TEMPORARY INNERVATION OF A PRIMARY COVERAGE MUSCLE:A NEW TECHNIQUE TO OPTIMIZE FUNCTION IN A SUBSEQUENTFUNCTIONAL MICROVASCULAR MUSCLE TRANSPLANT

DARRELL BROOKS, M.D.*

The author describes the simple technique of innervating the coverage muscle in the staged reconstruction of an upper-extremity crush-avulsion injury with a functional microvascular muscle transplant (FMMT). The thoracodorsal nerve was repaired to the mixed motor-sensoryradial nerve above the elbow. Contraction of the latissimus muscle at 8 months after nerve repair signaled the adequacy of the 10-cmthoracodorsal nerve graft as a target motor nerve for the eventual FMMT. Excursion of the latissimus muscle created a septo-alveoloar planesimilar to the plane between two healthy muscles into which the FMMT could be placed. The author also discusses the potential advantagesof early thoracodorsal nerve repair for successful nerve regeneration. This simple technique helped overcome the potential limitations tofunctional muscle transplantation in the severely traumatized upper extremity, and deserves applied study. ª 2005 Wiley-Liss, Inc.Microsurgery 25:310�315, 2005.

Functional microvascular muscle transplants (FMMTs)require artery and vein repair to restore viability, andmotor nerve repair to restore voluntary contraction.Obtaining maximal functional return in the transplantedmuscle, however, depends on additional variables, suchas the status of the recipient bed, time to reinnervation,and quality of the recipient nerve.1�3 These variables arecontrollable in many situations in which a FMMT isindicated. However, the recipient nerve and wound bedmay be suboptimal in major traumatic soft-tissue loss.

A staged approach to functional microvascularmuscle transplantation is preferred in cases of traumaticsoft-tissue loss. In the acute first stage, wound debride-ment is followed by primary coverage, usually with afasciocutaneous, myocutaneous, or muscle flap. In anelective secondary stage, after the wound is controlledand an adequate target motor nerve is defined, FMMTis performed. Unfortunately, even with this staged ap-proach, the success of functional microvascular muscletransplantation is less predictable in situations involvingmajor soft-tissue and/or nerve avulsion.1,4 We report onan innovation to the staged reconstruction of a trau-matized upper extremity with FMMT. This innovationinvolves the temporary innervation of a latissimus dorsimuscle transplant used for primary coverage of a largeupper-extremity defect in the acute phase. We also de-scribe how this simple technique can optimize both therecipient bed and motor nerve, maximizing the potentialof a functional muscle transplanted in the second stage.

CASE REPORT

A 37-year-old man was transferred to our facilityafter sustaining a right dominant forearm injury in amotor vehicle rollover accident. On examination, an 18· 22 cm soft-tissue defect extended from the dorsum ofthe forearm onto the proximal volar forearm. The pa-tient was unable to extend his wrist, thumb, or fingers,and he had weak flexion of his wrist and fingers.

Surgical exploration confirmed loss of the dorsalmusculature in the proximal two-thirds of the forearmand loss of the brachioradialis and pronator teres, withpartial loss of the flexor carpi radialis, palmaris longus,flexor digitorum superficialis, and flexor digitorumprofundus muscle bellies (Fig. 1). The radial artery,posterior interosseous, and radial sensory nerves weredestroyed throughout the extent of the wound. Al-though the extensor pollicis longus and extensor indicesproprius were intact at the distal forearm, no targetnerve was available for grafting. The radial nerve injuryextended proximal to the level of its sensory branch.

Continuous subatmospheric pressure dressings(VAC, Kinetic Concepts, Inc., San Antonio, TX) wereapplied between serial debridements to decrease swellingand reduce inflammatory factors. When the wound wasdetermined to be free of all nonviable tissue 11 daysafter the injury, an interposition saphenous vein graftwas used to reconstruct the radial artery. This was doneto provide a vascular target for future tissue transplan-tation. The contralateral latissimus dorsi muscle washarvested and transplanted to provide coverage. Theradial collateral artery provided inflow. The thoraco-dorsal nerve was repair to the mixed motor sensory ra-dial nerve stump, which had been debridedapproximately 2 cm proximal to healthy-appearingfascicles, well out of the zone of injury. The total

Department of Microsurgical Transplantation and Replantation, CaliforniaPacific Medical Center, San Francisco, CA

*Correspondence to: Darrell Brooks, M.D., 45 Castro St., Suite 140,San Francisco, CA 94114. E-mail: dbrooks@buncke.org

Received 20 August 2004; Accepted 7 October 2004

Published online 24 May 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/micr.20121

ª 2005 Wiley-Liss, Inc.

thoracodorsal nerve graft length was approximately 10cm. A split-thickness skin graft provided cover for thelatissimus muscle.

The postoperative course was uncomplicated. Attime of arterial anastomosis, intravenous dextran-40was started at 25 cc an hour. After venous anastomosis,an implantable Doppler (Crystal Biotech, Inc., Hopk-inton, MA) was placed on the vein to monitor the ve-nous signal. These modalities were continued for 5postoperative days. The patient was discharged onpostoperative day 6. After the skin grafts and all woundswere healed, physical therapy was started to maintainpassive range of motion at the elbow, wrist, and digitswithout traction on the muscle, tendon, nerve, or vas-cular repairs.

An advancing Tinel sign was followed distal to thenerve repair. Between weeks 6�8 after muscle trans-plantation, the latissimus was noted to contract at theproximal forearm. The patient was instructed to practice

extending his wrist and fingers multiple times a day tostrengthen these contractions. The contractions pro-gressed, and by postoperative week 13, the patient couldvoluntarily contract the entire length of his latissimusmuscle. In addition, the patient had regained activeflexion of his wrist and fingers. There was no palpableevidence of flexor carpi radialis or brachioradialisfunction. The flexor digitorum superficialis and palm-aris longus functions were weak compared to theopposite hand.

Eight months after latissimus transplantation, thecontralateral gracilis muscle was harvested to restorefinger and wrist extension. During elevation of the la-tissimus muscle, a plane, similar to that found betweenhealthy muscle bellies, could be developed by gentlefinger dissection. The latissimus was elevated over thisplane as a bipedicle flap, allowing the gracilis muscle tobe drawn under it and positioned at a level appropriatefor distal tendon weave (Fig. 2). The extensor digitorumcommunis (EDC) tendons were transected proximal tothe extensor retinaculum and tenodesed in a side-to-sidefashion. The gracilis tendon was then repaired to theEDC tendons outside of the extensor retinaculum sothat the functional microvascular muscle transplantwould ‘‘bowstring’’ across the wrist and restore bothwrist and finger extension. The gracilis was then drawnproximally, extending the wrist and digits. The excessproximal muscle was trimmed, and the origin of themuscle was sewn to the lateral epicondylar fascia andremnants of the extensor carpi radialis origin, placingthe muscle on stretch.

The gracilis muscle was harvested so that its arteryhad a ‘‘T-shaped’’ cuff from the deep femoral artery.The thoracodorsal artery was transected, and the‘‘T-shaped’’ cuff was interposed, not only restoringarterial flow to the gracilis, but also creating a flow-through conduit preserving arterial flow to the latissimusmuscle. The vein was repaired end-to-end to a deep veinin the antecubital fossa. An implantable Doppler wasplaced to monitor the vein. The thoracodorsal nerve wastransected, revealing eight bulging fascicles. It was thentransposed to a position which allowed obturator nerverepair as close as possible to its entrance into the gracilismuscle. Fascicles in the thoracodorsal nerve wererepaired to three small fascicles in the obturator nerve.Excessive proximal fascicles were buried in the gracilismuscle for direct neurotization. The palmaris longus wastransferred to the extensor pollicis longus. All incisionswere closed except for an area at the proximal forearmwhich required a small split-thickness skin graft. Thewrist and digits were splinted in extension. The thumbwas splinted in extension and abduction.

Immediate passive extension was begun to the digitsand wrist, placing no tension on the muscle transplant

Figure 1. (A). Traumatic loss of proximal 2/3 of dorsal muscle

bellies, with (B) radial sided loss of the volar muscle bellies. [Color

figure can be viewed in the online issue, which is available at

www.interscience.wiley.com.]

Temporary Innervation of Primary Coverage Muscle 311

or tendon repairs. At 1 week, external direct currentstimulation was begun on the gracilis muscle. At 3

weeks, passive range in flexion was begun. At 7 weeks,palpable contraction was noted at the proximal forearm,but there was no translation or active extension of thewrist or fingers. Direct current stimulation was discon-tinued, and alternating current stimulation was begun.At 8 weeks, the patient had regained active extension ofhis wrist and digits. Seven months after functionalgracilis muscle transplantation, the patient had regainedexcellent active extension of his wrist and fingers (Fig. 3).

One year after FMMT, gracilis muscle strength wasmeasured at 4+/5, based on the modification by Merled’Aubigne et al. of the Medical Research Council(MRC) grading scale (Table 1).5 We graded the strengthof the gracilis muscle at 4+, based on the fact that thepatient could extend his wrist and finger against 3pounds of resistance vs. 6 pounds of resistance in thecontralateral uninjured wrist and fingers. The patienthad one-half of his measured total active excursionagainst 5 pounds of resistance. Finger flexion wasincomplete, secondary to tethering of the extensor dig-itorum communis tendons (Table 2). Grip strength wasmeasured at 62.5% of the dominant uninjured righthand (Table 3). At time of measurement, finger/wristflexion and grip strength continued to improve. Thumbextension was incomplete and appeared to be the resultof a tenodesis rather than active palmaris longus con-traction. Sixteen months after the injury, the patient wasallowed to return to his previous employment as a car-penter, without restriction.

DISCUSSION

Successful functional microvascular transplantswere defined as muscles that attained M4 strength(based on the MRC grading scale), or the ability toresist 1 kg of weight.4 Experimental6,7 and clinical8�11

studies increased our understanding of functionalmicrovascular muscle transplants. They led us to astaged approach for FMMTs, and defined severalfactors which are necessary to ensure maximum func-tional potential after muscle transplantation. Regard-less, there continues to be a discrepancy in themaximum functional return of functional microvascu-lar muscles transplanted to the traumatized vs. non-traumatized upper extremity.

Chuang reviewed his series of FMMTs in the upperextremity between 1986�1994.4 Twenty-one FMMTswere transplanted in 17 patients who had major soft-tissue and nerve loss. Fourteen of 21 flaps (66%) werecharacterized as successful (muscle strength greater thanM3 according to the MRC grading scale). This wascompared to 15 FMMTs transplanted in 13 patients forVolkmann’s ischemic contracture. Thirteen of 15 flaps(86%) were successful.

Figure 2. A: Latissimus muscle contour prior to FMMT. Incissions

for bipedicle flap are marked. B: Bipedicle flap is elevated in prep-

aration for gracilis placement. C: Gracilis FMMT inset. [Color figure

can be viewed in the online issue, which is available at www.inter-

science.wiley.com.]

312 Brooks

In a different series between 1985�1990,1 5 of 11(45%) flaps transplanted for major soft-tissue traumawere successful (strength >M3), and 8 of 10 (80%) flapstransplanted for Volkmann’s ischemic contracture were

successful. This experience mirrors our own whentreating traumatic injury of the upper extremity withFMMTs.11 The questionable quality of the wound bedand recipient nerve were two factors cited whichdistinguished traumatic injury from other nontraumaticinstances where functional microvascular transplantswere indicated.

Providing an adequate wound bed requires meticu-lous debridement of the wound in the acute setting. Thisoften requires multiple trips to the operating room toexcise demarcating muscle and soft tissue. In woundsthat cannot be closed primarily, Chuang et al.12 pre-ferred immediate free-tissue transplantation for cover-age, namely a myocutaneous or fasciocutaneous flap.They preferred this to early skin grafting and delayedflap coverage or free muscle flap coverage with a skingraft. They claimed that skin grafting followed by laterreconstruction is hazardous, and a transplanted musclewhich has been skin grafted is difficult to handle andwould lead to closure problems.

Smaller soft-tissue defects can be managed withfasciocutaneous or myocutaneous tissue transplanta-tion. In both instances, a healthy plane can be developedinto which a functional muscle can be placed. Aftertransplanting a fasciocutaneous flap, the muscle can beplaced in the healthy subcutaneous plane. After trans-planting a myocutaneous flap, it can be placed in thehealthy plane between the subcutaneous layer and theatrophied underlying muscle. In the case presented, thelarge soft-tissue deficit precluded use of a myocutaneousor fasciocutaneous flap. For this situation, we chose alatissimus muscle transplant. Repair of the thoraco-dorsal nerve resulted in subsequent reinnervation andmaintenance of muscle integrity. Eventual contractionand muscle excursion developed a septo-alveolar planebetween the latissimus muscle and the underlying woundbed. This plane was easily developed with finger dis-section and comparable to the plane between two heal-thy muscles. The latissimus muscle was compliant andwas easily elevated as a bipedicle flap, allowing passageof the gracilis muscle underneath. Closure of the woundrequired only a small skin graft. Denervation of the la-tissimus muscle does not lead to unstable coverage or

Figure 3. Active wrist motion in extension (A) and flexion (B).

C: Comparison of extension. Note mild ‘‘bowstringing’’ of extensor

tendon over extensor retinaculum. [Color figure can be viewed in the

online issue, which is available at www.interscience.wiley.com.]

Table 1. Modification by Merle d’Aubigne’s et al.5 of MRC MuscleGrading System*

M0, no contractionM1, flicker (no joint motion)M2, contraction with mobility with gravity eliminatedM3, contraction against gravityM4, contraction with active motion against gravity and some

resistanceM5, normal power

*MRC, Medical Research Council.

Temporary Innervation of Primary Coverage Muscle 313

restriction of gracilis muscle function, as evidenced byits simultaneous wrist and digit extension, somethingwhich is usually not achieved by a single functionalmicrovascular muscle transplant.1

Assessing the adequacy of the recipient nerve inacute traumatic lesions can be difficult. Delayed nerverepair was advocated.13 After 2�3 weeks, a neuromawill develop and help define the level of reliable nerve.Before neuroma formation in the acute setting,debridement of the nerve proximal to the point that thenerve appears healthy under microscopic magnificationwas advocated.14 Traditionally, nerve repair is per-formed at the time of functional muscle transplantation,after the recipient nerve has formed a neuroma. How-ever, if the nerve is no longer in the area of the proposedFMMT (such as a situation in which the nerve has beenavulsed), strategies to bridge the gap must be employed.

Nerve grafting can bridge this gap prior to FMMT,or it can be done at time of FMMT. The latter situationprolongs the time to reinnervation of the muscle andwas reported to show less than optimal results.15 In ourexperience, the former can be problematic when thetarget is a mixed motor-sensory nerve, because althoughhealthy fascicles may be noted at the end of the nervegraft, histologic evaluation usually reveals nonmyeli-nated axons which are difficult to identify as motor vs.sensory. In the case reported, active contraction of thelatissimus muscle heralded the adequacy of the motornerve at a level appropriate for functional microvascularmuscle nerve repair prior to muscle transplantation.

Our understanding of the mechanisms which regu-late nerve regeneration after nerve injury is based on thework of Forssman16 and Ramon y Cajal,17 who alludedto diffusible factors or neurotrophs, which can affect therate, direction, and overall success of nerve regenera-tion. Since that time, multiple studies have reinforcedtheir conclusions.18�20 Others clarified the mechanism ofaction of these neurotrphic factors.

After nerve injury, there is not only immediate loss ofboth sensory and motor functions, but also a loss ofneurons (axon and cell body). Wilberg et al.21 found thatacute transection of the facial nerve in monkeys resultedin a 15% motor neuron death even if immediate repair

was performed. This 15% loss was associated with a 51%decrease in ultimate functional return. Tornqvist andAldskogius,22 studying neuron loss in the adult rat fol-lowing hypoglossal nerve injury, found that significantlymore neurons were lost if the nerve was specifically notrepaired. Presumably after nerve injury, neurotophicfactors which normally have a protective effect on theproximal axon and cell body are lost. Repair can rees-tablish this factor’s axonal transport proximally andreestablish this protective effect. Early repair of thethoracodorsal nerve innervating the latissimus dorsimuscle in the case presented reestablished peripheral-to-central axonal communication, potentially reducingneuron death and improving ultimate motor function.

When placed in the proximal limb of a ‘‘Y cham-ber,’’ a peripheral nerve will preferentially grow towardnerve tissue rather than non-nerve tissue in an animalmodel. Neurotrophic factors released from degeneratingdistal nerve Schwann cells induce preferential growthtoward neural vs. non-neural tissue. Tsubokawa et al.23

defined a more specific role for neurotrophism’s influ-ence over the regenerating axon. While examining theinductive ability of Schwann cells in a rat model, theyfound that motor Schwann cells had a neurotrophiceffect on motor axons but none for sensory neurons.Theoretically when repaired to a mixed motor-sensoryradial nerve, such as in the case reported, the intactthoracodorsal nerve (pure motor Schwann cells) wouldpreferentially attract motor fascicles, decreasing thepotential mismatching of motor and sensory axonregenerates.

Transplanting the latissimus muscle as a functionalmotor unit with nerve repair also potentiated the‘‘pruning effect’’ in the cases reported.24,25 After radialnerve injury, individual axons sprout, creating regener-ating units.26 Each axon produces multiple unmyeli-nated fibers, increasing the chances that some of themwill find the correct distal nerve and appropriate distalreceptor, restoring function. Axons which reestablishperipheral to central communication will be preservedand myelinated. Inappropriate axons which do notestablish the proper connection will be pruned. Earlyrepair of the thoracodorsal nerve and reinnervation of

Table 2. Total Active Motion, L. Hand at 18 Months*

MP PIP DIP L. TAM R. TAM %a

Index 0/70 0/75 0/50 195 240 81Long 0/70 0/85 0/75 230 255 90Ring 0/60 0/95 0/65 220 255 87Small 0/50 0/95 0/70 215 245 88Wrist Ext 60 65 92Wrist Flex 40 70 57

*TAM, total active motion.aPercentage of right uninjured hand.

Table 3. Strength in pounds at 12 and 18 Months After FMMT

12 months 18 months % (12months)a

% (18months)a

R L R L

Lat 24 14 24 17 58.0 71.0Pinch 3 pt 22 10 28 14 45.0 50.0

2 pt 24 10 24 10 42.0 42.0Jamar IIb 80 30 80 50 37.5 62.5

aPercentage of right uninjured hand at 12 and 18 months.bGrip strength.

314 Brooks

the motor end-plates allow the pruning phase to beginmuch earlier, and therefore increase the probability thathealthy myelinated motor axons might be establishedprior to gracilis muscle transplantation.

Considerable energies are being expended to controlneuron survival, axon regeneration, and target organreinnervation by the introduction of exogenous trophicfactors in animal models.27�29 Until these efforts be-come a clinical reality, simple innovations such as theearly innervation of a coverage muscle will allow us tomanipulate readily available endogenous trophic factorsand influence neuron regeneration and end-organ func-tion.

In summary, our technique offers a number ofadvantages in selected cases of FMMTs. The techniquerequires planning; early nerve repair is simple andtechnically easy. No additional tissues are expended,since the thoracodorsal nerve would be sacrificed withthe latissimus dorsi muscle harvest and transplantation.Active contraction of the latissimus dorsi muscle signalsthe adequacy of the motor nerve graft. Contraction ofthe latissimus dorsi muscle creates an excellent plane forfuture functional muscle insertion. Finally, repair of thethoracodorsal motor nerve creates a ‘‘baby-sitter’’ ef-fect, initiating multiple positive neurotrophic influences,which potentially increase appropriate nerve regenera-tion and neuron survival, and can ultimately maximizefunction of the microvascular muscle transplant.

REFERENCES

1. Chuang DCC. Functioning free muscle transplantation. In: PeimerCA, editor. Surgery of the hand and upper extremity. New York:McGraw-Hill; 1996. p 1901�1910.

2. Manktelow RT, McKee NH. Free muscle transplantation to pro-vide active finger flexion. J Hand Surg [Am] 1978;3:416�426.

3. Manktelow RT. Functioning free muscle transfers. In: Green DP,editor. Operative hand surgery. New York: Churchill Livingston;1993. p 1159�1177.

4. Chuang DCC. Functioning free muscle transplantation for theupper extremity. Hand Clin 1997;13:279�289.

5. Merle d’Aubigne R, Benassy J, Ramadier JO. Chirurgie orthope-dique des paralysis. Paris: Masson; 1956. p 8.

6. Tamai S. Free muscle transplantation in dogs with microsurgicalneurovascular anastomoses. Plast Reconstr Surg 1970;46:219�225.

7. Kubo T, Ikuta Y, Tsuge K. Free muscle transplantation in dogs bymicroneurovascular anastomoses. Plast Reconstr Surg1976;57:495�501.

8. Manktelow RT, Zucker RM. The principles of functional muscletransplantation: applications to the upper arm. Ann Plast Surg1989;22:275�282.

9. Doi K, Sakai K, Ihara K, Abe Y, Kawai S, Kurafuji Y. Rein-nervated free muscle transplantation for extremity reconstruction.Plast Reconstr Surg 1993;91:872�883.

10. McKee NH, Kuzon WM. Functioning free muscle: making itwork? What is known? Ann Plast Surg 1989;23:249�254.

11. Brooks D, Buncke HJ, Buntic RF, Kind GM, Buncke GM.Functional microvascular transplants (FMMTs) in mutilating

injuries of the upper extremity.In:Frey M, Giovanoli P, Koller R,editors. Fifth International Muscle Symposium, May 19�21, 2000.Vienna, Austria, proceedings. Vienna: Division of Plastic andReconstructive Surgery, University of Vienna Medical School;2000. p 82�83.

12. Chuang DCC, Lai JB, Cheng SL, Jain V, Lin CH, Chen HC.Traction avulsion amputation of the major upper limb: a proposednew classification, guidelines for acute management, and strategiesfor secondary reconstruction. Plast Reconstr Surg2001;108:1624�1638.

13. MacKinnon SE, Dellon AL. Nerve repair and nerve grafting.In: MacKinnon SE, Dellon AL, editors. Surgery of theperipheral nerve. New York: Thieme Medical Publishers; 1988.p 89�129.

14. Meyer VE, Stallmach TH, Burg D. Assessment of the nerve qualityat the coaptation site by nerve function evaluation. In: Frey M,Giovanoli P, Koller R, editors. Fifth International Muscle Sym-posium, May 19�21, 2000. Vienna, Austria, proceedings. Vienna:Division of Plastic and Reconstructive Surgery, University ofVienna Medical School; 2000. p 23�26.

15. Chuang DCC, Carver N, Wei FC. Results of functioning freemuscle transplantation for elbow flexion. J Hand Surg [Am]1996;21:1071�1077.

16. Forssman J. Ueber de Ursachen, welche die Wachstumstrichungder peripheren Nervenfasern bei der Regeneration bestimmen.Beitr Pathol Anat 1898;24:55�100.

17. Ramon y Cajal SR. Mechanismo de la degeneracion y regenerationde nervos. Trab Lab Inbest Biol Madrid 1905;4:119�210.

18. MacKinnon SE, Dellon AL, Lundborg G, Hudson AR. A study ofneurotrophism in the primate model. J Hand Surg [Am]1986;11:888�894.

19. Seckel BR, Ryan SE, Gagne RJ, Tin Ho Chiu CST, Watkins E.Target specific nerve regeneration through a nerve guide in the rat.Plast Reconstr Surg 1986;78:793�798.

20. Brushart TM, Seiler WA. Selective reinnervation of distal motorstumps by peripheral motor axons. Exp Neurol 1987;97:289�300.

21. Wilberg M, Vedung S, Stalberg E. Neuronal loss after transsectionof the facial nerve: a morphological and neurophysiological studyin monkeys. Scand J Plast Reconstr Hand Surg 2001;35:135�140.

22. Tornqvist E, Aldskogius H. Motoneuron survival is not affected bythe proximo-distal level of axotomy but by the possibility ofregenerating axon to gain access to the distal nerve stump. Neu-rosci Res 1994;39:159�165.

23. Tsubokawa N, Maki Y, Yoshizu T, Narisawa H. Comparison ofthe neurotropic effects of motor and sensory Schwann cells duringregeneration of peripheral nerves. Scand J Plast Reconstr HandSurg 1999;33:379�385.

24. Denburg JL. Elimination of inappropriate axonal branches ofregenerating cockroach motor neurons as detected by the retro-grade transport of horseradish peroxidase conjugated withwheatgerm agglutinin. Brain Res 1982;248:1�8.

25. MacKinnon SE, Dellon AL, O’Brien J. Changes in nerve fibernumbers distal to a nerve repair in the rat sciatic nerve model.Muscle Nerve 1991;14:1116�1122.

26. Morris JH, Hudson AR, Weddell G. A study of degeneration andregeneration in the divided rat sciatic nerve based on electronmicroscopy II. The development of the ‘‘regeneration unit.’’ ZZellforsch Mikrosk Anat 1972;124:103�130.

27. Derby A, Engleman VW, Frierdich GE, Neises G, Rapp SR,Roufa DG. NGF facilitates regeneration across nerve gaps: mor-phological and behavioral studies in rat sciatic nerve. Exp Neurol1993;119:176�191.

28. He C, Chen Z, Chen Z. Enhancement of motor nerve regenerationby NGF. Microsurgery 1992;13:151�154.

29. Pu LL, Syed SA, Reid M, Patwa H, Goldstein JM, Forman DL,Thompson JG. Effects of nerve growth factor on nerve regenera-tion through a vein graft across a gap. Plast Reconstr Surg1999;104:1379�1385.

Temporary Innervation of Primary Coverage Muscle 315