Faculty Bibliography

In a consecutive series of 1,067 patients entered into the data base of the Melanoma Cooperative Group at New York University School of Medicine between 1972 and 1980, 120 men were of draft age (18-31 years) during World War II (1941-1945). Questionnaires were sent to these 120 individuals; 89 responded. Simultaneously, a control (nonmelanoma) population of 65 men of similar age was queried. Each subject in both groups was asked whether he had served in the armed forces during World War II and, if so, what were his theaters of operation. Based on the response, 83% (74 of 89) of the melanoma group compared with 76% (49 of 65) of the control group had served in the armed forces during World War II; however, a significantly (p = 0.0002) greater percent of the melanoma patients (34%) served in the tropics than did the control subjects (6%). Further, overrepresented in the melanoma group that served in the tropics (compared with the melanoma group who served in the armed forces in nontropical theaters) were malignant melanomas that had their origin in nevocytic nevi. The findings suggest that Caucasian individuals heavily exposed to sunlight in the tropics for several years during early life may be at higher risk to the subsequent development of cutaneous malignant melanoma. In some individuals this may be a two-step phenomenon, in which the first step is the solar induction of nevocytic nevi and the second is malignant transformation within them

A multivariate analysis was performed on 20 clinical and histologic variables from 327 Stage I prospectively studied melanoma patients who underwent elective regional lymph node dissection (ERLD). Primary tumor thickness, microscopic satellites, and the elapsed interval between diagnosis and ERLD, were selected as the combination of variables that were most highly associated with clinically occult regional lymph node metastases (P = 10(-15), model chi-square). Microscopic satellites were defined as tumor nests, greater than 0.05 mm in diameter, in the reticular dermis, panniculus, or vessels beneath the principal invasive tumor mass but separated from it by normal tissue on the section in which the Breslow measurement was taken. The probability of finding nodal metastases for melanomas less than 0.75 mm thick was 0% (0/41 patients); for those 0.76-1.50 mm, 4% (4/108); 1.51-3.0 mm, 14% (14/102); and greater than 3.0 mm, 39.5% (30/76). Primary melanomas greater than 1.50 mm thick with microscopic satellites were more often associated with nodal metastases than those of similar thickness without satellites (30/57 (53%) versus 14/121 (12%), P = 0.01). Some satellites probably represent intraspecimen metastases, while others do not. Any predictive model for occult regional lymph node metastases based on data from ERLD done less than 50 days after diagnosis may underestimate the prevalence of metastases

A study of the influence of the anatomical location of malignant melanoma on the prognosis of 971 patients with stage I disease disclosed specific high-, intermediate-, and low-risk sites. High-risk sites included scalp, mandibular area, midline of trunk (anterior and posterior), upper medial thighs, hands, feet (except the arches), popliteal fossae, and genitalia. The life-table-adjusted five-year disease-free survival was 54% in the high-risk locations, 79% in intermediate-risk locations, and 93% in low-risk sites. A Cox proportional hazards analysis demonstrated that the grouping of lesions by their anatomical risk location had prognostic value that was significant in a model of eight other known predictive variables (thickness, sex, age, type, level, mitotic index, ulceration, and presence of preexistent nevus). The results indicate that anatomical location of the primary melanoma is significantly associated with five-year disease-free survival