A morphological study of peripheral nerves in human and experimental diabetes mellitus : effect of various therapies
Light and electron microscopical techniques were employed to make qualitative and quantitative observations on peripheral nerve morphology. Sural nerve biopsies from 4 groups of diabetic patients with different syndromes of polyneuropathy and 6 non-diabetic controls were evaluated. Observations indicated that unequal rates of successful fibre regeneration may underlie an apparent difference in myelinated fibre (MF) loss between painful and painless diaebetic polyneuropathy. MF and unmyelinated fibre (UF) degeneration/regeneration per se are unlikely to be the cause of neuropathic pain in diabetic polyneuropathy since evidence of each was present in patients with remission from painful symptoms and in patients with painless neuropathy. Axonal atrophy may have a role in neuropathic pain generation. Abnormal endoneurial capillary morphology was manifest in all the diabetic patients. Alterations in capillary basement membrane thickness were interdependent with the type and extent of neuropathological change and the clinical expression of the neuropathy. Two patients presented with intractible vomiting due to diabetic gastropathy. The findings of a study of their anterior and posterior abdominal vagus nerves, obtained as biopsy specimens at vagotomy and gastroenterostomy to prevent stomal ulceration, were compared with those of 2 diabetic and 2 non-diabetic patients undergoing vagotomy for duodenal ulceration. Autovagotomy, and its causative role in gastropathy, was not supported since there was persistence of intact and regenerating fibres and evidence of similar MF and UF changes in both groups of diabetic patients. Male Sprague-Dawley rats were employed to make a comparison of the effects of conventional insulin therapy (CIT) and continuous subcutaneous insulin infusion therapy (CSII) in streptozotocin-induced diabetes. CSII was superior to CIT in diurnal and circadian blood glucose homeostasis and in ameliorating a diabetes-related deficit in body weight and skeletal length. Although CSII performed better, neither form of therapy was completely successful in reversing the effects of untreated-diabetes on parameters indexing MF morphology in the tibial nerve. Potentially serious degenerative changes in MFs following the initiation of either form of insulin therapy were found to correlate with the number of detected hypoglycaemic episodes. Normal maturation of MFs in the sural nerve involved the formation of a proximo-distal taper for fibre and axon size. MFs and axons in diabetic rats did not assume a size taper. The effect of diabetes was manifest proximally in the nerve with no evidence of selective involvement of axon or Schwann cell. Male albino Lewis rats were made diabetic with streptozotocin and used to study the effect of early (within 3 weeks= EIT) and delayed (after 6 months= DIT) pancreatic islet transplantation on MF morphology. Optimal diurnal and circadian blood glucose control was achieved for 6 months following EIT and DIT. A body weight deficit in diabetic rats was ameliorated by EIT but not DIT. EIT completely prevented the occurrence of abnormal fibre and axon morphology in the tibial nerves of diabetic rats. DIT completely reversed a deficit in fibre size but not axon size. If these observations were applicable to man the best therapeutic option would be for candidates to receive islet transplantation soon after the onset of diabetes. The occurrence of neuropathy might then be prevented.