Surprisingly, male gender was associated with larger treatment effects, but this association may be a consequence of the presence of confounding variables. Most HIV-infected men in high-income settings are men who have sex with men, have longer histories of exposure

to antiretroviral drugs, and thus have fewer active drugs in their OBT regimens. The association between male gender and treatment outcome is probably confounded by GSS. In fact, when we adjusted our results for GSS, this association was no longer significant (data not shown). Our study used indirect comparison to demonstrate that the use of CCR5 inhibitors was not associated with higher increases in CD4 cell counts. This result contradicts the meta-analysis of Wilkin et al. which showed www.selleckchem.com/products/obeticholic-acid.html greater CD4 cell count increases among CCR5 inhibitor users at week 24, even when controlling for degree of virological suppression [14]. Wilkin et al. Small molecule library supplier used a multivariate linear regression model to evaluate predictors of CD4 cell count gains. In their analysis, each clinical trial arm was assigned a single data point. Our analysis also used a meta-regression model, but we included both clinical

trial arms as a single data point and considered the difference in CD4 gains between arms. Our analysis probably accounted for potential confounding variables more accurately. Nevertheless, we acknowledge that our findings are observational, and therefore vulnerable to bias. Baseline patient characteristics were heterogeneous in both treatment and placebo groups, with large

variations in the proportion of patients with AIDS, the median CD4 cell count, the median HIV RNA level and the OBT regimen GSS. We could not adjust our results for these differences. Even if we had done so, we would only have been able to adjust for information aggregated at the trial level. Moreover, Nintedanib (BIBF 1120) our results cannot be extrapolated to immunological nonresponders, who have weak immunological responses despite virological suppression [33], or to treatment-naïve patients initiating cART at very low CD4 cell counts. However, two recent studies that assessed immunological responses to adding maraviroc to existing cART regimens among patients with undetectable HIV RNA and CD4 counts ≤250 cells/μL did not find significant CD4 count improvements at week 24 [34,35]. Our systematic review demonstrates that including new antiretroviral drugs in cART regimens improves outcomes among treatment-experienced patients. This review also demonstrates that the most important predictive factor for achieving undetectable HIV RNA or higher CD4 cell count increases is the number of fully active drugs included in the regimen.