Highlights
Early initiation of direct oral anticoagulants (DOACs) within 4 days of acute ischemic stroke showed no differential effect based on whether atrial fibrillation (AF) was diagnosed before or after the qualifying stroke. The timing of AF diagnosis—prior to stroke onset versus newly identified during the acute event—does not appear to be a treatment-effect modifier for anticoagulation strategy. Similarly, AF subtype classification as paroxysmal versus persistent did not significantly alter outcomes with early versus delayed anticoagulation approaches. Persistent AF emerged as an independent predictor of adverse outcomes, associated with approximately twice the risk of recurrent ischemic stroke, symptomatic intracranial hemorrhage, or systemic arterial embolism compared with paroxysmal AF.
Background
Atrial fibrillation represents one of the most significant modifiable risk factors for cardioembolic stroke, accounting for approximately 15-20% of all ischemic cerebrovascular events. The optimal timing for initiating anticoagulation after an acute ischemic stroke in patients with AF remains a subject of ongoing clinical investigation and debate. Current guideline recommendations acknowledge the need to balance the competing risks of early anticoagulation—particularly hemorrhagic transformation of the infarct—against the heightened risk of early recurrent embolization in this patient population.
The OPTIMAS (Optimal Anticoagulation Timing in Atrial Fibrillation-Related Acute Ischemic Stroke) trial was designed to address this critical knowledge gap by comparing early DOAC initiation (within 4 days) with a delayed approach (days 7-14) in patients presenting with acute ischemic stroke and concurrent atrial fibrillation. This subgroup analysis specifically examined whether two clinically relevant factors—time of AF diagnosis and AF subtype—might influence the relative benefits and risks of early anticoagulation strategy.
The clinical importance of this investigation extends beyond academic interest. Patients in whom AF is first detected during hospitalization for acute stroke represent a distinct clinical scenario compared with those with a pre-existing AF diagnosis. Likewise, the distinction between paroxysmal and persistent AF carries implications for thrombotic risk, left atrial remodeling, and potentially treatment response. Understanding whether these factors modify anticoagulation treatment effects could inform more personalized therapeutic decision-making.
Study Design
OPTIMAS was conducted as a randomized, parallel-group, open-label trial with blinded outcome assessment, representing the gold standard methodology for comparative effectiveness research. The trial enrolled participants presenting with acute ischemic stroke and confirmed atrial fibrillation, randomizing them in a 1:1 allocation ratio to either early DOAC initiation (target initiation within 4 days of stroke symptom onset) or delayed DOAC initiation (initiation scheduled for days 7-14).
The prespecified subgroup analysis focused on two primary stratification factors. First, the time of AF diagnosis was classified as either “before stroke” (participants with known AF prior to the qualifying event) or “after stroke” (participants in whom AF was first identified during the acute hospitalization). Second, AF subtype was categorized as either paroxysmal (self-terminating episodes) or persistent (sustained episodes requiring medical intervention for termination).
The primary outcome composite comprised three major adverse events: recurrent ischemic stroke, symptomatic intracranial hemorrhage, and systemic arterial embolism. This composite endpoint was selected to capture the full spectrum of thrombotic and hemorrhagic complications relevant to anticoagulation management in this population.
Statistical analysis employed mixed effects logistic regression models incorporating interaction terms between subgroup variables and treatment allocation to formally test for effect modification. Multivariable logistic regression was utilized to investigate independent associations between AF characteristics and outcomes, adjusting for relevant confounders.
Key Findings
The analysis included 3,619 participants with a mean age of 78.0 years (standard deviation 9.9 years), and 45.3% were female. This population-based cohort reflects the typical demographics of AF-related stroke, occurring predominantly in older individuals with multiple vascular comorbidities.
Time of AF Diagnosis and Treatment Effect
Among participants with AF diagnosed before the qualifying stroke (n=1,838), the primary outcome occurred in 37 of 918 (4.0%) participants assigned to early DOAC compared with 32 of 920 (3.5%) assigned to delayed DOAC, yielding an odds ratio of 1.17 (95% confidence interval, 0.72-1.89). In contrast, participants with AF diagnosed after the stroke (n=1,781) demonstrated primary outcome rates of 22 of 895 (2.5%) with early DOAC versus 27 of 886 (3.0%) with delayed DOAC, corresponding to an odds ratio of 0.79 (95% confidence interval, 0.45-1.40). The P-value for interaction was 0.312, indicating no statistically significant evidence that timing of AF diagnosis modifies the treatment effect of early anticoagulation.
These findings suggest that clinicians should not base anticoagulation timing decisions on whether AF represents a new diagnosis made during the acute stroke admission or a pre-existing condition. The risk-benefit profile of early versus delayed initiation appears similar across both scenarios.
AF Subtype and Treatment Effect
Analysis of AF subtype revealed no significant modification of treatment effect. For participants with persistent AF, the odds ratio for early versus delayed DOAC was 1.06 (95% CI, 0.71-1.58), while for those with paroxysmal AF, the corresponding odds ratio was 0.66 (95% CI, 0.25-1.72). The P-value for interaction was 0.377, confirming that AF subtype does not appear to influence the relative benefits of early anticoagulation.
Independent Predictors of Outcome
Perhaps the most clinically significant finding concerned the independent association between AF subtype and outcomes. Persistent AF demonstrated a robust and statistically significant association with increased primary outcome risk compared with paroxysmal AF, with an adjusted odds ratio of 2.10 (95% CI, 1.19-3.68). This finding persisted after adjustment for multiple potential confounders, suggesting that persistent AF confers substantially higher thrombotic risk independent of anticoagulation timing strategy.
The time of AF diagnosis, by contrast, was not independently associated with outcome events in multivariable analysis. This suggests that the discovery of AF during acute stroke hospitalization, rather than representing a distinct high-risk phenotype, likely reflects similar underlying cardiac pathology that may have been intermittently present but undetected.
Expert Commentary
These findings carry important implications for clinical practice and future research directions. The absence of effect modification by AF diagnosis timing provides reassurance that the discovery of new AF during acute stroke hospitalization should not alter anticoagulation management strategies. Clinicians can apply evidence from early versus delayed anticoagulation trials to this growing patient population with similar confidence as to those with established AF diagnoses.
The independent association between persistent AF and doubled outcome risk despite anticoagulation raises intriguing questions about the underlying mechanisms of cardioembolic stroke in AF. Persistent AF is characterized by sustained atrial remodeling, persistent blood stasis within the left atrium, and ongoing thrombus formation risk. These pathophysiological changes may not be fully captured by standard stroke risk stratification tools, which often treat AF as a binary variable regardless of subtype.
From a practical standpoint, these results may inform more nuanced risk stratification approaches that incorporate AF burden and subtype classification. Patients with persistent AF might warrant more aggressive secondary prevention strategies, closer monitoring, or earlier and more sustained anticoagulation intensification.
Several limitations merit consideration. The subgroup analyses, while prespecified, may have been underpowered to detect modest treatment-effect modifications. The open-label design introduces potential bias in outcome assessment, although blinded outcome adjudication partially mitigates this concern. Additionally, the trial population comprised predominantly older adults, and generalizability to younger patients with AF-related stroke requires further investigation.
Conclusion
The OPTIMAS subgroup analysis provides valuable evidence that neither the timing of AF diagnosis nor AF subtype significantly modifies the treatment effect of early versus delayed DOAC initiation following acute ischemic stroke. These findings support a unified approach to anticoagulation timing decisions regardless of AF diagnosis circumstances or subtype classification.
However, the discovery that persistent AF carries approximately double the independent risk of adverse outcomes compared with paroxysmal AF represents a critical insight for risk stratification. This finding suggests that AF burden and permanence merit consideration in future stroke prevention strategies and clinical decision-making frameworks.
Further research is needed to elucidate optimal anticoagulation strategies tailored to AF subtype, to investigate mechanisms underlying the differential risk between persistent and paroxysmal AF, and to determine whether additional interventions may reduce the elevated risk observed in persistent AF despite anticoagulation.
Funding and Registration
This study was registered at ClinicalTrials.gov (NCT03759938). Full funding details and complete author affiliations are available in the original publication.
References
1. Lyon J, Nash PS, Ahmed N, et al. Early Versus Delayed Anticoagulation in Acute Ischemic Stroke According to Atrial Fibrillation Subtype and Time of Diagnosis: Subgroup Analysis of the OPTIMAS Randomized Controlled Trial. Stroke. 2026-04-01. PMID: 41919368.
2. Hart RG, Diener HC, Coutts SB, et al. Embolic strokes of undetermined source: the case for a new clinical construct. Lancet Neurol. 2014;13(4):429-438.
3. January CT, Wann LS, Calkins H, et al. 2019 AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation. Circulation. 2019;139(11):e1-e492.
