Participant Disposition and Baseline Characteristics

Of the 2801 participants included in the analysis, 290 experienced at least one viremic event on B/F/TAF. Participant inclusion in the analysis is shown in Fig. 2. A total of 145 participants had been treatment naïve at study entry, and 145 were virologically suppressed. In total, 14 participants were randomized to receive a comparator ART, had viremia at their last visit on study drug, and subsequently switched to B/F/TAF in the open-label extension phase of the studies. A total of 6 of these 14 participants also had viremic events during B/F/TAF treatment and were included in the group of 145 virologically suppressed participants who experienced ≥ 1 viremic event on B/F/TAF.

Fig. 2

Participant inclusion in the analysis. aIncludes six participants randomized to receive a comparator ART who had viremia at their last visit on comparator drug and subsequently switched to B/F/TAF in the open-label extension phase of the studies. ART antiretroviral therapy, B bictegravir, F emtricitabine, TAF tenofovir alafenamide

Overall, compared with participants without any viremic events, those with ≥ 1 viremic event were significantly more likely to be treatment naïve, younger, Black, have baseline HIV-1 RNA ≥ 50 copies/mL, have lower baseline cluster of differentiation 4 (CD4) count, and have CD4 < 200 cells/μL) (Table 1). Sex at birth and being of Hispanic or Latine ethnicity did not significantly predict the occurrence of viremic events. Participants with viremic events on a comparator ART had a median (quartile [Q]1, Q3) age of 38 (27, 50) years and were predominantly male at birth (9/14; 64%) and Black (9/14; 64%).

Table 1 Baseline demographics and clinical characteristicsEfficacy OutcomesOverview of Viremic Events

In total, 411 viremic events were identified among 290 participants receiving B/F/TAF, with 219 (75.5%) experiencing one viremic event. Of those with multiple viremic events (i.e., separated by periods of virologic suppression [HIV-1 RNA < 50 copies/mL]), 49 (16.9%) experienced two viremic events and 22 (7.6%) experienced three or more. These values were similar between treatment-naïve participants and those who were virologically suppressed at their first dose of B/F/TAF (data not shown). At the participant level, the first viremic event occurred a median (Q1, Q3) of 254 (135, 541) days after first virologic suppression following B/F/TAF initiation.

Resuppression after a Viremic Event

Among participants with ≥ 1 viremic event, 250/290 (86.2%; 95% CI: 81.7–90.0%) achieved resuppression after their last viremic event (Fig. 3). When nonevaluable events were excluded, resuppression was achieved in 250/263 participants (95.1%; 95% CI: 91.7–97.3%). At the event level, 371/411 of viremic events (90.3%; 95% CI: 87.0–93.0%) were followed by resuppression, rising to 371/384 (96.6%; 95% CI: 94.3–98.2%) when nonevaluable viremic events were excluded (Fig. 3).

Fig. 3

Virologic outcomes following viremic events on B/F/TAF. Main analysis with viremic event threshold ≥ 50 copies/mL: (a and b) outcomes per participant, based on outcome of last viremic event; (c and d) outcomes per viremic event. Sensitivity analysis with viremic event threshold ≥ 1000 copies/mL: (e and f) outcomes per participant, based on outcome of last viremic event. Data were analyzed using the “not evaluable = continued viremia” approach (a, c, e) and using the “not evaluable = excluded” approach (b, d, f). aNot evaluable = virologic event at last assessment. B bictegravir, F emtricitabine, TAF tenofovir alafenamide

The median (Q1, Q3) time from a viremic event to documented resuppression at a subsequent clinic visit was 22 (18, 36) days (Table 2). The median (Q1, Q3) HIV-1 RNA at the start date of the viremic event for all viremic events (n = 411) was 2.25 (1.88, 3.01) log10 copies/mL. The median (Q1, Q3) follow-up time (and sustained suppression on B/F/TAF) after resuppression of the last viremic event, based on HIV-1 RNA measurements, was 319 (127, 797) days.

Table 2 Characteristics of viremic events on B/F/TAFTime to Resuppression following Viremic Events on B/F/TAF

Among participants who experienced a viremic event on B/F/TAF, and for whom complete HIV-1 RNA data were available for the subsequent visits, 192/242 (79.3%; 95% CI: 73.7–84.3%) had achieved resuppression by week 4 (baseline reset at the date of last viremia) and 197/207 (95.2%; 95% CI: 91.3–97.7%) had achieved resuppression by week 12 following the viremic event (Supplementary Digital Content Fig. S1). All participants with complete HIV-1 RNA data up to week 36 after the viremic event achieved resuppression (160/160).

At week 24 after the first viremic event (baseline reset at the date of first viremia), 205/225 participants (91.1%; 95% CI: 86.6–94.5%) had HIV-1 RNA < 50 copies/mL. By week 48, 149/157 participants (94.9%; 95% CI: 90.2–97.8%) had HIV-1 RNA < 50 copies/mL.

Participants with Continued Viremia

Among the 13 participants with continued viremia, the median (Q1, Q3) duration of viremia from the first instance of HIV-1 RNA ≥ 50 copies/mL to the last HIV-1 RNA assessment on B/F/TAF was 72 (43, 87) days (Table 2). In total, 9 of the 13 participants (69.2%) had continued viremia with ≥ 200 copies/mL (i.e., not low-level viremia) and a median (Q1, Q3) duration of 79 (30, 127) days. Of these nine participants, seven had continued viremia with ≥ 1000 copies/mL. In total, 11 of the 13 participants (84.6%) with continued viremia prematurely discontinued B/F/TAF (4 lost to follow-up, 4 due to participant’s decision, and 3 due to lack of efficacy), 1 participant remained on B/F/TAF to study completion and continued on B/F/TAF (Study 4030), and 1 participant was still on B/F/TAF in Study 1474 at the time of this analysis.

Resuppression with B/F/TAF after a Viremic Event on Comparator ART

Among the 14 participants who had viremia on a comparator ART before switching to B/F/TAF, 100% achieved resuppression on B/F/TAF, although 6 had a subsequent viremic event on B/F/TAF. The median (Q1, Q3) HIV-1 RNA at the first assessment of the viremic event for these 14 participants was 2.5 (2.0, 4.3) log10 copies/mL. The median (Q1, Q3) time from a viremic event to documented resuppression at a subsequent clinic visit was 25 (21, 42) days.

As expected, the median (Q1, Q3) baseline HIV-1 RNA at study entry for treatment-naïve participants (from Studies 1489 and 1490, n = 634) of 4.42 (4.00, 4.88) log10 copies/mL was higher than that reported above for participants who were virologically suppressed prior to experiencing a viremic event (either on B/F/TAF or a comparator ART). For these treatment-naïve participants, the median (Q1, Q3) time from study day 1 to documented virologic suppression was 29 (29, 33) days.

Treatment Naïve versus Virologically Suppressed at First Dose of B/F/TAF

Among 145 participants who were treatment naïve at first dose of B/F/TAF who experienced ≥ 1 viremic event, 134 (92.4%; 95% CI: 86.8–96.2%) achieved resuppression after their last viremic event. When nonevaluable events were excluded, resuppression was achieved in 134/138 participants (97.1%; 95% CI: 92.7–99.2%).

Among the 145 participants who were virologically suppressed on their first dose of B/F/TAF and who had ≥ 1 viremic event, 116 (80.0%; 95% CI 72.6–86.2%) achieved resuppression after their last viremic event. When nonevaluable events were excluded, resuppression was achieved in 116/125 participants (92.8%; 95% CI: 86.8–96.7%).

For comparison, the median (Q1, Q3) HIV-1 RNA for treatment-naïve participants at the date of last viremia was 2.29 (1.93, 3.29) log10 copies/mL, compared with 2.19 (1.85, 2.96) for virologically suppressed participants.

Baseline Genotype Resistance Data for Participants with a Viremic Event

Analysis of preexisting primary resistance substitutions for the overall population (not limited to participants with viremic events) has previously been published both for the treatment-naïve [25] and virologically suppressed [15, 26,27,28,29,30] participants in this analysis.

As part of our analysis, baseline genotyping data were available for 284/290 participants (97.9%) with viremic events on B/F/TAF. The most frequent preexisting primary resistance substitutions overall were K103N/S (n = 28), E138A/G/K/Q/R (n = 16), M184I/V (n = 14), and G190A/E/Q/S (n = 11) in RT (Table 3). The proportion of participants (with available data) with M184V/I among those with viremic events (4.9%; 14/284) was similar to the proportion in the overall population (6.6%; 177/2690).

Table 3 Baseline genotype data for participants with viremic events on B/F/TAFa

Of the 14 participants with M184V/I receiving B/F/TAF, 10 (71.4%) had virologic suppression after their last viremic event, 3 (21.4%) had continued viremia, and 1 (7.1%) did not have an evaluable outcome after viremia. Of the eight participants with INSTI resistance receiving B/F/TAF, 7 (87.5%) had virologic suppression after their last viremic event and 1 (12.5%), with T97A, had continued viremia.

Postbaseline Virology Data

Of the 13 participants who had continued viremia after their last viremic event, 8 (61.6%) met the criteria for postbaseline resistance testing and had data available for PR/RT/IN. None had treatment-emergent resistance to study drugs.

Study Drug Adherence

Among participants without viremia who had available adherence data (n = 2491), the median (Q1, Q3) adherence rate was 98.2% (95.5%, 99.4%); in those with viremia (n = 290), it was 96.6% (92.0%, 99.0%). A significantly higher proportion of participants with viremic events had low adherence (< 85%) to B/F/TAF compared with those without viremic events (10.0% and 4.2%, respectively; p = 0.0003) (Fig. 4). Similar to the overall study population, resuppression after last viremia was achieved in 21/23 participants (91.3%; 95% CI: 72.0–98.9%) with < 85% adherence when nonevaluable data were excluded.

Fig. 4

B/F/TAF adherence. Participants who returned ≥ 1 pill bottle and had calculable drug adherence were included. B bictegravir, F emtricitabine, TAF tenofovir alafenamide

The adherence rate, duration of viremia on B/F/TAF, and last HIV-1 RNA measurement on B/F/TAF for the 13 participants with continued viremia are shown in Supplementary Digital Content Table S1. Two of these participants (15.3%) had < 85% adherence.

Among the 14 participants with viremia on a comparator ART before switching to B/F/TAF, 3 (23.1%) had < 85% adherence to B/F/TAF.

Sensitivity Analyses

Similar overall findings were observed when the analysis was repeated with the HIV-1 RNA threshold for a viremic event set to ≥ 1000 copies/mL (Fig. 3, e and f). This was also the case for results from an additional analysis with the threshold set to ≥ 200 copies/mL. Full details can be found in Supplementary Digital Content Sensitivity Analyses.