AEGiS-15IAC: Estimating the optimal threshold for initiating antiretroviral therapy in HIV disease.

15th International AIDS Conference

Bangkok, Thailand - July 11-16, 2004

Estimating the optimal threshold for initiating antiretroviral therapy in HIV disease.

Int Conf AIDS 2004 Jul 11-16; 15:(abstract no. MoOrB1079)

Braithwaite RS, Roberts MS, Chang CY, Justice AC
University of Pittsburgh, Pittsburgh, United States

BACKGROUND: The optimal clinical threshold for initiating therapy in HIV patients is unclear. Our purpose was to estimate the CD4 initiation threshold resulting in longest survival.

METHODS: We have created a probabilistic computer simulation that has predicts time to treatment failure and survival for HIV patients in the current treatment era. It has been calibrated on a large observational cohort and validated in a separate population. It represents genotypic resistance and nonadherence, the two main factors underlying treatment failure. We simulated newly diagnosed anti-retroviral naïve patients with CD4 counts of 500 and varying ages and viral loads, and predicted mean treatment duration and life expectancy. We explored initiation thresholds of 500, 350, and 200. All simulations were repeated 10,000 times. We performed sensitivity analyses varying assumptions regarding adherence, anti-retroviral toxicity, and number of HIV mutations at baseline.

RESULTS: Among 40 year old patients with baseline viral loads of 100,000, starting anti-retroviral therapies at a CD4 count of 350 rather than 500 decreased mean survival by 0.2 years and decreased mean treatment duration by 0.7 years. Starting at a CD4 count of 200 rather than 350 decreased survival by 1.0 years and decreased treatment duration by 1.5 years. Sensitivity analyses suggested that these benefits would be reduced or eliminated for many subgroups of patients. In particular, survival benefits from earlier treatment were reduced greatly if patients were older, less adherent, more prone to drug toxicity, or had higher viral loads. Survival benefits from earlier treatment were eliminated if patients were assumed to have highly mutated strains of HIV at the start of therapy. [table: see text] Conclusion Earlier anti-retroviral initiation may prolong survival for some patient groups, but benefits may be attenuated or eliminated by numerous patient factors. Sensitivity analyses from computer simulations may suggest patient subgroups to target for prospective data collection.

Keywords: AEGIS, HIV Infections, CD4 Lymphocyte Count, Viral Load, Antiretroviral Therapy, Highly Active, Acquired Immunodeficiency Syndrome, HIV Seropositivity, Treatment Failure, Drug Therapy, Combination, Humans, therapy, drug therapy

040711
MoOrB1079