Longitudinal Trends in Myocardial Metabolic Uptake Following Stereotactic Arrhythmia Radioablation Using 2-Deoxy-2-[18F]Fluoro-D-Glucose Positron Emission Tomography-Computed Tomography

Abstract

Purpose: Ventricular tachycardia (VT) is a life-threatening rhythm disturbance of the heart that has been shown to be amenable to treatment with stereotactic body radiation therapy, known as stereotactic arrhythmia radioablation (STAR). The metabolic impact of directly targeting arrhythmogenic regions of scarred left ventricle (LV) is unknown. In this study, the relationship between radiation dose and 2-deoxy-2-[¹⁸F]fluoro-D-glucose positron emission tomography (¹⁸FDG-PET) uptake in LV segments following STAR was interrogated.

Methods and materials: Patients with VT in the ENCORE-VT trial received a 25 Gy single-fraction of STAR to an electrophysiology guided target volume planned on 4-dimensional computed tomography (CT). Patients underwent ¹⁸FDG-PET/CT ± magnetic resonance imaging at baseline, day 3, and day 90. Auto-segmentation of the LV into the American Heart Association 17-segment model was performed on planning CTs and PET/CTs, allowing median radiation dose and standardized uptake value (SUV) to be extracted. Change in segment median SUV was correlated with dose and clinical outcomes (implanted cardiac defibrillator shock, death).

Results: Nineteen patients received STAR. Median VT episodes were reduced from 119 in the 6 months before versus 3 in the 6 months after STAR (P < .001). Targeted LV segments had lower baseline ¹⁸FDG SUV than nontargeted segments (P < .001). Median SUV of targeted segments at baseline, day 3, and day 90 were 1.43, 1.67, and 2.52, respectively (P = .012). At day 3, a ≥20% SUV increase in targeted segments (n = 7/19) was associated with improved survival (hazard ratio [HR], 0.10; P = .005). At day 90, there was a trend toward a weak association between increased median whole LV myocardium SUV and improved LV ejection fraction (R² = 0.355, P = .158). In logistic regression models, SUV values, together with cardiomyopathy type, VT storm occurrence, and amiodarone use, predicted implanted cardiac defibrillator shocks (P = .027) and death (P = .016).

Conclusions: In this small hypothesis-generating study, late SUV increases (day 90) were observed in targeted segments, consistent with VT burden reduction. Further longitudinal studies with ¹⁸FDG-PET/CT after STAR will define how SUV dynamics can be optimally incorporated as a biomarker in practice.