Limiting Cumulative HIV Viremia Copy-Years by Early Treatment Reduces Risk of AIDS and Death

Abstract

Background: Viremia copy-years (VCY), a time-updated measure of cumulative HIV exposure, predicts AIDS/death; although its utility in deciding when to start combination antiretroviral therapy (cART) remains unclear. We aimed to assess the impact of initiating versus deferring cART on risk of AIDS/death by levels of VCY both independent of and within CD4 cell count strata ≥500 cells per cubic millimeter.

Methods: Using Concerted Action on Seroconversion to AIDS and Death in Europe (CASCADE) data, we created a series of nested "trials" corresponding to consecutive months for individuals ≥16 years at seroconversion after 1995 who were cART-naive and AIDS-free. Pooling across all trials, time to AIDS/death by CD4, and VCY strata was compared in those initiating vs. deferring cART using Cox models adjusted for: country, sex, risk group, seroconversion year, age, time since last HIV-RNA, and current CD4, VCY, HIV-RNA, and mean number of previous CD4/HIV-RNA measurements/year.

Results: Of 9353 individuals, 5312 (57%) initiated cART and 486 (5%) acquired AIDS/died. Pooling CD4 strata, risk of AIDS/death associated with initiating vs. deferring cART reduced as VCY increased. In patients with high CD4 cell counts, ≥500 cells per cubic millimeter, there was a trend for a greater reduction for those initiating vs. deferring with increasing VCY (P = 0.09), with the largest benefit in the VCY ≥100,000 copy-years/mL group [hazard ratio (95% CI) = 0.41 (0.19 to 0.87)].

Conclusions: For individuals with CD4 ≥500 cells per cubic millimeter, limiting the cumulative HIV burden to <100,000 copy-years/mL through cART may reduce the risk of AIDS/death.

FIGURE 1.

A diagram of the “trials” construction. Individuals are assessed for eligibility at the beginning of each month (respective trial baseline). Each eligible individual is classified as having initiated or deferred cART in the baseline month. Time is measured from the beginning of the following month until AIDS, death, or censoring for each eligible individual (excluding any with an outcome during the baseline month). Cox proportional hazards models are used to assess the effect of initiating compared with deferring cART on time to AIDS/death, pooled across all trials.

FIGURE 2.

The effect of initiating compared with deferring cART on time to AIDS/death by VCY and CD4 cell count modeled continuously with 3 knot splines using the CASCADE data set.