Search

The silicone chamber model for nerve regeneration is suitable to test the effects of exogenous agents or surgical manipulations on nerve regeneration. The total 16-day regeneration period used in this model makes it possible to analyze the effects of certain manipulations on the sequential advancement of the individual cellular components (circumferential perineurial-like cells, vessels, Schwann cells, axons, and myelin) into the chamber fibrin matrix. In the present study we compared the effects on cellular migration of a 7 day delayed chamber repair vs. chamber repair immediately after transection (control chambers) of the rat sciatic nerve. Regeneration was evaluated with light and electron microscopic techniques. Chambers implanted after a delay of 7 days had a statistically significant more advanced migration of vessels, Schwann cells, and axons from the proximal nerve stump and also a significantly increased vascular density as compared to control chambers. We conclude that a 7 day delayed nerve repair stimulates nerve regeneration in this specific silicone chamber model.

Nerve regeneration through predegenerated (PNG) or fresh (FNG) autografts in either fresh or delayed recipient nerve beds were studied in the rat sciatic nerve. Grafts 10 mm in length were excised either immediately or following a 7-day period of predegeneration. They were sutured into gaps on the contralateral side which were either freshly made or had been made 7 days previously. The early recovery (10 days at 2-day intervals) was evaluated by the sensory pinch test to measure the rate of regeneration and the delay period. The PNG group had an improvement in axonal regeneration as evidenced by a reduced delay period, a reduced number of regeneration failures, and less variability in regeneration distances compared to the FNG group. Conditioning the host site blocked some of the delay reduction in the PNG group, but had no effect on regeneration rate or delay in the FNG group. The presence of axons in the graft was confirmed by immunocytochemistry for neurofilament protein and by electron microscopy. The results suggest that proliferated cells, primarily Schwann cells, promote regeneration through the suture line and graft into the distal segment.

A mathematical model is presented by which the regeneration rate and initial delay for peripheral nerve regeneration can be calculated from sensory pinch test data following surgical repair of peripheral nerves. The model is based on the assumption that experimental variations in regeneration distances between animals is due to the initial delay period--the time period before the regenerating fibers cross the suture line whereas the rate of regeneration is constant. This model which accounts for all observed data including 'regenerating failures' showed that nerve fibers in the rat sciatic nerve repaired with a fresh nerve graft regenerated at a rate of 1.5 mm/day after an initial delay of 3.6 days.

The ulnar nerve of a 21-year old man was repaired at the wrist by a silicone chamber technique 10 days after a traumatic transection. A 3 mm gap was left between the nerve ends inside the chamber. At follow-up three years later, motor and sensory recovery was excellent. At exploration at that time a macroscopically normal nerve was found in the tube.

Immediate and long-term microvascular effects of chondroitinase ABC, 200 unit/ml, were analyzed in ten hamsters. The immediate effects on the microcirculation were studied by vital microscopy following local injection in the cheek pouch. There were no detectable effects on the microvascular blood flow during the 60 minutes of observation for chondroitinase ABC or the control. A therapeutic concentration (2000 pKat/ml) of chymopapain stopped the microcirculation in the injected area immediately, with numerous microbleedings at the border zone. Long-term effects were studied after subcutaneous injections in the ears of six rabbits. Chondroitinase ABC and the control did not cause any macroscopic or microangiographic effects. However, light microscopy showed a moderate inflammatory reaction in the subcutaneous layer for both chondroitinase ABC and the control. Chymopapain induced severe effects on the cartilage and surrounding tissues. Microangiography revealed a vessel-free zone at the injection site. Since 200 units/ml of chondroitinase ABC is four to eight times higher than the concentration that might be used for chemonucleolysis, i.e., dissolution of intervertebral discs by local enzyme injection, the present investigation suggests a wide margin of safety regarding the potential effects on blood vessels in tissues surrounding the disc.

The silicone nerve regeneration chamber is a useful model to investigate the cellular and molecular events underlying successful regeneration in the peripheral nervous system. In this model a transected rat sciatic nerve with a 10-mm interstump gap, is repaired with a silicone chamber. The spatial-temporal sequence of regeneration in the silicone chamber has been examined in detail. The chamber rapidly becomes filled with fluid which contains neurotrophic activity for neurons in vitro. The second event to occur is the formation of a fibrin matrix connecting the two nerve stumps. This matrix is then invaded by cellular elements in the following order: perineurial-like cells, vasculature, Schwann cells, and axons. The silicone chamber model also allows manipulation of the regeneration process. Prefilling the chamber at the time of implantation with phosphate-buffered saline or dialyzed plasma stimulates nerve regeneration. Multiple injections into the chamber of a mixture containing laminin, testosterone, and ganglioside GM1 increase the size and the vascularization of the regenerate. Specially designed chambers divided into two compartments by a longitudinally inserted nitrocellulose strip have been used to examine the effects of substrate-bound trophic factors on nerve regeneration. Fibroblast growth factor containing chambers have an improved regeneration and vascularization as compared to controls.

The mechanisms influencing regeneration of peripheral nerves are incompletely known, but growth factors are supposed to play a key role. In the present study, we demonstrate, with the aid of immunohistochemical methods, that somatomedin C (Sm-C/insulin-like growth factor I/IGF-I) rapidly increased from low to high concentrations, reaching peak values in 2 weeks, in regenerating sciatic nerves of adult rats. In addition, IGF-I was demonstrated extracellularly, never observed in the control nerves. Reactive Schwann cells appeared to be the major source for IGF-synthesis. Higher concentrations were seen in tubulated nerves as compared to sutured ones. It is proposed that IGF-I exerts important growth supporting effects on regenerating peripheral nerves.

Regeneration of severed peripheral nerves is unfortunately often incomplete, due to loss of nerve fibers and neuroma formation. A new approach is presented with the intention of improving the conditions for nerve repair. In the first of the two stages, a pseudosynovial tube is formed around a silicone rubber rod, surrounded by a stainless steel spiral, which was placed in the backs of rats. This tube, in the second stage, is used as a free "tube graft" to bridge gaps of about 10-12 mm lengths in the severed sciatic nerve. The tube was kept open by the metal spiral. Regenerating nerve fibers with their sprouts grew into the initially open space in the tube. A new nerve trunk was formed, comprised of closely packed myelinated and unmyelinated axons, organized into fascicles. Demonstration by electron microscopy and by EMG recording of reinnervation of foot muscles supported successful long-term results. The fascicles were delimited by perineurial and epineurial sheaths and, furthermore, showed signs of maturation. It was also demonstrated that the nerve-fiber regeneration ceased after a few weeks if there was no distal nerve inserted into the tube. The importance of optimizing the interaction between local factors and regenerating nerve fibers for reestablishment of functionally valuable motor units is discussed.

Compound action potentials of both myelinated (A) and non-myelinated (C) fibres in the common peroneal nerve of rabbits were studied during and after acute, graded compression of the nerve at 200 or 400 mmHg applied for 2 h or during ischaemia created by nitrogen inhalation or aortic occlusion. Compression of the nerve at 200 mmHg blocked the AI component (large myelinated fibres) after about 23 min, while compression at 400 mmHg shortened this time to 11 min. The A2 component (thinner myelinated fibres) had a lower conduction velocity and a higher resistance to compression. There was just a slight decrease in conduction velocity of the non-myelinated fibres when the nerves were compressed at 200 mmHg for 2 h. However, compression at 400 mmHg for 2 h induced a marked deterioration of amplitude and conduction velocity of the C-fibres. There was an incomplete restitution of function of A- and C-fibres during 2 h of recovery. The thinner myelinated fibres were more susceptible to deprivation of oxygen than the thicker ones, while non-myelinated fibres differed in response according to method of ischaemia induction. It is concluded that non-myelinated fibres are very resistant to compression and a very high pressure (greater than 400 mmHg) is needed to affect these fibres.

The axonal transport and distribution of the fast phase of [3H]leucine-labeled proteins were used to monitor the outgrowth delay and regeneration rate in rabbit hypoglossal nerves 5-21 days after crush or transection. The transected nerves were repaired with mesothelial chambers or epineurial sutures. Radiolabeled proteins were transported into regenerating axons in the distal nerve segment after an initial delay of 2.5 days for crushed nerves and after a delay (initial and scar delays) of 4.8 and 5.7 days for sutured and mesothelial chamber-reconnected nerves, respectively. Regeneration rate was 3.5 mm/day after a crush and 2 mm/day after a transection with either type of repair. Total radioactivity was greater in both crushed and repaired nerves than in their contralateral controls. Transported radioactivity accumulated at the site of the lesions. This accumulation was greater and persisted longer in repaired nerves than in crushed ones. The difference in regenerative response after different types of trauma with respect to changes in axonal transport is emphasized.

Gore-tex chambers were used to bridge a 6 mm gap between the proximal and distal nerve stumps of a rat sciatic nerve. The wall structure of these chambers is characterized by "nodes" interconnected by smaller fibrils. Chambers with internodal distances of 5, 10 and 30 microns were used. Some 30 microns chambers were coated from the outside with Gore-tex (0.2 micron internodal distance) and others were coated from the inside. Regeneration after 12 weeks, as evidenced by muscle action potential recordings and light microscopy, was successful regardless of what type of chamber had been used. The organization of the nerve structure varied among different chamber types. A well organized coaxial nerve structure with myelinated axons was observed if inner-coated chambers were used. In chambers that were not coated or in outer-coated chambers tissue completely filled the chambers, and myelinated axons were arranged in mini-fascicles surrounded by loose connective tissue.

In 1944 Weiss & Taylor presented experimental evidence against "neurotropism" in nerve regeneration. We used a silicone Y-chamber system to repeat some of those experiments. The proximal stump of transected rat sciatic nerve was introduced into the proximal inlet of the Y. One of the distal outlets was left empty, plugged or occupied by a tendon graft, the other outlet being occupied by a nerve graft. Analysis after 4 and 12 weeks showed in all cases a preferential or exclusive axonal growth towards the nerve piece. The results, indicating the existence of "tissue specificity", are contradictory to the results reported by Weiss & Taylor (30). The findings are discussed with respect to possible influence of humoral, cellular and molecular factors associated with the distal nerve stump as well as the matrix, formed between both nerve segments.

Preformed, autologous mesothelial chambers were utilized to study axonal growth following selective predegeneration of the distal nerve stump and/or preconditioning of the proximal nerve stump. The left and/or right sciatic nerve of rats was exposed and transected in the thigh. Two weeks after transection, the left proximal nerve stump was cross-anastomosed with the right distal nerve stump by using a mesothelial chamber leaving a 15-mm gap between the two nerve stumps. Previous studies have shown that axonal overgrowth normally does not occur over this gap distance to the distal stump. Three months after cross-anastomosing, regeneration across the 15-mm gap was evaluated by muscle action potential recordings and light microscopical examination. In experiments in which a distal nerve stump was selectively degenerated and the proximal segment was freshly cut, axons had bridged the 15-mm gap in six of seven rats. When a proximal preconditioned nerve stump was matched with a freshly cut distal stump, axonal overgrowth occurred in only 4 of 10 experiments. In experiments including a proximal preconditioned nerve stump and a distal predegenerated stump, axons bridged the gap in 6 of 8 experiments. We concluded that a priming lesion, including manipulation with proximal and/or distal stump, enhances axonal growth in mesothelial chambers.

An experimental model is presented for studying nerve regeneration over gaps of various lengths between the both ends of a severed nerve. After transferring left and right sciatic nerves of rat to the back, the gap between the two nerve ends could be bridged by a preformed, tube-shaped mesothelial chamber of a desired length. When the gap length was 10 mm or less, a well developed nerve structure was generated in the chamber between the nerve ends, and axons from the left sciatic nerve reinnervated muscles in the right limb via the right sciatic nerve. When the gap length was extended to 15 mm or more no such regeneration occurred. When no distal nerve end was introduced into the chamber, there was a limited growth into this chamber over only 5-6 mm. This "limited growth phenomenon" is discussed with respect to a lack of "trophic" or cellular support from a distal nerve segment. It is also proposed that the termination of growth, seen under these circumstances, may suggest a new principle for avoiding the development of neuromas after nerve transections.

An experimental model is presented for studying axonal growth after experimental hyperthyroidism and hypothyroidism. The left sciatic nerve of the rat was transected and transposed to the back. The proximal nerve stump was inserted into a 50-mm-long mesothelial chamber leaving the distal end of the chamber open. Different groups of young adult rats were given daily injections of thyroxine (10 micrograms/100 g body weight) or the goitrogen, thiamazol, in the drinking water (0.125 g/liter) for 12 weeks. Thyroxine treatment increased significantly the extent of axonal outgrowth from the proximal nerve stump compared with untreated rats. Experimental hypothyroidism (thiamazol treatment), evidenced by a retarded body growth, did not affect the extent of axonal outgrowth. In other experiments the left proximal nerve stump was cross-anastomosed with the right distal nerve stump. The two nerve stumps were bridged with a mesothelial chamber leaving a 15-mm gap. This gap distance is known from our previous studies to inhibit axonal overgrowth to the distal nerve stump. As evidenced by histological evaluation, in three of six thyroxine-treated rats, axons had bridged the 15-mm gap. We conclude that experimentally induced hyperthyroidism enhances axonal growth in mesothelial chambers.

We have compared the anatomic and functional regeneration of a transected sciatic nerve following regrowth from its proximal stump through either preformed empty mesothelial chambers or autologous nerve grafts bridging a 10 mm gap. Within the mesothelial chambers an organized multifascicular nerve trunk forms between the proximal and distal stumps. After 3 months, distal segment cross sections from the mesothelial chamber and nerve graft groups did not differ with respect to axonal density or distribution of axonal diameters. Mean conduction velocities across the gaps were also similar, although the nerve graft group had a wider distribution of velocities. Little or no regeneration was evident when the gap between the nerve stumps was left empty. These results suggest that if the regrowing proximal stump is in an appropriate environment, it can form a well organized and oriented nerve trunk. In the mesothelial chambers, the regenerating nerve is surrounded by a loose cellular stroma and a small amount of interstitial fluid, which was found to contain trophic activity for cultured rodent sensory neurons. Such factors may also support nerve regeneration in vivo.

When silicone regeneration chambers are implanted empty, axonal regeneration fails if the interstump gap length is greater than 10 mm. Previous experiments using the 10-mm gap model demonstrated that regeneration success correlated with the dimension and/or consistency of the naturally formed acellular fibrin matrix. Both spatial and temporal parameters of regeneration could be stimulated through modifications of the fibrin matrix by prefilling the chambers at the time of implantation either with phosphate-buffered saline or plasma dialyzed against phosphate-buffered saline. In the present experiments, similar modification of matrix formation was found to promote successful regeneration across 15-mm and 20-mm interstump gap lengths. In addition, prefilling 15-mm-gap chambers with dialyzed plasma resulted in a 3.5-fold increase in the incidence of functional restitution detected at 8 weeks after implantation over the outcome with chambers prefilled with phosphate-buffered saline.

The distribution of fast migrating [3H]leucine-labelled proteins was studied in transected and repaired rabbit hypoglossal nerves. The nerves were repaired 90 days earlier with mesothelial chamber or epineurial suture technique. Fast migrating radiolabelled proteins were transported into the distal nerve segment and neurophysiological recordings from the tongue as well as the presence of myelinated axons in the distal nerve segment verified successful regeneration. The total amount of radioactivity was increased in repaired nerves as compared to contralateral nerves. In both groups there was a significant accumulation of radiolabelled proteins at the site of lesion. Nerves repaired with mesothelial chambers showed significantly more radioactivity in the distal nerve segment as compared to sutured nerves. The present study indicates long-standing effects on axonal transport system after both types of nerve repair. It is our opinion that axonal transport studies are a valuable complement when evaluating experimental nerve repair.

In the nerve regeneration silicone chamber model, the regenerate which forms across a 10-mm gap between proximal and distal nerve stumps is a monofascicular structure with an outer perineurial-like cell sheath. Recent work has provided indications that the geometry of the regenerate within a silicone chamber can be altered by experimental modifications of the chamber matrix. In the present study we modified the standard silicone chamber into a two-compartment chamber by inserting a 6- or 10-mm-long siliconized nitrocellulose strip in order to obtain two separate regenerates. Light microscopy 16 days after implantation revealed that two separate nerve structures had formed, one on each side of the nitrocellulose partition and adjacent to it, and each with its own perineurial-like cell sheath. In chambers with 6-mm-long strips a monofascicular regenerate started from the proximal stump and divided into two separate structures as it approached the proximal end of the strip: the two fascicles joined again into a monofascicular structure in the distal portion of the chambers. The new two-compartment silicone chamber model appears suitable for future examinations of experimental fasciculation. In addition, the nitrocellulose partition should allow one to study specific effects of growth factors on axonal regeneration in vivo, as growth factors bind strongly to untreated nitrocellulose while retaining their biological activity.

An extracellular matrix preparation, the human amnion membrane matrix (hAMM) can serve as a neurite-promoting substratum for cultured peripheral and central neurons, and also as a support for axonal growth in experimentally injured adult brain in vivo. In the present study, we tested similar materials as bridges in a silicone chamber model for the regeneration of sciatic nerve in the adult rat. Since hAMM elicited an inflammatory response, we developed a rat amnion membrane matrix (rAMM), which proved to be an excellent neurite-promoting substratum for cultured ganglionic and spinal cord neurons. The rAMM was coiled and inserted in the 10 mm gap between the two nerve stumps from the silicone chambers. At 16 days after implantation, temporal progress of regeneration was grossly similar as in saline-prefilled control chambers. However, rAMM-prefilled chambers displayed significantly higher number of vessels and a markedly different geometry of the regenerate: an endoneurium, surrounded by a perineurial-like cell layer, was formed outside the largely preserved central portion of the rAMM coil. After longer regeneration times (28 days), a rAMM core was no longer detected, but some rAMM-like materials remained interspersed in the endoneurium. The overall organization of the regenerate and the number of myelinated axons at this time were similar to those of control chambers, although the endoneurial cross-sectional area was larger in the rAMM chambers. One specimen, however, displayed the very patterns for which the experiments were designed, namely an array of numerous, myelinated axons tracing the spiraling spaces between consecutive lamellae of the rAMM coil.