Empagliflozin

證據等級: L5

預測適應症: 3 個

Empagliflozin: From Type 2 Diabetes Mellitus to Classic Stiff Person Syndrome

One-Sentence Summary

Empagliflozin is a well-known SGLT2 (sodium-glucose cotransporter-2) inhibitor whose original indication (type 2 diabetes mellitus and related cardiorenal use) is not captured in this evidence pack. The TxGNN model predicts possible efficacy in Classic Stiff Person Syndrome, but the prediction is currently supported by 0 clinical trials and 0 publications — this is a model-only signal (Evidence Level L5) with a "Hold" recommendation.

Quick Overview

Item	Content
Original Indication	Not recorded in this evidence pack (data gap — `original_indications` is empty). Empagliflozin is publicly known as an SGLT2 inhibitor for type 2 diabetes mellitus.
Predicted New Indication	Classic Stiff Person Syndrome
TxGNN Prediction Score	99.06%
Evidence Level	L5 (model prediction only, no supporting studies)
Taiwan Market Status	✗ 未上市 (Not Marketed)
Number of Authorizations	0
Recommended Decision	Hold

Why is This Prediction Reasonable?

Currently, detailed mechanism of action data is not available (flagged in this evidence pack as data gap DG002, severity: High). Based on known information, Empagliflozin acts as an SGLT2 inhibitor, reducing renal glucose reabsorption; its established action is confined to renal glucose transport physiology and downstream metabolic/cardiorenal effects.

Classic Stiff Person Syndrome (SPS) is an autoimmune neurological disorder driven primarily by anti-GAD65 antibodies that impair GABAergic neurotransmission. There is no established mechanistic overlap between SGLT2 inhibition and GABAergic signaling or autoimmune neuro-modulation. Because original_moa is a data gap, we cannot cross-validate whether an indirect (e.g., metabolic–neuroimmune) pathway underlies this prediction.

Two observations weaken confidence in this specific signal: (1) the near-identical TxGNN scores between rank 1 (classic stiff person syndrome, 99.06%) and rank 2 (focal stiff limb syndrome, 99.06%) suggest the model is scoring an entire disease cluster rather than generating an indication-specific signal; and (2) no clinical trials, registry entries, or literature exist to corroborate the prediction. The score most likely reflects knowledge-graph node-embedding proximity (e.g., topological closeness to other neuromuscular or metabolic-comorbidity nodes) rather than a validated pharmacological rationale.

Clinical Trial Evidence

Currently no related clinical trials registered.

Literature Evidence

Currently no related literature available.

Taiwan Market Information

No Taiwan market authorization records are currently available for Empagliflozin in this evidence pack (total_licenses = 0, market status: 未上市 / Not Marketed).

Safety Considerations

Please refer to the package insert for safety information.

Note: this evidence pack flags DG001 (TFDA package insert warnings/contraindications — severity: Blocking), meaning key warnings, contraindications, and drug-drug interaction data could not be retrieved. This blocks progression to the S1 safety pre-screening stage and must be resolved before any further evaluation.

Additional Predicted Indications (Same Evidence Pack)

Two further TxGNN predictions were returned alongside the primary candidate. Both share the same evidentiary profile — no clinical trials, no literature, Evidence Level L5, decision stage S0, recommendation Hold — and are noted here for completeness:

Rank	Disease	TxGNN Score	Note
2	Focal Stiff Limb Syndrome	99.06%	Near-identical score to rank 1; likely reflects the same disease-cluster embedding, not a distinct signal
3	Opsismodysplasia	99.03%	Rare pediatric skeletal dysplasia (INPPL1/SHIP2-driven); any link to SGLT2 inhibition is speculative and unsupported by data

Conclusion and Next Steps

Decision: Hold

Rationale: The prediction rests solely on a TxGNN model score (Evidence Level L5) with zero corroborating clinical trials or literature, and the drug's own mechanism of action data is unavailable. A Blocking-severity data gap (TFDA warnings/contraindications) also prevents even a preliminary safety assessment, so this candidate cannot advance past S0.

To proceed, the following is needed:

Resolve DG001: retrieve TFDA package insert (warnings, contraindications, DDI) to enable S1 safety pre-screening
Resolve DG002: obtain confirmed original MOA from DrugBank/product label to assess mechanistic plausibility
Establish original indication and Taiwan/international licensing status to support the "original vs. predicted indication" comparison
Conduct a targeted literature/preclinical search for any SGLT2i–GABAergic or SGLT2i–autoimmune neurological mechanism, given none currently exists
Given the clustering pattern across ranks 1–2, consider re-examining whether the underlying knowledge graph node structure is driving a shared false-positive signal across the stiff-person-syndrome spectrum before committing further evaluation resources
Disclaimer

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