← Back
NeuroTrials.ai
Neurology Clinical Trial Database

STROKE-AF 3-Year

Atrial Fibrillation In Patients With Stroke Attributed to Large- or Small-Vessel Disease: 3-Year Results From the STROKE AF Randomized Clinical Trial

Year of Publication: 2023

Authors: Richard A. Bernstein, Hooman Kamel, Christopher B. Granger, ..., for the STROKE AF Investigators

Journal: JAMA Neurology

Citation: JAMA Neurol. 2023;80(12):1277-1283

Link: https://clinicaltrials.gov/ct2/show/NCT02700945


Clinical Question

Does the superiority of ICM over standard monitoring for detecting atrial fibrillation persist beyond 1 year in patients with stroke attributed to large- or small-vessel disease, and for how long does AF detection continue to increase?

Bottom Line

In patients with ischemic stroke attributed to LAD or SVD, the risk of AF detection continues to rise beyond 1 year, reaching 21.7% at 3 years with ICM monitoring vs only 2.4% with usual care (HR 10.0). Most AF episodes are asymptomatic and would be missed by standard intermittent monitoring. One year of negative monitoring should not reassure clinicians that patients will not develop AF over subsequent years.

Major Points

  • At 3 years, AF was detected in 21.7% of ICM patients vs 2.4% of controls (HR 10.0, 95% CI 4.0-25.2, P<0.001)
  • AF detection increased progressively: 2.6% at 1 month, 7.9% at 6 months, 12.5% at 12 months, 18.5% at 2 years, 21.7% at 3 years
  • 87% of AF detected at 3 years would have been missed with only 30 days of continuous monitoring
  • 88% of all AF episodes in the ICM group were asymptomatic
  • 67.4% of patients with AF had at least one episode lasting >1 hour; median longest episode was 176 minutes
  • AF burden progressed over time: 66% of patients with ≥6 minutes of AF had burden progression
  • Patients with CHF, LAE, BMI >30, or QRS >120ms had 4-fold higher AF detection (30.0% vs 8.6%, HR 4.1)
  • OAC prescription was higher in ICM group: 24.4% vs 8.0% (OR 3.7, P<0.001)
  • Recurrent stroke rates were similar: 17.0% ICM vs 14.1% control (HR 1.1, P=0.71)
  • 24% of patients with detected AF in ICM group were not anticoagulated

Design

Study Type: Randomized controlled trial, multicenter, parallel-group, open-label (prespecified long-term secondary outcome analysis)

Randomization: 1

Blinding: Open-label; not feasible due to nature of intervention

Enrollment Period: April 2016 to July 2019

Follow-up Duration: Up to 3 years (through July 2022); mean follow-up 29.4 months

Centers: 33

Countries: United States

Sample Size: 492

Analysis: Intention-to-treat analysis. Kaplan-Meier incidence estimates with Cox proportional hazards regression for hazard ratios. Multivariable analysis for AF predictors using variables with P<0.10 in univariate analysis. SAS version 9.4 and R version 4.2.1.


Inclusion Criteria

  • Age ≥60 years, or age 50-59 years with at least 1 additional stroke risk factor
  • Index ischemic stroke attributed to large-artery atherosclerosis (LAD) or small-vessel occlusion (SVD) per TOAST criteria
  • Index stroke within 10 days prior to ICM insertion
  • Stroke risk factors included: congestive heart failure, hypertension, diabetes, prior ischemic stroke >90 days before index stroke, or other ischemic vascular disease

Exclusion Criteria

  • Cryptogenic stroke or embolic stroke of undetermined source
  • Cardioembolic stroke
  • History of documented atrial fibrillation or atrial flutter
  • Known indication or contraindication for long-term oral anticoagulation

Arms

FieldControlICM
InterventionSite-specific usual care with intermittent cardiac monitoring at physician discretion (162 ECGs in 93 patients, 34 Holter/event recorders in 30 patients, 4 mobile cardiac telemetry in 4 patients = 200 total recordings)Insertable cardiac monitor (Reveal LINQ, Medtronic) inserted within 10 days of qualifying stroke. Continuous monitoring with automatic AF detection.
DurationUp to 3 yearsUp to 3 years or end of battery life

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Time to first detection of atrial fibrillation (episode lasting >30 seconds, adjudicated by Clinical Events Committee) through 3 years of follow-up (prespecified secondary outcome of original trial)Primary2.4% (5 patients)21.7% (46 patients)10<0.001
AF detection at 1 month (ICM group)SecondaryN/A2.6% (6 patients)
AF detection at 6 months (ICM group)SecondaryN/A7.9% (18 patients)
AF detection at 12 monthsSecondary~2%12.5% (28 patients)
AF detection at 2 years (ICM group)SecondaryN/A18.5% (40 patients)
AF detection in LAD subgroup at 3 yearsSecondary3.5% (4 patients)20.4% (25 patients)6.4<0.001
AF detection in SVD subgroup at 3 yearsSecondary1.0% (1 patient)23.4% (21 patients)24.7<0.001
OAC prescription through follow-upSecondary8.0%24.4%3.7<0.001
Recurrent ischemic/hemorrhagic strokeSecondary14.1%17.0%1.10.71
Recurrent TIASecondary3.7%5.4%1.480.38
Hemorrhagic strokeSecondary0.5%2.3%3.830.23
Asymptomatic AF episodes (ICM group)SecondaryN/A88%
AF episodes >1 hour (among those with AF in ICM group)SecondaryN/A67.4% (31/46 patients)
AF burden ≥1 hour/day at 3 years (ICM group)SecondaryN/A20.5%
Death through 3 years - ICM groupAdverseN/A26 patients (12 by 12 mo + 14 by 36 mo)
Death through 3 years - Control groupAdverse13 patients (5 by 12 mo + 8 by 36 mo)N/A

Subgroup Analysis

AF incidence was significantly higher in ICM vs control regardless of stroke subtype: LAD (20.4% vs 3.5%, HR 6.4) and SVD (23.4% vs 1.0%, HR 24.7). Within the ICM group, AF detection rates were similar between LAD and SVD subtypes (log-rank P=0.55). Patients with ≥1 predictor (CHF, LAE, BMI >30, QRS >120ms) had AF detection of 30.0% vs 8.6% without any predictors (HR 4.1, P<0.001).


Criticisms

  • Treatment for AF was neither randomized nor prescribed by study protocol, limiting conclusions about clinical impact
  • Reasons for OAC use/nonuse were not captured
  • Study was not powered to detect differences in recurrent stroke rates
  • 24% of patients with detected AF were not anticoagulated, limiting potential clinical benefit
  • Presumed etiology of index and recurrent stroke was determined by site investigators, not centrally adjudicated
  • Control group did not include recently approved noninvasive wearable devices for AF detection
  • Cannot exclude that some AF detected represents background asymptomatic AF unrelated to stroke
  • Clinical significance of short AF episodes (minutes long) remains uncertain
  • 36.2% of patients did not complete 3-year follow-up

Funding

Medtronic Inc

Based on: STROKE-AF 3-Year (JAMA Neurology, 2023)

Authors: Richard A. Bernstein, Hooman Kamel, Christopher B. Granger, ..., for the STROKE AF Investigators

Citation: JAMA Neurol. 2023;80(12):1277-1283

Content summarized and formatted by NeuroTrials.ai.