Abstract Body

A randomized study showed higher efficacy of maintenance dual therapy (DT) with protease inhibitors(PI)/ritonavir(r) + 3TC despite the presence of M184V. Aim of this study was to compare virological efficacy of DT in pts with a history of M184V detection (M184V+) and without (M184V-).

We retrospectively analyzed HIV+ pts with HIV-RNA50 copies/ml switching to DT and with at least one previous genotype in the Italian ARCA database. Pts were followed from baseline (BL, starting of DT) until discontinuation or virological failure (VF). The primary endpoints were: VF [viral load (VL)>50 cps/mL in 2 consecutive determinations or one 200 cps/mL] and virological blips (VB) (VL 51-199 cp/mL not confirmed). M184V was assessed in the historical genotypic resistance tests (hGRT).

436 pts were included, 45(10%) switched to 3TC with lopinavir/r, 106(24%) with atazanavir/r, 155 (36%) with darunavir/r, 126(29%) with dolutegravir,4(1%) with raltegravir. 87/436 pts were M184V+ according to hGRT, 85/436 according to last GRT. The M184V+ group more frequently included females, IDU, had older age, lower CD4+ at nadir, longer duration of ART and of viral suppression. Median follow-up was 1.2 years (ys) (IQR 0.6;2.4). VF occurred in 17/369 M184V- pts and in 7/87 M184V+ pts. The 3-year probability of remaining free from VF was 91.9% (95% CI 86.6;97.2) in M184V- and 87.8% (78.4;97.2) in M184V+ (log rank p=0.323). In a multivariate model adjusting for M184V, class of 2nd drug and duration of viral suppression only zenith VL (aHR 2.07 per 1-log higher, p=0.034) was independently associated with VF. In pts without VF, VB occurred in 18/332 (5%) M184V- and 10/80 (13%) M184V+ pts. The 3-y probability of not having VB was 90.1% (84;96.2) in M184V- and 79.8% (67.8;91.8) in M184V+ (p=0.016). At multivariate analysis, only M184V resulted to be predictor of VB (aHR 2.84, p=0.040). Similar results were found with different definitions of VF and VB. Selecting pts with viral suppression 6.6ys (median of M184V+ group) (n=308), the 3-y probability of remaining free from VF was 92.5% in M184V- and 82.9% in M184V+ (p=0.080): zenith VL was the only predictor of VF; VB were 13/239 and 8/37, respectively, and the probability of remaining free of VB was 91.1% – and 69.4% (p<0.001).

M184V did not seem to be associated to higher risk of VF with 3TC+PI/r or DTG, but with a higher probability of VB. The shorter time of viral suppression appears to increase the risk of VF and of VB in those with M184V.