This study proposes to assess the hypothesis that, in HIV-infected patients with treated and healed CMV retinitis, an increase in CD4+ T-cells after initiation of potent antiretroviral therapy is either directly related to, or a marker of, immunologic protection for CMV retinitis and is associated with a recovery in specific proliferation responses to CMV antigens.

In this study, 100 patients [AS PER AMENDMENT 7/2/99: 50 patients] with treated and healed, non-immediate sight-threatening CMV retinitis will discontinue maintenance therapy for suppression of CMV retinitis. Patients are studied in 2 groups. Patients enrolled in Group 1 have CD4+ counts greater than 100 cells/mm3. Group 2 patients have CD4+ counts of 50-100 cells/mm3 and a minimum of a 2 log10 decrease in plasma HIV-1 RNA level or plasma HIV-1 RNA levels below the limit of detection while receiving potent antiretroviral therapy for at least 8 weeks prior to entry [AS PER AMENDMENT 7/2/99: Group 2 has been withdrawn]. An additional 25 patients who meet eligibility requirements but who choose to continue to receive maintenance therapy may also participate. All patients are followed to evaluate the relationship between reactivation or progression of CMV disease and changes in CMV DNA, HIV-1 RNA, and CD4+ cell counts. Patients are seen at Weeks 2, 4, 6 and 8, and every 4 weeks until study closure or for 12 months after the last subject is enrolled. [AS PER AMENDMENT 12/24/98: Patients with confirmed moderate to severe "immune recovery vitritis" should receive a 3-week course of systemic steroids (oral prednisone recommended). Moderate immune recovery vitritis is defined as symptomatic decrease in visual acuity of 2 or more Snellen lines along with, in the absence of active CMV disease, either 2+ or greater vitreous haze as defined by Nussenblatt et al., or cystoid macular edema.] [AS PER AMENDMENT 7/2/99: During the course of the study in patients with confirmed cystoid macular edema and a concomitant reduction in visual activity below 20/40, both attributable to immune recovery vitritis/uveitis only, a 21-day course of oral prednisone is recommended. This initial course of steroids helps to determine whether there is an improvement in vision or a decrease in macular edema. Long-term management of immune recovery vitritis/uveitis may include intraocular injection of steroids. Ophthalmoscopic examinations and laboratory tests are performed as per protocol.]