To test the hypothesis that epidermal rather than dermal mosaicism det
ermines Blaschko's lines in hypomelanosis of Ito (HI), we studied the
distribution of chromosomal mosaicism in four patients. In two, mosaic
ism had not been detected in lymphocytes or dermal fibroblasts, but wa
s clearly shown in epidermal keratinocytes; furthermore, the abnormal
cell line was confined to the hypopigmented epidermis and the normal e
pidermis contained only normal cells. Negative findings in the other t
wo patients might be because of mosaicism which was undetected either
because it was submicroscopic or because it was present in melanocytes
, which have not yet been studied. These preliminary results support t
he ideas that (1) Blaschko's lines represent single clones of epiderma
l cells; (2) in patients with HI and severe neurological involvement m
osaicism, if detectable, is best shown in keratinocytes; and (3) the c
ytogenetic defect in epidermal cells may be directly responsible for t
he failure of pigmentation in HI.