Dosimetric Predictors of New-Onset Atrial Fibrillation Following Thoracic Radiotherapy for Lung Cancer

Background

While the association between thoracic radiotherapy and atrial fibrillation (AF) is established, specific risks linked to the radiation dose delivered to critical cardiac substructures, such as the cardiac conduction system, remain unclear and require dedicated investigation.

Purpose

The aim of this study was to determine the incidence of new-onset atrial fibrillation in lung cancer patients receiving thoracic radiotherapy and to investigate the relationship between dosimetric values related to the irradiation of specific cardiac substructures and AF incidence.

Methods

Data from primary lung cancer patients who received thoracic radiotherapy at our center between 2021 and 2024 were retrospectively analyzed. Patients were followed with 12-lead surface electrocardiograms (ECGs) at baseline (pre-radiotherapy) and at 3- and 6-month control visits post-radiotherapy. Symptomatic patients also underwent 24-hour rhythm Holter monitoring.

All doses were converted to equivalent doses in 2-Gy fractions with an ratio of 3. We calculated the mean and maximum (dmax) doses for each cardiac substructure of interest, including the sinoatrial node (SAN) and atrioventricular node (AVN) (Figure 1b).

Spline analysis was performed to identify potential dosimetric cutoff points for the SAN dmax, AVN dmax and mean heart dose. Fine-Gray regression analysis was conducted both univariable and multivariable for the endpoint of new-onset AF.

Results

The average follow-up duration for the 84 included patients was 18.1±13.0 months. Of these patients, 21 had small-cell lung cancer (SCLC) and 63 had non-small-cell lung cancer (NSCLC). The overall clinic AF incidence was 11%. The mean age of the patients was 68.1±8.7 years, and 83% were male (Table 1a).

Univariable and multivariable Fine-Gray regression analysis demonstrated that increased AF incidence was strongly associated with a higher SAN dmax (sHR: 1.03, ) and advanced age (sHR: 1.09, ) (Table 1b). Spline analysis for the new-onset AF endpoint determined a cutoff point of 17.23 Gy for SAN dmax and 14.47 Gy for AVN dmax.

When comparing the two groups defined by the SAN dmax cutoff (≤ 17.23 Gy vs > 17.23 Gy), the cumulative AF incidence showed a statistically significant difference (χ2 =4.295, p=0.038) (Figure 1a). Despite the group with SA Node Gy being younger (p=0.01), the two groups were similar in terms of comorbidities, histology, medical treatment, LVEF, and left atrial (LA) diameter.

Conclusion(s)

Although an increase in AF incidence is typically associated with age, the group exceeding the determined SA Node cutoff value was found to be at significantly higher risk for AF incidence despite being younger. These findings suggest that examining the radiation doses received by components of the cardiac conduction system, specifically the SA node and AV node, could provide opportunities for early detection and screening of AF in this patient population.FigureTable