Magnetocardiographic Rotation Score and Arrhythmic Endpoints in Primary-Prevention ICD Recipients: An Exploratory Substudy of the Magneto-SCD Trial

Background/Objectives: Primary-prevention implantable cardioverter-defibrillator (ICD) selection remains dominated by left ventricular ejection fraction and clinical heart failure markers, which incompletely distinguish between patients who will experience ventricular arrhythmia and those whose prognosis is limited by competing non-arrhythmic mortality. The Magneto-SCD trial recently reported novel rotation-based magnetocardiography (MCG) indices associated with future appropriate ICD therapies. This substudy explores whether MCG Rotation Score provides incremental information in primary-prevention ICD recipients, particularly beyond NYHA status and MADIT-arrhythmic and non-arrhythmic risk scores. Methods: We included all Magneto-SCD trial participants with a primary-prevention indication with non-missing follow-up/MCG Rotation Score data. The full primary-prevention cohort comprised 54 participants; the post-MI subgroup comprised 28. The primary endpoint was time to first appropriate ICD therapy. Death without prior therapy was used as a pragmatic competing endpoint and not assumed to represent adjudicated non-arrhythmic death. Appropriate shock was a secondary endpoint. Two-year cumulative incidence functions were estimated using Aalen–Johansen methods. Cause-specific Cox models assessed Rotation Score alone and after addition to MADIT-derived scores; nested likelihood-ratio tests evaluated model fit. Results: Of 104 participants, 90 had analysable MCG data and 65 had a primary-prevention ICD indication; after exclusions for missing patients, 54 patients remained (mean age 63.8 ± 13.2 yrs, 45(83.3%) male). Appropriate ICD therapies occurred in 10 (18.5%), including appropriate ICD shock in eight (14.8%); 10 patients (18.5%) died without prior therapy. The median follow-up was 1211.5 days (IQR 793.5–1396.0). Each SD increase in Rotation Score was associated with appropriate ICD therapy with HR 1.64 (95% CI 0.93–2.87; p = 0.086), appropriate ICD shock with HR 2.07 (95% CI 1.15–3.72; p = 0.015), death without prior therapy with HR 1.08 (95% CI 0.54–2.17; p = 0.824), and death without prior ICD shock with HR 0.95 (95% CI 0.48–1.90; p = 0.890). Adding Rotation Score did not significantly improve the primary therapy model beyond MADIT components (LRT p = 0.065) or MADIT benefit score (LRT p = 0.109). Model fit improved for the secondary appropriate ICD shock endpoint when Rotation Score was added to MADIT components (LRT p = 0.016), based on eight shock events. Conclusions: Rotation Score showed an association pattern more closely aligned with arrhythmic endpoints, particularly appropriate ICD shock, than with death without prior therapy. Incremental prognostic value for the primary endpoint was not established. These findings are hypothesis-generating and require confirmation in adequately powered, prospectively designed and externally validated cohorts before any role in clinical risk stratification or ICD decision-making can be determined.