Plasma HIV viral load in patients with hemophilia and late-stage HIV disease: A measure of current immune suppression

Background: For patients infected with HIV, plasma HIV viral load in early disease predicts long-term prognosis. However, the implications of viral lo ad measurements late in HIV disease are uncertain. Objective: To evaluate the relation between plasma HIV viral load and subse quent risk for disease progression in patients with late-stage HIV disease. Design: Retrospective cohort study. Setting: 16 treatment centers for patients with hemophilia. Patients: 389 patients with hemophilia and late-stage HIV disease (CD4 coun t < 200 cells/mm(3)). Measurements: Plasma HIV viral load was measured at baseline. Patients were followed for AIDS-related illnesses (primary outcome) and, specifically, P neumocystis carinii pneumonia (secondary outcome). Results: HIV viral load strongly predicted AIDS-related illness. For patien ts with viral loads less than 4.00 log(10) copies/mL, the 1-year actuarial risk was 0% and the 5-year risk was 25%. For patients with viral loads of a t least 6.00 log(10) copies/mL, the 1-year actuarial risk was 42% and the 5 -year risk was 78%. A linear relation existed between viral load a nd risk for AIDS-related illness (hazard ratio, 2.37 per log(10) copies/mL; P < 0.0 01). In addition, viral load most strongly predicted risk for illness immed iately after viral load testing; this predictive relation attenuated over t ime (P = 0.002). These findings changed little after adjustment for CD4 cel l counts that were updated during follow-up. In the first year after viral load was measured, it predicted occurrence of P, carinii pneumonia (hazard ratio, 4.69 per log(10) copies/mL; P < 0.001). Conclusions: In patients with hemophilia and late-stage HIV disease, viral load predicts disease progression independently of CD4 cell counts. Because viral load most strongly predicts progression immediately after load is me asured, it seems to reflect the current level of immunosuppression.