Carcinogenesis and nicotine in malignant melanoma of the choroid.

The most abundant carcinogen in tobacco smoke or its condensate has been identified as N-nitrosonornicotine. This is formed predominantly from nicotine in tobacco during curing and smoking but may also be generated within the body. Intravenous injection of 14C NNN appears highly concentrated in the choroid of pigmented mice. Like nicotine, it crosses the placental barrier and appears in the choroid of the embryonic eye in pregnant female mice. Other studies of the male melanoma population show a trend toward earlier life appearance of malignant melanoma in smokers and significantly less of disease-free melanoma-bearing males at five years among smokers. Immunoglobulin G and natural killer leukocyte activities appear to be reduced in male smokers with or without melanomas. Although species and strain differences in susceptibility to carcinogenic agents are well known, these findings led us to a retrospective analysis of 126 white adults undergoing enucleation as immediate therapy for unilateral malignant melanoma of the choroid between 1958 and 1978. Histologic verification and adequate follow-ups were obtained in 105 patients. Minimum follow-up in 65 survivors was 3 years and maximum was 24 years. In 40 patients with metastatic melanomas, death occurred at 3 months to 13 years or at an average of 44 months following surgery. An adequate history of tobacco use was obtained in 94 patients, 44 of which were positive for tobacco use (32 male; 12 female) and 50 were denials (16 male; 34 female). Questions to female patients or families of deceased women evoked more prompt and uncompromising denial than was experienced in questioning of male patients. Straight (non-actuarial) five year tumor mortality in white males with choroidal malignant melanoma showed 12 of 32 or 37% dead among male smokers and 1 of 13 or 8% dead among male non-smokers. Actuarial calculation of five year survival showed only 62% in tobacco using males and 85% in non-tobacco using males. The distribution of cytology was essentially similar in groups both by sex and other subgroupings. There were, however, more large tumors (greater than 10 mm in diameter) in the tobacco group (61%) vs the non-smokers (31%). These preliminary findings merit further analysis by prospective tobacco histories in malignant melanoma patients, particularly among the male population.