Chinese Medical Journal 2008; 121(2):99-104 99

Original article

Repair of brachial plexus lower trunk injury by transferring brachialis muscle branch of musculocutaneous nerve: anatomic feasibility and clinical trials ZHENG Xian-you, HOU Chun-lin, GU Yu-dong, SHI Qi-lin and GUAN Shi-bing Keywords: brachialis muscle branch; median nerve; finger flexion; nerve transfer; brachial plexus injury

Background There are few effective methods for treating injuries to the lower trunk of brachial plexus, and the curative effect is usually poor. The purpose of this study was to provide anatomic references for transferring the brachialis muscle branch of musculocutaneous nerve (BMBMCN) for selective neurotization of finger flexion in brachial plexus lower trunk injury, and to evaluate its clinical curative effects. Methods Microanatomy and measurement were done on 50 limbs from 25 adult human cadavers to observe the origin, branch, type of the BMBMCN and median nerve, as well as their adjacent structures. Internal topographic features of the fascicular groups of the median nerve at the level of the BMBMCN were observed. In addition, the technique of BMBMCN transfer for selective neurotization of finger flexion of the median nerve was designed and tested in 6 fresh adult human cadavers. Acetylcholinesterase (AchE) staining of the BMBMCN and median nerve was done to observe the features of the nerve fibers. This technique was clinically tried to restore digital flexion in 6 cases of adult brachial plexus lower trunk injury. These cases were followed up for 3, 6, 9 and 12 months postoperatively. Recovery of function, grip strength, nerve electrophysiology and muscle power of the affected limbs were observed and measured. Results The brachialis muscle was totally innervated by the musculocutaneous nerve (MCN). Based on the Hunter’s line, the level of the origin of the BMBMCN was (13.18±2.77)cm. AchE histochemical staining indicated that the BMBMCN were totally made up of medullated nerve fibers. At the level of the BMBMCN, the median nerve consistently collected into three fascicular groups as shown by microanatomy in combination with AchE stain. The posterior fascicular group was mainly composed of anterior interosseous nerves and branches to the palmaris longus. The technique was tested in six fresh cadavers successfully, except that stoma split occurred in one case. Five of the six cases recovered digital flexion 12 months after operation, and at the same time grip strength, muscle power, and nerve electrophysiology also recovered markedly. Conclusions The technique of transferring the BMBMCN for selective neurotization of finger flexion is anatomically safe and effective, with satisfactory clinical outcomes.

Chin Med J 2008;121(2):99-104

rachial plexus injury is a common and frequently-occurring clinical problem, the incidence

of which has been on the rise in recent years. At present, nerve transfer is the principal approach to repair brachial plexus injury.1,2 Compared with injuries to the upper trunk of brachial plexus, there are relatively few methods for repairing injuries to the lower trunk, which mainly include transferring intercostal nerve, phrenic nerve or contralateral C7 nerve for the purpose of reconstructing finger flexion, but the therapeutic outcome was unsatisfactory and less than 50% of the cases could resume muscle power above M3.3-5 There is therefore an urgent need for seeking a novel operation modality to help reconstruct finger flexion quickly and effectively. In the present study, we conducted an intensive study on the feasibility of transferring the brachialis muscle branch with musculocutaneous nerve (BMBMCN) to repair the median nerve and reconstruct finger flexion, based on which we applied this technique to clinical cases and observed the therapeutic outcome.

METHODS Anatomic study Subjects Used in this experimental study were 50 upper limbs from 25 adult human cadavers fixed in 10% formalin and 12 upper limbs from 6 adult fresh human cadavers (obtained from Department of Anatomy, Fudan University, Shanghai, China). Procedures A z-shaped longitudinal incision extending to the infra-

B

Department of Orthopedics, Changzheng Hospital, SecondMilitary Medical University, Shanghai 200003, China (Zheng XY,Hou CL and Guan SB) Department of Hand Surgery, Huashan Hospital, Fudan University,Shanghai 200040, China (GU YD and Shi QL) Correspondence to: Dr. GU Yu-dong, Department of Hand Surgery,Huashan Hospital, Fudan University, Shanghai 200040, China (Teland Fax: 86-21-62489999. Email: zhengxianyou@126.com) This study was supported by a grant from the Major State BasicResearch Development Program of China (973 Program, No.G1999054206).

Chin Med J 2008;121(2):99-104 100

Figure 1. Incision. Figure 2. Findings during the operation: BMBMCN (green arrow), lateral antebrachialcutaneous nerve (yellow arrow), and median nerve (blue arrow). Figure 3. Transferring of the BMBMCN (green arrow) to the posterior fascicular group of the median nerve (blue arrow) to reconstruct the finger flexion. Figure 4. The BMBMCN (green arrow) was anastomosised to the posterior fascicular group of the median nerve (blue arrow).

clavicular part was made to expose the cords and branches of the brachial plexus located in the medial part of the upper arm. After marking the natural position of the median nerve, microdissection and measurement were carried out carefully with assistance of the headset surgical loupe (2.5 power). Observations included: (1) The origin, branch, and type of the BMBMCN and median nerve, and their adjacent structures; (2) Level of the BMBMCN origin on the basis of the Hunter’s line; (3) Internal topographic features of the fascicular groups of the median nerve at the level of BMBMCN separated retrogradely from the trunk and its muscular branches of the median nerve in the arm. 10% acetic acid solution was applied to the nerve fibers in order to loosen the connective tissue and for the convenience of observation; (4) Acetylcholinesterase (AchE) staining of the BMBMCN and median nerve at the level of BMBMCN of the 12 limbs from the 6 fresh cadavers to observe features of the nerve fibers. Mimic operation Mimic operation of transferring the BMBMCN for selective neurotization of finger flexion was performed on the 12 limbs from the 6 fresh cadavers. Postoperative abduction, elevation, flexion, extension, external rotation and internal rotation of the shoulders, and flexion and extension of the elbows were tested passively to see whether there was any collapse of the nerve stomas. Every action was repeated at least three times. Clinical study Patients A retrospective study was done on the cases of injuries to the brachial plexus lower trunk from October 2001 to August 2004. The subjects included 5 males and one female whose age ranged from 28 to 49 years with a mean of 35 years. The course of injury ranged from 3 to 20 months with a mean of 6.88 months. According to the level and severity, the injuries were classified as simple lower trunk injury in 2 cases, and middle trunk + lower trunk complete injury in 4 cases including one with combined partial upper trunk injury whose elbow flexion

had been normal but electromyographic study showed a simple phase of the brachialis muscle. Surgical procedures A 10–15 cm long longitudinal incision was made in the middle and inferior part of the medial upper arm. The BMBMCN was found at 9–18 cm proximal to the Hunter’s line, and was disconnected after local blockade with 1% lidocaine. The dissociable length of the BMBMCN was then measured. Furthermore, the median nerve was detected in the medial of the musculocutaneous nerve. The fascicular group selectively innervating finger flexion of the median nerve was cut off after disassociation beneath the myelin sheath. Finally, the nerves were sutured by 6-8 stitches with 9-0 no-healing suture with the assistance of a 10 power surgical microscope (Figures 1–4). Follow-up and measurements The cases were followed up for 3, 6, 9 and 12 months postoperatively for observation of (1) functional recovery of the limbs, including impediment of elbow flexion, impediment of two-point discrimination and wrist flexion; (2) grip strength as measured by the Biometrics 5.0 dynamometer, based on which the grip recovery rate was calculated by comparing it with the unaffected side. (3) nerve electrophysiological parameters, including the latency phase (LAT), the maximum evoked potential amplitude (AMP), motor nerve conduction velocity (MNCV) of the median nerve, by using Dantec 2000M myo-electrophysiograph, based on which the conduction recovery rate was calculated by comparing them with the unaffected side; (4) muscle power of the thumb and the index finger using Medical Research Council (MRC) standard, based on which the fineness rate was assessed. According to MRC standard, muscle power of the thumbs and index fingers was measured 12 months after operation. Muscle power was classified as excellent (above M4), good (M2–M3), and poor (below M1). Statistical analysis The values obtained are expressed as mean ± standard deviation (SD), and data were analyzed by the SPSS 12.0

Chinese Medical Journal 2008; 121(2):99-104 101

Figure 5. The results indicated the median nerve consistently collected into three fascicular groups at least in the level of the BMBMCN, which were located at the anterior (thick green arrow), middle, and posterior (slender green arrow) parts of the median nerve trunk. The schematic diagram in the bottom left had shown the cross-sectional features in the level of the BMBMCN (green part referred to motor nerve fibers; and blue part referred to sensory nerve fibers.) Figure 6. The natural position of the median nerve must be marked preoperatively (arrow referred to the posterior of the median nerve), then the separation were performed. Figure 7. After the separation of the median nerve, the posterior fascicular group (blue arrow) was cut to be prepare to anastomosise with the BMBMCN (green arrow). Figure 8. The BMBMCN (green arrow) was anastomosised to the posterior fascicular group of the median nerve (blue arrow. Yellow arrow referred to the site of the anastomosis).

statistical software. The significance of the difference was calculated by analysis of variance (ANOVA). A P value <0.05 was considered statistically significant.

RESULTS Anatomic study Origin, branch and adjacent structures of the nerves The microanatomical study of the 50 limbs from the 25 adult human cadavers showed that the brachialis muscle was totally innervated by the musculocutaneous nerve and doubly innervated by the radial nerve in the anteroinferior part of the brachialis muscle in about 22 cases (44%). The length of BMBMCN was (4.84±1.29) cm, the diameter was (1.69±0.30) mm, and the number of branches was (1.16±0.55). According to the type of BMBMCN, there were three types as observed in our study: type I of single branch in 45 limbs (90%), type II of two branches in 3 limbs (6%), and type III of multiple branches in 2 limbs (4%). At the same time, microanatomy of the median nerve indicated that there were 44 limbs (88%) of superior position type, 4 limbs (8%) of inferior position type, and 2 limbs (4%) of simple-root type. At the level of the BMBMCN origin, the distance between the BMBMCN and the median nerve was about (2.23±0.97)cm, so that anastomosis of them could be done freely. Localization of BMBMCN Based on the Hunter’s line, the level of the BMBMCN origin ranged from 8.7 cm to 18.2 cm with a mean of (13.18±2.77) cm. Internal topographic features of the median nerve After retrograde separation, the internal topographic features of the fascicular groups of the median nerve were observed at the level of BMBMCN. The results indicated the median nerves consistently collected at least into three

fascicular groups at the level of BMBMCN, which were located at the anterior, middle, and posterior parts of the median nerve trunk. The anterior fascicular group was composed of the branches of the pronator teres and the flexor carpi radialis; the posterior fascicular group was composed mainly of the anterior interosseous nerve (the branch of the long flexor muscle of thumb, the branch of the proximal part of deep flexor muscle of fingers and the branch of the quadrate pronator muscle), the branches of the palmaris longus and the branch of the partial superficial flexor muscle of fingers; and the middle fascicular group was made up mostly of the sensory and motion branches to the hand (the trunk of the median nerve) and the branches of the partial flexor digitorum superficialis (Figure 5). Histochemical stain of AchE AchE histochemical stain indicated that the BMBMCN was totally made up of medullated nerve fibers, whose stain area was confined to the axon, and the medullary sheath was not AchE stained. According to AchE histochemical stain of the median nerve at the level of BMBMCN, the internal topographic features of the fascicular groups of the median nerve were observed using the Leica FW4000 imaging analysis system, which geared to the microanatomic findings. The anterior and posterior fascicular groups were mainly composed of motor nerve fibers, and the middle fascicular group was mostly made up of sensory nerve fibers. Results of the mimic operation Figures 6–8 show the procedure of the mimic operation of transferring the BMBMCN for selective neurotization of finger flexion in the 12 limbs from the 6 fresh cadavers. All cases were successfully incised through a 10–15 cm long incision without tension at the site of anastomosis, or occurrence of secondary injuries to the nerves and vessels. After operation, passive activity of the shoulders and elbows was tested. Stoma split was observed in one case, because the BMBMCN origin in this limb was located 16.2 cm above the Hunter’s line. This indicates the

Chin Med J 2008;121(2):99-104 102

importance of secure postoperative external fixation. Clinical study The 6 cases were followed up for 12–28 months with a mean of 16 months. No secondary injury to the nerves and vessels occurred during the operation, and no postoperative complication such as infection and ulceration in the limbs was noticed. Functional recovery of the affected limbs No impediment of elbow flexion or change of two-point discrimination in the thumbs and index fingers occurred in any case. Digital flexion was recovered in 5 of the 6 cases 12 months after operation. For the patient who failed to recover digital flexion within 18 postoperative months, contralateral C7 was transferred to repair the median nerve. Recovery of grip strength Grip strength recovered gradually with the lapse of time. There was a significant difference in the recovery rate of grip strength between the two groups using SNK method (P <0.05, Table 1). Electrophysiological study Nerve electrophysiological parameters recovered in varying degrees with the lapse of time (Table 1). There was no significant difference in the recovery rate of evoked potential of the median nerve between 3- and 6-month groups (P>0.05). There was no significant difference in the recovery rate of motor nerve conduction of the median nerve between 3- and 6-month groups (P>0.05). The differences were otherwise significant (P<0.05) using SNK method.

Table 1. The recovery rate of the grip power and nerve electrophysiological items after lower trunk

injury (%, mean ± SD) Time (m) Grip strength LAT AMP MNCV 3rd 26.14±3.56 18.68±2.32 28.53±2.12* 23.65±2.21#

6th 36.04±4.66 32.33±3.36 33.58±2.54* 32.98±3.57#

9th 48.91±5.78 44.68±2.56 53.26±1.24 49.87±2.2412th 66.04±3.21 63.89±4.12 70.32±3.66 60.23±5.03

All comparisons of the the grip strength recovery rate between every two groups were statistically significant, P<0.05. As for the electrophysiological study, significant differences were noted between the every two groups except for *the evoked potential amplitude recovery rate between third month and group sixth month, #and the motor nerve conduction velocity recovery rate between third month and sixth month (P>0.05). LAT, latency phase; AMP, amplitude; MNCV, motor nerve conduction velocity. Muscle power study In this study, 5 cases were excellent, and one case was poor.

DISCUSSION Significance of the technique At present, the principal approach to the treatment of brachial plexus injury is nerve transfer. A new chapter has been opened in the repair of brachial plexus injury since Dr. GU YD created phrenic and contralateral C7 nerve

transfer and applied them in clinical practices.6,7 For brachial plexus upper trunk injuries, there have been many repairing methods aimed at reconstructing the functions of the shoulders and elbows, including traditional or video-assisted thoracoscopical phrenic nerve, intercostal nerve, the fascicular groups of ulnar nerve or median nerve transfer to MCN to reconstruct elbow flexion; accessory nerve transfer to suprascapular nerve to reconstruct shoulder abduction with the anterior or posterior approach; and ipsilateral C7 nerve transfer to the upper trunk to reconstruct shoulder abduction and elbow flexion.8-15 On the whole, there are numbers of methods to choose for the repair of upper trunk injuries. The curative effect is usually good, and 50%–84.6% cases are able to recover muscle power above M3. But there have been fewer methods for the repair of injuries to the brachial plexus lower trunk, which mainly include intercostal nerve, phrenic nerve or contralateral C7 nerve transfer to reconstruct finger flexion. In addition, the curative effect is poor, and fewer than 50% cases can recover muscle power above M3. Patients with lower trunk injury often lose most of their hand functions, especially paralysis of the flexion muscle of the arm, impairment of finger flexion, and incapacitation of the intrinsic muscles of the hand. It is therefore urgent to seek a new method of reconstructing finger flexion quickly and effectively.16,17 Gu YD et al advanced the idea of constructing finger flexion by transferring the BMBMCN from the upper trunk. But more deep-going study on the anatomic feasibility of this technique is required, and more clinical cases are needed to testify the clinical feasibility and effectiveness of this technique. Anatomic feasibility of the technique Firstly, the components of the nerve fibers in the BMBMCN and internal topographic features of the fascicular groups of the median nerve at the level of BMBMCN determine the effectiveness of the new technique of transferring the BMBMCN for selective neurotization of finger flexion in injury to the brachial plexus lower trunk. This is one of the anatomic foundations of the technique. On the one hand, the nerve fibers in the BMBMCN are simply motor nerve fibers. AchE histochemical stain revealed that the BMBMCN were totally made up of medullated nerve fibers. So it is beneficial to the enhancement of the curative effects by transferring the BMBMCN as the power source of repair due to decrease in the axonal mismatching rate of the motor nerve fibers and reduction in irregular growth of the fibers. On the other hand, the median nerves consistently collect into three fascicular groups at the level of BMBMCN. In their study of the internal topographic features of the fascicular groups of the median nerve in the arm, Zancolli et al18 found that there was a distinguished boundary between the sensory fascicles and the motor fascicles in the inferior parts of the upper arm. Furthermore, the microanatomic observation of this study revealed that the median nerves consistently collect at least into three fascicular groups at

Chinese Medical Journal 2008; 121(2):99-104 103

the level of BMBMCN, which were located at the anterior, middle, and posterior parts of the median nerve trunk. Among them, the posterior fascicular group was mainly composed of the anterior interosseous nerve, the branch to the palmaris longus and the branch of the partial superficial flexor muscle of fingers.19,20

Transferring the BMBMCN to the posterior fascicular group of the median nerve with a distance of (2.23±0.97) cm could reconstruct finger flexion definitely. In a word, this study suggests that transferring the BMBMCN for selective neurotization of finger flexion is an effective and anatomically feasible technique. Secondly, elbow flexion remains normal after BMBMCN transfer. Less residual function impairment after the procurement of posterior fascicular groups of the median nerve ensures the safety of the new technique. This is another anatomic foundation of the technique. On the one hand, microanatomy revealed that the brachial muscle was a secondary muscle for elbow flexion and was compensated by the covered biceps muscle in the 2/3 of the anteriomedial side and the brachioradial muscle in the 1/3 of the anteriolateral side.21 Moreover, this study showed that 44% of the middle and inferior part of the brachialis muscle was doubly innervated by the radial nerve. So no significant influence on elbow flexion was induced by BMBMCN transfer. On the other hand, our cases demonstrated that less residual function impairment after the procurement of posterior fascicular groups of the median nerve. In this study, slight sensory defects in the thumbs occurred in 2 cases within 3 months postoperatively, but the impaired sensation did not recover to the preoperative level until 6 months postoperatively, suggesting that there was little impairment to the residual function of the median nerve. In short, our anatomical study indicates that the new technique is safe. Advantages and disadvantages of the technique Our clinical study demonstrated the following advantages of the technique: The procedure can be completed conveniently and practicably by a single incision; the nerves can be anastomosed at a go without tension, thus promoting the curative effects of the technique; and the obvious shortening of the distance between the donor nerve and the recipient nerve allows for quick recovery of finger flexion due to acceleration of nerve regeneration. This study also indicates that the BMBMCN is an effective and safe donor nerve due to less impairment to elbow flexion. This study also found some disadvantages of the technique. First of all, it has strict indications, thus limiting its clinical use. The precondition of this technique is that there is no impairment to BMBMCN as a donor nerve. In addition, it is unable to repair the function of the internal muscle of the hand. Operative cautions of the technique Firstly, clinical studies have shown the indications of the

new technique, including single lower trunk injury or proximal median nerve injury, especially in cases with a long course of disease but without occurrence of irreversible fibrosis in the innervated muscle. When functional impairment or loss occurs in the upper or middle trunk, the donor BMBMCN is likely to be affected, and the outcome is usually poor. In this study, there was one case where the brachial plexus lower trunk was damaged completely and the upper and middle trunk was injured incompletely, digital flexion failed to recover within 18 months after operation possibly due to preoperative injury to the donor BMBMCN or sensory impairment of the hand. Secondly, it plays a key role in maintaining the function of median nerve. During the procurement of the posterior fascicular group of the median nerve, the residual function of the median nerve should be protected to the best. Furthermore, functional preservation of the median nerve helps recovery of digital flexion because it reduces impairment to the sensation of the radial half in the palm.22 Thirdly, separation and electric stimulation can be used during the operation to distinguish the BMBMCN and lateral antebrachial cutaneous nerve of the BMN more accurately. Fourthly, distinguishing the finger flexion fascicular group of the median nerve during operation is of great significance. The finger flexion fascicular group lies in the posterior 1/3 part according to the preoperatively marked natural position of the median nerve. The harvest should be done with the assistance of the surgical microscope and testified by SEP examination. Finally, the study recommends postoperative external fixation of elbow flexion by 90 degree to prevent stoma split.

REFERENCES 1. Gu YD, Wu MM, Zheng YL, Zhang GM, Yan JG, Cheng XM,

et al. Microsurgical treatment for root avulsion of the brachial plexus. Chin Med J 1987; 100: 519-522.

2. Terzis JK, Vekris MD, Soucacos PN. Outcomes of brachial plexus reconstruction in 204 patients with devastating paralysis. Plast Reconstr Surg 1999; 104: 1221-1240.

3. Xu W, Lu J, Xu J, Gu Y. Full-length ulnar nerve harvest by means of endoscopy for contralateral C7 nerve root transfer in the treatment of brachial plexus injuries. Plast Reconstr Surg 2006; 118: 689-693.

4. Mcguiness CN, Kay SP. The prespinal route in contralateral C7 nerve root transfer for brachial plexus avulsion injuries. J Hand Surg [Br] 2002; 27: 159-160.

5. Okinaga S, Nagano A. Can vascularization improve the surgical outcome of the intercostal nerve transfer for traumatic brachial plexus palsy? A clinical comparison of

Chin Med J 2008;121(2):99-104 104

vascularized and non-vascularized methods. Microsurgery 1999; 19: 176-180.

6. Gu YD, Wu MM, Zhen YL, Zhao JA, Zhang GM, Chen DS, et al. Phrenic nerve transfer for brachial plexus motor neurotization. Microsurgery 1989; 10: 287-289.

7. Gu YD, Zhang GM, Chen DS, Cheng XM, Zhang LY, Yan JG, et al. Cervical nerve root transfer from contralateral normal side for treatment of brachial plexus root avulsions. Chin Med J 1991; 104: 208-211.

8. Gu YD, Wu MM, Zhen YL, Zhao JA, Zhang GM, Chen DS, et al. Phrenic nerve transfer for treatment of root avulsion of the brachial plexus. Chin Med J 1990; 103: 267-270.

9. Gu YD, Ma MK.Use of the phrenic nerve for brachial plexus reconstruction. Clin Orthop Relat Res 1996; 323:119-121.

10. Xu WD, Gu YD, Xu JG, Tan LJ. Full-length phrenic nerve transfer by means of video-assisted thoracic surgery in treating brachial plexus avulsion injury. Plast Reconstr Surg 2002; 110: 104-109.

11. Gu YD. Contralateral C7 root transfer over the last 20 years in China. Chin Med J 2007; 120: 1123-1126.

12. Oberlin C, Béal D, Leechavengvongs S, Salon A, Dauge MC, Sarcy JJ. Nerve transfer to biceps muscle using a part of ulnar nerve for C5-C6 avulsion of the brachial plexus: anatomical study and report of four cases. J Hand Surg [Am] 1994; 19: 232-237.

13. Bertelli JA, Ghizoni MF. Improved technique for harvesting the accessory nerve for transfer in brachial plexus injuries. Neurosurgery 2006; 58: 366-370.

14. Guan SB, Hou CL, Chen DS, Gu YD. Restoration of shoulder abduction by transfer of the spinal accessory nerve to suprascapular nerve through dorsal approach: a clinical study. Chin Med J 2006; 119: 707-712.

15. Gu YD, Cai PQ, Xu F, Peng F, Chen L. Clinical application of ipsilateral C7 nerve root transfer for treatment of C5 and C6 avulsion of brachial plexus. Microsurgery 2003; 23: 105-108.

16. Gu Y, Wang H, Zhang L, Zhang G, Zhao X, Chen L. Transfer of brachialis branch of musculocutaneous nerve for finger flexion: anatomic study and case report. Microsurgery 2004; 24: 358-362.

17. Gu Y, Zhang L, Zhang G, Zhao X, Chen L. Anatomical study of transferring musculocutaneous nerve branches to brachialis muscle for restoration of digital flexion：a case report. Chin J Hand Surg (Chin) 2003; 19: 213-215.

18. Zancolli EA, Cizzi EP. Atlas of surgical anatomy of the hand. 1st ed. New York, Edinburgh, London: Churchill Livingstone; 1992:580-583.

19. Zhao X, Lao J, Hung LK, Zhang GM, Zhang LY, Gu YD. Selective neurotization of the median nerve in the arm to treat brachial plexus palsy. J Bone Joint Surg Am 2005; 87-A: 122-135.

20. Zhao X, Lao J, Hung LK, Zhang GM, Zhang LY, Gu YD. Selective neurotization of the median nerve in the arm to treat brachial plexus palsy. An anatomic study and case report. J Bone Joint Surg Am 2004; 86-A: 736-742.

21. Naito A. Electrophysiological studies of muscles in the human upper limb: the biceps brachii. Anat Sci Int 2004; 79: 11-20.

22. Hagg M, Larsson B. Effects of motor and sensory stimulation in stroke patients with long-lasting dysphagia. Dysphagia 2004; 19: 219-230.

(Received January 23, 2007)

Edited by HAO Xiu-yuan