Abstracts

January 12-13, 2002 Cancun, Mexico

Optimization of Schwann-Cell Adhesion for Peripheral-Nerve Tissue Engineering. Ranjan Gupta, Dara Chafik, David Bear, Arush Patel, Neil F. Jones, Brian Kim, and Clark T. Hung.

Interest in tissue engineering of the peripheral nervous system (PNTE) has developed, as current methods to promote peripheral nerve regeneration have plateaued in improving clinical function. PNTE calls for regeneration by creating a network of channels that guide extending axons to their distal components. Axonal regeneration is encouraged by the neurotrophic factors that the myelinating Schwann cells produce, as well as the molecular components of the basal lamina. Prosthetic tubes have been engineered from individual components of the basal lamina, to encase Schwann cells. These tubes provide both structure and neurotrophic factors, in addition to being guidance conduits. It remains unclear as to how the components of the prosthetic tubules perform, when exposed to physiologic stresses. The purpose of this study was to examine Schwann-cell attachment to assorted scaffoldings, when exposed to shear stress, to help determine the ideal substrate for prosthetic tubes.

Pure Schwann-cell cultures were developed from 3-day-old neonatal Sprague-Dawley rats with the use of the anti-mitotic agent, cytosine b-D arabinouranoside, rat anti-mouse Thy 1.2 antibody and rabbit complement. Schwann-cell-specific mitogens, neuregulin-a and forskolin, were used to expand the Schwann-cell culture. Pre-confluent cells were plated on either poly-L-lysine, collagen, or laminin. Shear stress in the form of laminar fluid flow was generated with the use of custom-designed bioreactors. The laminar flow rate was controlled by a computer-controlled roller pump using LabView software. A sterile fluid reservoir was vented to the CO₂ incubator humidified air. Dose response curves for each substrate were created. Images were captured using an Olympus CK40 microscope attached to a dedicated computer, with a video capture board pre- and post-flow to evaluate cellular lift-off.

Cells on captured images of pre- and post-flow were counted in each picture, and compared to calculate a percentage of retained cells. Schwann-cell adhesion data demonstrated that laminin maintained a higher cellular retention than poly-L-lysine and collagen.

To the authors' knowledge, these were novel experiments that analyzed the effectiveness of the interaction between Schwann cells and scaffolding, as determined by cellular adhesion. Although the results showed that laminin provided optimal Schwann-cell retention, in vivo studies are required to determine the effectiveness of a laminin scaffolding in promoting axonal regeneration. As PNTE is in its infancy, this study can serve to add additional insights as to the optimal design for nerve guidance channels.

Effect of Insulin-Like Growth Factor I (IGF-I) on Axonal Regeneration in Nerve Autografts and Tissue-Engineered Grafts with Cultured Schwann Cells. Hisham Fansa, W. Schneider, G. Wolf, and G. Keilhoff.

To overcome the problems of limited donor nerves for reconstruction, the authors established nerve grafts made from cultured Schwann cells and basal lamina from acellular muscle. These grafts bridged a 2-cm defect of the rat sciatic nerve. Due to their basal lamina that acts similarly to endoneurial tubes, and due to viable Schwann cells, these grafts offer a regeneration that is comparable to autologous nerve grafts. In order to enhance regeneration, IGF-I was locally applied via osmotic pumps. Autologous nerve grafts with and without IGF-I served as controls. Muscle weight ratio was significantly increased in the IGF-treated autograft group, compared to no treatment; no effect was evident in the tissue-engineered grafts.

Autografts with IGF-I application revealed a significantly increased axon count and an improved g-ratio, as indicator for ``maturity'' of the axons, compared to autografts without IGF-I. IGF-I application to the engineered grafts resulted in a decreased axon count, compared to grafts without IGF-I. However, the g-ratio revealed no significant difference between the groups. Local administration of IGF-I improved axonal regeneration in regular nerve grafts, but not in tissue-engineered grafts. Seemingly, in these grafts, the interactive feedback mechanism of neuron, glia cell, and extracellular matrix is not established, and IGF-I cannot exert its action as a pleiotrophic signal.

(This work was supported by grants from the Deutsche Forschungsgemeinshaft, DFG, KE 488/8-1.)

In Vitro Evaluation of Schwann-Cell Growth on Bioabsorbable Materials. Franz Lassner, Gary Brook, Ingo Heschel, Veronique Maquet, Michael Becker, and Norbert Pallua.

Methods of tissue engineering may provide answers to the growing demand for grafting material in peripheral-nerve surgery. The concept of tissue engineering for peripheral nerves aims at providing an artificial scaffold, which serves as a conduit, and seeding this scaffold with Schwann cells to enhance regeneration. In this reported study, the authors have evaluated the growth of Schwann cells on different bioresorbable materials.

Schwann cells were obtained from predegenerated nerves of adult Sprague Dawley rats. They were seeded with a concentration of 1×106 onto different materials. The following materials were evaluated: 1) collagen sheets with smooth surface; 2) collagen sheets with rough surface; 3) collagen sponges with a longitudinal tubular architecture; 4) polylactid filaments; 5) polylactid sheets with a smooth surface; 6) polylactid sheets with a rough surface; 7) multiblock copolymer sheets with a smooth surface; and 8) multiblock copolymer sheets with a rough surface. Schwann-cell growth was evaluated immunohistologically and with viability and proliferation tests.

The best growth of Schwann cells was observed on collagen and polylactid sheets. Very small alterations of the surface could induce directional orientation of the Schwann-cell processes. The coating of the material with laminin and fibronectin enhanced the number of adherent Schwann cells. Only small numbers of Schwann cells could be implemented into the collagen sponges.

The structural prerequisites for Schwann-cell scaffolds as peripheral-nerve conduits were discussed.

Nerve Allografting without Immunosuppression in a Long-Gap Model. David M. Godat, J.G. Yang, L.L. Zhang, P. Narini, I. Shehadi, and H.S. Matloub.

The successful transplantation of allograft nerves without the use of immunosuppression will provide readily available material for nerve reconstruction without donor-site morbidity. Success with various materials has been demonstrated in short-gap models; however, processed allograft has not been tested in a long-gap model. This reported study investigated the feasibility of transplanting processed allograft nerves without immunosuppression in a long-gap (6-cm) model using physiologic, as well as histomorphometric, data.

Sciatic nerves were harvested from outbred rabbits to yield an acellular, freeze-dried nerve graft. The nerves were hydrated in saline just prior to transplantation. Twenty rabbits, weighing 2.5 to 3.0 kg, were divided into four groups. A 6.0-cm gap was created in the right peroneal nerve prior to transplantation. Group 1 (autograft) animals received the contralateral saphenous nerve; Group 2 (fresh allograft) received an unprocessed allograft from an outbred rabbit; Group 3 (untreated group) received no graft; and Group 4 (processed allograft) received the processed allograft. Multiple outcome measures were performed after 6 months, and included moist muscle weight of the extensor digitorum longus muscle, tetanic muscle test, electromyelograms (EMG), and histomorphometric analysis.

Group 4 animals with processed allograft, showed no statistically significant difference in maximal muscle force generated, compared to Group 2, with fresh allograft. The mean muscle force for Group 4 was 78 g, compared to 47 g for Groups 2 and 1.1 g for Group 3. These were all significantly less than the autograft Group 1, 285 g. EMG data were consistent with the muscle force data. Muscle weight was less in Groups 2-4, compared to Group 1. Histomorphometric data were discussed.

Among conclusions were the following. The rabbit long nerve gap is an important model for assessing nerve regeneration. Animals receiving processed nerve allografts with a 6-cm gap did not demonstrate a significant functional improvement. Processed allograft alone appears not to be sufficient to stimulate functional regeneration across longer gaps.

Genetic Engineering of FK520 (Ascomycin) to Produce Non-Immunosuppressant Analogs for Nerve Regeneration. Bruce G. Gold, Andreas Schrimer, Jan Voda, Erin K. Tucker, Dana H.Z. Robinson, Lolita M. Chung, Rika Regentin, John R. Carney, Peter L. Licari, Gary Ashley, Sue Dillon, Leonard Katz, and W. Peter Revill.

FK520 (ascomycin), like FK506 (tacrolimus), is a polyketide that is produced by a Streptomycetes strain of bacteria with immunosuppressant and nerve regenerative properties. Both contain methoxy groups at their C-13 and C-15 positions that are important for immunosuppression, and are also major sites for metabolism by the P450 enzyme CYP-3A4. Based on the authors' speculation that these groups are not necessary for immunosuppression, they generated novel structures with selected modifications at these positions, with the goal of making non-immunosuppressant analogs that retain the nerve regenerative property. Since these methoxy groups are inert to chemical modification, they used genetic engineering to reprogram the FK520 polyketide synthase, to incorporate malonyl, methylmalonyl, and ethylmalonyl extender units (instead of methoxymalonyl) at modules 7 and 8. Several of these compounds potently (0.1-3 nM) increase neurite outgrowth in cell cultures systems (human neuroblastoma SH-SY5Y cells and primary hippocampal neurons). One lead non-immunosuppressant compound (KOS 104-90) was also shown to significantly accelerate functional recovery and nerve regeneration in the rat sciatic nerve crush model, following daily subcutaneous injections (5 mg/kg). The genetic engineering of FK520 offers a practical means to eliminate its immunosuppressant property, and to produce potentially clinically useful drugs for nerve regeneration in large quantities.

Nerve Repair at Different Angles of Attachment: Experimental Study in Rats. Yu-Hui Yan, Ji-Geng Yan, Hani S. Matloub, James R. Sanger, and Lin-Ling Zhang.

To improve nerve regeneration after nerve repair is still the authors' goal. This study attempted to modify traditional transverse nerve repair with a new method of oblique repair.

Sixteen Sprague Dawley male rats were divided into two groups: Group 1 with oblique nerve repair-on the left side, the peroneal nerve was obliquely cut at a 30-degree angle to its axis, which was below the bifurcation between the tibial and peroneal nerves; then, an oblique attachment in suite was carried out with three sutures and 10-0 nylon. In Group 2, the traditional transverse nerve repair was utilized. On the right side in the same animal, the peroneal nerve was transversely cut at a 90-degree angle to its axis, which was at the level equal to the distal point of the oblique line on the right side. A transverse attachment was carried out with three sutures and 10-0 nylon. After a 3.5-month survival period, the following tests were carried out: EMG, muscular tetanic force of the extensor digitorum longus (EDL), EDL muscle moist weight, maximum contractility of EDL tetanic force induced by electronic stimulation. Findings were transferred to a computerized system for measurement and analysis. The proximal and distal portions of the peroneal nerve were harvested and fixed in 2.5% glutaraldehyde in PBS, and then processed routinely. Semi-thin 0.5-um transverse sections were cut and stained with toluidine blue, and were used for nerve-fiber counting and for light microscopy study. Nerve-fiber counts were carried out by means of an image-analyzing computer, and rechecked by hand tally counting. A double-blind approach was used, whereby the surgeon and the technician in the muscle tetanic force test, EMG test, and nerve-fiber counting were blinded to each other. Data analysis was carried out with a computer system, using the t-pair test.

All values on the oblique nerve repair side were greater in amplitude, EDL weight and force, and nerve-fiber count than on the transverse repair side, and the difference was statistically significant in every evaluated value. Among the conclusions were the following. The results of the oblique attachment in nerve repair were superior to traditional transverse attachment. The oblique attachment increased the number of regenerative nerve fibers because of an increase in the area of the attached surface.

Schwann-Cell Migration and Peripheral-Nerve Regeneration in the Autogenous Venous Nerve Conduit (AVNC). Charles Y. Tseng, GuoLi Hu, and David T.W. Chiu.

Axon regeneration following a peripheral-nerve injury requires extrinsic factors that promote growth and provide guidance to the target. Clearly, efforts to promote regeneration would benefit greatly, if the identity of these factors and the cells in which they are produced were known. Toward this end, the authors have characterized the spatial and temporal progression of cellular events that occur during peripheral-nerve regeneration using an autogenous venous nerve conduit (AVNC).

In a rodent model, a 10-mm segment of right sciatic nerve was resected and a segment of autogenous femoral vein was used to bridge the nerve gap. Histologic examination of AVNC samples harvested at intervals up to 2 months after repair, revealed that the process of peripheral-nerve regeneration occurs in four sequential phases. In the hematoma phase (days 1-7), bleeding from endoneurial and epineurial vessels results in the formation of a blood clot that sustains the vital space of the vessel lumen. As phagocytes resorb red blood cells, a pauci-cellular fibrin lattice is left behind, that also serves to maintain vessel patency. In the cell migration phase (days 7-15), Schwann cells migrate into the lumen of the vein conduit from both the proximal and distal nerve stumps. These migratory Schwann cells completely traverse the vital space by about 13 days. Silver staining reveals that no axonal advancement is observed in this phase. In the axonal advancement phase (days 11-31), regenerating nerve fibers grow from the proximal stump into the venous conduit at about 11 days. While the advancing margin of the regenerating fibers is not uniform, it typically reaches the distal nerve stump by about 19 days. In the myelination and maturation phase (day 31 to months), nerve fibers completely fill the vessel lumen and exhibit a wavy pattern. The Schwann cells that had migrated into the conduit now begin to myelinate the advancing axonal fibers. As more time passes, these Schwann cells begin to manifest a more mature morphology consistent with the formation of the nodes of Ranvier.

This study demonstrated that the vessel lumen remains patent throughout the process of nerve regeneration; the vessel walls do not collapse onto themselves. The fibrin lattice, formed as phagocytes resorb red blood cells, facilitates cellular migration into the venous conduit but, by itself, is not amenable to axonal advancement. The findings also suggested that the distal Schwann cells play a critical role in the regenerative process, most likely through the elaboration and secretion of various neuronotrophic and neuronotropic factors.

Neuroma Prevention by End-to-Side Neurorrhaphy: Experimental Study in Rats. Oscar C. Aszmann, Klaus J. Korak, Tobias Winkler, and Manfred Frey.

Management of the painful neuroma has been a subject of controversy since the earliest descriptions of this problem. Currently, treatment has been limited to either resection only, or to resection of the neuroma and implantation of the nerve into a muscle at a location in which it is safe from irritation and/or trauma. However, it is the authors' experience that nerves without a home remain a source of discomfort and pain. In the reported study, the authors explored the feasibility of neuroma prevention through end-to-side neurorrhaphy into adjacent sensory and motor nerves, to provide a target for axons deprived of their end organs.

Experiments were performed in a total of 20 Sprague-Dawley rats, weighing 250 g, under Nembutal anesthesia. Two groups of 10 animals each were prepared. Group 1 served as an anatomic control. In Group 2, the right sensory branch of the femoral nerve was transected at knee level, rerouted through the adductor compartment, and implanted end-to-side, without a perineurial window, to the tibial nerve distal to the trifurcation of the sciatic nerve. After 8 weeks, corresponding sensory neurons were identified by retrograde labeling techniques. Three-dimensional analysis of the end-to-side coaptation was done, and histomorphometric analysis of the proximal and distal tibial nerve segments was performed.

The results of the retrograde labeling of the corresponding sensory neuron pool of the saphenous and tibial nerves showed that the saphenous nerve had contributions L1-L4. The neuron pool of the tibial nerve encompassed L4-L6, with massive labeling of the L5 root. The morphology of the end-to-side coaptation site validated that no neuroma developed; instead, a normal epineurial continuum was formed, and the axons obviously regenerated along the Schwann-cell tubes toward either muscle or skin. Retrograde labeling in the experimental group clearly revealed that the saphenous nerve regenerated along the tibial nerve, with some type of end-organ contact.

The results reported suggest that the axons of a severed peripheral nerve, which are provided with a pathway and target through an end-to-side coaptation, will either be pruned or establish some type of end-organ contact, so that a neuroma can be prevented. Whether these regenerating axons will lead to disturbing sensations such as par- or dysaesthesia in the newly found environment, or remain silent co-dwellers, the experiments reported cannot answer. The long-term results of future clinical work must decide whether the prevention of the neuroma through end-to-side coaptation will be the appropriate therapy for this difficult problem.

Sciatic Nerve Repair with a Single Fascicular Graft. Bettina Meirer and Maria Siemionow.

Clinical strategies to repair injured peripheral nerves concentrate on efforts to attain primary suture of transected nerve ends. Nerve autografts are used when direct repair is impossible. In extensive or multiple site injuries, the availability of autograft material is limited. These authors hypothesized that using a single nerve fascicle bridging the nerve gap, without exact axonal matching, is sufficient to obtain a good functional recovery, and will expand the availability of autograft material.

Twenty-four male Lewis rats were divided into four groups of six animals each. In each group, 1.5 cm of the sciatic nerve was excised. In Group 1 (control), the gap was not repaired; in Group 2 (conventional repair), a 1.5-cm nerve segment was reinserted as a conventional graft; in Group 3 (single fascicle repair, 50% diameter), one fascicle, 1.5-cm long, 50% of a normal-sized nerve segment was used to bridge the nerve gap; in Group 4 (single fascicle nerve repair, 25% diameter), a 1.5-cm-long segment of the sciatic nerve, 25% of a normal sized nerve segment, was used for repair. The rats were evaluated at 3 and 6 weeks after repair. Walking track analysis and toe-spread tests were used to evaluate recovery of motor function. A pin-prick test and somatosensory evoked potentials were used to evaluate recovery of sensory function.

In Group 1, the pin-prick test demonstrated no response at 3 and 6 weeks. SEP latencies (P1: 35.215 N2: 45.16) were significantly prolonged, compared to the fascicle groups (p<0.05). In Group 2, pin-prick sensation was present at the knee level at 3 and 6 weeks. At 6 weeks, P1 (17.85) and N2 (27.22) were significantly prolonged (p<0.05), compared to Groups 3 and 4 (P1: 17.29 and 14.56; N2: 22.84 and 20.20). In Groups 3 and 4, pin-prick response was present at the ankle level in 50% of the animals as early as 3 weeks. At 6 weeks, pin-prick was at the toe level in 50% of the animals in both groups. P1 (17.29 vs. 14.56) and N2 (22.84 vs. 20.20) demonstrated no statistically significant difference between Groups 3 and 4 after 6 weeks.

In this study, sciatic nerve repair using a single fascicle graft resulted in better functional recovery, compared to the conventional repair (p<0.05). If the described method proves to be successful in a larger animal series, it could revolutionize the field of peripheral-nerve surgery.

Biologic Activity of Nerve Growth Factor Slowly Released from Microspheres. Tessa A. Hadlock, Cathryn Sundback, Joseph P. Vacanti, and Mack L. Cheney.

Administration of neurotrophins enhances both the rate and quality of neural regeneration following injury. However, the efficient and sustained delivery of these factors in biologically active form presents a challenge. Delivery of neurotrophic proteins in a biodegradable polymer microsphere vehicle has recently been studied. Protein stability, both during the loading phase and during the extravasation phase, has been questioned. The purpose of the reported study was to load nerve growth factor (NGF) into fully biodegradable polymer microspheres, and to follow the biologic activity of extravasated NGF over time, using both ELISA and a cell culture assay.

NGF was co-loaded with albumin into polylactic-co-glycolic acid (PLGA, MW 60,000) microspheres, using a standard double emulsion technique, with CHC13 as the organic solvent and 1% polyvinyl alcohol for the emulsion step. Microspheres were prepared with an average diameter of 100 microns, and lyophilized. For extravasation studies, aliquots of 10 mg of spheres were suspended in 1 mL of buffered saline, and incubated at 27 degrees C. At 1, 8, 24, and 48 hr, the spheres were centrifuged, the supernatant removed, and the spheres resuspended in 1 mL of fresh saline. The supernatants were stored at -20 degrees C for bioactivity studies. For the ELISA measurements, supernatants were diluted and assayed, using a commercially available ELISA kit. For the in vitro bioactivity measures, the supernatants were added to primed PC-12 culture plates, and the number of neurite extensions per field was quantified. Concentration of biologically active NGF was then discerned, based on control curves establishing the relationship between known NGF concentration and neurite extension.

ELISA and PC-12 neurite extension assays demonstrated good agreement for the initial extravasation time points. However, by 24 hr, a large discrepancy was found between the concentration of NGF predicted by ELISA and that predicted by the cell assay, suggesting a loss of biologic activity over time, despite maintenance of a recognizable epitope for the ELISA assay.

NGF can be stored and released from biodegradable PLGA microspheres. While the protein appears to maintain enough structural integrity to bind its antibody for ELISA measurements, actual biologic activity, as determined using quantitative PC-12 neurite outgrowth measurements, appears to drop off relatively rapidly. The authors' efforts will now focus on methods to preserve growth factor bioactivity over prolonged time courses in the microsphere environment.

Novel Biocompatible Tube Enhanced with Growth Factors Supports Peripheral-Nerve Regeneration. Rajiv Midha, Catherine A. Munro, Paul D. Dalton, Michael K.K. Wong, Charles H. Tator, and Molly S. Shoichet.

As an alternative to nerve autografts, these authors have been developing an artificial tube graft to be used in the repair of significant peripheral-nerve-injury gaps. One goal was to incorporate neurotrophic factors into the grafts. To date, they have investigated 1) the incorporation of three different factors into the graft; and 2) the appropriate concentration of a given neurotrophic factor to be incorporated. Factors studied have included brain derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), and alpha fibroblastic growth factor (FGF-1). Biocompatible tubes, made from poly(2-hydroxyethyl methacrylate-co-methyl methacrylate, PHEMA-MMA), 1-cm long with an inner diameter of 1.6 mm and outer diameter of 2.1 mm, were used to repair gaps surgically created in the rat sciatic nerve.

Prior to surgery, the inner lumen of the tubes was filled with a collagen matrix, with or without neurotrophic factors (NT-3, BDNF, or FGF-1, 1.0 ug/ml concentration). FGF-1 was also tested at three concentrations: 0.1, 1.0, and 10.0 ug/ml (low, medium, and high). Eight weeks following entubulation, nerve regeneration through the tubes was assessed by morphometric analysis of tissue at mid-graft level, and in the nerve distal to the tube repair. While nerve regeneration was observed across all types of tube repairs, it was improved with the addition of neurotrophic factors, compared to empty tubes and those containing collagen gel alone. High concentration of FGF-1 significantly increased the total number of myelinated axons regenerating into the tube graft, compared to low concentration FGF-1 (t&equals;3.2, p<0.05).

Neurotrophic factors appear to improve peripheral-nerve regeneration into and through the tube grafts. This improvement appears to be dose-dependent, with increasing concentration being more favorable to axonal regeneration.

Comparison of Neuromuscular Junction Density and Structural Differences between Juvenile and Adult Rats. Jianjun Ma, Thomas L. Smith, Beth P. Smith, Eileen Rosencrance, and L. Andrew Koman.

Botulinum A toxin (BTX-A) blocks neurotransmission at the neuromuscular junction, produces a variable degree of muscle denervation, and promotes the improved balance between agonist and antagonist muscles. BTX-A injections have been used to manage muscle spasticity arising from cerebral palsy or nerve palsy, such as sixth cranial nerve palsy. The appropriate pediatric dose regimen is unknown and, currently, dosages are empirically based on results from adults. Any differences in NMJ structure or density between juvenile and adult animals could help define the appropriate pediatric dose.

Bilateral biceps and gastrocnemius muscles were harvested from 10 Sprague-Dawley rats (five 1-month-old rats and five 6-months-old rats). Left-sided muscles were analyzed for NMJ (neuromuscular junction) density, and right-sided muscles for NMJ structure. For NMJ density, a modified cholinesterase staining was used to identify the motor endplates in frozen muscle sections. The NMJ density was calculated as total number of motor end plates in each muscle divided by the muscle weight. For NMJ structure, TRITC-alpha-bungarotoxin labeling was used to label the postsynaptic side of NMJs. Confocal microscopy was used to measure gutter depth, surface area, and length of the NMJs.

Adult and juvenile rats were found to have a similar distribution of NMJs. However, NMJs were smaller in size and had shorter gutter depths in the juvenile rats. The total number of NMJs in both age groups was nearly equal, resulting in a two-fold increase in NMJ density or more than 100,000 NMJs/gram muscle tissue in the juvenile biceps.

BTX-A dosing is currently determined by total body weight. Based on the reported results, dosing based on the number of NMJs (the density of NMJs), rather than on the mass of the muscle being targeted, may be more appropriate when determining the correct dosage for pediatric patients. The higher NMJ density documented in juvenile rats suggests that the appropriate pediatric BTX-A toxin dose is probably higher than that required for adult patients.

Microsurgical Neurolysis of Old Peripheral-Nerve Injuries: Intra- and Postoperative Analysis. Shimon Rochkind and Malvina Alon.

A study was conducted of patients suffering from old peripheral-nerve injuries, in order to demonstrate that functional improvement was possible after microsurgical management.

The prospective study was carried out in 57 patients with old peripheral-nerve injury. The time period from injury until surgery ranged from 1.2 to 50 years. Responsible mechanisms included gunshot wounds, stab wounds, traction, contusion, and compression injuries. Patients were treated by external and/or interfascicular neurolysis. Intraoperative nerve conduction studies and compound muscle action potentials (CMAP) were used during surgery to identify functional activity of the nerve and fascicles.

Intraoperative electrophysiologic recordings of CMAP demonstrated low amplitude at the onset of surgery and a significant increase on completion of surgery, especially after interfascicular neurolysis (p&equals;0.0001), with improvement of latency (p&equals;0.001). The statistical analysis of recruitment of muscles showed a 60% improvement 1 year after neurolysis. There were no statistically significant differences in the results from patients under 40 years of age, compared to those above 40 years of age. Thirty-six (63%) of the 57 patients showed moderate-to-good results in motor function from 1 year to 2 years after surgery.

Intraoperative and postoperative electrophysiologic analysis and clinical follow-up suggested that the survival and viability of injured nerve tissue are of longer duration than previously considered, thus extending the time in which nerve recovery can occur. This study demonstrated that the use of microsurgical treatment can result in the functional improvement of patients suffering from old injuries of the peripheral nerve.

Surgical Repair of Brachial Plexus Injury: Multinational Survey of Experienced Peripheral Surgeons. Allan J. Belzberg, L. John Moriarity, and Michael Dorsi.

Surgical repair of the brachial plexus is one of the most complex problems facing the peripheral-nerve surgeon. The purposes of this study were multiple: 1) to determine what, if any, is the consensus regarding general and specific approaches to brachial plexus repair; 2) to provide the less experienced practitioner with a brief and rapid overview of a variety of brachial plexus injuries; and 3) to help define questions for future clinical study by identifying areas in which multiple experts use significantly different approaches to the same injury.

A literature review and the membership of various professional organizations generated a list of 126 peripheral-nerve surgeons. Each received a survey containing a list of general questions and textual/graphical descriptions of four clinical cases (birth-related upper plexus palsy, and three adult cases: total, upper, and lower plexus avulsion). In total, 49 surgeons (39%) returned completed questionnaires.

As a group, the 49 respondents represented three continents and 22 countries. In response to general issues, there was good agreement. For example, 80% (40/59) of respondents preferred CT myelography to detect preoperative root avulsion, while 73% (36/49) of respondents thought that elbow flexion was the most important for motor recovery. The authors tabulated and graphically represented the most frequent nerve transfers employed in each of the clinical cases. As presented, there was a significant divergence of opinion in each.

Surgical repair of the brachial plexus is a complex problem and, not unexpectedly, experienced practitioners often disagree about what, if any, is the best surgical approach. This study helped to underscore that fact and helped identify areas worthy of future clinical study.

Disability and Quality of Life Measures across Traumatic Upper Extremity Peripheral-Nerve Lesions Using the DASH and SF-36. Marc Dupere, Dorcas Beaton, and Dimitri Anastakis.

Several instruments have been designed to measure impairment or disability after upper extremity injury. Using the SF-36 and DASH questionnaires, this study measured disability and quality of life in patients with upper extremity peripheral-nerve injuries.

A total of 51 patients with injuries to one or more peripheral nerves of the upper extremity were included in the study. The SF-36 and DASH questionnaires were self-administered at initial presentation and at 6, 12, and 24 months following treatment. Statistical analysis was performed using the Statistical Analysis System version 6.12. ANOVA and descriptive univariate analysis were used to compare data across lesions.

There were 45 males and 6 females. Patient ages ranged from 17 to 68 years, with a mean age of 35 years. There were 27 low nerve lesions (median nerve: 14, ulnar nerve: 11, radial nerve: 2), 18 high lesions (median nerve: 10, ulnar nerve: 5, radial nerve: 3), and six combined nerve injuries. The majority of nerves were divided at the time of injury (72%). Treatment included primary nerve repair (35%), neurolysis (20%), and nerve grafting (18%).

With the DASH, grouped high nerve lesions were associated with greater disability than grouped low nerve lesions (p<0.05). Radial nerve lesions were associated with the greatest degree of disability, followed by median nerve lesions, and last by ulnar nerve lesions. Over a 2-year period following treatment, there was a consistent decrease in disability, as measured with the DASH, for each individual nerve lesion.

The SF-36 analysis showed a marked decrease in quality of life (QOL) for all individual nerve lesions, compared to normative Canadian values, particularly regarding physical functioning and bodily pain. SF-36 scores improved slightly at 12 months post-treatment, but values never reached normative levels.

High nerve lesions were associated with the greatest amount of disability. Radial nerve lesions were associated with the greatest amount of disability. With treatment, the amount of disability decreased with time for all lesions. Patients with peripheral-nerve injuries had a poor quality of life. Despite decreasing disability with time, SF-36 scores never returned to normative values. The DASH and SF-36 questionnaires may be useful outcome measures in patients with peripheral-nerve injuries.

Diagnosis and Documentation before and after Neurolysis and Scalenectomy for Brachial Plexus Compression (Thoracic Outlet Syndrome). Michael Howard, A. Lee Dellon, Kathy Lee, and Janice Conolley.

Diagnosis and documentation of brachial plexus compression in the thoracic inlet, ``thoracic outlet syndrome,'' remain difficult. This is because electrodiagnostic testing is inadequate for this purpose, and the syndrome overlaps with peripheral-nerve compression syndromes and cervical disc disease. In 1999, Gilbert Lee and co-workers applied the use of neurosensory testing with the Pressure-Specified Sensory Device (PSSD) to this problem, with excellent results. The hypothesis was that the cutaneous pressure threshold for discriminating one from two points would increase in the index finger (representing the upper trunk of the plexus) and the little finger (representing the lower trunk of the plexus), if the plexus is provoked by elevating the arms in the air for 3 min. Their study demonstrated a significant difference between normal controls and 11 preoperative (no postoperative) patients.

In this reported study, in addition to the PSSD, pinch and grip strength (Digit-Grip) were tested and, in contrast to the previous study, patients were not tested while their hands were still in the air. Sixteen controls (mean age: 34.2 years, range: 11-48 years) were tested, and demonstrated a 99% confidence limit of <10% change in any parameter after elevation of the hands for 3 min. Fifty-four patients symptomatic for brachial plexus compression (mean age: 41.0 years, range: 21-62 years) were also tested. The mean increase in pressure threshold (gmsqmm) was more than 50% for one-point static touch, more than 100% for two-point static touch, and more than 25% for two-point discrimination (mm). The mean decrease in pinch and grip strength was more than 50% in the dominant, and more than 15% in the non-dominant, hand. These differences between control and clinical populations were significant (p<0.001). Fourteen patients were tested before and after plexus neurolysis and anterior scalenectomy. Patients who had an excellent clinical result had improvement in PSSD testing, documented by return to a normal pattern (no increase in thresholds with hand elevation). Sensitivity, specificity, and positive predictive values for the PSSD and brachial plexus compression were given.

It was concluded that neurosensory (PSSD) and motor testing (Digit-Grip) can help in the diagnosis and documentation, and therefore in the management, of this difficult clinical peripheral-nerve problem.

Ultrasonographic Detection of Iatrogenic Accessory Nerve Palsy. Peter Kovacs, Gerd Bodner, Alexander Gardetto, Hildegunde Piza-Katzer, and Werner Jaschke.

The aims of this study were to report the ultrasonographic findings in four cases of iatrogenic palsy of the external branch of the accessory nerve after lymph-node biopsy and neck dissection, and to demonstrate the feasibility of ultrasonography (US) in detecting the accessory nerve in three cadaveric specimens.

Four patients (mean age: 54 years, range: 42-65 years) presented with neck and shoulder pain after surgical intervention in the neck region (two after lymph-node biopsy, two after neck dissection). US was performed with a linear broad band transducer (5-12 MHz) working on an HDI 5000 (ATL). In three fresh cadavers, the external branch of the accessory nerve was detected by means of a 5-12 MHz broad band linear transducer, and marked with blue ink on entering the trapezius muscle. Subsequently, the lateral cervical region was carefully dissected and the accessory nerve was exposed.

In the right lateral cervical region of two patients who underwent lymph-node biopsy, a hypoechoic mass was found, in which a tubular structure ended. A cut accessory nerve was suspected. In two patients with unilateral neck dissection, US revealed a tubular structure ending in a hypoechoic scar measuring 3 cm in length in the right neck region. In all cases, the trapezius muscle demonstrated hyperechoic texture, suggesting muscular atrophy. The US findings of accessory nerve lesions and trapezius atrophy were confirmed by electroneurographic testing in all cases, and surgical nerve inspection in two cases.

On US evaluation of the accessory nerve, there was a small tubular structure in a transverse plane (diameter: 0.7 mm), lying just under the superficial layer of the cervical fascia. In all specimens, the nerve was correctly detected, with US demonstrating blue-stained accessory nerves in all cadavers after dissection.

The capability of high resolution ultrasonography allows the depiction of a normal accessory nerve and detection of iatrogenic accessory nerve palsy.

Technical Considerations in Intraoperative Recording of Motor Nerve Action Potentials from the Brachial Plexus after Transcranial Electrical Stimulation. Marco Vennettilli, David A. Houlden, and Rajiv Midha.

Ventral nerve root integrity is essential for beneficial nerve grafting within the brachial plexus. Motor nerve action potentials (MNAPs) recorded from the brachial plexus after transcranial electrical stimulation (TCES) may be useful in assessing the functional integrity of ventral nerve roots during brachial plexus surgery. It has been shown that repetitive TCES is more effective in evoking muscle responses than single-pulse TCES. This case study demonstrated the stimulating and recording parameters for reliable acquisition of brachial plexus MNAPs and muscle responses after repetitive TCES in a 21-year-old woman with thoracic outlet syndrome.

Intraoperative TCES was performed, using a train of three stimuli, each stimulus separated by an interval of 1.6 msec. The anode was placed at C3 and the cathode at C4 (International 10-20 system). The stimuli were delivered by a Digitimer D185 Stimulator. MNAPs were obtained from hook electrodes placed on the upper, lower, and middle trunks of the exposed right brachial plexus. The amplifier gain was 50 uV/div and the recording bandpass was 100-2000 Hz. Right hypothenar (HT) MEPs were concomitantly obtained from needle electrodes. The amplifier gain was 200 uV/div and the recording bandpass was 10-2000 Hz. The sweep time was 100 msec for both MNAP and MEP recordings. Display scale and sweep time were adjusted for optimal presentation of MNAP and MEP waveforms.

MNAPs were successfully recorded from the upper, lower, and middle trunks of the brachial plexus using a TCES intensity of 250V. The repetitive TCES stimulus artifact did not interfere with accurate interpretation of the MNAP waveform. The TCES threshold (225V) was the same for both the lower trunk MNAP and the right HT MEP. Both MNAP and right HT MEP amplitude increased when the TCES intensity increased to 250V. When the TCES intensity increased to 275V and 300V, the MNAP amplitude continued to increase, but the HT MEP did not.

Intraoperative TCES using a short train of stimuli was effective in evoking MNAPs from all trunks of the brachial plexus. This short train of stimuli resulted in a decreased stimulus artifact which insured that all brachial plexus NAPs could be completely visualized and, therefore, accurately interpreted. At higher TCES intensities, the MNAP became larger, but the HT response did not. This was likely due to recruitment of more motor axons supplying lower-trunk innervated muscles other than the HT.

Correlation between Histopathology of C5 and C6 Nerve Stumps and Motor Recovery Following Nerve Grafting in Brachial-Plexus Repair. Martijn J.A. Malessy, Sjoerd G. van Duinen, Hans K.P. Feirabend, and Ralph T.W.M. Thomeer.

The correlation between the histopathology of C5 and C6 nerve stumps and the attained force of the target muscle following nerve grafting in brachial-plexus repair was analyzed. Four histopathologic parameters were examined in frozen-tissue sections of 31 C5 or C6 stumps by a neuropathologist. The total quantity of myelin was compared to normal. Also, thick myelinated fibers, fibrosis, and misguided axons were assessed. Plastic-embedded stumps were used in a morphometric study of myelinated fiber profiles. The fiber density, mean size, and size distribution in five donor stumps were determined. Three normal C5 spinal nerves obtained at autopsy served as controls. Finally, the relative area occupied by fiber profiles and interspace was computed. Linear regression was used as a multivariate analysis, adjusting the outcome of surgical repair for effects of age, interval between trauma and surgery, and graft length.

Histopathologic examination showed the total quantity of myelin in donor stumps used for biceps reinnervation to be considerably reduced. On morphometric examination, the fiber density did not differ significantly between patients and controls; however, a significant reduction of the area occupied by myelinated fibers was measured: from 46% in controls to 13% in patients (p<0.0001). Likewise, a significant reduction was found in mean fiber size: from 7.4 um in controls to 3.7 um in patients (p<0.0001). The relation between myelin quantity of the proximal stump and the grade of biceps recovery was statistically significant (p&equals;0.02). From the 95% confidence interval, it was concluded that the estimated effect of a mean increase of myelinated fibers by 25% almost equaled an increase of one point on the 0-5 Medical Research Council scale.

Because it is possible to assess the quantity of myelin by intraoperative frozen-section examination, this correlation is potentially useful in the decision process as to whether to use stumps for grafting, or to apply a nerve transfer procedure, in order to restore biceps muscle function.

Changing Central Control Following Intercostal Nerve Transfer. Martijn J.A. Malessy, Ralph T.W.M. Thomeer, and J. Gert van Dijk.

Changing central nervous system control, following intercostal (IC) to musculocutaneous (MC) nerve transfer in patients with root avulsions of the brachial plexus, was analyzed. Physiologically, IC nerves are controlled by central nervous system (CNS) mechanisms involved in respiration and posture control. Early in the reinnervation process, biceps contraction is effected only by means of volitional vigorous respiratory effort or by coughing and sneezing. However, at the end stage of reinnervation, voluntary contraction of the biceps muscle can be initiated and sustained independent of respiratory activity. Because the control of the reinnervated muscles appears to change with time, some kind of ``rewiring'' of CNS connections must take place. The goal of this study was to find which central nervous system pathways are involved in volitional control over reinnervated biceps muscles.

The facilitating effects of respiration and voluntary contraction on cortical motor-evoked potentials of biceps muscles were used. In addition, maps of magnetically excitable cortical areas of the affected arms were compared with maps of their healthy arms, and maps of both arms of healthy control subjects. The intercostal cortical area was also studied, requiring needle electromyography mapping. The time course of the facilitating effect of respiration and voluntary contraction differed significantly. In the end stage of nerve regeneration, the facilitating effect of voluntary contraction was significantly larger than that of respiration. The cortical areas of affected arms were smaller and less excitable than those of healthy arms. The locations of these areas could not be distinguished from that of the normal cortical biceps area, but seemed to differ from that of the intercostal cortical area.

These results indicated that the CNS control network over the muscle comes to resemble that of the recipient MC nerve, rather than that of the IC donor nerve. The strengthening of previously sub-threshold synaptic connections in a CNS network connecting IC to MC neurons, may underlie changing excitability. Cerebral plasticity does occur in IC to MC nerve transfers, and may be crucial for their clinical success.

Co-Contractions in Obstetrical Brachial Plexus Palsies: Treatment Concepts and Results. Michael H.-J. Becker, Shahab Hamzavi, Norbert Pallua, and Giulio Ingianni.

In cases of obstetrical brachial plexus (OBP) palsies, after spontaneous regeneration as well as nerve revision, co-contractions of agonists and antagonists often minimize the functional result. In these situations, range of movement is usually reduced by deficits of extension and flexion. Spontaneous movement itself is usually delayed. Therapeutic possibilities include corrections on a neural level, a musculotendinous level, or a temporary weakening of antagonists, depending on the age of the patient and the level of co-contractions.

These authors have performed 31 corrections of co-contractions in patients with OBP palsies during the last 3 years. Three different methods were used. Neural corrections were carried out in young patients (<2.5 years), in whom the weaker muscle group was corrected by extraplexus neurotization (11x accessory nerve). Secondary procedures were carried out in older patients (>2.5 years) with severe co-contractions, combining muscle groups (12x). For temporary weakening, in cases with mild co-contractions, botulinus toxin (8 pat a 30-100 U Botox/muscle) was injected in the dominant muscle group.

Regarding neural corrections, in nine cases the problem was adequately solved. The best cases regained a complete flexion and extension of the elbow with quick and spontaneous movements. In secondary procedures, mainly indicated and performed in co-contractions of the shoulder muscles, the active range of movement in respect to abduction and anteversion increased about 60 degrees. Active internal rotation was reduced in six patients. Using injections, with temporary weakening of the antagonist, the force of the corresponding agonist increased temporarily; permanent success was seen in only three patients.

Extraplexic neurotization is a safe method for solving the problem of co-contractions in young patients. Combining muscle groups (e.g., subscapularis tendon with supraspinatus tendon) increased the active ROM in the aimed direction. A simultaneous functional defect of the transposed musculature could be observed in 50% of the patients. In the patients treated by injection, fewer than 50% received a permanent solution (mainly the younger patients). Secondary problems such as shrinking of the musculature, adhesions of capsules at the joint level, etc., appeared to negatively modify the results in late cases.

Establishing a Positivity Criterion in Determining the Utility of the Ten Test in Diagnosing Carpal Tunnel Syndrome. Brent E. Faught and Nancy H. McKee.

Alterations in hand sensation are typically the initial complaints from patients with median nerve compression (Szabo et al., 1984). In 1997, Strauch et al. described the ten test, an instrumentless quantitative sensibility test, that had good inter- and intra-observer reliability. Patel and Bassini (1999) further validated the ten test by comparing it to the Weinstein enhanced sensory test, and static and moving two-point discrimination. They determined that the ten test could grade minimal loss of sensation in hands with carpal tunnel syndrome. To date, sensitivity, specificity, and likelihood ratio have been reported for Phalen's test, Tinel's sign, and the pressure provocative test as related to the diagnosis of carpal tunnel syndrome, but not for the ten test. Therefore, this reported study hypothesized that the ten test would demonstrate utility in diagnosing carpal tunnel syndrome.

Patients referred for examination of possible carpal tunnel syndrome formed the subject base for this study. The attending surgeon systematically administered clinical tests in the following order: 1) Phalen's wrist flexion test; 2) pressure provocative test; 3) ten test; and 4) Tinel's sign. Patients returned for an independent and blinded electrodiagnostic evaluation by the attending physiatrist. A positive electrodiagnosis, combined with a ``classic'' or ``probable'' rating from Stirrat's self-administered symptom reporting questionnaire, was the gold standard in the current study.

The sample of 92 patients (173 hands) included 65 females (mean age: 48±15 years) and 27 males (mean age: 53±13 years). Overall, the hand prevalence of carpal tunnel syndrome was 0.62±0.07 (108 +CTS and 65 -CTS hands). The average reported duration of symptoms was 1.96±0.02 years. ROC analysis established a positivity criterion for carpal tunnel syndrome in patients reporting sensibility scores <10 in at least three of the thumb, index, middle, and ring fingers. Tinel's sign (sensitivity&equals;79%, specificity&equals;65%, likelihood ratio&equals;2.22), the ten test (sensitivity&equals;87%, specificity&equals;52%, likelihood ratio&equals;1.8), and Phalen's test (sensitivity&equals;80%, specificity&equals;48%, likelihood ratio&equals;1.52) demonstrated significant kappa agreement (p<0.001) with the gold standard. The pressure provocative test (sensitivity&equals;76%, specificity&equals;34%, likelihood ratio&equals;1.15) did not indicate significant kappa agreement (p>0.05). The combination of the ten test with Tinel's sign (sensitivity&equals;73%, specificity&equals;50%, likelihood ratio&equals;1.46) demonstrated the highest significant kappa agreement (p<0.01) among combined clinical tests.

The ten test demonstrated a high degree of utility in the current study. The positivity criterion for carpal tunnel syndrome was identified as hand sensibility scores <10 in at least three of the thumb, index, middle, and ring fingers.

Radial Approach to Carpal Tunnel Release in Conjunction with Thumb Carpometacarpal Arthroplasty. Michael S. Wong, Makoto Tamai, and Tsu-Min Tsai.

The prevalence of carpal tunnel syndrome (CTS) in those patients with thumb carpometacarpal (CMC) arthritis may be as high as 43%. Because unrecognized CTS can result in additional postoperative pain and weakness, and can even precipitate a reflex sympathetic dystrophy following thumb CMC arthroplasty, some advocate carpal tunnel release (CTR) at the same time. To avoid the added morbidity of a second incision, the authors have begun using a radial approach to CTR, utilizing the same incision as for thumb CMC arthroplasty.

Between June, 2000 and May, 2001, eight patients (one male, seven females), with an average age of 53.8 years (range: 38 to 70 years), who had both thumb CMC arthritis and CTS, underwent CMC arthroplasty with ligament reconstruction and CTR via a radial approach through the standard thumb CMC arthroplasty incision. All patients had a history of basal thumb joint pain, a positive thumb CMC grind test, and x-rays consistent with thumb CMC arthritis. All patients also had symptoms of numbness and tingling in the median nerve distribution, positive provocative tests (i.e., Phalen's and Tinel's), and nerve conduction studies consistent with CTS. Patients had all failed conservative therapy, consisting of oral anti-inflammatory medications, splinting, and steroid injection, and thus were recommended thumb CMC arthroplasty, combined with CTR and using a radial approach. In all procedures, the trapezius was resected and the joint debrided. CTR was then performed, dividing both the superficial and deep portions of the radial side of the transverse carpal ligament (TCL).

There were six procedures performed on the right side and two on the left. Three patients were operated on for recurrent CTS, following previous open procedures. Four patients had additional procedures performed under the same axillary block, including two ganglion excisions, a trigger finger release, an excision of a subcutaneous forearm mass, and a DeQuervain release. Patients were followed for an average of 14 weeks (range: 1.7 to 24.8 weeks) postoperatively. All eight patients had improvement in their pain and numbness following surgery. One patient had pillar pain persisting at 19 weeks follow-up. One patient had basilar thumb pain at 19 weeks, although this was improved over preoperative condition. No recurrent motor branch injuries were noted.

It is not uncommon to have patients with both thumb CMC arthritis and CTS. These authors have found that a thumb CMC arthroplasty and CTR may be safely performed together through the same incision, by using a radial approach to the division of the TCL.

Treatment of Painful Neuromas Using Vein Implantation. Ryosuke Kakinoki, Taiichi Matsumoto, Ryosuke Ikeguchi, and Takashi Nakamura.

These authors have often experienced difficulty in treating painful neuromas. Here, they described the treatment of six patients with six painful neuromas, using the vein implantation method described by Herbert and Filan. They also carried out an experimental study using rats, in order to identify the morphometric time-course changes of a nerve stump that was inserted into a vein stump after the removal of a neuroma.

In the experimental study, 20 Sprague-Dawley rats were used. The left femoral nerve was exposed and ligated with 5-0 nylon suture to form a neuroma. The nerve was exposed again 3 weeks after ligation. The femoral vein was incised, and the proximal nerve stump was inserted into the lumen of the proximal stump of the vein after removal of the neuroma. The epineurium was sutured to the vein stump using 10-0 nylon suture. Five rats each were sacrificed at 2, 4, 6, 8, and 12 weeks after vein implantation. The neurovascular junction was harvested from each rat and examined by both light and electron microscopy.

In the clinical study, six patients with six neuromas were treated using the vein-implantation method. One had a neuroma in the superficial branch of the radial nerve, one in the saphenous nerve at the knee, one in the saphenous nerve in the lower extremity, one in the palmar branch of the median nerve, one in the dorsal branch of the ulnar nerve, and one in the superficial peroneal nerve. Four neuromas were caused by traumatic injuries, another by attempted suicide, and the other by TKA procedure. The period between injury and surgery varied from 3 months to 3 years 5 months (mean: 1 year 2 months). Surgery was performed as described by Herbert and Filan.

In the experimental study, no nerve stumps formed neuromas in the vein lumens. A layer of endothelial cells covered the nerve stumps in the veins 12 weeks after vein implantation. In the clinical study, pain was relieved, and Tinel's sign disappeared in four of the six patients. Slight pain remained in one patient. In the remaining patient with a neuroma in the saphenous nerve, followed by a TKA procedure, pain and the Tinel's sign disappeared immediately after vein implantation. However, 4 months later, the patient again noticed pain. She later underwent vein implantation again, which relieved the pain. These results showed that vein implantation is an excellent procedure for the treatment of painful neuroma.

Peroneal Intraneural Ganglia-Importance of the Articular Branch: A Unifying Theory. Robert J. Spinner, John L.D. Atkinson, David G. Kline, and Robert L. Tiel.

Peroneal intraneural ganglia have received disproportionate attention over the past century, largely due to the controversy regarding their pathogenesis. The authors' unifying theory explains intriguing clinical observations: universal occurrence near joints and proximal extension; the high percentage of trauma and frequent joint abnormality; the predominance of deep peroneal nerve dysfunction; the pedicle to the superior tibiofibular joint in 40% of reported cases; and the high recurrence rate.

A retrospective review of 12 consecutive patients was carried out, and then a prospective study of eight patients with peroneal intraneural ganglia. All 20 patients had predominantly deep peroneal nerve loss clinically, electrically, and radiographically, but several patients had more complete peroneal nerve dysfunction. In all cases, MRI review demonstrated a connection to the joint (not always noted on initial interpretation by the radiologist). At surgery, a communication to the superior tibiofibular joint was identified via an enlarged, cystic articular branch. The articular branch was oversewn, and the cyst capsule excised. At a minimum of 1-year follow-up, patients had significant pain relief, only mild improvement in neurologic function, but no recurrences.

The authors believe that intraneural ganglia are derived from an articular origin. Gelatinous fluid passes from the joint through a capsular defect, likely in patients with abnormal, traumatized joints. The articular branch serves as a conduit for the cyst fluid to egress, and enlarges under pressure. The fluid dissects proximally within the common peroneal nerve by the path of least resistance. Because the articular branch originates from the deep peroneal nerve, it is predictable that deep peroneal nerve dysfunction (particularly anterior tibialis weakness) occurs initially and preferentially. The cyst is formed eccentrically within the common peroneal nerve, and this can compress the neighboring superficial peroneal nerve fascicle(s). Successful surgery must address the articular branch. Since the presence and importance of this branch is not widely known, it is not typically explored at surgery. Recurrences may occur due to continued production and dissection of cyst fluid.

Brachial Plexus Neurolysis with Adhesiolytic Agents for Neurogenic Thoracic Outlet Syndrome Patients Diagnosed by MR Neurography: Outcome Study Results. Aaron Filler and Jodean Haynes.

This study applied MR image neurography to obtain specific diagnoses in cases of possible neurogenic thoracic outlet syndrome.

MR neurography with three-dimensional reconstructions was used as the primary diagnostic test in patients with a history and physical examination findings consistent with thoracic outlet syndrome. Image-guided scalene injection was used as a confirmatory test in equivocal cases. Surgery was directed at the lesion identified in the imaging test, and adhesiolytic agents (Seprafilm or Adcon-L) were used intraoperatively. Patients were then followed for outcome over periods ranging from 6 months to 5 years.

One hundred and thirty-five patients were referred for imaging and, of these, 71 were selected for surgery. The surgical approach was either supraclavicular (62 procedures) or transaxillary (41 procedures) for neurolysis, and none of the patients had first rib resections. Sixteen of the 71 patients had both supraclavicular and transaxillary surgery, and six had bilateral surgery. Eight of the surgical procedures were reoperations after recurrence commencing as early as 3 months and as late as 9 months after initial surgery (recurrence rate of 9%). Of the eight reoperative patients, only three failed to retain lasting benefits from the reoperation. Good or excellent outcomes were obtained in 77%, slight improvement in 9%, and no improvement in 14%. Complications included two patients with allergy to the adhesiolytic agent, one pneumothorax, and two wound infections.

These results demonstrated that MR neurography reliably predicts which patients will benefit from thoracic outlet surgery, and that recurrence rates are low when adhesiolytic agents are used.

Multi-Component T2 Relaxation as a Quantitative Measure of Demyelination. Stephanie Webb, Catherine A. Munro, Jeffrey Mason, Rajiv Midha, and Greg J. Stanisz.

Changes in tissue microstructure affect the tissue magnetic resonance (MR) properties, such as T2 relaxation. While average T2 values cannot be used to distinguish the different processes occurring within the injured nerve, multi-component T2 spectra may be able to do so. The purpose of this study was to evaluate the ability of multi-component T2 relaxation to quantitatively assess the degree of nerve damage, particularly demyelination, following nerve injury.

In male Lewis rats, the left sciatic nerve was cut, and the right crushed. Cutting the nerve causes irreversible degeneration in the portion of the nerve distal to the injury. The crushed nerve undergoes initial degeneration, followed by regeneration. At 1 to 6 weeks post-injury, samples of the nerve were collected for MR measurements and histomorphometric analysis. MR measurements were performed at 1.5T and at 20 degrees C on a 20-cm-bore superconducting magnet (Nalorac, Martinex, CA), controlled by an SMIS spectroscopy console (Surrey, England). The T2 data were fitted to a multi-component T2 model in which the relaxation of each T2 component has a Gaussian distribution on a logarithmic time scale.

The T2 spectrum of normal nerve showed three well-distinguished components. These components may be related to water found mainly in: myelin (short T2 component), intracellular space (intermediate T2 components), and extracellular space (long T2 component). The size of the short T2 component correlates very well with the morphometric myelin content for all time points. Initially (1 and 2 weeks post-injury), the size of the short T2 component decreased in both cut and crushed nerve. While degeneration was evident in both the cut and crushed histology sections, early regeneration was seen in the crushed sections by week 2. Two to 4 weeks after injury, the cut nerve short T2 component remained reduced, while that of the crushed nerve increased. By 3 weeks, no myelinated axons were seen in the cut nerve; remyelinated axons were seen in the crushed nerve. At 6 weeks, T2 spectra and histology of the crushed nerve resembled that of normal (uninjured) nerve. In contrast, the changes in the T2 spectra observed following injury in the cut nerve persisted, and complete demyelination and axonal loss were observed on histopathology sections.

Multi-component T2 spectra provide more information than average T2 values about the degree of damage to an injured nerve. Changes in the T2 spectra correlate well with observed changes in tissue microstructure. The size of the short component of the T2 spectra reflects the amount of myelin present.

Electrical Stimulation of Denervated EDL Muscles of Rats Maintains Mass, Maximum Force, Total Protein, and Myofibrillar Protein Contents. M.S. Calderon, J.A. Faulkner, and W.M. Kuzon, Jr.

Denervation of skeletal muscle leads to atrophy and loss of contractile function. After 1 month of denervation, the atrophic muscle has 38% of the mass and less than 15% of the specific force of an innervated muscle. A tetanic contraction protocol that produces 200 near-maximal tetanic contractions per day prevents the atrophy and loss of contractile force of denervated skeletal muscle. The prevention of denervation atrophy by this tetanic contraction protocol indicates a restoration of protein balance in the denervated skeletal muscle. The working hypothesis was that the tetanic contraction protocol prevents denervation atrophy by suppressing denervation-induced protein degradation, and thereby maintains total protein and myofibrillar protein content.

EDL muscles of adult Wistar rats were denervated by transection of the sciatic nerve. An implantable electrical stimulator was placed at the time of denervation, and a tetanic contraction protocol was initiated. A sham group was denervated and had a sham stimulator placed. Innervated and denervated animals provided positive and negative controls. After 14 days of stimulation, muscle mass, total protein, and myofibrillar protein contents were determined. Forces were measured by an in vitro technique.

After 2 weeks of denervation, the EDL muscle lost 40% of muscle mass, nearly 60% of total protein and myofibrillar protein contents, and 55% of maximum tetanic force, compared with control innervated muscles. Electrical stimulation maintained muscle mass, protein content, myofibrillar content, and maximum tetanic force of the denervated EDL muscle at values not different from the control innervated muscles.

These data support the hypothesis that electrical stimulation maintains the mass and force-generating capacity of denervated skeletal muscle by maintaining total protein and myofibrillar protein content. Experiments are in progress to determine if the protein degradation via the ubiquitin-proteasome pathway is responsible for the denervation atrophy, and if electrical stimulation can suppress the pathway.

Incomplete Nerve Repair Affects both Whole Muscle and Single Fiber Maximal Force. Jack van der Meulen, Dennis Claflin, Melanie Urbanchek, Paul Cederna, and William Kuzon, Jr.

A deficit in whole muscle maximal force generation (P_o) is observed after motor nerve injury and repair. These authors studied the contribution of individual muscle fibers to this force deficit after muscle reinnervation. They hypothesized that, compared with fibers from control muscle, single fibers from reinnervated muscles would generate a lower Po, especially if the muscle was reinnervated by a reduced number of motor axons.

In adult rats, the peroneal nerve was divided and repaired with either all (NR-non-reduced) or a reduced number (DR&equals;drastically reduced) of proximal motor axon stumps included in the repair. Sham-operated animals served as controls. After a recovery period of 4 months, P_o of whole muscles was measured in situ at optimal muscle length by supramaximal stimulation of the peroneal nerve. The EDL muscles were then removed, and bundles of fiber segments were dissected and stored in a membrane permeabilizing solution. Fibers were randomly selected from stored bundles and their diameter and maximum calcium-activated force (pCA∼4.0) were measured at optimal sarcomere length.

Compared with sham, no deficit was found in whole muscle or single fiber Po after complete nerve repair (NR group). However, compared with sham, P_o of whole muscle and single fibers showed a deficit in Po after DR nerve repair. Based on these data, the authors accepted their hypothesis that single fibers generate a lower P_o after incomplete nerve repair.

Lower force generation by single fibers can contribute to the lower force generation by whole muscles after nerve repair.

(Research supported by a Merit Review grant from the Veterans Administration and a grant from the NIH.)

No Donor Muscle Force Deficit after Muscle-Nerve-Muscle Neurotization. Melanie G. Urbanchek, Paul R. Kileny, Johns Michael, Norman D. Hogikyan, and William M. Kuzon, Jr.

Muscle-nerve-muscle (MNM) neurotization is reported to restore some functional capability to denervated muscle. There are no published results reporting the functional effects of MNM graft surgery on the donor muscle. The authors tested the hypotheses that 1) MNM neurotization restores force-producing capability to denervated skeletal muscle; and 2) MNM neurotization does not diminish donor muscle force capacity.

They tested muscle function following MNM neurotization from an innervated donor extensor digitorum longus (EDL) muscle to a denervated recipient peroneus digiti quinti (PDQ) muscle. Rats were assigned to one of four groups: normal control (CN), positive control (C+), negative control (C-), or MNM repair (MNM). Rat sciatic nerve was harvested from the right thigh for all groups but CN. On the left side, branches of the rat peroneal nerve leading to the medial compartment and innervating the PDQ were either: not operated (CN), sham exposed (C+), or dissected free, cut, reflected proximally, and stitched to the biceps (C- and MNM groups). Ends of the harvested sciatic nerve were implanted into pockets in the EDL and PDQ muscles, as an MNM graft in only the MNM repair group. Following 7 months for recovery, maximal isometric forces (P_o) were simultaneously measured in situ for the EDL and PDQ muscles, using sciatic nerve stimulation.

The mean Po for PDQ muscles in the MNM repair group was approximately 40% of the P_o for the C+ group. Po for the PDQ muscles from animals in both the C- and MNM groups was significantly reduced, compared with PDQ muscles from rats in the CN and C+ groups. There were no differences in the main effect for EDL muscle P_o.

These data supported the authors' hypotheses. Following MNM repair, 1) some force-producing capability was restored to the denervated muscle; and 2) no reduction in donor muscle Po was observed.

Use of End-to-Side Nerve Grafts to Reinnervate the Paralyzed Orbicularis Oculus Muscle. Michael J. Sundine, Edwin E. Quan, Vikas Dhawan, Oslan Saglam, Thomas G. Harralson, Janou W.J.M. Bardoel, Peter Quesada, Thanos Kakoulidis, Lynn Ogden, M. Douglas Grossman, and John H. Barker.

Facial paralysis is a serious neurologic disorder, particularly when it affects the eye. The purpose of this study was to determine whether end-to-side coaptation of nerve grafts to an intact facial nerve would provide reanimation to the paralyzed eyelid.

Seventeen adult dogs (25 kg) were allocated into four groups. Denervation of the left hemi-face was performed in all dogs. Group 1 (n&equals;2) animals were controls with no nerve grafts. Group 2 (n&equals;5) animals had end-to-side coaptation of the nerve graft to the intact palpebral branch and end-to-end coaptation to the denervated palpebral branch. Group 3 animals (n&equals;5) had end-to-end coaptation of the nerve graft to the intact palpebral branch and end-to-end coaptation of the denervated palpebral branch. Group 4 animals (n&equals;5) had end-to-side coaptation of the nerve graft to the intact and denervated palpebral branches. Video motion measurements were done every 3 months, to quantify the percentage of eyelid closure.

The animals were followed 9 months postoperatively to allow adequate time for reinnervation. At the end of the study, the nerve grafts were exposed surgically and mapped with an electrical probe, to confirm nerve conduction and eyeblink. The nerve grafts were harvested and histologically evaluated.

The results for each group, expressed as percent of eye closure: to percent of neuronal regeneration were: Group 1 - 0:0; Group 2 - 35±16.25 : 52±4.3; Group 3 - 74±13.57 : 68±4.9; Group 4 - 42±10.87 : 44±14.7. End-to-end coaptation provided the best function in cross-facial nerve grafting; however, there was a loss of function in the donor nerve. End-to-side coaptation produced a synchronous and spontaneous blink without the morbidity of sacrificing a donor nerve. However, the functional results were not as good as with end-to-end coaptation.

Intraneural Response to the Perineurial Window: A Functional and Histomorphometric Analysis. Jonathan M. Winograd, Irv M. Wiesman, J. Clinton Walker, Daniel A. Hunter, and Susan E. Mackinnon.

Both experimentally and clinically, there has been a renewed interest in end-to-side neurorrhaphy, often performed with the creation of a perineurial window. Since one of the main advantages of end-to-side neurorrhaphy is the lack of injury to the donor nerve, understanding the intraneural injury induced by the perineurial window is essential. Since Spencer's work in 1975, the perineurial window has been thought to cause focal demyelination with little Wallerian degeneration. More recently, there has been evidence that the size of the perineurial window may alter the severity of neural injury. In the reported study, the injury induced by 1- and 5-mm perineurial windows was examined in the rat tibial nerve.

One hundred forty-four Lewis rats were divided equally into three groups, with 1- and 5-mm perineurial windows, and a crush group as a positive control. A 1- or 5-mm perineurial window or a crush injury was created surgically on the left tibial nerve just distal to the sciatic trifurcation. Rats were assessed by walking-track analysis each week for 8 weeks postoperatively (n&equals;10 per group). The tibial function index (TFI) was calculated from digitized walking-track data. Histomorphometric analysis was performed up to 12 weeks postoperativeely (n&equals;8 per group per time point). Blood-nerve barriers were examined by fluorescent microscopy after parenteral Evans blue-labeled albumin injection at up to 9 weeks postoperatively (n&equals;12 per group per time point).

TFI means showed a severe decline postoperatively for all three groups: -20.1 for 1 mm; -20.7 for 5 mm; and -35.1 for crush. By 3 weeks postoperatively, the groups had leveled off at -7.0, -10.0, and -8.7 for 1 mm, 5 mm, and crush, respectively, representing complete recovery. Preliminary results of the histomorphometry for 1- and 2-week groups showed that 25% of the endoneurial contents adjacent to the perineurial window were injured. Demyelination and Wallerian degeneration were present. The total fiber number, fiber density, and fiber width within the zone of injury were markedly decreased. From 1 week to 2 weeks postoperatively, the fiber density increased, possibly from decreased edema or axon collateral sprouting. No differences were seen between the 1- and 5-mm groups. Blood-nerve barrier analysis at 2 weeks showed leakage for all three groups.

No obvious differences between 1- and 5-mm windows were noted at the site of injury. The findings are consistent with axonal collateral sprouting, which may be induced by the injury of the perineurial window. Analysis of later time points and the distal portions of the tibial nerves is still pending.

Functional Evaluation of the Epineurial Sleeve Technique in Nerve Grafting. Maria Siemionow, Przemyslaw Lubiatowski, Murat Unsal, Blazenka Skugor, and Dileep Nair.

Nerve grafting is a well-accepted method used for reconstruction of nerve defects when end-to-end repair is not feasible. Previous studies have shown superior recovery when a sleeve technique was applied following nerve transection. This study was conducted to compare the sleeve technique for nerve grafting with conventional repair.

Twenty-four Lewis rats were divided into three groups: Group 1-nerve graft (NG), with conventional epineurial repair used. With the sleeve technique, a segment of epineurium was dissected as a sleeve into which the opposite nerve stump was inserted; the sleeve was anchored with two epineurial sutures. Group 2-sleeve graft (SG) group, in which an epineurial sleeve was created from the graft. Group 3-sleeve nerve (SN) group, in which a sleeve was created from the proximal and distal sciatic nerve stumps. Nerve regeneration in each rat was evaluated at 3, 6, and 12 weeks by pin-prick, toe-spread test, and walking-track analysis (sciatic function index, SFI). At 12 weeks, somatosensory evoked potentials (SEP) and gastrocnemius muscle weight were measured, and nerve samples were stained with toluidine blue for histologic evaluation. Using a computerized digital system (Image^oVPro Plus, Media Cybernetics), the total numbers of myelinated axons, axon diameter, and myelin sheath thickness were calculated.

At 3 weeks, no significant differences were found between groups. At 6 weeks, a significant difference in SFI was seen in the SN group, compared to the SG and NG groups (p<0.05). At 12 weeks, the SN and SG groups had better motor recovery than the control NG group. SEP demonstrated better sensory recovery for the SN group, in which P1 and N2 latencies were shorter, compared to the SG and NG groups (p<0.05). Muscle weight ratio increased significantly in the SN group, compared to the NG group (p<0.05), and did not differ from the SG group. However, the axon number distal to the nerve graft showed greater diameters in the SN and SG groups, compared to the NG group (p<0.05). The myelin thickness was significantly greater in the SN group, compared to the NG group (p<0.05).

Nerve defects repaired with the epineurial sleeve technique, derived either from the nerve graft epineurium or the nerve stump, showed an improved functional nerve recovery, compared to the conventional nerve-graft repair. This was confirmed by shortened P1 and N2 latencies in somatosensory evoked potentials, walking-track analysis, and histomorphometric analysis.

Forearm Median and Ulnar Nerve Reconstruction with the Neurotube. Patrick Swier, Jaesuk Kim, and A. Lee Dellon.

Peripheral-nerve reconstruction with the bioabsorbable PGA conduit, the neurotube, has been reported recently to give excellent results when used for the digital sensory nerves. There have no reports of major nerve reconstructions with this device. This report described the long-term results in two adult patients (ages 43 and 61 years) of median-nerve reconstruction in the distal forearm.

One patient was 3 and one was 1 year post median-nerve injury. One case was work-related and had undergone an attempt at primary repair; the other was an iatrogenic injury. In each of these patients, the neuroma-in-continuity was resected, and a 2.8-cm defect reconstructed with four separate neurotubes, each placed as an interfascicular graft. The report also described long-term results in a 7-year-old boy of reconstruction of the motor component of the ulnar nerve in the distal forearm after a dogbite 3 months previously. In this case, one neurotube was used to reconstruct a 2.9-cm defect.

At 14 months after reconstruction, the 43-year-old patient had excellent localization in all median-nerve-innervated fingers, and he has M4 abduction of the thumb. Moving two-point discrimination has returned to the thumb, but not yet to the index or middle fingers. He is in late-phase reeducation. There is a slight tenderness at the reconstruction site. The patient is in vocational rehabilitation to become a graphic designer. At 12 months after reconstruction, the 61-year-old patient has moving two-point discrimination recovered in the index finger, an excellent localization to all median-innervated fingers. There is M4 thumb abduction. The patient is in late-phase reeducation. There is no tenderness at the reconstruction site, and he has been back to work for 4 months at his regular job. At 7 months after ulnar motor reconstruction, the 7-year-old patient has M2 little finger abduction, and can initiate crossing of the index and middle fingers. There is no clawing of the hand at rest, and Froment's sign remains. The first dorsal interosseous is still M0.

It was concluded that major nerve reconstruction in the forearm is possible, safe, and successful, using the 2-mm in diameter neurotube as an interposition interfascicular conduit.

BDNF-Enriched Gel-Filled Collagen Tube as a Substitute for Autologous Nerve Grafts. David J. Terris, Kenneth Toft, Melinda Moir, and Joann Lum.

Autologous nerve interposition grafts are frequently harvested by head and neck surgeons. The sacrifice of the donor nerves guarantees some degree of morbidity, including sensory loss, additional incision sites with associated potential complications, and prolonged operating time. An alternative to autologous nerve grafting is therefore desirable. The authors' objective was to determine if a collagen tube filled with either a plain collagen gel or brain-derived neurotrophic factor-enriched collagen gel (BDNF-gel) could be utilized, to achieve functional and histologic outcomes equivalent to an autologous nerve graft, in bridging a 15-mm nerve gap in the rabbit facial nerve.

Thirty rabbit facial nerves were resected (15-mm segments) to create nerve gaps. The gaps were bridged using one of three methods assigned randomly: reversed facial nerve (control); collagen gel-filled collagen tube (CT-gel); or BDNF-gel-filled collagen tube (CT-BDNF). The animals were evaluated after 6 weeks in a blinded fashion, for functional nerve recovery, axon count, and axonal diameter.

There were no significant differences between the autologous nerve-graft group (n&equals;10), the CT-gel group (n&equals;10), or the CT-BDNF group (n&equals;10) with respect to functional nerve recovery (p&equals;0.94). The mean axon count and mean axonal diameter were highest in the CT-BDNF group, but these differences failed to reach statistical significance.

Collagen tubes filled with BDNF-enriched collagen gel appeared to be at least as effective as autologous nerve grafts for bridging short facial nerve gaps. Larger experimental studies are warranted, to determine if clinical trials are justified.

Biodegradable Poly (DLLA-episilon-CL) Nerve Guides Filled with Modified Denatured Muscle Tissue or Phosphate Buffered Saline. M.F. Meek, J.F.A. van der Werff, and A. Gramsbergen.

Although autologous nerve grafting is the most widely used technique for bridging nerve gaps, this technique has several disadvantages, such as the loss of donor-nerve function and the risk of neuroma formation. Therefore, many alternative techniques have been developed, including the use of biodegradable nerve guides. A nerve guide eliminates the problem of a second operative site necessary to harvest a donor nerve (in the case of a nerve autograft). The nerve guide directs the outgrowing nerve fibers toward the distal nerve stump, while preventing neuroma formation and ingrowth of fibrous tissue into the nerve gap. After serving these functions, the nerve guide degrades. Furthermore, a nerve guide permits the outgrowing nerve fibers from the proximal nerve stump to respond to neurotropic and trophic substances, which can accumulate in the nerve guide. The aim of this reported study was to compare the return of nerve function after bridging 15-mm gaps in the rat sciatic nerve with either a biodegradable p(DLLA0-epsilon-CL; Neurolac, Polyganics Bv, Groningen, The Netherlands) nerve guide filled with modified denatured muscle tissue (MDMT) or phosphate-buffered saline (PBS).

The left sciatic nerve of 24 male Wistar rats was exposed through a gluteal muscle-splitting incision. In Group 1, a 15-mm gap was created, and continuity reestablished by interposing a 20-mm nerve guide filled with MDMT. In Group 2, a 15-mm gap was created and continuity reestablished by interposing a 20-mm nerve guide filled with PBS. Both the proximal and distal ends of the sciatic nerve were telescoped into the ends of the nerve guide and fixed with a single 9-0 nylon epineurial suture. Group 3 consisted of six non-operated control rats. To evaluate both motor and sensory nerve recovery, walking-track analysis and electrostimulation tests were carried out after implantation periods ranging from 2 to 15 weeks postoperatively. In addition, the animals' hindfeet were examined for signs of automutilation. All data for Groups 1 and 2 were submitted to linear regression analysis.

Recovery of sciatic nerve function was more rapid in Group 1 (nerve guide+MDMT), compared to Group 2 (nerve guide+PBS). Automutilation was more pronounced in Group 2. Linear regression analysis demonstrated that motor and sensory nerve recovery in Groups 1 and 2 during the study was highly significant (p<0.001). There was no statistically significant difference in the return of sensory nerve function between Groups 1 and 2.

The authors concluded that their experiments demonstrated that in the case of a short nerve gap, reconstruction was best carried out using a nerve guide filled with MDMT.

Inflammatory Reaction Following the Implantation of Polylactide and Collagen into a 2-cm Gap of the Rat Sciatic Nerve. Franz Lassner, Hans Mueller, Gary Brook, Ingo Heschel, Michael Becker, Michael Schroeder, and Norbert Pallua.

Methods of tissue engineering may provide an answer to the problem of the limited availability of nerve-grafting material: a resorbable matrix seeded with Schwann cells may serve as an axonal conduit to bridge the nerve gap. The resorbable material used for this purpose should elicit only minimal inflammatory changes during the resorption process. The authors evaluated two such materials, collagen and polylactide, for tissue reactions after implantation into a gap of rat sciatic nerve.

A 2-cm gap of the rat sciatic nerve was created by subepineurial resection of the nerve. This allowed for the introduction of different materials and the subsequent closure of the epineurium over the matrix. Collagen scaffolds with longitudinal tubular architecture and polylactide filaments were used as templates for axonal regeneration. Histologic evaluation was performed at intervals of 1, 2, 3, and 6 weeks. The results were compared to those of conventional nerve grafts.

Within 1 week, capillarization of the matrices occurred. After 2 and 3 weeks, inflammatory changes with lymphocyte infiltration, but no signs of necrosis or abscess formation, were present. The inflammatory reaction extended until the sixth week. Only small areas of myelinization were evident after 6 weeks.

The authors observed a foreign body reaction, inflammatory changes, and poor regeneration with polylactide filaments and collagen sponges. The histologic evaluation showed inferior regeneration, compared to the conventional nerve grafts. Further experiments are in progress, using matrices seeded with Schwann cells, and the results were presented.

Microsurgical Management of Old Cauda Equina Syndrome Following Injury or Previous Surgery. Shimon Rochkind, Malvina Alon, and Khalil Salame.

The authors' purpose was to demonstrate that functional improvement is feasible after microsurgical management, in patients suffering from old cauda equina syndrome.

Eleven patients identified as having old cauda equina syndrome (from 1.5 to 17 years after onset of lesion), with severe neurologic deficit and intractable pain, were presented in this prospective study. Responsible mechanisms included gun shot wound, road accident, fall injury, and previous lumbar surgical procedures. MRI and CT changes showed conglomerations of adhering roots residing centrally within the thecal sac or roots adhering peripherally to the meninges, giving rise to an ``empty sac'' appearance or an entrapment cyst mass replacing the subarachnoid space. CT scan demonstrated arachnoiditis ossificans in three patients. The patients were treated microsurgically by intradural cauda equina nerve-root release and neurolysis. The extensive scars with ossific lesions and various degrees of adhesion were freed and removed from the cauda equina nerve roots. Intraoperative spinal-cord monitoring was performed, to control the response from the conus medullaris. Compound muscle action potentials were used during surgery, to identify functional cauda equina nerve roots.

The appearance or increasing amplitude of evoked responses were recorded in four of the 11 patients during cauda equina nerve-root neurolysis. Six of the 11 patients experienced significant motor improvement, as well as a significant decrease or actual disappearance of pain from 6 months to 1 year after surgery. Four of 9 patients with previous sphincter disturbances experienced significant improvement of sphincter function.

This study demonstrated that the use of microsurgical management results in pain relief and functional improvement of patients suffering from old cauda equina syndrome.

Neurolysis of the Common Peroneal Nerve: Cadaver vs. Patient Fascial Variations. A. Lee Dellon, Johannes Ebmer, and Patrick Swier.

Neurolysis of the common peroneal nerve at the fibular neck will increase in frequency, as peripheral-nerve decompression is more extensively done to restore sensory and motor function in patients with symptomatic diabetic neuropathy. The anatomic variations related to compression in this area have not been previously described.

Dissection was carried out on 29 pairs of embalmed cadavers, to identify the incidence of fascial bands deep to the peroneus longus muscle, and connections of the soleus origin to the peroneus muscle. These fascial variations had been noted clinically, retrospectively, to be present in patients with symptomatic compression at this level that was not related to trauma. The incidence of these anatomic variations was noted prospectively in a series of 65 patients having neurolysis of the common peroneal nerve in this area over the past 2 years.

The most significant finding was a fascial band that was a mean of 9.1 mm (range: 5 to 15 mm) on the deep surface of the peroneus longus muscle in 30% of the cadavers. In contrast, this band was present, with the same dimensions, in 78.5% of the patients requiring peroneal-nerve decompression. There was no significant difference in the presence of a fascial band deep to the common peroneal nerve (30%) or a conjoined origin of the soleus musccle and the peroneus longus muscle (8%).

The study provided a guideline for fascial structures that must be identified, in order to have a successful outcome from decompression of the common peroneal nerve at the fibular neck.

Bridging of Long Nerve Defects Using Autogenous Nerve Grafting, Interfascicular Suture, and Microsurgery. Harilaos T. Sakellarides.

The author had the opportunity to treat peripheral-nerve injuries with large defects in the arm, but the results were disappointing. In analyzing the reasons for the poor results following nerve suture under tension, with extensive mobilization, the blood supply to the nerve segments was impaired and scarring resulted. With the sutures under tension, the arm had to be immobilized for much longer periods, resulting in further atrophy of the muscles. Since conventional epineurial sutures between the divided nerve ends and previously applied nerve grafts were not successful, a new method has been attempted over the last 20 years. Dr. Hanno Millesi, who has tried this new type of nerve grafting, concluded, after clinical research: 1) that there was a detrimental role of tension at the suture line; and 2) that there was a deleterious effect of postoperative stretching on successful functional recovery. Millesi found that regenerating axons passed more easily through nerve grafts of 5 cm, with two tension-free anastomoses, than through a single suture under tension, and that the epineurium was the primary source of connective tissue proliferation. The technique used is resection of the epineurium, resection of fascicles and excision of neuroma, interfascicular suture, resection of the fascicles at different levels, grafts slightly longer than defects, two-to-three necessary sutures of 10-0 nylon, and four or five grafts required for the median and ulnar nerves.

Of 130 cases, 70 involved the median nerve, 40 the ulnar nerve, and 20 the radial nerve. There were 85 male and 45 female patients, with ages ranging from 20 to 60 years. The time from injury to grafting was from 6 months to 5 years. Follow-up ranged from 2 to 15 years. Evaluation of nerve repairs was performed according to the British method. Motor recovery for median-nerve low lesions (recovery of thenar muscles from M3-4 level) was excellent in 40% of cases, M2-M3 was good in 40%, M1-M2 was fair in 20%. Ulnar-nerve motor recovery for intrinsics to level M3-M4 was excellent in 38% of cases; for level M2-M3, it was good in 40%, and fair in M1-M2 in the remaining 22%. Motor recovery for the radial nerve, extensors of the wrist, fingers, and thumb to level M3-M4, was excellent in 42%, good in 38%, and fair in 20% of cases.

Pitfalls and Omissions in Brachial Plexus Surgery. Hanno Millesi, Tarkan Basar, Klaus J. Korak, and Dagmar Millesi.

Over the past 35 years, brachial plexus surgery has developed from a salvage procedure to a well-established surgical approach. In the early days, the surgeon was happy with whatever function returned, and accepted a high rate of failure. Currently, different techniques are available, at least to attempt a useful recovery at different levels. In cases of avulsion of all five roots, it is evident that there is a certain failure rate which must be accepted. On the other hand, all available sources for improving function should be utilized, and it is not clear why one or another possibility is neglected, without good reason (omission). Ever so often, patients are seen with failure to achieve elbow flexion, not because of lack of neurotization, but because of co-contractions of the biceps and triceps. If the original surgeon had recognized this situation, the patient could have been advised to have, for example, a triceps transfer, and useful elbow flexion would have been achieved (pitfall). Analysis of similar cases produced thoughts about standardization of the surgical approach to brachial plexus lesions, which was presented.

Radial Nerve Palsy: Sonographic Findings. Gerd S. Bodner.

The author attempted to demonstrate the feasibility of ultrasonography (US) to evaluate radial nerve palsy. In a prospective study, 25 consecutive patients with sensomotor radial deficiency (distal humerus fracture, metastatic disease, muscular overuse) were investigated by x-ray, US (broad band linear array, 5-12 MHz), electroneurography, and electromyography. Surgical repair of fracture and nerve inspection were carried out in 14 cases. The remaining 11 patients were successfully treated conservatively. In 12 cases, US findings of a severely damaged radial nerve were confirmed by surgical nerve inspection. Radial nerve compression by metastatic lymph nodes was seen in two cases. In two other cases, a swollen radial nerve (ramus profundus) was found in the area of the supinator muscle. The conclusion was that US may be useful for accurate evaluation of the radial nerve associated with nerve palsy.

In Vivo Quantitative T2 MRI of Nerve Injury. Stephanie Webb, Catherine A. Munro, Rajiv Midha, and Greg J. Stanisz.

Previous work has shown that the magnetic resonance (MR) properties of damaged nerve differ from those of normal nerve, and thus can be used to quantitatively assess the extent of nerve damage. Specifically, in vitro measurement of T2 relaxation can be used to obtain a measure of the amount of myelin present in a tissue sample. However, for this technique to be of clinical significance, it must be applied in vivo. The purpose of this study was to make in vivo measurements of the T2 relaxation properties of nerve, and to evaluate whether in vivo techniques can be used, as can in vitro techniques, to quantify nerve damage.

The sciatic nerve of male Lewis rats was imaged on a 1.5T GE Signa whole body scanner. Both normal (uninjured) and injured rats were scanned. Injured rats were imaged at various time points following a crush injury to the right sciatic nerve and a cut injury to the left sciatic nerve. Scan sequence parameters were optimized for T2-weighted imaging of nerve tissue. The reconstructed images were analyzed on a pixel-by-pixel basis. T2 data were fitted to a multi-component T2 model in which the relaxation of each T2 component has a Gaussian distribution on a logarithmic time scale. The data were combined to create an overall T2 spectrum for the region of interest (ROI).

On T2-weighted images, the nerve appears bright and is easily distinguished from the surrounding muscle and fat tissue. The T2 spectrum for the ROI containing the sciatic nerve shows three well-distinguished components at short, intermediate, and long T2 values. The components of the in vivo T2 spectra are similar in size and position to those of the in vitro spectra. Thus, the short T2 component may be identified with myelin-associated water, with the size of the component indicating the amount of myelin present. The size of the myelin component in the injured group was statistically significantly different from the size of the myelin component in the control group, suggesting that the in vivo analysis successfully measured demylination.

It is possible to obtain multi-component T2 relaxation data about the nerve in vivo. The in vivo T2 spectrum compares favorably with data obtained in vitro. T2-weighted MR imaging shows great promise as a technique for the quantitative evaluation of nerve damage in vivo.

Utility of Somatosensory Evoked Potentials for the Early Diagnosis of Peripheral-Nerve Injuries in Comatose Patients with Head Injuries. Amanda B. Taylor, David A. Houlden, and Rajiv Midha.

A valid neurologic examination is often not possible when attempting to detect the presence of peripheral-nerve injury in trauma patients. These patients may be unable to cooperate with the examination, due to coma resulting from head injury, pharmacologic sedation, or pharmacologic paralysis. Unfortunately, EMG studies may not be useful in detecting peripheral-nerve injury in the early stage after injury, because muscle denervation may not appear until weeks later. For this reason, these authors used somatosensory evoked potentials (SSEPs) to detect peripheral-nerve injury affecting the upper extremity in comatose patients with head injuries.

Median-nerve SSEPs were performed within 24 hr after head injury. Left then right median nerves were stimulated at a rate of 2.13 Hz. Recordings were obtained from the ipsilateral Erbs point (EP)-Fpz, C2 spinous process-Fpz, C3'-C4', contralateral somatosensory (SS) cortex-Fpz, and the contralateral SS cortex-contralateral EP (International 10-20 system). The total sweep time was 75 msec. The amplifier gain was 20 uV/division and the recording bandpass was 10-1000 Hz. A peripheral-nerve injury was suspected when the amplitude of the peripheral-nerve response recorded from the EP was less than 0.5 uV or 2.5 times smaller than that obtained from the opposite side.

SSEPs were obtained from 52 consecutive comatose patients with head injuries. SSEPs detected peripheral-nerve injury affecting the upper limb in two of those patients. One was suspected of peripheral-nerve injury on clinical grounds. The other was not suspected of having peripheral-nerve injury on clinical grounds, but was discovered by SSEP. Both patients were able to cooperate with a neurologic examination 6 weeks after injury, and the SSEP results were clinically verified in both. One other patient was suspected of having a peripheral-nerve injury based on clinical examination, but this was not confirmed by SSEP. The SSEP result was verified by a normal clinical examination at follow-up. Another patient had a preexisting polyneuropathy that was detected by a significantly delayed peripheral-nerve response from EP bilaterally.

Despite the limitations of SSEP testing (i.e., only one nerve is stimulated), SSEPs were useful in confirming, ruling out, and discovering peripheral-nerve injuries affecting the upper extremity in comatose patients during the early stage after head injury. Accordingly, SSEP results may be important when considering the differential diagnosis of neurologic injury in patients who cannot cooperate with a neurologic examination.

Sciatica of Non-Disk Origin: Diagnosis by MR Neurography and Interventional MRI with Outcome Study of Resulting Treatment. Aaron Filler, Jodean Haynes, J. Pablo Villablanca, Joshua Prager, Duncan McBride, Ulrich Batzdorf, J. Patrick Johnson, and Keyvan Farahani.

This study applied MR neurography and interventional MRI to diagnose 223 consecutive patients with sciatica, whose lumbar MRI reports did not reveal a herniated disk. All patients had MR neurography, a physical examination, and injections, where appropriate. After conservative care, percutaneous therapeutic injections, or surgery, patients were followed from 6 months to 6 years, with outcome assessment questionnaires.

Patients had piriformis syndrome (143), ischial tunnel syndrome (14), distal foraminal root entrapment (14), discogenic pain with referred leg pain (9), lumbosacral plexus entrapment (6), pudendal nerve entrapment with referred pain (6), distal sciatic entrapment (5), unappreciated disk herniation (5), sciatic tumor (3), tumor in lumbosacral plexus (3), lumbar stenosis presenting with radiculopathy (2), sacral fracture (1), or no diagnosis (8). If MR neurography did not reveal another diagnosis, piriformis syndrome was confirmed by open MR guided Marcaine injection. Injected patients (143) responded as follows: relief at least 8 months (35), relief 2-4 months, with lasting relief after second injection (46), relief 2-4 months not responding to further injection (2), relief for 1-14 days with full recurrence (42). The last two groups had piriformis surgery (62 operative procedures) done through a 3-cm incision by transgluteal approach, 55% outpatient, 40% under local or epidural anesthesia, resulting in excellent outcomes (48%), good outcomes (22%), limited benefit (17%), and no benefit (13%).

Application of MR neurography and open MR injection procedures was highly effective in diagnosing and treating patients who failed routine workup and treatment for sciatica.

Endoscopic Decompression of Intermetarsal Nerve Entrapment. Stephen Barrett.

``Morton's neuroma'' or metatarsalgia is treated by both orthopedic and podiatric surgeons as a neuroma, by resecting the involved interdigital nerve, despite awareness that this painful common foot disorder is due to chronic nerve compression. An open surgical approach to decompression of the interdigital nerve was first described by Gauthier in 1983, and confirmed by Dellon in 1992, to give excellent relief of pain without resecting the nerve. The purpose of the presented report was to present the author's experience with an endoscopic technique to neurolyse the interdigital nerve.

This approach was used in 69 patients with interdigital ``neuromas'' in 90 webspaces. The study was begun in 1993 and completed in 1999. The interdigital neuromas were in 38 second and 52 third webspaces. The average patient age was 50.6 years. Preoperative symptoms lasted a mean of 23.3 months and did not respond to conservative measures, including steroid injections into the webspace, and orthotics. Good-to-excellent results occurred in 87.8% of the patients. There were five interspaces that did not improve, and these patients obtained relief with an open approach in which a neurectomy was utilized.

The author concluded that results for the surgical management of ``Morton's neuroma'' can be obtained with the same success rate using an endoscopic approach, as that reported for an open approach to decompress the interdigital plantar nerve.

Role of the Vastus Lateralis Muscle in the Treatment of Lumbosacral Plexus Lesions. Hanno Millesi, Tarkan Basar, Klaus J. Korak, and Dagmar Millesi.

Lesions of the lumbosacral plexus occur much less frequently than lesions of the brachial plexus; a complete lesion is rather rare. In the majority of cases, there are isolated lesions of either the lumbar plexus or the sacral plexus. The most frequent causes of lesions of the sacral plexus are fractures of the pelvis. The most difficult problem is represented by a lesion of the lumbosacral trunk which contains fibers to the superior gluteus nerve and to parts of the sciatic nerve. For the patient, paralysis of the gluteus medius muscle has grave consequences. These patients are not able to stabilize the pelvis and have difficulties walking without support. Therefore, in reconstructive procedures, attention is focused on the superior gluteus nerve, in order to neurotize the gluteus medius muscle. If these attempts fail, the function of the gluteus medius can be replaced by transfer of the vastus lateralis muscle as an island flap. This is possible only if the lumbar plexus has remained intact. The repair of the lumbosacral trunk, including the gluteus superior muscle, and the transfer of the vastus lateralis muscle, in cases of failed repair, were described in detail.

Surgical Techniques to Improve Shoulder Joint Function after Paralysis and Partial Recovery. Hanno Millesi, Tarkan Basar, Klaus J. Korak, and Dagmar Millesi.

There are several useful techniques available to improve insufficient external rotation after paralysis and partial recovery. However, there are very few techniques to achieve significant improvement of function in the glenohumeral joint, if abduction is insufficient. Trapezius transfer is a useful tool which, at least in the authors' hands, provides static improvement, but not much better improvement of the scapula humoral angle. Among the techniques that have been applied to achieve this goal are: 1) Medianization of the deltoid muscle. In a certain number of cases, the acromial portion of the deltoid muscle is completely atrophic, but the clavicular and spinal portion has preserved a certain function. In this case, the activity of the two parts is counterproductive, because they are adducting, at least at the beginning of abduction. Transfer of the origin of the clavicular and spinal portions of the deltoid muscle toward the mid line, and reinsertion at the acromion have produced improvement. 2) Regarding the supraspinatus muscle, in certain cases, the supraspinatus tendon is adherent and, even if the muscle has preserved some function, this is not transmitted to the humerus. In this case, tenolysis is helpful. In other cases, when the supraspinatus muscle is paralyzed, the descendent and horizontal portions of the trapezius are disinserted and connected to the supraspinatus tendon in various ways. 3) Regarding the pectoralis major muscle, transfer may take place in various ways. The clavicular portion can be transferred to the lateral aspect of the clavicle, in order to replace the clavicular portion of the deltoid after this muscle had been transferred to the acromion. A well-defined section of the sternocosal portion based on its blood supply can be transferred to replace the acromial portion of the deltoid muscle. 4) Regarding passive suspension, one of the problems is subluxation of the head of the humerus due to muscular suspension. Several of the muscle transfers described aim at an improvement of muscular suspension, in addition to achieving abductive function. However, a large part of the muscle power is lost statically, and does not contribute to the dynamics of the shoulder joint. The authors attempt to suspend the head of the humerus to the acromion by transplanting fascia lata or tendons, in order to achieve passive suspension, and to have all available muscle power free for active movement. Cases subject to the described guidelines were analyzed and presented in detail.

How to Evaluate Postoperative Results after Shoulder Paralysis. Hanno Millesi, Tarkan Basar, Klaus J. Korak, and Dagmar Millesi.

To evaluate shoulder joint function after surgery for obstetrical brachial plexus lesion, the scheme of Mallet is very practical. In other studies, the angle of so-called ``global'' abduction has been demonstrated. Both systems are useful for clinical purposes. However, for scientific studies of the outcome of different surgical techniques, a more detailed and reproducible quantitative system is to be desired. The authors discussed the following landmark measurements.

1) External rotation in adduction. The hand is kept close to the body, the elbow flexed, and the forearm in the sagittal plane. The angle between the sagittal plane and maximal external rotation is measured in plus degrees. The angle of maximal internal rotation is measured in minus degrees. 2) External rotation in abduction. The upper arm is kept horizontal, the elbow flexed, and the forearm in a horizontal plane. By external rotation, the forearm is lifted normally up to 90 degrees. The degree of actual external rotation in abduction is measured in positive degrees. Movement from the horizontal plane down describes internal rotation in abduction, and is measured in minus degrees. 3) The angle between the axes of the humerus and the contour line of the trunk is measured with a pendant arm. The same angle is measured with maximal lateral abduction (THORHUM). 4) The angle between the lateral margin of the scapula and the axes of the humerus is measured with a pendant arm. The same angle is measured with maximal lateral abduction (SCAHUM). This angle allows for the evaluation of movement in the glenohumeral joint. 5) The lateral margin of the scapula is marked with a pendant arm. The new position of the lateral margin of the scapula is marked with maximal lateral abduction. The angle between the two marks is measured. This angle (SCASCA) allows for the evaluation of movements of the scapula.

The measurement of these angles (3,4,5) allows for evaluation of the contribution of both the movement of the scapula and that of the glenohumoral joint in ``global'' abduction. The value of these measurements was demonstrated in the presentation of clinical cases.

Pisiform Arthrokinematics and Carpal Tunnel Syndrome. G.R. Williams and G.M. Cuka.

Normal palmar force carpal translation is 5-7 pounds during co-contraction of the fingers and thumb in extension. In carpal tunnel patients, force transducer measurements indicate deformation forces between 20-30 pounds of palmar force translation, resulting in shear forces to carpal ligaments. This translational force applies an insidious, low-load force that alters carpal arthrokinematics through remodeling of ligamentous structures. The altered carpal environment changes the moment arm of the pisiform. Concomitantly, dysfunctional pisotriquetral arthrokinematics by transmission forces of the flexor carpi ulnaris via the pisiform continue and perpetuate the progression of deformation of the carpal tunnel, resulting in increased intracarpal pressure.

Pisotriquetral behavior is critical for coordinating intrinsic and extrinsic power functions of the hand. The purpose of this reported study was to define normal and abnormal pisiform arthrokinematics relating to carpal tunnel syndrome in three distinct variations: immobility of the pisiform proximal (type I), distal (type II), and proximal/distal (type III).

Co-contraction, as defined in this study, is variable positioning of the fingers and thumb until some form of resistance occurs against these motions. When resistance is absent, end-range exertion in extension of the fingers and thumb results in co-contraction that requires inter-muscle communication. Co-contraction is maintained throughout the arc of motion in flexion or extension until resistance is encountered. Excessive contraction or co-contraction alters pisotriquetral function.

Type I (contraction): pisotriquetral dysfunction occurs with remodeled pisohamate and pisomatacarpal ligaments, producing lack of distal pisiform gliding during grasp. The pisiform does not track distal during composite fisting. Type II (co-contraction): pisotriquetral dysfunction is charracterized by a lack of proximal gliding of the pisiform during co-contraction, as the flexor carpi uulnaris moves the pisiform proximally. Attentuation of the pisometacarpal and pisohamate ligaments promotes a proximal position of the pisiform. The FCU shortens and further perpetuates deforming forces. Type III (contraction/co-contraction): presentation is characterized by an almost unpalpable pisoform. These individuals have greater dorsal translation force, decreased ROM, periarticular fibrosis, and more taut FCU insertion.

Improved arthrokinematics of the pisiform are consistent with complete relief of carpal tunnel symptoms in 60% of patients treated. A new method of testing carpal translation was defined and treatment described.

Sensitivity and Specificity of Clinical Testing in Carpal Tunnel Syndrome. Irvin M. Wiesman, Christine B. Novak, Susan E. Mackinnon, and Jonathan M. Winograd.

The diagnosis of carpal tunnel syndrome (CTS) is based on patient symptoms, physical examination, and confirmed with nerve conduction studies. The purpose of this study was to evaluate the sensitivity and specificity of clinical testing in the diagnosis of carpal tunnel syndrome.

The patient group included patients with complaints of sensory alteration in the median nerve distribution and abnormal nerve conduction studies across the carpal tunnel to confirm the diagnosis of CTS. Patients with previous surgery on the median nerve, including carpal tunnel release, median nerve repair or graft, were excluded. The control group included subjects with no history of paraesthesia or numbness in the median nerve distribution of the hand, and subjects with no previous distal radius fractures. The clinical testing was done by a single examiner and included a Tinel's sign, wrist flexion with fingers extended, wrist flexion with fingers flexed, wrist extension, combined wrist extension/median nerve pressure, and combined wrist flexion/median nerve pressure. Each test was held for a total of 1 min with a 1-min rest between each maneuver, and a positive response was recorded with reported sensory alteration in the median nerve distribution. Data were analyzed with 2 by 2 tables to examine the sensitivity, specificity, and predictive values for each clinical test. These data were presented.

Anatomic Comparison of Donor-Nerve Sites Utilized for Reconstruction of Traumatic Hand Injuries: A Cadaver Study. James P. Higgins, Greg S. Orlando, and Joseph M. Serletti.

Reconstruction of the traumatized upper extremity often requires the repair of segmental defects in sensory nerves distal to the wrist. Several donor nerve-graft sites are commonly utilized. Criteria for selection of donor sites for nerve harvest have never been standardized. The purpose of this study was to determine the cross-sectional area and number of nerve fascicles in different donor nerves commonly harvested for traumatic hand reconstructions. Similar data were also collected for digital nerve segments at standardized intervals from wrist to fingertips. By comparison of the donor and recipient data, guidelines were provided for selection of appropriate nerve-graft harvest sites, when treating segmental nerve defects at particular levels.

Nerve segments were harvested from 10 fresh cadavers (20 upper extremities). Five sites of nerve-graft harvest were included: lateral (LABCN) and medial (MABCN) antebrachial cutaneous nerves, posterior (PIN) and anterior (AIN) interosseous nerves, and sural nerves. Four sites of typical segmental nerve defects were harvested in a standardized zone protocol: common digital nerve (zone 4), proper digital nerve (zone 3), digital nerve distal to main dorsal branch at MCP (zone 2), and digital nerve distal to trifurcation at fingertip (zone 1). Fresh specimens were examined under light microscopy, recording the number of fascicles and cross-sectional area (determined by microscopic images transferred to computer software).

Appropriate defect/donor nerve matches were determined according to the two criteria of cross-sectional area and number of fascicles. The sural nerve is the most anatomically similar nerve graft for defects in zone 4 by both criteria. The LABCN is most appropriate for zones 2 and 3 injuries by both criteria. Fingertip grafts for zone 1 injuries displayed cross-sectional area similarity to PIN, AIN, and MABCN. With regard to number of fascicles, zone 1 digital nerves are most similar to LABCN donors.

Data were graphically represented displaying t-test results and mean/standard deviation values. A discussion of ease of harvest and donor deficits from the various sites was included, as well as suggestions for future study.

Prevention of Oxidative Stress in Diabetic Skeletal Muscle by Insulin-Like Growth Factor (IGF-1). Eric Rabinovsky and Saleh Shenaq.

Oxidative stress is a major factor in causing diabetic complications; however, the role of oxidative stress in causing motor-nerve neuropathies and myopathies is not well understood. In this study, the authors reported that skeletal muscle undergoes oxidative stress in response to diabetes, but that this can be prevented in transgenic mice overexpressing IGF-1 in skeletal muscle. Wild-type (non-IGF-1) diabetic mice exhibited increased DNA damage in skeletal muscle, as shown by immunoreactivity to 8-hydroxy-2-deoxyguanosine (oh8dG). DNA damage may be due to the observed increased expression of inducible nitric oxide synthase (iNOS) in these mice. Although there is increased expression of CUSOD (a critical scavenger for superoxide, hydroxyl ions, and peroxides in the cytoplasm), there is decreased expression of MBNSOD in muscle. MNSOD is a scavenger in the mitochondria (mt), and the loss of this molecule could result in significantly decreased protection against oxidative stress to mt DNA. In contrast, oxidative stress is significantly decreased in diabetic mice overexpressing IGF-1 in skeletal muscle. The levels of iNOS expression and DNA damage are reduced, and expression levels of MNSOD are retained. Together with results showing that muscle weight is better maintained in IGF-1 transgenic mice in the diabetic state, the results also showed that local expression of IGF-1 in skeletal muscle acted to protect muscle from diabetes-induced oxidative damage.

Effects of Myobacterium leprae on Schwann Cell/ Neuron Interactions In Vitro. Deanna Hagge, David Scollard, and Diana L. Williams.

Globally, millions of patients suffer from irreversible nerve damage, resulting in disabilities or blindness as a consequence of infection with Myobacterium leprae (ML). The mechanisms of nerve damage have not been fully elucidated, but appear to be the result of the presence of ML within Schwann cells, the neural target of ML in the peripheral nervous system, and an aggressive immune response to ML or its antigens within nerves. A major reason for this lack of understanding is that there are no well-developed in vivo or in vitro models available which maintain ML viability and closely mimic in vivo disease conditions.

Recently, these authors have developed an in vitro model and have used it to define the effects of ML on Schwann cell/neuron interactions. Results demonstrated that, although ML infection affects Schwann-cell morphology and specific gene expression, it does not appear to affect the ability of infected cells to associate with and myelinate axons of embryonic neurons in culture. In addition, preliminary results of experiments to define the effects of infection on intact myelinated Schwann cell/neuron co-cultures, suggest that ML does not affect the ability of these cells to maintain myelin-sheath architecture.

The conclusions of this study were that ML infection alone does not appear to adversely affect the ability of Schwann cells to participate in nerve regeneration processes. Preliminary data suggested that infection does not adversely affect the ability of these cells to function in intact Schwann cell/axon units. These data supported previous reports suggesting that nerve damage in leprosy results primarily from the immune response to the presence of ML within the nerves.

Sciatic Nerve Transection in the Adult Rat: Abnormal EMG Patterns during Locomotion by Aberrant Innervation of Hindleg Muscles. Albert Gramsbergen, Jos Ijkema-Paassen, and Marcel F. Meek.

Severe traumas to the extremities in humans often lead to lesions of the peripheral nerves. Functional recovery after such lesions usually is poor. Reconstruction with a poly(DL-polylactide-e-capralactone) artificial nerve guide in rats has been shown to speed up nerve outgrowth and to enhance functional recovery, compared to treatment with nerve grafts. In this study, EMG patterns during locomotion in rats in which a nerve gap was repaired with as graft, were compared with those in which the defect was reconstructed with an artificial nerve guide.

In the left hindleg of adult rats, a segment 12-mm long was resected from the proximal part of the nerve. In one group, this (reversed) segment was used as a nerve graft; in another group, a poly(DL-lactide-e-capralactone) nerve guide was applied. EMG patterns in the tibialis anterior (TA) and the gastrocnemius (GC) muscles at both sides were recorded during locomotion, and behavior was taped on video after recovery periods varying from 15 to 21 weeks. The specificity of axonal outgrowth was studied in grafted rats with unconjugated cholera toxin subunit B (CTB).

EMG patterns on the side at which the nerve gap was repaired with a graft were irregular, and coactivation of the GC and TA muscles was often observed. Burst activity was badly adjusted to the phases of the stepcycle. Similar patterns were observed after repair with the nerve guide, although these rats behaviorally showed improved locomotion. Retrograde labeling in the grafted rats indicated that both genuine and foreign motoneurons innervated the respective muscles on the left side.

The authors concluded that axons sprouting from the proximal stump of the nerve innervated the muscles aselectively, and that the motoneurons of origin maintain their original activation pattern. The improved behavioral outcome in the latter rats might be effected by central compensational mechanisms, that might be more effective after a shorter period of denervation, due to effective guidance of the outgrowing axons by the nerve guide.

(The artificial nerve-guide material is manufactured by Polyganics, The Netherlands, patent holder.)

Effect of GM-1 Ganglioside on Recovery of Motor Function Following Nerve Root Injury in a Rat Model. Mark Gonzalez, Thomas Atkins, David Bierbrauer, and James Kerns.

Nerve root avulsion injuries are sensory preganglionic, but motor postganglionic, as the anterior horn cell is preserved and is proximal to the injury. The purpose of this reported experiment was to investigate the effect of GM-1 ganglioside, a neural-sprouting agent, on motor recovery after nerve root avulsion in a rat model.

Twenty-seven Sprague-Dawley rats were separated into control (n&equals;15, 1 death) and treatment (n&equals;10, 1 death) groups. A straight posterior incision was made centered over the C4 vertebra, and a hemilaminectomy was performed. Using a nerve hook, the right side C5 nerve root was avulsed. Avulsion of the C5 nerve root achieves denervation of the ipsilateral forepaw biceps muscle. The treatment group received subcutaneous injections of GM-1 ganglioside at 30 mg/kg for 30 days. Recovery was based on the grooming response, which is a test of the return of biceps function (Bertelli and Mira, 1993). Water was dripped on the head of the rat, in order to elicit the innate grooming response. This response was graded into five groups, A-E: A, paw does not reach mouth; B, paw between mouth and nose; C, paw between nose and eye; D, paw between eye and ear; E, paw behind ear. The grooming response was elicited at prescribed intervals of 2 days, 1 month, 3 months, and 6 months, in order to determine functional return. Each of the grades of the grooming response was assigned a numeric value from 1-5, and the total value of all rats in each group at each interval was then summed. Values were analyzed using a one-way ANOVA.

The control group averaged 1.0 at 2 days and 1 month, and 1.4 at 3 and 6 months. None of the animals attained a grade greater than 3 at 6 months. The experimental group averaged 1.0 at 2 days, 1.4 at 1 month, 2.3 at 3 months, and 3.2 at 6 months. Four of the 10 animals in the experimental group achieved a score of 3 and four achieved a score of 4 at 6 months. This was highly statistically significant, p&equals;.0001. The results are promising, warranting further study of GM-1 ganglioside in the treatment of nerve root avulsions.

Gait Analysis in Composite Tissue Allotransplantation. Warren Breidenbach, Vijay Gorantla, Gustavo Perez-Abadia, Pascal Brouha, Marieke Vossen, Luis Laurentin-Perez, Cedric Francois, Claudio Maldonado, and John H. Barker.

Transplantation tolerance allows composite tissue allotransplantation (CTA) without immunosuppression but, in the authors' rat limb transplant model, also renders hosts susceptible to graft-vs.-host disease (GvHD). Limbs are rich in mature lymphocytes and cause GvHD when transplanted to chimeric hosts. In the authors' model, they have eliminated GvHD by irradiating donor limbs or by surgically removing lymph nodes from limbs. The purpose of the reported study was to determine the effect of irradiation on functional recovery of transplanted limbs.

Group 1 animals (n&equals;5) were normal WF rats. In Group 2 (n&equals;8), WF rats were made chimeric and tolerant by irradiation with 950 cGY and reconstitution with 100x106 T-cell depleted ACI bone marrow. Chimeras were transplanted with lethally irradiated (1050 cGY) ACI hindlimbs. Group 3 (n&equals;5) WF rats were made chimeric, as in Group 2, and were then transplanted with ACI hindlimbs after lymph node removal. All animals were observed daily for rejection and GvHD. Skin and tissue biopsies were taken regularly. Functional recovery in the transplanted limbs was assessed 3 months after transplantation using walking-track analysis and sciatic function index (SFI). Histomorphometry was performed on plastic-embedded nerve specimens collected at animal sacrifice.

In Group 2, no signs of rejection or GvHD were noted. Animals in Group 3 showed no clinical signs of rejection or GvHD. However, the histology showed mild polymorphonuclear infiltration in the skin of two recipients. The SFI of normal rats (Group 1) was 10.2±6.3. The SFI in Group 3 (-14.1±3.2) was significantly better than Group 2 (-69±10.4, p<0.05). Nerve histology (axon counts, myelin thickness) demonstrated good correlation with functional indices.

While a good method to prevent GvHD, irradiation appeared to impair functional recovery in this rat hindlimb transplant model. Surgical lymphadenectomy is an alternative to radiation in prevention of GvHD, resulting in improved functional return. Other strategies to prevent GvHD in CTA, that allow optimal functional recovery of limbs, should be explored. This will improve the clinical applicability of tolerance protocols in CTA.

Electrical Stimulation of Denervated EDL Rat Muscles for 15 Weeks Does Not Enhance Recovery of Maximum Force Following Reinnervation. Douglas E. Dow, Paul S. Cederna, Cheryl A. Hassett, Robert G. Dennis, and John A. Faulkner.

Following peripheral-nerve injury and repair, residual muscle weakness hinders effective rehabilitation. Recovery is especially poor, if reinnervation is delayed for many weeks. Prolonged periods of denervation result in greater muscle atrophy, which may inhibit functional recovery. Electrical stimulation of denervated muscles reduces atrophy. EDL muscles of rats denervated for 17 weeks maintain only 27% of muscle mass and 2% of maximum force of control muscles. In contrast, electrical stimulation of denervated muscles maintains 99% of the mass and 75% of the force. The authors' hypothesis was that electrical stimulation of denervated EDL muscles of rats during 15 weeks of peroneal nerve division would enhance recovery of muscle mass and maximum force following nerve repair.

The EDL muscles of 16 rats received one of four experimental treatments. Group 1-EDL muscle remained innervated (n&equals;6); Group 2-immediately repaired-peroneal nerve that innervates the EDL muscle was divided and immediately repaired (n&equals;10); Group 3-denervated/repaired-peroneal nerve remained divided for 15 weeks, then repaired (n&equals;7); Group 4-stimulated/denervated/repaired-peroneal nerve remained divided for 15 weeks, while EDL muscle was electrically stimulated (100 Hz, 20 pulses/contraction, 200 contractions/day by an implanted stimulator), then the nerve was repaired (n&equals;9). Functional evaluation occurred 6 months following nerve repair. Maximum force was measured by supramaximal electrical stimulation of the peroneal nerve, to generate isometric tetanic contractions measured by a force transducer tied to the exposed and divided distal tendons of the EDL muscle. Afterward, the EDL muscle was harvested and weighed.

Reinnervation failed in two muscles, one each from the denervated/repaired and stimulated/denervated/ repaired groups. Results from all other EDL muscle groups were presented. Muscle mass recovered to a higher level in the stimulated/denervated/repaired group than in the denervated/repaired group, but maximum force recovered no better.

The reduced level of atrophy in the denervated EDL muscles that were electrically stimulated during the 15 weeks of nerve division did not enhance recovery of maximum force following nerve repair. The critical issue for recovery following long-term denervation and nerve repair appears to be not just the maintenance of muscle mass and maximum force. Other factors not adequately affected by the stimulation protocol limit recovery.

Predicting Outcome in Obstetrical Brachial Plexus Palsy. Nancy de Kleer, Christine G. Curtis, Derek Stevens, and Howard M. Clarke.

The purpose of this study was to identify factors at an early age, which reliably predict outcome in cases of obstetrical brachial plexus palsy. Currently, lack of spontaneous recovery of elbow flexion at 3 months of age is used as an indication for surgery in centers around the world. This has not been scientifically validated, and there is no current gold standard to predict which children will improve with conservative management and which children will benefit from primary surgery.

Six hundred and four children assessed at the authors' brachial plexus clinic over the last 10 years were evaluated in a retrospective review of data which had been collected prospectively. Seventy-nine children, all of whom underwent a standardized brachial plexus physical examination at 3 months of age and had a minimum follow-up of 2 years, were analyzed in this study.

Four test scores, based on various combinations of validated physical examination maneuvers, were developed and calculated for children at their 3-month physical examination. One of the test scores was elbow flexion alone, while others incorporated additional movements such as wrist extension. The 3-month physical examination test scores were analyzed with respect to final functional outcome, regardless of whether or not surgery was undertaken. The ability of the test scores to predict outcome was analyzed with logistic regression, receiver operating characteristic curves, and tree-based analyses.

Outcome in obstetrical brachial plexus could be predicted using two recently developed test scores. Elbow flexion at 3 months of age does not predict outcome. Future studies will evaluate the test scores prospectively, and should yield a formula to calculate the child who will benefit from conservative vs. surgical management.