Bedaquiline

證據等級: L5

預測適應症: 10 個

Bedaquiline: From Pulmonary MDR-TB Treatment to Multidrug-Resistant Tuberculosis

One-Sentence Summary

Bedaquiline is a diarylquinoline antimycobacterial agent approved for the treatment of pulmonary multidrug-resistant tuberculosis (MDR-TB) as part of combination therapy. The TxGNN model predicts it to be effective for Multidrug-Resistant Tuberculosis with a score of 99.98%, which effectively validates the drug’s existing approved indication. With 50 clinical trials and 20 publications supporting this direction, the evidence base is exceptionally strong — and the model further identifies potential extensions to inactive tuberculosis, tuberculoma, bovine tuberculosis, and cutaneous tuberculosis as repurposing candidates.

Quick Overview

Item	Content
Original Indication	Pulmonary multidrug-resistant tuberculosis (MDR-TB), as part of combination therapy
Predicted New Indication	Multidrug-resistant tuberculosis
TxGNN Prediction Score	99.98%
Evidence Level	L1 (≥2 completed Phase 3 RCTs)
Malaysia Market Status	✓ Marketed
Number of Registrations	1
Recommended Decision	Proceed with Guardrails

Why is This Prediction Reasonable?

Bedaquiline (formerly TMC207) is a first-in-class diarylquinoline that selectively inhibits the mycobacterial F₀F₁-ATP synthase, blocking the oxidative phosphorylation pathway that Mycobacterium tuberculosis depends on for energy production. This novel mechanism of action is distinct from all existing anti-TB drugs and is effective against both actively replicating and dormant bacilli, making it particularly valuable against drug-resistant strains that have acquired resistance to conventional agents like isoniazid and rifampicin.

The TxGNN model’s top prediction of multidrug-resistant tuberculosis for bedaquiline serves as a strong model validation signal — the algorithm independently identifies the drug’s already-approved indication with near-perfect confidence (99.98%). This finding confirms the model’s ability to detect genuine drug-disease relationships through knowledge graph analysis.

Beyond this validation, the model identifies several biologically plausible repurposing candidates within the tuberculosis spectrum: inactive (latent) tuberculosis (rank 5, L2 evidence), tuberculoma (rank 6, L4), bovine tuberculosis (rank 2, L4), and cutaneous tuberculosis (rank 10, L4). These predictions are mechanistically coherent, as all involve Mycobacterium species sharing the conserved ATP synthase target. Notably, the prediction for inactive tuberculosis is actively being tested in a landmark Phase 2/3 trial (BREACH-TB, NCT06568484, n=2,530), further validating the model’s predictive utility.

Clinical Trial Evidence

Trial Number	Phase	Status	Enrollment	Key Findings
NCT02754765	Phase 3	Completed	754	endTB trial: largest multi-country RCT evaluating five all-oral shortened regimens containing bedaquiline for MDR-TB efficacy and safety
NCT02409290	Phase 3	Completed	588	STREAM Stage 2: evaluated standard treatment regimens including bedaquiline-containing arm for MDR-TB
NCT02589782	Phase 2/3	Completed	552	TB-PRACTECAL: multi-arm RCT evaluating bedaquiline + pretomanid combination regimens for MDR-TB; demonstrated superiority of BPaL-based regimens
NCT04062201	Phase 3	Completed	402	BEAT-TB: 6-month BDQ+DLM+LZD+LVX+CFZ regimen vs. South African standard of care for RR-TB
NCT03896685	Phase 3	Completed	323	endTB-Q: two all-oral shortened regimens for MDR-TB with fluoroquinolone resistance
NCT00910871	Phase 2	Completed	241	C209 pivotal open-label trial: bedaquiline + individualized background regimen for MDR-TB; key efficacy and safety data supporting regulatory approval
NCT00449644	Phase 2	Completed	208	C208: first randomized, placebo-controlled trial demonstrating bedaquiline’s antibacterial activity superiority over placebo in newly diagnosed MDR-TB
NCT03086486	Phase 3	Completed	181	ZeNix trial: optimized linezolid dosing within BPaL regimen for XDR-TB and treatment-intolerant MDR-TB
NCT02333799	Phase 3	Completed	109	Nix-TB: BPaL regimen (bedaquiline + pretomanid + linezolid) for XDR/treatment-intolerant MDR-TB; results led to pretomanid FDA approval
NCT05081401	Phase 3	Recruiting	1,050	INSPIRE-TB: pragmatic RCT evaluating seven 9-month oral regimens for RR-TB, the largest ongoing bedaquiline combination trial

Literature Evidence

PMID	Year	Type	Journal	Key Findings
32130813	2020	Landmark Trial	N Engl J Med	BPaL regimen achieved high treatment success in highly drug-resistant pulmonary TB with historically poor outcomes
35649043	2022	Systematic Review / Meta-Analysis	J Bras Pneumol	Bedaquiline-containing regimens for MDR-TB demonstrated high treatment success rates across pooled studies
40172415	2025	Systematic Review / Meta-Analysis	J Bras Pneumol	Meta-analysis confirmed bedaquiline + linezolid regimens enhance MDR-TB treatment outcomes
32192585	2020	IPD Meta-Analysis	Lancet Respir Med	Individual patient data analysis of drug-associated adverse events in MDR-TB treatment, informing risk-benefit assessment
36331978	2023	Systematic Review / Meta-Analysis	Clin Infect Dis	Concomitant bedaquiline + delamanid use showed favorable treatment outcomes with manageable QTc prolongation risk
31526739	2019	Review	Lancet	Comprehensive management of drug-resistant TB, positioning bedaquiline as a cornerstone of modern MDR-TB regimens
39431709	2024	Narrative Review	J Clin Lab Anal	Overview of bedaquiline and delamanid antimicrobial characteristics, resistance mechanisms, synergism, and side effects
32700565	2020	Clinical Review	Future Microbiol	Clinician’s perspective on bedaquiline and delamanid use in drug-resistant TB, summarizing observational and trial evidence
33837767	2021	Cohort Study	Clin Infect Dis	Real-world comparison of bedaquiline vs. delamanid treatment outcomes in MDR/RR-TB patients in South Korea
39537610	2024	Basic Science	Nat Commun	Catalase activity deficiency sensitizes MDR-TB to bedaquiline, elucidating a novel mechanism for enhanced efficacy against isoniazid-resistant strains

Malaysia Market Information

Authorization Number	Product Name	Dosage Form	Approved Indication
(Registration number not available)	Bedaquiline product	(Not specified)	(Indication text not available in registry data)

Note: Bedaquiline is registered and marketed in Malaysia (1 authorization). Detailed product information including authorization number, product name, dosage form, and approved indication text were not available in the current data extraction. Please refer to the NPRA database for complete registration details.

Safety Considerations

Please refer to the package insert for safety information.

Known class concerns based on published literature (not from registry data):

QTc prolongation: Bedaquiline is associated with QT interval prolongation; ECG monitoring is recommended during treatment

Hepatotoxicity: Liver function monitoring is advised

Drug interactions: As a CYP3A4 substrate, co-administration with strong CYP3A4 inducers (e.g., rifampicin) significantly reduces bedaquiline exposure; co-administration with CYP3A4 inhibitors may increase exposure

Mortality signal: An early Phase 2 trial (C208) noted a numerical imbalance in deaths in the bedaquiline arm, though subsequent larger studies have not confirmed a causal relationship

⚠️ Formal safety data (package insert warnings, contraindications, DDI) were not available in the evidence pack. The above points are derived from published clinical literature and should be verified against the approved prescribing information.

Additional Repurposing Candidates

Beyond the primary validation prediction, the TxGNN model identified the following repurposing candidates of potential clinical interest:

Rank	Disease	TxGNN Score	Evidence Level	Recommendation	Key Rationale
2	Bovine tuberculosis	99.96%	L4	Research Question	M. bovis shares highly conserved ATP synthase with M. tuberculosis; in vitro activity confirmed but no clinical data
5	Inactive (latent) tuberculosis	99.96%	L2	Proceed with Guardrails	Dormant M. tuberculosis relies on oxidative phosphorylation; a landmark Phase 2/3 trial (BREACH-TB, n=2,530) is underway
6	Tuberculoma	99.96%	L4	Research Question	Granulomatous TB lesion; case reports support BPaL regimen use, but CNS/tissue penetration data needed
10	Cutaneous tuberculosis	99.70%	L4	Research Question	Same pathogen; skin drug penetration and local concentration require verification

Note: Predictions for vulvovaginal candidiasis (rank 7), fascioliasis (rank 8), and urea cycle disorder (rank 9) lack mechanistic rationale and are classified as Hold — bedaquiline’s ATP synthase inhibition is highly selective for mycobacteria and has no established basis for activity against fungi, parasites, or metabolic disorders.

Conclusion and Next Steps

Decision: Proceed with Guardrails

Rationale: The TxGNN prediction for bedaquiline in MDR-TB is a strong model validation case, supported by L1-level evidence including multiple completed Phase 3 RCTs (endTB, STREAM, TB-PRACTECAL, BEAT-TB, ZeNix, Nix-TB) enrolling thousands of patients. Bedaquiline is already a WHO-recommended cornerstone of modern MDR-TB treatment. The more novel and actionable repurposing opportunity lies in inactive (latent) tuberculosis prevention, which has L2-level evidence and a major ongoing Phase 2/3 trial (BREACH-TB).

To proceed, the following is needed:

Complete Malaysia regulatory data: Obtain full NPRA registration details (authorization number, product name, approved indication text)
Package insert safety data: Download and parse the approved prescribing information for formal warnings, contraindications, and drug-drug interactions
Mechanism of action documentation: Retrieve detailed MOA data from DrugBank (ATP synthase subunit c inhibition)
Latent TB indication development: Monitor BREACH-TB trial (NCT06568484) results for potential label expansion to tuberculosis preventive therapy
Pharmacokinetic gap analysis: For tissue-specific indications (tuberculoma, cutaneous TB), assess drug penetration data in granuloma, CNS, and skin compartments

Disclaimer: This report is for research purposes only and does not constitute medical advice. Drug repurposing candidates require clinical validation before application. Data cutoff: 2026-04-09.

Disclaimer

This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.