Vulvar melanoma at the M. D. Anderson Cancer Center: 25 years later.

The purpose of this study was to review the clinical course of patients diagnosed with vulvar melanoma. Charts of patients diagnosed between 1970 and 1997 were reviewed for demographics, lesion characteristics, disease duration and extent, and treatments. Actuarial survival curves were computed by the Kaplan Meier method and compared by Cox proportional hazards regressions. Fifty-one patients (median age 54) with vulvar melanoma presented with a vulvar mass (39%), pain (30%), bleeding (24%), and itching (20%). Anatomical distribution was mucosa of the vulva (65%), vulvar epidermal site (21%), or unspecified vulva (14%), with 20% having multifocal disease at diagnosis. Histologic types were superficial spreading or nodular (50% each). Median lesion characteristics were diameter 2 cm, Breslow index 4.4 mm, and Clark level IV. Distribution of patients per American Joint Committee on Cancer (AJCC) stage was 29%, 50%, 16%, and 7% for stages I, II, III and IV, respectively. Inguinal node metastases were unilateral in 16% and bilateral in 7%. Despite complete surgical resection, 32 patients (63%) recurred. Median survival for all patients was 41 months (range, 5-324), with 91% 5-year survival for patients with stage I and 31% for stage >or= IIA (P = 0.0002). As with cutaneous melanoma, the AJCC classification, Breslow's thickness, and Clark's levels are the major predictors of overall survival (P = 0.0001 each) and disease-free survival (P <or= 0.0001, 0.0004, and 0.0002, respectively). Surgical techniques do not seem to alter the prognosis. Because vulvar melanoma carries a poor prognosis at early stages, the use of adjuvant therapies needs to be studied prospectively in women affected with this disease.