May 27, 1996

Duration of the Survival Benefit of Zidovudine Therapy in HIV Infection

Arch Intern Med. 1996;156(10):1073-1077. doi:10.1001/archinte.1996.00040041073005

Background:  Zidovudine therapy improves survival in advanced human immunodeficiency virus (HIV) infection and delays progression from earlier stages to advanced stage of HIV disease. The duration of the benefit of zidovudine therapy, however, may be limited.

Objective:  To quantitate the duration of the survival benefit of zidovudine therapy in a heterogeneous patient population receiving care for HIV infection in an urban clinic.

Methods:  We analyzed data from 393 HIV-infected patients with CD4+ cell counts of 0.5 × 109/L (500 cells/μL) or less who first presented for care at The Johns Hopkins HIV Clinic, Baltimore, Md, from July 1989 through December 1993. Follow-up extended to a maximum of 3 years (median, 2 years). Survival probabilities in patients who received and who did not receive zidovudine therapy were analyzed by Kaplan-Meier methods and by multivariate Cox proportional hazards regression analysis adjusting for both time-dependent and fixed prognostic covariates.

Results:  Adjusting for baseline differences in CD4+ cell count, clinical stage of HIV disease, and prophylaxis for Pneumocystis carinii pneumonia, Cox regression analysis showed a significant effect of zidovudine compared with no treatment on the risk of dying during the first year of therapy (relative hazard for death, 0.32; 95% confidence interval [CI], 0.18 to 0.59). However, analysis of the time-dependent effect of zidovudine therapy showed that there was a diminishing relative hazard between treatment and no treatment of 0.75 (95% CI, 0.45 to 1.26) at 1 to 2 years of therapy and a relative hazard of 1.61 beyond 2 years (95% CI, 0.70 to 3.71).

Conclusion:  The survival advantage of zidovudine therapy is time dependent, lasting between 1 and 2 years in patients with CD4+ cell counts of 0.5 × 109/L or less. Alternative antiretroviral treatment may be indicated at that time.(Arch Intern Med. 1996;156:1073-1077)