12.26 Brachial plexus injuries

Brachial plexus injuries are uncommon but very serious (Box 12.26.1). They may be closed or open, or occur as part of birth trauma. The treatment of these injuries has progressed but in most cases long-term disability persists due to the unpredictability of nerve regeneration. Secondary reconstruction is only of partial benefit. These injuries typically affect young males and they present a major challenge both to medicine and society as a whole.

The majority of brachial plexus injuries are closed, usually as a result of motorcycle accidents. This is demonstrated in Table 12.26.1 which shows experience from a tertiary referral centre in Switzerland. Open injuries occur more commonly in violent societies such as parts of South Africa and the United States of America. A significant percentage are iatrogenic lesions following removal of a lump from the posterior triangle of the neck.

The incidence of brachial plexus injuries is not known. There were estimated to be about 300–350 patients who suffer severe and permanent damage as a result of closed supraclavicular brachial plexus injuries each year in the United Kingdom. Recently the frequency appears to have diminished. There are many other milder lesions of the brachial plexus that recover well. The patients are typically inexperienced motorcyclists under 25 years old.

The brachial plexus is made up of five nerve roots from C5 to T1, but there is considerable variation and the plexus may be pre- or postfixed, i.e. coming from C4 or from as low as T2. The typical brachial plexus is shown in Figure 12.26.1.

Four types of injury can occur to the structures of the brachial plexus:

Complete recovery should occur.

This is a postganglionic lesion, i.e. one distal to the dorsal root ganglion. Therefore it represents a peripheral nerve injury with an opportunity for recovery provided macroscopic continuity is re-established, such as at surgery.

This is also a postganglionic lesion. It represents stretching of a large segment of nerve but without rupture. This should, in theory, be a favourable lesion, but intense peri- and intraneural fibrosis leads to poor results.

This is a preganglionic lesion, i.e. a central lesion often associated with a direct spinal cord injury. Typically both dorsal and ventral roots are avulsed from the spinal cord but one or other alone may be avulsed. Historically this has been an irreparable lesion, but there are now some opportunities for surgical repair (see later). Preganglionic lesions are commonly associated with severe neuralgic pains.

The main associated injuries are to bone and blood vessels.

Typically there may be one or more fractures of the head or shaft of the humerus, fracture of the scapula, clavicle, or ribs, or a dislocation of the shoulder. There may also be other limb injuries associated with severe polytrauma.

Major vascular injury, namely rupture of the subclavian artery, occurs in about 10% of supraclavicular and over 20% of infraclavicular lesions of the brachial plexus. The incidence is even higher in open wounds.

Head, spinal cord, and chest injuries occur in up to 10–15% of severe brachial plexus injuries.

The purpose of classification of brachial plexus injuries (Table 12.26.2) is to delineate the severity of the lesion. Two questions to be addressed are: Which nerve roots are involved? Are the injuries pre- or postganglionic? These are mainly addressed to closed injuries, as open injuries, typically with knives, are almost always postganglionic divisions of some or all of the brachial plexus.

The diagnosis of a brachial plexus injury is usually obvious—typically a young traffic accident victim with a flail insensate arm. In some patients, however, the diagnosis may be less clear cut, for example patients after reduction of an anterior dislocation of the shoulder who cannot abduct their shoulder, which may be due to pain, a rotator cuff lesion, or an axillary nerve or infraclavicular plexus injury.

In severe brachial plexus injuries the history is often incomplete, especially from patients who have suffered a concomitant head injury. There will often be a history of high-energy trauma.

The initial examination will often show a completely flail insensate limb but is often not very specific and is complicated by associated pathology. There may be rapid return of function within 7–10 days after the initial neurapraxia has resolved. Nonetheless with the severe injuries that often require early surgical intervention the picture rapidly becomes clear. Specifically, sensory testing should be done as per the dermatome map (Figure 12.26.2).

Muscle testing should ideally be done for each muscle in the upper limb but there are general patterns which can be recognized (Table 12.26.3).

Certain clinical findings are of particular significance.

Investigations can be subdivided into those for the neural injury and those for associated pathology. Obviously patients involved in severe polytrauma will require appropriate blood tests and other investigations.

The key radiographs are those of the chest and cervical spine. Elevation of the ipsilateral hemidiaphragm on the chest radiograph suggests a phrenic nerve injury and possible C5 and C6 nerve root avulsion. Fracture or dislocation of the first rib or fracture of the transverse process of the seventh cervical vertebra are associated with C8 and T1 root avulsions. On the anteroposterior radiograph the cervical spine may be abducted away from the site of injury suggesting root avulsion.

Myelography, computed tomography (CT) scanning with enhancement, and magnetic resonance imaging (MRI) scanning are all of value but have significant false-positive and false-negative rates and so are not yet regularly used. Ultimately, however, MRI scanning to look at both the spinal cord and the brachial plexus may be of value.

These are of little value for 2–3 weeks until Wallerian degeneration has begun. Most severe brachial plexus injuries require surgery. This is recommended earlier than 2–3 weeks, hence neurophysiology has little role acutely.

Neurophysiology has been extended to use intraoperatively with scalp or cervical electrodes. These are of assistance in evaluation of the injured nerves at surgery. They are, however, not entirely reliable.

Plain radiographs of skeletal injuries: these should be done as part of standard polytrauma assessment.

The treatment of brachial plexus injuries can be divided into initial, non-operative and primary, and secondary operative treatments.

The initial management of the brachial plexus injured patient is along advanced trauma life support guidelines. Vascular injuries are emergencies and require accurate diagnosis and urgent surgery. Surgical approach depends upon the site of injury. The vascular injuries specifically to the subclavian artery should be repaired with reversed vein grafts. Prosthetic grafts have a high incidence of thrombosis and lead to extensive local fibrosis. Skeletal injuries should be stabilized early, usually with rigid internal fixation. Humeral nails for humeral shaft fractures are easier and quicker to use but associated with higher non-union rates and do not permit examination of the radial nerve which can be ruptured in association with a brachial plexus injury.

Neurapraxic lesions should be treated conservatively. It may be difficult to delineate these lesions absolutely. Indicators include mild lesions with a limited number of nerve roots involved. There is often patchy sparing of sensation or motor function and early recovery within the first 7–10 days. The treatment aims to prevent further damage to the brachial plexus by keeping the arm adducted to the chest for 4 weeks whilst preventing distal joint stiffness with physiotherapy.

A conservative approach should not be adopted for too long because of the much greater benefit of early over late operative repair. Thus, if the anticipated recovery is not occurring at 2–3 weeks, neurophysiological studies should be performed in conjunction with spinal cord imaging, and the decision for further treatment based upon these findings and the clinical progress.

Primary operative treatment aims to restore nerve function whilst secondary operative treatment is that of muscle transfers and bone operations to overcome specific problems. There is only a very limited role for late nerve surgery.

The standard surgical approaches are the transverse supraclavicular incision for proximal lesions and the deltopectoral approach for infraclavicular lesions (Box 12.26.6).

In the supraclavicular approach flaps of skin and platysma are raised, then after division of the omohyoid between stay sutures, whilst avoiding damage to the cervical plexus and vessels and the phrenic nerve, the brachial plexus is delineated posterior and lateral to the scalenus anterior muscle.

In the infraclavicular approach the deltoid and the pectoralis major are separated, the pectoralis minor is detached from the coracoid and reflected, and the plexus is then seen wrapped around the subclavian arteries.

Once the plexus has been displayed the lesions may be obvious. With increasing time from injury, however, this is less so and may be complicated by the degree of recovery that is shown. Diagnosis may be enhanced by intraoperative electrical testing both stimulating orthodromic across the lesion and antidromic recording from cervical or scalp electrodes. For neuromas and lesions in continuity, when there is a partial response to electrical stimulation across the lesion, the decision to resect and graft, or to leave is very difficult.

The following specific surgical treatments are available:

Direct suture is rarely used and is only a possible option in early treatment of clean sharp injuries of the plexus. Even in such injuries, grafting is usually recommended. In clean injuries suitable for either direct suture or small grafts, the best results are achieved.

The technique is as for standard nerve grafting (see earlier). The nerves used are typically one or more of the medial cutaneous nerve of the forearm, medial cutaneous nerve of the arm, and the sural nerve. It is essential that these are put in without tension. They are typically secured with fibrin glue and sutures. Postoperatively the arm is immobilized in a sling full time for 6 weeks to protect the repairs. Thereafter, physiotherapy is begun.

Vascular grafts were first used in 1976 to provide a well-vascularized tube for nerve regeneration. It was hoped that this would overcome the problem of ischaemic scarred nerve beds. The reported results show that vascular grafts are only equivocally better than conventional grafting. The surgery is more complex and, therefore rarely used except in very scarring recipient sites.

These were first performed in the 1950s and are now well established. The most common are the accessory to suprascapular nerve transfers and transfers of intercostal nerves to median or musculocutaneous nerves. Their role is twofold. One is to provide further motor input to nerves distally to improve function. The accessory to suprascapular transfer can result in significant improved shoulder control and intercostal nerve transfer to the lateral cord can give grade III and even grade IV elbow flexion. They do require some relearning, but most patients manage this. The other role is to increase the sensory input to the distal nerves which can have a dramatic benefit in relief of pain especially with preganglionic injuries.

The results of nerve surgery are very variable and depend upon many factors, as with all peripheral nerve repair work (see Chapter 12.25). In essence, the recovery of the proximal muscles is best because there is the shortest distance from the nerve to recover and these muscles have a mass action so that fine innervation is not as important as in the finer actions of the small muscles of the hand. Furthermore, with nerve recovery to distal muscles there is often significant end-organ failure prior to the nerve reaching the muscle. Sensory recovery in the hand is at best protective.

The reported results are mainly with conventional nerve grafting and nerve transfers.

Of particular importance for the recovery after brachial plexus injury is the delay from treatment. Delays over 3 months are especially detrimental and after 6 months the chances of worthwhile regeneration are minimal. Repair of C5 and C6 by conventional nerve grafting gives functional flexion of the elbow and some control of the shoulder in 60% of cases.

Useful functional gain is achieved with the shoulder and elbow with significant relief of pain in over 60% of patients operated on within 3 months. Nerve transfers are particularly successful for preganglionic injuries of C5 and C6 and sometimes C7. They are much less effective for preganglionic injuries of C8 and T1.

There are various techniques of secondary operative reconstruction of brachial plexus injuries. The main ones are muscle transfers, arthrodesis, amputation, and surgery for the treatment of pain. These are always second best for ‘good nerve regeneration will result in far better function than musculotendinous transfers’.

The role of each reconstructive procedure is considered further in terms of the four main areas of use: the shoulder, elbow, forearm, and hand.

The complexity of surgery at the shoulder is often dictated by the level of function in the rest of the upper limb. Hence in injuries of C5 and C6 only, when nerve recovery has been poor, considerable efforts at improving the shoulder are worthwhile in order to use the very good hand and forearm function that the patient will have. For impaired shoulder function muscle transfer particularly of latissimus dorsi into the external rotators may be of value in regaining some active external rotation. If this is not possible, an external rotation osteotomy of the humerus may allow a more functional range for use of a good forearm and hand. For the flail or very weak shoulder arthrodesis is the most useful treatment. Gleno-humeral arthrodesis can be performed with internal or external fixation. Rarely, scapulothoracic fusion may be performed for poor thoracoscapular control.

The lack of elbow flexion is a major disadvantage to patients who still have a good hand. Lack of elbow extension is less disabling as gravity can be used to achieve extension. The patients most troubled by lack of elbow extension are those who can abduct their shoulder above 90 degrees as the forearm then tends to fall, or those patients who need extension strength to get out of a chair or wheelchair.

Most elbow surgery is directed at regaining elbow flexion. The three main types of transfer are the triceps to biceps transfer, the pectoralis major to biceps transfer, and the Steindler flexorplasty, advancing the brachioradialis proximally up the arm. The triceps to biceps transfer gives the best results but with loss of elbow extension. The Steindler flexorplasty tends to give the worst results with weakness and risk of damage to the ulnar nerves.

Recently there have been cases of use of contralateral latissimus dorsi as a free muscle graft and in Japan in particular, surgeons are using gracilis as a free muscle transfer for both elbow and wrist and hand flexion. The latter typically gain M4 or M3 elbow flexion.

The main problem is lack of forearm rotation. Active forearm rotation is very difficult to re-establish although re-establishing some pull on the distal tendon of the biceps may give some active supination. Surgery is mainly directed at improving the position of the rest of the forearm with a rotational osteotomy.

The most common problem that is treated is loss of finger and wrist extension. The extensors of the arm generally have more proximal root values than the flexors so for proximal brachial plexus injuries there may be retention of wrist and hand flexion but loss of extension. Tendon transfers are based upon the Robert Jones transfer (see Chapter 6.9) but obviously directed to which muscles are functioning well. Failure is more common that with peripheral nerve injuries as there is often partial involvement of the flexors. Retraining is also more difficult.

Pain can be the most severe consequence of a brachial plexus injury. It is more common with preganglionic injuries and with injuries of C8 and T1 than C5 and C6. It can lead to depression, opiate dependence, and suicide. It is typically crushing, burning, and vice-like often with superimposed shooting or lightening pains. These patients are difficult to treat and involvement in other activities such as work and the major role of successful rehabilitation back into the community. Specific treatments include analgesics, transcutaneous electrical nerve stimulation, nerve transfers, and spinal cord intervention.

Analgesics range from simple oral analgesics up to opiates. The latter, however, should be avoided if possible as they tend not be that effective and are associated with dependence. Neuralgical pains may respond in particular to carbamazepine, phenytoin, or amitriptyline.

Transcutaneous electrical nerve stimulation gives a significant reduction in pain in up to 50% of patients.

Nerve transfers, particularly intercostal nerve transfers, have been shown to be effective in reducing neuralgic pain. This was felt to be only with early surgery but even late surgery is of benefit in up to 80% of cases.

Attempts have been made at spinal cord intervention with interruption of spinal tracts and thermocoagulation of the dorsal root entry zone. Some patients have dramatic relief of symptoms and up to two-thirds are improved but there is a significant risk, of up to 10%, of lasting neural deficit, particularly affecting the lower limbs and possibly sexual function. Therefore spinal cord surgery is mainly used as a last resort.

Rehabilitation of patients following brachial plexus injuries is fundamental to their long-term function and integration back into their former lives. The key is a multidisciplinary approach with the surgeon or rehabilitation physician as the coordinator. Other team members include physiotherapists, occupational therapists, nurses, and retraining officers. The physiotherapists strengthen muscles, mobilize joints, and set up the transcutaneous electrical nerve stimulation. Occupational therapists make and fit splints and organize the retraining activities of daily living. The nurses provide ongoing support on the ward. The retraining officer is crucial in planning return to future work.

Predictions about the future of brachial plexus injuries are difficult and unreliable. Adult brachial plexus injuries are becoming less common as motorcycle use appears less frequent especially in poor weather. Nonetheless as the evidence continues to accumulate of the efficacy of early intervention it is hoped that these patients will be referred earlier and will have more effective nerve surgery.

Recent work from Sweden has shown regrowth of implanted ventral rootlets into the spinal cord. This allows some regeneration of motor function into preganglionic injuries. Initial work in humans has shown some promise.

Nerve growth factors have recently been studied and shown to be present only in low levels in brachial plexus neuromas. In future, application of nerve growth factors at the time of operation may improve nerve recovery.

In reconstructive surgery the role of free muscle transfers has not been fully evaluated but is likely to broaden the scope of reconstruction.

Brachial plexus injuries sadly affect some of the youngest and most active members of the population and are a severe and debilitating drain on both patients and society. Too often they have been treated with a degree of therapeutic nihilism based on the belief that surgery has no role, as espoused at the SICOT Meeting in Paris in the 1960s. However, there is no doubt that early aggressive treatment of these patients can give significant benefit and even late attention to their problems can be very helpful.

When a man has nothing a little is a lot.

Sterling Bunnell