Beta-Blocker Use Shows Interaction With AF Recurrence Radiomics

A retrospective study of 555 eligible patients with atrial fibrillation found that long-term preoperative beta-blocker use was associated with markedly different performance from a left atrial appendage radiomics model designed to predict recurrence after catheter ablation. Among the 555 eligible participants, 91 experienced recurrence, corresponding to 16.4%. The model’s area under the receiver operating characteristic curve was 0.742 in patients who had not received beta-blockers before ablation but 0.501 in those who had, a statistically significant difference with P < 0.001.

Statistical tests also identified an interaction between beta-blocker use and the radiomics score, indicating that the score’s relationship with recurrence depended on preoperative medication status. Researchers screened 707 patients with atrial fibrillation and selected 555 who met the study’s eligibility criteria. They stratified the eligible cohort by long-term preoperative beta-blocker therapy and analyzed baseline clinical and echocardiographic characteristics.

Because the groups differed in measured baseline characteristics, the investigators applied inverse probability of treatment weighting before evaluating the radiomics score and its interaction with beta-blocker use. This adjustment was intended to address observed clinical imbalances between the groups, allowing the analysis to compare the score’s predictive relationship within the two treatment strata. The principal prediction model was developed in the 295 patients who had not received beta-blockers before the procedure, 50 of whom experienced atrial fibrillation recurrence.

Within this group, patients with recurrence had lower lymphocyte counts, higher body mass index and a lower prevalence of diabetes than patients without recurrence. These three clinical features were integrated with a radiomics score, or Rad-score, calculated from 11 selected left atrial appendage features. The resulting nomogram combined imaging and clinical information to estimate the probability of post-ablation recurrence.

After inverse-probability-weighting adjustment, the Rad-score was strongly associated with recurrence among patients without preoperative beta-blocker therapy, with a hazard ratio of 2.73 and P < 0.001. The corresponding association was not statistically significant among beta-blocker users: the hazard ratio was 0.91 and the P value was 0.338. The formal interaction test between Rad-score and beta-blocker use was significant at P < 0.001.

Kaplan–Meier analysis showed a related pattern: the radiomics score significantly stratified 12-month recurrence-free survival in the weighted cohort without beta-blocker therapy, with a log-rank P value below 0.001, but not in the beta-blocker group. These findings do not show that beta-blocker therapy prevents atrial fibrillation recurrence, nor do they establish that the medication removes biologically meaningful information from left atrial appendage imaging. Instead, the reported interaction shows that the radiomics score had different predictive performance and different statistical associations with observed outcomes in patients with and without long-term preoperative beta-blocker use.

The model developed without beta-blocker users therefore did not transfer effectively to the treated group, whose discrimination was close to chance at an AUC of 0.501. The results support accounting for preoperative medication history when developing or evaluating imaging-based post-ablation risk models. The study has important limitations.

The model requires multicenter external validation before its applicability can be established in independent populations, according to the researchers. Within these constraints, the clearest finding is that long-term preoperative beta-blocker use was linked to a marked difference in the ability of left atrial appendage radiomics to predict atrial fibrillation recurrence after ablation.

Subject of Research: Interaction between preoperative beta-blocker therapy and left atrial appendage radiomics for predicting post-ablation atrial fibrillation recurrence

Article Title: Interaction between long-term preoperative beta-blocker therapy and left atrial appendage radiomics in predicting post-ablation atrial fibrillation recurrence

Article References: Yang, Y., Zhong, W., Zhang, L., Zhang, X., Fang, Y., Wang, H., Jiang, X., Yu, Y., & Huang, D. (2026). Interaction between long-term preoperative beta-blocker therapy and left atrial appendage radiomics in predicting post-ablation atrial fibrillation recurrence. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02838-7

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02838-7

Keywords: atrial fibrillation, catheter ablation, beta-blockers, left atrial appendage, radiomics, recurrence prediction, clinical risk stratification, medical imaging, retrospective study, Interaction, between, long-term