Bedaquiline

證據等級: L5

預測適應症: 10 個

Bedaquiline: From MDR-TB Treatment to Inactive Tuberculosis (Latent TB Prevention)

One-Sentence Summary

Bedaquiline (Sirturo) is a diarylquinoline antibiotic approved globally for multidrug-resistant tuberculosis (MDR-TB) treatment, though not currently registered in New Zealand. The TxGNN model predicts it may be effective for Inactive Tuberculosis (latent TB prevention), with 3 registered clinical trials — including one Phase 2/3 trial enrolling 2,530 participants — and 20 publications currently supporting this direction, representing a meaningful clinical shift from treating active disease to preventing reactivation.

Note on TxGNN rankings: The top-ranked prediction by TxGNN score is tuberculosis, bovine (rank 1, score 99.96%), followed closely by tuberculous ascites (rank 2), tuberculoma (rank 3), inactive tuberculosis (rank 4), and avian tuberculosis (rank 5) — all with near-identical scores (~99.96%). This report focuses on inactive tuberculosis (rank 4) as the primary subject because it carries the strongest clinical evidence (L2, Phase 2/3 RCT) and the most actionable repurposing rationale. Other TB-spectrum predictions are summarised in the Predicted Indication Landscape section below.

Quick Overview

Item	Content
Original Indication	Multidrug-resistant tuberculosis (MDR-TB) — globally approved; not registered in New Zealand
Predicted New Indication	Inactive Tuberculosis (Latent TB Prevention)
TxGNN Prediction Score	99.96% (rank 660/total)
Evidence Level	L2
New Zealand Market Status	Not marketed
Number of Authorizations	0
Recommended Decision	Proceed with Guardrails

Why Is This Prediction Reasonable?

Bedaquiline works by selectively inhibiting the mycobacterial F₀F₁-ATP synthase — specifically binding the c-subunit of the F₀ membrane rotor — thereby blocking the proton translocation required for ATP generation. This mechanism is lethal even to dormant, non-replicating M. tuberculosis bacilli that are in a low-energy metabolic state, a property that sharply distinguishes bedaquiline from most first-line TB drugs (isoniazid, rifampicin) whose bactericidal activity depends on active bacterial replication. The drug's selectivity is exceptional: human mitochondrial ATP synthase shows >20,000-fold lower sensitivity compared to mycobacterial ATP synthase, providing a large therapeutic window.

This unique activity against dormant bacilli is precisely why the leap from treating MDR-TB to preventing reactivation of latent TB (LTBI/inactive TB) is mechanistically coherent. Latent TB exists in a dormant state sustained by low-energy metabolism; bedaquiline's ability to sterilize even non-replicating organisms suggests it could clear the bacterial reservoir that drives reactivation. A 2022 mouse model study confirmed that long-acting bedaquiline formulations showed sustained antituberculosis activity for preventive therapy scenarios (PMID 34939891), directly validating the animal proof-of-concept.

The most significant safety challenge for this repurposing is the risk-benefit recalibration: current LTBI standard regimens (isoniazid, rifapentine) have well-established short-course safety profiles, while bedaquiline carries a risk of QT prolongation and hepatotoxicity that was accepted in the context of life-threatening MDR-TB, but requires more rigorous justification for a preventive indication in otherwise healthy or HIV-positive contacts. Defining the specific subpopulations — drug-resistant TB contacts, people living with HIV (PLHIV) — where the benefit clearly outweighs this risk is the central clinical question driving BREACH-TB.

Clinical Trial Evidence

(Inactive Tuberculosis — Rank 4)

Trial Number	Phase	Status	Enrollment	Key Findings
NCT06568484	Phase 2/3	Not Yet Recruiting	2,530	BREACH-TB: Seamless Phase 2/3 non-inferiority trial comparing 4-week bedaquiline vs. standard preventive regimen in PLHIV and high-risk contacts of DS-TB or RR-TB cases; 72-week follow-up for confirmed/probable TB disease
NCT05766267	Phase 2/3	Active, Not Recruiting	288	CRUSH-TB: 17-week bedaquiline + moxifloxacin + pyrazinamide ± rifabutin/delamanid vs. standard 6-month regimen for pulmonary TB; efficacy/safety data directly informs dosing parameters relevant to preventive use
NCT07069582	Phase 1	Not Yet Recruiting	60	Sub-study of SSTARLET: PK profiling of bedaquiline in breastfeeding women after single dose; provides safety data for a population relevant to TB preventive therapy (TPT) expansion

Literature Evidence

(Inactive Tuberculosis — Rank 4; up to 10 most relevant)

PMID	Year	Type	Journal	Key Findings
33299175	2021	Mechanism / Structural	Nature	Cryo-EM structure of mycobacterial ATP synthase bound to bedaquiline; confirms drug can sterilize latent M. tuberculosis by targeting dormant-phase energy metabolism
39766559	2024	Review / Mechanism	Antibiotics	Comprehensive review of Mtb F-ATP synthase inhibitors; details bedaquiline's activity under dormant/low-energy conditions directly relevant to latent TB clearance
34939891	2022	Preclinical	Am J Respir Crit Care Med	Long-acting bedaquiline formulation shows sustained antituberculosis activity for ≥12 weeks in validated mouse model of preventive therapy; proof-of-concept for LTBI indication
39301910	2025	Review	Infect Disord Drug Targets	Reviews bedaquiline delivery systems for MDR-TB; explicitly notes ability to target persistent/latent TB forms that remain viable despite conventional therapy
36982277	2023	Review	Int J Mol Sci	TB pathogenesis and treatment review; covers latent infection (~25% of global population) and emerging drug targets including bedaquiline for dormant bacilli
39887565	2025	Review	Respirology	Updates TB disease spectrum concept (latent → subclinical → active); supports rationale for preventive therapy in subclinical/inactive TB using newer agents
36915977	2022	Review	J Zhejiang Univ Med Sci	Progress on LTBI diagnosis and treatment; discusses current limitations of isoniazid/rifamycin regimens and need for novel agents
29187395	2018	Review	Clin Microbiol Rev	Comprehensive review of therapeutic approaches to dormant M. tuberculosis; bedaquiline identified as a key candidate due to ATP synthase inhibition in non-replicating state
38003836	2023	Review	Pathogens	Pediatric drug-resistant TB management; bedaquiline and delamanid highlighted for safe use in children — relevant to preventive therapy in pediatric contacts
28256380	2017	Review	Presse Médicale	TB/HIV co-infection challenges; identifies PLHIV as highest-priority group for preventive therapy — directly corresponds to BREACH-TB target population

Predicted Indication Landscape

All 10 TxGNN predictions are summarised below for clinical decision-making context:

Rank	Disease	TxGNN Score	Evidence Level	Decision	Rationale Summary
1	Tuberculosis, Bovine	99.96%	L4	Research Question	M. bovis carries homologous ATP synthase target; 3 in-vitro papers confirm selectivity. No in vivo or clinical data. Veterinary regulatory pathway required.
2	Tuberculous Ascites	99.96%	L5	Hold	No evidence. Peritoneal PK (drug penetration) completely unknown.
3	Tuberculoma	99.96%	L4	Research Question	Case reports of bedaquiline in MDR/XDR-TB with CNS involvement. BBB penetration is poor (low CSF/plasma ratio) — the critical unknow.
4	Inactive Tuberculosis	99.96%	L2	Proceed with Guardrails	Phase 2/3 BREACH-TB trial (n=2,530); mouse model proof-of-concept; 20 publications. Best evidence in this pack.
5	Tuberculosis, Avian	99.96%	L5	Hold	M. avium ATP synthase has lower bedaquiline affinity (high MIC in vitro); no supporting evidence.
6	Vulvovaginal Candidiasis	99.88%	L5	Hold	Fungal ATP synthase structurally distinct; no mechanism; TxGNN graph noise (TB/immunosuppression co-occurrence).
7	Fascioliasis	99.88%	L5	Hold	Eukaryotic helminth — no mechanistic basis whatsoever.
8	Urea Cycle Disorder	99.78%	L5	Hold	Genetic metabolic disorder; completely unrelated to antimicrobial mechanism.
9	Cutaneous Tuberculosis	99.70%	L4	Research Question	M. tuberculosis causative — mechanism applies. Phase 3 LEOPARD trial (n=124,000) tests bedaquiline for M. leprae prophylaxis, providing cross-mycobacterial safety data. Skin penetration PK needed.
10	Esophageal Candidiasis	99.68%	L5	Hold	Same as rank 6 — fungal, no mechanism, graph confound via HIV co-infection.

Safety Considerations

Detailed New Zealand / Taiwan-specific package insert data is not available (bedaquiline is not registered in New Zealand). The following safety information reflects globally available regulatory and clinical data:

QT Prolongation: Bedaquiline prolongs the QT interval; cardiac monitoring (baseline ECG and regular follow-up) is mandatory. Risk is amplified when co-administered with other QT-prolonging agents (e.g. fluoroquinolones, clofazimine, azithromycin).
Hepatotoxicity: Clinically significant liver enzyme elevations have been reported. Liver function tests (AST, ALT, bilirubin) should be monitored at baseline and throughout treatment.
Long Half-life: Terminal half-life is approximately 5.5 months (due to redistribution from tissues). Adverse effects and drug interactions may persist for months after the last dose — a particular consideration when designing LTBI preventive regimens that are inherently shorter.
Drug Interactions: Bedaquiline is metabolised primarily by CYP3A4. Strong CYP3A4 inducers (e.g. rifampicin, efavirenz) significantly reduce bedaquiline plasma levels. This interaction is clinically critical in the TB/HIV co-infection setting targeted by BREACH-TB.
Mortality Signal: A higher rate of all-cause mortality was observed in the bedaquiline arm vs. placebo in the pivotal Phase 2b trial (C208); the mechanistic explanation remains unclear. This signal requires careful monitoring in any preventive-use protocol.

Conclusion and Next Steps

Decision: Proceed with Guardrails (for Inactive Tuberculosis / Latent TB Prevention)

Rationale: BREACH-TB (NCT06568484) is a rigorously designed Phase 2/3 non-inferiority trial specifically targeting this indication (n=2,530, PLHIV and high-risk contacts), demonstrating that the field has cleared the proof-of-concept threshold and entered formal clinical validation. The mechanistic basis — bedaquiline's unique killing activity against dormant mycobacteria — is supported by structural biology data and a validated mouse preventive therapy model. The path to regulatory evaluation is clear, albeit requiring a compelling safety-benefit analysis against the established LTBI standard of care.

To proceed, the following is needed:

Await BREACH-TB primary results (estimated completion: September 2027)
Clarify QT prolongation and hepatotoxicity risk in healthy or HIV-positive contacts receiving bedaquiline for prevention (vs. MDR-TB treatment context where higher risk is tolerated)
Define subpopulations with net benefit: drug-resistant TB contacts and PLHIV are the leading candidates
Investigate CYP3A4 drug interaction management in patients on antiretroviral therapy (ART)
Evaluate long-acting injectable formulation feasibility (NCT34939891 mouse data) to improve adherence in preventive settings
Obtain bedaquiline New Zealand regulatory registration for active TB first, as a prerequisite pathway to preventive indication

For secondary indications requiring further research (Research Question):

Tuberculoma / CNS TB: Measure CSF/plasma ratio in MDR-TB patients receiving bedaquiline; evaluate whether adjunctive steroids + BDQ improves CNS penetration
Cutaneous TB: Pharmacokinetic study of bedaquiline in skin tissue; LEOPARD trial safety data (n=124,000) can provide a safety bridge
Bovine TB: Requires veterinary pharmacology programme; in vivo cattle model efficacy and food safety assessment before any regulatory consideration
Disclaimer

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