Changes in outcome of persons initiating highly active antiretroviral therapy at a CD4 count less than 50 cells/mm3

BACKGROUND: Although HIV treatment guidelines recommend highly active antiretroviral therapy (HAART) initiation before reaching a CD4 count of 200 cells/mm, many people in resource-rich settings, and a substantial proportion in resource-limited settings, present at levels <50 cells/mm. METHODS: Using UK Collaborative HIV Cohort data, we assessed virologic response to HAART for antiretroviral-naive persons initiating therapy at a CD4 count <50 cells/mm. We also investigated changes in the probability of having a viral level <400 copies/mL at 48 weeks over calendar time adjusting for gender, age, exposure category, ethnicity, baseline CD4 count and viral load, and whether the regimen contained a protease inhibitor. RESULTS: At 12, 24, 36, and 48 weeks, 80%, 83%, 85%, and 83% of participants, respectively, had a viral level <400 copies/mL. This proportion rose from 1997 to 1998, falling slightly in the most recent calendar period. By far the most important predictor of virologic suppression was calendar year of starting HAART (odds ratio [OR] = 2.49, 4.28, and 3.28 for 1999 to 2000, 2001 to 2002, and 2003 to 2005, respectively, compared with 1997 to 1998). Women were more likely to have a viral level <400 copies/mL at week 48 compared with men (OR = 1.74, 95% confidence interval [CI]: 1.07 to 3.02), as were older individuals (OR = 1.46, 95% CI: 1.11 to 1.96 for every 10 years older). There was marginal or no evidence that other factors were associated with outcome. The estimated corresponding probabilities of achieving a viral level <50 copies/mL at week 48 were 71%, 75%, and 79% for a woman aged 25, 35, and 45 years, respectively, initiating HAART in the most recent calendar period. The respective probabilities for a man at those ages were 68%, 73%, and 78%. CONCLUSIONS: These data, albeit under conditions of good infrastructure for care delivery, are a useful comparator for other populations starting therapy at similar levels of immunodeficiency and may be valuable for evaluating the success of antiretroviral therapy rollout programs.