Objective: To examine the independent association of discordant virologic and immunologic responses to highly active antiretroviral therapy (HAART) with mortality.
Methods: A population-based study of 1527 treatment-naive individuals initiating HAART used Cox proportional hazards modeling to determine the independent association of treatment response at 3 to 9 months with nonaccidental mortality. Logistic regression was used to examine associations with discordant responses.
Results: Viral load (VL)/CD4 discordant responses were seen in 235 (15.4%) of subjects, and VL/CD4 responses were seen in 179 (11.7%) of subjects. In adjusted Cox regression models, discordant responses were found to be independently associated with an increased risk of mortality (VL/CD4: relative hazard [RH] = 1.87, 95% confidence interval [CI]: 1.15 to 3.04; VL/CD4: RH = 2.47, 95% CI: 1.54 to 3.95). VL/CD4 discordance was found to be associated with increasing age, baseline HIV RNA load <100,000 copies/mL, baseline CD4 counts <50 cells/muL, the use of lamivudine (3TC)/zidovudine (ZDV), and poor adherence to therapy. VL/CD4 discordance was associated with younger age; injection drug use; baseline HIV RNA load >100,000 copies/mL; the use of 3TC/ZDV, didanosine (ddI)/3TC, or ddI/stavudine; and poor adherence to therapy.
Conclusion: Discordant responses are independently associated with an increased risk of mortality and are, in turn, associated with poor adherence to therapy.