Bypassing spinal cord injury: surgical reconstruction of afferent and efferent pathways to the urinary bladder after conus medullaris injury in a rat model

Afferent and efferent nerve function in the atonic bladder caused by conus medullaris injury in a rat model was established by intradural microanastomosis of the left L5 ventral root (VR) to right S2 VR to restore pure motor-to-motor reinnervation coupled with extradural postganglionic spinal nerve transfer of L5 dorsal root (DR) to S2 DR for pure sensory-to-sensory reinnervation. Early function of the reflex arc was evaluated by electrophysiological study, as well as by intravesicular pressure measurement and histological examination. The results demonstrated that single focal stimulation of the left S2 DR elicited evoked potentials at the left vesicular plexus before and after horizontal spinal cord damage between the L6 and S4 level. Bladder contraction was successfully initiated by trains of stimuli targeting the left L5-S2 DR anastomosis. Achievable bladder pressures and amplitude of bladder smooth muscle complex action potentials were unchanged before and after induced paraplegia and comparable to those of the control. Prominent axonal sprouting was seen in the distal part of nerve graft. Both afferent and efferent nerve pathways in the atonic bladder can be reconstructed by suprasacral motor-to-motor and sensory-to-sensory nerve transfer after spinal cord injury in rats. This reconstructive strategy has significant potential in clinical application.     

Dumbbell C2 schwannomas involving both sensory and motor rootlets: report of two cases

OBJECTIVE AND IMPORTANCE: Intradural-extradural dumbbell C2 schwannomas are rare. This report concerns two such cases with the intradural compartment located ventral to the spinal cord and involving both sensory and motor rootlets. CLINICAL PRESENTATION: One patient was a 57-year-old woman with sensory disturbances in the right extremities and hyperreflexia in the left extremities. The other patient was a 73-year-old man who presented with tetraparesis, walking disability, atrophy of the nuchal and bilateral shoulder muscles, and pain in the right C2 dermatome. INTERVENTION: The extradural component of the tumor was removed first; next, the intradural component was removed successfully via the posterior approach combined with a C1-C2 laminectomy. The patients experienced symptomatic improvement without further deficits except for sensory impairment of the C2 dermatome in one of the patients. CONCLUSION: Intradural-extradural dumbbell C2 schwannomas can be satisfactorily managed with a posterior approach. Removal of the extradural component and opening of the dural ring of the C2 nerve root are necessary for safe extraction of the intradural ventrally located component after debulking. These tumors may arise extradurally within the nerve sheath, extend intradurally and ventrally toward the spinal cord, and involve both sensory and motor rootlets.     

The feasibility study of extradural nerve anastomosis technique for canine bladder reinnervation after spinal cord injury

Objective: Intradural nerve anastomosis for bladder innervation has been demonstrated to be useful. However, its clinical application remains limited because of the complex surgery, its complications and extensive bony destruction. The purpose of the current study was to demonstrate the feasibility of extradural spinal root anastomosis for bladder innervation in canines.

Methods: Ten beagle dogs were used. The length of the extradural segment of the nerve root, upper nerve root outlet (the point at which it emerges from the spinal dura mater) to S2 (dS2), the S3 (dS3) nerve root outlet distance, and the diameters of the extradural spinal roots were measured. The numbers of nerve fibers from L6 to S3 ventral roots were calculated using immunohistochemical staining.

Results: The extradural spinal roots could be divided into a ventral root (VR) and a dorsal root (DR) before the ganglionic enlargement of the dorsal root, and the extradural motor nerve roots situate ventrally to their corresponding sensory nerve roots. The extradural nerve root lengths of S1 and parts of L7 were longer than the corresponding dS2. The numbers of nerve and motor nerve fibers, and the diameters of extradural nerve roots, were gradually descending from L6 to S3.

Conclusion: The S1 VRs and parts of the L7 VRs can be extradurally anastomosed to the S2 nerves without tension. A nerve graft was needed for extradural anastomosis of L6 VRs and parts of L7 VRs to S2 VRs. This study demonstrated the feasibility of extradural spinal nerve anastomosis for treating neurogenic bladder in canines.     

Sensory sequelae of injuries to the median nerve]

Median nerve wounds are frequent because of the superficial course of this nerve, especially in the wrist. The sensory sequellae of such lesions, either complete or dissociated, represent a major handicap which may be associated to a motor impairment of the external thenar muscles. The curative procedures such as neurolysis and, above all, sutures or grafts, must be very widely used, even some time after the initial injury and in spite of motor reinnervation. The role of palliative surgery is therefore limited, although it is sometimes useful. There are many procedures, which may be summed up by two techniques: use of sensitive neurovascular skin flaps, the most frequent grafts, neurotization of the median nerve with a transfer of sensory rami from the radial nerve, a less-known technique. The treatment of the sensory sequellae of median nerve lesions is primarily based on a good initial management of the nerve lesions by a good-quality emergent microsurgical suture.     

Sensory neurotization of muscle: past,present and future considerations

Research has shown that temporary innervation by a sensory neuron can provide trophic support to a denervated muscle and stave off muscular atrophy until motor neuron transfer is viable. This so called ‘sensory protection’ allows for improved outcomes when motor reinnervation able to occur. The theoretical benefit of sensory neurotization is hypothesized to maintain tissue architecture of the end organ due to tropic effects of stimulation. While the literature supports direct motor neurotization from 2 to 4 months post-injury, patient factors including the location of the injury and loss of nerve can preclude this therapeutic window. When direct neurotization is not possible, or there is a long distance to traverse for reinnervation, sensory neurotization may be beneficial. The theorized trophic stimulation enabling end organ architectural maintenance provided by sensory neurotization has been shown to allow for delayed direct motor neurotization without the irreversible sequelae of prolonged denervation. This is a review of the pathogenesis of nerve injury and a literature review of sensory neurotization. An analytical search of the literature in PubMed was performed in order to find articles pertinent to the topic of sensory neurotization, including experimental data from both animal models and case reports in humans.     

Functional reinnervation of the canine bladder after spinal root transection and genitofemoral nerve transfer at one and three months after denervation

ABSTRACT In the immediate management of patients with spinal cord injury (SCI), patients are typically observed for a period of time to determine whether voluntary control of bladder function returns. Therefore, bladder reinnervation surgeries are not likely to be performed immediately after the injury. We performed genitofemoral to pelvic nerve transfer (GF NT) surgery in canines at 1 and 3 months after bladder denervation (transection of S1 and S2 spinal roots) to determine whether this type of bladder reinnervation surgery has potential clinical feasibility. Nerve cuff electrodes were implanted on the genitofemoral nerves proximal to the pelvic nerve transfer site. Evidence for bladder reinnervation includes (1) increased bladder pressure and urethral fluid flow following electrical stimulation in four out of 20 nerve cuff electrodes implanted on the transferred GF nerves, (2) bilateral pelvic nerve stimulation induced bladder pressure and urethral fluid flow in three of four denervated animals with 1-month delay GF NT, and in five of six denervated animals with 3-month delay GF NT, and (3) abundant L1 and L2 spinal cord cell bodies (the origin of the GF nerve) retrogradely labeled with fluorogold injected into the bladder in all 10 of the GF NT animals, except one animal on one side. This study presents initial proof of concept that GF NT is a potentially viable clinical approach to reinnervation of the lower motor neuron-lesioned urinary bladder.     

Brachial plexus avulsion injury repairs with nerve transfers and nerve grafts directly implanted into the spinal cord yield partial recovery of shoulder and elbow movements

OBJECTIVE: Complete avulsion of the brachial plexus is a devastating injury that primarily affects young adults. The current treatment is based on nerve transfers, which yield very limited recovery. In this study, brachial plexus injuries were repaired with nerve transfers and nerve grafts directly implanted into the spinal cord. METHODS: Eight patients with complete brachial plexus avulsion injuries were surgically treated. Roots or target nerves of the brachial plexus were repaired with peripheral nerve grafts directly implanted into the spinal cord and with extraplexal nerve transfers. RESULTS: Muscle reinnervation was observed for six patients who received spinal implants. Among those patients, one recovered M4 muscle power. Reinnervation was observed only in proximal upper limb muscles. CONCLUSION: Muscle reinnervation through nerve grafts directly implanted into the spinal cord was demonstrated. It seems that the combination of intra- and extradural neurotizations improves the proximal muscle function results. However, the extent of this improvement is limited and, in our opinion, does not justify the use of spinal implants.     

Restoration of elbow flexion in brachial plexus avulsion injury: comparing spinal accessory nerve transfer with intercostal nerve transfer.

This study was performed to compare the clinical outcome of 2 types of commonly used nerve transfers, the spinal accessory nerve transfer and the intercostal nerve transfer. This study was a prospective randomized parallel trial involving 205 patients presenting between 1989 and 1994. All patients were males ranging in age from 16 to 43 years. All patients underwent surgery within 6 months of injury. Spinal accessory nerve transfer was performed in 130 patients; better results were obtained in terms of less operative time, fewer blood transfusions, fewer immediate complications, and better motor function (very good and good power in 83% of patients). Intercostal nerve transfer was performed in 75 patients; better results were observed in terms of earlier electromyographic evidence of motor reinnervation, improvement in protective sensation, and reduction of pain. However, very good and good motor recovery was observed in only 64% of patients. There was no significant difference with regard to tidal volume, vital capacity, and the FEV1 to FEV ratio before and after surgery in either group. Smoking adversely affected the rate of recovery. Spinal accessory nerve transfer should be used when motor function of the elbow flexors is the major concern. Intercostal nerve transfer should be performed in patients who need both motor and sensory reconstruction and in those who have chronic pain syndrome after brachial plexus injury.     

Muscle preservation by prolonged sensory protection

The functional recovery of a muscle target following nerve repair is inversely related to the denervation time: i.e., the longer the muscle denervation, the poorer the functional outcome following nerve reconstruction. The trophic and protective effects of sensory innervation to a motor nerve, following prolonged denervation (greater than 6 months), have been studied. Following proximal transection of the musculocutaneous nerve (MC) close to its C6 origin in 10 adult male Sprague-Dawley rats, the severed nerve was coapted to supraclavicular purely sensory nerves originating from C3 and C4 (sensory protection [SP] group). In another 10 Sprague-Dawley rats, the transected MC nerve was not protected by coaptation to sensory nerves (control group). After prolonged denervation or "sensory protection" (6 months), the MC nerve was then coapted in both groups to the purely motor medial pectoral nerve. Behavioral testing (grooming test) was performed on a weekly basis during the reinnervation time, which lasted 4 weeks. Statistically significant differences (p<0.05) favoring the SP group, were found at the second week of the reinnervation period, but not at the end of the experiment. Evaluation also included intraoperative electrical stimulation of the MC nerve, biceps muscle dry weights, motor endplate counts, and nerve axon counts of the MC nerve. The biceps muscle dry weights were statistically higher in the SP group, along with a trend for a higher number of motor endplates. No statistically significant difference was found in the nerve axon counts of the MC nerve between the two groups. Statistically better intraoperative electrical stimulation results were also encountered in the sensory protection group. An interpretation of the results favors the hypothesis that sensory reinnervation of a motor target may provide the necessary trophic environment to minimize muscle atrophy, until a motor donor nerve becomes available.     

Median to radial nerve transfer for treatment of radial nerve palsy. Case report

The purpose of this study is to report a surgical technique of nerve transfer to restore radial nerve function after a complete palsy due to a proximal injury to the radial nerve. The authors report the case of a patient who underwent direct nerve transfer of redundant or expendable motor branches of the median nerve in the proximal forearm to the extensor carpi radialis brevis and the posterior interosseous branches of the radial nerve. Assessment included degree of recovery of wrist and finger extension, and median nerve function including pinch and grip strength. Clinical evidence of reinnervation was noted at 6 months postoperatively. The follow-up period was 18 months. Recovery of finger and wrist extension was almost complete with Grade 4/5 strength. Pinch and grip strength were improved postoperatively. No motor or sensory deficits related to the median nerve were noted, and the patient is very satisfied with her degree of functional restoration. Transfer of redundant synergistic motor branches of the median nerve can successfully reinnervate the finger and wrist extensor muscles to restore radial nerve function. This median to radial nerve transfer offers an alternative to nerve repair, graft, or tendon transfer for the treatment of radial nerve palsy.     

End-to-side neurorrhaphy of motor nerves: reinnervation of free muscle transplants—first clinical application

End-to-side neurorrhaphy may offer a practical solution in limited cases of nerve reconstruction when no donor nerve or nerve transfer is available for direct end-to-end nerve suture, or when extremely long distances for nerve regeneration cause irreversible atrophy of the targets. We report our experience with the successful clinical use of a completely new technique of end-to-side neurorrhaphy for motor reinnervation of free functional muscle transplants. Since 1995 we have used end-to-side nerve repair for motor or sensory reinnervation in 13 cases. A free functional muscle graft was reinnervated by an end-to-side neurorrhaphy in four patients after tumor resection (two myocutaneous latissimus dorsi flaps), Volkmann's contracture (one myocutaneous latissimus dorsi flap), and in a long-standing brachial plexus lesion (one gracilis muscle flap). All four patients showed reinnervation of their muscle graft through the end-to-side nerve suture site. Two functional muscle grafts to the upper extremity had positive outcome of M3-M4 for elbow and finger extension in one case, and M4 for finger flexion in one case more than 2 years after transplantation. The transplant in the brachial plexus lesion is too early for a final functional result. One patient had a M3 for knee extension 8 months postoperatively. In the clinical cases there was no downgrading of the muscle functions supplied by the "donor" nerve. Good and clinically relevant reinnervation of a functional muscle graft is possible through an end-to-side nerve suture. Proximal avulsion, missing proximal nerve stumps, partial recovery, shortening of long distances for nerve regeneration, and prevention of long nerve grafts in proximal lesions are good indications for reinnervation of a functional muscle transplant by end-to-side neurorrhaphy without harm to the donor nerve. Donor nerves supplying synergists should be preferred.     

Transfer of fascicles from the ulnar nerve to the nerve to the biceps in the treatment of upper brachial plexus palsy

BACKGROUND: The transfer of one or more ulnar nerve fascicles to the nerve to the biceps can restore elbow flexion in patients with upper brachial plexus palsy. The purposes of the present retrospective study were to evaluate the results of this procedure, to measure the delay in reinnervation of the biceps muscle, and to define the indications for a secondary Steindler flexorplasty. METHODS: Thirty-two patients with an upper nerve-root brachial plexus injury were reviewed at an average of thirty-one months after the nerve fascicle transfer. The average age of the patients was twenty-eight years. The average time between the injury and the operation was nine months. Patients were evaluated with regard to reinnervation of the biceps, ulnar nerve function, elbow flexion strength, and grip strength. RESULTS: The average time required for reinnervation of the biceps after nerve fascicle transfer was five months. No motor or sensory deficits related to the ulnar nerve were noted clinically. The average grip strength at the time of the last follow-up was 25 kg (an improvement of 9 kg compared with the preoperative value). After the nerve transfer, twenty-four patients achieved grade-3 elbow flexion strength or better according to the grading system of the Medical Research Council. A Steindler flexorplasty was performed as a secondary procedure in ten patients with persistent grade-3 flexor strength or worse. In eight of these cases, elbow flexion strength improved after nerve transfer and flexorplasty. Overall, thirty of the thirty-two patients achieved a good result (grade-4 strength) or a fair result (grade-3 strength). CONCLUSIONS: We recommend this procedure for brachial plexus injuries involving the C5-C6 or C5-C6-C7 nerve roots. This procedure spares the C5 nerve root and other nerves for grafting or transfer elsewhere. A secondary Steindler flexorplasty is indicated for patients who have persistent grade-3 elbow flexion strength or worse for at least twelve months after nerve fascicle transfer.     

腺病毒介导的LacZ基因靶向性导入脊髓及示踪周围神经的动态过程

目的由损伤的周围神经靶向性导入腺病毒介导的LacZ基因(AdLacZ)至脊髓，动态观察病毒裁体逆行输送到脊髓前角运动神经元、脊髓后根神经节(dorsal root ganglia，DRGs)感觉神经元及基因产物顺行标记周围神经的全过程和特点。方法分别将AdLacZ转染大鼠正中神经和胫神经近断端，然后以10-0无创线吻合神经。在转染后9周内的24个不同时间点取出正中神经组的C5～T1脊髓节段、DRGs连同臂丛，胫神经组的L2～L6脊髓节段、DRGs连同骶丛。将脊髓和DRGs的50μm横切片行X—gal染色和免疫组织化学染色，臂丛和骶丛的整个标本分别行X—gal染色。计数阳性脊髓前角运动神经元、DRGs神经元及周围神经轴突数，研究转基因表达在脊髓前角运动神经元、DRGs神经元及正中神经、胫神经的最早时间、高峰时间和持续时间。结果LacZ基因能特异性、高效表达在损伤周围神经的感觉和运动神经元。转基因在脊髓和周围神经的表达严格限于感染神经的同侧。正中神经组标本各部位的转基因表达均早于胫神经组。胫神经组被标记的运动神经元和感觉神经元数均高于正中神经组。表达持续的时间在运动神经元最短，然后是感觉神经元，在周围神经持续时间最长。在同一组内，转基因表达在DRGs神经元最早．然后是运动神经元，最后是周围神经干，而且被标记的感觉神经元数多于运动神经元数。结论由损伤的周围神经导入的AdLacZ不但在靶神经元高效特异性表达，而且能高效顺行标记神经元突起、周围神经直至吻合口远端的再生轴突。这对周围神经损伤的基因治疗和神经示踪研究有实用价值。     

Histochemical staining of nerve endings as an aid to free muscle transplantation.

Histochemical staining techniques that identify intact motor nerve fascicles are available to aid free muscle transplantation. Cholinesterase activity of myelinated axons can be identified by Karnovsky and Roots's technique. Axon viability can be assessed based on the presence of axoplasmic enzyme activity. By reacting serial sections for cholinesterase activity and carbonic anhydrase activity, which labels sensory axons, an accurate cross-sectional map of regenerating or functional sensory and motor nerve fibers can be constructed. Resolving the motor and sensory identities of fascicles in a mixed peripheral nerve should lead to more precise coaptation of recipient motor fibers to the motor nerve of the transferred muscle and enhance reinnervation.     

End-to-side neurorrhaphy of sensory nerves

One century passed before end-to-side neurorrhaphy was rediscovered, and now it finds more frequent use in clinical practice. Experimental studies have improved our understanding of the underlying mechanism and its potential. However, still discussed is whether reinnervation by end-to-side neurorrhaphy works as well in sensory nerves as in motor nerves. The digital nerves are sensory nerves and therefore an ideal model to investigate this question. Two cases of successful sensory reinnervation by end-to-side nerve suture are reported. We began to use end-to-side nerve repair clinically in 1995 and have used it for motor or sensory reinnervation in a total of 13 cases. In two patients primary nerve repair using end-to-side neurorrhaphy was performed in digital avulsion injuries. In one patient the avulsed ulnar nerve of the thumb was sutured end-to-side to the median nerve; in the other the ulnar digital nerve of the ring finger had been destroyed over a distance of 20 mm, and the distal stump was joined end-to-side to the radial nerve of the same finger. Sensory recovery was obtained in both patients. The static two-point discrimination was 3.0 mm, and dynamic two-point discrimination was 2-3 mm for the reinnervated finger compared to 2 mm for static and dynamic two-point discrimination in the adjacent "donor" finger. The sensation of the finger supplied by the "donor nerve" was not altered in relation to the corresponding contralateral finger site. Excellent sensory reinnervation is possible through an end-to-side nerve suture. Proximal avulsion, missing proximal nerve stumps, partial recovery, and prevention of nerve grafts are good indications for resensitization using end-to-side neurorrhaphy. No harm to the donor nerve is expected. Preference should be given to donor nerves that supply skin areas near to the anesthetic area.     

Results of reinnervation of the biceps and brachialis muscles with a double fascicular transfer for elbow flexion

PURPOSE: To report the results of a surgical technique of nerve transfer to reinnervate the brachialis muscle and the biceps muscle to restore elbow flexion after brachial plexus injury. METHODS: Retrospective review was performed on 6 patients who had direct nerve transfer of a single expendable motor fascicle from both the ulnar and median nerves directly to the biceps and brachialis branches of the musculocutaneous nerve. Assessment included degree of recovery of elbow flexion and ulnar and median nerve function including pinch and grip strengths. RESULTS: Clinical evidence of reinnervation was noted at a mean of 5.5 months (SD, 1 mo; range, 3.5-7 mo) after surgery and the mean follow-up period was 20.5 months (SD, 11.2 mo, range, 13-43 mo). Mean recovery of elbow flexion was Medical Research Council grade 4+. Postoperative pinch and grip strengths were unchanged or better in all patients. No motor or sensory deficits related to the ulnar or median nerves were noted and all patients maintained good hand function. No patients required additional procedures to further improve elbow flexion strength. CONCLUSIONS: Transfer of expendable motor fascicles from the ulnar and median nerves successfully can reinnervate the biceps and brachialis muscles for strong elbow flexion. The reinnervation of the brachialis muscle, the primary elbow flexor, as well as the biceps muscle provides an additional biomechanical advantage that accounts for the excellent elbow flexion strength obtained using this technique. Direct coaptation of the nerve fascicles was performed without the need for nerve grafts and there was no functional or sensory donor morbidity.     

Nerve transfers in brachial plexus birth palsies: indications, techniques, and outcomes

The advent of nerve transfers has greatly increased surgical options for children who have brachial plexus birth palsies. Nerve transfers have considerable advantages, including easier surgical techniques, avoidance of neuroma resection, and direct motor and sensory reinnervation. Therefore, any functioning nerve fibers within the neuroma are preserved. Furthermore, a carefully selected donor nerve results in little or no clinical deficit. However, some disadvantages and unanswered questions remain. Because of a lack of head-to-head comparison between nerve transfers and nerve grafting, the window of opportunity for nerve grafting may be missed, which may degrade the ultimate outcome. Time will tell the ultimate role of nerve transfer or nerve grafting.     

Utility of direct stimulation of roots in spinal surgery

Prevention of postoperative neurological deficits is a major concern of spinal surgeons and has led to the introduction and current development of intraoperative neurophysiological monitoring. We have used motor evoked potentials and somatosensory evoked potentials as routine monitoring techniques and, in some cases, added optional methods such as direct stimulation of nerve roots and spinal evoked potentials. We report our experience of direct nerve root stimulation as an optional monitoring method during spinal surgeries in 7 patients with lesions affecting the proximal nerve roots aged from 1 day to 78 years (mean 23.5 years). Four patients had anomalous lesions, two had spinal nerve root schwannomas, and one had a far-lateral lumbar disc herniation. Direct stimulation was used for detection of motor nerve roots in the anomalous lesions and schwannomas, and to distinguish the nerve root from the paraspinal soft tissues in the case of a far-lateral herniated disc at the L5-S1 level. Although some patients had slight transient neurological symptoms such as motor weakness and sensory disturbance, none developed severe permanent neurological impairment. Direct stimulation allows detection of the motor nerve during spinal surgery in real time. Our limited experience suggests that the direct stimulation technique could reduce the risk of motor or vesicorectal disturbance after surgery of lesions affecting or involving the spinal nerve roots.     

Anatomical and functional connectivity of the transected ulnar nerve after intercostal neurotization in cats.

OBJECT: Acute transfer of three intercostal nerves to the ulnar nerve was performed in cats for histological and clinical evaluation of a distal muscle reinnervation. METHODS: Infraclavicular intercostal-ulnar communications were created after dividing the motor branches of the upper intercostal nerves in 14 adult cats. Reinnervation of distal forelimb muscles in the ulnar territory was assessed by electromyographic (EMG) studies and motor function rating each month until 18 months postsurgery. In five of these treated animals, and in tour controls, horseradish peroxidase (HRP) was applied to the ulnar or intercostal nerves to study the amount and distribution of retrograde motor neuron labeling in the spinal cord. Also, samples of reinnervated muscles and neurotized ulnar nerves were processed to assess regeneration. Simple ulnar transection without reconstruction led to permanent atrophy of ulnar muscles, lack of recovery according to EMG or clinical studies, and disappearance of the ulnar motor neuron pool. In contrast, ulnar neurotization with the intercostal nerves led to a high rate of functional recovery, which began 5 months postsurgery, and progressed from muscle activity synchronized with ventilatory movements to spontaneous movements that were independent of respiration. This recovery was accompanied by substantial retrograde labeling of intercostal motor neurons after HRP application in the ulnar nerve. Cell counts showed that practically the whole motor neuron pool of the involved intercostal nerves contributed to reinnervation of the transected ulnar nerve. CONCLUSIONS: These findings demonstrate that the use of intercostal nerves to neurotize long brachial plexus nerves can achieve long-lasting and successful reinnervation of distal forelimb muscles.     

Reconstruction of finger and elbow function after complete avulsion of the brachial plexus

Simultaneous reconstruction of elbow and finger function with free muscle and nerve transfers after complete avulsion of the brachial plexus (nerve roots C5 to T1) and its long-term results are presented. The basic procedure combined free or vascular pedicle latissimus dorsi muscle transfer with reinnervation by the spinal accessory nerve to obtain elbow and finger flexion, intercostal nerve transfer of the radial nerve to activate elbow and wrist extensors, and suture of the supraclavicular nerve or intercostal sensory rami to the median nerve to restore hand sensibility. Six patients had some or all of these procedures. Postoperative follow-up ranged from 2 to 5 years. Elbow function was restored completely, and some finger flexion was achieved in all cases, although a dynamic splint was necessary to straighten the digits. Patients have continued to improve in grasp power and finger control. This procedure appears to be promising for the restoration of basic hand function in severely handicapped patients.