Pretomanid
| Evidence Level: L5 | Predicted Indications: 10 |
Table of Contents
Pretomanid: From Drug-Resistant Tuberculosis to Candidiasis
One-Sentence Summary
Pretomanid is a nitroimidazooxazine antimycobacterial agent, used as part of the BPaL regimen (bedaquiline + pretomanid + linezolid) for extensively drug-resistant tuberculosis (XDR-TB) and treatment-intolerant/non-responsive multidrug-resistant tuberculosis (MDR-TB). The TxGNN model predicts it may be effective for Candidiasis, but 0 clinical trials and 0 publications currently support this direction, and the evidence pack explicitly flags the mechanistic link as biologically implausible.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Drug-resistant tuberculosis (XDR-TB / treatment-intolerant or non-responsive MDR-TB), as part of the BPaL regimen — no formal marketing authorization text available; drug is not marketed in this jurisdiction |
| Predicted New Indication | Candidiasis |
| TxGNN Prediction Score | 99.69% |
| Evidence Level | L5 |
| EU Market Status | ✗ Not marketed |
| Number of Authorizations | 0 |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Currently, detailed mechanism of action data is not available in the structured original_moa field (marked as a data gap). However, based on mechanistic notes elsewhere in the evidence pack, pretomanid is a nitroimidazooxazine prodrug that requires activation by a mycobacteria-specific F420-coenzyme-dependent deazaflavin nitroreductase system. This activation pathway is unique to mycobacteria and is central to pretomanid’s bactericidal activity against both replicating and non-replicating M. tuberculosis.
Candida species are fungi and do not possess this mycobacterial F420-dependent activation system. The evidence pack’s own rationale for this prediction states explicitly: “No plausible mechanism: Pretomanid’s antimicrobial activity depends on the mycobacteria-specific F420 coenzyme reductase system; Candida (a fungus) does not have this metabolic pathway, and there is no clinical or preclinical evidence supporting this.”
Given the complete absence of clinical trials, literature, or mechanistic rationale, this prediction should be treated as a likely embedding-similarity artifact of the TxGNN model rather than a genuine repurposing signal. For context, two other predictions in this evidence pack (leprosy, rank 2; HIV infectious disease, rank 6) did generate clinical trial and literature hits, but both were also found to lack support on closer review: a direct in vitro study (PMID 17005816) shows M. leprae is naturally resistant to pretomanid (PA-824), and the HIV-related trials/literature concern TB treatment in HIV-co-infected patients rather than any direct antiretroviral activity of pretomanid. No candidate in this evidence pack currently clears the bar for further development.
Clinical Trial Evidence
Currently no related clinical trials registered.
Literature Evidence
Currently no related literature available.
EU Market Information
Pretomanid is currently not marketed in this jurisdiction (total authorizations: 0). No license records are available to summarize.
Safety Considerations
Please refer to the SmPC for safety information.
(Note: key_warnings, contraindications, and drug-drug interaction data are all marked as data gaps in this evidence pack, including a Blocking-severity gap for TFDA label warnings/contraindications.)
Conclusion and Next Steps
Decision: Hold
Rationale:
- The top-ranked predicted indication (candidiasis) has no supporting clinical trials, no literature, and an explicitly stated implausible mechanism (fungal vs. mycobacteria-specific drug activation pathway). This is best interpreted as model noise rather than a genuine repurposing signal.
- The only two predictions in this evidence pack with actual trial/literature support (leprosy, HIV) were independently found to be non-viable — one due to direct evidence of natural drug resistance, the other due to a data-labeling mismatch (TB treatment in HIV+ patients, not anti-HIV activity).
To proceed, the following is needed:
- Resolve the Blocking-severity data gap (DG001): TFDA/official label warnings and contraindications must be obtained before any safety pre-screening (S1) can begin for any candidate.
- Resolve the High-severity data gap (DG002): formal MOA data from DrugBank API to support or refute mechanistic plausibility assessments.
- If candidiasis repurposing is to be pursued further, in vitro antifungal susceptibility data for pretomanid against Candida spp. would be the minimum evidence needed to justify moving past S0.
- No further action recommended on this candidate set without new primary evidence.
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.