I read with interest the recent report of two cases of lichenoid chronic graft-vs-host disease in a dermatomal distribution reported by Freemer et al1 in the January 1994 issue of the ARCHIVES. I disagree, however, with the authors' conclusion that these cases are likely to represent an association with herpes zoster virus infection.
It is generally believed that herpes zoster is the result of antedromic migration of varicella-zoster virus along sensory fibers, with deposition into the skin at the sites of nerve endings.2 Due to the arborization of peripheral nerves, this produces cutaneous lesions that are morphologically round but grouped along linear dermatomes. The photographs and description of lesions in this report (particularly those of case 1) suggest that these patients had confluent linear lesions, not round lesions in linear distribution.
I believe that the lesions in these patients have a distribution that corresponds quite well to Blaschko's