Analyzing the impact of AZT on the progression of HIV infection: Proportional or constant delay over three years of treatment

The optimum time to begin antiretroviral therapy (ART) in HIV infected individuals is controversial. Using a Markov chain approach, the authors have re-analyzed several apparently inconsistent studies designed to measure the effect of early AZT on the progression of HIV disease. We find that, for at least up to three years, early AZT therapy is consistent with increasing the time to AIDS by a multiplicative factor (rate change) rather than a fixed delay. The negative judgments of early AZT, based on results reported by Hamilton et al. for survival and the Concorde study group in regard to both AIDS and death, may be owing to comparison with the null hypothesis that treatment has no effect rather than with the rate-change hypothesis. These results can lead to important policy decisions as early treatment might enhance treatment efficacy by decreasing the progression hazard by a factor of up to 1.7, especially if multi-drug ART is used. Multi-drug ART is needed for long-term therapy (greater than three years) because of the development of resistant viruses.