Virological outcomes and genotypic resistance on dolutegravir-based antiretroviral therapy versus standard of care in children and adolescents: a secondary analysis of the ODYSSEY trial.

Background: ODYSSEY showed superior efficacy for dolutegravir-based antiretroviral therapy (ART) versus standard of care (SOC) in children living with HIV starting first-line or second-line ART aged 4 weeks or older. Here, we aim to compare virological outcomes and resistance in the dolutegravir group versus SOC for first-line and second-line ART up to 96 weeks.

Methods: ODYSSEY was an open-label, multicentre, randomised, non-inferiority trial done in 29 centres in seven countries (Germany, Spain, South Africa, Thailand, the UK, Uganda, and Zimbabwe). ODYSSEY recruited children living with HIV aged at least 28 days and younger than 18 years, weighing at least 3 kg, starting first-line ART (ODYSSEY A), or switching to second-line therapy after treatment failure (ODYSSEY B). Children were randomly assigned (1:1) to dolutegravir plus two nucleoside or nucleotide reverse transcriptase inhibitors (NRTIs; dolutegravir group) versus the SOC group (non-nucleoside reverse transcriptase inhibitor [NNRTI], boosted protease inhibitor, or non-dolutegravir integrase strand-transfer inhibitor, plus two NRTIs). Two randomised cohorts were combined in this exploratory analysis: children weighing at least 14 kg were enrolled between Sept 20, 2016, and June 22, 2018, and children weighing less than 14 kg were enrolled between July 5, 2018, and Aug 26, 2019. Virological failure was defined as an inadequate virological response at week 24 with an ART switch or confirmed HIV-1 RNA viral load of at least 400 copies per mL after week 36. Virological suppression was defined as two consecutive viral loads of less than 400 copies per mL and was compared between groups, including an ART switch and death as competing risks. Children with virological failure were tested for post-failure genotypic resistance, with baseline results used to identify emergent resistance. Development of emergent resistance was a secondary trial outcome and all other outcomes are exploratory. ODYSSEY was registered with ClinicalTrials.gov (NCT02259127), EUDRACT (2014-002632-14), and ISRCTN (ISRCTN91737921).

Findings: In ODYSSEY at enrolment, 381 participants started first-line ART (ODYSSEY A: 189 in the dolutegravir group and 192 in the SOC group) and 407 participants started second-line ART (ODYSSEY B: 202 in the dolutegravir group and 205 in the SOC group). 72 participants in ODYSSEY A and 13 participants in ODYSSEY B weighed less than 14 kg. 401 (51%) of 788 participants were female and 387 (49%) were male. Virological suppression occurred significantly earlier in the dolutegravir group (adjusted [cause-specific] hazard ratio [HR] 1·57 [95% CI 1·35 to 1·83]; p<0·0001). Overall, 51 (13%) participants had virological failure by 96 weeks in the dolutegravir group versus 86 (22%) in the SOC group (including 18 [10%] vs 43 [22%] in ODYSSEY A and in 33 [16%] vs 43 [21%] in ODYSSEY B; adjusted HR 0·56 [0·40 to 0·79]; p=0·0011). Among ODYSSEY B participants starting dolutegravir, virological failure was higher in children starting zidovudine (HR 2·22 [1·01 to 4·88]; p=0·048) and similar in those starting tenofovir disoproxil fumarate (1·19 [0·50 to 2·83]; p=0·70) compared with abacavir. Time to virological suppression was marginally faster in participants receiving second-line dolutegravir and abacavir with high-level abacavir resistance at baseline compared with those with no, low-level, intermediate-level resistance (cause-specific HR 1·70 [1·01 to 2·85]; p=0·046); and failure rates by week 96 were similar (HR 0·90 [0·23 to 3·61]; p=0·88). An estimated 1% (95% CI 0 to 2) of participants in the dolutegravir group versus 20% (14 to 26) in the SOC group in ODYSSEY A had emergent resistance to at least one drug-class within their first-line regimen (risk difference -20% [-25 to -14]; p<0·0001); 4% (1 to 6) versus 5% (2 to 8) had resistance to drug within their initial second-line regimen (risk difference -1% [-5 to 3]; p=0·60). 3% (0 to 5) of participants in the dolutegravir group had emergent integrase strand-transfer inhibitors resistance compared with 3% (1 to 6) of participants in the SOC group who had emergent resistance to the anchor drug (risk difference 0% [-4 to 3]; p=0·78).

Interpretation: Dolutegravir led to faster virological suppression and lower risk of virological failure than NNRTIs and boosted protease inhibitor-based SOC. Participants starting second-line dolutegravir-based ART with an abacavir or tenofovir backbone were at lower risk of virological failure than those starting zidovudine. During first-line therapy, dolutegravir protected against emergent resistance; starting second-line therapy, the risk of emergent resistance to nucleoside reverse transcriptase inhibitor backbone, and anchor drugs, was similar among participants starting dolutegravir within their second-line regimen and those starting mainly boosted protease inhibitor-based SOC.

Funding: Penta Foundation, ViiV Healthcare, and UK Medical Research Council.