Non-invasive assessment of the ventricular arrhythmogenic substrate beyond scar in non-ischemic cardiomyopathy: NIAVAS-2 study
J Reventos-Presmanes, E Guillen, E Ayauja, E Arbelo, E Guasch, J M Tolosana, L Mont, A M Guillem, A M Climent, A Porta-Sanchez, I Roca-LuqueAbstract
Background
Ventricular tachycardias (VTs) are a major cause of morbidity and mortality in patients with non-ischemic cardiomyopathy (NICM), yet accurate risk stratification remains a clinical challenge. Fibrosis in NICM is often mid-wall or patchy and frequently adjacent or dissociated from regions of slow conduction or re-entry, making substrate localization based on imaging particularly challenging. A non-invasive marker capable of identifying the functional arrhythmogenic substrate, rather than just structural changes is therefore of particular interest.
Purpose
To determine if a non-invasive functional metric, regional activation dispersion (rAD), can identify the arrhythmogenic substrate defined by deceleration zones (DZs) and distinguish it from non-viable scar in NICM patients.
Methods
We prospectively enrolled 47 NICM patients: 27 (55.0±20.2 years, 33.3% dilated) referred for VT ablation (VT+) and 20 (66.2±8.8 years, 60% dilated) with an implantable cardioverter-defibrillator (ICD) for primary prevention without documented VT (VT–). All underwent late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) before ablation or ICD implantation. Non-invasive ECGI was acquired periprocedural (VT+) or during medical consultation in (VT–). The ventricles were segmented into 15 regions (9 LV, 6 RV), and rAD was computed during sinus rhythm (SR) and right ventricular pacing (RVp). Scarred (LGE+) and non-scarred (LGE–) regions were identified in all subjects, and in the VT+ group, arrhythmogenic substrate was defined as DZs identified by invasive electroanatomic mapping (EAM).
Results
In VT+ patients, regions containing deceleration zones (DZ+) exhibited higher rAD compared to regions without (DZ–) during both SR (76.9 ± 7.5 vs 47.5 ± 3.4 ms; p < 0.001) and RVp (81.3 ± 5.7 vs 55.7 ± 3.8 ms; p < 0.001), as shown in Figure 1A. Importantly, this association was independent of scar (Figure 1B), indicating that rAD captures arrhythmogenic substrate remodelling beyond what is detectable by LGE-CMR (Figure 1C). In contrast, VT– patients with structural alterations showed no significant rAD differences between LGE+ and LGE– regions in either SR or RVp, suggesting that their substrate was electrically quiescent. Representative VT– and VT+ patients with ventricular segments presenting scar based on LGE-CMR are shown in Figure 2.
Conclusion
Non-invasive regional activation dispersion identifies the arrhythmogenic substrate defined by EAM (DZs) in NICM patients, even in the absence of scar on MRI. This ability to distinguish between functionally active arrhythmogenic substrate and quiescent scar based on ECGI- rAD may improve risk stratification and patient selection for ICD in primary prevention.