Evaluating BDLC Regimen in Pre-XDR Tuberculosis: Findings from the endTB-Q Phase 3 Trial

Highlight

– The endTB-Q phase 3 trial evaluated an all-oral regimen combining bedaquiline, delamanid, linezolid, and clofazimine (BDLC) in patients with rifampicin-resistant and fluoroquinolone-resistant tuberculosis.
– The BDLC regimen demonstrated favourable treatment outcomes comparable to the WHO-recommended standard of care but did not establish overall non-inferiority.
– Safety profiles were similar across the BDLC and control groups, with high rates of grade 3 or higher adverse events in both arms.
– Findings suggest that pre-XDR tuberculosis may require prolonged, reinforced treatment regimens beyond shortened ones.

Study Background and Disease Burden

Multidrug-resistant tuberculosis (MDR-TB), particularly pre-extensively drug-resistant tuberculosis (pre-XDR-TB), defined by resistance to rifampicin and additional resistance to any fluoroquinolone, poses significant treatment challenges worldwide. The limitations of existing regimens include lengthy duration, substantial toxicity, and variable efficacy. New treatment combinations are urgently needed to improve cure rates, reduce treatment duration, and optimize safety in this patient population. The endTB-Q trial was designed to assess the efficacy and safety of a novel, shortened, all-oral regimen combining bedaquiline, delamanid, linezolid, and clofazimine (BDLC) compared with the individualized WHO-recommended longer standard of care in pre-XDR TB patients.

Study Design

This open-label, multicentre, stratified, non-inferiority, randomized controlled phase 3 trial was conducted at ten hospitals across six countries—India, Kazakhstan, Lesotho, Pakistan, Peru, and Vietnam—between April 2020 and March 2023. Participants aged 15 years or older, with culture-confirmed pulmonary tuberculosis resistant to rifampicin and fluoroquinolones, were eligible. A total of 1030 individuals were screened, and 324 patients (31%) were enrolled and randomized in a 2:1 ratio to receive either the BDLC regimen or the control regimen.

Randomization was stratified by country and baseline extent of disease (limited vs extensive). The BDLC treatment consisted of all-oral bedaquiline (initial 400 mg daily for 2 weeks followed by 200 mg thrice weekly), delamanid 100 mg twice daily, linezolid 600 mg daily for 16 weeks followed by dose reduction, and clofazimine 100 mg daily. Duration was 24 weeks (6 months) for limited disease and 39 weeks (9 months) for extensive disease. The regimen could be extended if sputum cultures remained positive or culture results were missing at 8 weeks.

The control group received individualized WHO-recommended longer regimens, typically based on second-line drugs, stratified by resistance profiles.

Masking was partial: site staff and participants were unblinded, but investigators and laboratory staff assessing endpoints were blinded to treatment assignments.

The primary outcome was a favorable treatment response at week 73, defined as either two consecutive negative sputum cultures (including one between weeks 65 and 73) or favorable bacteriological, radiological, and clinical evolution. Both modified intention-to-treat (mITT) and per-protocol populations were analyzed, with a pre-specified non-inferiority margin set at -12%. The trial is registered under ClinicalTrials.gov NCT03896685.

Key Findings

Of the 324 randomized participants, 219 received BDLC and 105 received the control regimen. The median age was 30.5 years, with a balanced gender distribution (46% female, 54% male), and 64% had extensive disease at baseline.

In the BDLC arm, 29% received the 6-month regimen and 71% the 9-month regimen based on disease extent. In the control group, 91% received BDLC plus additional drugs as core regimen.

At 73 weeks, favorable outcomes were observed in 87% of participants in the BDLC group and 89% in the control group in the mITT population (adjusted risk difference 0.2%, 95% CI -9.1 to 9.5; p for non-inferiority=0.0051), demonstrating non-inferiority in this analysis. However, in the per-protocol population, favorable outcomes were 88% for BDLC and 93% for control (adjusted risk difference -3.5%, 95% CI -12.8 to 5.9; p=0.037), failing to meet overall criteria for non-inferiority.

Figure 3 Kaplan–Meier plot of time to relapse (from treatment completion) by baseline extent of tuberculosis disease in the mITT population

Adverse events of grade 3 or higher occurred in 68% of the BDLC group and 73% of controls. Mortality was low but slightly higher in the BDLC group (4% versus 2%). These safety findings highlight substantial toxicity risks in this complex population regardless of regimen.

Expert Commentary

The endTB-Q trial represents a pivotal effort to evaluate novel all-oral regimens for a highly challenging subgroup of tuberculosis patients with resistance to rifampicin and fluoroquinolones. The high favorable outcome rates in both arms underscore improvements in therapeutic options compared to historical controls. However, the lack of definitive overall non-inferiority for the shortened BDLC strategy suggests that treatment duration and potential regimen reinforcement remain critical considerations in pre-XDR TB management.

While bedaquiline and delamanid have demonstrated substantial antimycobacterial activity, and linezolid and clofazimine provide synergistic effects, treatment of extensive disease may necessitate prolonged therapy to secure durable cure. The safety profile observed aligns with known toxicities of linezolid and other agents, underscoring the need for frequent monitoring and supportive care.

Limitations include the open-label design potentially introducing bias, heterogeneity in baseline disease and treatment regimens in the control group, and the relatively young median age limiting generalizability to older or comorbid populations. Furthermore, the necessity of complex individualized regimens in the control arm complicates direct comparisons.

Current tuberculosis guidelines increasingly recommend all-oral regimens incorporating bedaquiline and delamanid, but these findings highlight that careful patient selection and potential regimen tailoring based on disease extent and resistance profiles are essential.

Conclusion

The endTB-Q trial evaluated the efficacy and safety of a shortened, oral BDLC regimen in treating pre-extensively drug-resistant tuberculosis, compared with individualized longer WHO-standard regimens. Although the BDLC regimen showed promising efficacy with favorable outcomes comparable to standard care, it did not meet overall non-inferiority criteria, particularly in the per-protocol analysis. High rates of adverse events emphasize the clinical complexity of this patient population.

These data suggest that shortened regimens, while appealing, may not suffice for patients with extensive pre-XDR TB. Extended or reinforced treatment durations incorporating effective drug combinations may be necessary to optimize outcomes.

Further research to refine regimen components, duration, and patient stratification strategies remain priorities to improve management of drug-resistant tuberculosis worldwide.

References

1. Guglielmetti L, Khan U, Velásquez GE, et al; endTB-Q Clinical Trial Team. Bedaquiline, delamanid, linezolid, and clofazimine for rifampicin-resistant and fluoroquinolone-resistant tuberculosis (endTB-Q): an open-label, multicentre, stratified, non-inferiority, randomised, controlled, phase 3 trial. Lancet Respir Med. 2025 Sep;13(9):809-820. doi: 10.1016/S2213-2600(25)00194-8 IF: 32.8 Q1 B1. Epub 2025 Jul 16. PMID: 40683298 IF: 32.8 Q1 B1.

2. World Health Organization. WHO consolidated guidelines on tuberculosis treatment. 2022 update. Geneva: WHO; 2022.

3. Nahid P, Mase SR, Migliori GB, et al. Treatment of drug-resistant tuberculosis. An official ATS/CDC/ERS/IDSA clinical practice guideline. Am J Respir Crit Care Med. 2019;200(10):e93-e142. doi:10.1164/rccm.201909-1874ST IF: 19.4 Q1 B1.

4. Lange C, Dheda K, Chesov D, et al. Drug-resistant tuberculosis: an update on disease burden, diagnosis and treatment. Respirology. 2022;27(2):157-172. doi:10.1111/resp.14132 IF: 6.3 Q1 B2.