Sensory neuropathy is a prominent component of diabetic neuropathy. It is not entirely clear how diabetes influences skin innervation, and whether these changes are correlated with clinical signs and laboratory findings. To investigate these issues, we performed skin biopsies on the distal leg of 38 consecutive type 2 diabetic patients with sensory symptoms in lower limbs (25 males and 13 females, aged 56.2 +/- 9.4 years) and analysed the correlations of intraepidermal nerve fibre (IENF) densities in skin with glycaemic status (duration of diabetes, HbA1C, and fasting and post-prandial glucose levels), and functional parameters of small fibres (warm and cold thresholds) and large fibres (vibratory threshold and parameters of nerve conduction studies). Clinically, 23 patients (60.5%) had signs of small-fibre impairment, and 19 patients (50.0%) had signs of large-fibre impairment. IENF densities were much lower in diabetic patients than in age- and gender-matched controls (1.794 +/- 2.120 versus 9.359 +/- 3.466 fibres/mm, P < 0.0001), and 81.6% (31/38) of diabetic patients had reduced IENF densities. IENF densities were negatively associated with the duration of diabetes (standardized coefficient: -0.422, P = 0.015) by analysis with a multivariate linear regression model. Abnormal results of functional examinations were present in 81.6% (warm threshold), 57.9% (cold threshold), 63.2% (vibratory threshold) and 49% (amplitude of sural sensory action potential) of diabetic patients. Among the three sensory thresholds, the warm threshold temperature had the highest correlation with IENF densities (standardized coefficient: -0.773, P < 0.0001). On nerve conduction studies in lower-limb nerves, there were abnormal responses in 54.1% of sural nerves, and 50.0% of peroneal nerves. Of neurophysiological parameters, the amplitude of the sural sensory action potential had the highest correlation with IENF density (standardized coefficient: 0.739, P < 0.0001). On clinical examination, 15 patients showed no sign of small-fibre impairment, but seven of these patients had reduced IENF densities. In conclusion, small-fibre sensory neuropathy presenting with reduced IENF densities and correlated elevation of warm thresholds is a major manifestation of type 2 diabetes. In addition, the extent of skin denervation increases with diabetic duration.

No abstract

To characterize the pathology of epidermal nerve degeneration and regeneration, we investigated temporal and spatial changes in skin innervation of the mouse footpad. Within 24 hours after sciatic nerve axotomy, terminals of epidermal nerves appeared swollen and there was a mild reduction in epidermal nerve density (5.7 +/- 2.8 vs 12.7 +/- 2.2 fibers/mm, p < 0.04). Epidermal nerves completely disappeared by 48 hours (0.2 +/- 0.2 vs 14.2 +/- 0.9 fibers/mm, p < 0.001). Concomitant with the disappearance of epidermal nerves, the immunocytochemical pattern of the subepidermal nerve plexus became fragmented. At the electron microscopic level, the axoplasm of degenerating dermal nerves was distended with organelles and later became amorphous. Beginning from day 28 after axotomy, collateral sprouts from the adjacent saphenous nerve territory extended into the denervated area with a beaded appearance. They never penetrated the epidermal-dermal junction to innervate the epidermis. In contrast, 3 months after nerve crushing, the epidermis on the surgery side resumed a normal innervation pattern as the epidermis on the control side (10.3 +/- 3.9 vs 10.6 +/- 1.5 fibers/mm, p = 0.1). This study demonstrates the characteristics of degenerating and regenerating nerves, and suggests that successful reinnervation mainly originates from regenerating nerves of the original nerve trunks. All these findings provide qualitative and quantitative information for interpreting the pathology of cutaneous nerves.

To generate an experimental neuropathy model in which small-diameter sensory nerves are specifically affected and to test a potential treatment, adult mice were given a single injection (50 microg/kg, i.p.) of the capsaicin analog resiniferatoxin (RTX). On Day 7 after RTX treatment, there was a 53% reduction in unmyelinated nerve density in the medial plantar nerve (p = 0.0067) and a 66% reduction in epidermal nerve density of hind paw skin (p = 0.0004) compared with vehicle-treated controls. Substance P-immunoreactive dorsal root ganglion neurons were also markedly depleted (p = 0.0001). These effects were associated with the functional deficit of prolonged withdrawal latencies to heat stimuli (p = 0.0007) on a hot plate test. The potential therapeutic effects of 4-methylcatechol (4MC) on this neuropathy were then tested by daily injections of 4MC (10 microg/kg, i.p.) from Days 7 to 35 after neuropathy induction. On Day 35, 4MC-treated mice had an increase in unmyelinated (p = 0.014) and epidermal nerve (p = 0.0013) densities and a reduction in thermal withdrawal latency (p = 0.0091) compared with RTX-only controls. These results indicate that 4MC promoted regeneration of unmyelinated nerves in experimental RTX-induced neuropathy and enhanced function.