Skin disease and malignancy. An epidemiological study.

The aim of this work was to develop a computer program (CANEST) to estimate the risk of cancer in patient populations and to use this program to investigate cancer risk associated with several dermatological disorders. Patients seen at the dermatology departments at the Karolinska Hospital and South Hospital were used for the study of chronic urticaria, condylomata acuminata, basal cell carcinoma, lichen planus and positive patch tests. The national Swedish In-Patient Register was used to find all patients hospitalized for dermatomyositis or polymyositis since 1964. From eleven large dermatological centers in Sweden, details of close to 5,000 PUVA-treated patients were obtained for study. The computer program CANEST was developed and used to calculate the expected number of malignant tumors in these patient populations, based on incidence data from the Swedish Cancer Register for the years 1958-1987. By matching the patients' records with the Cancer Register the actual number of cancers was obtained. Of 1,155 patients with chronic urticaria, a malignancy was diagnosed in 36, while the expected number was 41: clearly there is therefore no association between chronic urticaria and malignancy. In 3,260 patients with condylomata acuminata there was no increased risk of cancer in situ of the cervix (relative risk = 1.5; 95% confidence interval 0.9 to 2.5) and the number of genitourinary cancers in males was almost three times higher than expected (2.6; 1.2 to 5.0). These results indicate that the risk of developing cervical carcinoma in situ is less than previously thought, but the implications of the increase in genitourinary tumors in males are uncertain. Patients with basal cell carcinoma had an increased risk of malignancy in general. Melanoma risk was seven times greater in males (6.6; 3.0 to 12.5) after the basal cell carcinoma diagnosis. Risks of squamous cell carcinoma of the skin, lung cancer, thyroid cancer and cancer of the uterine cervix were also increased. No increased risk of cutaneous malignancy was found in 2,071 patients with lichen planus, but for oral cancer it was six times greater in males (5.9; 2.5 to 11.4). A slight general increase in malignancy risk was found in 2,183 males (1.3; 1.1 to 1.5) with positive patch tests, but not in 3,675 females. When individual sites were analyzed, cancers of the lung, larynx, uterine cervix and prostate were significantly increased. The implications of this are uncertain, but might indicate a common failure of the immune system which might predispose for both conditions, or be a marker of certain occupational exposure.(ABSTRACT TRUNCATED AT 400 WORDS)