A total of 799 individuals who switched to BIC/FTC/TAF were included in the analysis. Baseline characteristics of the overall population and of the subgroup switching from RPV/FTC/TAF (n = 103) are summarised in Tables 1 and 2, respectively. The median age was 53 (IQR 45, 58) years, and most participants were male (83%). Most individuals acquired HIV through sexual exposure (heterosexual 42.2% and men who have sex with men 31.7%). Comorbidities were common, including dyslipidaemia (38.7%) and hypertension (24.9%), and nearly half of the cohort was overweight or obese. Polypharmacy was frequently observed at baseline, with approximately 14% of participants receiving concomitant medications classified as “orange” or “red” according to the Liverpool drug-drug interaction database in relation to their pre-switch antiretroviral regimen.

At the time of switch, the median CD4 cell count was 701 cells/mm3 (IQR 502–903) in the overall cohort and 730 cells/mm3 (IQR 548–911) in the subgroup of individuals switching from RPV/FTC/TAF, with comparable immunological profiles between groups.

Reasons for switching to BIC/FTC/TAF are summarized in Fig. 1. The most common reasons for switching to BIC/FTC/TAF were the need to increase the genetic barrier of the regimen, reported in 489 individuals (61.2%), and treatment simplification, reported in 209 individuals (26.2%). Other reasons included potential or documented drug-drug interactions in 34 individuals (4.3%) and other clinical reasons in 48 individuals (6.0%). Less frequent reasons included central nervous system toxicity, dyslipidaemia, gastrointestinal toxicity, hepatic or renal toxicity, patient choice, osteoporosis, lipodystrophy, and adherence-related considerations. Reasons for switching were recorded as a single primary reason per participant.

Reasons for switching to BIC/FTC/TAF. Bar chart showing the number of participants by reason for switching to bictegravir/emtricitabine/tenofovir alafenamide in the overall study population (N = 799). Reasons were recorded as a single primary reason per participant. CNS, central nervous system; BIC/FTC/TAF: bictegravir/emtricitabine/tenofovir alafenamide

Virological suppression was maintained in the majority of participants throughout the follow-up period. At 6 months, 746 of 766 individuals (97.4%; 95% CI 96.3–98.5%) had HIV-RNA < 50 copies/ml, and at 12 months, 728 of 743 (98.0%; 95% CI 97.0–99.0%) were virologically suppressed (observed-case analysis). In the prespecified subgroup switching from RPV/FTC/TAF, virological suppression was 93.9% at 6 months and 98% at 12 months.

Virological events were uncommon. The cumulative incidence of virological failure was 0.5% (95% CI 0.2–1.3%) at 6 months and 1.5% (95% CI 0.8–2.6%) at 12 months. Episodes of detectable HIV-RNA were infrequent and mostly consisted of transient low-level viraemia, generally occurring in the context of documented suboptimal adherence. In most cases, these episodes did not lead to treatment discontinuation or modification, and virological suppression was subsequently re-established. Only one participant discontinued BIC/FTC/TAF because of virological failure.

In individuals switching from RPV/FTC/TAF, the cumulative incidence of virological failure was 0% at 6 months and 2.1% (95% CI 0.5–8.3%) at 12 months, with only two confirmed virological failures.

Treatment persistence with BIC/FTC/TAF was high, as shown by the Kaplan-Meier analysis (Fig. 2). Detailed Kaplan-Meier estimates for the overall cohort and subgroup analyses according to age and baseline antiretroviral regimen are reported in Supplementary Table S1. The estimated probability of remaining on treatment without discontinuation for any cause was 96.6% (95% CI 95.3–97.9) at 6 months and 94.6% (95% CI 93.0–96.2) at 12 months. Treatment persistence remained consistently high across age subgroups at the time of switch, with no clinically meaningful differences observed (Fig. 2B). When stratified by age, treatment persistence was 99.1% (95% CI 97.5–100) at 6 months and 98.1% (95% CI 95.4–100) at 12 months among individuals aged < 40 years; 96.4% (95% CI 94.9–98.0) and 94.4% (95% CI 92.4–96.4), respectively, among those aged 40–60 years; and 95.1% (95% CI 91.6–98.6) and 92.6% (95% CI 88.1–97.1), respectively, among those aged > 60 years. Differences across age groups were not statistically significant by log-rank test (p = 0.310). In analyses stratified by prior regimen, treatment persistence remained high. In the subgroup switching from RPV/FTC/TAF, treatment persistence was 97.1% (95% CI 93.8–100) at both 6 and 12 months, compared with 96.5% (95% CI 95.2–97.9) and 94.3% (95% CI 92.5–96.0) among individuals switching from other regimens, with no clinically meaningful differences between groups (Fig. 2C). Differences between groups were not statistically significant by log-rank test (p = 0.384).

Treatment durability of BIC/FTC/TAF. Kaplan-Meier estimates of the probability of remaining on bictegravir/emtricitabine/tenofovir alafenamide without treatment interruption for any cause. A Overall cohort, with 95% confidence intervals. B Estimates stratified by age group: < 40 years, 40–60 years, and > 60 years. C Estimates according to prior antiretroviral regimen: rilpivirine/emtricitabine/tenofovir alafenamide versus other antiretroviral regimens. Participants without treatment interruption were censored at the last available follow-up. ART antiretroviral therapy; CI confidence interval; RPV/FTC/TAF rilpivirine/emtricitabine/tenofovir alafenamide, BIC/FTC/TAF bictegravir/emtricitabine/tenofovir alafenamide

Treatment interruptions were uncommon and occurred for heterogeneous reasons (Fig. 3). Missing virological assessments during follow-up were mainly attributable to loss to follow-up (n = 8) and transfer of care to another clinical centre (n = 3) rather than documented virological failure. During the first 6 months after switch, the most frequently reported reasons were dropout (n = 7), dyslipidaemia (n = 5), simplification (n = 5), central nervous system toxicity (n = 4), and other reasons (n = 4). No treatment interruption due to virological failure was reported during the first 6 months. Between 6 and 12 months, the most frequently reported reasons were simplification (n = 6), hepatic or renal toxicity (n = 3), pregnancy (n = 3), dropout (n = 3), death (n = 2), and weight gain (n = 2). One treatment interruption due to virological failure was reported between 6 and 12 months.

Reasons for treatment interruption. Bar chart showing the number of participants who interrupted treatment for each reason during follow-up, stratified by time interval after switching to bictegravir/emtricitabine/tenofovir alafenamide. A total of 33 and 23 participants interrupted treatment during months 0–6 and 6–12, respectively. Reasons were recorded as a single primary reason per participant. CNS central nervous system

CD4 cell counts showed a modest increase during follow-up, without clinically meaningful changes in immunological status. In the overall cohort, the median CD4 count was 701 cells/mm3 (IQR 502–903) at baseline, 704 cells/mm3 (IQR 518–899) at 6 months, and 732 cells/mm3 (IQR 531–935) at 12 months. Although paired analyses identified statistically significant differences over time (median CD4 count was 701 [503–903] cells/mm3 at baseline, 704 [518–899] cells/mm3 at Month 6 and 732 [531–935] cells/mm3 at Month 12; p < 0.035 and p < 0.0001 respectively), the absolute magnitude of change was small and not considered clinically meaningful. Similar trends were observed in individuals switching from RPV/FTC/TAF, with median CD4 cell counts of 730 cells/mm3 (IQR 548–911) at baseline, 722 cells/mm3 (IQR 576–896) at 6 months, and 746 cells/mm3 (IQR 569–951) at 12 months. Laboratory parameters assessed during routine clinical care showed no clinically relevant variations over time and did not lead to treatment modification.

BIC/FTC/TAF was generally well tolerated. In the overall cohort, adverse events were reported in 6 individuals (0.8%) at 6 months and in 2 individuals (0.3%) at 12 months. Discontinuations due to adverse events were rare, and no unexpected safety signals emerged during the observation period.

Body mass index data were available for 437/799 participants at baseline and 433/743 at 12 months. Median body mass index was 24.9 kg/m2 (IQR 22.9–27.4) at baseline and 25.0 kg/m2 (IQR 22.8–27.5) at 12 months. In paired analyses, the median change from baseline to month 12 was 0.0 kg/m2 (IQR – 0.4–0.7), indicating no clinically relevant variation in body mass index over follow-up. Weight gain was reported as the reason for treatment discontinuation in two participants between Month 6 and Month 12.