Background: Poor virologic outcomes on antiretroviral therapy (ART) and HIV drug resistance pose the biggest threats to the success of ART scale-up for the individual and public health. We assessed virological response and emergence of drug resistance in a cluster randomised trial of universal test and treat strategy in South Africa comparing immediate ART initiation versus CD4-guided national guidelines.
Methods: HIV-positive individuals enrolled March 2012-June 2016 on ART ≥6 months were included and data from both trial arms were combined. Participants either transferred care from the public ART programme (Transferees; N=1389 or were new ART initiators within the ANRS 12249 trial (Initiators; N=983). A 1.3kb region of the pol gene was amplified, sequenced using the Sanger method, assembled in Geneious V9.1.5 and interpreted using the Stanford Drug Resistance algorithm. Chi-Squared or Mann-Whitney tests were used for comparisons as appropriate. Logistic regression was used to estimate odd ratios (OR) and their 95% confidence intervals (CI)
Results: 96% of Initiators were on fixed-dose combination of tenofovir+emtricitabine+efavirenz vs. 16% of Transferees. 6 months viral load (VL) was available for 908 Initiators; of whom 94% (851) achieved virologic suppression [VS] (VL< 400 copies/mL). Amongst Transferees, 80% had VS at their first trial clinic visit after a median ART duration of 4.0years (IQR 2.2-6.0). For both groups, the probability of VS increased with every 100 cells/mm3 increase in CD4 at initiation [adjusted OR=1.30 (95%CI=1.07-1.58) and 1.33 (1.14-1.55) for Initiators and Transferees respectively]
Of all 2372 individuals on ART ≥6 months, 227 (9.6%) experienced virologic failure [VF] (VL>1000 copies/mL ≥6 months on ART). Of 135 (59%) for which resistance testing was possible, 119 (88%) had ≥1 drug resistance mutation at VF: M184V/I (76%); K103N/S (56%); K65R (25%). The median plasma VL at VF was similar in individuals with and without the K65R mutation [35,504 copies/mL (IQR 4961-109,278) vs. 19,699 copies/mL (3,969-80,064); p=0.96, respectively].
Conclusions: Despite good levels of VS, we observed high levels of tenofovir resistance in individuals failing the recommended first-line ART. K65R did not confer a reduced infectivity (VL). Our data suggest that drug resistance may continue to emerge and spread during a test and treat strategy.

Download the e-Poster