Traumatically injured spinal cord (SC) displays structural damage that includes discontinuity of long tracts and cavitations. Axonal regrowth beyond the lesion is necessary to achieve functional recovery following SC injury. We report here the development of an artificial collagen-filament (CF) scaffold to replace the SC in 8-week-old female Fisher rats. Axonal sprouting and regrowth was very rapid following grafting of the CF. One week after implantation, the scaffold was filled with cells of host origin and with regenerated axons. Histological examination of SC adjacent to the scaffold showed little cavity formation or fibrous scarring. Eight weeks after implantation, myelinated nerve fibers were found in the scaffold and 10–25 % of rubrospinal tracts were repaired. Four to six weeks after transplantation, motor evoked potentials were recorded in CF-grafted rats but were not detectable in non-grafted rats. Electrophysiological and histological examinations revealed the grafted CF was likely to function as a nerve tract. In addition, these results suggest that collagen fibers may provide a permissive microenvironment for the elongation of SC axons and to support the process of spinal cord regeneration.

Na-x, a sodium concentration-sensitive sodium channel, is expressed in non-myelinating Schwann cells of the adult peripheral nervous system, but the pathophysiological role remains unclear. We found that functional recovery of the hind paw responses from the sciatic nerve transection was delayed in Na-x knockout (Nax-/- ) mice. Histological analyses showed a decrease in the number of regenerated myelinated axons in Nax-/- sciatic nerves. The delay in the recovery in Nax-/- mice was improved by lactate and inhibited by a monocarboxylate transporter inhibitor. In vitro experiments using cultured Schwann cells showed that lactate release was enhanced by endothelin (ET)-1 and blocked by an ET receptor type B antagonist. Here, it is conceivable that Na-x was activated by ET-1. The amount of lactate release by ET-1 was lower in Nax-/- mice than in wild-type mice. These results indicated that Na-x is functionally coupled to ET for lactate release via ET receptor type B and is involved in peripheral nerve regeneration.

A recent study of a rat model treated with grafted collagen filament (CF) after spinal cord transection showed dramatic recovery of motor function but did not report on the acute-stage phenomenon. In the present study, we describe molecular and histological aspects of the axonal regeneration process during the acute stage following spinal cord transection. The spinal cord of 8-week-old rats was completely transected, and a scaffold of almost the same size as the resected portion was implanted in the gap. Changes in the mRNA expression of four neurotrophic factors [nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), NT-3, and glial cell-derived neurotrophic factor (GDNF)] were analyzed after 72 h. The expression of BDNF and NT-3 mRNA increased significantly in the CF-grafted group compared to the nongrafted group. Immunostaining for BDNF and NT-3 revealed that cells positive for these neurotrophic factors extended along the collagen filaments in the CF-grafted group. Similarly, astrocytes extended into the collagen filament scaffold together with the neurotrophic factors and partly across a border line. These findings indicate that collagen filament helps to reduce scar tissue, supports the expression of neurotrophic factors, and serves as a scaffold for the outgrowth of regenerating axons.

We evaluated sciatic nerve regeneration in thy1-YFP transgenic mice selectively expressing a fluorescent protein in their axons. Using in vivo imaging, we observed the dorsal cutaneous renervation of the hind paw for 8 weeks. Three to four weeks after the operation, the length of the regenerated nerve treated with NGF tended to be longer than that of the regenerated nerve treated with saline. Functional recovery was evaluated by a withdrawal response of the hind paw to mechanical stimuli. In NGF and GDNF groups, mice started to resume a mechanical response 4 weeks after the operation, earlier than in the saline control group. Histological and ultrastructural analyses showed that the density of unmyelinated axons in the regenerated nerve of the NGF group was larger than that of those in the saline group. These results indicate that NGF accelerated the regeneration of the sciatic nerve and thus that the monitoring of cutaneous nerve regeneration in the dorsal foot is useful to evaluate the regeneration of the sciatic nerve in vivo.

The aim of the present study was to investigate the relationships between interleukin 6 (IL-6) and insulin resistance, serum leptin, serum adiponectin, or serum lipids including triglycerides in 98 nonobese Japanese type 2 diabetic patients. Insulin resistance was estimated by the insulin resistance index of homeostasis model assessment (HOMA-IR). Serum IL-6 concentration was negatively correlated to high-density lipoprotein cholesterol (r = −0.295, P = .004), but was not associated with HOMA-IR (r = 0.016, P = .871), body mass index (BMI) (r = 0.090, P = .375), systolic (r = 0.169, P = .116) and diastolic (r = −0.061, P = .570) blood pressures, leptin (r = 0.062, P = .544), and adiponectin (r = −0.020, P = .841) in these patients. In contrast, serum leptin level was positively correlated to HOMA-IR (r = 0.291, P = .004), BMI (r = 0.338, P < .001), and systolic blood pressure (r = 0.241, P = .025). Serum adiponectin level was negatively correlated to HOMA-IR (r = −0.288, P = .005), BMI (r = −0.308, P = .002), diastolic blood pressure (r = −0.269, P = .012), and triglycerides (r = −0.338, P < .001), and positively correlated to high-density lipoprotein cholesterol (r = 0.300, P = .003) in our patients. From these results, it can be suggested that fasting serum IL-6 is not a major factor responsible for the evolution of insulin resistance in nonobese Japanese type 2 diabetic patients.

The aim of the present study was to investigate the relationship between insulin resistance and tumor necrosis factor alpha (TNF-alpha) as well as soluble TNF receptors (sTNF-R), body mass index (BMI), leptin, adiponectin, and serum lipid profile including triglycerides in nonobese Japanese patients with type 2 diabetes. A total of 88 nonobese Japanese type 2 diabetic patients were studied. The duration of diabetes was 11.0 +/- 0.8 years. In conjunction with BMI, glycosylated hemoglobin (HbA1c), fasting concentrations of plasma glucose, serum lipids (triglycerides, high-density lipoprotein cholesterol, and total cholesterol), serum leptin, serum adiponectin, serum TNF-alpha, and soluble TNF receptors (sTNF-R1 and sTNF-R2) were also measured. Insulin resistance was estimated by the insulin resistance index of homeostasis model assessment. Insulin resistance was positively correlated with BMI, triglycerides, leptin, and total cholesterol and negatively correlated with adiponectin and high-density lipoprotein cholesterol. In contrast, insulin resistance was not associated with TNF-alpha, nor sTNF-R (sTNF-R1 and sTNF-R2) in our diabetic patients. There was no significant relationship between the 3 measures of TNF-alpha system (TNF-alpha, sTNF-R1, and sTNF-R2) and BMI, serum triglycerides, leptin, or adiponectin in these patients. From these results, it can be concluded that peripheral levels of TNF-alpha (system activity are not a major factor responsible for insulin resistance in nonobese Japanese type 2 diabetic patients. (c) 2005 Elsevier Inc. All rights reserved.

The aim of the present study was to investigate the relationships between insulin resistance and soluble E-selectin, body mass index (BMI), leptin, and serum lipid profile including triglycerides in nonobese Japanese type 2 diabetic patients. A total of 97 nonobese Japanese type 2 diabetic patients aged 43 to 84 years were examined. The duration of diabetes was 11.2 +/- 0.8 years. In conjunction with BMI and fasting concentrations of plasma glucose, serum lipids (triglycerides, total cholesterol, and high-density lipoprotein cholesterol) and serum insulin, soluble E-selectin, and leptin were also measured. The low-density lipoprotein (LDL) cholesterol level was calculated using the Friedewald formula. Insulin resistance was estimated by the homeostasis model assessment. The subjects were divided into 2 groups according to the value of insulin resistance estimated by the homeostasis model assessment. Values greater than 2.5 were indicative of the insulin-resistant state, and values less than 2.5 were indicative of the insulin-sensitive state. The insulin-resistant group had significantly higher levels of E-selectin, leptin, triglycerides, total and LDL cholesterol, and diastolic blood pressure as compared with the insulin-sensitive group. There was, however, no significant difference in age, sex, diabetes duration, BMI, systolic blood pressure, HbA1c, and high-density lipoprotein cholesterol between the 2 groups. Univariate regression analysis showed that insulin resistance was positively correlated to E-selectin (r = 0.305, P = .003), BMI (r = 0.283, P = .006), leptin (r = 0.296, P = .004), HbA1c (r = 0.241, P = .018), serum triglycerides (r = 0.385, P < .001), serum total (r = 0.240, P = .019) and LDL cholesterol (r = 0.254, P = .013) levels, and systolic (r = 0.247, P = .024) and diastolic (r = 0.305, P = .006) blood pressure. Multiple regression analyses showed that insulin resistance was independently predicted by serum E-selectin (F = 18.4), serum leptin (F = 14.0) and serum triglycerides (F = 20.0) levels, which explained 45.0% of the variability of insulin resistance. From these results, it can be concluded that in conjunction with serum triglycerides and serum leptin, serum E-selectin is another important independent factor associated with insulin resistance in nonobese Japanese type 2 diabetic patients.

We have developed a block of glass-ceramic to augment the dysplastic acetabulum. Four cases of acetabular dysplasia underwent implantation of a block of glass-ceramic on the lateral surface of the ilium, a pelvic bone, just above the hip joint. In these cases, the mean Harris hip score was 46+/-8 points pre-operatively and 85+/-13 points 4 to 7 years post-operatively. Our procedure is unique in that we use a bioactive glass-ceramic block as an artificial bone to construct the acetabular bone. The results appear to be satisfactory.

This article describes the regeneration of severed peripheral nerve axons along collagen filaments in the absence of the distal nerve stump. 22-mm long nerve guides made of collagen filaments were sutured to the proximal ends of severed rat sciatic nerves. The distal ends of the guides were sutured to the distal stumps of the nerves in a group and not sutured in the other. Nerve autografts and collagen tubes were used as controls. At 8 weeks postoperatively, the mean number and the mean diameter of myelinated axons were 5491+/-617 (mean-+/-SD) and 2.3+/-1.3 mum at the distal ends of the collagen filaments nerve guides those the distal ends were sutured to the distal stumps of the nerves, while in the nerve autografts these were 4837+/-604 and 3.3+/-1.4mum. These were 1992+/-770 and 2.7+/-1.2 mum at the distal ends of the collagen-filaments guides those the distal ends were not sutured to the distal stumps of the nerves, while in the nerve autografts these were 3041+/-847 and 2.3+/-1.1 mum. No axon was found at the distal ends of the collagen tubes. The results suggested that the contact guidance and the chemotaxis guided regenerating axons along the collagen filaments.

The restoration of function of transected adult mammalian spinal cord without living tissue has not been reported previously. We report the first success of functional restoration of transected spinal cord without living tissue. We grafted collagen filaments parallel or transverse to the axis of the spinal cord to bridge 5-mm defects of 47 adult rat spinal cords. Twenty-five rats were used as a control. Of the 72 rats, 42 rats survived the experimental period. At 4 weeks postoperatively, regenerated axons crossed the proximal and distal spinal cord-implant interfaces in all 5 rats of the parallel-grafted group. At 12 weeks postoperatively, the rats in the parallel-grafted group (8 rats) could walk, run, and climb with hind-forelimb coordination. The somatosensory-evoked potentials were seen. Results suggest that the collagen filaments support the axonal regeneration of the transected spinal cord and the restoration of function when grafted parallel to the axis of the spinal cord. The functional restoration appeared to be permanent, raising the possibility of therapeutic application in humans.

This article describes a 30-mm regeneration of severed peripheral nerve axons along collagen filaments. Two thousand or 4000 31-mm-long collagen filaments were grafted to bridge a 30-mm defect of the rat sciatic nerve. A collagen tube was grafted as a control. The mean number and mean fiber diameter of regenerated myelinated axons were 330 +/- 227 and 2.7 +/- 0.9 microm in the distal end of the 2000 collagen-filaments nerve guide, and 564 +/- 275 and 2.5 +/- 1.1 microm in the distal end of the 4000 collagen-filaments nerve guide at 12 weeks postoperatively, whereas in the distal end of the collagen tube, no regenerated axon was found. These results suggest that the collagen filaments guide axons of the rat's sciatic nerve to regenerate for 30 mm and act as a scaffold for axonal regeneration. Thirty-millimeter nerve regeneration of a 1-mm-diameter rat sciatic nerve by an artificial nerve guarantees a clinical application of the implant which should be very important for patients and surgeons.

Background: The aim of the present study was to investigate the relationships between serum leptin levels and regional adipose fat area, BMI, and the measures of variables including serum insulin in nonobese Japanese type 2 diabetic patients.Methods: A total of 121 nonobese Japanese type 2 diabetic patients [aged 35 to 83 years, body mass index (BMI) (15.4 to 26.8 kg/m(2))] were studied. They all were male patients. In conjunction with serum leptin level, BMI, glycosylated hemoglobin (HbA(1c)), and fasting concentrations of plasma glucose and serum insulin and lipids (triglycerides, total and HDL cholesterol) were measured.Results: Univariate regression analysis showed that serum leptin levels were positively correlated to subcutaneous (r = 0.566, P < 0.0001) and visceral (r = 0.481, P < 0.001) fat area in our diabetic patients. Furthermore, serum leptin levels were positively correlated to serum insulin (r = 0.517, P < 0.0001), BMI (r = 0.428, P < 0.0001), serum triglycerides (r = 0.279, P < 0.005), and age (r = 0.225, P < 0.05). There was, however, no relationship between serum leptin levels and measures of other variables including total and HDL cholesterol. Multiple regression analyses showed that serum leptin levels were predicted by subcutaneous fat area (F = 5.92, P < 0.0001) and serum insulin level (F = 5.60, P < 0.0001), which explained 29.0% of the variability of serum leptin concentrations in our nonobese Japanese type 2 diabetic male patients. Visceral fat area, BMI, serum triglycerides, and age, however, were not independently associated with serum leptin levels in our patients.Conclusions: These results indicate that serum leptin levels are reflective of subcutaneous fat area in nonobese Japanese type 2 diabetic male patients.

The aim of the present study was to investigate whether non-obese Japanese type 2 diabetic patients with porphyromonas gingivalis infection have atherosclerotic vascular diseases. A total of 134 non-obese Japanese type 2 diabetic patients (96 men and 38 women, aged 36 to 84 years, body mass index [BMI] 20.1 to 26.9 kg/m(2)) were studied. In conjunction with BMI, glycosylated hemoglobin (HbA(1c)), fasting glucose, and serum lipids (triglycerides, total cholesterol, high-density lipoprotein [HDL] cholesterol, low-density lipoprotein [LDL] cholesterol) were measured. LDL cholesterol was calculated using the Friedewald formula. Using high-resolution B-mode ultrasound scan, we measured intimal medial thickness (IMT) in plaque-free segments of bilateral common carotid arteries, and the mean of IMT in 2 vessels was used for the analysis. Furthermore, we calculated the degree of stenosis in plaque segments of bilateral common carotid arteries. The degree of carotid atherosclerosis was expressed as a percentage ratio between the area of plaque and that of the lumen using the formula (Lumen Area Residual - Lumen Area)/Lumem Area x 100. Both the areas were automatically measured by the system on a frozen transverse scanning plane at the site of maximal narrowing. When 2 or more plaques were present in the vessel, only that causing the greatest degree of stenosis was considered for analysis. Values represent mean+/-SEM unless otherwise stated. Immunoglobulin G (IgG) titer against porphyromonas gingivalis was 245 +/- 65 (mean +/- 2 SD) in nondiabetic healthy subjects. In contrast, there was a wide variation in IgG titer against porphyromonas gingivalis in type 2 diabetic patients studied (range, 16 to 26,800). Thus, we classified our type 2 diabetic patients into 2 subpopulations according to the value of mean +/- 2 SD (= 310) of nondiabetic healthy subjects: one with high IgG titer against porphyromonas gingivalis (>310) (1,422 +/- 408) and the other with normal IgG titer against porphyromonas gingivalis (<310) (152 +/- 10, P =.002). The populations did not differ with respect to age, sex, BMI, fasting glucose, HbA(1c), serum triglycerides, total, HDL, and LDL cholesterol levels. Although the mean IMT in plaque-free segments was not different between the 2 groups (0.73 +/-0.03 v 0.68 +/- 0.02 mm, P =.098), the degree of stenosis in plaque segments was significantly higher in the high IgG titer group (12.0% +/- 2.2%) than in normal one (5.5% +/- 1.4%, P =.009). From these results, it can be concluded that porphyromonas gingivalis infection, although still a subclinical infection, is associated with atherosclerotic vascular disease in non-obese Japanese type 2 diabetic patients.

STUDY DESIGN:A rat model of spinal cord defect was designed to evaluate the effect of collagen filament implant on nerve regeneration in the spinal cord defect. OBJECTIVES:To bridge a spinal cord defect and restore the function in adult mammals. SUMMARY OF BACKGROUND DATA:Resection of the spinal cord in mammals is always followed by motor paralysis and loss of voluntary function below the lesion. Partial success in bridging the ends of the spinal cord after complete resection was reported. However, restoration of function has not been reported in adult mammalian. MATERIALS AND METHODS:Four thousand collagen filaments 5-mm-long were grafted to bridge a 5-mm defect of rat spinal cord. Controls had their spinal cord defect left ungrafted after resection. At 1-week intervals, animals were evaluated functionally. After 4 and 12 weeks, animals were evaluated histologically. After 12 weeks, animals were evaluated electrophysiologically. RESULTS:The severed spinal cord axons regenerated along the collagen filament implant crossing the proximal and distal spinal cord implant interfaces at 4 weeks after surgery. The rats with collagen filament grafts could walk, run, and climb with hind forelimb coordination at 12 weeks after surgery. Sensory-evoked potential waveform was found in the rats with collagen filament at 12 weeks after surgery. CONCLUSIONS:The collagen filaments support the axonal regeneration of the transected spinal cord and the restoration of function.

The aim of the present study was to investigate the relationship between platelet count and insulin resistance in non-obese Japanese type 2 diabetic patients. A total of 163 non-obese Japanese type 2 diabetic patients (112 men and 51 women, aged 36 to 84 years, body mass index [BMI] 16.2 to 26.9 kg/m(2)) were studied. In conjunction with BMI, glycosylated hemoglobin (HbA(1c)), fasting concentrations of plasma glucose and serum lipids (triglycerides, low-density lipoprotein [LDL] cholesterol, high-density lipoprotein [HDL] cholesterol, and total cholesterol), and hematological parameters (platelets, white blood cell count, red blood cell count, hematocrit, hemoglobin) were measured. LDL cholesterol was calculated using the Friedewald formula. Insulin resistance was estimated by the insulin resistance index of homeostasis model assessment (HOMA-IR). Univariate regression analysis showed that HOMA-IR was positively correlated to BMI (r = 0.465, P <.0001), HbA(1c) (r = 0.423, P <.0001), platelet count (r = 0.310, P <.0001), triglycerides (r = 0.277, P <.0005), white blood cell count (r =.222, P =.005), red blood cell count (r = 0.210, P =.008), hematocrit (r = 0.156, P =.047), total cholesterol (r = 0.178, P =.023), and systolic (r = 0.216, P =.011) and diastolic (r = 0.263, P =.002) blood pressure, and inversely correlated to HDL cholesterol (r = -0.312, P <.0001) level in our diabetic patients. Multiple regression analysis showed that HOMA-IR was independently predicted by BMI (P <.0001, F = 22.45), HbA(1c) (P <.0001, F = 16.15), platelet count (P <.0001, F = 10.75), and serum triglycerides (P <.0001, F = 10.47) levels, which explained 34% of the variability of HOMA-IR in non-obese Japanese type 2 diabetic patients. These results indicate that not only BMI, HbA(1c), and triglycerides levels but also platelet counts are independent predictor of insulin resistance in non-obese Japanese type 2 diabetic patients.

The aim of the present study was to investigate the relationship between C-reactive protein (CRP) and insulin resistance in non-obese Japanese type 2 diabetic patients. A total of 135 non-obese Japanese type 2 diabetic patients (96 men and 39 women, aged 36 to 83 years, with a body mass index [BMI] of 16.2 to 26.8 kg/m2) were studied. BMI, glycosylated hemoglobin (HbA(1c)), fasting concentrations of plasma glucose, serum lipids (triglycerides, low-density lipoprotein [LDL] cholesterol, high-density lipoprotein [HDL] cholesterol, and total cholesterol), CRP, and fibrinogen were measured. LDL cholesterol was calculated using the Friedewald formula. Insulin resistance was estimated by the insulin resistance index of homeostasis model assessment (HOMA-IR). Univariate regression analysis showed that CRP value was positively correlated to age (r = 0.218, P =.012), BMI (r = 0.239, P =.006), HOMA-IR (r = 0.397, P <.0001), triglycerides (r= 0.310, P <.005), LDL cholesterol (r= 0.179, P =.038), and fibrinogen (r = 0.371, P <.0001) levels and inversely correlated to HDL cholesterol (r = 0.174, P =.044) level in our diabetic patients. Multiple regression analysis showed that CRP was independently predicted by HOMA-IR (P<.0001, F = 11.6) and fibrinogen (P<.0001, F = 34.2), which explained 23.5% of the variability of CRP in our non-obese Japanese type 2 diabetic patients. These results indicate that insulin resistance and fibrinogen level are independent predictors of CRP in non-obese Japanese type 2 diabetic patients.

The aim of this study was to investigate the relationships between insulin resistance and regional abdominal fat area, body mass index (BMI), and serum lipid profile in nonobese Japanese type 2 diabetic patients. A total of 63 nonobese Japanese type 2 diabetic patients aged 45 to 83 years were examined. The duration of diabetes was 8.4 +/- 0.8 years. BMI, glycosylated hemoglobin (HbA(1c)) levels, and fasting concentrations of plasma glucose, serum lipids (total cholesterol, high-density lipoprotein [HDL] cholesterol, and triglycerides), and serum insulin were measured. The low-density lipoprotein (LDL) cholesterol level was calculated using the Friedewald formula (LDL cholesterol = total cholesterol - HDL cholesterol - 1/5 triglycerides). Insulin resistance was estimated by the homeostasis model assessment (HOMA-IR). Computed tomography (CT) was used to measure cross-sectional abdominal subcutaneous and visceral fat areas in all the patients. Adipose tissue areas were determined at the umbilical level. Subcutaneous and visceral abdominal fat areas were 136.5 +/- 6.0 and 86.0 +/- 4.1 cm(2), respectively. Univariate regression analysis showed that insulin resistance was positively correlated with subcutaneous (r =.544, P <.001) and visceral (r =.408, P =.001) fat areas, BMI (r =.324, P =.009), HbA(1c) (r =.254, P =.001), serum triglycerides (r =.419, P <.001), and serum LDL cholesterol (r =.290, P =.019) levels and was negatively correlated with serum HDL cholesterol level (r =.254, P =.041). Multiple regression analyses showed that insulin resistance was independently predicted by the areas of subcutaneous (F = 6.76, P <.001) and visceral (F = 4.61, P <.001) abdominal fat and serum triglycerides (F = 8.88, P <.001) level, which explained 36.9% of the variability of insulin resistance. Moreover, the present study demonstrated that whereas BMI was positively correlated with visceral (r =.510, P <.001) and subcutaneous (r =.553, P <.001) fat areas, serum triglyceride level was positively associated with visceral (r =.302, P =.015), but not with subcutaneous (r =.222, P =.074) fat area. From these results, it can be suggested that (1) both subcutaneous and visceral abdominal fat areas are independently associated with insulin resistance and (2) visceral fat area, but not the subcutaneous one, is associated with serum triglyceride levels in our nonobese Japanese type 2 diabetic patients.

This paper describes 30 mm regeneration of peripheral nerve axons along collagen filaments; 31-mm-long collagen filaments or collagen tube were grafted to bridge a 30-mm defect of rat sciatic nerve. The mean number and the diameter of regenerated myelinated axons were 330+/-227 and 2.7+/-0.9 microm at the distal end of the collagen-filaments 12 weeks postoperatively; while at the distal end of the tube no axon was found.

This article describes repair of peripheral nerve defect using collagen filaments instead of tubes. Many tube-shaped nerve guides induce regeneration of severed peripheral nerve axons within a limited distance. Substantial regeneration of nerve axons has not been reported without a tubular conduit. Here we show the regeneration of peripheral nerve axons along filaments of collagen without a tube. Cables of collagen filaments were grafted to repair 20-mm defects of rat sciatic nerves. Nerve autografts and collagen tubes were grafted as controls. The mean number and the mean fiber diameter of regenerated myelinated axons were approximately 4800 and 3.3 microm in the distal end of the nerve autograft at 8 weeks postoperatively while in the distal end of the collagen-filaments nerve guide, they were approximately 5500 and 2.3 microm. Collagen tubes failed to bridge the nerve defect. Histologic studies suggest that nerve axons regenerated substantially along the collagen filaments.

This paper describes the regeneration of severed peripheral nerve axons along collagen filaments without a tube. Two thousand collagen filaments were grafted to bridge 20 mm defects of rat sciatic nerve. The number of myelinated axons was approximately 4800 in the distal end of the nerve autograft at 8 weeks postoperatively; while in the collagen-filaments nerve guide it was 5500. The results suggested the collagen filaments guided regenerating axons effectively.