Tadalafil

證據等級: L5

預測適應症: 8 個

Tadalafil: From an Undocumented Original Indication to Ambras Type Hypertrichosis Universalis Congenita

One-Sentence Summary

Tadalafil is a PDE5 inhibitor (per the evidence pack's own rationale text, already approved for WHO Group 1 pulmonary arterial hypertension); its original indication record for this evaluation is not documented. The TxGNN model's top-ranked prediction is Ambras type hypertrichosis universalis congenita, a rare congenital multi-hair syndrome, but this prediction is currently supported by 0 clinical trials and 0 publications — the model's own rationale flags it as a likely embedding-space artifact rather than a mechanistic signal.

Quick Overview

Item	Content
Original Indication	Not documented in this evidence pack (Finland licensing data absent)
Predicted New Indication	Ambras type hypertrichosis universalis congenita
TxGNN Prediction Score	99.98%
Evidence Level	L5 (model prediction only, no supporting studies)
Finland Market Status	Not marketed (未上市)
Number of Authorizations	0
Recommended Decision	Hold

Why is This Prediction Reasonable?

Detailed mechanism-of-action data for tadalafil is flagged as a gap in this evidence pack (DG002). The only mechanistic context available comes from the model's own rationale text across the eight candidate indications: tadalafil is described there as a PDE5 inhibitor that raises cGMP levels and relaxes vascular smooth muscle, and is already an approved therapy for WHO Group 1 pulmonary arterial hypertension.

For the top-ranked candidate — Ambras type hypertrichosis universalis congenita, a rare genetic syndrome typically linked to chromosome 8q rearrangements — the evidence pack's rationale explicitly states there is no known pathological connection to PDE5 inhibition, cGMP signaling, or vascular smooth muscle relaxation. It characterizes the high TxGNN score as most likely an artifact of embedding-space similarity rather than a genuine mechanistic inference.

A weaker, secondary hypothesis appears for the related rank-2 prediction (hypertrichosis, general): topical vasodilators such as minoxidil can promote follicular growth, and PDE5 inhibitors share vasodilatory pharmacology in principle. However, this remains speculative extrapolation with no clinical or case-report support in the current evidence, and does not apply to the rank-1 Ambras-syndrome prediction, which is a distinct and much rarer genetic entity.

Clinical Trial Evidence

Currently no related clinical trials registered.

Literature Evidence

Currently no related literature available.

Finland Market Information

Tadalafil is currently not marketed in Finland per this evidence pack (0 authorizations recorded), so no product/license table can be generated.

Safety Considerations

Please refer to the package insert for safety information. (Key warnings, contraindications, and drug interaction data are all currently unavailable — obtaining the TFDA/Fimea package insert is flagged as a blocking data gap, DG001, required before any safety pre-screening can proceed.)

Other Predicted Signals & Safety Note

The evidence pack screened 8 TxGNN predictions for tadalafil; all scored in a similarly high but narrow range (99.42%–99.98%), and all but two returned zero supporting literature/trials:

Kyphoscoliotic heart disease (rank 7, L4, stage S1 "Research Question") is the most mechanistically plausible of the set: severe kyphoscoliosis can cause restrictive lung disease with Group 3 pulmonary hypertension, and tadalafil is an approved Group 1 PAH therapy. However, PDE5 inhibitors in hypoxia/lung-disease-related (Group 3) PH are controversial — they can worsen ventilation-perfusion mismatch — and no direct trial evidence exists for this population.
Migraine with brainstem aura (rank 8, L4) surfaced one case report (PMID 17059442) describing tadalafil-associated/induced typical aura without headache — this is an adverse-effect signal, not therapeutic evidence, and the rationale explicitly warns TxGNN may have inverted the drug–disease relationship direction. This should be read as a caution, not a repurposing lead.
The "malformation syndrome with odontal/periodontal component" candidate returned 20 literature hits, but all are general periodontitis pathophysiology reviews with no tadalafil-specific content — a keyword co-occurrence artifact, not real evidence.
The remaining candidates (hair-shaft abnormality, Dandy-Walker syndrome, familial trichomegaly) have no known mechanistic link and no supporting data.

Conclusion and Next Steps

Decision: Hold

Rationale: The top-ranked prediction (Ambras type hypertrichosis) has zero clinical trials, zero literature, and no plausible mechanistic link per the model's own rationale — this is a model-score-only signal (L5) with a strong internal warning that it is a statistical artifact. No candidate among the 8 screened reaches even L2/L3 evidence.

To proceed, the following is needed:

Resolve the blocking gap (DG001): obtain the TFDA/Fimea package insert for warnings, contraindications, and DDI before any safety pre-screening (S1) can begin
Obtain confirmed original MOA and original indication data from DrugBank (DG002) to properly ground any repurposing rationale
If pursuing further, redirect evaluation toward the more mechanistically defensible candidate — kyphoscoliotic heart disease / Group 3 PH — rather than the top TxGNN-ranked but unsupported hypertrichosis prediction
Treat the migraine-aura literature signal as a pharmacovigilance flag for existing tadalafil use, not a repurposing opportunity
Disclaimer

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