Author + information
- Tor Biering-Sørensen, MD∗ (, )
- Louisa M. Christensen, MD,
- Derk W. Krieger, MD, PhD,
- Rasmus Mogelvang, MD, PhD,
- Jan Skov Jensen, MD, PhD, DMSci,
- Søren Højberg, MD, PhD,
- Nis Høst, MD, PhD,
- Finn Michael Karlsen, MD, PhD and
- Hanne Christensen, MD, PhD, DMSci
- ↵∗Department of Cardiology, Gentofte Hospital, Copenhagen University, Niels Andersensvej 65, DK-2900 Post 835, Copenhagen, Denmark
In 25% of patients with ischemic stroke, no etiologic factor is identified: so-called cryptogenic strokes (CS) (1). Asymptomatic paroxysmal atrial fibrillation (PAF) is often suspected to be the cause of stroke in these patients.
Echocardiographic estimates of left atrial (LA) function, such as the left atrial emptying fraction (LAEF), can easily be determined by echocardiography and could potentially be a surrogate marker for already present PAF (2).
The aim of this study was to establish if echocardiographic investigation of the LA function would identify patients with subsequent documentation of PAF following a CS.
A total of 58 patients with CS were included in the SURPRISE (Stroke Prior to Diagnosis of Atrial Fibrillation Using Longterm Observation with Implantable Cardiac Monitoring Apparatus Reveal) study (clinical trial number NCT01498146) and underwent conventional 2-dimensional echocardiography a median of 42 days (interquartile range [IQR]: 14 to 134 days) after the stroke as part of the clinical work-up to identify if the heart was the source of the ischemic stroke. All patients had an insertable cardiac monitor implanted within a time interval after their stroke as per the current practice recommendations (median 69 days, IQR: 42 to 127 days). Left atrial volumes (LAV) were estimated by the biplane area-length method in the apical 4- and 2-chamber views. The maximal LAV was measured during left ventricular (LV) end-systole and minimal LAV during LV end-diastole. The LAEF was calculated as: (maximal LAV − minimal LAV)/(maximal LAV). During the course of follow-up (median 20 months, IQR: 11 to 29 months) 13 (22%) patients were diagnosed with PAF. Follow-up was 100%. Atrial fibrillation was diagnosed a median of 59 days (IQR: 8 to 109 days) after implantation of the insertable cardiac monitor. No patient experienced any symptoms in conjunction with the diagnosis of atrial fibrillation.
Patients subsequently diagnosed with PAF were significantly older (65 ± 8 years vs. 51 ± 12 years, p < 0.001) and had significantly higher mean CHADS2 (Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Stroke) score (2.9 ± 1.1 vs. 2.4 ± 0.6, p = 0.022) than patients without PAF. Other than age and CHADS2 score, no baseline characteristic (sex, hypertension, diabetes, cholesterol, blood sugar, or body mass index) was associated with PAF.
No patient was diagnosed with an evident cardioembolic source as a consequence of the echocardiographic examination. None of the conventional echocardiographic parameters (LV ejection fraction, LV volumes, LV mass index, E/A ratio, deceleration time, e′, a′, LA volume index, LA volume index/a′, tricuspid annular plane systolic excursion, and grade of diastolic dysfunction) were significant predictors of PAF. However, the LAEF was significantly reduced (37 ± 17% vs. 47 ± 11%, p = 0.016) in patients subsequently diagnosed with PAF. The risk of subsequently being diagnosed with PAF increased with decreasing tertile of the LAEF (Figure 1), and was approximately 10× higher for patients in the first tertile (with an LAEF ≤41%) compared with patients in the third tertile (with an LAEF >50%) (hazard ratio: 9.6; 95% confidence interval: 1.2 to 77.3; p = 0.033).
Age <60 years and an LAEF >50% individually predicted the absence of PAF with 94% and 95% accuracy, respectively. Furthermore, an age ≥60 years and an LAEF ≤50% correctly identified 85% and 92% of the PAF patients, respectively.
Using joint criteria by combining both cutoff values, the diagnostic utility was improved. If the patient was <60 years of age and had an LAEF >50%, the patient had no risk of PAF, corresponding to a negative predictive value of 100%. Furthermore, when using ≥60 years of age and/or an LAEF ≤50% as warning signals for the presence of PAF, no patient subsequently diagnosed with PAF would have been misclassified, corresponding to a sensitivity of 100%. In addition, if a patient was ≥60 years of age and had an LAEF <50%, the risk of suffering from PAF was 59%, corresponding to a positive predictive value of 59%.
This is the first study, using long-term rhythm monitoring, to demonstrate that a simple echocardiographic assessment of the left atrium function contributes significantly to risk stratification for PAF in patients with CS.
We found that a decrease in the LAEF was associated with an increased risk of PAF (Figure 1). Furthermore, using joint criteria by combining both cutoff values of age and LAEF, the diagnostic accuracy was increased to 100% for excluding PAF as the causal reason for the cryptogenic ischemic stroke. However, these results are from a relatively small sample size, and unless they are repeated in larger cohorts, the results cannot be extrapolated to the entire cohort of cryptogenic stroke patients.
Please note: This study was financially supported by the Faculty of Health Sciences, University of Copenhagen, Denmark; the capital region of Denmark; Medtronic Denmark; Sophus Jacobsens; and the Arvid Nilssons Foundation. The sponsors had no role in the study design, data collection, data analysis, data interpretation, or writing of the manuscript. Dr. Christensen has received a research grant from Medtronic Denmark. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation