ORIGINAL ARTICLE

Posterior thoracic osteotomies

Ferran Pellise • Alba Vila-Casademunt •

European Spine Study Group (ESSG)

Received: 6 March 2014 / Accepted: 9 April 2014

� Springer-Verlag France 2014

Abstract Spinal osteotomies are used to treat partially

flexible and fixed deformities. Fixed thoracic spinal

deformities have been traditionally treated with anterior

release and posterior correction with fusion. In recent

decades, it has been shown that posterior-only osteotomies

might be sufficient to achieve proper deformity correction

with lower complication rates than with combined anterior

and posterior procedures. Different types of osteotomies

have been described to treat spinal deformities through a

single posterior approach. These include posterior column

osteotomies such as the Smith-Petersen osteotomy and the

Ponte osteotomy, and three-column osteotomies such as the

pedicle subtraction osteotomy, the posterior vertebral col-

umn resection and the posterior vertebral column decan-

cellation. In general, three-column osteotomies are most

commonly performed in the lumbar spine, where the vast

majority of reports have focused on. They can also be

performed in the thoracic spine in the setting of rigid

thoracic deformity. A progressive increase in complica-

tions has been reported with more aggressive osteotomies.

The aim of this article was to describe the most common

posterior spinal osteotomies used to treat adult thoracic

spinal deformities, with special emphasis on the technical

aspects, complications and outcomes, based on current

publications and European Spine Study Group (ESSG)

data.

Keywords Thoracic spine osteotomy � Posterior column

osteotomy � Pedicle subtraction osteotomy � Posterior

vertebral column resection � Spinal osteotomy

complications

Introduction

Normal alignment of the thoracic spine is straight in the

coronal plane and kyphotic in the sagittal plane. Studies

using standing lateral radiographs of the entire spine have

reported values for both segmental and total kyphosis and

lordosis in an asymptomatic population [1–9]. It is well

known that each person has their own physiological upright

standing posture and individual pattern of spinopelvic and

sagittal alignment [9]. Although a ‘‘normal’’ mean value

cannot be determined, some authors describe normal tho-

racic kyphosis as ranging from 20� to 40� and normal lumbar

lordosis as ranging from 30� to 80� and always being pro-

portionally balanced. With normal ageing, these physiolog-

ical curves change: thoracic kyphosis increases and lumbar

lordosis decreases, resulting in an anterior shift of the C7

plumb line [8, 10, 11]. The most common aetiologies of adult

spinal deformities in the thoracic region are idiopathic,

congenital, post-traumatic or inflammatory [12–15].

Surgical strategies for treating spinal deformities are

determined based on severity, flexibility, apex location and

deformity shape [16]. Osteotomies are currently used to

treat partially flexible and fixed deformities. Thoracic

deformities with an apex at the cord level are more tech-

nically challenging than lumbar deformities with an apex at

the cauda equina level due to the risk of spinal cord

compromise. Osteotomies should be performed using

intraoperative neuromonitoring. Long, sweeping, rounded

deformities are easier to treat with multiple osteotomies.

F. Pellise (&)

Unitat de Raquis, servei de cirurgia ortopedica i traumatologia,

Hospital Universitari Vall d’Hebron, Passeig Vall d’Hebron

119-129, 08035 Barcelona, Spain

e-mail: 24361fpu@comb.cat

A. Vila-Casademunt

Vall d’Hebron Institut de Recerca, Passeig Vall d’Hebron

119-129, 08035 Barcelona, Spain

123

Eur J Orthop Surg Traumatol

DOI 10.1007/s00590-014-1463-7

Surgical correction can be performed on multiple spinal

segments. Short, angular deformities require a more

aggressive approach in which surgical correction is con-

centrated in a few spinal segments.

Fixed thoracic spinal deformities have been traditionally

treated with anterior release and posterior correction with

fusion [17, 18]. In recent decades, different authors have

shown that posterior-only osteotomies with instrumented

fusion is sufficient to achieve proper deformity correction

with lower complication rates than with combined anterior

and posterior procedures [19, 20].

Different types of osteotomies have been described to

treat fixed or partially flexible spinal deformities through a

single posterior approach. These include posterior column

osteotomies such as the Smith-Petersen osteotomy (SPO)

and the Ponte osteotomy, and three-column osteotomies

such as the pedicle subtraction osteotomy (PSO), the pos-

terior vertebral column resection (PVCR) and the posterior

vertebral column decancellation (PVCD). The variable use

of the terminology to define osteotomy type has led to a

more systematic grading of the complexity of posterior

spinal osteotomies [21]. There is a paucity of information

describing the use of osteotomies to correct adult thoracic

spinal deformities.

In July 2010, a group of European surgeons founded the

European Spine Study Group (ESSG) to evaluate clinical

outcomes for conservative and surgical adult spinal

deformity (ASD) treatment. To do this, the ESSG members

share a comprehensive prospective database into which

data from all consecutive patients meeting pre-established

inclusion criteria and signing informed consent is uploa-

ded. ESSG inclusion criteria are age 18 or over at the time

of surgery or initial consultation, having a spinal coronal

curvature of C20� or a sagittal vertical axis (SVA) of

[5 cm or a pelvic tilt of [25� or thoracic kyphosis of

[60�. By January 2014, the ESSG database included a

total of 897 patients (mean age 49.9 years, 241 females)

from six different European sites: the Ankara Spine Center

in Ankara (Turkey), the Vall d’Hebron Hospital in Barce-

lona (Spain), the CHU de Bordeaux university hospital

centre in Bordeaux (France), the Acibadem Maslak Hos-

pital in Istanbul (Turkey), La Paz University Hospital in

Madrid (Spain) and the Schulthess Klinik in Zurich

(Switzerland). Of these 897 patients, 586 were not surgical

candidates and 311 had undergone surgery. Spinal osteot-

omies had been performed on 145 patients. Thoracic os-

teotomies had been performed on 33 (22.7 %) patients

(mean age 42.1 years, 18 women).

The aim of our paper was to describe the most common

posterior spinal osteotomies used to treat adult thoracic

spinal deformities, with special emphasis on the technical

aspects, complications and outcomes, based on current

publications and ESSG data.

Schwab Grade 2, Smith-Petersen osteotomy (SPO)/

Ponte osteotomy

In a Schwab Grade 2 osteotomy, both inferior and superior

facets of an articulation at a given spinal segment are

resected, as well as the ligamentum flavum; other posterior

elements of the vertebra including the lamina, or the spi-

nous processes may also be resected [21]. The Smith-Pet-

ersen osteotomy can be considered a Schwab Grade 2

osteotomy. In most cases, an SPO is described as an

osteotomy in which the posterior column is resected

between the facet joints at one or more levels to create

additional lordosis [16]. SPOs were first described in 1945

and were used principally for ankylosing spondylitis [22–

24]. Correction was obtained by rupturing and opening the

anterior column. Schwab Grade 2 osteotomies incorporate

resection of bone beyond what was originally described by

Smith-Petersen. The Ponte procedure is also considered a

Schwab Grade 2 since it corresponds to the resection of

multiple facets joints and spinous processes and involves

substantial bone and ligament resection for deformity

correction. It was first described by Alberto Ponte in 1984

to treat flexible Scheuermann’s kyphosis [25].

In all of these osteotomies, the middle and anterior

columns are such that the anterior column opens through

the disc space when closing the osteotomy posteriorly to

correct kyphotic deformities or laterally to correct scoliotic

deformities. It is therefore necessary to have a potentially

mobile disc space to achieve correction with an SPO or

Schwab Grade 2 osteotomy. This is usually done at mul-

tiple levels to create gradual correction in a large number

of spinal segments.

The ideal candidates for multiple Grade 2 osteotomies

are patients who have long, sweeping, rounded deformities

with some anterior column mobility, such as Scheuer-

mann’s kyphosis and idiopathic thoracic scoliosis [16]

(Figs. 1, 2). Schwab Grade 2 SPOs may provide 5�–10�lordosis or 1� per millimetre of correction [16]. The extent

of correction will be proportional to and depend on disc

space or anterior column mobility. Thirteen (13) ESSG

patients underwent multiple Schwab Grade 2 thoracic

SPOs for the following aetiologies: Scheuermann’s ky-

phosis (5), idiopathic kyphoscoliosis (6), proximal junc-

tional kyphosis (1) and syndromic kyphoscoliosis (1). They

experienced a mean correction of 7� per segment with an

overall deformity correction rate of 43.8 %. Schwab Grade

2 SPOs can also be used concurrently with more focused,

aggressive three-column osteotomies to obtain some addi-

tional correction.

Surgical technique starts with complete removal of the

spinous process and inferior facets of the superior vertebra.

Careful opening of the epidural space is followed by a

complete flavectomy and the bilateral removal of the

Eur J Orthop Surg Traumatol

123

superior articular facets. In kyphotic deformities, the spinal

cord will be located anteriorly and subsequently, SPOs can

be performed easily. In scoliotic deformities, the cord will

be located in the concavity of the curve and care should be

taken when removing the concave articular facets. Concave

pedicle screws may be inserted before or after completing

the osteotomy. If inserted before, the screw head may

hinder concave facet joint removal in scoliotic cases. If

inserted once, the osteotomy is completed, pedicle screws

can be placed under direct visual control, but care should

be taken to protect neural structures during screw insertion.

In kyphotic deformities, correction is obtained by closing

the posterior column at the osteotomy sites. This can be

done through sequential segmental compression and pos-

terior osteotomy closure or by thoracic spine extension

with a cantilever manoeuver, or with a combination of

Fig. 1 Scheuermann’s, pre-operative and post-operative X-rays. PSFI T2-L1 with multiple T5–T12 SPOs

Fig. 2 Idiopathic scoliosis, pre-operative and post-operative X-rays. PSFI T2-L1 with multiple T6–T10 SPOs

Eur J Orthop Surg Traumatol

123

both. Scoliotic deformity correction is generally obtained

by closing, compressing or shortening the convexity and

opening the concavity. This can be done using periapical

derotation, translation and compression–distraction

manoeuvres.

The Morbidity & Mortality (M&M) Committee of the

Scoliosis Research Society (SRS) studied and reported

short-term complications associated with the surgical

treatment of thoracolumbar fixed sagittal plane deformities

[26]. Osteotomies were performed in 402 of the 578 ana-

lysed cases and included 135 SPOs, 215 PSOs and 18

VCRs. Cases with thoracic osteotomies were not specifi-

cally identified. The SPO procedure complication rate was

28.1 %, and the most common complications were dural

tears (6.7 %). Neurological deficits developed in 3.7 %.

These data were comparable with the results extracted from

the Cho et al. report [27], in which they reported an SPO

complication rate of 36.6 % and a new neurological deficit

rate of 3.3 %. Our data on 13 patients with Grade 2 tho-

racic osteotomies demonstrated two transient, minor sen-

sory deficits and one posterior arch fracture without further

consequences. No other osteotomy-related complications

were identified.

Schwab Grade 3, thoracic pedicle subtraction

osteotomy (TPSO)

A pedicle subtraction osteotomy (PSO) is a V-shaped

resection through the posterior structures, pedicles and

vertebral body. The hinge point is located anteriorly. In

most cases, the resection is performed entirely through

bone [16]. This type of osteotomy is categorised as Schwab

Grade 3. With this technique, no anterior column length-

ening is performed. A variant of this involves resecting the

disc space above, but this variation would then be con-

sidered a Schwab Grade 4 osteotomy. Michele and Krueger

[28] or Thomasen [29] have been alternately credited with

developing PSO.

Although PSO is a commonly used and reported tech-

nique for lumbar flat back syndrome, it can also be used for

thoracic spine deformities [12–15, 30]. The ideal candi-

dates for a PSO are patients with fixed sagittal (Fig. 3a)

sharp angular deformity, such as what is seen in cases of

post-traumatic kyphosis [16]. Severe, symptomatic roun-

ded and rigid thoracic kyphosis ([80�), for which poten-

tially less-risky osteotomies such as SPOs are not feasible

due to the lack of sagittal curve flexibility, may also be a

good indication for thoracic PSO [12]. The unilateral PSOs

commonly used to treat hemivertebrae in paediatric

patients are seldom used in adults, but could be used to

treat coronal plane deformities. O’Shaughnessy et al. [12]

reported on 15 patients with proximal junctional kyphosis

(n = 5), ankylosing spondylitis (n = 3), post-traumatic

kyphosis (n = 2), idiopathic kyphoscoliosis (n = 3), post-

laminectomy kyphosis (n = 1) and degenerative kypho-

scoliosis (n = 1) who underwent 25 thoracic PSOs. Thir-

teen patients underwent thoracic PSO only to correct a

sagittal deformity and two patients underwent PSO com-

bined with another three-column osteotomy. Yang et al.

[13] reported on seven patients who underwent thoracic

PSO to correct degenerative kyphoscoliosis (5), late-onset

adolescent idiopathic scoliosis (1) and ankylosing spon-

dylitis (1). Lafage et al. [15] analysed eighteen (18)

patients who underwent thoracic PSO. Lehmer et al. [30]

reported the results of four distal thoracic PSOs as part of a

larger series of primarily lumbar PSOs. The ESSG database

includes nine patients who underwent nine thoracic PSOs

to treat thoracolumbar Scheuermann’s kyphosis (4), post-

traumatic kyphosis (4) and idiopathic kyphoscoliosis (1). In

eight cases, we performed a single PSO, and in one patient

with Scheuermann’s kyphosis, the thoracic PSO was

associated with six SPOs (Fig. 3b).

PSOs are typically performed in the apical region of the

deformity [12]. If a PSO is performed in the thoracic spine,

it is important to refrain from retracting the thecal sac and

to resect portions of ribs on both sides in an effort to

approach the vertebral body more laterally [16]. The

proximal 2-to-3 cm of the ribs that articulate with the discs

above and below the planned PSO are exposed and cut. The

thoracic nerve root may or may not be cut. The lateral

vertebral body wall is bilaterally dissected. A posterior,

V-shaped resection that includes the pedicles and part of

the vertebral body is performed. The medial aspect of the

pedicles and the posterior vertebral body wall are preserved

until the end. The final structure resected is the posterior

vertebral body wall. This is performed using an impaction

technique [12]. When closing a PSO, it is necessary to

observe the spine for subluxation. The use of temporary

rods and progressive correction through sequential rod

exchange are advisable [31]. Safe posterior osteotomy

closure should be performed without excessive dural

kinking. Twenty (20) to 25 mm of posterior laminar clo-

sure may be tolerated in a thoracic PSO without adverse

neurological consequences [12]. It is advisable to keep the

canal centrally enlarged when performing a thoracic PSO

[16].

In general, lumbar and thoracic PSOs have a

39.1–58.5 % complication rate [26, 27]. Dural tears, new

neurological deficits and pseudoarthrosis are the most

commonly reported complications. Procedures with a PSO

have been found to have almost twice the rate of new

neurological deficits as procedures with an SPO (7 vs

3.7 %). O’Shaughnessy et al. [12] reported on three tho-

racic PSOs out of 25 that were associated with early

structural complications, including two cases of vertebral

Eur J Orthop Surg Traumatol

123

collapse and one case of segmental translation associated

with osteotomy closure. In the patient who experienced

segmental translation, there was a decline in intraoperative

somatosensory-evoked potentials at the time of the oste-

otomy closure. This was the only case of intraoperative

neurophysiological change. Yang et al. [13] reported one

surgical complication and one late-onset complication

requiring corrective surgery in the thoracic PSO group.

However, none of them were related to the resection

technique. None of the ESSG patients undergoing thoracic

PSO sustained complications directly related to the

osteotomy.

Lumbar PSOs are associated with significant improve-

ments in local, segmental and global measures of sagittal

balance, whereas thoracic PSOs are only associated with

local improvement and fail to produce the same degree of

segmental or global correction [13]. A lumbar PSO will

produce approximately 20�–35� of lordosis. A thoracic

PSO will produce approximately 10�–30� of lordosis [12,

13, 16, 30, 32]. Our patients achieved a 73.9 % overall

deformity correction rate, with a mean 26.6� correction per

thoracic PSO. The extent of thoracic PSO correction may

be dependent on the region in which the osteotomy is

performed [12]. As one moves from distal to proximal,

vertebral bodies become shorter and more triangulated in

the axial plane. Shorter vertebral bodies may offer less

corrective potential than longer ones. O’Shaughnessy et al.

[12] recommended PVCR for corrections between T2 and

T10. PVCR was found to be better than PSO in terms of

stability during closure and offering a more robust and

predictable amount of correction for a single osteotomy. In

a PSO, excessive posterior shortening may result in buck-

ling of the dura and spinal cord. The more posteriorly

located the hinge is, the lesser the need is for spinal cord

shortening, and thus, the safer the correction is. PVCR has

a more posterior hinge and provides correction with less

posterior shortening. In post-traumatic cases, an extended

PSO or Schwab Grade 4 osteotomy is very frequently

needed to ensure good bone contact and anterior column

fusion. In cases where most of the vertebral body has

already collapsed due to the fracture, standard PVCR or

unilateral PVCR preserving the posterior wall and half

posterior arch (described later) may be a better option than

PSO. Lafage et al. [15] compared the correction obtained in

18 thoracic PSOs and 23 thoracic PVCRs and found no

significant difference. Mean focal sagittal correction was

14.4� in all patients and 20.8� in patients with primarily

sagittal or multiplanar deformities. The mean correction

obtained with PSO was 12.8� (SD 14.4�) and with PVCR

was 15.6� (SD 14.9�).

Schwab Grade 5 and Grade 6, thoracic posterior

vertebral column resection (PVCR)

Vertebral column resection (VCR) is an osteotomy of all

three columns of the spine and is normally reserved for

severe rigid spinal deformities. VCRs provide the transla-

tion and shortening necessary to correct multiplanar

deformities [31]. VCR refers to completely resecting one or

more vertebrae, including posterior elements, pedicles and

the entire vertebral body with the discs above and below.

Reconstruction of both the posterior, and frequently, the

anterior column is usually necessary with a VCR. VCR was

initially suggested for tumours. This type of resection can

be included within Schwab osteotomy Grade 5. When more

than one vertebral body is resected, it may be included in

Grade 6. Despite the satisfactory outcomes of the two-stage

anterior–posterior technique, several series of patients have

been treated with posterior-only VCR (PVCR) in recent

years [20]. The osteotomy is usually performed at the apex

of the deformity to increase the effectiveness of the

resection. Despite both being circumferential posterior

osteotomies, the primary difference between a PVCR and a

Fig. 3 a Fixed Scheuermann’s, pre-operative Fulcrum X-ray. b Fixed Scheuermann’s, pre-operative and post-operative X-rays. PSFI T2-Iliac

with T10-L4 SPO and T12 PSO

Eur J Orthop Surg Traumatol

123

PSO is that during a PVCR, both the spinal cord and the

impinging wedge fragment are identified under lateral direct

vision, thus confirming complete decompression. Therefore,

PVCR can be safely performed at the level of the cord. This

provides the powerful translation and shortening necessary to

correct extensive rigid deformities. VCR is a complete de-

stabilising spinal osteotomy, for which the amount of cor-

rection is only limited by the spinal cord. VCR indications

include severe, rigid biplanar deformities with coronal and

sagittal misalignment and sharp angular kyphosis [15, 20].

The majority of reports concerning thoracic PVCR include

paediatric patients [31]; few are focused on adult thoracic

PVCR [15]. Eleven patients in the ESSG database underwent

thoracic PVCR for idiopathic kyphoscoliosis (3), congenital

kyphoscoliosis (5), post-traumatic kyphosis (2) or Scheuer-

mann’s kyphosis (1) (Fig. 4a–c).

After typical midline posterior spinal exposure, seg-

mental instrumentation is performed. Vertebrae are then

instrumented both cephalad and caudally to the osteotomy

site. PVCR is performed through a bilateral costotrans-

versectomy. We recommend dissecting circumferentially

and exposing the anterior aspect of the spine as much as

possible while maintaining the spinal canal closed and the

cord protected. The transverse processes, head of the ribs

and proximal portion of the ribs are excised. Using the

lateral wall of the pedicles as guidance, the parietal pleura

is detached and reflected back from the anterior aspect of

the vertebral body, exposing both the apical and the

immediately adjacent vertebrae circumferentially until the

anterior surface of the vertebral body is palpable. The

sharper and more angular the thoracic kyphosis is, the

easier the circumferential exposure of the spine will be

through a posterior-only approach. A temporary rod is

placed. Following a total laminectomy and bilateral total

foraminal unroofing under direct visual control, the pedi-

cles and the lateral portion of the vertebral body are

Fig. 4 a Congenital scoliosis, pre-operative clinical photographs. b Congenital scoliosis, post-operative clinical photographs. c Congenital

scoliosis. Pre-operative and post-operative X-rays. PSFI T2-Iliac with T10-L4 SPO and T12 PSO

Eur J Orthop Surg Traumatol

123

removed. The vertebral body and discs are removed in a

piecemeal fashion, keeping a thin shell of posterior verte-

bral wall beneath the dural sac. This portion of the pos-

terior wall is the last portion of vertebrae to be removed. In

principle, the minimal amount of resection necessary for

safe, adequate correction of the deformity should be per-

formed. Deformities are corrected either with in situ rod

bending or the sequential exchange of temporary rods with

progressively less deformed pre-contouring [31]. The ver-

tebral column is initially shortened through compression.

The deformity is gradually corrected with repeated addi-

tional compression to shorten the vertebral column until the

exposed cord looks redundant. After correcting the defor-

mity, any residual anterior interbody gap should be filled

with a cage and bone graft to reconstruct anterior column

support and improve the chances of fusion [20]. We would

recommend using a final four-rod construct to avoid early

rod breakage in case bony healing is delayed. Cord pro-

tection should be considered if, after osteotomy closure, the

spinal cord is still the posterior-most spinal structure.

Protection can be provided with bone (rib or allograft) or

metal [33] (Fig. 5).

Papadopoulos et al. [31] reported intraoperative moni-

toring changes in 22 % of patients due to hypotension or

excessive cord manipulation [20]. All but one of these

patients returned to normal after appropriate manoeuvres.

Common strategies to avoid cord damage include main-

taining blood supply by preserving as many neurovascular

bundles as possible and avoiding hypotensive anaesthesia.

Excessive cord shortening may be dangerous. In an

experimental study, Kawahara et al. [34] demonstrate that

during acute spinal shortening, the spinal cord and dura

pass through three phases. Phase 1 is considered safe and

indicates that the vertebral segment is shortened by less

than one-third. There is no dural sac or spinal cord defor-

mity. In Phase 2, a warning stage, the vertebral segment is

shorted by one-third to two-thirds and is accompanied by

shrinking and buckling of the dural sac, but no spinal cord

deformity. In Phase 3, a dangerous stage, vertebral

shortening is in excess of two-thirds and is accompanied by

spinal cord deformity and compression of the buckled dura.

In Phase 3, cord damage is highly probable. Spinal cord

blood flow is markedly increased in Phases 1 and 2.

The SRS M&M reported an overall complication rate

for procedures with a VCR of 61.1 % [26]. Suk reported a

34.3 % complication rate in 70 spinal deformity patients

who had a total of 143 resected vertebrae (76 thoracic and

67 lumbar). The most serious complications were two

complete cord injuries. Kim et al. [35] reviewed 233

patients treated with PVCR and found an overall incidence

of complications of 40.3 %. Xie et al. [36] reported 46

intraoperative and post-operative complications in 18 out

of 28 patients receiving PVCR. There were five cases of

neurological complications including one case of late-onset

paralysis and four cases of thoracic nerve root pain, all of

which resolved during the early follow-up period. Non-

neurological complications occurred more often in ky-

phoscoliotic patients (41 complications). Eleven (11)

patients included in the ESSG database underwent thoracic

PVCR. We identified 19 complications, five of which were

considered major. Two patients suffered rod breakage that

required corrective surgery and one other patient experi-

enced unresolved proximal junctional kyphosis. Two

patients suffered three neurological complications, two of

which were transient and one of which ended up being

permanent paralysis. Overall reported deformity correction

obtained with PVCR ranges from 45 to 68 % [15, 20, 31,

35, 36]. Our patients achieved a mean deformity correction

of 67.6 %, 32.9� per osteotomy.

Other types of thoracic osteotomies

Posterior vertebral column decancellation (PVCD)

Wang [37] suggested a modified PVCR to reduce spinal

column instability, blood loss and excessive cord manip-

ulation. In PVCD, pedicles of apical vertebrae are enlarged

using a high-speed drill. Then, multilevel vertebral body

decancellation and residual disc removal are performed.

After vertebral body decancellation, the posterior elements

are removed. Deformity correction is obtained through

osteoclasis of the concave cortex and compression of the

convex cortex using posterior pedicle instrumentation. In

kyphotic deformities, anterior cortex osteoclasis is

achieved through gentle extension. Autogeneous bone graft

is then applied to fill gaps and ensure bony fusion. In most

cases, segmental vessel exposure is not needed. Wang

reported a mean of 2.2 decancelled vertebrae in 45 patients

with severe sharp angular spinal deformities (29 cases of

congenital kyphoscoliosis and 16 cases of Pott’s disease)

who underwent vertebral decancellation, and an average

Fig. 5 Metal cord protection

Eur J Orthop Surg Traumatol

123

post-operative kyphosis correction of 82.2�. All patients

showed evidence of solid fusion at the last follow-up.

There was a reported surgical complication rate of 17.8 %.

These complications included one case of CSF leak, one

case of deep wound infection, one case of epidural

haematoma and four cases of transient neurological

deficits.

Unilateral PVCR preserving posterior wall and half

posterior arch

In some cases, especially in elderly patients with post-

traumatic kyphosis in which deformity is due to vertebral

body collapse with some movement expected at the disc

level, circumferential reconstruction can be performed

through a unilateral costotransversectomy (Fig. 6). Once

the unilateral costotransversectomy approach is performed

and pedicle screws have been inserted proximally and

distally to the osteotomy site, the ipsilateral posterior arch

is removed exposing the dural sac and the ipsilateral nerve

roots. Ligation of the interfering nerve root is recom-

mended to safely mobilise the spinal cord and create

enough space for bony and soft tissue removal. A tempo-

rary rod is inserted contralaterally. The discs above and

below the resected vertebral bodies should be identified

and removed taking care not to damage the end-plates of

the adjacent vertebral bodies. The collapsed vertebral body

is gradually removed, preserving most of the posterior

wall. Under direct visual control, the anterior vertebral

body wall can be removed using a Kerrison rongeur. At this

stage, rod exchange is performed to progressively correct

the deformity. Anterior release is amplified until full cor-

rection is obtained. The anterior gap is then filled with a

cage and autologous bone to restore anterior column sup-

port and obtain interbody fusion. Compression is provided

to capture the cage and the ipsilateral rod is inserted.

Contralateral posterior arches are decorticated and the bone

graft added. This procedure can be done to reconstruct one

or more spinal segments (Fig. 7).

Conclusions

Spinal osteotomies are well known and highly effective

surgical techniques used to correct partially flexible and

rigid spinal deformities. The recent SRS M&M [26] report

Fig. 6 CT scan. Unilateral PVCR preserving posterior wall and half

posterior arch

Fig. 7 Post-traumatic kyphosis, pre-operative and post-operative X-rays

Eur J Orthop Surg Traumatol

123

clearly demonstrates a progressive increase in complica-

tions with more aggressive osteotomies. There is still very

limited data on thoracic three-column osteotomies in the

adult population. A careful case-by-case analysis of the

magnitude, location and flexibility of deformities is needed

to understand what type of osteotomy or combination of

osteotomies will accomplish various surgical goals. It is

essential to take into consideration the significantly higher

complication rates associated with osteotomies as well as

the relevance of patient characteristics such as age and

comorbidities in determining the best surgical strategy for

each case type and for each patient.

Acknowledgments We acknowledge DepuySynthes Spine for its

funding support.

Conflict of interest Dr. Pellise reports grants from Depuy Synthes

Spine, during the conduct of the study; grants and personal fees from

Depuy Synthes Spine, grants from K2 M, personal fees from Biomet,

outside the submitted work. Dr. Vila-Casademunt reports grants from

Depuy Synthes Spine, during the conduct of the study; grants from

Depuy Synthes Spine, outside the submitted work. Dr. (ESSG) reports

grants from Depuy Synthes Spine, during the conduct of the study;

grants from Depuy Synthes Spine, outside the submitted work.

References

1. Stagnara P, Mauroy JC, Dran G, Gonon GP, Costanzo G, Dimmet

J, Pasquet A (1982) Reciprocal angulation of vertebral bodies in a

sagittal plane: approach to references for the evaluation of ky-

phosis and lordosis. Spine 7:335–342

2. Bradford DS, Moe JH, Montalvo FJ, Winter RB (1974) Scheur-

mann’s kyphosis and roundback deformity; results of Milwaukee

brace treatment. JBJS 56A:740–758

3. Fon GT, Pitt MJ, Thies AC Jr (1980) Thoracic kyphosis: range in

normal subjects. Am J Radiol 134:979–983

4. Moe JH, Winter RB, Bradford DS, Lonstein JE (1978) Kyphosis;

general principles. Chap 13. Scoliosis and other spinal deformi-

ties. WB Saunders Company, Philadelphia, pp 325–330

5. Bernhard M, Bridwell K (1989) Segmental analysis of the sagittal

plane alignment of the normal thoracic and lumbar spines and

thoracolumbar junction. Spine 14(7):717–721

6. Jackson RP, McManus AC (1994) Radiographic analysis of

sagittal plane alignment and balance in standing volunteers and

patients with low back pain matched for age, sex, and size. Spine

19(14):1611–1618

7. Jackson RP, Hales Chris (2000) Congruent spinopelvic aligne-

ment on standing lateral radiographs of adult volunteers. Spine

25(21):2808–2815

8. Gelb DE, Lenke LG, Bridwell KH, Blanke K, McEnery KW

(1995) An analysis of sagittal spinal alignment in 100 asymptomatic

middle and older aged volunteers. Spine 20(12):1351–1358

9. Vialle R, Levassor N, Rillardon L, Templier A, Skalli W, Guigui

P (2005) Radiographic analysis of the sagittal alignment and

balance of the spine in asymptomatic subjects. JBJS 87A(2):

260–267

10. Schwab F, Lafage V, Boyce R, Skalli W, Farcy JP (2006) Gravity

line analysis in adult volunteers age-related correlation with

spinal parameters, pelvic parameters, and foot position. Spine

31(25):959–967

11. Hammerberg EM, Wood KB (2003) Sagittal profile of the

elderly. J Spinal Disord Tech 16:44–50

12. O’Shaughnessy BA, Kuklo TR, Hsieh PC, Yang BP, Koski TR,

Ondra SL (2009) Thoracic pedicle substraction osteotomy for

fixed sagittal spinal deformity. Spine 34(26):2893–2899

13. Yang BP, Ondra SL, Chen LA, Jung HS, Koski TR, Salehi SA

(2006) Clinical and radiographic outcomes of thoracic and lum-

bar pedicle substraction osteotomy for fixed sagittal imbalance.

J Neurosurg Spine 5:9–17

14. Kim KT, Suk KS, Cho YJ, Hong GP, Park BJ (2002) Clinical

outcome results of pedicle substraction osteotomy in ankylosing

spondylitis with kyphotic deformity. Spine 27:612–618

15. Lafage V, Smith JS, Bess S, Schwab FJ, Ames CP, Klineberg E,

Arlet V, Hostin R, Burton DC, Shaffrey CI, ISSG (2012) Sagittal

spino-pelvic alignement failures following three column thoracic

osteotomy for adult spinal deformity. Eur Spine J 21:698–704

16. Bridwell KH (2006) Decision making regarding Smith Petersen

vs pedicle substraction osteotomy vs vertebral column resection

for spinal deformity. Spine 31(19, suppl):5171–5178

17. Boachie-adjei O, Bradford DSV (1991) Vertebral column resec-

tion and arthrodesis for complex spinal deformities. J Spinal

Disord 4:193–202

18. Coe JD, Smith JS, Berven S, Arlet V, Donaldson W, Hanson D,

Mudiyam R, Perra J, Owen J, Marks MC, Shaffrey CI (2010)

Complications of spinal fusion for scheuermann kyphosis a report

of the scoliosis research society morbidity and mortalitycom-

mittee. Spine 35(1):99–103

19. Lee SS, Lenke LG, Kuklo TR, Valente L, Bridwell KH, Brenda

S, Blanke KM (2006) Comparison of Scheuermann kypho-

sis correction by posterior-only thoracic pedicle screw fixation

versus combined anterior/posterior fusion. Spine 31(20):2316–

2321

20. Suk SI, Kim JH, Kim WJ, Lee SM, Chung ER, Nah KH (2002)

Posterior vertebral column resection for severe spinal deformi-

ties. Spine 27(21):2374–2382

21. Schwab F, Blondel B, Chay E, Demakakos J, Lenke LG, Tro-

piano P, Ames C, Smith JS, Shaffrey CI, Glassman S, Farcy JP,

Lafage V (2014) The comprehensive anatomical spinal osteot-

omy classification. Neurosurgery 74(1):112–120

22. Smith-Petersen MN, Larson CB, Aufranc OE (1945) Osteotomy

of the spine for correction of flexion deformity in rheumatoid

arthritis. JBJS 27:1–11

23. Mc Master MJ (1985) A technique for lumbar spinal osteotomy in

ankylosing spondilytis. JBJS Br 67:204–210

24. Kostuik JP, Maurais GR, Richardson WJ et al (1988) Combined

single stage anterior and posterior osteotomy for correction of

iatrogenic lumbar kyphosis. Spine 13:257–266

25. Ponte A, Vero B, Siccardi G (1984) Surgical treatment of

Scheuermann’s hyperkyphosis. In: Winter R (ed) Progress in

spinal pathology: kyphosis. Aulo Gaggi, Bologna, Italy, pp 75–81

26. Smith JS, Sansur CA, Donaldson WF III, Perra JH, Mudiyam R,

Choma TJ, Zeller RD, Knapp DR Jr, Noordeen HH, Berven SH,

Goytan MJ, Oheneba Boachie-Adjei Shaffrey CI (2011) Short-

term morbidity and mortality associated with correction of tho-

racolumbar fixed sagittal plane deformity: a report from the

Scoliosis Research Society Morbidity and Mortality Committee.

Spine 36(12):958–964

27. Cho KJ, Bridwell KH, Lenke LG, Berra A, Baldus C (2005)

Comparison of Smith Petersen versus pedicle substraction oste-

otomy for the correction of fixed sagittal imbalance. Spine

30(18):2030–2037

28. Michele AA, Krueger FJ (1949) Surgical approach to the verte-

bral body. J Bone Joint Surg Am 31(4):873–878

29. Thomasen E (1985) Vertebral osteotomy for correction of ky-

phosis in ankylosing spondylitis. Clin Orthop Relat Res

(194):142–152

Eur J Orthop Surg Traumatol

123

30. Lehmer SM, Keppler L, Biscup RS et al (1994) Posterior trans-

vertebral osteotomy for adult thoracolumbar kyphosis. Spine

19:2060–2067

31. Papadopoulos EC, Boachie-Adjei O, Hess WF, Sanchez Perez-

Grueso FJ, Pellise F, Gupta M, Lonner B, Paonessa K, Faloon M,

Cunningham ME, Kim HJ, Mendelow M, Sacramento C, Yazici

M, and Foundation of orthopedis and Complex Spine, New York,

NY (2013):Early outcomes and complications of posterior ver-

tebral column ressection. Spine J. doi:10.1016/j.spinee.2013.03.

023

32. Ondra SL, Marzouk S, Koski T, Silva F, Salehi S (2006) Math-

ematical calculation of pedicle substraction osteotomy size to

allow precision correction of fixed sagittal deformity. Spine

31(25):973–979

33. Boachie-Adjei O (2010) Expert’s comment concerning Grand

Rounds case entitled Severe Kyphoscoliosis after primary echi-

nococcus granulosus infection of the spine (by M. Thaler, M.

Gabl, R. Lechner, M. Gstottner and C.M. Bach). Eur Spine J

19(9):1423–1424. doi:10.1007/s00586-010-1408-8

34. Kawahara N, Tomita K, Kobayashi T, Abdel-Wanis ME, Mura-

kami H, Akamaru T (2005) Influence of acute shortening on the

spinal cord: an experimental study. Spine 30(6):613–620

35. Kim SS, Cho BC, Kim JH, Lim DJ, Park JY, Lee BJ, Suk SI

(2012) Complications of posterior vertebral resection for spinal

deformity. Asian Spine J 6(4):257–265

36. Xie JM, Zhang Y, Wang YS, Bi N, Zhao Z, Li T, Yang H (2014)

The risk factors of neurologic deficits of one-stage posterior

vertebral column resection for patients with severe and rigid

spinal deformities. Eur Spine J 23(1):149–156. doi:10.1007/

s00586-013-2793-6

37. Wang Y, Lenke LG (2011) Vertebral column decancellation for

the management of sharp angular spinal deformity. Eur Spine J

20:1703–1710

Eur J Orthop Surg Traumatol

123