K65R development among subtype C HIV-1-infected patients in tenofovir DF clinical trials

Abstract

A recent study demonstrated a greater propensity for HIV-1 subtype C to develop a K65R mutation under in-vitro selection with tenofovir compared with subtype B or other non-B subtypes [1]. The mechanistic basis for this in-vitro observation was not clear because the single nucleotide change is identical for the K65R substitution in either subtype B or subtype C. For subtype B, a switch from AAA to AGA occurs. For subtype C, a switch from AAG to AGG occurs. The third codon position, however, is different for subtype C, reflecting redundancy in the genetic code for lysine and a natural variation among different HIV-1 subtypes. The authors speculate that this third codon position difference or other genetic changes may influence the propensity for the development of K65R in subtype C HIV-1. We evaluated patients with subtype C, subtype B, and other non-B HIV-1 subtypes for virological failure and the development of resistance in two phase III clinical trials of tenofovir disoproxil fumarate (DF). These studies enrolled patients primarily from the United States, Europe and South America. Within these studies, approximately 7% of the 1200 patients enrolled were infected with non-B HIV-1 subtypes. Study 903 assessed the combination of tenofovir DF with lamivudine and efavirenz, and study 934 assessed the combination of tenofovir DF with emtricitabine and efavirenz. Although neither study was statistically powered to address efficacy in patients with non-B subtypes, the virological failure rates were similar between patients with subtype B or non-B HIV-1 subtypes (15.5 versus 19%, respectively, P = 0.75 for study 903 at week 144; 19 versus 13%, respectively, P = 1.0 for study 934 at week 48) [2,3]. Moreover, among the 10 patients with subtype C HIV-1 treated with tenofovir DF in both studies, only one patient was classified as a virological failure, and this patient did not develop a K65R mutation or any other resistance mutations. One patient developed K65R with a non-B subtype, a subtype AG patient from study 903. This patient did not show the AAG polymorphism at codon 65, but rather the AAA K65 codon typically found among subtype B patients. In summary, in highly suppressive antiretroviral regimens that include tenofovir DF, efavirenz and either emtricitabine or lamivudine, there was no evidence of a higher rate of virological failure or the development of K65R among patients with non-B HIV-1 subtypes or in particular among patients with subtype C HIV-1. The studies were limited by the number of patients enrolled with non-B subtypes. A larger study of tenofovir DF with zidovudine and lamivudine in African patients infected with subtypes A, C and D, however, concluded similarly that patients with subtypes A, C and D respond effectively to that drug combination with a low incidence of the development of K65R (three out of 20 virological failures; 15%) [4]. Another study [5], conducted exclusively among subtype C-infected patients in Botswana, demonstrated a higher rate of K65R development among patients treated with didanosine, stavudine, and efavirenz (seven out of 23 virological failures; 30%). The rate of thymidine analogue mutation development was also high in that trial, with eight out of 13 (62%) virological failure patients who were taking combivir developing thymidine analogue mutations, suggesting an overall higher rate of resistance development in that study. Overall, additional studies with the prospective enrollment of patients with subtype C and other comparator HIV-1 subtypes will be necessary to rule out any differential propensity for the development of K65R among subtype C patients in the clinical setting. Currently available clinical data, however, do not suggest a higher rate of K65R development among patients with subtype C.