Clinical manifestations

Presentation and course

The usual initial presentation of a patient with brachial plexus injury occurs in the emergency room and is polytraumatic. In patients with severe closed traction injuries, other systemic injuries are common. Webb and colleagues reported 18 (12%) associated significant spinal column injuries in 149 patients with brachial plexus injuries (97). Dorsi and colleagues reported spinal fractures in 13 (11.5%) and spinal cord injury in five (4.4%) of 113 children with brachial plexus injuries (29). Patients may have potentially life-threatening injuries, and the diagnosis of plexus injury might be delayed until the patient is stabilized and out of sedation. Early evaluation and diagnosis of these injuries is important because of the long-term morbidity produced by them. An initial record is extremely important because most injuries show some improvement over a period of time, and a comparison may be made possible at the time of surgical intervention. A thorough examination is aimed at accurate anatomical localization of the injury as much as possible. A good observation of the gait and attitude of the extremity, especially as the patient is getting undressed, gives useful information concerning the existing function and the patient’s ability to cope with the disability. It is important that good psychological support is provided to the patients during examination because most patients are young and suffer from significant disability due to loss of a limb function.

Paralysis of rhomboid function indicates a proximal lesion at C5 because the dorsal scapular nerve arises directly from the C5 nerve root. Sometimes, this may be difficult to elicit in the case of obese patients and patients with excessively developed trapezius muscles. Paralysis of the serratus anterior, which is supplied by long thoracic nerve from C5, C6, and C7, also indicates a proximal location of the lesion. A proximal lesion has a grave prognostic importance, though it does not contraindicate surgery.

A paralyzed deltoid would leave a groove between the tip of the acromion and humerus (a subluxating shoulder). If the rhomboid is also weak, the lesion is in the C5 root or spinal nerve. It is an upper trunk lesion if there is paralysis of supra and infraspinati muscles (supplied by the suprascapular nerve, the only nerve from the upper trunk of the brachial plexus). This setting will have an associated weak elbow flexion because the musculocutaneous nerve supplying brachialis and biceps comes from the upper trunk fibers. If the rhomboid is intact and elbow flexion is normal with isolated deltoid paralysis, the lesion is at the axillary nerve.

The posterior cord in the axilla supplies the latissimus dorsi, and it is important to elicit its function in the presence of deltoid palsy. Most of the time, a periarthritis of the shoulder joint also results in a confusing clinical picture and may require an orthopedic opinion. Function of the trapezius is important in the case of a fused shoulder because the function of the pectoral girdle is controlled by trapezius, pectoralis major, latissimus dorsi, and the serratus anterior. More distally, elbow flexion (supplied by the lateral cord) is important to examine because it is an important movement for upper limb function. Weakness of the triceps is most often the result of posterior cord injury and not a radial nerve injury. Elbow extension, however, may not be a crucial function. A wrist drop indicates posterior cord injury, whereas an associated radial drift means injury to the posterior interosseous nerve.

Evidence of Horner syndrome is extremely important and must be considered both during history taking and clinical examination. Examination in a brightly lit room might hinder this crucial finding that localizes the lesion to C8-T1 and proximal to the spinal foramen (preganglionic). It is also evidenced by weak or paralyzed intrinsic muscles of hand. This is a virtually helpless situation that generally cannot be reversed by surgery. Though the presence of Horner syndrome is not a contraindication of surgery, any procedure planned given this background must have a strong indication.

Weakness and wasting of the intrinsic and long muscles of the hand, in the absence of Horner syndrome, usually point to a lesion of the medial cord (through its supply to the ulnar and medial head of the median nerves). Weak elbow flexion and anesthesia of the thumb would mean an injury to the lateral cord that divides into the musculocutaneous nerve and the lateral head of the median nerve.

A positive Tinel sign in the form of tingling in the anesthetic arm or hand differentiates a root rupture from a root avulsion. Tinel sign may be elicited by applying mechanical stimulation at the level of the exit of various cervical roots in a craniocaudal fashion. In the case of ruptured roots, the patient typically points at the site of tingling, and distal advancement of the tingling is presumptive evidence of axonal regeneration.

Careful examination of passive movements of all joints is necessary. Upper plexus injury produces deranged shoulder joint movements, whereas the lower ones result in deranged motion at the small hand joints. Evaluating the vascularity of the upper limb is essential because the injury force, strong enough to disrupt the nerve roots, can damage the subclavian vessels also. Suspected injury to the vascular structures would indicate angiographic evaluation for pseudoaneurysm of the subclavian artery. Palpation of the clavicle for an old injury and callus formation also provides useful hints in localization. The classic triad of an anterior shoulder dislocation, expanding axillary hematoma, and diminished peripheral pulse is virtually pathognomonic for associated axillary artery transection, of which severity of brachial plexus injury is the most important long-term consequence (47).

Biological basis

Etiology and pathogenesis

The most common cause of brachial plexus injury is the high-velocity motor vehicle crash (01; 14; 34; 77). Gunshot wounds, however, have increased and are the second most common cause. Other causes include industrial injuries, penetrating trauma, snowmobile injury, pedestrian injury, iatrogenic trauma, and obstetrical trauma in newborns (61; 64; 20; 68). A combination of crush, traction, and compression injuries exist in the vast majority of cases. Severe permanent brachial plexus injuries can occur in sports like American football. Most of these injuries are caused by brachial plexus stretching or direct injury and contusion (24). Brachial plexopathy can also be caused by radiation injury or metastatic spread of tumor in breast cancer patients (101; 55). Endotracheal tube tie and head-turning has been proposed to be the cause of upper trunk brachial plexus compression in a patient undergoing spinal surgery (02). Delayed fixation of a clavicle fracture, especially between 2 and 4 weeks after injury, can result in iatropathic brachial plexus injury (43). Root avulsion injury caused by occipital condyle fracture is a rare cause of brachial plexus injury (07). Brachial plexopathy is rarely identified as a first presentation of hereditary neuropathy to pressure palsies. Wedderburn and colleagues describe a young man who developed a painless flail upper limb with a clinical diagnosis of a brachial plexopathy after his partner slept on his arm; a PMP22 deletion was found (98). Chang and colleagues reported the first case of brachial plexus injury associated with massage therapy in a 58-year-old woman who developed sudden unilateral paralysis of her left shoulder girdle after a session of massage therapy. The possible mechanism for this injury was direct and vigorous pressure at Erb’s point during massage therapy, causing compressive injury of the brachial plexus (17). Another unique case of brachial plexus palsy has been reported that occurred during manual spinal manipulative therapy performed by a chiropractor (23). Backpack palsy (BPP) is a common etiology of brachial plexopathy in military hospitals. Low body mass index is found to be the major risk factor (48). Rarely, proximal humeral physeal separation due to birth injury can also be associated with brachial plexus injury (73).

Injury to the plexus has the same responses as a peripheral nerve injury. Traction in the axis of the plexus can tear the roots out of the spinal cord, which may be observed as a brownish staining of pia mater in line with the affected roots at surgery. Avulsion of roots, however, does not mean that all the roots involved in the plexus formation from that spinal level are lost. Root avulsion can coexist with injury at a different level of the plexus. Microscopic examination of the root entry zone at the time of surgery would disclose disruption and gliosis, which might be a trigger zone for a chronic pain syndrome.

Axial traction may also tear the dura and arachnoid, resulting in the more often described pseudomeningocele of radiographic imaging. Nerve root avulsion and pseudomeningocele formation can occur independent of each other. An axial injury can also produce a proximal lesion. The spinal roots can get damaged at their exit zone of the spinal canal, especially those overlying the gutter of transverse processes, which can be identified intraoperatively by palpating over the tips of the transverse processes.

Histological confirmation of the spinal root (differentiating it from scar tissue) is important before embarking on repair by graft placement. Proximal axotomy induces a greater anterior horn cell death than peripheral axotomy. Stretch injury to the trunks in the posterior triangle usually results in neuromas-in-continuity, and disruption is rare.

Though fracture of the clavicle is extremely common, its association with brachial plexus injury is not seen that often because of the padding provided by the subclavius. However, tethering and compression due to callus are reported to damage the brachial plexus (08). A preexisting plexus injury may get aggravated with an overlying fracture of the clavicle, and a proper exposure might require division of the clavicle. In the axilla, the most common cause is shoulder dislocation, producing stretch and resulting in avulsion or neuroma formation of the axillary nerve, transmitting the force to the posterior cord frequently producing incomplete injury (84).

Characteristically, stretch injuries, the most common brachial plexus injuries, produce a cascade of insult from roots to nerves, trunks, divisions, and cords. The lower spinal roots of the plexus are intimately related to the subclavian vessels, and an aneurysm formation may be one of the causes for the progressive neurologic disability. Hemorrhage into the brachial plexus sheath during an axillary angiography may cause significant compression, and hematoma requires early evacuation to save hand function. Similarly, vascular surgery procedures involving passage of grafts around the plexus without proper visualization may result in permanent injury. The close anatomical relationship with the first rib has to be considered during resection of this structure.

A characteristic feature of brachial plexus injuries is that the sites of peripheral nerve injury are multiple within the plexus structures, each containing injuries of various intensities. A nerve fiber exhibiting segmental demyelination will have an intact axon and a completely intact peripheral extension, synapse, and end organ. It will merely require local remyelination for the restoration of function. At the other extreme, a nerve may contain fascicles whose integrity has been completely disrupted so that the pathway to regenerating axon sprouts may be blocked by an impregnable scar. Injury causing axon loss is far more common (99). Two great contributors providing breakthroughs in this area are Kline, with his electrophysiological studies of the nerve fibers, and Millesi, who emphasized microsurgical dissection.

Epidemiology

There are different patient populations with some variance in the causation and other demographics. The most common causes usually include high speed motor vehicle accidents, typically involving motorcycles and affecting young adults (01; 37; 52; 14; 34; 77). Published epidemiology from a polytrauma population had an incidence of brachial plexus injury between 0.67% and 1.3% (60). Dorsi and colleagues found that brachial plexus injury occurred in 0.1% of pediatric polytrauma patients and that the injuries were most often caused by motor vehicle accidents involving passengers (32%) and pedestrians (17%). Gunshot wounds caused brachial plexus injuries in 12% of the patients (29). In a selective population, Terzis and colleagues reported that 59% of the cases had sustained injury in motor vehicle crashes. In some studies, motorcycle accidents produced brachial plexus injury twice as often as automobile crashes (20). This increased injury incidence is due to the large amount of force applied to the unprotected victim on a motorcycle, most often producing a traction injury, unlike automobile crashes where a crush injury is often found (89). Kaiser and Haninec observed that complete plexus injuries were more common in those not wearing seatbelts, whereas upper plexus injuries were more likely in those wearing seatbelts (45). Direct compression by hematoma in vascular trauma, metastatic tumors, tumors of the neck, tumors arising from the neural sheath, fractures, and callus can produce brachial plexus injury (26; 99). Due to its firm attachments to the cervical, fascia brachial plexus is subject to stretch injury or retraction injury during anesthesia or surgery. The incidence of brachial plexus injury during delivery is 0.4 to 4.6 per 1000 live births (65). Young men are most commonly affected, and there is no right-to-left predilection (60; 89). Portions of the brachial plexus are more commonly affected by traumatic injury. Chuang reported root and rootlet injury in 70%, postganglionic spinal nerve (interscalene space, proximal to the suprascapular nerve) injury in 8%, trunk and division injury in 5%, and cord and terminal branch injury (proximal to the axillary fossa) in 17% (22). Chamata and colleagues reported 1.72% prevalence of brachial plexus injuries in patients with scapular fractures (18). In patients with multiple scapular fractures, the prevalence of brachial plexus injury was 3.12%, and it ranged from 1.52% to 2.22% in patients with single scapular fractures, depending on the specific anatomical location of the fracture. Of the 426 injuries with detailed information on nerve injury, 208 (49%) involved the radial nerve, 113 (26.5%) the ulnar nerve, 65 (15%) the median nerve, 36 (8.5%) the axillary nerve, and four (1%) the musculocutaneous nerve. The prevalence was similar across anatomical regions for single scapular fracture and was higher with multiple fractures. The largest percentage of nerve injuries was to the radial nerve (18).

Prevention

Like many other injuries, motor vehicle crashes and gunshot wounds result in a majority of brachial plexus injuries. Prevention of automobile injuries by traffic rules and regulation, as well as regulation of firearms, can result in a significant reduction of these injuries. In the polytrauma population, early diagnosis and identification of brachial plexus injury holds promise. Iatrogenic trauma and birth trauma can be significantly reduced with proper care and preprocedural investigations. Management of orthopedic trauma of the shoulder and clavicle with protective measures for brachial plexus will be helpful. Intraoperative monitoring with somatosensory evoked potentials can detect impending brachial plexus injury due to arm positioning; repositioning the arm helps in preventing the brachial plexus injury (25; 56; 58). Similarly, intraoperative impending brachial plexus injury can also be detected by transcranial electrical motor-evoked potentials (42).

Differential diagnosis

Confusing conditions

Proper clinical examination and electrophysiological investigations can differentiate plexus injury from the more commonly diagnosed nerve lesions. A number of diseases can mimic the presentation of brachial plexus injury, such as axillary nerve injury, radial nerve injury, ulnar nerve injury, posterior interosseous nerve injury, carpel tunnel syndrome, Horner syndrome, and periarthritis of the shoulder joint.

Diagnostic workup

Most often, a reasonable idea about the clinical level of injury can be obtained after a thorough clinical examination. For planning appropriate management, it is necessary to have electrophysiological and radiological studies. Electrophysiological studies require expertise in identifying the temporal sequence of electrical events following axonal discontinuity. Wallerian degeneration produces spontaneous electrical discharges (recorded as fibrillations) approximately 3 weeks following the injury; hence, needle electromyography must be delayed until this time. Denervation of paraspinal muscles supplied by dorsal rami of spinal roots provides strong evidence of avulsion of these roots. Paraspinal EMG is important in differentiating between root avulsion and distal rupture, and it helps in surgical planning when combined with routine EMG and clinical examination (06). Sensory conduction studies, along with motor conduction studies, are useful in differentiating ruptured roots from avulsed ones. Intraoperative stimulation producing reproducible, cortical, somatosensory evoked potentials definitely excludes root avulsion. Intraoperative electrophysiological studies demonstrating nerve action potentials may indicate nerve activity months before traditional electrophysiological studies due to direct nerve stimulation. Somatosensory evoked potentials during surgery also help in planning surgical procedures (49). Somatosensory evoked potentials and motor evoked potentials indicate continuity of the dorsal roots and ventral roots respectively, with the spinal cord.

Appropriate radiologic evaluation contributes to better preoperative diagnosis and diagnostic accuracy compared to clinical examination alone (01). Abul-Kasim and colleagues reported that radiologic work-up, including CT myelogram or MRI with or without 3-dimensional constructive interference in steady-state (3D CISS) images showed accuracy of 88%, sensitivity of 90%, positive predictive value of 0.9, negative predictive value of 0.87, and specificity of 87% relative to intraoperative findings. Clinical examination showed an accuracy of 65%, positive predictive value of 0.37, specificity of 56%, and negative predictive value of 1.

Radiological evaluation must include a complete study of the cervical spine, upper limb, and chest (for the clavicle and first rib). Fractures of transverse processes might indicate avulsion of the corresponding cervical roots. Fractures of the clavicle and scapula may indicate a worse supraclavicular injury to the plexus. Bony spicules or callus from the clavicle might indicate a laceration or compression of the brachial plexus. Inspiratory-expiratory chest films, diaphragm ultrasounds, and fluoroscopy can help diagnose abnormal phrenic nerve function that would indicate potential damage to the C5 nerve root. Arteriography or magnetic resonance arthrography is indicated when the physical examination suggests the possibility of vascular injury as brachial plexus injuries are associated with 10% to 25% incidence of arterial injuries (37).

CT myelography utilizing water-soluble contrast media has evolved with a 95% positive predictive value (96). It was found to have an overall sensitivity of 95% and a specificity of 98%. It is considered the gold standard for evaluating root-level injuries (37; 77). The main advantage of this technique is the visualization of the pseudomeningocele, which indicates an avulsion of roots (either one can exist without the other). Magnetic resonance imaging (MRI) is best for visualization of the plexus beyond the spinal foramina (62; 99). High-field strength MRI with multiplanar views can easily distinguish the nerves from the neighboring vascular structures. MRI has an overall sensitivity of 81% (39). The more advanced MRI myelography has improved the sensitivity to 89%, with 95% specificity and 92% diagnostic accuracy (36).

Doi and colleagues reported an MRI technique (the overlapping coronal-oblique slice MRI) with 92.9% sensitivity of detection of the cervical root avulsion that was the same as myelography/CT myelography (28). Several new imaging modalities may prove helpful in the future. New MRI sequences such as 3D CISS allow for the attainment of thin slices with the possibility of 3-dimensional reconstruction (01). The Bezier surface technique allows reformatting of volumetric data from CT myelogram to depict an individual nerve root in a single image (01). MRI-based techniques, such as diffusion tensor imaging and proton magnetic resonance spectroscopy, are beginning to be applied to assess brachial plexus and associated spinal cord changes (86; 95; 44). A study has compared the utility of ultrasound and MRI in patients suspected of having a brachial plexus lesion and claimed high sensitivity and specificity of ultrasound in detecting brachial plexus injuries (87). In this context, another study reports that ultrasound may be able to distinguish preganglionic and postganglionic traumatic brachial plexus lesions noninvasively (19). Upadhyaya and colleagues performed a study to evaluate the correlation of results of MR neurography in patients with traumatic brachial plexopathy with their operative findings (92). MR neurography is an extremely useful modality to image the traumatized brachial plexus, and it helps to accurately image the complete brachial plexus and localize as well as describe the lesions from the roots to the terminal nerves. It has a high correlation with surgical exploration and, thus, it influences both surgical planning and outcome/prognosis (92). Caution is warranted, as all of these techniques are relatively new, and further evaluation is needed to determine their utility in brachial plexus evaluation.

Management

The crucial issues in the surgical management of brachial plexus nerve injuries include: (a) whether or not segments of nerve adjacent to injury are in physical continuity; (b) extent of Wallerian-like degeneration in the nerve segment distal to the injury site; (c) initiation of regeneration in injured axons proximal and across the injury site; and (d) regeneration of injured axons in distal segment. Nerve discontinuity is often established during surgical exploration of patients with traumatic nerve injuries or MR neurography can identify this, thus, confirming the need for surgical repair. However, the majority of injuries do not disrupt the continuity of injured nerve (Sunderland grade I to IV) (83). In Sunderland grade II to IV injuries, a fundamental dilemma is whether or not the nerve needs repair and, if so, what is the appropriate time of repair or graft procedures? Protocols and algorithms about whether or not to operate at specific intervals after nerve injury are not widely or strictly followed, largely due to the lack of reliable noninvasive measures to assess ongoing regeneration of injured nerve fibers into the distal segment of the injured but continuous nerve trunks (75). Caution is warranted in interpreting the surgical studies because published studies have used extremely varied clinical and diagnostic criteria for patient selection and timing of the surgery, which has precluded emergence of evidence- or consensus-based recommendations for surgical intervention in these patients. Further, these studies are limited by lumping of patients with varied spectrum of brachial plexus injuries into the same surgical procedure/intervention groups, which prevents systematic analysis of benefit. Moreover, natural history of different brachial plexus injuries based on pathophysiologic information does not exist, partly reflecting lack of relevant noninvasive measures, which makes it extremely difficult to assess claims of benefit after specific surgical procedures.

A number of factors, including mechanism of injury, time since injury, type of injury (preganglionic or postganglionic), and associated injuries, should be considered when planning treatment (37). There are sequential steps in the treatment of brachial plexus injury: Phase I: acute therapy; Phase II: diagnosis of neurologic injury; Phase III: neurosurgical treatment; Phase IV: postoperative treatment; Phase V: reconstructive operation (70). The timing of the surgery depends on the type of injury. Any brachial plexus injury that has not shown substantial spontaneous recovery in 3 months deserves to be explored. Some investigators propose exploration of any brachial plexus injury if there is no spontaneous recovery within 3 months because over time, muscular targets can lose muscle endplates and develop fibrofatty changes that can interfere with effective reinnervation of denervated muscles (32). Associated vascular injuries need emergency treatment. Open injuries usually need sooner surgical exploration than closed injuries. For sharp open injuries, acute exploration and surgical repair of the injured brachial plexus is necessary (77). For blunt open injuries, exploration with tagging of lacerated nerves to be repaired later (3 to 4 weeks) should be performed. Gunshot wound patients with no associated vascular or thoracic injuries should initially be treated conservatively, especially low-velocity gunshot wounds that result in neuropraxia. If there is no recovery for 2 to 4 months, further exploration may be necessary (30). High-velocity gunshot wounds require more acute surgical exploration. The objective of surgical treatment is to minimize the effects of chronic denervation because the longer a muscle is denervated, the lower the chance of successful reinnervation. Patients with total palsy have almost no chance for spontaneous recovery. Irreparable lesions must be corrected as soon as possible. Bertelli and Ghizoni recommend surgery between 3 and 6 months in these patients (13). Evidence of recovery may be looked for in closed plexus injuries. It may be appropriate to wait for 4 months with frequent clinical and electrophysiological reviews (77). Routine exploration between 3 to 6 months is indicated if there are no adequate signs of reinnervation. Avulsed or grossly attenuated plexus elements could be repaired by early exploration, but this is unusual. Shin and colleagues report that exploration and reconstruction between 3 to 6 weeks is indicated when there is high suspicion for root avulsion (76). In closed injury cases, as the recovery pattern is recorded, the patient is followed for 6 months with monthly follow-up examinations. In the case of no or minimal recovery, further tests would be helpful after 16 weeks of injury. Following radiological and electrical studies, an appropriate treatment is planned. In the case of multiple root avulsions, direct reconstructive procedures may be performed without exploring the brachial plexus for neurotization. In general, most authors recommend early surgical procedures performed within 6 months (01; 33). Reconstructive surgery is better planned 18 months after brachial plexus surgery because, by then, stabilization of the plexus sets in, and the reconstructive procedure might be altogether unnecessary or minimal.

Supraclavicular exposure. By sectioning the sternocleidomastoid muscle and exposing the scalenus anticus, one would identify the phrenic nerve from where the point of the fifth and sixth nerve junction can be reached. These nerves could be dissected up to the transverse processes. If nerve resection at this proximal level is contemplated, histological verification of the spinal nerves is essential to distinguish scar tissue from nerves. In the case of excessive scarring, the phrenic nerve is a useful guide to trace the fifth nerve and then move on to the other nerves. In the case of exposure limitations, the clavicular head of pectoralis major may be removed.

Infraclavicular exposure. The deltopectoral grove is reached by sectioning the cephalic vein. The pectoralis tendon can be dissected free or divided (to be reapproximated at the end of the procedure) to expose or palpate the coracoid process, which is a landmark for the axillary nerve that leaves the posterior cord. At this point, the median nerve is also identified along with the musculocutaneous nerve. The medial head of the median nerve would lead to the ulnar nerve and the medial cutaneous nerve of the forearm that serves as a donor. In the presence of scar tissue, care must be taken to spare the nerve to the latissimus dorsi close to the posterior cord along with the nerve to the subscapularis muscle.

After adequate exposure, intraoperative conduction studies and microsurgical techniques are utilized for best results. In the case of a positive electrical response of the muscle groups, circumferential external neurolysis is performed. The prognosis for these cases is good. In the absence of contraction, a distal stimulus is given to evoke a response across the lesion. By microsurgical splitting of the nerves, appropriate elements can be isolated for repair. In the absence of a nerve action potential, the lesions are resected and grafted under a microscope. Both external and internal neurolysis coupled with nerve resection and grafting is usual.

Contralateral motor rootlet and ipsilateral nerve transfers to reconstruct shoulder abduction and external rotation and elbow flexion have been used with some success (11). Bertelli and Ghizoni reported some success with nerve transfer in upper plexus injuries, including root avulsions. Neurotizations or nerve transfers using the spinal accessory nerve, phrenic nerve, intercostal nerves, or the contralateral C7 nerve root are the common donors that have proved useful (14). Huan and colleagues, in a case series, described a method of double neurotization from a single spinal accessory nerve to two target nerves, including suprascapular and axillary nerves, to restore shoulder abduction. According to these authors, the technique of double neurotization from a single donor nerve provides favorable results in restoring shoulder abduction in avulsion brachial plexus injuries. A mean range of shoulder abduction of 91 degrees was achieved through this procedure in their study of 13 cases, of which seven cases were C5-C7 root avulsion and six cases were C5-T1 root avulsion brachial plexus injuries (41). In another study by Dr. Chu and colleagues, dual nerve transfer for restoration of shoulder function was done using spinal accessory nerve transfer to the suprascapular nerve and two intercostal nerves transfer to the anterior branch of the axillary nerve (21). They included 19 patients in the study; most had C5, C6, and C7 nerve root avulsions. Satisfactory shoulder function recovery was achieved with this technique, and patients were able to achieve shoulder abduction of 93.83 degrees and external rotation of 54.00 degrees on average. The combined use of nerve transfers and root grafting may improve outcomes in the reconstruction of C5-C6 brachial plexus injuries (12). Sköld and colleagues similarly reported that reimplantation of avulsed ventral roots after total brachial plexus avulsion injury restores useful proximal limb function and can even restore some hand function (78).

Arthrodesis of the shoulder is an option in patients sustaining severe traction injuries to the supraclavicular brachial plexus with a flail shoulder. In addition to open procedures, now arthroscopic arthrodesis of the shoulder in brachial plexus palsy is feasible. Lenoir and colleagues report that arthroscopic arthrodesis of the shoulder without open surgery can be a reliable and safe procedure and is associated with a low rate of complications (53). Open-shoulder arthrodesis has been associated with high rates of complications, including non-union and soft-tissue infections. In this study, all eight patients achieved glenohumeral fusion with arthroscopic procedure. The patients on average achieved 80 degrees flexion (range, 60 to 90 degrees) and 59 degrees abduction (range, 40 to 80 degrees) postoperatively. Mean blood loss during the perioperative period was 432 ml, which is less than with open surgical techniques. There was one superficial wound infection and one migration of an acromiohumeral screw. Because of advances in shoulder arthroplasty, there are limited indications for glenohumeral arthrodesis. This procedure provides a stable and strong shoulder with reduced pain and allows active shoulder elevation through scapulothoracic motion.

Injuries to the upper plexus and posterior cord are managed aggressively. Lesions of the lower plexus and medial cord are handled in a conservative fashion. A younger patient also deserves a more enthusiastic approach compared to an older one. Open plexus injuries are often managed by early exploration and repair, unless the general condition contraindicates. Sharply lacerated elements can be sutured without the use of grafts. In the case of a combined crushing injury, a 2 to 3 week waiting period might allow delineation of transverse and longitudinal extent of the lesion. Gunshot wounds also most often require some exploration when the neurologic function fails to improve. Other indications are aneurysm formation, hematoma formation, or intractable pain.

Other options. Articles have appeared regarding the tendon transfer technique as a method of allowing movement where brachial plexus repair is otherwise not a feasible option. Tendon transfers redistribute the forces around the joint and theoretically promote glenohumeral joint remodeling (51). This is an important feature because secondary joint abnormalities, especially deformities around the shoulder, are fairly characteristic of longstanding brachial plexus injuries such as obstetrical brachial plexus injury (69). In one such study (03) for wrist mobility impairment following Erb palsy in 20 patients, a good result was obtained in 18 patients and a fair result in 2. The choice of tendon transfer to reconstruct the wrist drop deformity in various conditions, including adult traumatic brachial plexus injuries is an option that is discussed as well. In addition, use of the long head of the biceps for shoulder function restoration has been analyzed as an option (88). Kozin and colleagues found an improvement in shoulder abduction and external rotation at 1 and 3 years after tendon transfer but no reduction in humeral head subluxation or improvement in glenohumeral joint realignment (51). Van Heest and associates report that such surgeries to correct such deformities may have better outcomes if performed at an earlier age in patients with a lack of active external rotation at the shoulder due to brachial plexus injury sustained at birth (94). Eleven children treated at 24 months of age or younger had greater improvement in the CT scan measurements of glenohumeral dysplasia than the 15 children treated after 2 years of age. Nath and colleagues similarly found better modified Mallet scores when such bony deformities were corrected before the age of 2, compared to later than the age of 2 (66). In a retrospective review of 37 patients with global avulsion of the brachial plexus, Liu and colleagues report that patients who underwent nerve transfers showed better functional outcomes (57).

Brachial plexus injuries with lower root avulsions can lead to a permanently functionless hand. The neurologic deficit with accompanying loss of muscle activity of the hand and forearm or the trophic state of the hand might be so poor that biological reconstruction is not possible. These defects pose a major reconstructive challenge, and the existing secondary procedures for brachial plexus injuries including tendon transfer, free functioning muscle transfer, arthrodesis, tenodesis, and corrective osteotomy might not adequately restore hand function. Additionally, these patients do not qualify for a homologous hand transplant because the muscles and nerves of the extremity lack the power to move the transplant. Replicating hand function with prostheses requires the creation of independent electromyographic signals to achieve multifunctional control. To achieve this surgically, nerves can be selectively transferred to optimize the number of electromyographic sites and to maximize the number of potential prosthetic functions. This procedure, combined with a carefully planned elective amputation, signal processing, advanced prostheses to substitute the amputated body part, and a comprehensive rehabilitation program is defined here as bionic reconstruction. Aszmann and colleagues presented the first case series of three patients with global brachial plexus injury, including lower root avulsions that underwent bionic reconstruction. All patients received timely brachial plexus reconstruction but still had poor hand function. Prerequisite for reconstruction with a prosthetic hand included useful shoulder and elbow function and at least two cognitively separate electromyographic signals in the forearm. Treatment occurred in two stages: first, to identify and create useful electromyographic signals for prosthetic control, and second, to amputate the hand and replace it with a mechatronic prosthesis. Before amputation, the patients had a specifically tailored rehabilitation program to enhance electromyographic signals and cognitive control of the prosthesis. Final prosthetic fitting was applied as early as 6 weeks after amputation. Bionic reconstruction successfully enabled prosthetic hand use in all three patients (04). Furthermore, Hruby and colleagues proposed a treatment algorithm for bionic hand reconstruction and reported data on the management and long-term functional outcome of patients with global brachial plexopathy undergoing this innovative treatment (40). The study included 34 patients with global brachial plexopathy treated in a single center. In 16 cases, in which primary and secondary biological reconstruction failed to improve hand function, bionic hand reconstruction was initiated. Treatment algorithm includes the following steps: (1) Profound case history and clinical examination of stiff, insensate, and functionless hand. (2) Identification of two cognitively separate EMG signals to govern a myoelectric prosthesis. (3) Nerve transfer to create new muscle targets for intuitive EMG signal control or free functional muscle transfer (FFMT). (4) Rehabilitation and signal training. (5) Hybrid hand fitting, in which the hybrid hand consists of a splint-like construction with a myoelectric hand prosthesis mounted onto or below the impaired hand. (6) Elective amputation of functionless hand. (7) Final prosthetic fitting. Functional outcome measurements were obtained in these patients using sophisticated tests such as Action Arm Research Test (ARAT), the Southampton Hand Assessment Procedure (SHAP), and the Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire. All patients showed significant improvement of scores in ARAT and SHAP; DASH scores decreased, indicating decreased disability.

Management of pain following brachial plexus injury. In about 10% to 20% of patients, pain is a prominent symptom. Almost 40% of patients with avulsions, especially of the C8 and T1 roots have pain due to the high content of sensory fibers in the lower roots (90). Early intense pain suggests deafferentation and root avulsion (77). The incidence of pain is nearly 100% in brachial plexopathy due to neoplasms, multiple root avulsions, and injuries from neurogenic thoracic outlet syndrome surgery. The pain associated with brachial plexopathy is extremely difficult to treat, and most of the medications and blocks do not provide relief. Anticonvulsants, tricyclic antidepressants, selective serotonin inhibitors, and muscle relaxants are often used with some degree of success. Sympathetic nerve blocks and transcutaneous electrical nerve stimulation are used in select patients. There is a statistically significant relationship between returning motor function and pain reduction (10). The dorsal root entry zone lesion is an effective neuroablative procedure in pain alleviation associated with avulsion injuries (35). Amputation for pain relief is seldom indicated currently, and good relief from pain is seen with early exploration and reconstruction of the brachial plexus with statistical significance (90). Maldonado and colleagues performed a retrospective study to evaluate the role of elective amputation of arm or forearm and hand after brachial plexus injury to relieve intractable pain (59). The study included a chart review of 2140 patients with brachial plexus injuries treated with elective amputation between 1999 and 2012 at a single institute. Nine patients in the study fulfilled the inclusion criteria. Three conditions were observed in all 9 patients who requested an elective amputation: (1) Pan plexus injury; (2) non-recovery 1 year after all other surgical options were performed; and (3) one chronic complication, eg, chronic infection, nonunion fractures, full-thickness burns, etc. Pain improvement was found in five patients. Four patients reported that their shoulder pain felt better than it did before the amputation, and two patients indicated that chronic pain was completely cured after surgery (59).

Subedi and colleagues performed a study to compare persistent neuropathic pain after brachial plexus injury between patients with and without co-morbid conditions (81). They analyzed the medical records of patients diagnosed with brachial plexus injury were referred to a pain center between 2006 and 2010. The study included 45 patients, 24 of them presented with one of the following co-morbid conditions during a 2-year follow-up period: myofascial pain (21%), psychiatric disorder (17%), phantom limb pain (4%), complex regional pain syndrome (21%), and insomnia (37%). Tramadol was required as a second-line agent in combination with an antiepileptic or antidepressant by 20 patients with co-morbidity and 9 patients without co-morbidity. The mean pain score (evaluated by Douleur Neuropathique 4 and a numerical rating scale) was higher in patients with co-morbidity than those without.

Outcomes

There is no unanimous agreement on the quantification of the end result according to the particular plexus element that was injured. None of the microsurgical techniques have been submitted for any randomized clinical trials. Narakas reported 64% good or fair results following autologous nerve grafting of the supraclavicular plexus and 73% good or fair results following grafting of the infraclavicular plexus (63). Terzis and colleagues had good or excellent results in 75% of suprascapular nerve reconstructions, 40% of deltoid reconstructions, 48% of biceps reconstructions, 30% of triceps reconstructions, 35% of finger-flexion reconstructions, and 15% of finger-extension reconstructions (90). These were all devastating adult brachial plexus injuries. Sulaiman and colleagues evaluated recovery of elbow flexion and shoulder abduction after nerve transfer surgeries for brachial plexus injuries (82). Ninety percent of their patients with medial pectoral to musculocutaneous nerve transfers and 60% of their patients with intercostal to musculocutaneous nerve transfers recovered to Medical Research Council grade 3. Shoulder abduction recovery to Medical Research Council grade 3 after spinal accessory to suprascapular or thoracodorsal to axillary nerve transfer was 95% and 36%, respectively. They also noted improved recovery in those patients who had both nerve transfers and direct repair of plexus elements. Bertelli and Ghizoni found that with “upper type palsies” (C5-C6± C7), root avulsions treated with triple neurotization with root grafting to the upper trunk had better outcomes than those treated with neurotization alone (13). Flores similarly found better average of the final result of surgery (AFRS) score with surgeries involving nerve root grafting, regardless of the root used for reconstruction (34). Bhatia and colleagues reported restoration of biceps power to ≥3 in more than two thirds of their patients with global brachial plexus injury by using nerve transfer with rehabilitation. Flores found better surgical outcomes as measured by the average of final results of surgery (AFRS) score in patients with good preoperative hand function who had surgery before 6 months (33). Vascularized nerve grafts are supposed to yield better results, especially with a scarred bed. Three prognostic aspects showing progress are neurobiology of regeneration, an internationally recognized classification of injury and treatment, and prevention. Assessment of muscle atrophy and fatty degeneration in brachial injury could yield valuable insight into the pathophysiology and could also be used to predict clinical outcome. Duijnisveld and colleagues performed a study to quantify and relate fat percentage and cross-sectional area (CSA) of the biceps to range of motion and muscle force of traumatic brachial plexus injury patients (31). They used a 3 Tesla magnetic resonance scanner to acquire T1-weighted TSE sequence and 3-point Dixon images of the affected and non-affected biceps brachii to determine the fat percentage as well as total and contractile CSA of 20 adult patients with brachial plexus injury. The mean fat percentage of affected biceps was higher than the non-affected biceps. The mean contractile CSA of the affected biceps was lower than non-affected biceps. The contractile CSA of the affected biceps contributed most to the reduction in active flexion, active supination, and muscle force.

A retrospective study completed by Stiasny and Birkeland suggests that nerve transfers may result in a better functional outcome than nerve grafting (79). Twelve patients were included in their analysis. The six patients with upper brachial plexus palsies all regained shoulder function and useful elbow flexion. Of the six patients with complete brachial plexus palsies, four regained shoulder function, and only one regained useful elbow function, and this was after nerve transfers (79). Nerve transfer surgery has revolutionized the management of traumatic brachial plexus injuries. Schreiber and colleagues investigated the axon count ratio of ulnar and median fascicular transfers to restore elbow flexion. Ten cadaveric specimens were used for a histomorphologic analysis of fascicular nerve transfers. A review of previously published axon counts and clinical results following transfer to the musculocutaneous nerve to restore elbow flexion was performed for the medial pectoral, spinal accessory, intercostal, thoracodorsal, ulnar, and median fascicular donor nerves. The average number of fascicles identified was 7.9 in the ulnar nerve and 8.0 in the median nerve. The mean fascicular axon count was 1318 for the ulnar nerve and 1860 for the median nerve. Mean recipient nerve axon count was 1826 for the musculocutaneous biceps branch and 1840 for the brachialis branch. A significant correlation between axon count and clinical results of transfers to restore elbow flexion was observed. Donor-to-recipient nerve axon count ratios below 0.7:1 were associated with a decreased likelihood of a successful outcome (71).

Shankar and colleagues presented a case report of a patient with an intact extremity, suffering from what is best described as phantom limb pain following a motor vehicle accident during which he experienced avulsion of his left brachial plexus (74). Electrodiagnostic studies revealed a left brachial plexopathy localized between the dorsal rami and the takeoff of the long thoracic nerve consistent with multilevel nerve root avulsion sparing the dorsal rami. Their case report demonstrates that with severe deafferentation injury, it is possible to experience phantom sensations and pain even with an intact limb. Patients with such injuries, as well as their treating clinicians, face the same challenges, which are inherent in the management of phantom limb pain (74).

Concomitant rotator cuff tears are present in approximately 1 in 10 patients with traumatic brachial plexus injury. These injuries may contribute to shoulder dysfunction; therefore, evaluation of the rotator cuff with imaging studies is appropriate when formulating treatment strategies. Restoration of shoulder function is a primary goal when treating patients with traumatic brachial plexus injury. A concomitant rotator cuff tear may alter the treatment approach and prognosis for these individuals (16).

Souza and colleagues performed a study to investigate balance impairments after brachial plexus injury (80). Eleven patients with a unilateral brachial plexus injury and 11 healthy subjects were included in the study. They assessed balance using the Berg Balance Scale, which is considered the gold standard for testing the dynamic and static balance abilities of a person. Other tests included the number of feet “touches” on the ground while performing a 60-second single-leg stance and posturographic assessment (eyes open and feet placed hip-width apart during a single 60-second trial). The body weight distribution (BWD) between the legs was estimated from the center of pressure (COP) lateral position. The variability of COP was quantified in the anterior-posterior and lateral directions. Patients with brachial plexus injury had lower BBS scores and a higher frequency of feet touches during the single-leg stance compared with those of the healthy subjects. Seventy-three percent of patients with brachial plexus injury showed asymmetric BWD toward the side opposite the affected arm. Patients with brachial plexus injury also had higher COP variability compared with healthy subjects for anterior-posterior direction.

Special considerations

Pregnancy

Brachial plexus injury is not life-threatening enough to warrant emergency exploration. However, in the case of penetrating trauma or gunshot trauma to the plexus, the urgency of the procedure and the risks may be explained to the patient in detail.

Perinatal brachial plexus injury. The incidence ranges from 0.4 to 4.6 per 1000 live births (54; 65). Risk factors include shoulder dystocia, macrosomia, assisted delivery, breech delivery, prolonged labor, excessive maternal weight, and previous delivery with brachial plexus injury (27). Most frequently, damage occurs to the upper (C5-C6) roots. Twenty percent of the time, there is complete brachial plexus injury (65). It is often an iatrogenic closed traction injury, often due to excessive lateral traction applied to the fetal head during delivery. Lesions are often unilateral and occur slightly more commonly on the left (99). Therapy is initially conservative. Hale and colleagues report that infants who recover some antigravity upper-trunk muscle strength during the first 2 months of life will likely have a full and complete recovery over the first 1 to 2 years of life (38). Those who don’t have return of biceps function until after 3 months rarely have complete recovery. A prospective study of 48 newborns with paralysis of elbow flexion found that absence of motor unit potentials by needle electromyography at age 1 month predicts paralysis of elbow flexion at 3 months in obstetric brachial plexus lesions (93). O’Brien and colleagues suggested that complete recovery by natural history occurs only if return of upper extremity motor function is present by 6 months of age (67). Bjӧrkman and colleagues performed a study to determine cerebral activation patterns in patients with brachial plexus birth injury and also residual symptoms from the shoulder (15). They monitored cerebral response to active movement of the shoulder and elbow of the injured and healthy arm using FMRI in a 3T MRI scanner. The authors report that movements, including shoulder rotation or elbow flexion and extension, of the injured side resulted in a more pronounced and more extended activation of the contralateral primary sensorimotor cortex compared to the activation seen after moving the healthy shoulder and elbow. Movement of the injured shoulder or elbow also caused increased activation in ipsilateral primary sensorimotor areas along with increased activation in associated sensorimotor areas, in both hemispheres. Hence, it was concluded that patients with brachial plexus birth injury and residual shoulder problems show reorganization in sensorimotor areas in both hemispheres of the brain. Surgical treatment is considered in cases of recovery failure in 3 to 6 months of life. In one study, Zaidman and colleagues evaluated ultrasound measurements of muscle thickness and backscatter in newborns with brachial plexus palsy and showed that muscle thickness differentiates moderate from severe impairment after newborn brachial plexus palsy (100).

Anesthesia

All surgical procedures are performed under general anesthesia with special consideration for light anesthetics, and there is an avoidance of long acting muscle relaxants to make allowance for intraoperative electrophysiologic studies.

Interscalene brachial plexus block is associated with the risk of permanent upper trunk plexopathy. Avellanet and colleagues report long-term neurologic deficit after a nerve stimulator-assisted brachial plexus block (05). There is a strong probability that the intraneural injection, together with massive intrafascicular infiltration of a local anesthetic solution at the anterior rami of C5 and C6 nerve roots, plays a major role in the neural injury involved in interscalene brachial plexus block. Increased use of ultrasound-guided techniques in regional anesthesia reduces the incidence of postoperative neurologic deficits (09; 85). Sonoelastography demonstrated increased brachial plexus stiffness during head and arm positioning changes, indicating its potential for assessing surgery-related injury risk (46). Although it shows promise for surgical positioning guidance, additional research is necessary to assess its real-world clinical advantages across diverse scenarios.