EFFICIENT ASSESSMENT OF PROGRESSION IN CLINICAL TRIALS IN RETINITIS PIGMENTOSA.

Abstract

To conduct a simulation study to evaluate the effect of disease stage, frequency, and clustering of visual field tests, and inclusion of one or both eyes on sample size calculation in clinical trials on retinitis pigmentosa. A series of VFs were simulated on the basis of test-retest visual field data in the early, moderate, and advanced stages of retinitis pigmentosa. Visual field measurements were scheduled 8 times in approximately 2 years. The probability to detect significant difference between treatment and control groups was measured. The investigation was conducted for 50%, 40%, 30%, 20%, and 10% treatment effects. When only one eye was used in each patient with a treatment effect of 50%, 80% probability of significance was observed in the moderate stage when sample size was 70 eyes in each arm (140 eyes in both arms). Early stage disease and inclusion of both eyes decreased this number to 30 eyes (60 eyes in both arms); these decreasing effects were larger than performing additional VFs at the beginning and at the end of the observation. A similar tendency was observed with 40%, 30%, 20%, and 10% treatment effects. When planning a clinical trial, it is important to consider disease stage, in addition to visual field schedule and inclusion of one or both eyes.