PRINCIPLES, TECHNIQUES, AND BASIC SCIENCE
CHAPTER 9 PRINCIPLES AND TECHNIQUES OF PERIPHERAL NERVE REPAIR, GRAFTS, AND TRANSFERS
SUSAN E. MACKINNON AND STEPHEN H. COLBERT
Injuries to peripheral nerves may be devastating due to the incomplete nature of nerve healing and the possibility of permanent functional impairment. Peripheral nerve injuries require appropriate management to optimize motor and sensory recovery and to minimize pain. The surgeon must accurately identify the injury, determine the primary therapeutic goal, and decide if and when to operate. The management of peripheral nerve injuries has benefited from clinical experience gained in World War II, the evolution of microsurgical technique, improvements in surgical equipment, and the consistently advancing field of neuroscience.
NERVE ANATOMY AND PHYSIOLOGY
In the normal nerve (Figure 9.1), axons are either unmyelinated or myelinated. Unmyelinated axons are ensheathed by a single Schwann cell–derived double basement membrane, whereas myelinated axons are surrounded by a multilaminated, laminin-rich, myelin sheath with stacks of individual Schwann cells along the length of the axon. Individual nerve fibers are surrounded by the thin collagen of the endoneurium. Fibers destined for a specific anatomic location are grouped together in fascicles surrounded by the perineurium. The connective tissue that surrounds the peripheral nerve is the epineurium. A thin layer of loose areolar tissue, the mesoneurium, connects the epineurium to the surrounding structures and allows for the uninhibited excursion of nerves within the extremities. Regional arteries and veins supply the vasa nervorum, longitudinal vessels running along the epineurium that communicate with intraneural vessels running within the perineurium and the endoneurium. Bidirectional axonal transport within the nerve fiber is responsible for structural support of the nerve and delivery of neurotransmitters and trophic factors. In the normal nerve, the intrinsic blood supply is substantial, allowing mobilization and elevation of nerves over a long distance (bipedicle width:length ratio of 64:1).
Traumatized peripheral nerves are characterized by specific changes both proximal and distal to the site of injury. Proximally, axons retract a variable distance depending on the degree of injury and after a brief period of quiescence elongate as a hydra-like regenerating unit in which a single parent axon gives rise to multiple daughter axons. In myelinated nerves, axons sprout at unsheathed gaps known as the nodes of Ranvier and progress to their sensory or motor targets. Observations and elegant studies by Cajal, Sunderland, Lundborg, Brushart, Mackinnon, and others have shown that regenerating axons do not always take a direct course but do preferentially target their appropriate end-organ receptors.1-5 Once a functional synapse is made, the remaining daughter axons degenerate, or are “pruned back.” In the distal nerve segment, Schwann cells, fibroblasts, myocytes, and injured axons express a host of neurotrophic factors, including glial and brain-derived neurotrophic factors at discrete concentrations and time points as the degrading neural elements are phagocytosed in a process termed Wallerian degeneration. Neurotrophism, which literally means food for nerves, is the ability of neurotrophins secreted in an autocrine or paracrine fashion to enhance the elongation and maturation of nerve fibers. Schwann cells assume a pro-regenerative phenotype instrumental in remyelinating and guiding regenerating axons to their appropriate targets along residual endoneurial tubes. The orderly arrangement of these Schwann cells along the endoneurium forms the bands of Bungner. Functional recovery depends on the number of motor fibers correctly matched with motor endplates and the number of sensory fibers correctly matched with sensory receptors.
Experimental studies show that regenerating fibers can demonstrate both tissue and end-organ specificity.5 This process is called neurotropism. The preference of a nerve fiber to grow toward a nerve versus other tissue depends on a critical gap across which the fiber responds to the influences of the distal nerve. Current research suggests that the expression of various Schwann cell and myelin-associated glycoproteins may facilitate or impede the regeneration of damaged axons to their correct targets.6
CLASSIFYING NERVE INJURIES
The classification of nerve injuries, originally proposed by Seddon in 19437 and Sunderland in 1951,8 was subsequently expanded by Mackinnon9 to include a sixth category representing a mixed injury pattern (Figure 9.2). The level and degree of injury are important in determining treatment. First-, second-, and third-degree injuries have the potential for recovery and for the most part do not require surgical intervention. A first-degree injury recovers function quickly (within 3 months). A second-degree injury recovers slowly (1 inch per month) but completely, whereas recovery after third-degree injuries is slow and incomplete. Fourth- and fifth-degree injuries will not recover without surgical intervention. A sixth-degree injury shows a variable recovery.
First-degree injury (neurapraxia). A localized conduction block is produced that may result in segmental demyelination. Because the axons are not injured, regeneration is not required and remyelination and complete recovery occur within 12 to 16 weeks.
Second-degree injury (axonotmesis). Axonal injury occurs and the distal segment undergoes Wallerian degeneration. Proximal nerve fibers will regenerate at a rate of 1 inch per month. By definition, the connective tissue layers are uninjured. Recovery will be complete. The progress of regeneration can be followed by the advancing Tinel sign.
Third-degree injury. Wallerian degeneration is combined with some fibrosis of the endoneurium. Recovery will be incomplete because scar within the endoneurium may block or cause mismatching of regenerating fibers with the appropriate end organs. Surgery is indicated if the lesion localizes to a known area of entrapment where nerve regeneration is delayed. The recovery is uniformly better than that seen with a repair or graft unless it is associated with severe causalgia.
FIGURE 9.1. Nerve regeneration. A. The normal nerve consists of myelinated and unmyelinated axons. B. When a myelinated axon is injured, degeneration occurs distally and for a variable distance proximally. C. Multiple regenerating fibers sprout from the proximal axon forming a regenerating unit. A growth cone at the tip of each regenerating fiber samples the environment and advances the growth process distally. D. Schwann cells eventually myelinate the regenerating fibers. E. From a single nerve fiber, therefore, a regenerating unit is formed that contains several fibers, each capable of functional connections.
Fourth-degree injury. The nerve is in continuity but with complete scar block resulting from injury to the endoneurium and perineurium. Regeneration will not occur unless the block is excised and the nerve is repaired or grafted.
Fifth-degree injury (neurotmesis). The nerve is completely divided and must be repaired before any regeneration can occur.
Sixth-degree injury. This represents a combination of any of the previous five levels of injury. Because of the longitudinal nature of crushing injuries, different levels of nerve injury can be seen at various locations along the nerve. This is the most challenging nerve injury for the surgeon as some fascicles will need to be protected and not “downgraded,” whereas others will require surgical reconstruction.
Proper clinical assessment is paramount to development of a treatment plan. The extent of motor nerve injury is determined by an evaluation of weakness, loss of motion, and atrophy. The extent of sensory nerve injury is determined by moving and static two-point discrimination, which are measurements of innervation density and the number of fibers innervating sensory end organs. Light moving touch, for example, evaluates the innervation of large A-β fibers and can be quickly screened with the valid and reliable “Ten Test.”10 Patients rank the quality of sensation in the affected digit compared with that in the normal contralateral digit using a scale from 0 to 10. Vibration instruments and Semmes-Weinstein monofilaments are also used as threshold tests to evaluate the performance level of nerve fibers and are more useful in evaluating chronic compressive neuropathies. Testing is also performed after nerve repair to assess the quality of nerve repair, determine the need for revision, and monitor recovery.
FIGURE 9.2. Classification of nerve injuries. A. Uninjured nerve consists of myelinated axons, surrounded by the endoneurium, grouped into fascicles surrounded by the perineum. The outer layer of the nerve is the epineurium. In a first-degree injury, the axons are only demyelinated, whereas in a second-degree injury, the axons are injured and undergo degeneration. A third-degree injury includes damage to the axons, myelin, and endoneurium. A fourth-degree injury is a complete scar block that prevents any regeneration, and a fifth-degree injury is a division of the nerve. B. The pattern of injury may vary from fascicle to fascicle along the nerve. This mixed pattern of injury is considered a sixth-degree injury.
Sharp nerve injuries are treated with repair or reconstruction in a timely fashion, generally with minimal delay unless required to achieve a healthy wound bed. Closed injuries are treated expectantly up to 12 weeks to allow for first-, second-, and third-degree injuries to show signs of recovery. Recovery is assessed with serial physical examinations and electrodiagnostic nerve studies at 6 and 12 weeks. This allows for the accurate assessment of the degree of injury and appropriate subsequent treatment plan. Fibrillations on electromyography (EMG) indicate axonal injury and will be present around 6 weeks postinjury (second-, third-, fourth-, and fifth-degree injuries). By contrast, the presence of motor unit potentials (MUPs) does not occur until about 12 weeks postinjury. MUPs are present in second- and third- but not fourth- and fifth-degree injuries. The presence of MUPs on EMG is a contraindication to surgery except for a simple decompression at distal sites of compression. MUPs indicate collateral sprouting of intact nerve fibers. Nascent units will occur later as actual injured axons regenerate to motor targets. MUPs and nascent units are not present in fourth- and fifth-degree injuries.
PRINCIPLES OF NERVE REPAIR
Basic principles of nerve repair include the use of meticulous microsurgical techniques with adequate magnification, microsurgical instruments, and sutures. When the clinical and surgical conditions allow, a primary nerve repair is performed in a tension-free manner. To facilitate the repair, the injured segments of the nerve can be mobilized or, in the case of the ulnar nerve at the elbow, transposed, to obtain length. Intrinsically, peripheral nerves do afford a limited degree of excursion. This property of intrinsic redundancy or elasticity gives the peripheral nerves a horizontal or spiral banded appearance called the bands of Fontana.11 The bands of Fontana are created by laxity in nerve fibers. Thus, their presence in an injured nerve will let the surgeon know that nerve fibers (first-, second-, or third-degree injury) are present. This finding is helpful in evaluation of in-continuity nerve injuries. These bands disappear when the nerve is compressed or stretched. Extremes in the range of motion of joints in the vicinity of the repair and facilitation of an end-to-end repair with postural positioning of the extremity are discouraged. If a tension-free repair cannot be achieved, an interposition nerve graft is preferable with the limb in a neutral position. In an effort to match sensory and motor modalities and to optimize the specificity of nerve regeneration, a grouped fascicular repair should be performed whenever the internal topography of the nerve is segregated into motor, sensory, or regional components. Otherwise an epineural repair is performed. Postoperative motor and sensory reeducation maximizes the surgical result.
The object of peripheral nerve repair is to restore the continuity of motor and sensory fascicles in the proximal segment with the corresponding fascicles in the distal segment. The internal organization of nerves is distinct even in the proximal extremity, although nerves in the proximal extremity are monofascicular. There is considerable plexus formation between the fascicles, which decreases in the distal extremity. As nerves progress distally, they become polyfascicular and the fascicles are further differentiated into motor or sensory components.12,13 In the proximal segment of the nerve, motor fibers are distinguished from sensory fibers by knowledge of the internal topography, intraoperative stimulation, or “neurolysis with the eyes.”14 Using this technique, the distal stump of the injured nerve is dissected to discern motor from sensory fascicles. These fascicles are then visually traced back to the level of injury.
Knowledge of the usual internal topography of the peripheral nerves can direct proper alignment of fascicles at the time of nerve repair. For example, the fascicles of the ulnar nerve in the mid- and distal forearm are divided into a dorsal sensory group, a volar sensory group, and a motor group. In the mid-forearm, the motor group is positioned between the ulnar dorsal sensory group and the radial volar sensory group (Figure 9.3). The dorsal sensory group separates from the main ulnar nerve approximately 8 to 10 cm proximal to the wrist. The motor group remains ulnar to the volar sensory group until the Guyon canal, at which time it passes dorsally and radially to become the deep motor branch to the intrinsic muscles. The motor group is two-thirds the size of the sensory group at this level. The median nerve topography is more complex because it contains more fascicles. In the forearm, the anterior interosseous nerve is situated in the radial or posterior aspect of the median nerve as a distinct group. The distal internal topography of the median nerve approximates the distal anatomy; the motor fascicles to the thenar muscles are on the radial side and the sensory fibers to the third web space are on the ulnar side. Our web site, nerveinjury.wustl.edu, details the internal topography of the various nerves.
FIGURE 9.3. Ulnar nerve fascicular topography. A. At the mid- forearm, the ulnar nerve is composed of three distinct fascicular groups. The dorsal sensory branch separates from the motor branch and the main sensory group. The motor branch remains ulnar to the sensory group until the Guyon canal, at which time it passes dorsally to the sensory branches of the little and ring fingers to innervate the intrinsic muscles. B. Knowledge of this topography can be used to accurately reconstruct distal forearm nerve injuries.
When repairing the radial nerve at or above the elbow, the priority is motor rather than sensory recovery (Figure 9.4). The distal sensory fascicles should be identified and can be excluded from the repair or harvested and used as a graft to repair the motor fibers. In a similar fashion, the sensory fibers of the peroneal nerve should be excluded from repair and all efforts directed toward repairing the motor fibers to the anterior tibialis muscle (Figure 9.4). The motor fibers to the anterior tibialis are located medially within the nerve as it crosses the knee and turns abruptly around the head of the fibula. Several histochemical techniques have been described that allow motor (acetylcholinesterase and choline acetyltransferase) or sensory (carbonic anhydrase) discrimination. However, these techniques require experienced histochemical personnel, are cumbersome, and are not universally available.
After the work of Sunderland, it was assumed that the motor and sensory fibers were diffusely scattered across the different fascicles and followed a tortuous course of plexus formation until they finally organized themselves into specific motor and sensory groups distally in the extremity (Figure 9.5). Recent work contradicts this theory, showing that fibers destined for a specific territory organize themselves into distinct groups proximally within the nerve.12,15
Fascicular identification can also be used to assist with nerve reconstruction after tumor extirpation.16 If it appears likely that a functioning nerve will have to be sacrificed during tumor extirpation, the individual fascicles proximal and distal to the resection site should be mapped. By performing direct nerve-to-nerve stimulation and recording, the proximal and distal corresponding fascicles can be identified. After resection of the involved nerve, the proximal fascicles are repaired to their corresponding distal fascicles using nerve grafts.
TIMING OF NERVE REPAIR
The best results are obtained after immediate repair of a sharply transected nerve. The fascicular pattern and vascular landmarks guide the proper orientation of the nerve ends. Retraction and neuroma formation, which may result in the need for grafting, are avoided, and within the first 72 hours after injury, motor nerves in the distal nerve segment still respond to direct electrical stimulation because of the presence of residual neurotransmitters within the nerve terminals. If the nerve was injured by a crush, avulsion, or blast injury, however, the surgeon must be cognizant of nerve injury proximal and distal to the site of transection. In the acute setting, the extent of injury is difficult to determine even using the operating microscope. In this situation, the two nerve ends should be tacked together to prevent retraction and repair delayed for 3 weeks or until the wound permits. At the time of re-exploration, the extent of injury will be defined by neuroma and scar formation. The neuroma must be excised in a bread loaf fashion until a healthy fascicular pattern is seen proximally and distally. The resultant defect usually requires nerve grafting. Occasionally, when there are other associated significant injuries that require acute management that might be compromised with secondary surgery, we will do an acute nerve graft. In these cases, we will make sure that we bread loaf proximally and distally enough to be well outside the zone of injury. Our current algorithms for the timing of nerve repair are shown in Figures 9.6 and 9.7.
FIGURE 9.4. Radial and peroneal nerve fascicular anatomy. A. In the radial nerve, the motor and sensory components are separated into discrete fascicles. Awake stimulation can be used to identify the motor and sensory components of the proximal nerve, whereas anatomic dissection is used to identify them distally. The sensory portion should be excluded from the repair and can be used as a source of donor graft material. If the sensory component cannot be separated from the distal stump because of plexus formation with the motor fascicles, it can be turned into the extensor carpi radialis brevis to neurotize this muscle. This ensures that regenerating motor fibers will not be lost in the sensory territory of the radial nerve. B. Foot dorsiflexion is the essential goal of peroneal nerve repair. Grafting may be limited to the motor branch of the anterior tibialis, which lies on the medial side of the nerve as it rounds the head of the fibula and travels transversely to reach the anterior tibialis. Again, the sensory portions of the nerve can be used as donor material.
FIGURE 9.5. Nerve topography. Early surgeons believed that the fibers destined for a distinct fascicular group in the distal limb gradually came together as the plexus formation decreased. Recent work shows that fibers of a distinct fascicular group are actually located adjacent to each other, even in the proximal limb.
Clinical studies have not shown a clear superiority of fascicular repair over an epineural repair. If the internal topography of the nerve is known to be segregated in discrete motor/sensory groups, however, a grouped fascicular repair should have benefit over an epineural repair; otherwise, the extra manipulation and suture material may actually decrease the functional results. Unless the surgeon is specifically trying to direct motor and sensory alignment because of a favorable internal topography, an epineural repair is standard. Bleeding from epineural vessels should be controlled with gentle pressure or fine bipolar coagulation under microscopic guidance. After transection of a nerve, healthy individual fascicles tend to herniate out from the epineural sheath because of the normally high endoneurial fluid pressure. At the time of epineural repair, the fascicles may bend inward or outward, causing a misdirection of the regenerating fibers (Figure 9.8). Appropriate trimming of the fascicles will allow them to lie end-to-end within the epineural sheath. The epineural sutures should be placed loosely so as not to cause any additional bunching of the fascicles and so that the nerve can be realigned appropriately.
During the primary repair of a nerve, the two ends of the nerve should lie in approximation without tension. If the repair will not hold with 9-0 suture, or if postural positioning is required, a nerve graft is preferable. One challenge with nerve grafting is to restore proper sensory/motor alignment. Often the internal topography of a nerve changes across a gap. The proximal nerve may contain mixed motor and sensory fascicles or a different number of fascicles compared with the distal nerve, and thus the alignment of the grafts cannot be specific. Proper orientation is aided by knowledge of the internal anatomy, longitudinal epineural vessel location, distal dissection, and “neurolysis with the eyes.” A second challenge is to maximize the number of axons that can traverse the nerve graft through both proximal and distal neurorrhaphy sites. To divert the maximal number of axons distally, nerve grafts are reversed in orientation. This maneuver is particularly important when a long graft that possesses branches is utilized. If the graft is placed anatomically, some regenerating axons travel along these branches instead of to the distal neurorrhaphy site. If the graft is reversed in orientation, it will funnel all regenerating axons distally.
When repairing long nerve defects, the surgeon may wish to prioritize the functions of the nerve and consider excluding nonessential branches. In both the radial and peroneal nerves, but not the median and tibial nerves, the sensory components are expendable and the surgeon can concentrate on restoring the motor function. If necessary, the sensory fascicles can be used as graft material. The distal end of the excluded sensory component may be repaired in an epineural, end-to-side fashion to a nearby donor sensory nerve, not necessarily to restore excellent sensation, but to provide some sensation and limit the potential for distally mediated nerve pain by allowing reinnervation of some sensory receptors.17
NEUROMA IN CONTINUITY
A complete neuroma in continuity that has no transmission of signals and no functioning component is treated with resection and nerve grafting. However, an incomplete neuroma in continuity or a mixed, sixth-degree injury may arise after a partial nerve injury or a previous nerve repair in which portions of the nerve are functioning while other critical components are not. The surgeon must be careful not to downgrade the patient’s function by sacrificing the functioning components in an attempt to repair the remainder of the nerve. Careful preoperative assessment will determine which fascicular components are functioning and should be preserved.
At the time of repair, the neuroma in continuity may involve the complete circumference of the nerve. Individual fascicles proximal and distal to the neuroma can be separated using a microneurolysis technique. A hand-held nerve stimulator or intraoperative nerve conduction testing is used to help identify functioning motor fascicles. If sensory fascicles are to be protected, intraoperative nerve conduction testing proximal and distal to the neuroma may be required.18
FIGURE 9.6. Algorithm for the management of closed nerve injuries. EMG, electromyography; MUPs, motor unit potentials; NAP, nerve action potential; NCS, nerve conduction studies.
Separating the functioning fascicles from within the neuroma may cause additional injury to functioning components. In this situation, the neuroma possessing functioning fascicles should be preserved, whereas the nonfunctioning proximal and distal fascicles can be reconstructed with nerve grafts “black boxing” around the neuroma (Figure 9.9).
DONOR NERVE GRAFTS
The sural nerve in the adult can provide 30 to 40 cm of nerve graft. In 80% of dissections, it is formed by a union of the medial sural cutaneous nerve and the lateral peroneal communicating branch. When a large amount of graft material is needed, the communicating branch can contribute an additional 10 to 20 cm. It can also be neurolyzed from the tibial and peroneal nerves well proximal to the popliteal fossa. The nerve is found adjacent to the lesser saphenous vein at the lateral malleolus and is usually harvested in a retrograde direction. The resultant area of numbness on the lateral side of the foot decreases in size over time. The disadvantages of the sural nerve are the separate distal donor site and the less favorable neural-to-connective tissue ratio as compared with upper extremity donor nerves.
When a limited amount of graft material is required, the medial or lateral antebrachial cutaneous nerve can be harvested from the injured upper extremity. The lateral antebrachial cutaneous nerve is found adjacent to the cephalic vein along the ulnar border of the brachioradialis muscle. A maximum of 8 cm of nerve graft can be obtained and the loss of sensation is negligible as a result of the overlap in distribution by the radial sensory branch. The donor scar on the volar aspect of the forearm may be objectionable to some patients. The medial antebrachial cutaneous (MABC) nerve, found in the groove between the triceps and biceps muscles adjacent to the basilic vein, has a posterior and an anterior division. Harvesting of the anterior branch is preferred because this results in loss of sensation over the anterior aspect of the forearm, whereas loss of the posterior branch causes numbness over the elbow and the resting portion of the forearm. If necessary, up to 20 cm of nerve graft can be obtained with the MABC, and the donor scar on the medial side of the upper arm is more acceptable. Patients are instructed that an initial broad area of donor sensory loss will gradually decrease in size over 2 to 3 years. We do an end-to-side transfer from the distal stump of the MABC to the sensory side of the median nerve to rapidly decrease sensory donor deficit.
FIGURE 9.7. Algorithm for the management of open nerve injuries. asurgeon uncertain as to proximal and distal extent of injury; bas soon as soft tissue status permits.
FIGURE 9.8. Nerve repair. A. In an epineural repair, the fascicles must be appropriately trimmed so that they do not buckle, which will result in misdirection of the regenerating fibers. Excessive tightening of the epineural sutures can also cause buckling of the fascicles. B.Well-performed nerve repairs will result in good alignment of fascicles without the need for fascicular sutures.
Patients with complete median nerve sensory loss have loss of sensation in the first, second, and third web spaces. Because sensation is not critical in the third web space, the third web space nerve can be harvested to reconstruct the median nerve defect, avoiding any additional morbidity caused by nerve harvesting. The third web space nerve can be neurolyzed from the main median nerve, providing up to 24 cm of nerve graft (Figure 9.10). In a similar manner, the dorsal branch of the ulnar nerve can be harvested to reconstruct the ulnar nerve. When possible, the distal stump of the donor nerve is sewn end-to-side to an adjacent normal nerve to restore improved sensation to the donor territory. Vascularized nerve grafts have a limited role in peripheral nerve reconstruction, and their use is typically limited to lengthy, large caliber nerve grafts such as the ulnar nerve.
Expendable motor nerves that can be used as motor nerve grafts include the distal anterior interosseous nerve to the pronator quadratus, the obturator nerve branch to the gracilis, and the thoracodorsal nerve branches to the latissimus dorsi. In fact, any nerve innervating a free muscle transfer could be used as a motor nerve graft. We use these motor nerve grafts for reconstruction of critical motor function when nerve transfer does not provide a better option, such as the intrinsic motor fascicle of the ulnar nerve at the hand or wrist.
The use of nerve transfers has expanded over the last decade based on a more detailed knowledge of the intraneural topography and branching patterns of peripheral nerves in the upper and lower extremities. Nerve transfers are indicated in very proximal peripheral nerve injuries or root avulsions where a proximal stump is unavailable for primary repair or grafting. Even when grafting is possible, the injury may be so proximal that a nerve transfer facilitates better reinnervation of motor endplates than does a nerve graft. Nerve transfers are also indicated to avoid operating in regions of severe scarring, when nerve injuries present in a delayed fashion, when partial nerve injuries present with a well-defined functional deficit, or when the level of injury is unclear such as in idiopathic neuritides or radiation-induced nerve injury.19
FIGURE 9.9. Neuroma in continuity. A. When reconstructing a neuroma in continuity with intact motor function, the motor fascicles through the neuroma must be preserved. B. Intraoperative nerve testing can identify the motor fascicles proximal and distal to the neuroma. C.The remaining sensory fibers are divided proximally and distally, then reconstructed with grafts bypassing the neuroma. Any attempt to dissect the motor fascicles out of the neuroma will only downgrade the function.
Motor nerve transfers require an expendable donor motor nerve with a large number of pure motor axons that are located in close proximity to motor endplates, thus minimizing the distance and time regenerating axons need to travel to reinnervate their targets. It is also preferable that the donor nerve innervates a muscle that is synergistic with its target.13 The criteria for sensory nerve transfers include an expendable donor sensory nerve that innervates a noncritical sensory distribution, contains a large number of pure sensory axons, and is located near its sensory end organs.
The most common applications of motor nerve transfers include the restoration of elbow flexion, shoulder abduction, ulnar-innervated intrinsic hand function, forearm pronation, and radial nerve function.13 To restore elbow flexion, the medial pectoral, thoracodorsal, or intercostal nerves can be transferred to the musculocutaneous nerve. The flexor carpi ulnaris branch of the ulnar nerve and the flexor digitorum superficialis/flexor carpi radialis branch of the median nerve can also be transferred to the biceps and brachialis branches of the musculocutaneous nerve to more specifically restore elbow flexion and limit donor nerve morbidity (Figure 9.11). To restore shoulder abduction, the distal accessory nerve can be transferred to the suprascapular nerve, or the medial head triceps branch of the radial nerve can be transferred to the axillary nerve. To restore intrinsic hand function, the distal anterior interosseous nerve can be transferred to the ulnar nerve. Transferring redundant fascicles of the flexor carpi ulnaris branches of the ulnar nerve or the extensor carpi radialis brevis nerve to the median nerve–innervated pronator teres can restore forearm pronation. Alternatively, the flexor digitorum superficialis, or palmaris longus branches of the median nerve, can be transferred to its pronator branch. The radial nerve may be reconstructed by transferring median nerve donors including redundant flexor digitorum superficialis branches and flexor carpi radialis branches to the nerve to extensor carpi radialis brevis and the posterior interosseous nerve, respectively, perhaps in combination with a pronator teres to extensor carpi radialis brevis tendon transfer.20 The site nervesurgery.wustl.edu has all our nerve transfers available.
FIGURE 9.10. Median nerve at the wrist. The nerve to the 3rd webspace can be used as a nerve graft to assist in the reconstruction of the more critical nerves. The proximal portion is harvested as a graft (green). The distal end of the 3rd webspace nerve is repaired in an end-to-side epineural fashion to the nerve to the 2nd webspace (yellow).
FIGURE 9.11. A double fascicular transfer for elbow flexion. A. Transfer of a redundant fascicle of the ulnar nerve to the biceps branch of the musculocutaneous nerve and a redundant fascicle of the median nerve to the brachialis branch of the musculocutaneous nerve. B.Transfer of a redundant fascicle of the median nerve to the biceps branch of the musculocutaneous nerve and a redundant fascicle of the ulnar nerve to the brachialis branch of the musculocutaneous nerve. FCR, flexor carpi radialis; FCU, flexor carpi ulnaris.
Studies show that nerves will regenerate across a short nerve gap through various conduits, such as veins, pseudosheaths, and bioabsorbable tubes.21 The characteristics of the ideal nerve conduit include low antigenicity, availability, and biodegradability. Vein grafts have been used to reconstruct distal sensory nerve defects of less than 3 cm. Sensory results with vein grafts have been acceptable but not as good as conventional grafting. For this reason vein grafts are recommended only for reconstruction of noncritical nerve gaps of less than 3 cm.22
Nerve regeneration across a 3-cm gap through a biodegradable polyglycolic nerve tube has been demonstrated in the primate model and in a clinical trial.13 Clinical recovery was comparable to that across a standard nerve graft. The insertion of a short piece of nerve graft material into the center of the conduit will enhance regeneration by providing a local source of trophic factors. The ready availability of biodegradable synthetic grafts to span short nerve gaps would eliminate the morbidity associated with nerve graft harvest and would capitalize on the potential benefits of neurotropism in directing nerve regeneration. Synthetic nerve conduits are now available for reconstruction of small diameter nerves with a gap ≤3 cm, or with large diameter nerves with gaps ≤0.5 cm. We recommend limiting the use of nerve conduits to bridging small sensory gaps and as nerve wraps and we would advise the addition of some proximal minced nerve to the center of the conduit to provide a source of Schwann cells and trophic factors.
Nerve allografts have demonstrated clinical usefulness in the setting of extensive peripheral nerve injuries where there is a paucity of donor nerve material. Because the nerve allograft serves as a scaffold that is repopulated by host axons and Schwann cells over time, its challenge to the immune system is of limited duration. The agent FK506 (tacrolimus) is most ideally suited for treating patients with peripheral nerve allografts based on its dual role as an immunosuppressive and a neuroregenerative agent. By accelerating the rate at which axons traverse the nerve allograft, FK506 shortens the duration of immunosuppression and the period during which complications can develop. The optimal timing and dose of FK506 therapy has been identified in rodents, and a synergistic effect with nerve allograft cold preservation, as well as an ability to rescue nerve allografts undergoing acute rejection, established. Based on these findings, FK506 is now the mainstay of clinical peripheral nerve allotransplantation.
Potential candidates for peripheral nerve allotransplantation receive nerve allografts from donors that have been screened for ABO blood typing, HIV, and cytomegalovirus. These grafts are stored in the University of Wisconsin cold storage solution at 41°F (5°C) for at least 7 days. This solution is supplemented with penicillin G, dexamethasone, and insulin. Immunosuppression of the nerve allograft recipient begins 3 to 5 days prior to nerve transplantation and consists of FK506 whose dose is titrated to appropriate steady-state blood levels, azathioprine, and prednisone. The prednisone dose is tapered in the first 4 to 8 weeks after surgery. Pneumocystis carinii prophylaxis is performed at the time of immunosuppression to minimize opportunistic pulmonary infections. Immunosuppression continues for 6 months after a Tinel sign is noted to pass the last distal neurorrhaphy site and some functional recovery has occurred. Peripheral nerve allograft rejection resembles a superficial phlebitis with inflammation and tenderness, but is localized over the underlying nerve allograft and not a vein.
Processed acellular cadaveric nerve allografts have become available for clinical use recently (AxoGen, Inc., Alachua, FL). These grafts are available in different diameters and lengths, are not immunogenic, and thus do not require immunosuppression. Some studies suggest that these allografts allow regeneration over longer nerve gaps than empty conduits, but they fail to be equivalent to nerve autograft.23 These acellularized allografts have largely replaced nerve conduits, but we limit their use to noncritical sensory nerve deficits ≤3 to 4 cm.
Nerve repair and grafting have benefited from the development of microsurgical techniques and advances in the neurosciences. State-of-the-art nerve repair requires not only precision techniques but also additional measures to direct nerve regeneration to its original function. Although nerve grafting remains the standard for reconstruction of the nerve gap, synthetic conduits, allografts, and nerve transfers now play a limited role in the peripheral nerve surgeon’s armamentarium.
1. Cajal SRY. Degeneration and Regeneration in the Nervous System. Vol 1. London: Oxford University Press; 1928.
2. Sunderland S. The capacity of regenerating axons to bridge long gaps in nerves. J Comp Neurol. 1953;99:481-497.
3. Lundborg G, Hansson HA. Nerve lesions with interruption of continuity: studies on the growth pattern of regenerating axons in the gap between the proximal and distal nerve ends. In: Gorio A, Millesi H, Mingrino S, eds. Posttraumatic Nerve Regeneration: Experimental Basis and Clinical Implications. New York, NY: Raven Press; 1981:229-239.
4. Mackinnon S, Dellon L, Lundborg G, Hudson A, Hunter D. A study of neurotropism in the primate model. J Hand Surg [Am]. 1986;11:888-894.
5. Brushart TM, Seiler WD. Selective reinnervation of distal motor stumps by peripheral motor axons. Exp Neurol. 1987;97:289-300.
6. Jabaley ME, Wallace WH, Heckler FR. Internal topography of major nerves of the forearm and hand: a current view. J Hand Surg [Am]. 1980;5:1-18.
7. Seddon HJ. Three types of nerve injury. Brain. 1943;66:237-288.
8. Sunderland S. A classification of peripheral nerve injuries producing loss of function. Brain. 1951;74:491-516.
9. Mackinnon SE, Dellon AL. Surgery of the Peripheral Nerve. New York, NY: Thieme; 1988.
10. Strauch B, Lang A, Ferder M, et al. The ten test. Plast Reconstr Surg. 1997;99:1074-1078.
11. Clarke E, Bearn JG. The spiral nerve bands of Fontana. Brain. 1972;95:1-20.
12. Williams HB, Jabaley ME. The importance of internal anatomy of the peripheral nerves to nerve repair in the forearm and hand. Hand Clin. 1986;2:689-707.
13. Dvali L, Mackinnon S. Nerve repair, grafting, and nerve transfers. Clin Plast Surg. 2003;30:203-221.
14. Hallin RG. Microneurography in relation to intraneural topography: somatotopic organization of median nerve fascicles in humans. J Neurol Neurosurg Psychiatry. 1990;53:736-744.
15. Greenberg MM, Leitao C, Trogadis J, et al. Irregular geometries in normal unmyelinated axons: a 3D serial EM analysis. J Neurocytol. 1990;19:978-988.
16. Lundborg G, Dahlin LB, Danielsen N, et al. Tissue specificity in nerve regeneration. Scand J Plast Reconstr Surg. 1986;20:279-283.
17. Dorsi[Q15] MJ, Chen L, Murinson BB, Pogatzki-Zahn EM, Meyer RA, Belzberg AJ. Pain. 2008;134(3):320-334.
18. Lee GW, Mackinnon SE, Brandt K, et al. A technique for nerve reconstruction following resection of soft-tissue sarcoma. J Reconstr Microsurg. 1993;9(2):139-144.
19. Mackinnon SE, Novak CB, eds. Nerve transfers. Hand Clin. November 2008;24(4):319-488.
20. Ray WZ, Mackinnon SE. Clinical outcomes following median to radial nerve transfers. J Hand Surg [Am]. 2011;36(2):201-208.
21. Weber RV, Mackinnon SE. Bridging the neural gap. Clin Plast Surg. 2005;32(4):605-616.
22. Moore AM, Kasukurthi R, Magill CK, Farhadi HF, Borschel GH, Mackinnon SE. Limitations of conduits in peripheral nerve repairs. Hand. 2009;4(2):180-186.
23. Whitlock EL, Tuffaha SH, Luciano JP, et al. Processed allografts and type I collagen conduits for repair of peripheral nerve gaps. Muscle Nerve. 2009;39(6):787-799.