Comparative Effectiveness of First-Line Antiretroviral Therapy

Results From a Large Real-World Cohort After the Implementation of Dolutegravir

Mariana Veloso Meireles; Ana Roberta P. Pascom; Elisabeth C. Duarte; Willi McFarland


AIDS. 2019;33(10):1663-1668. 

In This Article


Our study of over 100 000 ART-naive adults provides strong evidence for the superior effectiveness of dolutegravir-based regimens in a real-world setting after 1 year of ART initiation. More than 90% of those who initiated with TLD achieved viral suppression (<50 copies/ml) after the first year of treatment, a proportion higher than the 84% for TLE and 79% and lower for protease-inhibitor-based regimens.

Adherence levels were higher with TLD, which mirrors the greater tolerability observed in clinical trials.[5–7] We conducted both an intention-to-treat analogous analysis and a secondary per-protocol analogous analysis, in which patients who presented suboptimal adherence were excluded. Importantly, aOR for TLD relative to TLE was slightly higher in this secondary analysis, indicating that, although adherence to TLD is higher, the superiority of dolutegravir when compared with efavirenz does not seem to be driven primarily by lower rates of discontinuation and greater tolerability, as suggested previously.[6]

We recognize limitations of our study. First is the concern of indication bias, that is there may be confounding due to the reasons why physicians chose certain regimens, such as pretreatment genotyping or contraindications, most of which are not recorded in our databases. We expect this affects especially the results of regimens that were not recommended as first-line during the study period, whose interpretation requires caution. Nonetheless, we do not believe this bias substantially affects the comparison between TLE and TLD, as the profile of patients starting on those regimens were similar regarding known variables and our sensitivity analyses produced similar results. Another limitation is the use of secondary data, wherein important factors were not systematically recorded or had substantial missing information. However, the use of these databases also presents strengths of our study. The large number of PLHIV results in sufficient statistical power to detect small effect sizes, even among subgroups, and the possibility to assess effectiveness of multiple regimens in a real-world population. In addition, we defined having a 12-month viral load count as an inclusion criterion, thereby not assessing later loss to follow-up (LTFU). The SISCEL database only captures viral load examinations performed within Brazil's public health system (SUS), and people frequently switch between using SUS and private healthcare[18] making assessment of LTFU difficult. This issue has been discussed in a previous study using these same databases,[19] and, on the basis of the sensitivity analyses and robustness of results, we do not believe the level of LTFU changes our interpretation. In addition, temporal bias is always a concern in observational studies. As TLD was recommended in 2017, including the year of ART initiation as a covariate may not fully address this issue. Notably, there is a sharper increase in the proportion of viral suppression from 2016 to 2017, beyond the seemingly modest trend observed from 2014 to 2016. Moreover, between 2014 and 2017, viral suppression did not significantly improve after controlling for other covariates.

In conclusion, our study provides real-life, national-level evidence of the superiority of dolutegravir over efavirenz, lopinavir and atazanavir in suppressing viral replication in adults by the first year on ART. Findings endorse the decision made by the BMoH to switch its recommendations for first-line ART from TLE to TLD, and the recommendation of WHO for national programmes of lower/middle and higher- income countries to transition to dolutegravir-based first-line regimens.