• Prognosis of HIV-1 Infected Drug-Naive Patients Starting Potent Antiretroviral Therapy: Multi-Cohort Analysis of 12,040 Patients
  Abstract LB-18
  Presented by M. Egger, ART Cohort Collaboration (ART-CC); University of Bristol, Bristol, United Kingdom
  View the original abstract

This is a fascinating study from the United Kingdom presented in a late-breaker session. This study tries to answer the question: what is the probability that an HIV-naive patient starting HAART will develop an AIDS-related event or death?

The investigators collected retrospective information from an extremely large group of patients (12,040) from different cohorts across the U.S., Europe, and Canada. Several characteristics were identified such as gender, risk factors, disease stage, initial treatments used, baseline CD4 and HIV-RNA, and CD4 and HIV-RNA at six months after starting therapy.

Investigators reported 1,045 AIDS events and 344 deaths within 24,310 cumulative years of follow-up. Ten potential prognosis factors were identified: age (<50 or >50), risk (IVDA or no IVDA), CDC class of disease stage (A/B or C), CD4 counts (5 strata) and baseline HIV RNA (<5 logs or >5 logs).

As expected, the probability of an event-free survival was higher with higher CD4 counts and lower with lower HIV RNAs. But using a multivariate analysis, investigators were able to propose that CD4 count is the most dominant prognostic factor, followed by a baseline HIV RNA log >5. Also predictive were age >50, history of intravenous drug use (IDU) and stage C of disease. The initial drug regimen did not appear to affect prognosis.

In this study, the CD4 count had a very important prognostic role, but, interestingly, the HIV RNA was not predictive unless it was greater than 5 logs. The stratification of risk factors presented was quite clear and convincing. The study also had an extremely high volume of patients powering the statistical analysis.

This is another study that presents data to support the concept of starting antiretroviral therapy late rather than early. Although the current DHHS guidelines reflect some of that tendency in delaying treatment, this study, in my opinion, suggests that it may be OK to go even further.

On the other hand, the fact that this study was well-powered and statistically well-done can give us a "risky" sense of comfort that it is OK to further delay therapy. We need to consider that CD4 counts are highly variable and what we may perceive today as >200 may indeed be only 150 cells. There are many variables in the population that were not considered in the analysis that could potentially change the results presented. Among these variables are the multiplicity of treatments, change in antiretroviral therapy, treatment interruptions, compliance, and prevention of opportunistic infections.

The results in this study differ somewhat from the data presented by Mellors (Mellors, et al. Science 1996; 272:1167) in the risk of AIDS-defining illness at CD4 counts between 200-350. The differences between these two study results suggest, at least to me, how subjective the interpretation of the weight that we put in those variables can be. Nevertheless, this study gives us another tool that we can use with our patients when discussing prognosis and the right time for initiation of antiretroviral therapy.