Home > Clinical Alert

Clinical Alert

October 17, 1991

Clinical Alert To Physicians And Others Who Treat Patients With AIDS

Dear Colleague:

This letter is to inform you of findings from the Foscarnet-Ganciclovir CytomegalovirusRetinitis Trial. This randomized, multicenter clinical trial is supportedby the National Eye Institute, a component of the National Institutes of Health(NIH). It was conducted in collaboration with the AIDS Clinical Trials Groupsponsored by the National Institute of Allergy and Infectious Diseases atNIH.

The trial was designed to evaluate the relative efficacy and safety of foscarnetand ganciclovir for the initial treatment of cytomegalovirus (CMV) retinitisin patients with AIDS. On October 7, the trial’s independent Policy and DataMonitoring Board recommended suspending the protocol because the data indicatedthat patients treated with foscarnet lived on average four months longer thanthose treated with ganciclovir. The median survival for those treated withfoscarnet was approximately 12 months, compared to eight months for thosetreated with ganciclovir. The difference in survival could not be explainedby variations in disease severity at the time patients entered the study orto other chance factors.

Foscarnet and ganciclovir appeared to be equally effective in halting theprogression of CMV retinitis and preserving vision. While a survival benefitfor foscarnet was seen in most patients, in the group of patients who enteredthe study with a predicted creatinine clearance <1.2 ml/min/kg, a survivalbenefit was seen for ganciclovir (see the attached treatment administrationprotocol).

Within the trial, anti-retroviral therapy [zidovudine (AZT), ddI, and ddC]was used according to best medical judgment. The difference in mortality betweenthe foscarnet-treated patients and the ganciclovir-treated patients couldnot be fully explained by differential anti-retroviral use. However, the trialwas not designed to study possible interactions between anti-CMV and anti-HIVtreatments. Therefore, such an explanation cannot be ruled out.

These findings suggest that foscarnet may be the preferable initial treatmentfor CMV retinitis. A possible exception is the subgroup of patients with decreasedrenal function (predicted creatinine clearance <1.2 ml/min/kg) who appearedto do better on ganciclovir.

The complete findings from this study will be published soon. This announcementis being sent in advance of journal publication to physicians likely to treatpatients with AIDS who have CMV retinitis. A press conference will be heldon October 21 at the National Institutes of Health.


Carl Kupfer, M.D.
National Eye Institute

Douglas Jabs, M.D.
Study Chairman
Studies of the Ocular Complications of AIDS
Associate Professor of Ophthalmology and Medicine
Johns Hopkins University School of Medicine