Background
The WHO recommends dolutegravir (DTG) instead of ritonavir-boosted lopinavir (LPV/r) in second-line antiretroviral therapy (ART), and recycled tenofovir disoproxil fumarate (TDF)/emtricitabine or lamivudine (XTC) instead of zidovudine (AZT)/XTC as backbone. However, there is little long-term data from routine, low-resource settings, where DTG resistance may be emerging. We compared 24-month outcomes with these 3 regimens.
Methods
We analysed routinely collected, de-identified data from people with HIV at 125 public primary care clinics in KwaZulu-Natal, South Africa. We emulated a target trial among adults switching to second-line AZT/XTC/LPV/r, AZT/XTC/DTG, or TDF/XTC/DTG between December 2019 and February 2021 after viral failure with TDF/XTC/efavirenz. The co-primary endpoints were i) loss to follow-up (LTFU) including death through 24 months (time-to-event) and ii) viraemia >50 copies/mL at 24 months. We used logistic regression with inverse probability of treatment weights in intention-to-treat (ITT) analyses and with inverse probability of censoring weights in per protocol analyses, with bootstrapped confidence intervals (CIs).
Results
Of 2144 people, 917 (42.8%) switched to AZT/XTC/LPV/r, 343 (16.0%) to AZT/XTC/DTG, and 884 (41.2%) to TDF/XTC/DTG. Median age was 36 years (IQR 30-43), 1245 (58.1%) were female and median time with viraemia ≥1000 copies/mL was 371 days (IQR 235-588). ART changed during follow-up for 24.2%, 24.5%, and 15.4% of those switched to AZT/XTC/LPV/r, AZT/XTC/DTG, and TDF/XTC/DTG, respectively. In ITT analysis of LTFU (N=2144), the standardised risk was higher with AZT/XTC/LPV/r (33.5%; adjusted risk difference [aRD] 6.7%, 95% CI 0.2 to 12.3; strong evidence) and with TDF/XTC/DTG (32.4%; aRD -5.7%, 95% CI -11.3 to 1.4%; weak evidence) than with AZT/XTC/DTG (26.7%), and similar between AZT/XTC/LPV/r and TDF/XTC/DTG. The standardised risk of viraemia in those retained in care with a VL result (N=1190) was higher with AZT/XTC/LPV/r (48.0%) than AZT/XTC/DTG (39.6%; aRD 8.4%, 95% CI -0.2 to 16.9; weak evidence) or TDF/XTC/DTG (39.0%; aRD 9.0%, 95% CI 2.9 to 15.1; strong evidence), and similar for AZT/XTC/DTG and TDF/XTC/DTG. Per protocol analyses gave similar results (Table).
Conclusions
Compared with AZT/XTC/LPV/r, people on second-line DTG had less viraemia, with no evidence of worse viraemia with recycled TDF. Interventions to improve LTFU and viraemia with second-line therapy are needed, and drug resistance surveillance remains essential.
Table or Figure
