Search

The World Health Organization recommends dolutegravir with two nucleoside reverse-transcriptase inhibitors (NRTIs) for second-line treatment of human immunodeficiency virus type 1 (HIV-1) infection. Evidence is limited for the efficacy of this regimen when NRTIs are predicted to lack activity because of drug resistance, as well as for the recommended switch of an NRTI from tenofovir to zidovudine.In a two-by-two factorial, open-label, noninferiority trial, we randomly assigned patients for whom first-line therapy was failing (HIV-1 viral load, ≥1000 copies per milliliter) to receive dolutegravir or ritonavir-boosted darunavir and to receive tenofovir or zidovudine; all patients received lamivudine. The primary outcome was a week 48 viral load of less than 400 copies per milliliter, assessed with the Food and Drug Administration snapshot algorithm (noninferiority margin for the between-group difference in the percentage of patients with the primary outcome, 12 percentage points).We enrolled 464 patients at seven sub-Saharan African sites. A week 48 viral load of less than 400 copies per milliliter was observed in 90.2% of the patients in the dolutegravir group (212 of 235) and in 91.7% of those in the darunavir group (210 of 229) (difference, -1.5 percentage points; 95% confidence interval [CI], -6.7 to 3.7; P = 0.58; indicating noninferiority of dolutegravir, without superiority) and in 92.3% of the patients in the tenofovir group (215 of 233) and in 89.6% of those in the zidovudine group (207 of 231) (difference, 2.7 percentage points; 95% CI, -2.6 to 7.9; P = 0.32; indicating noninferiority of tenofovir, without superiority). In the subgroup of patients with no NRTIs that were predicted to have activity, a viral load of less than 400 copies per milliliter was observed in more than 90% of the patients in the dolutegravir group and the darunavir group. The incidence of adverse events did not differ substantially between the groups in either factorial comparison.Dolutegravir in combination with NRTIs was effective in treating patients with HIV-1 infection, including those with extensive NRTI resistance in whom no NRTIs were predicted to have activity. Tenofovir was noninferior to zidovudine as second-line therapy. (Funded by Janssen; NADIA ClinicalTrials.gov number, NCT03988452.).

BACKGROUND: The efficacy and toxic effects of nucleoside reverse-transcriptase inhibitors (NRTIs) are uncertain when these agents are used with a protease inhibitor in second-line therapy for human immunodeficiency virus (HIV) infection in resource-limited settings. Removing the NRTIs or replacing them with raltegravir may provide a benefit. METHODS: In this open-label trial in sub-Saharan Africa, we randomly assigned 1277 adults and adolescents with HIV infection and first-line treatment failure to receive a ritonavir-boosted protease inhibitor (lopinavir-ritonavir) plus clinician-selected NRTIs (NRTI group, 426 patients), a protease inhibitor plus raltegravir in a superiority comparison (raltegravir group, 433 patients), or protease-inhibitor monotherapy after 12 weeks of induction therapy with raltegravir in a noninferiority comparison (monotherapy group, 418 patients). The primary composite end point, good HIV disease control, was defined as survival with no new World Health Organization stage 4 events, a CD4+ count of more than 250 cells per cubic millimeter, and a viral load of less than 10,000 copies per milliliter or 10,000 copies or more with no protease resistance mutations at week 96 and was analyzed with the use of imputation of data (≤4%). RESULTS: Good HIV disease control was achieved in 60% of the patients (mean, 255 patients) in the NRTI group, 64% of the patients (mean, 277) in the raltegravir group (P=0.21 for the comparison with the NRTI group; superiority of raltegravir not shown), and 55% of the patients (mean, 232) in the monotherapy group (noninferiority of monotherapy not shown, based on a 10-percentage-point margin). There was no significant difference in rates of grade 3 or 4 adverse events among the three groups (P=0.82). The viral load was less than 400 copies per milliliter in 86% of patients in the NRTI group, 86% in the raltegravir group (P=0.97), and 61% in the monotherapy group (P