Dolutegravir Monotherapy as Maintenance Strategy: A Meta-Analysis of Individual Participant Data From Randomized Controlled Trials

Background Dolutegravir monotherapy (DTG-m) results in virological failure (VF) in some people with human immunodeficiency virus (PWH). We sought to identify the independent factors associated with the risk of VF and to explore the effect size heterogeneity between subgroups of PWH enrolled in DTG-m trials. Methods We searched for randomized clinical trials (RCTs) evaluating DTG-m versus combined antiretroviral therapy (cART) among PWH virologically controlled for at least 6 months on cART. We performed an individual participant data meta-analysis of VF risk factors and quantified their explained heterogeneity in random-effect models. Definition of VF was a confirmed plasma human immunodeficiency virus (HIV)-1 ribonucleic acid (RNA) >50 copies/mL by week 48. Results Among 416 PWH from 4 RCTs, DTG-m significantly increased the risk of VF (16 of 227 [7%] versus 0 of 189 for cART; risk difference 7%; 95% confidence interval [CI], 1%-2%; P = .02; I$^{2}$ = 51%). Among 272 participants exposed to DTG-m, VF were more likely in participants with the following: first cART initiated ≥90 days from HIV acute infection (adjusted hazard ratio [aHR], 5.16; 95% 95% CI, 1.60-16.65), CD4 T cells nadir <350/mm$^{3}$ (aHR, 12.10; 95% CI, 3.92-37.40), HIV RNA signal at baseline (aHR, 4.84; 95% CI, 3.68-6.38), and HIV-deoxyribonucleic acid (DNA) copy number at baseline ≥2.7 log/10$^{6}$ peripheral blood mononuclear cells (aHR, 3.81; 95% CI, 1.99-7.30). Among these independent risk factors, the largest effect size heterogeneity was found between HIV DNA subgroups (I$^{2}$ = 80.2%; P for interaction = .02). Conclusions Our study supports the importance of a large viral reservoir size for explaining DTG-m simplification strategy failure. Further studies are needed to link size and genetic diversity of the HIV-1 reservoir.