Brain natriuretic peptide and echocardiographic parameters as biomarkers for predicting new atrial fibrillation detection after acute ischemic stroke

Abstract

Background:

Detection of atrial fibrillation (AF) after ischemic stroke (IS) is crucial for starting anticoagulant therapy to prevent IS recurrence. Strategies to identify patients who might benefit the most from extended electrocardiography (ECG) monitoring would be highly desirable.

Objectives:

We aimed to investigate the role of brain natriuretic peptide (BNP) and echocardiographic parameters as biomarkers for predicting new AF detection in a large cohort of patients with acute IS.

Design:

We retrospectively analyzed data of 2411 consecutive patients admitted for IS to a single Stroke Center from January 1, 2018, to May 31, 2023.

Methods:

BNP levels were measured within 48 h of onset. Clinical and echocardiographic variables were evaluated. Single or multiple 12-lead ECG and continuous monitoring with external or implantable devices were used to detect new AF. The outcome measure was new AF detection at 3 months.

Results:

Of 2337 included patients, 1907 (81.6%) had not previously known AF, and new AF was detected in 422 (22%) patients. In the multivariate analysis, older age (odds ratio (OR) 1.03, 95% confidence interval (CI) 1.02–1.05, p < 0.001), higher National Institutes of Health Stroke Scale score (OR 1.05, 95% CI 1.03–1.07, p < 0.001), BNP ⩾150 pg/ml (OR 4.10, 95% CI 2.95–5.69, p < 0.001), and left atrium (LA) enlargement (OR 2.88, 95% CI 1.99–4.18, p < 0.001) were independently associated with new AF detection. The combination of both BNP ⩾150 pg/ml and LA enlargement showed a strong association (adjusted OR 4.74, 95% CI 3.47–6.48, p < 0.001) and a good discriminative performance for predicting new AF detection at 3 months (area under the receiver operating characteristic curve 0.74, 95% CI 0.71–0.77, sensitivity 0.70 (95% CI 0.65–0.74), specificity 0.78 (95% CI 0.76–0.80)).

Conclusion:

BNP ⩾150 pg/ml within 48 h of IS onset and LA enlargement, especially if combined, may be valuable biomarkers for predicting new AF detection and identifying patients who might benefit the most from extended ECG monitoring.

Keywords

atrial fibrillation biomarkers brain natriuretic peptide echocardiography ischemic stroke

Introduction

Cardioembolic stroke accounts for nearly 30% of ischemic stroke (IS) and atrial fibrillation (AF) is the most common cause of cardioembolism.¹ Early detection of AF after acute IS is crucial for starting an effective secondary prevention strategy based on anticoagulation treatment. Recent guidelines recommend extended electrocardiography (ECG) monitoring for AF detection after IS, but there is uncertainty on timing, duration, and patient selection.² Studies on extended ECG monitoring with implantable loop recorders in patients with cryptogenic IS have shown that AF is detected in about 20%–30% of cases after 2 years.^3–5 However, routine use of these devices in clinical practice presents several logistic and resource issues, and selection strategies to identify patients who might benefit the most from such interventions would be highly desirable. Previous studies have identified predictors of AF detection after IS, including age, vascular risk factors, and stroke occlusion pattern (embolic pattern, with cortical branch occlusion and multiterritorial involvement).^5–8 Among the most promising are blood and echocardiographic biomarkers. Blood biomarkers previously associated with heart failure, such as brain natriuretic peptide (BNP) and N-terminal pro-BNP (NT-proBNP), have also been related to an increased risk of AF detection after IS.^9–14 However, most studies were limited by small sample sizes, heterogeneous populations, and lack of a complete cardiologic assessment.

We aimed to investigate the role of BNP and echocardiographic parameters as biomarkers for predicting new AF detection in a large and homogenously studied cohort of patients with acute IS.

Materials and methods

Study population and data collection

We retrospectively analyzed data of 2411 consecutive patients admitted for IS to a single Stroke Center (Department of Neurology and Stroke Center, Maggiore Hospital, Bologna, Italy) from January 1, 2018, to May 31, 2023.

Inclusion criteria were acute IS admitted within 48 h of symptom onset, availability of BNP plasma levels within 48 h of symptom onset, and complete data on AF history. There were no specific exclusion criteria.

IS was clinically diagnosed by stroke neurologists and confirmed by neuroimaging. Computer tomography (CT) and/or magnetic resonance imaging (MRI) were used. In cases where CT imaging yielded negative results during hospitalization, and MRI was not performed, IS was defined based on clinical criteria as one or more focal neurological deficits of presumed vascular origin lasting more than 24 h, and exclusion of stroke mimics.

Patients received acute reperfusion treatments (intravenous thrombolysis, thrombectomy, or both), as indicated by the most recent guidelines on stroke treatment at that time.

We recorded demographic data (age, sex), vascular risk factors (hypertension, hypercholesterolemia, and diabetes mellitus), and medical history related to AF (previous AF), previous cerebrovascular events, coronary artery disease (CAD), and congestive heart failure (CHF). Stroke severity was assessed using the National Institutes of Health Stroke Scale (NIHSS) score at admission.

To evaluate stroke etiology, we conducted various examinations based on patient profile, test availability, and patient consent. These examinations always included neck and intracranial vessel studies (CT angiography and/or neck and transcranial echo color Doppler), and cardiologic assessment for heart rhythm (12-lead ECG, 24-h Holter-ECG, and prolonged ECG monitoring in selected cases) and cardiac morphology (transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) in selected cases). All patients performed routine blood exams, including complete blood cell count, glucose, hemoglobin glycated (HbA1c), prothrombin time, partial thromboplastin activated time, fibrinogen, lipid profile, renal function, C-reactive protein, and BNP within 48 h of symptom onset. Stroke subtypes were classified according to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria.¹⁵

Measurement of BNP plasma level

Whole venous blood samples were collected upon admission, within 48 h from symptoms onset, using EDTA-containing plastic tubes. Plasma BNP concentrations were promptly measured after collection through a commercially available immunoassay (Triage BNP Test, Biosite Inc., San Diego, California, USA). This assay employs a two-site immunoenzymatic approach, comprising two antibodies: a mouse monoclonal anti-human BNP antibody, immobilized on the solid phase, which binds to plasma BNP, and a mouse anti-human BNP antibody-alkaline phosphatase conjugate that specifically reacts with the bound BNP. A chemiluminescent substrate (Lumi-Phos 530) is added to the reaction and the light generated is measured with a luminometer.

Cardiologic assessment

To detect AF, patients underwent single or multiple 12-lead ECG, continuous ECG monitoring, single or multiple 24-h Holter monitoring with 7-day dynamic monitoring in most patients, or implantable loop recorder as feasible. Most patients received TTE. Patients under 60 years old with an embolic stroke pattern of undetermined origin received TEE for a more detailed cardiac characterization and exclusion of other cardioembolic sources. The following TTE parameters were recorded: left atrium (LA) volume index (LAVI), left ventricle (LV) end-diastolic volume index (LVVI), left ventricular ejection fraction (LVEF) with Simpson method, mitral valve (MV) regurgitation, LV filling pattern, and estimated pulmonary artery systolic pressure (ePASP). LA enlargement was defined as LAVI >28 ml/m². LV dilatation was defined as LVVI >75 ml/m². LV dysfunction was defined as LVEF ⩽50%. Severe MV regurgitation was defined as the presence of at least one among these parameters: central jet mitral regurgitation >40% LA or holosystolic eccentric jet, vena contracta ⩾0.7 cm, regurgitant volume ⩾60 ml, regurgitant fraction ⩾50%, effective regurgitant orifice ⩾0.40 cm². Restrictive LV filling pattern was defined as E/A wave ratio >2 or 1–2 with an E wave deceleration time ⩽140 ms. Pulmonary hypertension was defined as ePASP ⩾50 mmHg.

Outcome measure

The outcome measure was the detection of new AF at 3 months. New AF was defined as a dysrhythmia, lasting at least 30 continuous seconds, with irregular RR interval and no detectable P waves on ECG, in patients without previous AF.

Statistical analysis

Continuous variables were presented as median (interquartile range, IQR), while categorical variables were presented as frequencies (percentages, %). For comparison between groups, the Mann–Whitney U test was used for quantitative variables with skewed distribution, while the Chi-square test (χ²) and Fisher’s exact test were used for qualitative variables. Quantitative data were tested with the Kolmogorov–Smirnov test to assess whether they were normally distributed. We examined the association of the variables, including BNP levels, with our outcome measure (i.e., new AF detection at 3 months). Youden’s index was used to identify the BNP threshold with the best discriminative performance for new AF detection. Variables that showed a statistically significant association with the outcome measure were included in a logistic regression multivariate analysis to assess an independent association. Discrimination performance was assessed by calculation of the area under the receiver operating characteristic curve (AUC-ROC). The results are presented as odds ratios (OR) and 95% confidence intervals (CI). Complete case analysis was used for handling missing data. Data were analyzed using a commercially available software package (IBM SPSS, version 24, IBM, Armonk, New York, USA). We established statistical significance at a two-tailed 0.05 level (p < 0.05).

Results

Of 2411 consecutive patients with IS, we finally included 2337 patients, of which 430 (18.4%) had previously known AF and 1907 (81.6%) had not previously known AF (Figure 1).

Figure 1.

Flowchart of patients’ enrollment.

The characteristics of the population are summarized in Table 1. The median age was 78 years (IQR, 66–85), and 1145 (49%) patients were females. In comparison to patients without known AF, patients with known AF were older, and more often, females had more frequently a history of hypertension, CAD, CHF, and previous cerebrovascular events, had a higher baseline NIHSS score, and were less frequently treated with intravenous thrombolysis and more frequently with endovascular thrombectomy. As regards stroke etiology, patients with known AF had most of the time a cardioembolic stroke, while patients without known AF had more frequently a stroke of undetermined origin. The median BNP level was 160 (IQR, 62–372) pg/ml with a right-skewed distribution. Patients with known AF had a higher median BNP level than patients without known AF. As for echocardiographic parameters, patients with known AF had more often LA enlargement, LV dilatation, LV dysfunction, severe MV regurgitation, and PH in comparison to patients without known AF.

Table 1.

Characteristics of the population.

Characteristics	All patients(n = 2337)	With known AF(n = 430)	Without known AF(n = 1907)	p Value
Age, median (IQR)	78 (66–85)	83 (78–88)	76 (63–83)	<0.001
Female sex, n (%)	1145 (49.0)	251 (58.4)	894 (46.9)	<0.001
Diabetes, n (%)	490 (21.0)	101 (23.5)	389 (20.4)	0.155
Hypercholesterolemia, n (%)	874 (37.4)	158 (36.7)	716 (37.5)	0.756
Hypertension (%)	1638 (70.1)	337 (78.4)	1301 (68.2)	<0.001
CAD, n (%)	356 (15.2)	101 (23.5)	255 (13.4)	<0.001
CHF, n (%)	85 (3.6)	45 (10.5)	40 (2.1)	<0.001
Previous stroke/TIA, n (%)	320 (13.7)	77 (17.9)	243 (12.7)	0.005
NIHSS score, median (IQR)	6 (3–13)	10 (4–18)	5 (3–11)	<0.001
Intravenous thrombolysis, n (%)	1073 (45.9)	124 (28.8)	949 (49.8)	<0.001
Thrombectomy, n (%)	765 (32.7)	181 (42.1)	584 (30.6)	<0.001
TOAST classification				<0.001
LAA, n (%)	317 (13.6)	30 (7.0)	287 (15.0)
Cardioembolism, n (%)	734 (31.4)	355 (82.6)	379 (19.9)
Small vessel disease, n (%)	299 (12.8)	18 (4.2)	281 (14.7)
Other, n (%)	90 (3.9)	2 (0.5)	88 (4.6)
Undetermined, n (%)	897 (38.4)	25 (5.8)	872 (45.7)
BNP
median (IQR)	160 (62–372)	393 (224–649)	124 (53–291)	<0.001
⩾100 pg/ml (%)	1486 (63.6)	403 (93.7)	1083 (56.8)	<0.001
Echocardiographic parameters
LA enlargement, n (%)	1104 (63.4)	241 (85.8)	863 (59.1)	<0.001
LV dilatation, n (%)	106 (6.0)	29 (10.2)	77 (5.2)	0.001
LV dysfunction, n (%)	288 (16.2)	76 (26.2)	212 (14.3)	<0.001
Severe MV regurgitation, n (%)	41 (2.3)	16 (5.5)	25 (1.7)	<0.001
Restrictive LV filling pattern, n (%)	56 (3.2)	14 (4.9)	42 (2.8)	0.073
Pulmonary hypertension, n (%)	67 (3.8)	29 (10.1)	38 (2.6)	<0.001

AF, atrial fibrillation; BNP, brain natriuretic peptide; CAD, coronary artery disease; CHF, congestive heart failure; IQR, interquartile range; LA, left atrium; LAA, large artery atherosclerosis; LV, left ventricle; MV, mitral valve; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack; TOAST, Trial of ORG 10172 in Acute Stroke Treatment.

New AF was detected in 422 (22.1%) patients without previously known AF, in 304 cases before hospital discharge and 118 cases after hospital discharge, during the 3-month follow-up.

Median BNP levels were significantly higher in patients with new AF in comparison with patients without new AF (282 (IQR, 168–548) vs 95 (IQR, 42–213) pg/ml, p < 0.001; Figure 2). The maximum value of Youden’s index was 0.46 for a BNP cutoff ranging from 149 to 151 pg/ml. We, therefore, identified a threshold of BNP ⩾150 pg/ml with the best discriminative performance for new AF detection to include as a categorical variable in the multivariate logistic regression model.

Figure 2.

Levels of BNP in patients with and without new AF detection at 3 months.

The echocardiographic parameters associated with new AF detection in the univariate analysis that entered the multivariate logistic regression model were LA enlargement, LV dysfunction, restrictive LV filling pattern, severe MV regurgitation, and pulmonary hypertension.

The other variables associated with new AF detection in the univariate analysis that entered the multivariate logistic regression model were older age, female sex, history of hypertension, and higher NIHSS score.

In the multivariate analysis, older age (OR 1.03, 95% CI 1.02–1.05, p < 0.001), higher NIHSS score (OR 1.05, 95% CI 1.03–1.07, p < 0.001), BNP ⩾150 pg/ml (OR 4.10, 95% CI 2.95–5.69, p < 0.001), and LA enlargement (OR 2.88, 95% CI 1.99–4.18, p < 0.001) were the variables independently associated with a greater likelihood of new AF detection at 3 months (Table 2). BNP ⩾150 pg/ml had good discriminative performance for predicting new AF detection at 3 months, with high sensitivity and moderate specificity (AUC-ROC 0.73, 95% CI 0.70–0.76, sensitivity 0.80 (95% CI 0.76–0.84) specificity 0.66 (95% CI 0.63–0.68); Figure 3).

Table 2.

Univariate and multivariate analyses for new AF detection at 3 months.

Variables	Univariate analysis		Multivariate analysis
Variables	OR (95% CI)	p Value	OR (95% CI)	p Value
Age (per year)	1.06 (1.05–1.07)	<0.001	1.03 (1.02–1.05)	<0.001
Female sex	1.97 (1.58–2.46)	<0.001	1.20 (0.90–1.62)	0.217
Diabetes	0.93 (0.71–1.22)	0.576	/	/
Hypertension	1.61 (1.26–2.06)	<0.001	0.96 (0.69–1.34)	0.800
Hypercholesterolemia	0.85 (0.68–1.07)	0.157	/	/
CAD	1.10 (0.80–1.50)	0.563	/	/
CHF	1.52 (0.77–3.02)	0.230	/	/
Previous stroke/TIA	0.85 (0.61–1.19)	0.340	/	/
NIHSS score (per 1 point)	1.08 (1.06–1.09)	<0.001	1.05 (1.03–1.07)	<0.001
BNP ⩾150 pg/ml	7.72 (5.94–10.03)	<0.001	4.10 (2.95–5.69)	<0.001
LA enlargement	5.71 (4.11–7.94)	<0.001	2.88 (1.99–4.18)	<0.001
LV dilatation	1.14 (0.67–1.94)	0.633	/	/
LV dysfunction	1.49 (1.07–2.06)	0.017	0.69 (0.46–1.02)	0.061
Restrictive LV filling pattern	2.43 (1.30–4.56)	0.006	0.94 (0.46–1.91)	0.858
Severe MV regurgitation	2.77 (1.25–6.16)	0.012	0.95 (0.35–2.60)	0.918
Pulmonary hypertension	3.98 (2.08–7.62)	<0.001	2.04 (0.91–4.57)	0.084

Values in bold are statistically significant.

AF, atrial fibrillation; BNP, brain natriuretic peptide; CAD, coronary artery disease; CHF, congestive heart failure; CI, confidence interval; LA, left atrium; LV, left ventricle; MV, mitral valve; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; TIA, transient ischemic attack.

Figure 3.

ROC curves of BNP ⩾150 pg/ml (AUC 0.73, 95% CI 0.70–0.76, sensitivity 0.80 (95% CI 0.76–0.84), specificity 0.66 (95% CI 0.63–0.68)), and of the combination of BNP ⩾150 pg/ml and LA enlargement (AUC 0.74, 95% CI 0.71–0.77, sensitivity 0.70 (95% CI 0.65–0.74), specificity 0.78 (95% CI 0.76–0.80)) for predicting AF detection at 3 months.

A second logistic regression model was performed, entering the combination of BNP ⩾150 pg/ml and LA enlargement as a single variable and the other variables associated with new AF detection in the univariate analysis. In this analysis, the combination of BNP ⩾150 pg/ml and LA enlargement showed an adjusted OR of 4.74 (95% CI 3.47–6.48, p < 0.001; Table 3). The combination of BNP ⩾150 pg/ml and LA enlargement showed good discriminative performance for predicting new AF detection at 3 months, with less sensitivity but significantly increasing specificity compared to BNP ⩾150 pg/ml alone (AUC-ROC 0.74, 95% CI 0.71–0.77, sensitivity 0.70 (95% CI 0.65–0.74), specificity 0.78 (95% CI 0.76–0.80); Figure 3).

Table 3.

Univariate and multivariate analyses for new AF detection at 3 months, considering the combination of BNP ⩾150 pg/ml and LA enlargement as a single variable.

Variables	Univariate analysis		Multivariate analysis
Variables	OR (95% CI)	p Value	OR (95% CI)	p Value
Age (per year)	1.06 (1.05–1.07)	<0.001	1.04 (1.03–1.05)	<0.001
NIHSS score (per 1 point)	1.08 (1.06–1.09)	<0.001	1.06 (1.03–1.08)	<0.001
BNP ⩾150 pg/ml and LA enlargement	7.97 (6.07–10.48)	<0.001	4.74 (3.47–6.48)	<0.001
Female sex	1.97 (1.58–2.46)	<0.001	1.20 (0.90–1.61)	0.217
Hypertension	1.61 (1.26–2.06)	<0.001	1.00 (0.72–1.39)	0.992
LV dysfunction	1.49 (1.07–2.06)	0.017	0.72 (0.48–1.08)	0.110
Restrictive LV filling pattern	2.43 (1.30–4.56)	0.006	0.95 (0.46–1.96)	0.887
Severe MV regurgitation	2.77 (1.25–6.16)	0.012	0.94 (0.34–2.60)	0.897
Pulmonary hypertension	3.98 (2.08–7.62)	<0.001	2.05 (0.91–4.60)	0.084

Values in bold are statistically significant.

AF, atrial fibrillation; BNP, brain natriuretic peptide; CI, confidence interval; LA, left atrium; LV, left ventricle; MV, mitral valve; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio.

Discussion

Our study showed that older age, higher NIHSS score, BNP ⩾150 pg/ml, and LA enlargement were independently associated with a greater likelihood of new AF detection at 3 months in patients with acute IS.

These results are in line with previous evidence showing that older age and stroke severity are predictors of AF and cardioembolism.^6,16,17 Moreover, our study confirms previous evidence on the potential role of laboratory (BNP) and echocardiographic (LA enlargement) parameters as biomarkers strongly associated with new AF detection.

A BNP level ⩾150 pg/ml within the first 48 h of IS emerges as the strongest predictor, increasing the risk of new AF detection by up to fourfold (OR 4.10, 95% CI 2.95–5.69, p < 0.001). BNP is a peptide produced by cardiomyocytes in response to various stimuli and is an established biomarker of cardiac diseases.¹⁸ Previous studies showed elevated BNP levels in both paroxysmal and permanent AF.^19,20 Moreover, BNP elevation predicted the occurrence of AF in a cohort of apparently healthy individuals.²¹ Other studies found an association between natriuretic peptides and the risk of AF in patients with IS.^{9–14,22–26} BNP was the only laboratory biomarker externally validated in two studies,^9,10 and one small study showed that a baseline high BNP level reduced the number needed to screen with extended ECG monitoring.¹³ However, previous studies on BNP and AF detection after IS had important limitations due to small sample sizes, heterogeneous populations, and lack of a complete cardiologic assessment.

As regards echocardiographic parameters, multiple studies have detected an association between LA enlargement and the detection of AF.^27–32 LA enlargement is the main echocardiographic biomarker of atrial cardiomyopathy, which is considered the common pathological feature of AF and an independent risk factor of IS.³³ LA remodeling, including changes in atrial tissue, size, cellular ultrastructure, and fibrosis, is believed to be the main cause of AF, and conversely, AF itself can lead to atrial remodeling.^33,34 However, there is no single cardiac biomarker that is validated for AF screening after IS, and a recent international clinical consensus statement on atrial cardiomyopathy suggested, as a possible future direction, the use in combination of imaging and circulating biomarkers.³⁵

In this respect, the strength of our results is precisely the availability of both BNP and echocardiographic parameters in the largest population with acute IS ever studied to our knowledge. The combination of both BNP ⩾150 pg/ml and LA enlargement was strongly associated with new AF detection (adjusted OR 4.74, 95% CI 3.47–6.48, p < 0.001) and showed a good discriminative performance for predicting new AF detection within 3 months of IS (sensitivity 0.70 (95% CI 0.65–0.74), specificity 0.78 (95% CI 0.76–0.80)). Similar diagnostic accuracy was recently reported for NT-proBNP by a systematic review and meta-analysis on 2171 patients from 7 studies.³⁶ It is noteworthy that we included a larger sample size in a single-center study.

There is growing evidence supporting the use of prolonged ECG monitoring as an effective stroke prevention strategy in carefully selected patients. In particular, prolonged ECG monitoring is associated with a higher likelihood of AF detection and anticoagulant initiation to prevent stroke recurrence.^37,38 Taken together, the results of our study may help to identify a subgroup of patients with IS who are at higher risk of developing AF. In particular, patients with both high BNP levels and LA enlargement should be carefully taken into consideration for more prolonged and costly ECG monitoring with implantable loop recorders. The practical use of BNP as a biomarker is highly feasible since the test is noninvasive, cost-effective, provides rapid results, and is widely available in hospital laboratories.

Our study has several limitations. First, it is characterized by the inherent weakness of any retrospective study, such as collection and entry bias and possible residual confounding. In particular, the duration and methods of cardiac monitoring were not the same for all patients, and variations in monitoring strategy may have influenced the likelihood of AF detection across patients. Nevertheless, this is a single-center study performed in a Comprehensive Stroke Center with standardized diagnostic protocols, and the vast majority of patients had at least a 7-day ECG monitoring during hospitalization or after discharge. Second, we did not collect information about several factors (i.e., pharmacological therapies and renal function) that may have an impact on BNP levels. Third, BNP rises early after stroke independently of AF, presumably due to BNP release from the brain, and samples taken in the first hours after stroke contain BNP not only from the heart but also from the damaged brain tissue.²² Fourth, we did not collect data on thyroid disease, and therefore, we could not calculate the C₂HEST score,³⁹ which was recently shown to be superior to other scores for AF detection in patients with cryptogenic stroke.⁴⁰ Fifth, there is a minority of patients with missing data on echocardiographic assessment. Sixth, the study population was relatively old, which may limit application to younger patients with cryptogenic or embolic strokes of undetermined source. Seventh, the study is not externally validated, and further evidence from other cohorts is warranted to assess generalizability.

Nevertheless, our study also has several strengths, mainly the considerably large sample size, the largest ever studied to our knowledge, with complete data on BNP levels. Other strengths are the inclusion of consecutive and homogeneously studied IS patients and the high percentage (about 75%) of patients with complete echocardiographic assessment.

Conclusion

Our study suggests that elevated BNP levels (i.e., ⩾150 pg/ml) within 48 h of acute IS onset and LA enlargement may serve as valuable biomarkers for identifying patients with a higher likelihood of new AF detection and who might benefit the most from extended ECG monitoring. In particular, the combination of both elevated BNP levels and LA enlargement showed a good discriminative performance for predicting the detection of new AF within 3 months of IS.

Footnotes

Acknowledgements

None.

Declarations

ORCID iDs

Stefano Forlivesi

Andrea Zini

References

Sacchetti

Furie

Yaghi

Cardioembolic stroke: mechanisms and therapeutics. Semin Neurol 2017; 37: 326–338.

Rubiera

Aires

Antonenko

, et al. European Stroke Organisation (ESO) guideline on screening for subclinical atrial fibrillation after stroke or transient ischaemic attack of undetermined origin. Eur Stroke J 2022; 7: VI. (Erratum in: Eur Stroke J 2023; 8: 413.)

Sanna

Diener

H-C

Passman

, et al. Cryptogenic stroke and underlying atrial fibrillation. New Engl J Med 2014; 370: 2478–2486.

Ziegler

Rogers

Ferreira

, et al. Long-term detection of atrial fibrillation with insertable cardiac monitors in a real-world cryptogenic stroke population. Int J Cardiol 2017; 244: 175–179.

Ungar

Pescini

Rafanelli

, et al. Detection of subclinical atrial fibrillation after cryptogenic stroke using implantable cardiac monitors. Eur J Intern Med 2021; 92: 86–93.

Riordan

Opaskar

Yoruk

, et al. Predictors of atrial fibrillation during long-term implantable cardiac monitoring following cryptogenic stroke. J Am Heart Assoc 2020; 9: e016040.

Kwong

Ling

Crawford

, et al. A clinical score for predicting atrial fibrillation in patients with cryptogenic stroke or transient ischemic attack. Cardiology 2017; 138: 133–140.

Yushan

Tan

BYQ

Ngiam

, et al. Association between bilateral infarcts pattern and detection of occult atrial fibrillation in embolic stroke of undetermined source (ESUS) patients with insertable cardiac monitor (ICM). J Stroke Cerebrovasc Dis 2019; 28: 2448–2452.

Miyazaki

Toyoda

Iguchi

, et al. Atrial fibrillation after ischemic stroke detected by chest strap-style 7-day Holter monitoring and the risk predictors: EDUCATE-ESUS. J Atheroscler Thromb 2021; 28: 544–554.

10.

Yoshioka

Watanabe

Zeniya

, et al. A score for predicting paroxysmal atrial fibrillation in acute stroke patients: iPAB score. J Stroke Cerebrovasc Dis 2015; 24: 2263–2269.

11.

Pagola

Juega

Francisco-Pascual

, et al.; Crypto-AF Study Group. Predicting atrial fibrillation with high risk of embolization with atrial strain and NT-proBNP. Transl Stroke Res 2021; 12: 735–741.

12.

Fonseca

Brito

Pinho

Melo

, et al. N-terminal pro-brain natriuretic peptide shows diagnostic accuracy for detecting atrial fibrillation in cryptogenic stroke patients. Int J Stroke 2014; 9: 419–425.

13.

Wasser

Weber-Krüger

Gröschel

, et al. Brain natriuretic peptide and discovery of atrial fibrillation after stroke: a subanalysis of the Find-AF_RANDOMISED trial. Stroke 2020; 5: 395–401.

14.

Ward

Vail

Cameron

, et al. Molecular biomarkers predicting newly detected atrial fibrillation after ischaemic stroke or TIA: a systematic review. Eur Stroke J 2023; 8: 125–131.

15.

Adams

Jr Bendixen

Kappelle

, et al. Classification of subtype of acute ischemic stroke: definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993; 24: 35–41.

16.

Gao

Xie

, et al. Aging and atrial fibrillation: a vicious circle. Int J Cardiol 2024; 395: 131445.

17.

Ferro

JM.

Cardioembolic stroke: an update. Lancet Neurol 2003; 2(3): 177–188.

18.

de Lemos

McGuire

Drazner

MH.

B-type natriuretic peptide in cardiovascular disease. Lancet 2003; 362: 316–322.

19.

Seegers

Zabel

Grüter

, et al. Natriuretic peptides for the detection of paroxysmal atrial fibrillation. Open Heart 2015; 2: e000182.

20.

Silvet

Young-Xu

Walleigh

, et al. Brain natriuretic peptide is elevated in outpatients with atrial fibrillation. Am J Cardiol 2003; 92: 1124–1127.

21.

Takase

Dohi

Sonoda

, et al. Prediction of atrial fibrillation by B-type natriuretic peptide. J Atr Fibrillation 2013; 5: 674.

22.

Wachter

Lahno

Haase

, et al. Natriuretic peptides for the detection of paroxysmal atrial fibrillation in patients with cerebral ischemia—the Find-AF study. PLoS One 2012; 7: e34351.

23.

Fujii

Shibazaki

Kimura

, et al. A simple score for predicting paroxysmal atrial fibrillation in acute ischemic stroke. J Neurol Sci 2013; 328: 83–86.

24.

Shibazaki

Kimura

Fujii

, et al. Brain natriuretic peptide levels as a predictor for new atrial fibrillation during hospitalization in patients with acute ischemic stroke. Am J Cardiol 2012; 109: 1303–1307.

25.

Okada

Shibazaki

Kimura

, et al. Brain natriuretic peptide as a predictor of delayed atrial fibrillation after ischaemic stroke and transient ischaemic attack. Eur J Neurol 2010; 17: 326–331.

26.

Haeusler

Gröschel

Köhrmann

, et al. Expert opinion paper on atrial fibrillation detection after ischemic stroke. Clin Res Cardiol 2018; 107: 871–880.

27.

Jordan

Yaghi

Poppas

, et al. Left atrial volume index is associated with cardioembolic stroke and atrial fibrillation detection after embolic stroke of undetermined source. Stroke 2019; 50: 1997–2001.

28.

Lee

Moon

Cho

, et al. Left atrial diameter and atrial ectopic burden in patients with embolic stroke of undetermined source: risk stratification of atrial fibrillation with insertable cardiac monitor analysis. J Clin Neurol 2021; 17: 213–219.

29.

Baturova

Sheldon

Carlson

, et al. Electrocardiographic and echocardiographic predictors of paroxysmal atrial fibrillation detected after ischemic stroke. BMC Cardiovasc Disord 2016; 16: 209.

30.

Waldenhjort

Sobocinski Doliwa

Alam

, et al. Echocardiographic measures of atrial function may predict atrial fibrillation in stroke patients. Scand Cardiovasc J 2016; 50: 236–242.

31.

Stahrenberg

Edelmann

Haase

, et al. Transthoracic echocardiography to rule out paroxysmal atrial fibrillation as a cause of stroke or transient ischemic attack. Stroke 2011; 42: 3643–3645.

32.

Perlepe

Sirimarco

Strambo

, et al. Left atrial diameter thresholds and new incident atrial fibrillation in embolic stroke of undetermined source. Eur J Intern Med 2020; 75: 30–34.

33.

Ning

Tse

Luo

, et al. Atrial cardiomyopathy: an emerging cause of the embolic stroke of undetermined source. Front Cardiovasc Med 2021; 8: 674612.

34.

Nattel

Harada

Atrial remodeling and atrial fibrillation: recent advances and translational perspectives. J Am Coll Cardiol 2014; 63: 2335–2345.

35.

Goette

Corradi

Dobrev

, et al. Atrial cardiomyopathy revisited-evolution of a concept: a clinical consensus statement of the European Heart Rhythm Association (EHRA) of the ESC, the Heart Rhythm Society (HRS), the Asian Pacific Heart Rhythm Society (APHRS), and the Latin American Heart Rhythm Society (LAHRS). Europace 2024; 26(9): euae204.

36.

Patel

Bhaskar

SMM

. Diagnostic utility of N-terminal Pro-B-type natriuretic peptide in identifying atrial fibrillation post-cryptogenic stroke: a systematic review and meta-analysis. Pathophysiology 2024; 31(3): 331–349.

37.

Tsivgoulis

Palaiodimou

Triantafyllou

, et al. Prolonged cardiac monitoring for stroke prevention: a systematic review and meta-analysis of randomized-controlled clinical trials. Eur Stroke J 2023; 8: 106–116.

38.

Tsivgoulis

Triantafyllou

Palaiodimou

, et al. Prolonged cardiac monitoring and stroke recurrence: a meta-analysis. Neurology 2022; 98: e1942–e1952.

39.

Bisson

Bodin

, et al. C₂HEST score and prediction of incident atrial fibrillation in poststroke patients: a French nationwide study. J Am Heart Assoc 2019; 8: e012546.

40.

Palaiodimou

Theodorou

Triantafyllou

, et al. Performance of different risk scores for the detection of atrial fibrillation among patients with cryptogenic stroke. Stroke 2024; 55: 454–462.