Real-world use and outcomes of dolutegravir-containing antiretroviral therapy in HIV and tuberculosis co-infection: a site survey and cohort study in sub-Saharan Africa

Abstract

Introduction: Dolutegravir is being scaled up globally as part of antiretroviral therapy (ART), but for people with HIV and tuberculosis co-infection, its use is complicated by a drug-drug interaction with rifampicin requiring an additional daily dose of dolutegravir. This represents a disadvantage over efavirenz, which does not have a major drug-drug interaction with rifampicin. We sought to describe HIV clinic practices for prescribing concomitant dolutegravir and rifampicin, and characterize virologic outcomes among patients with tuberculosis co-infection receiving dolutegravir or efavirenz.

Methods: Within the four sub-Saharan Africa regions of the International epidemiology Databases to Evaluate AIDS consortium, we conducted a site survey (2021) and a cohort study (2015-2021). The cohort study used routine clinical data and included patients newly initiating or already receiving dolutegravir or efavirenz at the time of tuberculosis diagnosis. Patients were followed from tuberculosis diagnosis until viral suppression (<1000 copies/ml), a competing event (switching ART regimen; loss to program/death) or administrative censoring at 12 months.

Results: In the survey, 86 of 90 (96%) HIV clinics in 18 countries reported prescribing dolutegravir to patients who were receiving rifampicin as part of tuberculosis treatment, with 77 (90%) reporting that they use twice-daily dosing of dolutegravir, of which 74 (96%) reported having 50 mg tablets available to accommodate twice-daily dosing. The cohort study included 3563 patients in 11 countries, with 67% newly or recently initiating ART. Among patients receiving dolutegravir (n = 465), the cumulative incidence of viral suppression was 58.9% (95% confidence interval [CI]: 54.3-63.3%), switching ART regimen was 4.1% (95% CI: 2.6-6.2%) and loss to program/death was 23.4% (95% CI: 19.7-27.4%). Patients receiving dolutegravir had improved viral suppression compared with patients receiving efavirenz who had a tuberculosis diagnosis before site dolutegravir availability (adjusted subdistribution hazard ratio [aSHR]: 1.47, 95% CI: 1.28-1.68) and after site dolutegravir availability (aSHR 1.28, 95% CI: 1.08-1.51).

Conclusions: At a programmatic level, dolutegravir was being widely prescribed in sub-Saharan Africa for people with HIV and tuberculosis co-infection with a dose adjustment for the drug-drug interaction with rifampicin. Despite this more complex regimen, our cohort study revealed that dolutegravir did not negatively impact viral suppression.

Keywords: HIV integrase inhibitors; antiretroviral agents; antitubercular agents; drug interactions; observational study; rifampin.

Figure 1

Exposure group assignment by site start of dolutegravir use in the patient‐level cohort study. The horizontal axis represents time, beginning at 24 months before the start of dolutegravir use at the patient's site and ending at the site date of database closure. Each individual line represents hypothetical event‐free 12‐month follow‐up for the three exposure groups: historical efavirenz (H‐EFV; dashed line), contemporaneous efavirenz (C‐EFV; dotted line) and dolutegravir (DTG; solid line). Patients who were receiving or initiating efavirenz at the time of a tuberculosis diagnosis were divided into two control groups: C‐EFV, with tuberculosis diagnosis after the patient's site began using dolutegravir and H‐EFV with tuberculosis diagnosis before the patient's site began using dolutegravir.

Figure 2

Assessment for eligibility and reasons for exclusion, exposure group disposition and outcomes. Abbreviations: ART, antiretroviral therapy; C‐EFV, contemporaneous efavirenz; DTG, dolutegravir; EFV, efavirenz; H‐EFV, historical efavirenz; TB, tuberculosis. ^†Details on regimen switches for the DTG group: n = 13 switched to efavirenz and n = 6 switched to a protease inhibitor (PI)‐containing regimen (atazanavir/ritonavir or lopinavir/ritonavir); details on regimen switches for the C‐EFV group: n = 46 switched to dolutegravir, n = 20 switched to a PI‐containing regimen and n = 4 switched to nevirapine; details on regimen switches for the H‐EFV group: n = 41 switched to dolutegravir, n = 63 switched to a PI‐containing regimen and n = 33 switched to nevirapine. ^‡Specific outcomes related to loss to program/death for the DTG group: n = 47 were lost to follow‐up, n = 31 had a known reason for leaving care, n = 31 had death documented; specific outcomes related to loss to program/death for the C‐EFV group: n = 96 were lost to follow‐up, n = 41 had a known reason for leaving care, n = 64 had death documented; specific outcomes related to loss to program/death for the H‐EFV group: n = 400 were lost to follow‐up, n = 104 had a known reason for leaving care, n = 187 had death documented.

Figure 3

Crude cumulative incidence of viral suppression in the 12 months after tuberculosis diagnosis stratified by exposure group among (a) the entire sample and (b) patients with viral load ascertained in the 12 months after tuberculosis diagnosis. Abbreviations: C‐EFV, contemporaneous efavirenz; DTG, dolutegravir; H‐EFV, historical efavirenz. The horizontal axes represent month of follow‐up after tuberculosis diagnosis through 12 months. The vertical axes represent the crude cumulative incidence proportion of patients with viral suppression (viral load <1000 copies/ml) at each month during follow‐up. The left panel (a) includes the entire sample (n = 3563) regardless of viral load testing and the right panel (b) includes the subgroup of patients who had viral load ascertained in the 12 months after their tuberculosis diagnosis (n = 2079). In the graph, the solid lines represent the DTG group, the dotted lines represent the C‐EFV group and the dashed lines represent the H‐EFV group. Median (interquartile range) months until viral suppression among those with the outcome was 6.1 (4.9–7.5) for DTG, 6.1 (4.2–7.6) for C‐EFV and 6.1 (3.9–8.0) for H‐EFV.