Repair of Severe Traction Lesions of the Brachial Plexus

Transcription

1 Repair of Severe Traction Lesions of the Brachial Plexus LAURENT SEDEL, M.D. Since 1972, the author has performed 259 brachial plexus repairs and various associated secondary procedures. The best results were obtained with surgery delayed four to five weeks, because the preoperative assessment of the lesion is more accurate after wallerian degeneration has occurred. In addition, formation of a proximal neuroma allows definition of the exact limits for resection. In cases with associated vascular damage, the vessels should be repaired at the same time as the nerve grafts unless there is severe ischemia. Intraspinal exploration with cervical laminectomy is not justified because intraspinal avulsion is always due to rootlet avulsion. Division of the clavicle to facilitate exploration of the anatomy of the plexus where it is the most complex is advocated. In general, distal grafting allows the recovery of a single function, which is preferable to an attempt at total anatomic repair. The adverse effects of contractions must be avoided. The priority of restoration of functions is an important consideration. Elbow flexion should be the first priority, followed by wrist extension, finger flexion, and shoulder abduction, in that order. The results of grafting may be improved by ancillary operations such as shoulder fusion, flexor tendon tenodesis, humeral derotation, and other procedures that provide limited function for patients with various incomplete and complete avulsions. Microsurgical repairs of brachial plexus lesions currently offer the best results for patients with this type of injury. Microsurgical grafting came into prominent use for treatment of brachial plexus injuries in Europe in the late 1960s and early Reprint requests to Laurent Sedel, M.D., Professor in Orthopaedic Surgery, H6pital Saint-Louis, 40, rue Bichat, Pans Cedex 10, France. Received: August 17, s. Early reports suggested very favorable results, especially in restoring function to the supraspinatus, deltoid, biceps, and muscles supplied by the median and radial nerves.',' Only the ulnar nerve and interossei were exempt. Time, experience, and discussions with many patients revealed that the results were clinically good, but were generally poor in terms of patient satisfaction. Some patients, however, regained function of one muscle, ix., the biceps or triceps, which allowed them to perform some activities. These observations indicated the need for a more modest appraisal and a redefinition of the objectives of microsurgical grafting. In reevaluating this technique, the following questions were asked: (1) how much improvement can a patient expect given the extent of the lesion? (2) which nerves should be used for grafting? and (3) what palliative procedures are available? These questions were examined based on experiences with 259 brachial plexus repairs and related palliative procedures performed from 1972 until the present. INDICATIONS FOR SURGERY Emergency surgery is always indicated when multiple injuries have caused brachial plexus palsies. Isolated or minor injuries to the upper limb require more judgment in deciding when to perform surgery. Delayed surgery, preferably after four or five weeks, is generally more successful. The preoperative assessment of the lesion can be more accurate 62

2 Number 237 December Traction Lesions of the Brachial Plexus 63 after wallerian degeneration has occurred, and the formation of a proximal neuroma allows precise definition of the limits for resection. In 90% of cases, delaying surgery allows time for spontaneous recovery to occur, providing healthier tissues. The results of patients surgically treated after two to four months are not different from those treated after four to nine months. However, after nine months, the results are less successful and, after two years, surgery provides no benefit. This approach is not altered in patients with injuries with associated vascular damage, except in cases with severe ischemia. In most cases, the collateral circulation is adequate and vein grafting can be performed at the same time as nerve repair. Alnot et al.' suggested exploration of the intraspinal course of the nerve roots through a cervical laminectomy. This procedure is not necessary, however, because if avulsion is intraspinal, it is always due to rootlet avulsion. An intraspinal lesion with a graftable stump has not been reported. In contrast, Privat3 showed that it is possible to have rootlets in continuity associated with a dural tear, which explains the occurrence of meningoceles in the presence of residual function. TECHNICAL CONSIDERATIONS Whether the clavicle should be divided is a question that must be considered. Initially, the author preserved the clavicle and passed grafts beneath it. This technique made the dissection easier but increased the length of the resection gap, particularly when one of the nerves ended at the level of the ~lavicle.~ For the last five years, the author has divided the clavicle, allowing a more delicate exploration at the point where the anatomy of the plexus is most complex. This improves the accuracy with which the recipient sites can be prepared and reduces the length of the graft. The clavicle is most easily repaired with a sixhole compression plate if the screw holes have been predrilled. Certain general rules apply to the grafting technique and should be followed to achieve the best results. It is always preferable to restore a single function of a root than to attempt anatomic repair of all roots. For example, if the upper trunk has been ruptured at its division and only C5 or C6 is available for grafting, it is best to graft from the available root to the anterior division and achieve good elbow flexion rather than to attempt to graft all branches and achieve some function in the deltoid, some in the spinati, and some in the brachioradialis. In the latter situation, triceps cocontraction could render the repair useless. It is therefore always preferable to graft to a specific area as distally as possible, if necessary by dissecting the nerve distally and tracing it proximally to identify the bundles that supply a single muscle. In general, the goal is to restore function to individual muscles. Problems with muscle imbalance can impair the result. COMPARISON OF NERVE GRAFTING NERVE TRANSFER AND Whether nerve grafting or nerve transfer provided better results was considered. In the shoulder, a graft from C5 to the subscapular and axillary nerves often gave an excellent result with full abduction and Grade 4 power in the deltoid. But this graft never succeeded in restoring both abduction and external rotation. Many of these patients had to be subsequently treated with a derotation osteotomy of the humerus. The best result achieved in treatment of this injury was by a nerve transfer from the accessory nerve to C5, which provided Grade 3+ power in the deltoid. Considering the elbow, many patients achieved Grade 4 or 5 power in the biceps (8 kg in the hand) after a graft from C5 or C6 to the musculocutaneous or lateral cord. The best result after a three-strand intercostal nerve transfer to the musculocutaneous cord was Grade 4 power in the biceps (2 kg in the hand). Using the accessory nerve, the patient

3 Clinical 0-i 64 Sedel and Related Research with the best result could lift 2 kg with his hand and 8 kg using the elbow. A Grade 5 power in the triceps was possible. In the forearm, deep flexor power was restored in some patients, but never to more than Grade 4. Patients sometimes achieved Grade 3 power in the interossei and thenar muscles without achieving any useful function. Finger power was weak and never reached Grade 5. Protective sensation was usually restored if motor function recovered and exceptionally light touch and pain could be appreciated, but sensory discrimination in the hand was never achieved. CHOICE OF TREATMENT The choice of treatment is related to the type and extent of the lesion and to the number, quality, and length of the available grafts. Special needs of the patient should be assessed by detailed preoperative interviews with the patient. INCOMPLETE OR SIMPLE LESIONS In some cases, the decision regarding choice of treatment is relatively simple. For example, with an incomplete lesion such as the rupture of C5 or C5 and C6 where there are one or two good roots with large neuromas and a short gap after resection of the nerve ends, the defect should be grafted from C5 to the posterior part of the upper trunk and from C6 to its anterior part. With this procedure, the results are often outstanding with good recovery of the biceps and the deltoid and sensory recovery in the thumb and index finger, although external rotation of the shoulder is not always regained. A localized infraclavicular lesion with rupture of the posterior or posterior and lateral cords is another example of an injury that is relatively simple to treat. A complete repair with very good or excellent results is usually possible. COMPLETE LESIONS In situations where there is a complete avulsion of the nerve root, the age or level of activity of the patient may play a role in the selection of the type of treatment. While a patient with a relatively sedentary lifestyle may not choose surgery, a more active patient may elect surgical treatment. Possible procedures in these cases include transfer of the accessory nerve to the anterolateral aspect of the musculocutaneous nerve alone or in conjunction with transfer of two or three intercostal nerves. If there is good recovery of the biceps, but the shoulder is stiff although in a good position with 60" of internal rotation, nothing further can be done. If the biceps has Grade 3 or 3+ power, but the shoulder is unstable, the shoulder can be arthrodesed at 45" of internal rotation, 30" of abduction, and 15" of flexion, which will increase the power of the biceps. In cases with complete avulsion of all roots except one, usually C5, which is ruptured above the clavicle, two approaches are possible. One approach is to graft the C5 root to the musculocutaneous nerve, which usually involves bridging a defect of cm. If the C5 stump is of poor quality, the accessory nerve can also be grafted to the musculocutaneous nerve. If, however, the C5 stump is in very good condition and if the lateral cord is not too fibrotic, it is possible to graft C5 to the lateral cord in an attempt to recover some function in the pectorals and finger flexors and sensation in the hand. C5 can be grafted to the musculocutaneous nerve while either the accessory nerve or three intercostals can be transferred to the radial nerve. The results of this procedure, although good in one case, were generally poor. In cases with complete avulsion of all roots except two that are ruptured above the clavicle, several procedures are possible that lead to a good recovery. If the roots can be resectioned to good bundles, C6 may be grafted to the lateral cord, taking care to anastomose the anterior portion of C6 to the bundles destined for the musculocutaneous nerve to ensure recovery of the biceps. After dissection and removal of the branch to the triceps, the motor fibers of C5 are grafted to the radial nerve. This procedure improves the chance

4 Number 237 December Traction Lesions of the Brachial Plexus 65 of recovery of the radial nerve. Also, the biceps and the triceps should not recover strongly at the same time, because of the possibility of cocontraction. If the biceps is strong, the elbow can extend by gravity. This method also provides good recovery of the extensors, which is important because with an elbow that flexes, sensation in the fingers, and strong extensors, the fingers can be expected to regain their gripping function after flexor tenodesis. Also, stability may be restored to the shoulder by arthrodesis, and at the same time correct the excessive medial rotation that is often present. Cases that involve avulsion of two roots, with three available for grafting, are rare. Finding enough grafts to bridge the defects is the major problem in these cases. The dissection must be extremely precise and the clavicle must be divided to determine the shortest defect. The largest number of grafts must be used, and care must be taken not to graft two antagonists to a single proximal root. The results in these cases have been poor, either because of insufficient grafts or poor selection of distal distribution. Stored or prosthetic nerves would be particularly useful in these cases. In Erb-Duchenne palsies,6 the hand, the triceps, and the extensors are functional. When a typical lesion involving C5 and C6 is encountered, treatment depends on the condition of the roots. If C5 and C6 are ruptured distally and the stumps of both roots are present, both nerves should be grafted. In these cases, the deltoid and the biceps should easily recover. The external rotators will only occasionally recover and function synchronously with the deltoid. A derotation osteotomy of the humerus is sometimes necessary. If one of the two roots is avulsed from the cord and if the remaining root has a good stump, an attempt can be made to graft both the biceps and the axillary nerve. If the remaining stump is damaged, only the biceps should be grafted. In the latter case, it can also be useful to graft the accessory nerve to the musculocutaneous nerve. When C5 and C6 are both avulsed from the cord, the only option is to perform an accessory nerve transfer. Sometimes the accessory nerve can be sutured directly to the anterior division of the upper trunk with good results. With lesions of C5, C6, and C7, only the wrist and finger flexors, the interossei, and the pectoralis major remain functional. The prognosis is extremely good if elbow flexion can be restored, which requires grafting to the musculocutaneous nerve. Usually C6 and C7 have been avulsed from the cord and C5 is the only available root, all of which is grafted to the musculocutaneous nerve. Occasionally, C6 and C7 are available and should be grafted to the axillary nerve. It is preferable not to attempt to restore radial nerve function because it can be regained by tendon transfer under normal circumstances. If the deltoid has recovered with an active abduction of 45", a derotation osteotomy of the humerus is indicated to overcome the loss of external rotators. If the deltoid and the external rotators remain powerless, an arthrodesis of the shoulder should be performed. Infraclavicular lesions are generally extensive because of the mechanism of injury. The only option is multiple grafting over a long distance, taking into account the significance of the nerve in question and the length of the defect. A gap of more than 20 cm gives poor results and consumes a large quantity of valuable graft. For lesions of comparable seventy, the order of precedence for grafting should be the musculocutaneous, the median, the radial, the axillary, and the ulnar nerves. If the gaps are so large that both sural nerves and both medial cutaneous nerves of the arms are insufficient to fill the defects, then the ulnar nerve must be used as a graft. Specific procedures for other types of injuries cannot be recommended because of widely differing circumstances. The limiting factor is generally the number of grafts available. The lengths of the defects should be decreased as much as possible to maximize the number of grafts that can be obtained. Clearly, however, a ruptured nerve must always be resectioned to healthy bundles. In general, this type of repair of avulsed

5 66 Sedel nerves would become obsolete if techniques such as reimplantation of avulsed spinal rootlets and prosthetic nerve grafting, and accurate and practical methods of recognizing motor and sensory fibers were available. At present, these techniques are not available and thus careful individualization of the methods discussed here offer the best results for patients with brachial plexus injuries. PALLIATIVE PROCEDURES Recovery ofthese types of injuries takes between two and three years. Palliative procedures should only be undertaken after this period of time has elapsed. Indications should include the overall result of the surgical treatment, the patient s needs and motivations, the degree of pain, the condition of the upper limb, and the strength of the muscle to be tran~ferred.~ A muscle that has recovered af- Clinical m i c s and Related Research ter denervation makes a poor transfer, so the best operations are osteotomies, arthrodeses, or tenodeses. Arthrodesis of the wrist may improve comfort or appearance, and arthrodesis of the shoulder may avoid fixed internal rotation if the biceps and pectoralis major are powerful. Treatment of persistent pain may be necessary. REFERENCES I. Alnot, J. Y., Jolly, A., and Frot, B.: Traitement direct des lesions nerveuses dans les paralysies traumatiques du plexus brachial de I adulte. A propos d une seriede 100casoperes. 1nt.Orthop. 5:151, Narakas, A.: Surgical treatment of traction injuries ofthe brachial plexus. Clin. Orthop. 133:7 I, Privat, J. M.: Personal communication, Sedel, L.: Traitement palliatif d une sene de 103 paralysies par elongation du plexus brachial. Rev. Chir. Orthop , Sedel, L.: The results of surgical repair of brachial plexus injuries. J. Bone Joint Surg. 64B:54, Sedel, L.: The management of supraclavicular lesions. Clin. Plast. Surg. 11:121, 1984.