Diflunisal
| Evidence Level: L5 | Predicted Indications: 10 |
Table of Contents
- Diflunisal
- Diflunisal: From NSAID Pain and Arthritis Therapy to Acromesomelic Dysplasia, Hunter-Thompson Type
Diflunisal: From NSAID Pain and Arthritis Therapy to Acromesomelic Dysplasia, Hunter-Thompson Type
One-Sentence Summary
Diflunisal is a salicylate-derived NSAID (COX inhibitor) marketed in the US as generic tablets. The TxGNN model predicts it may be effective for acromesomelic dysplasia, Hunter-Thompson type, but no clinical trials and no publications support this prediction. The link most likely reflects knowledge-graph proximity rather than a real mechanism.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Not stated in the pack (approved-indication text is empty); diflunisal is generally used as an NSAID for pain and arthritis |
| Predicted New Indication | Acromesomelic dysplasia, Hunter-Thompson type |
| TxGNN Prediction Score | 99.99% (model rank 548) |
| Evidence Level | L5 |
| US Market Status | ✓ Marketed |
| Number of NDAs | 7 (the listed licenses are ANDAs) |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Currently, detailed mechanism of action data is not available in the pack. Diflunisal is an NSAID that inhibits COX-1 and COX-2 and reduces prostaglandin-mediated inflammation and pain.
The predicted disease is a monogenic skeletal dysplasia driven by the CDMP1/GDF5 pathway, which controls chondrogenic signaling. COX inhibition has no known effect on this pathway. I found no plausible mechanistic link. The high score most likely comes from the drug sitting close to musculoskeletal nodes in the knowledge graph, not from a causal relationship. Because the disease is a structural genetic defect, an NSAID could at most relieve secondary pain and would not modify the disease.
Clinical Trial Evidence
Currently no related clinical trials registered.
Literature Evidence
Currently no related literature available.
US Market Information
The pack lists 5 of the 7 licenses. Approved-indication text was not provided for any of them.
| Authorization Number | Product Name | Dosage Form | Approved Indication |
|---|---|---|---|
| ANDA202845 | Dolobid (INA Pharmaceutics Inc) | Tablet, film coated | Not provided |
| ANDA203547 | Diflunisal (Zydus Lifesciences Limited) | Tablet | Not provided |
| ANDA202845 | Diflunisal (Heritage Pharmaceuticals / Avet Pharmaceuticals) | Tablet, film coated | Not provided |
| ANDA073673 | Diflunisal (Teva Pharmaceuticals USA) | Tablet, film coated | Not provided |
| ANDA203547 | Diflunisal (Zydus Pharmaceuticals USA) | Tablet | Not provided |
Safety Considerations
Please refer to the package insert for safety information.
Conclusion and Next Steps
Decision: Hold
Rationale: There is no clinical or literature evidence (L5, model prediction only). The mechanism is implausible, since an NSAID does not act on GDF5/CDMP1 signaling. The score is most likely a graph-topology artifact.
Among the other top-10 predictions, ankylosing spondylitis has the most support (L3). It rests mainly on a 1986 comparison of diflunisal and phenylbutazone (PMID 3524970). NSAIDs are already established therapy for that disease, so it is a label-status question rather than a new repurposing finding. It is a better candidate for follow-up than the rank 1 prediction.
To proceed, the following is needed:
- Package insert warnings and contraindications, to complete safety screening
- Mechanism of action data from DrugBank
- Approved-indication text for the US labels
- For the ankylosing spondylitis lead: full-text review of PMID 3524970 to confirm randomization, blinding and sample size
- For the rank 1 disease: no action until independent preclinical or clinical evidence appears
This report is for research reference only and does not constitute medical advice. Repurposing candidates require clinical validation before any use.
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.