We thank Tong et al.  for their interest in the AIDS Clinical Trials Group study on the pharmacogenetics of plasma efavirenz exposure after treatment discontinuation . It is well established that, among white Americans and African Americans, the CYP2B6 516G→T (rs3745274) polymorphism is associated with delayed plasma clearance and increased plasma exposure with efavirenz [3, 4]. Limited data suggest a similar association with nevirapine . However, a causal relationship between this polymorphism and reduced CYP2B6 functional activity has not yet been convincingly demonstrated. The observation that the frequency of this variant may be increased among ethnic Chinese persons contributes to our understanding of the potential global impact of this polymorphism. A major impetus for pharmacogenomic research is the potential to predict responses to antiretroviral drugs, even among populations that were not represented in registrational trials. On the basis of the report by Tong and colleagues, we anticipate that the distribution of plasma area under the 24-h efavirenz concentration time curve (AUC0–24 h) values may be higher among ethnic Chinese than among other populations. Formal pharmacokinetic analyses involving ethnic Chinese persons are needed to confirm this prediction. However, the implications of such altered efavirenz pharmacokinetic profiles are presently uncertain. Although we previously reported an association between CYP2B6 516G→T and increased adverse effects in the CNS during the first week of efavirenz therapy, this association was not apparent during subsequent weeks, despite persistently elevated plasma efavirenz concentrations [3, 6]. Furthermore, CYP2B6 516G→T was not associated with differences in long-term virological response in a separate study involving 340 efavirenz recipients . Only in the context of a multilocus interaction with ABCB1 (also called MDR1) 2677G→T (rs2032582) did CYP2B6 516G→T help predict virological failure .
The AIDS pandemic is arguably one of the greatest infectious disease challenges in history. Without a cure on the horizon, tens of millions of individuals worldwide will soon require lifelong multidrug antiretroviral therapy. Importantly, these drugs will increasingly be prescribed for populations in which frequencies of human genetic polymorphisms that influence treatment response vary considerably and differ from populations for which optimal dosages, efficacy, and safety profiles were characterized (mostly individuals of European ancestry). Despite the relatively limited number and complexity of HIV pharmacogenomic investigations to date, many associations have already been described between human genetic variants, antiretroviral pharmacokinetics, and/or treatment response (some associations must still be validated) [8–28]. Many more associations are certain to be identified in the coming years. It is noteworthy that for each reported association, the implicated genetic variant differs considerably in frequency among racial and ethnic populations. As knowledge in HIV pharmacogenomics continues to increase in scope and precision, it is our view that information from well-designed and validated pharmacogenetic analyses should be used to help predict likely patterns of antiretroviral treatment response among different populations worldwide. Using this knowledge to optimize antiretroviral efficacy while minimizing toxicity could have substantial public health benefits, particularly in resource-limited countries.
Potential conflicts of interest. D.W.H. has received research grants from Bristol Myers Squibb, Boehringer-Ingelheim, Tanox, GlaxoSmithKline, and Gilead Sciences, and is a consultant for Boehringer-Ingelheim. H.J.R. and E.P.A.: no conflicts.