Metformin
| Evidence Level: L5 | Predicted Indications: 5 |
Table of Contents
Metformin: From Undocumented Indication to Classic Stiff Person Syndrome
One-Sentence Summary
This evidence pack does not include documented original indications or mechanism of action data for Metformin, and the drug is currently not marketed in the UK according to this dataset. The TxGNN model’s top prediction is Classic Stiff Person Syndrome, but this signal is supported by 0 clinical trials and 0 publications — it is a purely computational (knowledge-graph) association with no corroborating clinical or mechanistic evidence.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Not documented in this evidence pack |
| Predicted New Indication | Classic Stiff Person Syndrome |
| TxGNN Prediction Score | 99.45% |
| Evidence Level | L5 |
| UK Market Status | Not marketed |
| Number of Marketing Authorisations | 0 |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Detailed mechanism of action data is not available for Metformin in this evidence pack, so a full pharmacological rationale cannot be constructed here.
Based on the model’s own rationale, the association is weak: Classic Stiff Person Syndrome is an autoimmune neurological condition mediated by anti-GAD65 antibodies causing GABAergic transmission dysfunction. This has no established biological connection to Metformin’s commonly cited mechanisms (AMPK activation, mitochondrial complex I inhibition). The high TxGNN score (99.45%) reflects graph co-occurrence patterns within the knowledge graph rather than a validated pharmacological pathway, and the model’s rank (5,380th overall) also indicates this is a low-confidence tail prediction rather than a strong signal.
It is also worth noting that the other four candidates in this evidence pack show similarly weak or even concerning patterns — one candidate (thiamine-responsive dysfunction syndrome) is a mitochondrial/metabolic disorder where Metformin’s complex I inhibition could theoretically worsen lactic acidosis risk, suggesting a potential contraindication direction rather than therapeutic benefit. None of the five predictions has any supporting clinical trial or literature evidence.
Clinical Trial Evidence
Currently no related clinical trials registered.
Literature Evidence
Currently no related literature available.
UK Market Information
No UK marketing authorisations are recorded for Metformin in this dataset (market status: not marketed, 0 licenses on file).
Safety Considerations
Please refer to the SmPC and BNF for safety information. Report suspected adverse reactions via the Yellow Card Scheme.
Note: This evidence pack flags a blocking data gap (DG001) — MHRA/manufacturer label warnings and contraindications are not yet available for this candidate, which prevents a full S1 safety review.
Conclusion and Next Steps
Decision: Hold
Rationale: The prediction is evidence level L5 (model prediction only) with no clinical trials or literature support, and the proposed mechanistic link is not biologically well-founded. A blocking data gap (missing regulatory safety label data) also prevents progression to a formal safety assessment at this stage.
To proceed, the following is needed:
- MHRA-equivalent SmPC/label data — warnings and contraindications (blocking gap, DG001)
- Confirmed mechanism of action data (DrugBank query, DG002)
- Documented original indication(s) and current global regulatory status for Metformin
- Preclinical or case-level evidence establishing a plausible biological link to stiff person syndrome spectrum disorders before further evaluation is warranted
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.