BACKGROUND
Dolutegravir in second-line antiretroviral therapy (ART) is more effective with recycled tenofovir than switching to zidovudine. However, dolutegravir resistance is more frequent in second-line compared to first-line ART.
OBJECTIVES
We report long-term virologic outcomes from a clinical trial.
METHOD
AntiRetroviral Therapy In Second-line: investigating Tenofovir-lamivudine-dolutegravir (ARTIST) was a randomised, double-blind, phase II clinical trial. Eligible participants had two consecutive HIV-1 RNA ≥ 1000 copies/mL on first-line ART, mostly tenofovir-emtricitabine-efavirenz. Participants were switched to tenofovir-lamivudine-dolutegravir (TLD) with lead-in 50 mg dolutegravir twice daily in stage one (n = 62), and randomised to TLD with additional lead-in 50 mg dolutegravir or placebo for the first 14 days in stage two (n = 130). We present results up to 158 weeks, combining stages one and two.
RESULTS
We enrolled 192 participants: 127/176 (72%) had resistance (Stanford score ≥ 15) to both tenofovir and lamivudine. At week 48, 151/186 (81%; 95% confidence interval [CI] 75%, 87%) had HIV-1 RNA < 50 copies/mL. Of 127 participants with follow-up through week 158, 78% (95% CI 70%, 85%) maintained HIV-1 RNA < 50 copies/mL, 11% had HIV-1 RNA 50–999 copies/mL, and 11% had HIV-1 RNA ≥ 1000 copies/mL. Twenty-nine participants met criteria for resistance testing: one developed intermediate-level dolutegravir resistance (G118R mutation) at week 96, and one had high-level dolutegravir resistance (E138K, G118R, G163R, T66A mutations) detected at week 146.
CONCLUSION
Among adults switching to TLD with detectable HIV-1 RNA and substantial tenofovir and lamivudine resistance, a high proportion maintained virologic suppression up to 158 weeks. Emergent dolutegravir resistance occurred in ~1% of participants after 2–3 years on second-line TLD.
Tenofovir diphosphate (TFV-DP) concentration in dried blood spots is a marker of long-term adherence. We investigated the relationship between TFV-DP concentrations and virological outcomes in participants initiating tenofovir–lamivudine–dolutegravir (TLD) as first-line or second-line antiretroviral therapy.
SETTING
Three primary care clinics in Khayelitsha, Cape Town, South Africa.
METHODS
We conducted a post hoc analysis of 2 randomized controlled trials of participants initiating TLD. TFV-DP concentrations and viral loads were measured at 12, 24, and 48 weeks. Multivariable logistic regression was performed to assess the association with virological suppression (<50 copies/mL) per natural logarithm increase in TFV-DP concentration. Generalized estimating equations with logit link were used to assess associations with virological rebound. The Akaike Information Criterion and Quasi-likelihood Information Criteria were used to compare models built on continuous TFV-DP data to 4 previously defined concentration categories.
RESULTS
We included 294 participants in the analysis, 188 (64%) of whom initiated TLD as second-line therapy. Adjusted odds ratios (95% CIs) of virological suppression were 2.12 (1.23, 3.75), 3.11 (1.84, 5.65), and 4.69 (2.81, 8.68) per natural logarithm increase in TFV-DP concentration at weeks 12, 24, and 48, respectively. In participants with virological suppression at week 12, the adjusted odds ratio for remaining virologically suppressed was 3.63 (95% CI: 2.21 to 5.69) per natural logarithm increase in TFV-DP concentration. Models using continuous TFV-DP data had lower Akaike Information Criterion and Quasi-likelihood Information Criteria values than those using categorical data for predicting virological outcomes.
CONCLUSION
TFV-DP concentrations in dried blood spots exhibit a dose–response relationship with viral load. Analyzing TFV-DP concentrations as continuous variables rather than conventional categorization may be appropriate.