Abstract
From a consecutive, nationwide series of 219 females with primary vulvar malignant melanomas diagnosed in Sweden during 1960 to 1984 and followed up until 1994, we analyzed epidemiological, clinical, histopathological, prognostic and molecular genetic data. The age-standardized incidence among these patients, 75 % of whom were 60 years old or more, decreased by 3.2 % annually compared to an increase of almost 6 % for contemporary cutaneous melanomas in Sweden. Relative 5-and 10-year survivals were 47 % and 44 %, respectively.
The density of melanomas was roughly 2.5 times higher in the vulva than elsewhere; 46 % emerged in glabrous skin, 12 % in hairy skin, and 35 % extended in both. 57 % were mucosal lentiginous (MLM), 22 % nodular (NM), and only 2 % superficial spreading (SSM), an incidence reversing that of cutaneous melanomas. More than 94 % of the vulvar melanomas had a vertical growth phase; 27 % were macroscopically amelanotic, usually in the glabrous skin. Pre-existing nevocellular nevi were only in hairy skin and were adjacent to the SSM subtype, suggesting that glabrous skin melanomas emerge de novo.
Of malignant melanomas, in clinical Stage I patients, tumor thickness, ulceration and clinical/macroscopical amelanosis predicted short survival according to multivariate analyses that evaluated all clinical, histopathological and therapeutical data. The relative risk (RH) for short survival increased 5 times when those variables were combined.
Mutations in the Ras proto-oncogene family and in the TP53 gene of selected primary melanomas were traced in DNA extracted from tumor tissues dissected from histopathological archival materials. PCR/SSCP assays and nucleotide sequencing determined that 32 % of malignant melanomas from chronically sun-exposed skin, I I % from intermittently exposed and 7 % from unexposed skin displayed mutations in N-ras codon 6 1. The significant differences between those groups substantiate UV radiation as an inducer of N-ras mutation. 23 % of all melanomas from sun exposed and unexposed sites had TP53 mutations. C->T mutations at dipyrimidine sites, supposed to be typical of UV radiation -induced DNA damage, were found in melanomas from sun-exposed and unexposed areas, suggesting mutational factor(s) other than UV radiation.
The diversity of surgical methods was too great to allow adequate analysis. However, wide local resection did not seem to shorten the prognosis as to survival compared to radical vulvectomy.
We conclude that vulvar and cutaneous malignant melanomas differ markedly. Vulvar melanoma and its hairy and glabrous skin compartments are established here as a model for studying pathogenetic mechanisms of these tumors.
