Device-detected atrial fibrillation and risk for stroke: an analysis of >10 000 patients from the SOS AF project (Stroke preventiOn Strategies based on Atrial Fibrillation information from implanted devices)

Boriani G, Glotzer TV, Santini M, et al.
Eur Heart J 2014 35: 508-516

Background

Current guidelines recommend risk-based use of anticoagulant therapy in patients with both paroxysmal and persistent atrial fibrillation (AF) [1]. Length and frequency of episodes of AF are highly variable in paroxysmal AF, and many episodes are only detected through routine physical examinations, pre-operative assessments, diagnostic logs of implanted devices, rather than clinically [2]. Guidelines do not specifically address anticoagulation use in clinically ‘silent’ AF.
The effect of duration and frequency of AF on stroke risk is unclear. Modern implanted devices now allow precise, continuous long-term monitoring of heart rhythms, including AF. Implanted devices with an atrial lead can measure the total time spent in AF per day, thus the daily atrial tachycardia or AF burden. This study pooled data from three studies of patients with a variety of such devices to examine the relationship between the maximum daily AF burden and the risk of ischaemic stroke. Data from the TRENDS and PANORAMA studies and the Italian ClinicalService Registry Project were used. In total, 10016 patients were included in this analysis, and median duration of follow-up was 24 months (IQR: 14-40 months).

Main results

• 43% of patients experienced at least 1 day with at least 5 min of AF burden. Median time to maximum AF burden was 6 months (IQR: 1.3-14).
• Increasing AF burden was significantly associated with increasing age, CHADS2 score and presence of prior stroke. Among people who experience ischaemic stroke during follow-up, 26 (46%) had had at least 5 min of AF burden prior to the event.
• Ischaemic stroke risk was modelled against AF burden as a continuous variable, to define HRs at any given maximum AF burden. After adjustment for the CHADS2 score and oral anticoagulation at baseline, a significant association was seen between daily AF burden and ischaemic stroke, with HR= 1.03 per hour (95%CI: 1.00-1.05, P=0.04).
  HR could be calculated as: exponential (0.0263 * AF burden daily maximum hours). A maximum AF burden of 6 hours then gives a HR of 1.17 (95%CI: 1.01-1.36) and of 12 hours gives 1.37 (95%CI: 1.01-1.85).
• Ischaemic stroke HR point estimates were similar for different thresholds of AF burden examined. A threshold of >1 hour showed the highest point estimate of HR: 2.11 (95%CI: 1.22-3.64, P=0.008). AF burden of >5 min was already significantly associated with occurrence of ischaemic stroke, as compared to <5 min (HR: 1.76, 95%CI: 1.02-3.02, P=0.041). Thresholds of longer AF burden did not reach statistical significance.
• An analysis that excluded patients on oral anticoagulation at baseline, and adjusted for CHADS2 score, also showed an increased risk of stroke for a device-detected AF burden >1 hour, as compared to shorter than 1 hour (HR: 1.90, 95%CI: 1.00-3. 61, P=0.0487).

Conclusion

This pooled analysis shows that in relatively unselected patients with an arrhythmia detecting cardiac implanted electronic device , daily AF burden is associated with in increased risk of ischaemic stroke or TIA, even after adjustment for use of oral anticoagulation and CHADS2 score. Every additional hour in the daily maximum of AF burden, increased the relative risk of stroke by ~3%. In dichotomised analyses, a threshold of >1 hour yielded the highest risk.
These data support the idea that measuring daily AF burden is clinically relevant and that specific thresholds of daily AF burden may be useful to identify a substantial increase in the risk of ischaemic stroke.

Find this article on Pubmed

References

1. Kirchhof P, Lip GY, Van Gelder IC et al.. Comprehensive risk reduction in patients with atrial fibrillation: emerging diagnostic and therapeutic options—a report from the 3rd Atrial Fibrillation Competence NETwork/European Heart Rhythm Association consensus conference. Europace 2012;14:8–27.
2. Camm AJ, Corbucci G, Padeletti L. Usefulness of continuous electrocardiographic monitoring for atrial fibrillation. Am J Cardiol 2012;110:270–276.