Dofetilide

證據等級: L5

預測適應症: 10 個

Dofetilide: From Atrial Fibrillation to Stroke Disorder

One-Sentence Summary

Dofetilide is a Class III antiarrhythmic (selective Kv11.1/hERG/IKr potassium-channel blocker) originally indicated for pharmacologic cardioversion and maintenance of sinus rhythm in atrial fibrillation/atrial flutter. The TxGNN model predicts a very high association with Stroke Disorder (score 99.99%), and 5 clinical trials and 8 publications are currently linked to this pathway — however, on closer reading the evidence does not show dofetilide preventing stroke; it largely reflects the known AF–stroke co-occurrence in the knowledge graph, with one key trial (AFFIRM) actually showing rhythm-control strategies do not reduce stroke risk. This candidate should be treated as low-confidence pending further review.

Quick Overview

Item	Content
Original Indication	Atrial fibrillation / atrial flutter (rhythm conversion and maintenance of sinus rhythm) — per mechanistic evidence in this pack; not confirmed via a formal Taiwan/US label (see Data Gaps)
Predicted New Indication	Stroke Disorder
TxGNN Prediction Score	99.99% (rank 281 of all candidates)
Evidence Level	L3
US Market Status	✗ Not Marketed
Number of NDAs	0
Recommended Decision	Hold

Why is This Prediction Reasonable?

Dofetilide is a selective Kv11.1 (hERG/IKr) potassium-channel blocker — a Class III antiarrhythmic approved for pharmacologic cardioversion of atrial fibrillation/atrial flutter and for maintenance of sinus rhythm afterward. (Note: this mechanism-of-action description is drawn from the evidence pack’s repurposing rationale; the formal original_moa field is flagged as a data gap — DG002 — pending DrugBank confirmation.)

The link to “Stroke Disorder” is mechanistically indirect. TxGNN’s high score appears to reflect the well-known clinical association between atrial fibrillation and cardioembolic stroke, rather than any direct pharmacological effect of dofetilide on stroke pathophysiology. Critically, this pack’s own literature includes the AFFIRM trial sub-analysis (PMID 15007003), which found that a rhythm-control strategy (which includes Class III agents like dofetilide) offered no reduction in stroke or death compared with a rate-control strategy — both arms still required anticoagulation. This is evidence against interpreting rhythm restoration as a stroke-prevention mechanism.

In addition, dofetilide carries a well-established risk of QT prolongation and Torsades de Pointes, requiring in-hospital initiation and ECG monitoring — a safety burden that itself argues for caution rather than expanded use. One literature entry (PMID 30700466) even describes a case of dofetilide-associated facial paralysis being mistaken for stroke after cardioversion, underscoring how AF/stroke-adjacent search terms can surface adverse-event confounders rather than efficacy signals. Overall, the mechanistic case for repurposing dofetilide specifically for stroke is weak and is better explained as a knowledge-graph co-occurrence artifact than a therapeutic hypothesis.

Clinical Trial Evidence

Trial Number	Phase	Status	Enrollment	Key Findings
NCT00911508	N/A (catheter intervention)	Completed	2,204	CABANA trial: catheter ablation vs. antiarrhythmic drug therapy (dofetilide among comparator drugs) for AF; primary endpoint was a death/stroke/bleeding composite, not a dofetilide-specific stroke endpoint (relevance grade B).
NCT06096337	N/A	Active, not recruiting	484	Pulsed field ablation vs. antiarrhythmic drugs (dofetilide a possible comparator) as first-line therapy for persistent AF; endpoints focus on rhythm control and ablation safety, not stroke outcomes (relevance grade B).
NCT06783868	N/A	Not yet recruiting	100	SAVE STROKE Phase II: neurological outcomes after AF ablation vs. routine medication in recent-stroke patients; does not use dofetilide as an intervention (relevance grade C).
NCT00392106	Phase 3	Suspended	240	Device (HIFU pulmonary vein ablation) trial for paroxysmal AF vs. best medical therapy; a medical-device study, not a dofetilide drug trial (relevance grade C).
NCT05034432	Phase 4	Recruiting	100	Prophylactic ventricular arrhythmia ablation in high-risk LVAD candidates; population and endpoint (ventricular arrhythmia, not stroke/AF) have low relevance (relevance grade C).

None of these trials directly evaluate dofetilide’s effect on stroke incidence; at best two (grade B) include dofetilide as a comparator arm in AF rhythm-control studies.

Literature Evidence

PMID	Year	Type	Journal	Key Findings
15007003	2004	RCT (post-hoc/substudy)	Circulation	AFFIRM sub-analysis: rhythm-control strategy (incl. Class III drugs) showed no survival or stroke-risk advantage over rate control; sinus rhythm itself was associated with lower mortality only in an on-treatment analysis.
21955243	2012	Cohort	J Cardiovasc Electrophysiol	Dofetilide reduced frequency of ventricular arrhythmias and ICD therapies in patients with implanted defibrillators — an off-label use, not an approved indication.
32538135	2020	Cohort	Circ Arrhythm Electrophysiol	Dofetilide use in AF patients with reduced LVEF being considered for ICD; describes safety outcomes and LVEF improvement, not stroke prevention.
26233885	2016	Comparative effectiveness/Cohort	J Cardiol	Comparative effectiveness of antiarrhythmic drugs (including dofetilide) for AF rhythm control; limited comparative outcome data available.
11445058	2001	Review	Curr Treat Options Cardiovasc Med	General review of atrial flutter management strategies, including rate/rhythm control goals.
20638626	2010	Review	Gend Med	Review of gender differences in AF epidemiology and management.
11174354	2001	Review	Am Heart J	Review of pharmacologic AF management strategies and embolic-risk context.
32435191	2020	Preclinical (animal model)	Front Pharmacol	Pig model studying KCa2/Kv11.1 channel inhibition and AF conversion; mechanistic, not clinical, evidence.

No literature directly demonstrates a stroke-preventive or stroke-treatment effect of dofetilide; the strongest-tier evidence (AFFIRM, tier 1) argues against the rhythm-control-prevents-stroke hypothesis.

US Market Information

Dofetilide is currently not marketed under this evidence pack’s regulatory dataset (market_status: 未上市, 0 licenses recorded). No NDA/license records are available to summarize.

Safety Considerations

Please refer to the package insert for safety information. Formal TFDA warnings, contraindications, and drug-interaction data could not be retrieved for this evaluation (data gap DG001, severity: Blocking — this prevents the candidate from advancing past the S1 safety screening stage). Known class-level concerns (QT prolongation, Torsades de Pointes risk requiring in-hospital initiation) are referenced qualitatively in the mechanistic literature above but are not yet backed by a verified label source in this pack.

Conclusion and Next Steps

Decision: Hold

Rationale: The very high TxGNN score is not corroborated by direct clinical or mechanistic evidence of a stroke-preventive effect — the strongest available literature (AFFIRM) actually contradicts the rhythm-control-prevents-stroke hypothesis, and a blocking data gap (missing TFDA label/warnings, DG001) prevents even a basic safety screen.

To proceed, the following is needed:

Resolve DG001 (Blocking): obtain and parse the TFDA/official label for warnings and contraindications before any S1 safety evaluation can proceed
Resolve DG002 (High): confirm mechanism-of-action data via DrugBank API rather than relying on rationale-text inference
Drug-interaction (DDI) data, currently unavailable (query_status: not_found)
A mechanistic study or clinical endpoint that isolates dofetilide’s effect on stroke incidence independent of general AF rhythm-control strategy
Re-evaluation against the AFFIRM-type negative evidence before considering advancement beyond Hold
Disclaimer

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