We recommend therapy-naïve patients start combination Apoptosis Compound high throughput screening ART containing TDF and FTC as the NRTI backbone (1A). We suggest ABC and 3TC is an acceptable alternative NRTI backbone in therapy-naïve patients who, before starting ART, have a baseline VL≤100 000 copies/mL

(2A). ABC must not be used in patients who are HLA-B*57:01 positive (1A). Three RCTs have compared TDF-FTC with ABC-3TC as the NRTI backbone in combination with different third agents: ATV/r or EFV [2-6], EFV [7-9] and LPV/r [10]. Assessment of virological efficacy as a critical outcome was complicated by different definitions across the three studies. In our analysis for GRADE (see Appendix 3.1), there was no difference in rates of virological suppression at Protease Inhibitor Library chemical structure 48 weeks or 96 weeks but the analysis excluded the largest of

the three trials (ACTG 5202) and the quality of evidence for this outcome was assessed as low or very low. Assessment of the risk of protocol-defined virological failure at 48 weeks favoured TDF-FTC (RR 0.76, 95% CI 0.53–1.07); the effect was not statistically significant and heterogeneity in the analysis was relatively high (I2 46%). Assessment of protocol-defined virological failure at 96 weeks showed a significant difference favouring TDF-FTC (RR 0.73, 95% CI 0.59–0.92). Data were only available from one study [4] for this analysis; however, this was by far the largest of the three trials and the quality of evidence

assessment for this outcome was rated as high. The difference in virological failure was assessed by the Writing Group to be large enough to be above the clinical threshold for decision-making. The difference equates to a O-methylated flavonoid number needed to treat to prevent one case of virological failure of approximately 20 patients treated for 1 year. The results of ACTG 5202 [2-4] are complicated by early termination of those individuals with a baseline VL >100 000 copies/mL at the recommendation of the data and safety monitoring board due to significantly inferior performance in those subjects receiving ABC-3TC. No difference in virological efficacy between the TDF-FTC and ABC-3TC arms was seen in those in the lower VL stratum (baseline VL <100 000 copies/mL). The subsequent 96-week analysis, after discontinuation of those subjects in the higher VL stratum, may therefore underestimate the difference between the two backbones. HLA-B*57:01 screening was not routine in ACTG 5202 and this potentially may have influenced some of the safety endpoints, but appears not to have influenced the primary virological outcome. In the higher VL strata the number of patients with suspected hypersensitivity reactions was equal between both arms and virological failure in these patients was infrequent.