PURPOSE: To compare the efficacy of two regimens of intravitreous fomivirsen for treatment of cytomegalovirus (CMV) retinitis that had reactivated or was persistently active despite other anti-CMV therapies in patients with acquired immunodeficiency syndrome (AIDS).

DESIGN: Randomized treatment intervention clinical trials.

METHODS: Two multicenter prospective randomized open-label controlled clinical trials (USA/Brazilian study; EuroCanadian study) each compared two treatment regimens of intravitreous fomivirsen (330 μg/injection), a more intense schedule (regimen A) consisting of 3 weekly injections as induction, followed by injections every 2 weeks as maintenance therapy, and a less intense schedule (regimen B), consisting of two injections at day 1 and day 15 as induction, followed by injections every 4 weeks as maintenance therapy. A total of 61 patients (67 eyes) were treated with regimen A and a total of 32 patients (39 eyes) were treated with regimen B. Time to progression was determined by masked evaluation of serial fundus photographs. The proportion of patients whose lesions became less active or inactive is described for the USA/Brazilian study as an additional measure of efficacy.

RESULTS: Results were similar for both regimens. In the USA/Brazilian study, the median time to progression was 106 days for regimen A and 267 days for regimen B (P = .2179, Wilcoxon rank sum test). In the EuroCanadian study, the median time to progression was not determinable for regimen A (25th percentile: 91 days), but was 403 days for regimen B. In the USA/Brazilian study, lesion activity decreased in 80% of patients and became inactive in 55% of patients during therapy.

CONCLUSIONS: Both regimens of intravitreous fomivirsen can be effective treatment for patients with CMV retinitis that has not been controlled by other anti-CMV drugs. The less intense regimen may be preferable in terms of convenience and safety.