sign in sign up
<<
Home , Axonotmesis, Endoneurium, Epineurial repair, Epineurium, Nerve allograft, Nerve injury

Peripheral and Repair

Steve K. Lee, MD, and Scott W. Wolfe, MD

Abstract

Peripheral nerve are common, and there is no easily available formula sheath. The perineurial layer is the for successful treatment. Incomplete injuries are most frequent. Seddon classi- major contributor to nerve tensile fied nerve injuries into three categories: , , and neu- strength. The is the rotmesis. After complete axonal transection, the undergoes a number of innermost loose collagenous matrix degenerative processes, followed by attempts at regeneration. A distal growth within the fascicles. run cone seeks out connections with the degenerated distal fiber. The current surgi- through the endoneurium and are cal standard is with nylon suture. To span gaps that primary protected and nourished by this repair cannot bridge without excessive tension, nerve-cable interfascicular auto- layer.1 grafts are employed. Unfortunately, results of nerve repair to date have been no Sunderland has demonstrated better than fair, with only 50% of patients regaining useful function. There is that fascicles within major peripheral much ongoing research regarding pharmacologic agents, immune system modu- repeatedly divide and unite lators, enhancing factors, and entubulation chambers. Clinically applicable to form fascicular plexuses.1 This developments from these investigations will continue to improve the results of leads to frequent changes in the treatment of nerve injuries. cross-sectional topography of fas- J Am Acad Orthop Surg 2000;8:243-252 cicles in the peripheral nerves. In general, the greatest degree of fascic- ular cross-branching occurs in the lumbar and brachial plexus regions. Peripheral nerves were first distin- these strategies involve the use of Several studies have demonstrated guished from tendons by Heroph- pharmacologic agents, immune sys- greater uniformity of fascicular ilus in 300 BC. By meticulous dis- tem modulators, enhancing factors, arrangement in the major nerves of section, he traced nerves to the and entubulation chambers. A thor- the extremities; in fact, the palmar , demonstrating the con- ough understanding of the basic cutaneous and motor branches of tinuity of the .1 In concepts of and repair the median nerve may be dissected 900 AD, Rhazes made the first clear is necessary to evaluate the contro- proximally for several centimeters reference to nerve repair. How- versies surrounding these innova- without significant cross-branching. ever, not until 1795 did Cruikshank tive new modalities. demonstrate nerve healing and recovery of distal extremity func- tion after repair. In the early 1900s, Anatomy Dr. Lee is Major, United States Air Force, Cajal pioneered the concept that Section of Orthopaedic , Walson Air axons regenerate from and The cross-sectional anatomy of a Force Hospital, Fort Dix, NJ. Dr. Wolfe is Professor and Director, Hand and Upper are guided by chemotrophic sub- peripheral nerve is demonstrated in Extremity Center, Department of Orthopaedics stances. In 1945, Sunderland pro- Figure 1. The is the and Rehabilitation, Yale University School of moted microsurgical techniques to layer of the pe- , New Haven, Conn. improve nerve repair outcomes.1 ripheral nerve, which both encircles Since that time, there have been a and runs between fascicles. Its main Reprint requests: Dr. Wolfe, Department of Or- number of advances and new con- function is to nourish and protect thopaedics and Rehabilitation, Yale University School of Medicine, 800 Howard Avenue, New cepts in peripheral nerve recon- the fascicles. The outer layers of the Haven, CT 06510. struction. Research regarding the epineurium are condensed into a molecular of nerve injury sheath. Within and through the Copyright 2000 by the American Academy of has expanded the available strate- epineurium lie several fascicles, Orthopaedic Surgeons. gies for improving results. Some of each surrounded by a perineurial

Vol 8, No 4, July/August 2000 243 Peripheral Nerve Injury and Repair

nerve does not undergo distal de- Epineurium generation. Axonotmesis is defined as a loss of continuity of axons, with Endoneurium variable preservation of the connec- tive tissue elements of the nerve. is the most severe injury, equivalent to physiologic dis- ruption of the entire nerve; it may or may not include actual nerve tran- section. After injury (short of tran- section), function fails sequentially in the following order: motor, proprio- ception, touch, temperature, pain, and sympathetic. Recovery occurs sequentially in the reverse order. Sunderland1 further refined this sheath classification on the basis of the real- Axon ization that axonotmetic injuries had widely variable prognoses. He di- vided Seddon’s axonotmesis grade into three types, depending on the degree of connective tissue involve- Figure 1 Cross-sectional anatomy of the peripheral nerve. Inset at left shows an unmye- ment. Neurapraxia is equivalent to a linated fiber. Inset at bottom shows a myelinated fiber. (Adapted with permission from Lundborg G: Nerve Injury and Repair. New York: Churchill Livingstone, 1988, p 33.) Sunderland type 1 injury. Complete recovery follows this injury, which may take weeks to months. In a Sunderland type 2 injury, the In nerve repair, fascicular matching which accommodates strain and endoneurium, perineurium, and is critical to outcome, and strategies gliding of the nerve during motion.1 epineurium are still intact, but the for achieving this will be discussed. Endoneurial have the axons are physiologically disrupted. The supply of peripheral structural and functional features of Because the endoneurium is intact, nerves is a complex anastomotic the capillaries of the central nervous the regenerating axons are directed network of blood vessels (Fig. 2). system and function as an extension along their original course, and There are two major arterial systems of the blood- barrier. The en- complete functional recovery can be and one minor longitudinal system dothelial cells within the capillaries expected. The time for recovery de- linked by anastomoses. The first of the endoneurium are intercon- pends on the level of injury, as the major system lies superficially on nected by tight junctions, creating a axon must regenerate distally to the the nerve, and the second lies with- system that is impermeable to a end-organ. It can usually be mea- in the interfascicular epineurium. wide range of macromolecules, sured in months, as opposed to The minor longitudinal system is including proteins. This barrier is weeks for a Sunderland type 1 injury. located within the endoneurium impaired by , trauma, and Injuries to subsequent connective and perineurium. The major super- toxins, as well as by the mast-cell tissue layers upgrade the Sunder- ficial longitudinal vessels maintain a products histamine and serotonin. land classification. relatively constant position on the In a Sunderland type 3 injury, the surface of the nerve. The segmental endoneurium is also disrupted, but vascular supply consists of a num- Injury Classification the perineurium and epineurium ber of nutrient arteries that vary in are intact. Recovery is incomplete size and number and enter the nerve Seddon2 classified nerve injuries into in this grade of injury for a number at irregular intervals. They repeat- three major groups: neurapraxia, of reasons. First, there is more se- edly branch and anastomose with axonotmesis, and neurotmesis vere retrograde injury to cell bodies, the internal longitudinal system to (Table 1). Neurapraxia is character- which either destroys neurons or create an interconnected system. In- ized by local myelin damage, usual- slows their recovery. Second, with- jection studies have revealed the rel- ly secondary to compression. Axon out an intact endoneurium, intrafas- ative tortuosity of the blood vessels, continuity is preserved, and the cicular fibrosis occurs, which hin-

244 Journal of the American Academy of Orthopaedic Surgeons Steve K. Lee, MD, and Scott W. Wolfe, MD

Physiology of Nerve Vascular plexa Degeneration in epineurium Following axonal transection, a se- quence of pathologic events occurs in the cell body and axon. The cell body swells and undergoes chro- matolysis, a process in which the Nissl granules (i.e., the basophilic neurotransmitter synthetic machin- ery) disperse, and the cell body be- comes relatively eosinophilic. The Vascular system cell nucleus is displaced peripher- in endoneurium ally. This reflects a change in meta- Regional nutrient bolic priority from production of vessel neurotransmitters to production of structural materials needed for Extrinsic axon repair and growth, such as vessel messenger RNA, lipids, actin, tubu- Vascular system lin, and growth-associated proteins. in perineurium Shortly after axonal transection, the proximal axon undergoes trau- Figure 2 Blood supply of a peripheral nerve. (Adapted with permission from Lundborg matic degeneration within the zone G: Nerve Injury and Repair. New York: Churchill Livingstone, 1988, p 43.) of injury (Fig. 3). In most instances, the zone of injury extends proxi- mally from the injury site to the next node of Ranvier, but death of ders axonal regeneration. Third, Sunderland’s classification accurate- the cell body itself may occur, de- with longer delays, end-organs may ly describes the pathoanatomy of pending on the mechanism and undergo changes that may not allow nerve injury, it is seldom possible to energy of injury. full recovery. accurately subclassify an axonotmet- (i.e., Only the epineurium is intact in ic nerve injury on the basis of preop- breakdown of the axon distal to the the Sunderland type 4 injury. Ret- erative clinical and electromyo- site of injury) is initiated 48 to 96 rograde neuronal damage and intra- graphic data. The subtype is usually hours after transection. Deterioration fascicular fibrosis is intensified, discernible only by histologic exami- of myelin begins, and the axon be- which allows only minimal useful nation of the injured nerve. comes disorganized. Schwann cells recovery to occur. This type of in- jury requires excision of the dam- aged segment and surgical repair or Table 1 reconstruction of the nerve. Neurot- Injury Classification mesis (complete nerve disruption) is equivalent to a Sunderland type 5 Seddon2 Sunderland1 Pathophysiologic Features injury, and spontaneous recovery is negligible.1 Neurapraxia Type 1 Local myelin damage usually secondary Although Sunderland’s classifica- to compression tion provides a concise and anatom- Axonotmesis Type 2 Loss of continuity of axons; endoneurium, ic description of nerve injury, the perineurium, and epineurium intact clinical utility of this system is Type 3 Loss of continuity of axons and endoneurium; debatable. Many injuries cannot be perineurium and epineurium intact classified into a single grade. Mixed Type 4 Loss of continuity of axons, endoneurium, nerve injuries, in which all fibers are and perineurium; epineurium intact affected but to varying degrees, are Neurotmesis Type 5 Complete physiologic disruption of entire common among peripheral nerve nerve trunk injuries. Furthermore, although

Vol 8, No 4, July/August 2000 245 Peripheral Nerve Injury and Repair

Traumatic Wallerian Muscle Cell degeneration degeneration body Basal fiber Schwann cell lamina nucleus

+ + Nerve Ca2 K + End- sprout Na Protein Node of Myelin plate Microglial cell Ranvier A B

Growth cone Schwann cell

Basal Büngner lamina tube band C D

Figure 3 Degeneration and regeneration of the peripheral nerve. A, Transection of the axon. B, Traumatic degeneration in the zone of injury and wallerian degeneration distally. C, Growth-cone regenerating down the basal lamina tube. D, Schwann cells aligning to form Büngner bands. (Adapted from Seckel BR: Enhancement of peripheral nerve regeneration. Muscle Nerve 1990;13:785-800. Copyright 1989 Lahey Clinic. Reproduced with permission from John Wiley & Sons, Inc.)

proliferate and phagocytose myelin The is composed of a trophic factor is nerve growth fac- and axonal debris. structure of flattened sheets of cel- tor. This protein was seen to be Nerve injury may disrupt the lular matrix, called lamellipodia, released by a murine sarcoma and, nerve-blood barrier. Incompletely from which multiple fingerlike pro- when transplanted into chick em- injured nerves may then be ex- jections, called filopodia, extrude bryos, caused sensory and sympa- posed to unfamiliar proteins, which and explore their microenviron- thetic axons to grow toward the may act as antigens in an autoim- ment. The filopodia are electro- tumor. In addition to being trophic mune reaction. This mechanism philic and attach to cationic regions (i.e., promotes survival and growth), may propagate the cycle of nerve of the basal lamina. Within the is chemotropic degeneration.1 filopodia are actin polypeptides, (i.e., guides the axon) and also which are capable of contraction to affects growth-cone morphology. produce axonal elongation. The Other neurotrophic factors include Physiology of Nerve cone releases protease, which dis- ciliary neurotrophic factor3 and Regeneration solves matrix in its path to clear a growth factor,4 which way to its target organ. also have an important role in the After wallerian degeneration, the The growth cone responds to survival and regeneration of dam- Schwann cell basal lamina persists. four classes of factors: (1) neurotro- aged neurons. The Schwann cells align themselves phic factors, (2) neurite-promoting Unlike the neurotrophic factors, longitudinally, creating columns of factors, (3) matrix-forming precur- the neurite-promoting factors are cells called Büngner bands, which sors, and (4) metabolic and other substrate-bound glycoproteins that provide a supportive and growth- factors. Neurotrophic factors are promote neurite (axonal) growth. promoting microenvironment for macromolecular proteins present in Laminin, a major component of the regenerating axons. Endoneurial denervated motor and sensory re- Schwann cell basal lamina, is bound tubes shrink as well, and Schwann ceptors. They are also found within to type IV , proteoglycan, cells and fill the tubes. the Schwann cells along the regen- and entactin, and has been shown to At the tip of the regenerating eration path. These factors aid in accelerate axonal regeneration axon is the growth cone, a special- neurite survival, extension, and across a gap.5 Fibronectin is another ized motile exploring apparatus. maturation. The original neuro- neurite-promoting factor that has

246 Journal of the American Academy of Orthopaedic Surgeons Steve K. Lee, MD, and Scott W. Wolfe, MD been shown to promote neurite cles, and Merkel cells. Although ensures correct orientation of re- growth,6 as have neural cell adhe- there seems to be agreement that generating axons, there is little evi- sion molecule and N-cadherin.7 the sensory end-organs degenerate dence that it is superior to the less Fibrinogen, a matrix-forming pre- over time, there is debate as to how exact but simpler epineurial repair. cursor, polymerizes with fibronectin long they remain viable for reinner- Monofilament nylon suture is the to form a fibrin matrix, which is an vation, with estimates ranging from preferred suture type because of its important substrate for cell migra- 1 year to several years. As with ease of use and minimal foreign- tion in nerve regeneration.8 muscle reinnervation, however, it is body reactivity. Using a cadaveric The fourth class comprises a va- evident that early reinnervation pro- median nerve model, Giddins et al13 riety of factors that enhance nerve duces superior functional return.1 demonstrated that 10-0 nylon failed regeneration but cannot appropriate- under tension; that 9-0 nylon with- ly be placed in any of the first three stood the greatest distractive force classes. Among them are acidic and Neurorrhaphy before repair gapping; and that 8-0 basic fibroblast growth factors,9 nylon had a tendency to pull out of insulin and insulinlike growth factor, Historically, it was thought best to the repaired nerve ending. leupeptin, -derived protease wait 3 weeks before repair to allow A number of techniques are inhibitor, electrical stimulation, and the conclusion of wallerian degener- available to facilitate fascicular hormones such as thyroid hormone, ation. However, Mackinnon10 and matching. Visual alignment may be corticotropin, estrogen, and testos- other authors have shown that im- aided by topographic sketches of terone. mediate primary repair is associated both cut ends. With this method, it with better results. Prerequisites are can be determined which fascicular a clean wound, good vascular sup- group of the proximal stump corre- Distal Reinnervation ply, no crush component of the sponds to the fascicular group of injury, and adequate soft-tissue cov- the distal stump. Electrical stimula- After , distal structures erage. Skeletal stability is para- tion can be used to identify sensory undergo many changes. In major mount, and there should be mini- fascicles in the proximal stump in peripheral nerve injuries, such as mal tension on the nerve repair. an awake patient, but because wal- brachial plexus palsy, bone devel- Although the classic technique of lerian degeneration of the distal ops disuse osteoporosis, and joints neurorrhaphy is devoid of tension, axon begins within 2 to 4 days after and soft tissues become fibrotic and Hentz et al11 studied a primate transection, motor fascicles can be stiff. Muscle atrophies and under- model and showed that a direct re- identified reliably only by direct goes interstitial fibrosis but remains pair under modest tension actually nerve stimulation in fresh injuries. viable for at least 2 years. There is does better than a tension-free nerve Nerve ends can also be stained to an initial weight loss of 30% in the graft over the same regenerating differentiate between motor and first month and 50% to 60% by 2 distance. sensory axons. Initially, staining months, with muscle atrophy reach- With the advent of microsurgical was too time-consuming to be clini- ing a relatively stable state at 60% to instrumentation and technique, cally useful, but recent advances 80% weight loss by approximately 4 attempts at group fascicular repair, have been made. Gu et al14 reported months. Histologically, this is evi- rather than simple epineurial coap- on a 30-minute technique for blue- denced by a dramatic decrease in tation, have been attempted (Fig. 4). SAb staining of sensory fascicles muscle-fiber volume of approxi- Proponents argue that group fascic- and showed that staining does not mately 80% to 90%. The number of ular repair is better because axonal affect the growth and metabolism of motor endplates increases, and the realignment is more accurate with neurons. Sanger et al15 have reported muscle becomes hypersensitive and this technique. However, others on carbonic anhydrase staining and fasciculates clinically. As fibrosis have shown that there is no func- cholinesterase staining of sensory progresses, it is generally agreed tional difference in outcome be- and motor neurons, respectively. that the chances of functional rein- tween epineurial and group fascic- Carbonic anhydrase staining took nervation diminish if the nerve does ular repair. Furthermore, group 12 minutes, and cholinesterase not reach the motor endplates with- fascicular repair has the potential staining took 1 hour. The stain per- in approximately 12 months of de- disadvantage of increased scarring sisted for 35 days in the proximal nervation. and damage to the blood supply as stump and 9 days in the distal Distally, sensory nerves seek a result of the additional dissection. stump. These techniques may aid their target sensory “organs,” the Lundborg et al12 concluded that al- in both immediate and delayed pri- Meissner corpuscles, Ruffini corpus- though this technique purportedly mary nerve repair.

Vol 8, No 4, July/August 2000 247 Peripheral Nerve Injury and Repair

A B

Figure 4 A, Epineurial neurorrhaphy. B, Group fascicular neurorrhaphy. (Adapted with permission from Lundborg G: Nerve Injury and Repair. New York: Churchill Livingstone, 1988, pp 199-200.)

Nerve Grafting cularized nerve grafts have been out tension. If the injured nerve has used in the past, but with conflict- a large diameter relative to the nerve Autografts ing results. They may be consid- graft, several cable grafts are placed When primary repair cannot be ered if a long graft is needed in a in parallel to reconstruct the nerve. performed without undue tension, poorly vascularized bed. Because The grafts are matched to corre- nerve grafting is required. Auto- donor-site morbidity is an issue, sponding fascicles and sutured to grafts remain the standard for nerve vascularized grafts have been most the injured nerve with epineurial grafting material. Allografts have widely utilized in irreversible bra- sutures, as in the primary neuror- not shown recovery equivalent to chial plexus injuries. rhaphy technique. Fibrin glue may that obtained with autogenous The most common source of be used to connect the cable grafts, nerve and are still considered exper- autograft is the sural nerve, which is thus decreasing the number of su- imental. easily obtainable, the appropriate tures and minimizing additional The three major types of auto- diameter for most cable grafting trauma to the nerve grafts. The sur- graft are cable, trunk, and vascular- needs, and relatively dispensable. geon can make fibrin glue intraoper- ized nerve grafts. Cable grafts are Other graft sources include the ante- atively by mixing thrombin and fi- multiple small-caliber nerve grafts rior branch of the medial ante- brinogen in equal parts, as originally aligned in parallel to span a gap brachial cutaneous nerve, the lateral described by Narakas.16 between fascicular groups. Trunk femoral cutaneous nerve, and the Although nerve grafts have not grafts are mixed motor-sensory superficial radial .1 generally been considered polarized, whole-nerve grafts (e.g., an ulnar The technique of nerve grafting it is recommended that the graft be nerve in the case of an irreparable involves sharply transecting the placed in a reversed orientation in ). Trunk grafts injured nerve ends to excise the zone the repair site. Reversal of the nerve have been associated with poor of injury. The nerve ends should graft decreases the chance of axonal functional results, in large part due display a good fascicular pattern. dispersion through distal nerve to the thickness of the graft and con- The defect is measured, and the branches. A well-vascularized bed sequent diminished ability to revas- appropriate length of graft is har- is critical for nerve grafting. The cularize after implantation. Vas- vested to allow reconstruction with- graft should be approximately 10%

248 Journal of the American Academy of Orthopaedic Surgeons Steve K. Lee, MD, and Scott W. Wolfe, MD to 20% longer than the gap to be and lymphatic flow and prevent Later goals include recovery of tac- filled, as the graft inevitably short- tendon adherence. The extremity tile gnosis. ens with connective tissue fibrosis. must be kept warm, as cold expo- The graft repair site and the graft sure damages muscle and leads to itself have been demonstrated to fibrosis. Judicious bandaging pro- Evaluation of Recovery regain the same tensile strength as tects and limits venous congestion the native nerve by 4 weeks; there- and . Direct galvanic stimu- The most widely used grading sys- fore, the limb is usually immobilized lation reduces muscle atrophy and tem for nerve recovery is that devel- during this initial period to protect may be of psychological benefit to oped by the Medical Research Coun- the graft.1 the patient during the prolonged cil for the evaluation of both motor recovery phase, but has not been and sensory return (Table 2). Motor Allografts unequivocally demonstrated to en- recovery is graded M0 through M5, Allografts have several potential hance or accelerate nerve recovery and sensory recovery is graded S0 clinical advantages: (1) grafts can be or functional outcome. through S4 on the basis of the physi- banked; (2) there is no need for sacri- During reinnervation of the limb, cal examination. An excellent result fice of a donor nerve; and (3) surgi- continued motor and sensory reha- is described as M5,S4; a very good cal procedures are quicker without bilitation are critical. Pool result, M4,S3+; good, M3,S3; fair, the need to harvest a graft. How- can be helpful to improve joint con- M2,S2-2+; poor, M0-1,S0-1. Objective ever, allografts are not as effective as tractures and eliminate the effects of measurement of sensory recovery autografts, mainly due to the immu- gravity during initial motor recov- includes density testing by use of nogenic host response. Ansselin ery, thereby enhancing muscular moving and static two-point dis- and Pollard17 studied rat allograft performance. Biofeedback may pro- crimination and threshold testing by nerves and found an increase in vide sensory input to facilitate use of Frey or Semmes-Weinstein fila- helper T cells and cytotoxic/sup- motor reeducation. Early-phase ments. Measurement of grip and pressor T cells, implying immuno- sensory reeducation decreases mis- pinch strength is of limited use genic rejection. The cellular compo- localization and hypersensitivity because of inability to discriminate nent of allografts—and with it, their and reorganizes tactile submodali- among early levels of recovery and immunogenicity—can be destroyed ties, such as pressure and vibration. the fact that both the median and the by freeze-thawing. This leads to the production of cell debris, which in turn impairs neurite outgrowth. Dumont and Hentz18 reported on a Table 2 biologic detergent technique that Medical Research Council Grading System for Nerve Recovery removes the immunogenic cellular components without forming cell Motor recovery debris. Their experiments in rats M0 No contraction have shown that allografts processed M1 Return of perceptible contraction in the proximal muscles with this detergent had equivalent M2 Return of perceptible contraction in the proximal and distal muscles postrepair results compared with M3 Return of function in proximal and distal muscles to such a degree that all important muscles are sufficiently powerful to act against gravity autografts. M4 All muscles act against strong resistance, and some independent movements are possible M5 Full recovery of all muscles Rehabilitation of Nerve Injuries Sensory recovery S0 No recovery S1 Recovery of deep cutaneous pain The preoperative goals in a dener- S1+ Recovery of superficial pain vated extremity are to protect it S2 Recovery of superficial pain and some touch and to maintain range of motion, S2+ As in S2, but with overresponse so that it will be functional when S3 Recovery of pain and touch sensibility with disappearance of reinnervated. Splinting is useful to overresponse prevent contractures and deformity. S3+ As in S3, but localization of the stimulus is good, and there is imperfect Range-of-motion exercises are im- recovery of two-point discrimination perative while awaiting axonal re- S4 Complete recovery generation, so as to maintain blood

Vol 8, No 4, July/August 2000 249 Peripheral Nerve Injury and Repair ulnar nerve contribute to pinch and excellent) results in 21 patients. Pharmacologic agents work on grip function. Grafting was done in 21 cases and the molecular level to alter nerve re- resulted in useful recovery in 8. generation. Horowitz24 has shown Vastamäki et al22 reviewed the data the positive effects of gangliosides Results on 110 patients after ulnar nerve on rat sciatic nerve regeneration. repair and demonstrated useful Gangliosides are neurotrophic (i.e., The first large series of results of recovery in 57 patients (52%). they aid in the survival and mainte- nerve repairs came from Woodhall In a study by Wood23 of 11 pero- nance of neurons) and neuritogenic and Beebe in 1956; they reported on neal nerve reconstructions, 9 were (i.e., they aid in increasing the num- 3,656 injuries sustained during treated with nerve grafting and 2 ber and size of branching neural World War II, with an average 5- with direct neurorrhaphy. In the 9 processes). Klein et al25 have shown year follow-up.19 The results were patients treated with grafting, the forskolin to be an activator of ade- relatively poor, tainting the concept results were excellent in 2, good in nylate cyclase that increases neurite of nerve repair in the minds of sur- 2, fair in 3, and poor in 2. The only outgrowth in vivo. Wong and Mat- geons for years. It must be re- statistically significant prognostic tox26 have shown that polyamines membered that these injuries were factor was nerve graft length. All 4 work on the molecular level to in- pre–antibiotic era war injuries with patients with nerve grafts measur- crease the functional recovery of rat large areas of soft-tissue destruction ing 6 cm or less had good or excel- sciatic nerve. and wound contamination. Repairs lent results; in contrast, all 5 pa- Immune system modulators were performed without the benefit tients with grafts longer than 6 cm work by decreasing fibrosis and/or of modern microsurgical technique. had fair or poor results. Of the 2 histiocytic response. In a murine The results from subsequent patients treated with direct neuror- model, ganglioside-specific autoan- studies in which modern surgical rhaphy, 1 had an excellent result, tibodies have been demonstrated techniques were used have been and 1 had a good result. after nerve injury. In that ganglio- more encouraging. In a large compi- On the basis of 40 years’ experi- sides are neurotrophic and neurito- lation of data from a 40-year period, ence with nerve repairs, Sunder- genic, it is evident that antibodies to Mackinnon and Dellon19 reported land1 made a number of generaliza- them would be deleterious to nerve that very good results (M4,S3+) tions regarding nerve reconstruction regeneration.27 Azathioprine and were obtained in approximately results. He found that (1) young pa- hydrocortisone decrease the levels 20% to 40% of cases. Very few inju- tients generally do better than old of these autoantibodies, thereby ries recovered fully, and war inju- patients; (2) early repairs do better imparting a protective effect on ries generally did worse. than late repairs; (3) repairs of single- gangliosides after nerve-blood bar- A more recent series of primary function nerves do better than mixed- rier disruption. Regarding other repairs and fascicular grafts in 132 nerve repairs; (4) distal repairs do modulators, Sebille and Bondoux- patients with median nerve injuries better than proximal repairs; and Jahan28 have shown that cyclophos- showed good to excellent results in (5) short nerve grafts do better than phamides increase motor recovery 47 of 98 patients (48%) treated with long nerve grafts. in rat sciatic nerve. Bain et al29 have grafting and in 17 of 34 patients shown that cyclosporin A increases (50%) treated with secondary neu- nerve recovery in primate and rat rorrhaphy.20 Overall, 65 of 132 pa- Strategies to Improve models. tients (49%) had good to excellent Results The numerous enhancing factors results, 14 (11%) had fair results, include nerve growth factor, ciliary and 53 (40%) had poor results. Re- Because of the relatively large num- neurotrophic factor, motor nerve sults were poor in four situations: ber of fair to poor results still being growth factor, laminin, fibronectin, (1) the patient was more than 54 obtained in civilian injuries with neural cell adhesion molecule, N- years old; (2) the level of injury was modern microsurgical technique, cadherin, acidic and basic fibroblast proximal to the elbow; (3) the graft much research is being done to alter growth factor, insulinlike growth length was greater than 7 cm; or (4) regeneration mechanisms and im- factor, and leupeptin. Nerve growth the surgery was delayed more than prove results of nerve repair. The factor is chemotrophic to regenerat- 23 months. strategies to improve results fall ing neurons, as demonstrated by the In a separate series of 33 radial into four major categories: pharma- classic experiments first done by nerve repairs treated with grafting cologic agents, immune system Cajal in the early 1900s. Recent or secondary neurorrhaphy, Kallio modulators, enhancing factors, and studies lend support to these origi- et al21 demonstrated useful (good to entubulation chambers. nal theories. In animal studies simi-

250 Journal of the American Academy of Orthopaedic Surgeons Steve K. Lee, MD, and Scott W. Wolfe, MD lar to those of Cajal, a transected for loosely approximated nerve were not statistically different. nerve is allowed to regenerate to- ends. They allow decreased surgi- Research is under way to find a ma- ward appropriate and inappropriate cal handling of nerve ends and thus terial that will allow diffusion of receptor nerve segments on either decreased scarring. Use of entubu- nutrients, blood, and locally intro- end of a Y-shaped tubing. Axons lation chambers leaves a small in- duced factors; will prevent aberrant have been demonstrated to grow tentional gap between nerve ends, sprouting; and will resorb with time preferentially in a ratio of 2:1 to the which allows fascicular rerouting. to prevent nerve compression.34 appropriate nerve end.30 Other Entubulation chambers may also studies have used Y chambers to allow local introduction of some of show that nerves preferentially the previously mentioned pharmaco- Summary grow toward their distal stump, logic agents, immune system modu- rather than toward tendon.31 Proxi- lators, and enhancing factors.33 Despite more than 100 years of mal motor axons have been shown Entubulation chambers can be intense laboratory and clinical inves- to grow preferentially toward their made from a variety of materials. tigations, results of nerve repairs are distal motor axons instead of their Some that are currently being inves- somewhat discouraging, with only sensory axons.32 Although trophic tigated include silicone, Gore-Tex, 50% of patients regaining useful factors undoubtedly play a role in autogenous vein or dura, and poly- function. The current standard of nerve regeneration specificity, proper glycolic acid.34 Hentz et al33 have treatment is immediate epineurial end-organ reinnervation is essential stated that tubularization offers no repair with nylon suture. If primary to ultimate function. A considerable advantage over epineurial repair. repair would place more than mod- pruning effect has been demonstrated Lundborg et al12 reported on the est tension on the anastomosis, to occur after axonal mismatch and treatment of 18 patients with silicone nerve-cable autografts are employed initial reinnervation. tubes and a 3- to 4-mm repair gap. to bridge the gap. At this time, there Entubulation chambers are an They stressed the importance of is much research under way, and intriguing concept, and extensive using slightly larger tubes to prevent pharmacologic agents, immune sys- research is under way to better our nerve compression. Sensory and tem modulators, enhancing factors, understanding of their effects. motor testing after 1 year showed and entubulation chambers offer These chambers are hollow cylindri- improvement of tactile sensation promise for future improvement in cal tubes that serve as the conduit with tubularization; other variables nerve repair outcomes.

References

1. Sunderland S: Nerve Injuries and Their the patterning of neuronal projections and ulnar nerves in the human fore- Repair: A Critical Appraisal. New York: in vertebrates. Science 1988;242:692-699. arm: Early results from a prospective, Churchill Livingstone, 1991. 8. Williams LR, Varon S: Modification of randomized, clinical study. J Hand 2. Seddon HJ: Surgical Disorders of the fibrin matrix formation in situ enhances Surg [Am] 1997;22:99-106. Peripheral Nerves. Baltimore: Williams nerve regeneration in silicone cham- 13. Giddins GEB, Wade PJF, Amis AA: & Wilkins, 1972, pp 68-88. bers. J Comp Neurol 1985;231:209-220. Primary nerve repair: Strength of 3. Manthorpe M, Skaper SD, Williams 9. Cordeiro PG, Seckel BR, Lipton SA, repair with different gauges of nylon LR, Varon S: Purification of adult rat D’Amore PA, Wagner J, Madison R: suture material. J Hand Surg [Br] 1989; sciatic nerve ciliary neuronotrophic Acidic fibroblast growth factor en- 14:301-302. factor. Brain Res 1986;367:282-286. hances peripheral nerve regeneration 14. Gu XS, Yan ZQ, Yan WX, Chen CF: 4. Slack JR, Hopkins WG, Pockett S: in vivo. Plast Reconstr Surg 1989;83: Rapid immunostaining of live nerve Evidence for a motor nerve growth 1013-1021. for identification of sensory and motor factor. Muscle Nerve 1983;6:243-252. 10. Mackinnon SE: New directions in fasciculi. Chin Med J 1992;105:949-952. 5. Madison R, da Silva CF, Dikkes P, peripheral nerve surgery. Ann Plast 15. Sanger JR, Riley DA, Matloub HS, Chiu TH, Sidman RL: Increased rate of Surg 1989;22:257-273. Yousif NJ, Bain JL, Moore GH: Effects peripheral nerve regeneration using 11. Hentz VR, Rosen JM, Xiao SJ, McGill of axotomy on the cholinesterase and bioresorbable nerve guides and a KC, Abraham G: The nerve gap carbonic anhydrase activities of axons laminin-containing gel. Exp Neurol 1985; dilemma: A comparison of nerves in the proximal and distal stumps of 88:767-772. repaired end to end under tension rabbit sciatic nerves: A temporal study. 6. Gundersen RW: Response of sensory with nerve grafts in a primate model. Plast Reconstr Surg 1991;87:726-740. neurites and growth cones to patterned J Hand Surg [Am] 1993;18:417-425. 16. Narakas A: The use of fibrin glue in substrata of laminin and fibronectin in 12. Lundborg G, Rosén B, Dahlin L, repair of peripheral nerves. Orthop vitro. Dev Biol 1987;121:423-431. Danielsen N, Holmberg J: Tubular Clin North Am 1988;19:187-199. 7. Dodd J, Jessell TM: Axon guidance and versus conventional repair of median 17. Ansselin AD, Pollard JD: Immuno-

Vol 8, No 4, July/August 2000 251 Peripheral Nerve Injury and Repair

pathological factors in peripheral Surgical results. Clin Orthop 1991;267: Falk RE, Falk JA, Hunter DA: The rejection: Quantifica- 206-210. peripheral nerve allograft: A dose- tion of lymphocyte invasion and major 24. Horowitz SH: Therapeutic strategies response curve in the rat immunosup- histocompatibility complex expres- in promoting peripheral nerve regen- pressed with cyclosporin A. Plast sion. J Neurol Sci 1990;96:75-88. eration. Muscle Nerve 1989;12:314-322. Reconstr Surg 1988;82:447-457. 18. Dumont CE, Hentz VR: Enhancement 25. Klein HW, Kilmer S, Carlsen RC: En- 30. Seckel BR, Ryan SE, Gagne RG, Chiu of axon growth by detergent-extracted hancement of peripheral nerve regen- TH, Watkins E Jr: Target-specific nerve grafts. Transplantation 1997;63: eration by pharmacological activation nerve regeneration through a nerve 1210-1215. of the cyclic AMP second messenger guide in the rat. Plast Reconstr Surg 19. Mackinnon SE, Dellon AL: Surgery of system. Microsurgery 1989;10:122-125. 1986;78:793-800. the Peripheral Nerve. New York: Thieme 26. Wong BJF, Mattox DE: The effects of 31. Lundborg G, Dahlin LB, Danielsen N, Medical Publishers, 1988, pp 115-129. polyamines and polyamine inhibitors Nachemson AK: Tissue specificity in 20. Kallio PK, Vastamäki M: An analysis on rat sciatic and facial nerve regener- nerve regeneration. Scand J Plast Reconstr of the results of late reconstruction of ation. Exp Neurol 1991;111:263-266. Surg 1986;20:279-283. 132 median nerves. J Hand Surg [Br] 27. Mackinnon SE, Hudson AR, Bain JR, 32. Brushart TME: Preferential reinnervation 1993;18:97-105. Falk RE, Hunter DA: The peripheral of motor nerves by regenerating motor 21. Kallio PK, Vastamäki M, Solonen KA: nerve allograft: An assessment of regen- axons. J Neurosci 1988;8:1026-1031. The results of secondary microsurgical eration in the immunosuppressed host. 33. Hentz VR, Rosen JM, Xiao SJ, McGill repair of the radial nerve in 33 patients. Plast Reconstr Surg 1987;79:436-446. KC, Abraham G: A comparison of J Hand Surg [Br] 1993;18:320-322. 28. Sebille A, Bondoux-Jahan M: Motor suture and tubulization nerve repair 22. Vastamäki PK, Kallio PK, Solonen KA: function recovery after axotomy: techniques in a primate. J Hand Surg The results of secondary microsurgical Enhancement by cyclophosphamide [Am] 1991;16:251-261. repair of ulnar nerve injury. J Hand and spermine in rat. Exp Neurol 1980; 34. Terris DJ, Fee WE Jr: Current issues in Surg [Br] 1993;18:323-326. 70:507-515. nerve repair. Arch Otolaryngol Head 23. Wood MB: Peroneal nerve repair: 29. Bain JR, Mackinnon SE, Hudson AR, Neck Surg 1993;119:725-731.

252 Journal of the American Academy of Orthopaedic Surgeons