Distribution density of intraepidermal nerve fibers in normal human skin.

A total of 74 specimens was obtained from the normal human skin of patients from 3 to 90 years old. The specimens were roughly classified into 5 groups: 15 for the face group from the face; 15 for the abdomen group from the abdomen; 13 for the back group from the back; 14 for the arm group from the upper arm and forearm; and 17 for the leg group from the thigh and lower leg. They were all fixed in 4% paraformaldehyde and 14% saturated picric acid. Cryostat sections were examined by the immunoperoxidase method and indirect immunofluorescence (IF). Primary antibodies against neurofilament, neuron-specific enolase, protein gene product 9.5 (PGP 9.5) and S-100 protein were used. The most effective method was found to be the combination of IF with PGP 9.5; it visualized the intraepidermal nerve fibers easily and clearly. Of the 74 specimens, 32 (43%) had intraepidermal PGP 9.5-immunoreactive (or nerve) fibers (IPIF), and 42 (57%) did not have any. With reference to the different skin locations, the maximal rate of specimens having IPIF was 57% in the arm group, and the minimum was 23% in the back group. IPIF positive specimens had approximate surface lengths of 6 mm, in which the existence number of the IPIF was 1 to 75. Their distribution density per 1000 epidermal basal cells was highest at 9.63 in the arm group and lowest at 2.89 in the back group. Their thickness was 2.94 +/- 0.83 microns with no significant differences among the five groups. We concluded that intraepidermal nerve fibers may not be distributed evenly in the hairy portions of normal human skin, but they may be present focally. Physiologically, two-point discrimination of itch may be explained by the distribution mode of intraepidermal nerve fibers.