Comparative efficacy and safety of dolutegravir relative to common core agents in treatment-naïve patients infected with HIV-1: a systematic review and network meta-analysis

Abstract

Background: Network meta-analyses (NMAs) provide comparative treatment effects estimates in the absence of head-to-head randomized controlled trials (RCTs). This NMA compared the efficacy and safety of dolutegravir (DTG) with other recommended or commonly used core antiretroviral agents.

Methods: A systematic review identified phase 3/4 RCTs in treatment-naïve patients with HIV-1 receiving core agents: ritonavir-boosted protease inhibitors (PIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs), or integrase strand inhibitors (INSTIs). Efficacy (virologic suppression [VS], CD4⁺ cell count change from baseline) and safety (adverse events [AEs], discontinuations, discontinuation due to AEs, lipid changes) were analyzed at Week 48 using Bayesian NMA methodology, which allowed calculation of probabilistic results. Subgroup analyses were conducted for VS (baseline viral load [VL] ≤/> 100,000copies/mL, ≤/> 500,000copies/mL; baseline CD4⁺ ≤/>200cells/μL). Results were adjusted for the nucleoside/nucleotide reverse transcriptase inhibitors (NRTI) combined with the core agent (except subgroup analyses).

Results: The NMA included 36 studies; 2 additional studies were included in subgroup analyses only. Odds of achieving VS with DTG were statistically superior to PIs (odds ratios [ORs] 1.78-2.59) and NNRTIs (ORs 1.51-1.86), and similar but numerically higher than other INSTIs. CD4⁺ count increase was significantly greater with DTG than PIs (difference: 23.63-31.47 cells/μL) and efavirenz (difference: 34.54 cells/μL), and similar to other core agents. INSTIs were more likely to result in patients achieving VS versus PIs (probability: 76-100%) and NNRTIs (probability: 50-100%), and a greater CD4⁺ count increase versus PIs (probability: 72-100%) and NNRTIs (probability: 60-100%). DTG was more likely to result in patients achieving VS (probability: 94-100%), and a greater CD4⁺ count increase (probability: 53-100%) versus other core agents, including INSTIs (probability: 94-97% and 53-93%, respectively). Safety outcomes with DTG were generally similar to other core agents. In patients with baseline VL > 100,000copies/mL or ≤ 200 CD4⁺cells/μL (18 studies), odds of achieving VS with DTG were superior or similar to other core agents.

Conclusion: INSTI core agents had superior efficacy and similar safety to PIs and NNRTIs at Week 48 in treatment-naïve patients with HIV-1, with DTG being among the most efficacious, including in patients with baseline VL > 100,000copies/mL or ≤ 200 CD4⁺cells/μL, who can be difficult to treat.

Keywords: Antiretroviral therapy; Dolutegravir; HIV-1; Integrase strand inhibitors; Network meta-analysis; Non-nucleoside reverse transcriptase inhibitor; Protease inhibitor; Systematic review; Treatment-naïve.

Fig. 1

PRISMA flowchart of systemic literature review update. CADTH, Canadian Agency for Drugs and Technologies in Health; CSR, clinical study report; FDA, Food and Drug Administration; NCT, National Institute of Health clinical trial results published on ClinicalTrials.gov; NMA, network meta-analysis; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; Pubs, published articles; RCT, randomized controlled trial; SLR, systemic literature review; TE, treatment experienced; TN, treatment naïve

Fig. 2

NMA Network of treatment comparisons for efficacy outcomes contained within the identified clinical studies. The major classes of agents analyzed in the selected trials for NMA are indicated along the perimeter of the figure: NNRTIs in orange, INSTIs in green, PIs in purple and connectors in blue. Black lines connecting each of the treatments of interest (red dots) represent a publication or clinical trial containing those 2 agents of interest. A connector is a treatment not of interest that is compared with at least two different treatments of interest that are included in the analysis to provide additional data. ATV, atazanavir; ATV/r, ritonavir-boosted atazanavir; BIC, bictegravir; DTG, dolutegravir; DRV/r, ritonavir-boosted darunavir; EFV, efavirenz; EVG/c, cobicistat-boosted elvitegravir; FPV/r, ritonavir-boosted fosamprenavir; INSTIs, Integrase strand inhibitors; LPV/r, lopinavir-boosted ritonavir; NFV, nelfinavir; NMA, network meta-analysis; NNRTIs, non-nucleoside reverse transcriptase inhibitors; PIs, protease inhibitors; RAL, raltegravir; RPV, rilpivirine; SQV/r, ritonavir-boosted saquinavir; VS, virologic suppression

Fig. 3

a VS and b median CD4⁺ CFB at Week 48 with DTG versus comparators. FE model. *Indicates treatment comparisons are significantly different. ATV/r, ritonavir-boosted atazanavir; BIC, bictegravir; CFB, change from baseline; Crl, credible interval; DTG, dolutegravir; DRV/r, ritonavir-boosted darunavir; EFV, efavirenz; EVG/c, cobicistat-boosted elvitegravir; FE, fixed effect; INSTIs, Integrase strand inhibitors; LPV/r, lopinavir-boosted ritonavir; NNRTIs, non-nucleoside reverse transcriptase inhibitors; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; OR, odds ratio; PIs, protease inhibitors; RAL, raltegravir; RPV, rilpivirine; VS, virologic suppression

Fig. 4

(a) VS and (b) CD4⁺ CFB at Week 48 with DTG versus comparators

Fig. 5

a AEs, b discontinuations, and c discontinuations due to AEs with DTG versus comparators. RE model. *Indicates treatment comparisons are significantly different. AEs, adverse effects; ATV/r, ritonavir-boosted atazanavir; BIC, bictegravir; Crl, credible interval; DTG, dolutegravir; DRV/r, ritonavir-boosted darunavir; EFV, efavirenz; EVG/c, cobicistat-boosted elvitegravir; INSTIs, Integrase strand inhibitors; LPV/r, lopinavir-boosted ritonavir; NNRTIs, non-nucleoside reverse transcriptase inhibitors; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; OR, odds ratio; PIs, protease inhibitors; RAL, raltegravir; RE, random effect; RPV, rilpivirine

Fig. 6

a TC, b HDL, c LDL, and d TG CFB with DTG versus comparators. RE model. *Indicates treatment comparisons are significantly different. ATV/r, ritonavir-boosted atazanavir; BIC, bictegravir; CFB, change from baseline; Crl, credible interval; DTG, dolutegravir; DRV/r, ritonavir-boosted darunavir; EFV, efavirenz; EVG/c, cobicistat-boosted elvitegravir; HDL, high-density lipoprotein; INSTIs, integrase strand inhibitors; LDL, low-density lipoprotein; LPV/r, lopinavir-boosted ritonavir; NNRTI, non-nucleoside reverse transcriptase inhibitors; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; OR, odds ratio; PIs, protease inhibitors; RAL, raltegravir; RE, random effect; RPV, rilpivirine; TC, total cholesterol; TG, triglycerides

Fig. 7

VS at Week 48 by baseline VL (a) and CD4⁺ (b) with DTG versus comparators. NRTI-unadjusted, FE model. *Indicates treatment comparisons are significantly different. ^†RPV is not indicated for patients with a VL > 100,000 copies/mL. Studies that reported data on CD4⁺ subgroups used a mix of cutoff inequalities (</≥, ≤/> and </> 200 cells/μL); the convention ≤/> 200 cells/μL used here reflects those used in the majority of studies. ATV/r, ritonavir-boosted atazanavir; BIC, bictegravir; Crl, credible interval; DTG, dolutegravir; DRV/r, ritonavir-boosted darunavir; EFV, efavirenz; EVG/c, cobicistat-boosted elvitegravir; FE, fixed effect; INSTIs, Integrase strand inhibitors; LPV/r, lopinavir-boosted ritonavir; NNRTIs, non-nucleoside reverse transcriptase inhibitors; NRTI, nucleoside/nucleotide reverse transcriptase inhibitor; OR, odds ratio; PIs, protease inhibitors; RAL, raltegravir; RPV, rilpivirine; VL, viral load; VS, virologic suppression