Predictors of Atrial Fibrillation in Patients With Transient Ischemic Attack

Patients with transient ischemic attack (TIA) with high brain natriuretic peptide (BNP) values are more likely to have atrial fibrillation (AF), while other serum biomarkers were of limited predictive value. Carotid intima-media thickness (CIMT) and coronary artery calcium (CAC) score were also poor predictors of AF in these patients. Findings were published in the Journal of Stroke and Cerebrovascular Diseases.

While AF is an important risk factor for TIA, there are limited data on markers of cardiac arrhythmia in patients with TIA. The goal of this study was to assess predictors of AF in patients with TIA, including markers of subclinical atherosclerosis, such as CAC score and CIMT; serum biomarkers, including cardiac troponoin I, high-sensitive C-reactive protein, BNP, copeptin, and midregional proadrenomedulin were also evaluated.

Patients diagnosed with TIA and no known history of AF were included in this study. They received sequential screening for AF which included first ECG, 72-hour Holter monitoring and finally at 1 year of continuous monitoring with an implanted cardiac monitor. At baseline, cardiac computed tomography, carotid artery ultrasound, and blood sampling were done.

The final cohort included 114 patients (median age 66 years, 49% male), of which 18 patients had AF detected on ECG (4 patients), Holter monitoring (5 patients), or an implanted cardiac monitor (9 patients).

Patients with AF were significantly older than patients without AF (median age 71.1 vs 64.4 years, P =.008) and were more likely to have a history of prior TIA (22.2% vs 5.2%, P =.04).

There was no reported difference in median CAC score between patients with and without AF. Mean CIMT was significantly higher in patients with AF compared with patients without AF (0.84 vs 0.74 mm, P =.002) and patients with AF also had a higher proportion of carotid plaques (66.7% vs 37.9%, P =.04). However, CIMT was not a predictive marker for AF when analyzed in a multivariate logistic regression analysis.

When divided into tertiles, the biomarkers did not have any predictive value, except that patients with BNP in the upper tertile (>68.95 pg/mL) were more likely to have AF (odds ratio 5.96; 95% CI 1.04-34.07, P =.045) than were patients with BNP in the lowest tertile (<21.1 pg/mL).

The study had several limitations, according to the researchers, including the small sample size and possible operator error as all carotid ultrasound were performed by the same operator. Further, the fact that not all blood tests were performed at admission but almost 3 months after the TIA event may have further limited these findings.

“CIMT and CACS as markers of subclinical atherosclerosis were poor predictors of incident AF in patients with TIA,” concluded the researchers. They also noted that “Apart from BNP, the selected serum biomarkers were also of limited value in predicting AF.” The investigators believe that in order to identify more accurate predictors of AF in patients with TIA, further studies with larger sample sizes should be pursued.

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.

Reference

Pedersen KB, Madsen C, Sandgaard NCF, et al. Predictive markers of atrial fibrillation in patients with transient ischemic attack. J Stroke Cerebrovasc Dis. 2020;29(4):104643. doi: 10.1016/j.jstrokecerebrovasdis.2020.104643