Proflavine is a synthetic acridine dye which early in this century was found to have bacteriostatic and bacteriocidal properties when administered topically. During World War II it was widely used as a would antiseptic. With the advent of more specific and less toxic antibiotics, its clinical importance declined, until it was reintroduced recently in combination with ultraviolet light for the treatment of psoriasis and type-II herpesvirus infection.
A bioassay of the carcinogenicity of proflavine monohydrochloride hemihydrate was conducted using Fischer 344/CR rats and B6C3F1 mice. The compound was administered in the diet at concentrations of 300 and 600 ppm to groups of 50 rats for 109 weeks and at concentrations of 200 and 400 ppm to groups of 50 mice for 104 weeks. The animals were subjected to necropsy and histopathologic evaluation as they died or at the end of their periods of treatment.
Average weights attained by high-dose groups were consistently lower than those of control groups; weights of low-dose groups showed essentially no differences from those of the controls. Survival rates of the treated rats and mice did not differ from those of the controls except for a lower rate among the female mice.
Five malignant neoplasms of the intestinal tract consisting of three leiomyosarcomas of the small intestine, a sarcoma near the colon area, and an adenocarcinoma of the small intestine were observed in five of the high-dose male rats. None were observed in other treatment or control groups. If these five intestinal neoplasms are considered together, they aresignificant at the P=0.026 level using the Fisher exact test. A positive dose-related trend (P=0.034) was also present for the three leiomyosarcomas.
The observed incidence of hepatocellular carcinoma in female mice was 4/50 (8%) in the control group, 20/49 (41%) in the low-dose group, and 22/50 (44%) in the high-dose group. The test for dose-related trend showed a level of significance of P<0.001. In male mice, the observed incidence of hepatocellular carcinoma was 20/49 (41%) in the control group 28/49 (57%) in the low-dose group, and 30/50 (60%) in the high-dose group. The dose-related trend was significant at P=0.057, and the high dose was significant at P=0.044.
The unusually high incidence of hepatocellular carcinomas and hemangiosarcomas in control male mice and the unusually high incidence of malignant lymphomas in all groups of female mice in conjunction with the fact that a positive-control carcinogen was tested in the same room with these animals, raises a question of the validity of these bioassay results.