Multispline catheter mapping of the coronary ostia to guide aortic root ablation

INTRODUCTION AND OBJECTIVES: Safe ablation of ventricular arrhythmias from the left ventricular outflow tract (LVOT) requires accurate coronary ostia localization, but conventional imaging adds procedural complexity, radiation, and cost. The aim of this study was to determine whether electroanatomic coronary ostial localization using multispline-based catheters reproduces computed tomography (CT)-defined ostial anatomy and could serve as an anatomic reference. METHODS: Among 85 consecutive patients undergoing evaluation for LVOT ablation, 51 met the inclusion criteria and were prospectively enrolled in a 2-phase observational study: a CT-validated cohort (n = 26) followed by a non-CT implementation cohort (n = 25). Coronary ostia were identified electroanatomically using predefined protrusion criteria and compared with CT in a validation cohort. An exploratory electrocardiogram-based model was tested to predict ostial proximity. RESULTS: Electroanatomic identification of the left coronary artery (LCA) ostium was achieved in all patients (100%; 95%CI, 93.0%-100%), and of the right coronary artery (RCA) in 73% (95% CI, 58.3%-84.1%). In the CT-validation cohort, electroanatomic and CT-derived ostial locations demonstrated excellent 3-dimensional concordance (LCA: 1.30 ± 0.37 mm; RCA: 1.48 ± 0.43 mm). In the subsequent implementation cohort, electroanatomic coronary localization was performed without systematic CT use. No ischemic clinical events occurred during 404 days of follow-up; CT in the validation cohort showed no coronary lesions. The exploratory electrocardiogram-based model suggested artery-specific variables for predicting coronary proximity. CONCLUSIONS: Electroanatomic mapping with spline-based catheters enables accurate coronary ostia localization during LVOT ablation, consistently for the LCA and in most cases for the RCA. These findings support its use as a feasible anatomical reference to inform procedural decision-making.