Background: The optimal timing of highly active antiretroviral therapy (HAART) in human immunodeficiency virus (HIV)-infected patients with > or = 200 absolute CD4 lymphocytes/mm3 is unknown. CD4 lymphocyte percentage could add prognostic information.
Methods: Persons who initiated HAART between 1 January 1998 and 1 January 2003, received > or = 30 days of therapy, and had baseline CD4 lymphocyte data available were included in the study. The log-rank test for time to event and Cox proportional hazards models were used to determine predictors of a new acquired immunodeficiency syndrome-defining illness or death.
Results: A total of 788 patients met the inclusion criteria. At baseline, subjects had a median of 225 CD4 lymphocytes/mm3 and 17% CD4 lymphocytes. Subjects with < 17% CD4 lymphocytes had earlier disease progression, compared with subjects with > or = 17%, both in the entire cohort (P<.0001) and of those subjects with > 350 absolute CD4 lymphocytes/mm3 at baseline (P=.03). CD4 lymphocyte percentage < 17% was the strongest predictor of disease progression among subjects in this latter group (hazard ratio, 3.57; P=.045).
Conclusions: In this cohort, CD4 lymphocyte percentage predicted disease progression in HIV-infected subjects who initiated therapy with > 350 CD4 lymphocytes/mm3. This information may help identify persons who will derive the greatest benefit from initiation of HAART.