the literature havediscussed in detail forearm and hand recon-struction caused by OBPP (Zancolli 1967, 1981,1988, Adler 1967, Doi 1996, Hoffer 1991, Lamb1980). In this chapter, we present a retrospectivereview of our S-OBPP patients who had under-gone palliative surgery for sequelae deformitiesof the forearm and hand.

Materials and methods

We treated sequelae deformities of forearm andhand in S-OBPP in nearly 100 cases between1988 and 1997 (a nine-year period). Of these, 54were followed-up more than two years. They allhad deformities arising from spontaneous recov-ery without any prior nerve surgical intervention.There were 35 male patients and 19 female. Thepatients’ average age based on the first time ofsurgical correction for deformity of forearmand/or hand was 10 years old, ranging from 3.5to 21 years. The average age for shoulder orelbow reconstruction is 4–5 years old (Chuang etal, 1998b). Many of them might have undergone

28Palliative surgery:forearm and hand deformitiesDavid C-C Chuang

Page 303: Brachial Plexus Injuries

294 OBSTETRICAL PARALYSIS

Figure 1

(A,B) A 16-year-old girl of late obstetrical brachial plexus palsy with severe supination contracture of the left forearm andsensory disturbance of the hand. (C,D) After re-routing the biceps for forearm pronation in addition to separation of the

a b

c d

Page 304: Brachial Plexus Injuries

shoulder or elbow correction before. Theaffected limbs were 36 in the right and 18 in leftupper limb.

The common sequelae deformities in theforearm and hand in S-OBPP include: (1) forearmsupination contracture (Fig. 1a,b), or forearmpronation contracture, the former being of higherincidence than the latter; (2) weak or absent wristextension, MP joint extension and/or IP jointextension (Fig. 2a); (3) weak or absent fingerflexion (Fig. 3a); (4) wrist ulnar deviation; (5)thumb instability; and (6) sensory disturbance,especialy in ulnar 2 fingers.

The main surgical procedures performed forforearm and hand deformities in S-OBPP included:forearm flexor tendons transfer for extension ofwrist, MP joint or IP joint or thumb (26 patients);re-routing of the tendon for forearm pronation(treatment of forearm supination contracture) (16patients); re-routing of the tendon for radial devia-tion of the wrist (treatment of wrist ulnar devia-tion) (11 patients); functioning free muscle

transplantation for extensor digitorum communis(EDC) and extensor pollicis longus (EPL), using themusculocutaneous nerve as a transformator (fivepatients); functioning free muscle transplantationfor flexor digitorum profundus (FDP) (four patients);opponensplasty (six patients); plication of extensortendon (six patients); distal advancement of centralextensor mechanism for IP joint extension (fivepatients); wrist extensor tendon transfer for fingerflexion (three patients); for MP joint extension (onepatient); for lumbricales replacement (one patient);for MP joint flexion (A1 Lasso procedure) (onepatient); bone fusion including wrist fusion, orthumb fusion (three patients); reduction of volardislocation of the radial head (three patients) (seeTable 1).

Multiples of the above surgical procedures forthe deformity of forearm and hand were done onthe same patient. Many of them might had thereconstructive procedures for the shoulderand/or elbow at the same time as forearm andhand reconstruction.

PALLIATIVE SURGERY: FOREARM AND HAND DEFORMITIES 295

a

b

Figure 2

(a) An 8-year-old boy of late OBPP with deficits of the wrist and finger exten-sion of the right upper limb. (b) At 5 months after multiple tendon transfer withpronator teres-to-ECRB, FCU-to-EDC and PL-to-EPL.

Page 305: Brachial Plexus Injuries

296 OBSTETRICAL PARALYSIS

c d

a

b

Figure 3

(a) A 6-year-old boy had left late-OBPP with weak finger flexion.There was no regional tendon available for transfer. After rightcontralateral C7 transfer with two sural nerve grafts (b), and agracilis functioning free muscle transplantation for FDP replacementin a two stage procedure, he achieved finger flexion and can nowhold an object with the help of an IP-extension dynamic splint (c,d).

Page 306: Brachial Plexus Injuries

Surgical techniques

A. Correcting MP joint drop (and/orwrist drop)

When forearm flexors are powerful

Powerful wrist or finger flexors and/or pronatorteres are very important. Those muscles can beutilized for transfer for finger and/or thumbextension. Most late OBPP patients have more orless wrist extension, but drop the MP jointbecause of poor muscle strength of the extensordigitorum communis (EDC), associated withweak extensor pollicis longus (EPL) or weakabductor pollicis longus (APL). The incidence offlexor-to-extensor transfer was quite high in ourseries, 26/54 patients (48 per cent). The tradi-tional technique of tendon transfer for radialnerve palsy (Tsuge 1989a, Schneider 1991, Green1993, Wheeler 1996) can be applied here. Pre-operatively, careful examination of each forearmvolar muscle is critically important for a success-ful transfer. At least M4 muscle strength (againstresistance to 3 kg) is available for transfer. Theroute of tendon transfer should also consider thedeformity of wrist deviation. For example, inwrist ulnar deviation with weak wrist extension,FCU (flexor carpi ulnaris)-to-ECRL (extensor carpiradialis longus) transfer around the radial bonewill be more effective than passing through theinterosseous membrane. It not only correctswrist extension but also corrects wrist ulnardeviation. FCU-to-APL around the ulnar bone willbe better than PL (palmaris longus)-to-APL,which not only benefits thumb abduction butalso corrects wrist ulnar deviation.

When forearm flexors are not powerful

There were four methods utilized individually forcorrecting MP joint drop (ie for EDC replacement):

1. Utilizing the musculocutaneous nerve as atransformator. The musculocutaneous nervewas found and isolated between biceps andcoracobrachialis muscles in the upper armand axilla. It was cut at the distal third of itslength. The distal stump was re-innervated bydirect coaptation to the deep central branchof the three intercostal nerves (usually T3–T5)

(Chuang et al 1996). The proximal stump wasthen used as a recipient motor source toinnervate a fresh free muscle for EDC replace-ment. Gracilis myocutaneous functioning freemuscle transplantation was always the firstchoice. The proximal end of the transferredmuscle was fixed to the upper third of thehumerus, a space between the deltoid inser-tion posteriorly and biceps muscle anteriorly.The distal end was sutured to the EDC tendonby the end-to-side weaving method. Apostoperative splint, allowing elbow flexionand wrist plus finger extension was appliedfor six weeks. Electrical stimulation of themuscle could start at three weeks postopera-tively;

2. ECRL-to-EDC transfer. In pure but rareKlumpke’s palsy, finger flexion and MP jointextension are quite commonly seen as absentor weak. Only one wrist flexor, FCR, and onewrist extensor, ECRL, are preserved. ECRL-to-EDC transfer (a concept of one muscle for twofunctions) was quite effective. The EDC’s 4tendons were moved out of (or superficial to)the dorsal transverse carpal ligament andsutured together. The ECRL was detached fromits insertion and sutured to the EDC by the end-to-side weaving method. Postoperative splint-ing for 6 weeks was required;

3. Plication of EDC (and/or APL). We had 6patients treated by this when there was noavailable tendon for transfer. We preferredshortening by segmental resection andtendon repair, instead of side-to-side plicationwhich could easily become loose again. Inside-to-side tendon plication, there is not cuttendon surface allowing tendons to adheretogether;

4. Wrist fusion plus correcting drop in the MPjoint. Three older S-OBPP patients (more than15 years old) were treated by wrist fusion.Simultaneous tendon transfer with FCR-to-EDC (2 patients) or plication of EDC (1 patient)was performed.

B. Correcting finger flexion

In Klumpke’s palsy or total plexus palsy,weakness or partial absence of finger flexion iscommonly noted. When powerful wrist extensor

PALLIATIVE SURGERY: FOREARM AND HAND DEFORMITIES 297

Page 307: Brachial Plexus Injuries

are available, traditional extensor-to-flexortransfer is used as in treatment of high medianand/or ulnar nerve palsy (Omer 1993; Burkhalter1993; Tsuge 1989a). However, incidence of thistransfer is low (possibly in only three of ourpatients) because of a high frequency of wristextensor weakness. In cases where wrist exten-sor donors are unabilable, functioning freemuscle transplantation for FDP replacement ispreferred, either using intercostal donor nervesas a donor nerve (two patients), or contralateralC7 transfer followed by a free muscle trans-plantation in a two stage procedure (1 patient,Fig. 3c).

C. Opponensplasty

The use of forearm flexor transfer (four patients)is more common than extensor transfer (twopatients) for thumb opponensplasty.

D. Correcting forearm supinationcontracture

Forearm supination contracture varied in severityfor S-OBPP. In less severe cases (interosseousmembrane is not yet fixed), re-routing the biceps

298 OBSTETRICAL PARALYSIS

Table 1 Operative procedures for deformities of forearm and hand in 54 late obstetric brachial plexus palsy

Purpose Procedure No. Incidence*patients

Correcting MP joint and/or wrist dropWhen forearm flexors are powerful Flexor-to-extensor transfer 26 48%When forearm flexors are not powerful Plication of EDC, APL or ECR 6

ECRL-to-EDC (one muscle for 2 functions) 1Wrist fusion + tendon transfer or plication 3

28%

Utilizing MCN as a transformator 5

Correcting finger flexion Extensor-to-flexor transfer 3FFMT for FDP (ICN innervation) 2 11%Contralateral C7 + FFMT 1(a two-stage procedure)

Thumb opposition (opponensplasty) Using flexor transfer 4Using extensor transfer 2

11%

Correcting forearm supination contracture Re-route biceps 16Re-route biceps + additional procedure 4

37%

Correcting forearm pronation contracture Re-route pronator teres 3 5%

Correcting IP joint extension Distal advancement of EDC to middle phalanx 5Using dynamic splint 3

15%

Correcting wrist ulnar deviation FCU-to-ECRL 2ECU-to-ECRL 3ECU-to-APL 2

20%

Route ECRL 4

Correcting proximal head dislocationVolar dislocation Annuloplasty by distal biceps tendon 1

By triceps tendon 1 5%Dorsal dislocation Wedge osteotomy 1

*A patient may have two or multiple procedures.MCN, musculocutaneous nerve; FFMT, functioning free muscle transplantation; MPJ, metacarpophalangeal joint; IP joint, interphalangealjoint; ECRL, extensor carpi radialis longus; APL, abductor pollicis longus; EDC, extensor digitorum communis; FDP, flexor digitorumprofundi; FCU, flexor carpi ulnaris; ECU, extensor carpi ulnaris.

��

Page 308: Brachial Plexus Injuries

as described by Zancolli (1988) is often enough forforearm pronation, as in 12 patients of our seriesrequiring this treatment (16/54, 30 per cent).However, for severe supination contracture (90° ormore, with a fixed interosseous membrane)additional procedures, such as detaching or re-routing the supinator to become a pronator (atechnique similar to the Zancolli’s procedure of re-routing the biceps), extensive separation ofinterosseous membrane including the distalradioulnar joint, and rotational osteotomy of theradius were also required for further enhance-ment of forearm pronation, in addition to re-routing of the biceps (four patients).

E. Correcting forearm pronationcontracture

This was rarely necessary, and only performedwhen a patient was keen to acquire some supina-tion for hand-to-mouth movement. Re-routingthe pronator teres to simulate a supinator mayhelp in enabling some hand-to-mouth movement(three patients).

F. Correcting IP joint extension

Intrinsic palsy of the hand in Klumpke’s or totalplexus palsy is one of the biggest challenges tothe reconstructive surgeon. Some techniqueswere utilized:

1. Distal advancement of the central limb of theextensor mechanism to the base of the middlephalanx (five patients). This is only indicatedin patients with positive MP joint extensionbut deficits in proximal interphalangeal (PIP)joint extension. Loosening of the centralattachment of the EDC to the middle phalanxdue to prolonged deficit of PIP joint extensionis the main cause. Distal advancement of thecentral extensor mechanism to the base ofmiddle phalanx is a feasible procedure for PIPjoint extension (six patients). The surgicaltechnique is similar to the correction of aBoutonniere deformity (Tsuge 1989b);

2. Lumbricales reconstruction by FDS transfer(one patient), or by ECRL transfer with tendongraft (one patient);

3. Using IP-extension dynamic splint (3 patients,Fig. 3b,c).

G. Correcting wrist ulnar deviation

Wrist ulnar deviation is quite commonly seen inS-OBPP, caused by muscle imbalance (10/54patients, 18.5 per cent). Too strong FCU, ECU orEDC is the main cause. Surgical correction,mainly for cosmesis, included releasing wristulnar contracture, FCU transfer to the ECRL (2patients), or ECU transfer to the ECRL (threepatients), or ECU to the APL (two patients); or re-routing ECRL or APL. (After division of the ECRLor APL tendon, the proximal stump goes aroundthe radial bone and sutures back to the distalstump: four patients.)

H. Correcting a proximal radialhead dislocation

1. In volar dislocation. A long sling includingpart of the muscle was elevated from thedistal biceps tendon (one patient) or from thetriceps tendon (one patient), based on thedistal insertion. The sling wraps around theradial neck and pulls the neck posteriorly. Itwas fixed to the posterior distal tricepstendon under tension. A temporary K-wirewas fixed from radius to the ulna bone forfour weeks of immobilization;

2. In dorsal dislocation. One patient receivedwedge osteotomy of the ulnar side of theradius to reduce the radial head into thenormal position where it was fixed with aplate and screws. For complete dislocation ofradius and ulna at the elbow, no surgery isadvocated, as it will possibly cause elbowstiffness.

Results

The existence of powerful forearm flexors,especially wrist flexors, is very useful for recon-struction, and their transfer usually leads to goodresults (Fig. 3b). The success of FCR-to-EDC

PALLIATIVE SURGERY: FOREARM AND HAND DEFORMITIES 299

Page 309: Brachial Plexus Injuries

transfer is higher than FCU transfer (11/12, 92 percent vs. 5/8, 62 per cent). Success of PL (palmarislongus) transfer to EPL (extensor pollicis longus)was around 70 per cent; and pronator teres (PT)transfer, usually to the ECRB, was 60 per cent.But FDS, usually the ring finger (fourth FDS),transfer for digits or thumb extension has moreunpredictable results. Extensor tendon transferby using BR (brachioradialis), ECRL, or ECU forfinger flexion or thumb opposition usually gaveunreliable and poor results.

Re-routing the biceps for forearm pronation isquite useful. It increases forearm pronation fromaverage 24.5° (range, 0–80°) up to an average of77.5° (range, 0–90°). But if the supination defor-mity is very severe (more than 90°), the proce-dure is not enough. Detachment of the supinatorfrom the radius, separation of interosseousmembrane, release of distal radio-ulnar joint,and rotational osteotomy of the radius or thehumerus will be additionally required to get agood result (Fig. 1c,d).

Plication of the extensor tendon (ECR, APL orEDC) was only effective in half of the cases (6/12)and worked well only temporarily, ending inrecurrence due to getting loose. Tenodesis ofextensor tendon, EDC, to the radius in one casecaused wrist and finger extension contracturedue to growth, which eventually requiredrelease.

Distal advancement of the central extensormechanism was an effective method to improveIP joint extension in six cases. Lumbricalesreconstruction with ECRL and tendon graft in onepatient, and with fourth FDS in another onepatient were occasionally performed. Bothpatients showed also improvement of IP jointextension.

Using functioning free muscle transplantationto improve hand function is a more advancedand complex surgical procedure. We had fivepatients with the musculocutaneous nerve as atransformator. All five patients achievedimproved EDC function by the free muscle trans-plantation. Recovery of elbow flexion, innervatedby intercostal nerves, was fortunately as good asbefore. However, using an intercostal nerve toinnervate a free muscle transplantation for fingerflexion did not produce a good result in either oftwo patients. The muscle strength was only M2.

One child had contralateral C7 elongation withtwo sural nerve grafts which were embedded

into the biceps muscle. One year later, a gracilisfree muscle transplantation for FDP replacementwas performed at the second stage. He finallyachieved M3–M4 finger flexion strength.However, he still needs a dynamic splint for thefinger IP extension (Fig. 3c,d).

Discussion

Leffert (1985) wrote a good review about thehistory, pathogenesis, clinical presentation,musculoskeletal changes and prognosis ofobstetrical brachial plexus palsy (OBPP),although he misused the term ‘congenitalbrachial palsy’. A wide divergence of opinion asto the best treatment remains. He concluded thatthe treatment of the paralytic hand due to OBPPmust be highly individualized. Hentz (1991) hasmentioned that palliative treatment of the seque-lae of birth palsies is difficult, and the results arerarely totally satisfactory.

Questions related to the management ofsequelae deformities of the forearm and hand oflate OBPP include: (1) what is the best age forthe later management or palliative reconstruc-tion? (2) in triple deformities of shoulder, elbowand hand, what is the first priority for recon-struction? (3) is traditional tendon transfer orbone management enough for those deformi-ties? and (4) is there any difference betweenyounger and older late OBPP patients related topalliative reconstruction?

For reconstruction of the sequelae deformities inS-OBPP, most authors favor an age of 4 years ormore as optimal (Leffert 1985, Gilbert et al 1991,Chuang et al 1998b) because of no severe contrac-ture, active cooperation, easy clinical evaluation,better following order and performing the rehabil-itation program. Preoperative examination of eachfunctional muscle that is prepared for transfer iscritically important for a successful transfer.Advising the children that they should undertakecontinual muscle strength exercises before thetransfer takes place is also very important, whichusually requires parental supervision and encour-agement.

Shoulder and elbow muscles always recoverearlier and more maturely than the forearm andhands. Patients with Erb’s palsy are also morefrequent than those with Klumpke’s palsy. The

300 OBSTETRICAL PARALYSIS

Page 310: Brachial Plexus Injuries

muscles in the forearm or hand most showparesis (incomplete palsy with weak strength),paralysis (complete palsy with significantatrophy), or contracture due to muscle imbalanceor aberrant re-innervation. Aberrant re-innervationis less seen in the forearm and hand exceptsupinator muscles. The tendon transfer in theforearm and hand is no more based on therelationship between antagonist and agonist asseen as in the shoulder and elbow (due toaberrant re-innevation) (Chuang 1998). Therefore,although the treatment is highly individualized,shoulder and elbow reconstruction are usuallyperformed before forearm and hand. In ourexperience, the optimum time for forearm andhand reconstruction is at school age 6–13 yearsold, while for shoulder and elbow reconstructionthe optimum time is at pre-school age, 4–6 yearsold. There is usually a lack of powerful muscles inthe forearm and hand for transfer in the lateOBPP. Continuous physical therapy with rehabili-tation program to increase the muscle strength inthe forearm and hand are important for the latermanagement.

From our results evaluation, traditional tendontransfer is usually not enough (either number orstrength) for reconstruction of deformities offorearm and hand in S-OBPP. Powerful wrist andfinger flexors felt by clinical examination aremore reliable than extensors as potential candi-dates for transfer. A high failure rate wasencountered when using wrist extensor transfer,either for opponensplasty or for FDP reconstruc-tion. We usually required further complextechniques to overcome it, such as functioningfree muscle transplantation (Doi 1996).

Utilizing intercostal nerves as a transformatoris indeed a high-risk procedure, requiring anadvanced microneurovascular anastomosistechnique and surgical experience in intercostalnerve transfer and functioning free muscle trans-plantation. It is really beneficial for finger exten-sion when there is lack of powerful regionaltendons for transfer. Patients start to have abilityto extend their fingers towards targets, and theyreally feel that the hand is useful. The transferredmuscle for EDC replacement, innervated by themusculocutaneous nerve, is always strongerthan free muscle transplantation innervated byan intercostal nerve or spinal accessory nerve,because more muscle belly locates and functionsin the forearm. The other alternative for EDC

reconstruction is possibly only the dynamicsplint, which needs to be worn continuously andis inconvenient. Contralateral C7 transfer withlong sural nerve graft embedded into the bicepsmuscle, followed by a functioning free muscletransplantation as a two-stage procedure is ananother effective option for EDC reconstruction.

Tenodesis, such as ECR or EDC fixed to theradius, is absolutely no good as a procedure ingrowing children, as it ends in wrist or fingercontracture. Elongation of ECRL to EDC out ofdorsal transverse carpal ligament is an anothereffective option based on the concept of ‘onemuscle for two functions’ (Doi 1996). The ECRL,of course, needs to be strong enough.

Arthrodesis is only indicated in older children,more than 15 years old, or when there is nomore growing cartilage seen in the distal radiusor ulna by X-ray radiography.

Reference

Adler JB, Patterson RL (1967) Erb’s palsy. Long timeresults of treatment in eighty-eight cases, J Bone JointSurg 49A:1052–64.

Burkhalter W (1993) Median nerve palsy. In: Green D,ed. Operative Hand Surgery, 3rd edn. Churchill-Livingstone: New York: 1419–48.

Clarke HM, Curtis CG (1995) An approach to obstetri-cal brachial plexus injuries, Hand Clin 11:563–81.

Chuang DCC, Ma HS, Wei FC (1998a) A new evaluationsystem to predict the sequelae of late obstetric brachialplexus palsy, Plast Reconstr Surg 101:673–85.

Chuang DCC, Ma HS, Wei FC (1998b) A new strategyof muscle transposition for treatment of shoulderdeformity caused by obstetric brachial plexus palsy,Plast Reconstr Surg 101:686–94.

Chuang DCC, Carver N, Wei FC (1996) Results offunctioning free muscle transplantation for elbowflexion, J Hand Surg 21A:1071–7.

Doi K (1996) Obstetric and traumatic pediatric palsy. In:Peimer A, ed. Surgery of the Hand and UpperExtremity. McGraw-Hill: New York: 1443–63.

Doyle JR (1993) Extensor tendons – acute injuries. In:Green D, ed. Operative Hand Surgery, 3rd edn.Churchill-Livingstone: New York: 1941–5.

PALLIATIVE SURGERY: FOREARM AND HAND DEFORMITIES 301

Page 311: Brachial Plexus Injuries

Gilbert A, Razaboni R, Amar-Khodja S (1991) Surgicaltreatment of brachial plexus birth palsy, Clin Orthop264:39.

Green DP (1993) Radial nerve palsy. In: Green D, ed.Operative Hand Surgery, 3rd edn. Churchill-Livingstone: New York: 1401–17.

Hentz VR (1991) Operative repair of the brachial plexusin infants and children. In: Gelberman RH, ed.Operative Nerve Repair and Reconstruction. JBLippincott Company: Philadelphia: 1369–83.

Hoffer HM (1991) Assessment and natural history ofbrachial plexus injury in children. In: Gelberman RH,ed. Operative Nerve Repair and Reconstruction. JBLippincott Company: Philadelphia: 1361–8.

Lamb DW (1980) Tendon transfers for paralytic states.In: Barron JN, Saad MN, eds. The Hand. OperativePlastic and Reconstructive Surgery. Churchill-Livingstone: Edinburgh: 1163–77.

Leffert RD (1985) Brachial Plexus Injuries. Churchill-Livingstone: New York: 189–235.

Omer GE (1993) Combined nerve palsies. In: Green D,ed. Operative Hand Surgery, 3rd edn. Churchill-Livingstone: New York: 1401–17.

Schneider LH (1991) Tendon transfer for radial nervepalsy. In: Gelberman RH, ed. Operative Nerve Repairand Reconstruction. JB Lippincott Company:Philadelphia: 697–709.

Tsuge K (1989a) Comprehensive Atlas of HandSurgery. Year Book Medical Publishers: Chicago:498–501.

Tsuge K (1989b) Comprehensive Atlas of HandSurgery. Year Book Medical Publishers: Chicago:675–8.

Wheeler DR (1996) Reconstruction for radial nervepalsy. In: Peimer A, ed. Surgery of the Hand and UpperExtremity. McGraw-Hill: New York: 1363–79.

Zancolli EA (1967) Paralytic supination contracture ofthe forearm, J Bone Joint Surg 49A:1275.

Zancolli EA (1981) Classification and management ofthe shoulder in birth palsy, Orthop Clin North Am12(2):433–57.

Zancolli JER (1988) Palliative surgical procedures insequelae of obstetric palsy, Hand Clinics 4: 643–69.

302 OBSTETRICAL PARALYSIS

Page 312: Brachial Plexus Injuries

Introduction

Between 1981 and 1999, 1714 patients withbrachial plexus lesions were treated in our insti-tution. There were 418 obstetrical brachial plexuspatients. In order to optimize treatment and toenable a therapeutically relevant documentation,both as thorough as possible and internationallycomparable, we have developed the so-called‘Plexus evaluation system’ (PES) – an examina-tion and documentation procedure for obstetrical(Hierner et al 1999) and adult (Hierner et al 2000)brachial plexus lesions.

For the treatment of brachial plexus lesions weare using the so-called ‘integrated treatmentconcept’ (Berger et al 1991). Daily physical therapyis the basis of every treatment. Spontaneous orearly surgical nerve repair is given the highestpriority. Secondary procedures (tendon transfer,free functional muscle transplantation) alone or incombination with adjuvant procedures (joint

releases, botulinum toxin, capsulodesis, splints,arthrodesis, derotation osteotomies) often canimprove the overall result.

Epidemiology

Muscle co-contractions after nerve regenerationin proximal brachial plexus lesions, eitherspontaneously or postoperatively, occur in everypatient with a higher degree (SUNDERLAND III–V)nerve lesion.

Pathophysiology

Co-contractions after brachial plexus injury arethe result of an aberrant nerval re-innervation atthe site of lesion (or repair), thus anatomically

29Treatment of co-contractionRobert Hierner and Alfred C Berger

Figure 1

(a–c) Pathophysiology of muscleco-contraction after regeneration ofproximal brachial lesions. (FromSchliack and Memmenthaler 1987.)

a b c

Page 313: Brachial Plexus Injuries

defined (Mumenthaler et al 1998). Misled axonsafter spontaneous or postoperative nerve regen-eration in brachial plexus lesions occasionallycause severe muscle co-contractions (Roth 1983,Mumenthaler et al 1998) (Fig. 1).

Those severe co-contractions may not allowbasic movement functions like shoulder abduc-tion, shoulder external rotation, elbow flexion,etc and/or hamper coordinated movements ofthe whole extremity, like hand-to-mouth.Moreover, motor development is delayed andsecondary skeletal deformities will occur (Clarkeand Curtis 1995, Lone 1966, Gordon et al 1973,Tassin 1983, Greenwald et al 1984, Tada et al1984, Narakas 1986, Gilbert and Tassin 1987,Jackson and Hoffer 1988, Gilbert 1995, Sloof1995, Berger et al 1997, Rollnik et al 2000 Hentz1991, Michelow et al 1994).

Classification

For clinically relevant purposes muscle co-contractions can be classified according to theirlocalization and severity. Co-contractions mayoccur over the whole extremity, with the mostimportant being deltoid/teres major at the shoul-der level (Fig. 2a,b) and triceps/biceps co-contractions at the upper arm level (Fig. 3a,b).

With regard to severity of co-contraction, a classi-fication into mild, moderate, and severe turned outto be useful in clinical practice (Table 1).

Diagnosis

The leading symptom of muscular co-contractionis a contracting muscle without adequate motorfunction, although passive range-of-motion ofthe adjacent joint(s) is free or not significantlylimited. Further specific clinical examinationlooks for co-contraction of the antagonisticmuscle, either by palpation (moderate) or oninspection (severe). Further information can beobtained from an EMG (amplitude, traces). Assevere muscular co-contraction leads tosecondary deformities in the growing skeleton(joint deformities, length discrepancy), specialattention must be given to its early detection anddocumentation.

304 OBSTETRICAL PARALYSIS

Figure 2

Treatment of severe deltoid/teres major co-contractionwith repeated intramuscular injection of botulinum toxintype A: (a) pre-injection (‘trumpet sign’); (b) 24 monthspost-injection (shoulder abduction with hand to ceilingmovement ‘big man-test’).

a

b

Page 314: Brachial Plexus Injuries

Treatment

Specific options for the treatment of muscularco-contractions can be classified into non-opera-tive treatment and operative treatment. Non-surgical treatment options are physical therapy,biofeedback training, and intramuscular injec-tion of botulinum toxin type A. Surgical treat-ment has a palliative character and comprisestenotomy and tendon transfer (Table 3).

Non-operative treatment

Physical therapy

The specific aim of physical therapy in muscleco-contraction treatment is to strengthen theagonist in order to overcome the co-contractingantagonist. However, low patient compliance inchildren and persistent co-contractions oftenlead to disappointing results.

Biofeedback training

Biofeedback training has proved of good value inthe treatment of synkinesis after facial nerverepair. Preliminary results in post-traumatic adultbrachial plexus lesions are promising.

Botulinum toxin type A

Botulinum toxin type A is a potent drug in thetreatment of several diseases associated withincreased muscle tone or muscle overactivity,such as torticollis spasmodicus, blepharospasm,limb dystonia, facial hemispasm, and spasticity(Scott et al 1985, Bernius et al 1999). Spasticity

may reflect disabling muscle overactivity,including permanent activity in the absence ofany stretch or volitional command, inappropri-ate responses to cutaneous or vegetative input,and unwanted antagonistic co-contractions(Heinen et al 1999). Since botulinum is effectivein the treatment of the spasticity-associated co-contractions, it might be beneficial in postpara-lytic antagonistc movements after obstetricalbrachial plexus lesions (Hierner and Berger1998). Although there are no structural changesat the peripheral nerve, our results clearlyindicate that repeated intramuscular injectionsof botulinum toxin type A into the co-contract-ing antagonist result in a lasting reduction ofimbalance between antagonist and agonistactivity, with the possibility of adequate function(M4 power) (Alfonso et al 1998, Hierner andBerger 1998, Schubert et al 1998, Hierner et al1999). This functional change might beexplained by: (1) local interaction at the motorplate; (2) structural changes of the injectedmuscle and its antagonist; (3) changes at thefirst and second motor neuron.

Botulinum toxin type A is an extremely potentagent acting presynaptically by blocking therelease of acetylcholine at the neuromuscularjunction. The neurotoxin is actively taken up intothe nerve terminal. However, this ‘pharmacolog-ical denervation’ is temporary. Nevertheless, itshould be taken into account that no recurrenceof severe co-contractions has been observed upto now. These long-lasting effects cannot beexplained by a simple temporary palsy of the co-contracting antagonistic muscle. They might beexplained by effects at the muscle and spinalcord level and central effects like the so-called‘cortical plasticity’ (Fujimoto et al 1992).

At the skeletal muscle level botulinum toxininjection leads to a temporary muscle atrophy,showing variable fiber diameter and an increasing

TREATMENT OF CO-CONTRACTION 305

Table 1 Classification of muscular co-contractions after brachial plexus lesions

Co-contraction EMG Clinical examination Treatment

Mild + Not visible on clinical examination NoneModerate + Visible on clinical examination Physical therapy (biofeedback)

No significant functional disturbanceSevere + Visible on clinical examination Physical therapy biofeedback botulinum toxin

Significant functional disturbance Palliative surgery

Page 315: Brachial Plexus Injuries

number of type I fibers. Even after repeated injec-tions, no muscle fibrosis or persistent muscleatrophy like that after nerve damage could befound (Heinen et al 1999). From studying spasticpatients we have learned that agonist/antagonistco-contractions lead to functional impairment byreducing the joint excursion and the speed of themovement. Moreover, constantly increasedmuscle tone during growth leads primarily toreduction of muscle growth and secondarily tomuscles and joint contractures. This in itself leadsto secondary joint and skeletal deformities, thus avicious circle with further functional impairmentpersists (Bernius et al 1999). It is essential for thesuccess of botulinum toxin treatment that physi-cal therapy is intensified concurrently. As‘pharmacological denervation’ is temporary, theaim of physical therapy is: (1) strengthening of theagonist and stretching of the antagonist, whichleads to an increased total amplitude ofmovement (TAM), thus preventing secondaryjoint and/or skeletal deformities; (2) to learnand/or stabilize new simple (elbow flexion) orcomplex (hand-to-mouth) movements, and (3) toteach the patient to coordinate those movementswith weaker but still persistent co-contractions(Heinen et al 1999). Although experimental datashow that botulinum toxin is transported to thecentral nervous system by the retrograde axonaltransport mechanism, there is no clear evidencethat it has direct impact on the central nervoussystem. Its mode of action seems to be restrictedto the peripheral nerve system (Hambleton andMoore 1995).

Operative treatment

Palliative surgery is indicated if the non-operativetreatment fails or does not lead to a satisfactoryresult. In each and every case donor site morbid-ity (= loss of function of the co-contracting antag-onistic muscle) must be balanced against itsbenefit (increased active range-of-motion).

Tenotomy

The simplest palliative procedure is tenotomy ofthe co-contracting antagonistic muscle. However,by simple transsection of a tendon, movement

potential is lost. Therefore, simple tenotomy isonly indicated if no transfer is possible. This maybe the case of a co-contracting teres majormuscle in adult brachial plexus palsy.

Tendon transfer

If the co-contracting antagonistic muscle can beused to reinforce the action of the weak agonist,a tendon transfer is the treatment of choice.Without the possibility of shoulder abduction> 90° and lack of other possibilities to restoreelbow flexion, triceps transfer is a good optionin adults to restore the important elbow flexion,by giving up active elbow extension. Tricepstendon transfer (Carroll and Hill 1972, Marshallet al 1988) to the biceps is rarely indicated nchildren (Berger et al 1991, 1997) because ofpossible skeletal growth impairment by too earlya tendon transfer.

Personal experience and resultswith botulinum toxin type A

From 1997 to 2000, 14 patients with severe co-contraction after brachial plexus injury weretreated in cooperation with the neurologicaldepartment. There were eight children (age 2–4years) with severe triceps/biceps co-contraction,one child (age 8 years) with triceps/biceps co-contraction after spontaneous recovery andmodified Steindler-transfer, two children (age 4years) with severe deltoid/teres major co-contrac-tion, and two adults with severe triceps/biceps co-contractions after early microsurgical brachialplexus repair.

Six children (aged 2–4 years, mean 3.2, SD - 1.0)presenting severe biceps/triceps co-contractionsafter nerve regeneration (three cases with sponta-neous regeneration, three cases after early micro-surgical repair at the age < 6 months) were treatedby intramuscular injection of botulinum toxin typeA into the triceps. We used botulinum toxin A(DYSPORT, Ispen Pharma) with a concentration of 25Mouse Units (MU)/ml. An average of 39.2 MU(range: 25–50 MU, SD 8.0) was injected into thetriceps muscle at two sites. EMG with surfaceelectrodes (Viking II, Nicolet Instruments,Madison/Wisconsin, USA) was used to evaluate

306 OBSTETRICAL PARALYSIS

Page 316: Brachial Plexus Injuries

TREATMENT OF CO-CONTRACTION 307

Table 2 Personal experience and results after repeated intramuscular botulinum toxin type A injections for thetreatment of triceps/biceps co-contractions

Patient’s initials————— 1 2 3 4 5 6 Total/Criteria HM SJ BT SS SC NM mean

Age (years)

Nerveregenerationa

microsurgicalnerve repair

Active ROMelbow pre-Botox(Neutral-0-Method)

Elbow flexionpower pre-Botox (MRC)

Active ROMelbow post-Botox(Neutral-0-Method)

Elbow flexionpower post-Botox (MRC)18/24 months

Dose ofbotulinumtoxin (MouseUnits)

Duration oftricepsparalysis(weeks)

Number ofinjections

Complications

Recurrency

4

s

0–20–40°

M2 (hand-to-mouth)

18 months0–20–95°24 months0–10–100°

M4/M4(hand-to-mouth +)

40

36

2

no

3

s

0–30–50°

M2 (hand-to-mouth)

18 months0–30–100°24 months0–10–100°

M4/M4(hand-to-mouth +)

40

19

3

no

4

nr1⁄2XI-SSCneurolysis(C5–C7)

0–20–40°

M2 (hand-to-mouth)

18 months0–20–120°24 months0–10–120°

M4/M4(hand-to-mouth +)

40

44

2

no

2

nr2 1⁄2XI-SSCneurolysis 2 C5–(1x5) ax4 C6-(4x5)FasLat

0–30–40°

M1 (hand-to-mouth)

18 months0–30–110°24 months0–20–110°

M4/M4(hand-to-mouth +)

40

19

2

no

4

nr1⁄2XI-SSCneurolysis(C5–C7)

0–30–50°

M1 (hand-to-mouth)

18 months0–30–80°24 months0–10–95°

M2 +/M4(hand-to-mouth +)

50

16

2

no

2

s

0–20–50°

M2 (hand-to-mouth)

18 months0–20–100°24 months0–20–110°

M4/M4(hand-to-mouth +)

25

18

3

no

3.2

s =3nr = 3

0–25–50°

M1 = 2M2 = 4(hand-to-mouth 0/6)

18 months0–25–101°24 months0–13–106°

18 monthsM2+ = 1M4 = 5(hand-to-mouth 5/6)24 monthsM4 = 6(hand-to-mouth 6/6)

39

25.3

2.3

0/6

aS, spontaneous; nr, early microsurgical nerve repair. 21/2XI-SSC, direct neurotization of half spinal nerve onto suprascapular nerve; 3C5–1�5) ax, intraplexual neurotization from C5 root with 1 nerve graft of 5 cm length onto axillary nerve. 4 C6–(4�5), intraplexualneurotization from C6 root with 4 nerve grafts of 5 cm length onto lateral fascicle.

Page 317: Brachial Plexus Injuries

308 OBSTETRICAL PARALYSIS

Figure 3

Treatment of severe biceps/triceps co-contraction with repeated intramuscularinjection of botulinum toxin type A. (a)pre-injection; (b) 24 months post-injec-tion (elbow extension); (c) 24 monthspost-injection (elbow flexion with hand-to-mouth movement ‘cookie-test’).

a

b c

Page 318: Brachial Plexus Injuries

biceps/triceps co-contractions before and after thetreatment (Rossi et al 1982). All patients werefollowed-up at four-week intervals. Informationabout the onset of response and duration ofbotulinum toxin effects was obtained via theparents. Clinical testing (muscle power graded bythe British Medical Research Council classifica-tion) and measurement of the active range ofmotion using the Neutral-0-Method (Fig. 3a–c)were performed prior and after injections.Possible complications, like systemic or localmuscular paralysis (ptosis, strabismus, impair-ment of swolling, etc), and/or allergic responsesto the toxin therapy were checked. Moreover, theduration of the temporary ‘pharmacologicaldenervation’ (Erbguth and Claus 1996) and thenumber of injections were recorded (Table 2). Thepatients of all children had given written informedconsent to the treatment (Hierner et al 2001).

Onset of response following the first injectionwas seen after an average of 8.5 days (range: 4–14days; SD 4.4). Mean active elbow flexion prior toapplication was about 50° (range: 20–60°) andmuscle power was graded M1 (2 cases) to M2 (4cases). Hand-to-mouth movement was not possi-ble in any of the patients. At 18 months after injec-tion, mean elbow flexion was about 100° (range:80–120°) and muscle power was graded M2+ inone case and M4 in five cases. Five out of sixpatients were able to perform adequate hand-to-mouth movement. Two years after the last injec-tion elbow flexion power could be graded as M4in all patients and hand-to-mouth movement(‘cookie test’) was possible in all patients.Moreover, a clear reduction in extension lag of theelbow could be seen. On EMG examination a clearreduction of triceps contractions during bicepsactivity was observed. Temporary paralysis of thetriceps after injection persisted for 16–44 weekswith an average of 25.3 weeks. In order to achievea stable elbow flexion at the M4 level thebotulinum toxin Botox injection had to berepeated two or three times. The average time oftreatment took 8–21 months. There was no recur-rence of co-contraction in any of the patients atthe 18 month follow-up point. Mild-to-moderatediscomfort at the injection site was seen forseveral days after injection in two children. It hasto be pointed out that no severe adverse eventsoccurred, especially no bleeding, infection, exten-sive muscle weakness, or systemic complications(Table 2, Fig. 3a–c).

Discussion

There are specific differences in the treatment ofmuscle co-contraction in children and adults(Table 3). In babies and young children the aimsof treatment are: (1) treatment of co-contraction,and (2) treatment of secondary skeletal deformi-ties at their early stages. Programmed physicaltherapy and biofeedback training proved of lowvalue, because of low patient compliance. In thisage group the main aim is to provoke movementduring daily playing or daily living activities ofthe child. In contrast, based on personal experi-ence, repeated intramuscular injection ofbotulinum toxin type A proved of immense valuefor early treatment of co-contraction and thusprevention of secondary skeletal deformities.Moreover, when started at the age of 2–4 yearswith secondary skeletal deformities just begin-ning (ie elbow extension lag) it was surprising tosee that those early deformities disappearedafter treatment with growth. Tenotomies andtendon transfers are not indicated at this age(Table 3).

In older children a specific physical therapyprogram may be possible, as well as biofeedbacktraining. Again, botulinum toxin A is an impor-tant tool in combination with physical therapy. Inselected cases tendon transfers – especially atthe shoulder level – may be indicated.

In the adult patient, physical therapy andbiofeedback training proved of value. Botulinumtoxin is an adjunct to make physical therapymore easy or effective by blocking the co-contracting antagonistic muscle. If the non-operative treatment fails or does not lead to asatisfactory result, tenotomies, and especiallytransfers of the cut tendon to the agonisticmuscle, are the treatment of choice (Table 3).

TREATMENT OF CO-CONTRACTION 309

Table 3 Differential treatment ‘muscle co-contraction’ inobstetrical and adult brachial plexus lesions

Options Obstetrical Adultearly late

Physical therapy – + ++Biofeedback training – – ++Botulinum toxin type A ++ + +Tenotomy – – +Tendon transfer – + ++

Page 319: Brachial Plexus Injuries

References

Alfonso I, Papazian O, Grossman JA (1998) Clinicalpresentations, differential diagnosis and managementof obstetrical brachial plexus palsy, Rev Neurol27:258–63.

Berger A, Schaller E, Mailänder P (1991) Brachialplexus injuries: an integrated treatment concept, AnnPlast Surg 26:70–6.

Berger A, Hierner R, Becker MH-J (1997) SekundäreErsatzoperationen zur Wiederherstellung derEllenbogenbeugefunktion nach Läsionen des Plexusbrachialis, Orthopäde 26:643–50.

Berger A, Hierner R, Becker MH-J (1997) Die frühzeitigmikrochirurgischer Revision des Plexus brachialis beigeburtstraumatische Läsionen – Patientenauswahl undErgebnisse, Orthopäde 26:643–50.

Bernius P, Döderlein L, Siebel A et al (1999)Botulinumtoxin – Anwendung in der Orthopädie.Grundlagen, Indikationen, Therapieziele, Applikation,Verlaufskontrolle, unerwünschte Wirkungen, Therapie-kombinationen. Hoffmann-Verlag: Mainz.

Caroll RE, Hill RA (1972) Triceps transfer to restoreelbow flexion. A study of 15 patients with paralyticlesions and arthrogryposis, J Bone Joint Surg52A:239–44.

Clarke HM, Curtis CG (1995) An approach to ostetricalbrachial plexus lesions, Hand Clin 11:563–81.

Erbguth F, Claus D (1996) Therapie von Erkrankungenmit erhöhtem Muskeltonus durch Injektion vonBotulinumtoxin A, Psycho 22:810–16.

Fujimoto S, Yokochi K, Togari H et al (1992) Neonatalcerebral infarction: symptoms, CT findings and progno-sis, Brain Dev 14:48–52.

Gilbert A, Tassin JL (1987) Obstetrical palsy: a clinical,pathological and surgical review. In: Terzis JK, ed.Microreconstruction of Nerve Injuries. WB Saunders:Philadelphia: 529–53.

Gilbert A (1995) Paralysie obstétricale du plexusbrachial. In: Alnot JY, Narakas A, eds. Les paralysiesdu plexus brachial, 2nd edn. Expansion ScientifiqueFrançaise: Paris: 270–81.

Gordon M, Rich H, Deutschberger E (1973) Theimmediate and long-term outcome of obstetrical birthtrauma. I. Brachial plexus paralysis, Am J ObstetGynecol 117:51–6.

Greenwald AG, Schulte PC, Shiveley JL (1984) Brachialplexus birth palsy: A 10-year report on the incidenceand prognosis, J Pediatr Orthop 4:639–92.

Hambleton P, Moore AP (1995) Botolinum neurotoxins:origin, structure, molecular actions and antibodies. In:Moore P, ed. Botulinum toxin treatment. BlackwellScience: London: 16–27.

Heinen F, Kirschner J, Mall V, Berweck S et al (1999)Botulinumtoxin – Behandlung des Pes equinus imKindesalter mit Botulinumtoxin A. Theorie, Praxis,Checkliste. Hoffmann-Verlag: Mainz.

Hentz VR (1991) Operative repair of brachial plexus ininfants and children. In: Gelbermann RH, ed. Operativenerve repair and reconstruction, vol II. JB Lippincott:Philadelphia: 1369–83.

Hierner R, Berger A (1998) Leitlinien für diePatientenauswahl zur frühzeitigen mikrochirurgischenRevision geburtstraumatischer Läsionen des Plexusbrachialis. Faltblatt: Hannover.

Hierner R, Rollnik JD, Tröger MC et al (2001) Botulinumtoxin type A for the treatment of biceps/triceps co-contraction in obstetrical brachial plexus lesions:preliminary results after a follow-up of 24 months, EurJ Plast Surg (in press).

Hierner RF, Siebenbrodt F, Berger A (1999) Vorschlagzur standartisierten Untersuchung und Dokumentationbei geburtstraumaischen Läsionen des Plexusbrachialis (‘Baby-Plexus-Evaluations-System; Baby-PES’), Krankengymnastik 51:52–62.

Hierner R, Berger A, Matthäus H (2001) Vorschlag zurstandartisierten Untersuchung und Dokumentation beiposttraumatischen Läsionen des Plexus brachialis desErwachsenen (‘Plexus-Evaluations-System; Baby-PES’),Unfallchirurg (submitted for publication).

Jackson S, Hoffer MM (1988) Brachial plexus palsies innewborns, J Bone Joint Surg 70A:1217–20.

Lone G (1966) Obstetrical lesions to brachial plexus,Acta Neurol Scand (Suppl) 18:1–80.

Marshall RW, Williams DH, Birch R, Bonney G (1988)Operations to restore elbow flexion after brachialplexus injuries, J Bone Joint Surg 70:577–82.

Michelow BJ, Clarke HM, Curtis CG et al (1994) Thenatural history of obstetrical brachial plexus palsy,Plast Reconstr Surg 93:675–80.

Mummenthaler M, Schliack H, Stöhr M (1998) Läsionenperipherer Nerven und radikuläre Syndrome, 7th edn.Thieme: Stuttgart.

310 OBSTETRICAL PARALYSIS

Page 320: Brachial Plexus Injuries

Narakas A (1986) Injuries to the brachial plexus. In:Bora FW JR, ed. The pediatric upper extremity, diagno-sis and treatment. WB Saunders: Philadelphia: 247–58.

Rollnik JD, Hierner R, Schubert M et al (2000)Botulinum toxin treatment of cocontractions after birth-related brachial plexus lesions, Neurology 55:112–14.

Rossi LN, Vasella F, Mummenthaler M (1982)Obstetrical lesions of the brachial plexus. Naturalhistory in 34 personal cases, Europ Neurol 21:1–7.

Roth G (1983) Reinnervaton in obstetrical brachialplexus paralysis, J Neurol Sci 58:103–15.

Schliack H, Memmenthaler M (1987) Lasionen periph-erer Nerven, Thieme, Stuttgart.

Schubert M, Wohlfarth K, Hierner R et al (1998)Botulinum toxin type A in obstetrical brachial plexuslesions, Electroenceph Clin Neurophysiol 107:70P–71P.

Scott AB, Kennedy RA, Stubbs HA (1985) Botulinumtoxin injections as a treatment for blepharospasm,Arch Ophthalmol 103:347–50.

Sloof ACJ (1995) Obstetrical brachial plexus lesionsand their neurosurgical treatment, Microsurgery16:30–4.

Tada KY, Tsuyuguchi Y, Kawai H (1984) Birth palsy:natural recovery course and combined root avulsion, JPediatr Orthop 4:279–84.

Tassin JL (1983) Paralisies obstétricales di plexusbrachial. Evolution spontanée. Resultats des interven-tions réparatrices précoces. Thèse, Paris 1983.

Vojta V (1988) Die zerebralen Bewegungsstörungen imSäuglingsalter, 5th edn. Enke: Stuttgart.

TREATMENT OF CO-CONTRACTION 311

Page 321: Brachial Plexus Injuries
Page 322: Brachial Plexus Injuries

Special Lesions

Page 323: Brachial Plexus Injuries
Page 324: Brachial Plexus Injuries

Traumatic injuries of the brachial plexus inchildren are rare and the literature on this subjectis sparse. Merle d’Aubigne and Deburge (1967)reported on 13 cases over a 20-year period. Inthis series, only one patient was under 10 yearsold. Rigault (1969) presented seven cases thatwere treated and followed-up. Some large adultseries (Alnot 1977, Sedel 1977, Narakas 1982)include a few children but their treatment andresults are not reported separately. However, thebrachial plexus lesion is specific, usually moresevere than obstetrical lesions and even moreserious than the adult lesions.

Diagnosis

The diagnosis is obvious. The brachial plexuslesion often occurs in a multiple trauma situationwhere association with a head lesion, severalfractures and sometimes abdominal trauma iscommon. The diagnosis is sometimes delayedowing to the severity of the other lesions;patients are sometimes even in coma.

The clinical evolution will be monitored closelywith muscle testing and EMG. After two months,if the clinical improvement is absent or limited,and if the EMG does not show clear signs of re-innervation, a myelogram will be performed. Thebest information will be given by a CT-myelo-gram. In some cases of older children, MRI hasproven effective in determining the presence ofmeningoceles but in other cases it cannot beinterpreted.

The association of clinical testing, myelogramsand EMG with evoked potentials will allow adecision regarding surgical exploration to bemade.

Surgical exploration and repair

Exposure is often extensive. In over 30 per centof cases, the lesions are so severe that theincision needs to be extended distally to the mid-arm (Fig. 1). Dissection is very difficult as vascu-lar lesions (Fig. 2) are often associated which had

30Traumatic brachial plexus injuries in childrenAlain Gilbert and Christian Dumontier

Figure 1

Extensive lesions need wide exposure and long nerve grafts.

Figure 2

Severe lesions include vascular injuries associated with thebrachial plexus injury.

Page 325: Brachial Plexus Injuries

sometimes necessitated previous surgery. Thetechnique and repair of the lesions in childrenare very similar to adult brachial plexus repair.There are, however, differences:

• The size of the defect is often larger, necessi-tating long grafts (10–20 cm);

• Neurolysis is used more often as scar is exten-sive;

• The repair will be more ambitious as inchildren, particularly younger ones, the repairmay obtain recovery of function, even in thehand;

• The severity of the lesions and the number ofavulsions necessitate the use of neurotizationswhich explains the low quality of the endresults.

Follow-up and secondary surgery

The recovery time is long and the results of theoperation cannot be assessed before three oreven four years; although we are dealing withchildren, the time for re-innervation is not muchshorter than in the adult, and certainly not inproportion to age.

Long-term physiotherapy is necessary, in orderto keep the joints supple and allow a good mobil-ity when the muscles recover.

In most cases, the recovery is incomplete andwhen possible, secondary surgery may improvethe situation. Two years after the brachial plexusrepair, it is possible to have an idea whether theshoulder or elbow flexion will recover. Tendontransfers will be an integral part of the surgicalrepair. These transfers necessitate a suppleshoulder joint and especially a good passiveexternal rotation.

If the shoulder is internally rotated, asubscapularis release or an anterior release willallow excellent external rotation. This internalrotation contracture is rare in traumatic plexusinjuries because of the involvement of morenerve roots, which usually gives a flail shoul-der.

To improve the shoulder some tendon trans-fers are possible, such as latissimus dorsitransfer, but this muscle is often paralysed andthe most commonly used muscle will be thetrapezius.

This also applies to elbow flexion. Again thelatissimus dorsi, whose excellent results are wellknown, is often not usable and it is necessary tofind other donors: triceps or pectoralis major.

In C5-C6 lesions the forearm and hand flexionsare intact and a Steindler procedure may give anexcellent result.

Supination contracture is frequent, especiallyin complete paralysis. Re-routing of the bicepsis possible only with a good muscle and asupple pronosupination. Otherwise rotationalosteotomy of the radius will solve the problem.In both these cases, the prerequisite for pronat-ing the wrist and hand is to have an active wristextension.

Wrist and hand paralysis is a great challengein these patients as there are often no availablemuscles to transfer. If wrist extension is active,tenodeses are possible and may give somefunction.

Recovery of finger flexion is difficult. It issometimes possible to do complex operationssuch as extended latissimus dorsi transfer tothe finger flexors, albeit when that muscleexists, or biceps transfer, or even free muscletransplantation using the gracilis.

Clinical material

In line with the rare occurrence of brachial plexusinjuries in children, we have been able to treatand follow-up just 41 cases from 1977 to 1998,which compares and contrasts with 696 obstetri-cal cases operated during the same period. Theetiology shows a predominance of traffic injuries(32 cases) but two cases were due to stabinjuries (Fig. 3).

Of the 41 cases, 33 were boys and only eightwere girls. Age range shows that all periods ofgrowth are represented (Fig. 4) but that thelargest group is the pre-adolescent age, from 12to 14 years.

There is a large number of associated lesions,showing that the brachial plexus trauma occursduring a violent insult with multi-tissular lesions.Over 50 per cent of the cases (21/41) were associ-ated with head injury, sometimes coma. Multiplefractures are common (shoulder in 11 cases,lower limb in six cases, upper limb in five cases).The severity of these lesions, the association

316 SPECIAL LESIONS

Page 326: Brachial Plexus Injuries

with vascular injury (6 cases) which needimmediate treatment temporarily put thebrachial plexus injury on the backstage. It issometimes several weeks before the paralysis isstudied and a complete evaluation done.

The damage requiring clinical evaluation andtreatment shows a majority of complete anddistal lesions (Fig. 5). This initial severity of theclinical aspect will be found during operation,with a great number of avulsions of lower andeven upper roots. In only 31 cases was it possi-ble to use some kind of intraplexual repair bygrafts, sometimes over 15 cm long.

In 35 cases neurolysis was done, which atteststo the great difficulty in assessing and repairingthe roots. In 17 cases, it was necessary to addsome kind of extraplexual neurotization, from thespinal accessory, intercostals, controlateral andipsilateral or pectoralis major nerves. Recently,in two cases, we were able to use end-to-sidesutures of C6 and C7 avulsions to C8 and T1roots.

Re-innervation is a long process and it ispossible to evaluate the results only after aminimum of three years. In cases where theresults are poor it is tempting to use secondarytransfers. However, only very few were done(II–25 per cent), demonstrating the lack of gooddonor muscles (Figs 6–8).

The secondary transfers were: trapezius trans-fer (three); Steindler operations (three); wristextension transfers (two); vascularized freemuscle transfer (two); biceps to flexors transfer(one). We evaluated the final results according tofunction of shoulder, elbow and hand, using ourscales (Gilbert and Raimondi) (Fig. 9).

TRAUMATIC BRACHIAL PLEXUS INJURIES IN CHILDREN 317

Figure 3

A small stab injury with complete C5–C6 lesions at thelevel of the foramen.

Figure 5

The initial lesions in 41 cases of brachial plexus trauma inchildren.

C5-C6 C5-C6-C7 C5-D1 Distal

7

20

3

11

02468

101214161820

No.

of c

ases

Figure 4

Age at operation for brachial plexus repair in children.

840 16120

2

4

6

8

10

12

14

Age at injury (years)

No.

of c

ases 8

109

14

Page 327: Brachial Plexus Injuries

Conclusion

This series, as well as the scattered reports in theliterature, confirm the impression that traumaticbrachial plexus in children is a very severe lesionbecause of the multiple traumatic associationsand the extensiveness of the avulsion injuries ofthe nerve roots. Repair is difficult and oftenlimited; neurotizations are necessary but rarelygive satisfactory results. Secondary transfers areusually impossible and it may be necessary to usemicrovascular neurovascular muscle transfers.

318 SPECIAL LESIONS

a

b

Figure 7

C5–C6 avulsion injury treated by neurotizations. (a) Lack ofexternal rotation. (b) Acceptable abduction.

a

b

Figure 6

Result after C5, C6, C7 lesion treated by grafts and neuro-tizations. (a) Good elbow flexion. (b) Poor abduction butgood wrist extension.

Page 328: Brachial Plexus Injuries

References

Alnot JY (ed.) (1977) Paralysie traumatique du plexusbrachial (symposium proceedings), Rev Chir Orthop63:5–125.

Gilbert A, Tassin JL (1984) Réparation chirurgicale duplexus brachial dans la paralysie obstétricale, Chirurgie110:70–5.

Merle d’Aubigne R, Deburge A (1967) Etiologie, évolu-tion et pronostic des paralysies traumatiques du plexusbrachial, Rev Chir Orthop 53:23–42.

Narakas A (1982) Les neurotisations ou transfertsnerveux dans les lésions du plexus brachial, Ann ChirMain 1:101–18.

Narakas A (1985) The treatment of brachial plexusinjuries, SICOT 9:29–36.

Rigault P (1969) Paralysies traumatiques du plexusbrachial chez l’enfant (Etude de 7 cas), Rev Chir Orthop55:125–30.

Sedel L (1977) Traitement palliatif d’une série de 103paralysies par élongation du plexus brachial. Evolutionspontanée et résultats, Rev Chir Orthop 63:651–66.

TRAUMATIC BRACHIAL PLEXUS INJURIES IN CHILDREN 319

Figure 8

(a) Complete paralysis withavulsion of five roots.Treatment by cross-chestnerve transfer frompectoralis major and freevascularized gracilis trans-fer. (b) Elbow flexion ispossible with controlateralactivation of pectoralismajor.

a b

Shoulder Elbow Hand Sensation

35

30

25

20

15

10

5

0

911

20

30

6 5

19

8

14

31

62

No.

of c

ases

GoodFairPoor

Figure 9

The final results in our series of 41 cases of surgical repairof traumatic brachial plexus injury in children.

Page 329: Brachial Plexus Injuries
Page 330: Brachial Plexus Injuries

Introduction

In modern-day warfare weapons have becomeincreasingly more powerful and the injuries theycause are more extensive, especially if thebullets or fragments hit a bone such as the clavi-cle. In such a case there will usually be associ-ated injuries such as vascular injuries andextensive necrosis, and saving the patient’s lifebecomes the first priority while the neural repairsare usually delayed. Moreover, although some ofthe injuries caused by high-speed projectiles aredifficult to detect, it is important to understandthem fully. Therefore proper treatment of warinjuries of the brachial plexus requires full knowl-edge of two aspects: first, the full extent andnature of tissue damage caused by penetrationof high-speed projectiles, and second, themodern treatment techniques including micro-surgery. The purpose of this chapter is to conveysome of the author’s experience gained in thetreatment of these injuries (Gousheh 1995).

First, let us discuss briefly the damage causedby a penetrating projectile such as a bullet. Thisdamage is caused by two factors. The first is thephysical body of the bullet as it passes throughthe body. In particular, if the bullet hits a boneall of its kinetic energy is instantaneouslyreleased and this causes the most extensivedamage. In this case shattered bone fragmentsbecome secondary projectiles and travel inalmost all directions, causing extra damage tothe surrounding tissue. The second and lessobvious cause of damage is the compressionshock wave accompanying the bullet. The major-ity of such damage is caused by thephenomenon of cavitation and its aftershocks,which can damage organs far removed from thephysical path of the bullet. For further details, thereader is referred to the appendix.

Preoperative considerations

War injuries of the brachial plexus manifest aspartial and/or complete disruption of one or moreneural elements, and can be segregated into fourlevels: root, trunk, cord, and terminal branches.Here, by ‘terminal branches’ we mean lowerportions of the brachial plexus down to the lowerborder of pectoralis major muscle, in the axillaryfossa. In this region the injuries usually extendupward toward the cords in such a way that theproximal ends of the disrupted elements areusually located in the cord region. The causes ofwar injuries, in order of importance, are penetra-tion of high-speed projectiles, explosion injuries,and traction injuries. As mentioned earlier, exten-sive damage to the brachial plexus, accompaniedby significant vascular injuries usually occurswhen a high-velocity bullet or shell fragmentdirectly hits the clavicle, the scapula, the first ribor another bone adjacent to the brachial plexus,such as the humeral head. In this case, total paral-ysis may occur even without anatomical disruptionof the brachial plexus elements. Exploration ofpatients with these ‘blunt’ injuries usually reveals,upon opening the epineurium, multiple punctatehemorrhagic and edematous areas in the neuralelements. When the exploration is carried out afew months or more after the injury, neuralelements are often found to be stiff and cord-likewith extensive fibrosis. The condition of thesepatients gradually improves after neurolysis andrelease of adhesions from the surrounding tissues.

The majority of the patients have completeparalysis in the motor field of the brachialplexus before the treatment. Although EMGusually indicates a complete lesion in the major-ity of patients with brachial plexus injuries, atthe time of operative dissection few patientshave complete brachial plexus disruption.

31War injuriesJamal Gousheh

Page 331: Brachial Plexus Injuries

Injuries are usually only to one or a few nerveelements and are often partial. It is not uncom-mon that nerve elements that were thought to betransected are found intact, but surrounded andcompressed by scar tissue.

Operative timing

Clean uninfected wounds caused by sharpcutting objects, such as shattered glass pieces,must be repaired as soon as possible (Dunkertonand Bomme 1988). In our experience, if anyemergency operation is necessary (such as foruncontrolled hemorrhage) all of the proceduressuch as nerve and vessel repairs should beperformed in one session. We feel that delay ofthe nerve repair is contraindicated since the re-operations are usually found to be extremelydifficult due to the formation of extensive fibro-sis and dense scar tissue as a result of the previ-ous surgery. However, when there is extensivedamage infection usually develops in spite ofantibiotic therapy and performance of the neces-sary debridement, and the repair of the brachialplexus is usually postponed until the infectionhas been eliminated, the soft tissue repaircompleted, and the general condition of thepatient has improved. This waiting periodusually lasts 3–8 weeks. It has been observedthat for some patients with brachial plexusinjuries due to penetration of small and slowprojectiles (e.g. shell fragments from a distantsource), after a few days there are no signs ofassociated injuries or infection. These cases arein the same category as clean uninfected cases,and should be repaired as soon as possible.

Preoperative work-up

A complete clinical examination in the field ofthe brachial plexus should be performed, and allmotor and sensory abnormalities meticulouslydelineated and charted. This detailed investiga-tive method, when carried out serially, helps todifferentiate between neuropraxia with potentialspontaneous recovery and the clear-cut anatomicinjuries that would benefit from surgical repair.

Electromyography (EMG) should be performedin all patients and repeated monthly for those

patients who are being followed for suspectedbrachial plexus injuries. A complete vascularinvestigation should be performed, and thearterial pulses checked along the entire extrem-ity. If the radial pulse is absent or weakcompared with the contralateral side, or if thereis suspicion regarding integrity, angiographyshould be performed.

Operative techniques

General anesthesia is used in all brachial plexusoperations. The patient lies supine with a verysmall cushion placed under the shoulder. Theentire upper limb, neck, clavicle, and pectoralisregions are prepared and draped. The limb muststay free and be supported by an assistant asnecessary. The head faces contralateral to theinjured side. The skin incision depends on theentry point of the projectile. If this point is abovethe clavicle, the skin incision is made in the shapeof an ‘L’ or ‘J’ along the lateral border of thesternocleidomastoid muscle and continued belowthe clavicle. If the entry point is below the clavi-cle, the skin incision extends from the infraclav-icular region to the deltopectoral crease. In atypical war situation, on average 17 per cent ofcases have the entry point above the claviclecausing damage to the brachial plexus at the rootlevel, and in 83 per cent of cases it is below theclavicle, usually causing damage to the terminalbranches and the cords and rarely to the trunk. Ifthe brachial plexus exposure is difficult or ifassociated problems such as vascular injuries areprobable, a large incision in the shape of an ‘S’should be performed extending from the lateralborder of the sternocleidomastoid muscle to theupper arm infra-axillary region, and the clavicleosteotomized as necessary. The major and thenthe minor pectoral muscle tendons are transectedand retracted for exposure. This method exposesthe entire brachial plexus from top to bottom.

The first step of the procedure is to find andcontrol the proximal and distal parts of the arteryin the brachial plexus region. The most importantfactor in this procedure is precise recognition ofneural lesions. For this purpose a careful neuraldissection is necessary, and during this procedurecare must be taken to ensure that the principalvessels are carefully dissected and put aside. It is

322 SPECIAL LESIONS

Page 332: Brachial Plexus Injuries

best to start the dissection from the median nerve,continuing to reach the lateral and medial cords,and in this region there is the musculocutaneousnerve on the lateral side and the ulnar nerve onthe medial side. The posterior cord is found bylocating the radial nerve and continuing upwards.The posterior cord is posterior to the axillary arteryand lateral and medial cords. The axillary nerve isin this region. If the injury is a few weeks old thesearch for the musculocutaneous and axillarynerves requires more care and time, but it must bedone. In continuing the dissection of the cordsupwards, the trunk is reached. Dissection andrepair of the nerves in the trunk region is alwaystechnically difficult and requires precise anatomi-cal knowledge and extreme care. The nerve trunksdivide in a complicated fashion to give rise tolateral, medial and posterior cords. Injuries causedby high-velocity projectiles are often associatedwith fibrosis and adhesion, which make the recog-nition and identification of these elements a matterof experience, patience and care. Therefore if thedissection of the neural elements in this region isnecessary and scar tissue is present, there shouldbe no hesitation in cutting the clavicle. If the clavi-cle is cut for better exposure, the neural or vascu-lar repair should be preceded by preparation forthe osteosynthesis, since manipulation of the clavi-cle might damage the neural repairs. After identi-fication of the neural elements in the trunk region,the roots are also dissected and identified if neces-sary. In this way the entire brachial plexus alongwith the vascular elements is dissected andexposed from bottom to top, and the injuries canbe identified and repaired.

Intact elements should be found and protected.If a nerve element, which is diagnosed asdamaged in the preoperative investigation, isfound intact during exploration, it must bechecked very carefully to exclude ‘blunt’ injuries(Narakas 1985). Nerve stimulation is useful inthis regard. If some fascicles of a plexus elementare found to be damaged or severed, they shouldbe identified and prepared for grating. Neuralelements that are compressed in fibrous tissueare freed from the surrounding scar tissue andneurolysis is carried out by incision of theepineural cover (epineurium) and removal of allforeign bodies (debris of explosives).

The recognition and diagnosis of partial disrup-tions and ‘lesions in continuity’ are difficult. Twomethods aid in their detection: serial preoperative

EMG, and interneural neurolysis under magnifica-tion. Treatment of these ‘lesions in continuity’consists of nerve grafting of the severed portionsand neurolysis for the unsevered yet disruptedneural segments. The grafts are done using 10-0nylon suture under a microscope. Generally thefusiform neuromas are also treated by neurolysis.Sometimes a long segment of a nerve element isfibrosed and innumerable tiny pieces of explosivematerials are observed penetrating the nerve.Although nerve grafting can be considered, simpleneurolysis usually yields unexpectedly good results.

In cases of complete neural disruption of anyelement, the procedure of choice is direct epiperi-neural anastomosis whenever possible, suturingthe corresponding fascicles to each other withouttension and torsion. End-to-end neural anasto-mosis has been advised by Narakas and otherswhenever the distance between the two ends ofthe severed nerve is less than four to five timesthe external diameter of the disrupted nerve(Narakas 1985). In our experience, the techniqueof end-to-end anastomosis is not possible for theneural lesions above the clavicle region, but isotherwise indicated whenever the two nerve endscan be re-approximated with an 8-0 nylon suturematerial without flexion of the joints. Using thismethod the recovery period is shortened, lessmuscular atrophy ensues, and there are overallbetter end results. When this is not possible,nerve grafting is carried out (interfascicular nervegraft is usually constructed).

For the irreparable lower elements (C7, C8 andT1), it is possible to transfer the latissimus dorsimuscle for flexion or extension of the fingers(Gousheh et al 2000).

Autogenous sural nerve is usually used as thegraft material of choice. When it is not sufficient,the antebrachialis medialis nerve is chosen(Gilbert et al 1986). Sometimes it is possible touse the nerve elements harvested from the previ-ously amputated limbs of the same patients.When this procedure is performed, theepineurium of the grafted nerves should beexcised and the fascicles positioned to lay onhealthy well-vascularized tissue to prevent theproblem of ‘cable graft’.

The ulnar nerve as well as other lowerelements of the plexus should be repaired withthe same care as that devoted to the upper nerveelements. Although the intrinsic muscles of thehand in the territory of this nerve will not gener-

WAR INJURIES 323

Page 333: Brachial Plexus Injuries

ally recover in adult patients, the presence ofsensation in the field of the ulnar nerve is veryimportant for the usefulness of the hand andperforms at least a protective function. This isparticularly important when accompanied by therecovery of digital flexion, which can be accom-plished in the majority of cases. When this iscombined with a few tendon transfers, thepatient can have a useful hand. Therefore we donot recommend using the ulnar nerve as graftmaterial for the rest of the brachial plexuselements.

The repair of the musculocutaneous nerveusually yields good results. However the repairof the radial nerve for digital extension is usuallyless successful. Therefore, when one repairs theposterior cord at plexus level, one shouldconsider tendon transfer for digital extension atthe same time.

Associated injuries

There are usually some associated injuries,which should be treated in the first operatingsession, along with the neural repairs, wheneverpossible. These injuries can be categorized asfollows.

1. Vascular injuries related to the subclavian andaxillary vessels (McCready et al 1986):• Arteriovenous fistulae: these can be treated

by autogenous vein graft. In these casesthe graft can be obtained from the internalsaphenous vein. Rarely, the patient with alateral laceration of the subclavian arterycan undergo repair by simple suture.

• False aneurysms: these are usually locatedon the axillary artery and sometimes on thesubclavian artery. Due to the severity of thevessel lesions, generally the only feasibletreatment is with interpositional autoge-nous vein graft, rather than direct repair.

• Complete disruption of the axillary artery inits proximal position: this is usually due toligation of the bleeding artery in emergencyoperations in the front-line hospitals toprevent life-threatening hemorrhage.

2. Associated pulmonary injuries: these areusually due to emergency treatments bythoracotomy or chest tubes.

3. Skeletal injuries: these can be either to theclavicle, scapula or head of humerus,separately or in combination. Most of thecomplicated plexus injuries are associatedwith comminuted or shattered bone injuries.

4. Causalgia should be mentioned as a sequela.Most cases respond to medical therapy orsimple neurolysis. Severe causalgia can besuccessfully treated with upper thoracicsympathectomy.

For the vessel repair, the whole injured area isusually resected and repaired by vessel graft. Todo this, first the proximal and distal parts of theartery or artery and vein are controlled andtemporarily closed off. The whole area of thevessel and the vessel graft are rinsed withheparin normal saline solution. When perform-ing the graft, attention must be paid to the direc-tion of the valves in the vessel graft for bloodflow. Suturing is with 8-0 nylon. Usuallycomplete vascular disruptions are repaired or atleast ligated in the front-line hospitals. If theradial pulse is weak and the blood flow to thehand does not seem sufficient, the damagedvessel area should be re-examined and repairedagain if necessary.

Appendix: Injuries caused byhigh-speed projectiles

Penetration injuries of the brachial plexus due tohigh-velocity bullets or shell fragments are differ-ent from other penetrative injuries such as stabwounds or non-penetrative injuries such astraction injuries. This difference is due to thealmost spontaneous high-energy release to theregion by the penetrating agent. The amount ofenergy transfer depends on the following factors:

1. Speed. In modern weapons the speed of thebullet is several times greater than the speedof sound in air. The amount of kinetic energyassociated with the center of mass motion ofa projectile with mass m and speed v is givenby KE = 1⁄2 m.v2. As the projectile travelsthrough the air, its speed decreases due to airfriction. The extent of tissue damage causedby the these projectiles obviously dependson their speed at the point of the impact.

324 SPECIAL LESIONS

Page 334: Brachial Plexus Injuries

Shell fragments slow down faster thanbullets due to their irregular shape, whichcauses more air friction. However, shellfragments could have a much higher initialspeed – as much as 2000 m s–1. Generallytheir effective range as high-velocity projec-tiles is less than 100 m;

2. Auxiliary motion. The projectiles usually haveauxiliary motion, in addition to their averagevelocity towards the target. This additionalmotion carries extra kinetic energy which,when released in the target, causes extradamage. The additional motions of the bulletsconsist of the following:a. Spiraling: the center of mass of the bullet

actually moves on a spiral path, rather than astraight line, between the gun and the target;

b. Spin: the bullet spins or rotates about itsphysical axis. This spin is caused by thespiraling grooves of the gun barrel;

c. Precession: the physical axis of the bulletrotates about the axis defined by theinstantaneous velocity vector of its centerof mass;

d. Nutation: the axis of the bullet has smalloscillations perpendicular to the directionof its precession and its spin;

3. Shockwaves. In many war injury cases wehave observed tissue damage far removedfrom the physical path of the penetratingprojectile. This is due to the shockwavesaccompanying the projectiles. As is wellknown in the science of aerodynamics,objects traveling faster than the speed ofsound in a given medium produce shock-waves in the form of a cone-like ‘shell’consisting of compressed and high-pressurematerial comprising the medium. This ‘shell’accompanies the projectile and travels justbehind it, and inside it there is a partialvacuum. This whole structure, consisting ofthe bullet, the shockwave and the partialvacuum, is called a ‘Mach’s cone’. This isexactly the phenomenon that we observewhen an aircraft breaks the sound barrier.

When the bullet strikes the body, it suddenlyslows down. The shockwave accompanyingthe bullet also strikes the body, like the wagonsof a train giving aftershocks once the locomo-tive strikes an obstruction. The shockwaveusually does not stop there, and accompaniesthe bullet even after it enters the body. There

they both cause damage. The damage that theshockwave causes is complicated, and isrelated to the amount of energy that it trans-fers to the body. This in turn depends on manyfactors such as the resistance (or hardness ordensity) of the tissue and the length of tissueon the trajectory inside the body. For example,a bullet passes through an empty containerwithout appreciable energy loss, and thereforewithout inflicting appreciable damage to thecontainer, by simply making two holes.However, if the container holds water thebullet loses more energy, and therefore thedamage to the container is greater and the exithole is larger. An extreme case would be whenthe container holds tar. In this case the bulletloses all of its energy inside the container andis stopped there, imparting great damage tothe container.

One important mechanism by which theshockwave inflicts damage inside the body isby making an almost instantaneous cavity.This is due to high pressure contained in theshockwave. The boundary of the cavitybecomes necrosed tissue. This phenomenon iscalled cavitation. It is difficult to detect anddebride this necrosed area in the first few daysafter the injury. The remaining shockwave thentravels through the body and might damageorgans far from the physical path of the bullet.

References

Dunkerton MC, Bomme RS (1988) Stab wounds involv-ing the brachial plexus. A review of operated cases, JBone Joint Surg 70B:566–70.

Gilbert A, DeMouraw, Salazar R, Grossman J (1986)Prélèvement des greffes nerveuses. In: Tubiana R (ed).Traité de Chirurgie de la Main, Vol. III. Masson: Paris:451–7.

Gousheh J (1995) The treatment of war injuries of thebrachial plexus, J Hand Surg 20A:S68–76.

Gousheh J, Arab H, Gilbert A (2000) The extended latis-simus dorsi muscle island flap for flexion or extensionof the fingers, J Hand Surg 25B:160–5.

McCready RA, Procter CD, Hyde Gl (1986)Subclavian–axillary vascular trauma, J Vasc Surg3:24–31.

Narakas AO (1985) The treatment of brachial plexusinjuries, Int Orthop 9:29–36.

WAR INJURIES 325

Page 335: Brachial Plexus Injuries

Page numbers in italics denote figure legends. Wherethere is a textual reference to the topic on the samepage as a legend, italics have not been used.

3D-CISS 35, 36

abductor digiti minimi muscle 25abductor pollicis longus muscle 25accessory nerve 86, 98

transfers 220, 222, 223acromion 110, 141, 182acromion bone block 119acromion transfer 118, 229Active Movement Scale 161, 162, 163, 168adduction contractures 178, 179adductor pollicis brevis muscle 24adductor pollicis muscle 24adhesions 145, 229, 323aetiology 47–50, 151–7against-gravity movement 163age assessment 189age effects on nerve repair 102agnosia 177Albinus accessory muscle 9allografting, vascularized 53amputation 55, 97anaesthesia 189–90

general 49, 322analicular syndromes 11anatomy 21–9, 67–8

surface 192aneurysm, traumatic 31angiography 322ansa pectoralis 125, 126anterior scalenus muscle 7, 8, 194anterior serratus nerve 14antibiotics 245aneurysm 96apnea 189arm function, distal 108arterial injury 94arterial resection 96arteriovenous fistula 96arteriovenous fistulae 103arthrodesis 63, 107–12, 301

to provide key grip 89see also specific joints

arthrograms 242, 252articular process 5

associated lesions 156augmentation transfer of elbow 126autogenous grafting 323autologous grafting 95, 110, 196auxiliary motion of projectiles 325avascular necrosis 257avulsions 58, 198

of cervical roots 79intradural injury 53in situ 220, 224

axillary artery 100rupture of 97, 102transection 96traumatic occlusion 36

axillary cavity 11, 12axillary nerve 3, 13, 14

muscle innervation 21, 22neurotization 62, 85palsy 116–19repair 116see also circumflex nerve

axillary region see infraclavicular regionaxon donors 86–7axonal supercharging 265–6axonolateral regeneration 52axonotmesis 39, 40, 92axonotomy, permanent 92

behavioural outcomes of OBPP 182behavioural problems and OBPL 166Bell’s nerve see large thoracic nervebiceps brachii and brachialis muscles 29biceps function 174biceps neurotizations 54, 85biceps recovery in infants 168, 205–7, 240biceps re-routing 281, 282, 286, 288, 294, 298

results 300bilateral lesions 218, 224biofeedback 143, 268, 269, 305, 309bipolar transpositions 125, 128birth weight 153, 154, 217bleeding, control of 193botulinum toxin type A 264, 304, 305–9bowstring deformation 133, 141brachial plexus 3–15, 91

individual variation 52reconstruction of 11repair in children 316, 317trauma in children 317

brachial plexus exploration 140

Index

Page 336: Brachial Plexus Injuries

brachial plexus lesions 41–2, 86in continuity 323outside BP 42

Brachial Plexus Work Group 174–5, 182, 186brachioradialis muscle 29brachiothoracic pinch 70breast cancer 50breech deliveries 152, 153, 155, 156, 198, 208

in forearm supination deformity 277results of surgery 217–24

British Medical Council classification 309Brown–Sequard syndrome 20, 68

C5 plexus root 194C5 rupture 71–2C5–C6 avulsion injury 318C5–C6 injury 198–200, 218C5–C6 lesions 174–5, 211–13, 251, 276, 317C5–C6 palsies 57, 59, 62, 63

and elbow stiffness 263C5–C6 rupture 72C5–C7 lesions 174–5, 200–1, 276, 318

after breech delivery 218results of repairs 213

C5–C7 palsies 57, 59, 62C5–C7 rupture 73C5–C8–T1 palsies 57C5–T1 avulsion 69–71C5–T1 lesions 276C5–T1 rupture 73C6 plexus root 194C6–T1 avulsion 71–2C7 contralateral nerve root 54–5, 69, 85, 87

elongation 300transfer 296

C7 experimental grafting 200C7 plexus root 194C7–T1 avulsion 72, 289C8 plexus root 194C8–T1 avulsion 73, 201C8–T1 lesion 276C8–T1 palsies 57, 63cable graft 323Caesarean deliveries 156, 157Carliotz operation 241, 244carpometacarpal arthrodesis 143, 145causaglia 104, 324cavitation 325cellular adipose layer 11central apnoea 189central mechanism 33, 155, 156cephalic deliveries 153, 155cervical cutaneous nerve 193cervical nerve roots 47–8, 60cervical nerves 3, 4–5, 191

anatomy 9, 11, 12, 13, 14, 15avulsion 42, 86in classification of BP injuries 165damage 152, 154, 155and elbow paralysis 262foraminal anatomy 6, 7injury in OBPP 198muscle innervation 21–9

and pain 19, 20traction lesions 47

cervical plexus 69cervical surgical approach 59cervical transverse process fractures 20cervical–thoracic node see star-shaped nodechromatolysis 92circumflex nerve 100–1

palsy 100rupture 99see also axillary nerve

classificationof adult traumatic BP lesions 50of arthrodeses 108of BP injuries 57, 92, 123of co-contraction 304, 305of focal mechanical injury 93of function in OBPL palsy 163, 164of hand sequelae 289–90medial rotation contracture 251–2of OBPP 165, 293

sequelae 277of sensory response in infants 164, 165of shoulder deformities 178surgical findings 219of wounds 93–4

Claude–Bernard–Horner syndrome 8clavicle

divisions 51fracture/dislocation 31splitting 191, 192traction injury to 47war injuries 321

clawhand deformity 152, 160, 288clinical findings in complete palsy 68closed injuries 47, 79

with fracture/dislocation 93, 94traction 96–7traction rupture 94

coaption 87cock-up splints 176co-contractions 71

deltoid/teres major 304, 306treatment of 303–9triceps/biceps 126, 264, 266, 309

and toxin therapy 306, 307, 308co-contractures 81collateral branches of brachial plexus 13–14collum scapulae fracture 32combined treatments 83complete palsies 67–74

shoulder 107without recuperation 60–2see also total palsies

complete paralysis 321old 133–4, 135persistent 173of upper extremity 123see also total paralysis

complete root avulsion 133, 144from traction injury 48

complicationsarthrodesis 111OBPP surgery 202

INDEX 327

Page 337: Brachial Plexus Injuries

compound muscle action potential (CMAP) 39, 40compression 78

and traction 49compression lesions 49compression shockwave 321compression theory 151computer measurements of forces in delivery 156conduction block see neuropraxiaconservative management of OBPP 175–7contractures 77, 176, 244–6

capsular 226–7, 230, 233, 234and dislocation 240, 241–2in elbow paralysis 261flexion 261, 262intentional 143internal rotation 316medial rotation 249–58

causation 250–1classification 251–2diagnosis 252treatment 255–8

pronation 261, 263, 299releases 226soft tissue 273supination 88, 181, 261, 263, 316

forearm 214, 280, 298–9Volkmann’s 133, 263

cookie test 169, 170, 308, 309coracobrachialis muscle 29coracoid process 182, 256cryoprecipitate thrombin 197CT (computerized tomography) 242, 252CT scan arthrography 180, 226CT-myelography 33–4, 58, 60, 61, 208

repeated 59technical notes 34

cubital nerve 15cuff lesion 32Current Muscle Grading System 168curvature of forearm bones 280, 284cutaneous nerves 196

decortication 109, 110, 111deep muscles of neck, nerves for 13deep pressure sense test 19–20degenerative lesions 91–2, 93dehiscence 229Dejerine–Klumpke paralysis 151deltoid muscle 21demography of OBPL 152denervation 42, 156depressor muscles 115descending radio-ulnar fibres (DRUF) 278descending ulno-radial fibres (DURF) 278developmental outcomes of OBPP 182diagnosis

of co-contraction 304delayed 249differential, of OBPP 183of injuries to terminal branches 91–3lesion 79–80medial rotation contractures 252of traumatic BP injuries 315

diagnostic cervical laminectomy 37digital extension 132digital flexion 133, 324direct epiperineural anastomosis 323discriminants of recovery 167dissection

of BP 194, 195Millesi’s technique 84of upper trunk neuroma 193

distal fixation 118distal radio-ulnar joint 281

dislocation 276donor muscle tissue 139donor nerve tissue 86–7, 139, 140

availability of 73outside BP 82, 197sural nerve as 196, 323

dorsal artery of scapula 11dorsal interossei muscles 26dorsal scapular nerve 13, 87, 195, 227

muscle innervation 23double level lesions 68, 102Duchenne–Erb paralysis 276Duchenne–Erb–Klumpke paralysis 276Dutch Health Care Information Centre (SIG) 153dynamic stability 143Dysport 306

elbowextension 138

absence, total 261functional pathology 264and muscle transfer 126, 127, 146, 272secondary reconstruction 143

extensors 262fixation 124flexion 85, 111, 134, 137, 308

absence, total 261after total palsies 61, 62assessment after 169and BP trauma in children 316functional pathology 263–4improvement 88, 237late restoration 208and muscle transfer 126, 127muscle/tendon transfers for 124neurotizations 62, 198, 200reconstruction 123, 125, 135, 266recovery 70, 174, 214scores 167treatment of paralysis 264–72

flexors 262function 111, 117palliative surgery 123–9paralysis 261–73stability 138, 146stiffness 261, 262–3, 264, 272–3

elbow arthrodesis 107elbow deformities 181electrical stimulation 132, 142, 151

for nerve identification 270electrodiagnostic studies 173electroencephalography (EEG) 39

328 INDEX

Page 338: Brachial Plexus Injuries

electromyography (EMG) 39–43, 58, 59, 309biofeedback training 143in co-contraction 304of traumatic BP injuries 315vs clinical examination 42of war injuries 321, 322

electrophysiologic studies 79EMG see electromyographyEMG-triggered myofeedback 177endoscopic harvesting 140endoscopic investigations 37energy transfer 103, 104, 324epidemiology of co-contraction 303epineuriectomy 80epineuriotomy 80, 84epineurium 48Erb–Duchenne paralysis 151Erb’s palsy 152, 154, 155, 165, 263

typical position 159, 160evaluation, initial 159–64evoked potential (EP) measurement 39,

315examination

clinical 57, 58, 79, 243, 315charts 58neurophysiological 39–44repeated 59vs electromyography 42of war injuries 322

neurological 183OBPP 205–6paraclinical 57, 58physical 17–29, 157–64shoulder 185

exercisesfor medial rotation contractures 255range of motion (ROM) 175–6, 182, 185

extensor carpi radialus muscles 28extensor carpi ulnaris muscle 28, 280, 289

transfer 282extensor carpi ulnaris tendon 289extensor digiti minimi muscle 27extensor digitorum communis tendons 141extensor digitorum muscles 27extensor indicis muscle 27extensor pollicis longus muscle 25extensor-to-flexor transfer 298external rotation 85, 88, 112, 225, 244–5

active 180, 222, 232passive 176, 227shoulder paralysis 233, 234, 236

extraplexal neurotization 68, 197extra-scalenus anatomy 10–11

false aneurysms 96, 100, 101, 103fascia lata grafts 244fascicular grafts 52fasciculus lateralis 125fasciculus medialis 125–6fibrin glue 196, 197fibrosis 84, 101, 276, 323

post-ischaemic 97finger extension 144

finger flexion 134after complete palsy 72correction 297–8reconstruction 135, 141, 144restoration 71

finger function 137, 280fixed supination deformities 282, 283, 284fixed supination deformity 263, 279, 286, 287flail arm 77

in infants 168, 205, 207flail hand 205, 207flail shoulder 116, 225, 227flail upper limb 137flexible supination deformity 282, 283, 287flexion pronation posture 249flexor carpi muscles 27flexor carpi ulnaris muscle 28

in Steindler flexor-plasty 267–8flexor digiti minimi muscle 26flexor digitorum profundus muscles 27flexor digitorum superficialis muscles 27flexor pollicis brevis muscle 25flexor pollicis longus muscle 25fluids and anaesthesia 190fluoroscopy 160foetal flexion posture 263foraminate anatomy 5–8foraminate space 6forearm

flexion/pronation 126palliative surgery 123–9

forearm deformities 298forearm flexors 297, 299–300forearm supination deformity 275–6, 279–81, 287–8

classification 280operative procedures 281–9see also fixed supination deformity; flexible

supination deformityFormer Numerical Score 168Foundation of Perinatal Epidemiology in the

Netherlands (PEN) 153fracture of intima 96–7fractures 68

after breech delivery 217and BP trauma in children 316indications for intervention 95

free muscle transfers 127, 128, 133–4, 135, 271–2, 296complications 145–6double 144, 145palliative surgery 137–46

free muscle transplantation 297, 300functional muscle stimulation 268

Gerdy’s ligament 12Gilbert and Mallet scores 257Gilbert and Tassin Muscle Grading Systems

161Gilbert shoulder function staging 252–3, 257Gilbert/Raimondi assessment 186Gilbert’s classification 11glenohumeral arthrodesis 70, 138, 145glenohumeral deformity 246glenohumeral dissociation 20

INDEX 329

Page 339: Brachial Plexus Injuries

glenohumeral jointin arthrodesis 107, 108, 117deformity 242, 249stabilization 128, 146

glenoid deformity 178glues, biologic 53, 196gluteal aponeurosis 131, 132, 133gracilis muscle

harvesting 140, 141transfers 133

graft revascularization 53grafts/grafting 52–5, 195

donors 86–7grasp release 138gravity-eliminated movements 161, 162, 163greater auricular nerve 98, 193gripping function 85, 88guidelines for action in OBPP 183, 184, 185gunshot injuries 50

haematomas 36, 52, 217haemostasis 110hand

deformities 298function 111, 117, 214

after total palsies 61in partial palsies 62, 63in S-OBPP 293

palliative surgery 131–6recovery 207, 215sequelae 289–90

hand assessment 186handgrip reconstruction 137handgun injuries 104Hara’s technique 62harvesting from amputations 323hemidiaphragm paralysis 160hemilaminectomy 81hemiparalysis of diaphragm 32–3high velocity traction injuries 47high-speed projectiles 321, 324history, obstetric 159–64Horner’s syndrome 20, 57, 151, 152, 165, 185, 277,

279four signs 160in infants 159, 168, 170, 206, 207

Hospital for Sick Children Muscle Grading System161

humeral head 321humerus 32, 257humerus fractures 111, 117hypoglossal nerve 69, 221hypothermia 190Hyrtl’s intercosto-brachial nerve 15

iatrogenic injuries 49, 49–50immobilization, postoperative 60, 87–8, 227, 268,

269after muscle transfers 229, 231, 232, 237, 245

immunosuppression 53incisions 51, 59, 83, 191–2indications

for arthrodesis 107–10

for intervention 94, 95in OBPP 205surgery 206, 207

industrial traction injuries 49infant OBPP see I-OBPPinferior gleno-humeral angle 254inferior trunk 3, 4infraclavicular branch of BP 13, 14infraclavicular fossa 84infraclavicular injuries 50infraclavicular lesions 67, 94infraclavicular region 11–13, 14–15infraspinitus muscle 22initial OBPP see I-OBPPinnervation of muscles 42insertion of biceps, advancement of 271integrated therapy 128integrated treatment concept 303intercostal nerves 15, 60, 68–9, 86–7, 139

transfers 137, 141, 221, 223as transformator 301

intercostobrachialis nerve 87intermediate scalenus muscle 9internal rotation 112, 177, 226interneural sclerous nodules 51interosseous membrane 278

released 282, 286, 294retracted 279, 286

inter-scalenus artery 9intra-articular fracture 100intradural injury 33, 34, 91intraplexal neurotization 68, 197intra-plexus distribution 4intrinsic palsy joint extension 299, 300intrinsic palsy of hand 299I-OBPP 293IP-extension dynamic splint 296ischaemic necrosis 145

joint deformities 226joint mobility 139, 163joints of scapular belt 108

key grip 137kinetic energy of bullets 321Kirschner wire 281Kline and Hudson grading system 97Klumpke’s palsy 152, 160, 165, 276, 297

laceration repairs 52Langer’s arch 12Langer’s muscle 12large thoracic nerve 8, 9, 12

see also long thoracic nervelate exploration 207–8late OBPP with deformity see S-OBPPlateral cord 4, 12lateral pectoral nerve 22, 23lateral rotation 243lateral–cervical surgical approach 5latissimus dorsi muscle

bipolar transfer 268–9bipolar transposition 125

330 INDEX

Page 340: Brachial Plexus Injuries

latissimus dorsi muscle (cont.)innervation 22, 124–5island flap 131and plication of diaphragm 218–19transfers 133, 226, 233, 243

anterior approach 230–1for BP trauma in children 316indications 230posterior approach 231–2results 232, 236in war injuries 323

latissimus dorsi transfer 128latissimus to triceps transfers 272L’Episcopo transfers 243lesser occipital nerve 193levator scapulae muscle

innervation 23transfers 227, 236transposition 230

ligamentoplasty 63, 116Limb Motion Scores 166long thoracic nerve 13, 14, 194, 195, 227–8

muscle innervation 24neurotization 85palsy 120–1see also large thoracic nerve

low velocity traction injuries 49lower plexus palsy 165lower subscapular nerve 22, 23lower trunk 11, 12, 194, 195lumbricalis muscles 26

macrosomia 153, 154, 217magnetic resonance angiography (MRA) 36magnetic resonance imaging (MRI) 33, 35–6, 80, 226,

242glenoid visualization 252, 258

magnetic stimulation (Magstim) 39, 42magnification 191Mallet and Gilbert scores 257Mallet sum scores 222–3Mallet’s movement classification/scale 163, 164,

205and disability 223for infants 206, 221–2in OBPP management 180, 182, 185in shoulder dislocation 253, 254, 257

manual muscle tests 17mechanical injury in delivery 157Meckel’s adipose mass 10medial cervical fascia 10medial cord 4, 103medial epicondyle 267medial pectoral nerve 14, 22, 23medial rotation 249, 255median nerve 13, 15, 102–3, 126

muscle innervation 24, 25, 26, 27, 28outcome of nerve repair 103in Steindler flexor-plasty 268

Medical Research Council Muscle Grading System 17,161, 170

Medical Research Council system of outcomemeasurement 97

meningoceles 35, 61, 208CT-myelography 33, 34myelography 219, 223

Meyer’s guillotine knife 53microsurgery 52microsurgical neurolysis 51microsurgical repair, early 237microvascular muscle transplantation 82middle root lesions 213middle scalenus muscle 9middle trunk 194, 195, 200

anatomy 3, 4, 11, 12exposed 191

Millesi’s principles 52Millesi’s technique 53, 83–9Moberg-type reconstruction 137, 138modelling arthrodesis 109monopolar transpositions 128motor functions 19, 39, 160–3, 173motor unit potentials 42MP joint drop/wrist drop 297MR techniques, variant 35multiple arthrodeses 107multiple nerve transfers 137muscle

function 97imbalance 241, 242, 279innervations 115

muscle grading systems 161muscle grafting 83muscle test charts 17muscle transfers 77, 82, 117–19

for elbow flexion 124extensor carpi ulnaris muscle 282microvascular neurovascular 318multiple 236in posterior shoulder dislocation 250preconditions for success 123–4secondary 88, 317shoulder 118, 242

grade 0 227–30, 232, 236–7grade II-III 230–6

muscle transposition 243muscle–tendon transfers 271, 272–3muscular imbalance 177, 251musculocutaneous nerve 3, 13, 15, 59, 101–2,

297muscle innervation 29neurotization 62results of repairs 101transfers 221

myelograms 315myometric study 100

Nakamura brace 142Narakas’ Grading System for outcome in OBPP 170,

174Narakas’ Sensory Grading System 164, 165narcosis 42natural history of OBPP 173–4neck–shoulder angle 47, 48necrosed tissue 325needle EMG 40

INDEX 331

Page 341: Brachial Plexus Injuries

nervegrafts/grafting 60, 77, 82, 318, 323

bridging 81elbow reconstruction 126, 128by gluing 53, 197for sharp injury 49

growth 134regeneration 52repair 58, 95

evaluation 211grading of results 102, 211timing of treatment 102

retraction 92, 95rupture 155severance 39

nerve compression 18nerve conduction studies 39nerve root avulsion 47nerve transfers 81–2

in breech delivery BP lesions 224for complete palsy 68–9

nerve-crossing 142neurolysis 51–2, 78, 80–1, 84, 317, 323

in OBPP 195neurolysis, external 103neuromas 41, 134, 152, 194, 220

resection 195, 196, 220neurophysiological investigations 39–44, 219neuropraxia 39–40, 51, 92, 93, 322

physical examination 18, 20neurorrhaphy 78–9

direct 81, 82end-to-side 63for sharp injury 49

neurosurgerycriteria 185for OBPP 173, 174–5

neurotizations 53–4, 85, 193, 318after total palsies 70of biceps nerve 62by C7 contralateral nerve root 54–5in complete palsy 60, 61, 68–9distal stumps 81limited 82–3in OBPP 197, 198, 208

neurotmesis 39, 40, 92neurovascular pedicle 131Neutral-0-Method 309night-splinting 268non-degenerative lesions 91, 93

OBPL see obstetrics brachial plexus lesionsOBPP see obstetrical brachial plexus palsyobservation 160obstetric brachial palsy 42–3, 151obstetric characteristics of OBPL 152obstetric lesions 35obstetric patients 32–3, 34obstetrical brachial plexus palsy (OBPP) 173–86,

239obstetrics brachial plexus lesions (OBPL) 153, 155,

211–15obstructive apnea 189

Omnipaque 300 34omohyoid muscle 59, 194open injuries 47, 49–50, 79, 96, 102open reduction of dislocations 241operative techniques

free muscle transfers 134latissimus dorsi muscle island flap 131–3war injuries 322

opponens digiti minimi muscle 26opponens pollicis muscle 24opponensplasty 298orthosis, functional 142osteoclasis, closed 282osteosynthesis 51, 78, 323osteotomy

clavicle 73, 84, 192coracoid 230, 246exorotation 181glenoid deformity 258humeral 181, 242, 243, 244, 273lateral rotation 255radius 282rotational 294

outcomesbehavioural 182functional 146of nerve repair 97–8, 103of subscapularis recession 258

pain 19, 78persistent 95and poor lesion prognosis 57from radiation injury 50severe 93, 97

pain syndrome 19, 61, 78, 146palliative surgery

elbow and forearm 123–9elbow paralysis 261–73forearm and hand deformities 293–301free muscle transfers 137–46hand 131–6obsolete in OBPP 237prosupination 275–90shoulder paralysis in neonates 225–37tendon transfers to shoulder 115–21

in OBPP 239–46palmar interossei muscles 26palmar sensibility 61palmaris longus muscle 27paralysis

deep 95hand 60patterns 77serratus anterior muscle 120

paralytic shoulder 115–16paraneuriotomy 80, 84paraspinal EMG 41parental involvement 176, 183partial lesions 41, 52, 323partial palsies 62, 63partial paralysis 123, 128passive pronation of forearm 283, 284pathogenesis of BP injuries 152

332 INDEX

Page 342: Brachial Plexus Injuries

pathophysiologyof co-contraction 303–4of nerve lesion 39–40

patient re-education 131, 132patient selection 71, 138–9patient support groups 208patients with OBPL 153–7, 217pectoralis major muscle 126

free muscle transfers 134innervation 22, 125neurotization 85transfers 120, 128, 269–71

operative steps 269transplantation 121

pectoralis minor muscle 22, 23, 126pectoralis nerves 13, 14, 134, 221pectoralis nerve transfers 222, 223pectoris minor muscle 59penetrating missile injuries 103penetration injuries 324peripheral lesions 41phrenic hemiparalysis 218–19phrenic nerve 69, 86, 194, 195phrenic nerve lesions 224, 272phrenic palsy 218physical therapy 303, 305, 309physiotherapy 124, 132, 245, 316pinch test 137

see also deep pressure sense testplatysma colli muscle 10plexus evaluation system 303plication of diaphragm 218–19plication of extensor tendon 297, 300point of Erb 151poliomyelitis

arthrodesis after 117trapezius transfer after 229, 230

polyneural innervation 42position for arthrodesis 109posterior bone block 258posterior cervical triangle 10posterior cord 3, 4, 12, 13, 177, 194, 323posterior cord sequelae 277posterior gleno-humeral angle 254–5posterior scalenus muscle 8postganglionic fibres 92postganglionic lesions 67postoperative care 87–8, 202postoperative management 142post-traumatic arthritis 100posture, characteristic, in OBPP 239predictors of recovery 174prefixed plexus 4preganglionic lesions 67, 223prehension 137, 146preoperative considerations 321–2primary BP exploration 166primary interventions 220primary operations 223primary repair, urgent 96prime mover muscles 115profundus flexor tendons 132prognosis 57, 166–7, 173

progressiveness of deformities 279pronation deformity 288pronator quadratus muscle 28pronator teres muscle 28prosupination 275–90proximal axonal stump 40, 58, 195proximal joint instability 145–6proximal limb sequelae 277proximal middle-limb sequelae 277proximal radial head dislocation 299pseudoarthrosis 31, 111pseudomeningoceles 60, 63pseudoparalysis 160pseudotumours 160pulleys 124, 126pulmonary injuries 324pulped bone grafts 110, 111Putti’s sign 279

radial dislocation 176radial head, dorsal dislocation 299radial nerve 3, 13, 14, 101

and humerus fractures 32muscle innervation 25, 27, 28, 29results of repairs 101

radiation 47radiation injury 50radical neck surgery 98radicular artery 6radicular avulses 7radicular surgical approach 6–7radiculo-medullar artery 6radiological features 250radiological investigations 31–7, 219radio-ulnar fusion 284, 286, 287, 288Raimondi’s Grading System 171, 182–3, 186range of motion (ROM)

exercises 175–6passive 182, 185

range of movement 258, 309reconstruction 69–73, 88–9, 166

elbow 293in OBPP 195, 197–201shoulder 293

reconstructive goals 69reconstructive neurosurgery 152redressing splint 176re-education 266regenerated muscles for transfer 82rehabilitation 142–5, 268, 269reimplantation, nerve root 69re-innervation 42, 173, 266, 301, 317

after grafting 61assessment after 42delayed/failed 145electromyographic 142–3incomplete 225

residual function 123retraining elbow flexion 269retroclavicular injuries 50retroversion of humerus neck 256, 257rhomboid muscles 23, 120, 228, 229rhomboid nerve 9

INDEX 333

Page 343: Brachial Plexus Injuries

Richet’s clavicular–coracoaxillary fascia 12road traffic accidents 47, 49root avulsions 85–6

comparative incidences 48and EMG 41and pain 19, 20

rootlet avulsions 58, 152and CT-myelography 33, 34

rootlets 35roots, anterior 42rotation contractures 177, 178, 179, 180rotation osteotomy 63, 117, 181rotational lower arm 186rotator cuff

rupture 100system 225

Rouvière’s clavipectoral–coracoaxillary fascia 12rupture repair 86, 198

Saha’s modification 118scalenus muscles 8scalenus region 8–10scalenus sickle 9scapula

fracture 32lengthening of 241movement 109stabilization 85

scapular muscles 115, 239, 244scapular spine fractures 49scapular winging 120, 246scapulo-humeral angle 47, 176scapulo-humeral arthrodesis 107scapulo-thoracic arthrodesis 121scapulo-thoracic dissociation 31scapulo-thoracic joint 115, 227–8scar tissue/scarring 269, 322, 323score of ten system 293Sébileau scalenus–vertebro-pleural space 7secondary bone deformities 252, 304secondary joint deformities 306secondary operations 223secondary procedures 303secondary reconstruction

elbow flexion 124–7free muscle transfers 143, 145for grade 0 shoulders 236

secondary shoulder surgery 116–17secondary surgery 316

for OBPL 173, 214, 221Seddon’s classification 92sensibility assessment chart 18sensibility of hand 146, 324sensibility tests 18sensory evaluations 19sensory function assessment 164sensory functions 19, 138sensory nerve action potentials (SNAP) 39, 40,

140in neonates 219

sensory nerve function 39sensory reconstruction 142SEP see somatosensory evoked potentials

sequelae deformities of S-OBPP 295, 300shoulder 177–81

sequelae of OBPP see S-OBPPserial arthrodesis 108serial casting 176serial preoperative EMG 323serratus anterior muscle 24, 57, 99, 120

palsy 99Sever–L’Episcopo transfers 243sharp injury 49shockwaves from projectiles 325shotgun injuries 50, 103shoulder

abduction 237and latissimus dorsi transfers 233, 234scores 167and trapezius transfer 119

abnormalities 240arthrodesis 61, 62, 107, 117, 246

results 111X-rays 109, 110

deformity in OBPP 177, 178, 240–2, 263dislocation 20, 32, 240, 241–2

palsy after 57posterior 244, 249–58simple 257

dystocia 152, 156elevation 115function 174, 211, 240, 257

improvement 232neurotizations 198, 200recording 252–4and Test Scores 169

hyper-abduction 47hyperextension 67muscles 115, 239paralysis 225–37stabilization 61, 62, 85, 118, 137, 237subluxation 240, 241, 244

shoulder assessment 186shoulder examination 185shoulder instability, multidirectional 116, 119single-fibre EMG 42skeletal abnormalities 178, 306skeletal injuries 103, 324skin as monitoring device 134, 145skin moisture 19S-OBPP 293, 294, 296soft tissue damage 103soft tissue procedures 181somatosensory evoked potentials (SEP) 42sonography 37speed of projectile 324–5spinal accessory nerve 68, 70, 191, 193, 201

as donor tissue 139injury 98, 99in neurotization 60, 62, 140–1palsy 99, 121transfers 137

spinal cord 34, 35, 47spinal nerve 5, 6, 11

intraforaminal lesion of spinal nerve 31, 54spinal-evoked potentials 140

334 INDEX

Page 344: Brachial Plexus Injuries

splinting 176, 268spontaneous recovery 166, 167, 277

in adults 77–8of external rotation 227incomplete 225in OBPL 221in OBPP 174, 205–6, 233, 234

sporting activities 177sports traction injuries 48, 49stab injury 317star-shaped node 7, 8steering group muscles 115Steindler effect 262, 266Steindler flexor-plasty 266–8Steindler operation 82, 88Steindler transfer 126, 181, 306

modified 127, 128stepwise discriminant analysis 166–7sternocleidomastoid (SCM) muscles 59

in infants 159–60sternocleidomastoid to biceps transfer 271stretch injuries 52, 116stretching of BP during delivery 151, 152subclavian artery aneurysm, traumatic 31subclavian nerve 14subclavian vein 11subclavicular and axillary surgical approach 59subluxation

acquired 250complex 253, 256radius head 249shoulder 240, 241, 244typical posture 249

subluxation of distal radio-ulnar joint 276subscapular nerves 14subscapular slide/release 179, 180, 226–7,

230and latissimus dorsi transfers 233, 234

subscapular tendon lengthening 180, 181subscapularis muscle 23subscapularis operations 179–81subscapularis recession 255

obstacles to 256–7subscapularis tenotomy 180Sunderland’s classification 92superficial jugular vein 11superior trunk 3, 4, 12supinator muscle 28supraclavicular branch of BP 13supraclavicular fossa 84supraclavicular injuries 50supraclavicular lesions 57, 67

vs infraclavicular lesions 91supraclavicular nerves 86, 193supraclavicular–extrascalenius region 5suprascapular artery 11suprascapular nerve

repair 116rupture 100and scapula fractures 32transfers 222

suprascapular nerves 3, 13, 14, 194, 195compression of 49

innervation 21as landmark 59muscle innervation 22neurotization 85palsy 116

supraspinatus muscle 226innervation 21palsy of 108

sural nerve 134, 196, 323surface anatomy of BP 192–3surgery 190–202, 219

elbow/wrist/hand 181–2indications for 59, 167–70, 219for medial rotation contractures 255–8shoulder 178–9timing of 78–9, 219

surgical approach, Alnot’s 117surgical exploration 315–16surgical interventions 219–21surgical options 80–2surgical procedures 59, 295surgical techniques 51–5, 110–12

forearm supination deformity 281in OBPP 189–202shoulder muscle transfers 228–30subscapular release 226–7

sutures/suturing 95, 192, 196, 232, 245, 323direct 101end-to-end 51, 52end-to-side 52, 208, 317marking 269, 271of vessel grafts 324

synostosis 286, 287syrinx 41

T1 166avulsion 160

techniquefree muscle transfers 139–45

first 140–1secondary shoulder surgery 141–2

temperature during anaesthesia 190temporary muscle atrophy 305–6tendon lengthening 181, 230, 244tendon transfer 82tendon transfers 230, 231, 242–3, 301, 306

for BP trauma in children 316for elbow flexion 124multiple 295in OBPP 214shoulder 115–21, 240

tendon transfers to shoulder 115–21, 239–46tenodesis 82, 143, 301tenolysis 143, 145tenorrhaphy 141tenotomy 244, 306tension of muscle 134, 140, 142, 232

adjustment 141, 269teres major muscle 243

innervation 22, 236teres minor muscle 22terminal branches of BP 14, 91–104Test Scores 168, 169

INDEX 335

Page 345: Brachial Plexus Injuries

thoracic nerves 3, 152, 155anatomy 4, 7, 8, 12, 13, 14, 15muscle innervation 22–9and pain 19, 20traction lesions 47

thoracodorsal artery 131thoracodorsal nerve 14, 22, 124thoraco-scapular muscles 107, 108thumb function 133, 280thumb-in-palm deformation 176timing of treatment 42, 102, 185

arthrodesis 111contractures 242elbow paralysis 265, 266forearm supination deformity 281OBPP 206, 219for paralysis of upper extremity 123reconstructive procedures 139shoulder paralysis in neonates 226shoulder surgery 242of war injuries 322

Tinel’s sign 53, 79, 80, 93Tissucol 53total palsies 152, 155, 276

with avulsion of all roots 61with avulsion of lower roots 60–1in infants 160, 168see also complete palsies

total paralysis 289see also complete paralysis

total plexus injury 201, 202obstetric 235results of repairs 213–15

total plexus palsy 165, 293, 297traction

of BP 33and compression 49foetal 152lesions 47–9resistance to 7

transfers in hand 182transverse cervical artery/vein 194transverse cervical nerve 98transverse process of cervical vertebrae 5, 31transverse-radicular ligament 6transverso-pleuro muscles 9trapezium, palsy of 108trapezius muscle 120

innervation 23transfers 118, 119, 235, 243, 245

for grade 0 shoulders 236surgical techniques 229

traumatic arachnoid cysts see meningocelestraumatic birth 153, 154traumatic BP injuries 315treatment

of co-contraction 305–6, 309, 309of complete palsy 68–73

Trendelenburg position 49triceps 126, 262triceps brachii muscle 29triceps to biceps transfer 128, 271trumpet sign 264, 304

ulna dislocation 288ulnar deviation of wrist 275–6, 279–81, 284,

289ulnar nerve 13, 102–3

in biceps neurotizations 54muscle innervation 24, 25, 26, 28outcome of nerve repair 205in Steindler flexor-plasty 267, 268

upper humeral fractures 20upper plexus palsy 165upper respiratory tract infections 189upper root lesions 205, 211–12, 213upper subscapular nerve 23upper trunk 11, 191, 195

vascular compromise 145vascular damage 49vascular investigation 322vascular lesions 31, 96–7, 324

‘silent’ 36vascularized grafts 53vascularized nerve grafts 81vaso-motor paralysis 92Velpeau quadrilateral 14ventral ramus 23ventral terminal branches 15vertebral artery 7vertex delivery 251vessel repair 324volar dislocation 279, 284, 299Volkmann’s contracture 133, 263

waiter’s tip hand position 152, 218Waldeyer’s vertebral triangle 7, 8Wallerian degeneration 40, 79, 91, 92war injuries 101, 321–5whole-limb with paralysis sequelae 279whole-limb with partial recovery sequelae 278wrist arthrodesis 82, 107wrist deformity 280wrist dorsiflexion with ulnar deviation 289wrist extension 215, 267wrist flexion after complete palsy 72wrist fusion 297wrist ulnar deviation 299

X-rays 31–3, 109, 110, 160

336 INDEX


Recommended