Dronedarone
| Evidence Level: L1 | Predicted Indications: 10 |
Table of Contents
Dronedarone: From Atrial Fibrillation to Stroke
One-Sentence Summary
Dronedarone is a Class III antiarrhythmic agent used to maintain sinus rhythm in patients with atrial fibrillation (AF) or atrial flutter. The TxGNN model predicts it may reduce the risk of Stroke, as a downstream benefit of rhythm control in AF, with 19 clinical trials and 20 publications — including Phase 3/4 RCT data — currently supporting this direction.
⚠️ This candidate carries a Blocking data gap (no TFDA label/warning data available) and mixed efficacy/safety signals across trials — see Conclusion below.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Atrial Fibrillation / Atrial Flutter (established use, per clinical trial and literature context — TFDA license text unavailable) |
| Predicted New Indication | Stroke (TxGNN term: “stroke disorder”) |
| TxGNN Prediction Score | 99.97% |
| Evidence Level | L1 |
| Taiwan Market Status | ✗ Not Marketed (Not marketed) |
| Number of Authorizations | 0 |
| Recommended Decision | Proceed with Guardrails |
All TxGNN Predicted Indications for Dronedarone
Because this evidence pack bundles multiple candidate indications (TW-DB04855-multi), the full ranked list is shown for transparency. Only rank 1 (and, secondarily, rank 4, which is mechanistically identical) has actionable evidence; the rest are AI-only signals or likely embedding-space noise.
| Rank | Predicted Disease | TxGNN Score | Evidence Level | Recommendation |
|---|---|---|---|---|
| 1 | Stroke (“stroke disorder”) | 99.97% | L1 | Proceed with Guardrails |
| 2 | Obsolete susceptibility to ischemic stroke | 99.96% | L5 | Hold (obsolete/non-clinical UMLS term) |
| 3 | ABri amyloidosis | 99.92% | L5 | Hold (no biological rationale) |
| 4 | Cerebrovascular disorder | 99.71% | L1 | Proceed with Guardrails (overlaps with Rank 1) |
| 5 | Brain stem infarction | 99.71% | L5 | Hold (no dedicated evidence) |
| 6 | Sick sinus syndrome 2, autosomal dominant | 99.63% | L5 | Hold — safety signal, not an indication (dronedarone is negatively chronotropic; caution/contraindication territory) |
| 7 | Duodenal obstruction | 99.56% | L5 | Hold (no plausible mechanism) |
| 8 | Cerebral artery occlusion | 99.44% | L4 | Research Question (only indirect trial support) |
| 9 | Sarcoglycanopathy | 99.43% | L5 | Hold (no biological rationale) |
| 10 | Wildervanck syndrome | 99.37% | L5 | Hold (no biological rationale) |
Why is This Prediction Reasonable?
Currently, detailed mechanism-of-action data from DrugBank is not available for this candidate (Data Gap DG002). Based on known pharmacology, Dronedarone is a Class III, multichannel-blocking antiarrhythmic — structurally related to amiodarone but without the iodine moiety — that restores and maintains sinus rhythm in AF/atrial flutter, with additional negative chronotropic, antiadrenergic, and mild sodium/calcium-channel blocking effects.
The proposed new indication, stroke, is not a direct pharmacological target of dronedarone but a downstream consequence of its established AF indication: atrial fibrillation is the leading cause of cardioembolic stroke, and by preventing AF recurrence and sustaining sinus rhythm, dronedarone reduces left atrial blood stasis and thrombus formation — the proximate cause of AF-related stroke. This is consistent with the mechanistic rationale supplied with the prediction: “restoring and maintaining sinus rhythm reduces intra-atrial thrombus formation associated with AF, thereby lowering the risk of cardioembolic stroke.”
This pathway is directly supported by post-hoc analyses of the ATHENA trial and by the EAST-AFNET 4 trial (NCT01288352), both of which link early rhythm-control therapy to reduced cardiovascular/cerebrovascular events. However, the effect is AF-subtype dependent: the PALLAS trial (NCT01151137, Phase 3, terminated) found dronedarone increased the risk of stroke, MI, and death in patients with permanent AF and additional risk factors. This makes the mechanism plausible and evidence-backed for paroxysmal/persistent AF populations, but it defines a clear safety boundary that must guardrail any repurposing pathway.
Clinical Trial Evidence
| Trial Number | Phase | Status | Enrollment | Key Findings |
|---|---|---|---|---|
| NCT01151137 | Phase 3 | Terminated | 3,236 | PALLAS trial: dronedarone 400mg BID added to standard therapy in permanent AF with risk factors — stopped early due to a significant increase in stroke, MI, and cardiovascular death. Key negative safety signal defining the population boundary for use. |
| NCT01288352 | Phase 4 | Completed | 2,789 | EAST-AFNET 4: early rhythm-control therapy (including dronedarone) vs. usual care in AF — evaluated prevention of AF-related complications including stroke. |
| NCT01856075 | N/A | Completed | 1,015 | International observational cohort comparing real-world effectiveness of dronedarone vs. other antiarrhythmics, including cardiovascular/stroke-related outcomes. |
| NCT05130268 | Phase 4 | Completed | 339 | Pragmatic RCT of early dronedarone vs. usual care in first-detected AF, evaluating whether earlier antiarrhythmic therapy alongside stroke-prevention therapy improves outcomes. |
| NCT01266681 | N/A | Unknown | 100 | Amiodarone vs. dronedarone for maintenance of sinus rhythm post-cardioversion in persistent AF, a population with elevated stroke risk. |
| NCT00911508 | N/A | Completed | 2,204 | CABANA trial: catheter ablation vs. antiarrhythmic drug therapy (dronedarone as one arm) for AF, with cardiovascular/stroke-relevant follow-up. |
| NCT02618577 | Phase 3 | Terminated | 2,608 | NOAH-AFNET 6: anticoagulation vs. current therapy for stroke/systemic embolism prevention in atrial high-rate episodes; relevant background trial for the AF–stroke pathway. |
| NCT05293080 | Phase 3 | Not yet recruiting | 1,746 | Early comprehensive rhythm-control therapy vs. usual care specifically in patients with acute ischemic stroke and AF — direct future test of the stroke-prevention hypothesis. |
| NCT06096337 | N/A | Active, not recruiting | 484 | Pulsed field ablation vs. antiarrhythmic drug therapy (dronedarone as a comparator) as first-line treatment for persistent AF. |
| NCT02294955 | N/A | Unknown | 152 | Catheter ablation vs. optimized pharmacological therapy for symptomatic AF, informing relative-risk context for drug-based rhythm control. |
Literature Evidence
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 40387892 | 2025 | RCT (post-hoc) | Clin Res Cardiol | Long-term safety of amiodarone and dronedarone for early rhythm control in the EAST-AFNET 4 trial population. |
| 22082198 | 2011 | RCT (primary publication) | N Engl J Med | PALLAS trial: dronedarone in high-risk permanent AF — study halted due to increased major vascular events (stroke, MI, death). Defines the key safety boundary. |
| 20730068 | 2010 | Review | Vasc Health Risk Manag | Regulatory approval review of dronedarone; cites ATHENA trial post-hoc analysis suggesting reduced stroke risk in the approved (non-permanent) AF population. |
| 28496906 | 2013 | Cohort/Pharmacovigilance | J Atr Fibrillation | Real-world comparison of cardiovascular events, stroke, heart failure, and liver injury: dronedarone vs. amiodarone and other antiarrhythmics. |
| 37485722 | 2023 | Cohort (comparative effectiveness) | Circ Arrhythm Electrophysiol | Head-to-head comparison of dronedarone vs. sotalol in antiarrhythmic-naive veterans with AF. |
| 35293087 | 2022 | Cohort/Subgroup (post-hoc) | Eur J Heart Fail | ATHENA post-hoc analysis: dronedarone in AF/atrial flutter with concomitant HFpEF/HFmrEF. |
| 33888353 | 2021 | Cohort (real-world) | Clin Ther | Evaluates digitalis intoxication risk from concomitant dronedarone + digoxin use — relevant drug-interaction signal. |
| 28992468 | 2017 | Mechanistic/basic research | Atherosclerosis | Dronedarone shows anticoagulant/antiplatelet effects independent of its antiarrhythmic action, offering a possible additional mechanistic link to stroke reduction. |
| 22920480 | 2012 | Review | Curr Cardiol Rev | Broader review of stroke prevention concepts and controversies in atrial fibrillation. |
| 22166900 | 2012 | Review | Lancet | General review of AF management, including stroke risk stratification and thromboprophylaxis context. |
Taiwan Market Information
Dronedarone is currently not marketed in Taiwan (Not marketed); the evidence pack records 0 TFDA product licenses, so no approved-indication text, dosage form, or SmPC data is available locally. Any repurposing pathway would need to first establish a regulatory route (e.g., named-patient import, new drug application, or reference to EMA/FDA labeling) before a Taiwan-specific authorization table can be produced.
Safety Considerations
Official TFDA warnings, contraindications, and drug-interaction data are not yet available for this candidate (Data Gap DG001, Blocking severity — required before this candidate can enter Stage 1 safety screening). Please refer to the SmPC (from a market where dronedarone is approved) for authoritative safety information.
Evidence-derived safety signals (from clinical trial/literature evidence in this pack, not official labeling):
- AF-subtype dependent risk: the PALLAS trial (NCT01151137) was terminated early after dronedarone significantly increased stroke, MI, and cardiovascular death in patients with permanent AF — the drug should not be assumed protective outside paroxysmal/persistent AF populations.
- Sinus node caution: dronedarone’s negative chronotropic effect makes sick sinus syndrome a mechanistic contraindication/caution area (flagged independently by TxGNN as rank 6, “sick sinus syndrome 2, autosomal dominant”).
- Digoxin interaction: real-world data (PMID 33888353) indicate increased digitalis intoxication risk with concomitant dronedarone + digoxin use, likely via P-glycoprotein inhibition.
Conclusion and Next Steps
Decision: Proceed with Guardrails
Rationale: Multiple Phase 3/4 trials (EAST-AFNET 4, PALLAS, real-world cohorts) and ATHENA post-hoc analyses directly evaluate dronedarone’s effect on stroke-related outcomes in AF patients (L1 evidence), supporting a plausible and mechanistically consistent repurposing pathway. However, evidence is bidirectional — protective in paroxysmal/persistent AF, harmful in permanent AF (PALLAS) — so any use for stroke prevention must be strictly bounded to the appropriate AF subtype.
To proceed, the following is needed:
- TFDA-equivalent warnings, contraindications, and DDI data (Data Gap DG001, Blocking — currently prevents entry into Stage 1 safety screening)
- Confirmed mechanism-of-action documentation from DrugBank (Data Gap DG002)
- A regulatory pathway assessment, since Dronedarone currently holds zero Taiwan licenses (Not marketed)
- An explicit patient-selection guardrail excluding permanent AF, based on the PALLAS safety signal
- A formal evidence plan (e.g., real-world study or trial) using stroke incidence as a primary endpoint, since existing trials treat stroke as a secondary/composite outcome rather than the primary target
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.