Background and methods: We performed a multicenter, randomized, unblinded clinical trial (the Foscarnet-Ganciclovir Cytomegalovirus Retinitis Trial) designed to compare ganciclovir with foscarnet in the treatment of cytomegalovirus retinitis in patients with the acquired immunodeficiency syndrome (AIDS). Of 234 patients, 127 were randomly assigned to ganciclovir and 107 to foscarnet; the study drugs were administered according to a common protocol at the 11 participating clinical centers. Antiretroviral therapy (with zidovudine, didanosine, or dideoxycytidine) was given as dictated by best medical judgment. The patients were followed for the progression of retinitis, visual loss, and death.
Results: Excess mortality in the ganciclovir group (as compared with the foscarnet group) led the Policy and Data Monitoring Board to recommend suspension of the treatment protocol 19 months after the trial started. As of that time, 65 of the patients assigned to ganciclovir had died, as compared with 36 of those assigned to foscarnet (51 percent vs. 34 percent, P = 0.007; relative risk, 1.79; 95 percent confidence interval, 1.17 to 2.73). The median survival was 8.5 months in the ganciclovir group and 12.6 months in the foscarnet group. Although the patients assigned to ganciclovir received less antiretroviral therapy on average than those assigned to foscarnet, the excess mortality could not be explained entirely by differences in exposure to antiretroviral drugs. In the forscarnet group, the only subgroup of patients identified as having excess mortality were those whose renal function was compromised at entry. There was no difference between the two treatment groups in the rate of progression of retinitis (relative risk, 0.95; P = 0.751).
Conclusions: These results suggest that for patients with AIDS, and cytomegalovirus retinitis, treatment with foscarnet offers a survival advantage over treatment with ganciclovir, although the patients may not tolerate foscarnet as well as ganciclovir.