Background: Fenretinide, a synthetic derivative of retinoic acid, is under study in clinical trials for the prevention of breast, skin basal cell, bladder, and oral cancer in patients at risk. Although fenretinide is well tolerated even after prolonged use, it does lower plasma retinol levels and thus may affect night vision. Purpose: The purpose of this study was to measure changes in dark adaptation resulting from fenretinide administration, to compare the measured results with the patient's subjective perception, to define the association with plasma retinol levels, to assess the reversibility of alterations in night vision, and to assess the effects of fenretinide on the surface of the eye. Methods: The study involved 65 women who had been operated on for stage I breast cancer. Of the study group, 34 received 200 mg daily of fenretinide for a median of 32 months, while 31 control subjects did not. Dark adaptation was studied with the Goldmann-Weekers adaptometer and with a subjective questionnaire. Plasma retinol levels were measured at each test of dark adaptation. Effects of fenretinide on the ocular surface were evaluated through conjunctival impression cytology. Results: Of the patients on fenretinide, eight (23.5%) showed mild and nine (26.5%) showed moderate alterations of measured dark adaptability, compared with just two controls (6.5%) with mild alterations (cumulative odds ratio = 15.4; P = .0008). A significant inverse correlation exists between the final sensitivity threshold of dark-adaptometry and plasma retinol levels, with mild alterations arising below 16 mug/dL and moderate alterations below 10 mug/dL. Abnormal rod function improved significantly after 7 days and normalized 1 month after use of fenretinide was stopped or vitamin A supplementation was begun, while the conventional 3-day drug suspension, or drug half dose, did not allow sufficient recovery. Alterations of conjunctival cytology were slightly higher in patients receiving fenretinide, but no clinical disorders of the ocular surface were observed. Conclusions: The women treated with 200 mg fenretinide daily showed a relatively high incidence of mild-to-moderate alterations of dark-adaptometry as measured with the Goldmann-Weekers adaptometer. However, the real-life implication of the measurements is an open question, for the questionnaire shows that 50% of the patients with altered dark adaptometry were asymptomatic.
