Abstract
Background:
Ablation Index (AI) software has allowed better atrial fibrillation (AF) ablation results, but recurrence rates remain significant. Specific serum biomarkers have been associated with this recurrence.
Objectives:
To evaluate whether certain biomarkers could be used (either individually or combined) to predict arrhythmia recurrence after AI-guided AF ablation.
Methods:
Prospective multicenter observational study of consecutive patients referred for AF ablation from January 2018 to March 2021. Hemoglobin, brain natriuretic peptide (BNP), C-reactive protein, high sensitivity cardiac troponin I, creatinine clearance, thyroid-stimulating hormone (TSH) and free thyroxine (FT4) were assessed for their ability to predict arrhythmia recurrence during follow-up. Statistical significance was accepted for p values of<0.05.
Results:
A total of 593 patients were included - 412 patients with paroxysmal AF and 181 with persistent AF. After a mean follow-up of 24±6 months, overall single-procedure freedom from atrial arrhythmia was 76.4%. Individually, all biomarkers had no or only modest predictive power for recurrence. However, a TSH value >1.8 μUI/mL (HR=1.82 [95% CI, 1.89-2.80], p=0.006) was an independent predictor of arrhythmia recurrence. When assessing TSH, FT4 and BNP values in combination, each additional “abnormal” biomarker value was associated with a lower freedom from arrhythmia recurrence (87.1 % for no biomarker vs. 83.5% for one vs. 75.1% for two vs. 43.3% for three biomarkers, p<0.001). Patients with three “abnormal” biomarkers had a threefold higher risk of AF recurrence compared with no “abnormal” biomarker (HR=2.88 [95% CI, 1.39-5.17], p=0.003).
Conclusions:
When used in combination, abnormal TSH, FT4 and BNP values can be a useful tool for predicting arrhythmia recurrence after AI-guided AF ablation.
Keywords:
Biomarkers; Catheter Ablation; Atrial Fibrillation; Cardiac Arrhythmias
Thumbnail
Thumbnail
Thumbnail
Kaplan-Meier curves for freedom from atrial arrhythmia recurrence, according to the number of “abnormal” biomarkers present (87.1 % for no biomarker vs. 83.5% for one vs. 75.1% for two biomarkers vs. 43.3% for three biomarkers, p<0.001).
