Low-dose intravitreal cidofovir (HPMPC) therapy of cytomegalovirus retinitis in patients with acquired immune deficiency syndrome

Purpose: The authors have shown that long-term treatment of cytomegalovirus (CMV) retinitis with 20-mu g intravitreal injections of cidofovir (HPMPC) is highly effective but may be associated with iritis and profound hypotony. They evaluated the efficacy and safety of 10-mu g intravitreal injections of cidofovir and made comparisons with their findings of 20-mu g injections. Methods: The current study was conducted as a nonrandomized consecutive case series at the AIDS Ocular Research Unit of the University of California at San Diego. Twenty-seven eyes of 18 patients were injected with 10 mu g intravitreal cidofovir and had complete follow-up. These were compared with another consecutive series of 24 eyes of 17 patients injected with 20 Ixg of cidofovir. Main Outcome Measures: The main outcome in this study was the incidence of failure to respond to treatment with 10-mu g injections. The authors also compared the time to progression of CMV retinitis after the initial intravitreal injections of 10 mu g and 20 mu g of cidofovir. Secondary outcomes included incidence of iritis and changes in intraocular pressure (IOP) after cidofovir injections. Results: The median time to retinitis progression was 45 days after a single intravitreal injection of 10 mu g cidofovir compared with 55 days with the authors' series of 20-mu g injections. This difference was statistically significant (P = 0.033, log-rank test) and appeared to be due principally to a 26% incidence of primary failure in the 10-mu g group (progression greater than or equal to 750 mu m within 28 days, P = 0.0017 Wilcoxon test). Progression after a second injection of 10 mu g cidofovir was more rapid (32 days, P = 0.037). The incidence of iritis after 10-mu g injections was 2.2% compared with 23% with 20-mu g injections (P = 0.003, Fisher's exact test, two-tailed). There was less decrease in IOP between the baseline injection and subsequent visits in the 10-mu g group. Conclusions: Treatment of CMV retinitis with 10-mu g intravitreal cidofovir injection was not as effective as with 20 mu g and may allow development of drug resistance, but there were fewer side effects with the 10-mu g dose. The drug appears to have a narrow therapeutic index, and other attempts at reducing the side effects while preserving the long-acting effect, such as liposome delivery, may be warranted.