1. ^*Trescowthick Research Laboratories, Peter MacCallum Cancer Institute, Melbourne, Australia
2. ^†Department of Pathology, Mercy Hospital for Women, East Melbourne, Australia
3. ^‡Statistical Center, Peter MacCallum Cancer Institute, Melbourne, Australia

Correspondence: Dr Robert Ramsay, Group Leader Differentiation and Transcription Laboratory, Trescowthick Research Laboratories, Peter MacCallum Cancer Institute, Locked Bag #1, A' Beckett Street, Melbourne, Victoria 8006, Australia. Email: r.ramsay@pmci.unimelb.edu.au

Received 22 January 2002; Revised 23 June 2002; Accepted 31 July 2002.

Abstract

Human papillomavirus is a risk factor for vulvar cancer, whereas human papillomavirus-negative late onset vulvar carcinoma is associated with the dermatologic condition, lichen sclerosus. Human papillomavirus E6 protein targets TP53 for degradation and by inference it has been assumed that human papillomavirus-negative vulvar cancer is dependent upon the acquisition of p53 somatic mutations and subsequent allelic loss. To investigate this, TP53 expression, loss of heterozygosity, and p53 genomic sequence were examined in 29 cases of human papillomavirus-negative vulvar carcinoma with adjacent lichen sclerosus. We examined 37 cases of lichen sclerosus without vulvar carcinoma, 10 cases of nongenital lichen sclerosus, and 12 cases of normal vulvar epithelium served as controls. TP53 was evident in 72% of vulvar carcinoma, 48% in epithelium adjacent to vulvar carcinoma, but was minimal in normal samples. When lichen sclerosus cases were selected to exclude samples with absolutely no TP53 expression through probable failed antigen retrieval or homozygous p53 loss the number of epithelial cells expressing TP53 increased progressively from nongenital lichen sclerosus to lichen sclerosus without vulvar carcinoma, then to lichen sclerosus with vulvar carcinoma (p<0.0001). These data suggest elevated TP53 is a feature of vulvar lichen sclerosus. Seventy-four percent of vulvar carcinoma had chromosome 17p-linked loss of heterozygosity, whereas 47% of adjacent lichen sclerosus featured loss of heterozygosity, but only 31% of vulvar carcinoma had p53 mutations, a frequency less than reported previously. Seven percent of adjacent lichen sclerosus had mutations, showing for the first time the presence of an identical mutation to the matched vulvar carcinoma. These data, however, implicate p53 mutations as a later event in vulvar carcinoma and in marked contrast to the original expectation, our loss of heterozygosity data are consistent with loss of another locus (not p53) on 17p operating as a tumor suppressor in lichen sclerosus destined to develop vulvar carcinoma.