The CD4 cell count and other surrogate markers of human immunodeficiency virus (HIV) infection possess enormous potential value for drug trials and for patient management. This usefulness, however, depends on our ability to reliably use surrogate markers to predict clinical outcomes. Current knowledge indicates that in persons with the acquired immunodeficiency syndrome (AIDS) or asymptomatic HIV infection, the beneficial effects of zidovudine-related clinical progression can be only partially explained by the drug's effect on CD4 cell count. Consequently, a drug's clinical efficacy cannot be confidently predicted solely by its effect on CD4 count. These results show the need for additional correlative studies to better understand whether and how markers can be used to predict clinical outcomes in comparative therapeutic trials, in the early approval of new anti-HIV drugs, and for the management of persons infected with HIV.
