Predictors of Atrial Fibrillation After Thoracic Radiotherapy

Abstract

Background

Atrial fibrillation (AF) has been associated with thoracic radiotherapy, but the specific risk with irradiating different cardiac substructures remains unknown.

Objectives

This study sought to examine the relationship between irradiation of cardiac substructures and the risk of clinically significant (grade ≥3) AF.

Methods

We analyzed data from patients who underwent definitive radiotherapy for localized cancers (non–small cell lung, breast, Hodgkin lymphoma, or esophageal) at our institution between 2004 and 2022. The 2-Gy fraction equivalent dose was calculated for cardiac substructures, including the pulmonary veins (PVs), left atrium, sinoatrial node, and left coronary arteries (the left main, left anterior descending, and left circumflex arteries). Competing risk models (subdistribution HRs [sHRs]) for AF incidence were adjusted for the Mayo AF risk score (MAFRS).

Results

Among 539 patients, the median follow-up was 58.8 months. The 5-year cumulative incidence of AF was 11.1% for non–small cell lung cancer, 8.3% for esophageal cancer, 1.3% for breast cancer, and 0.8% for Hodgkin lymphoma. Increased AF risk was associated with a higher PV maximum dose (d_max) (sHR: 1.22; P < 0.001), larger left atrial volume (sHR: 1.01; P = 0.002), greater smoking history in pack-years (sHR: 1.01; P = 0.010), and higher MAFRS (sHR: 1.16; P < 0.001). PV d_max remained a significant predictor of AF across different MAFRS subgroups (P_interaction = 0.11), and a PV d_max >39.7 Gy was linked to a higher AF risk, even when stratified by MAFRS.

Conclusions

PV d_max is a significant predictor of grade ≥3 AF regardless of underlying risk factors. These findings highlight the importance of cardiac substructures in radiation toxicity and suggest that various PV dose metrics should be further validated in clinical settings.