Long-term follow-up analysis of a highly characterized arrhythmogenic cardiomyopathy cohort with classical and non-classical phenotypes–a real-world assessment of a novel prediction model: does the subtype really matter

To provide long-term outcome data on arrhythmogenic cardiomyopathy (ACM) patients with non-classical forms [left dominant ACM (LD-ACM) and biventricular ACM (Bi-ACM)] and an external validation of a recently proposed algorithm for ventricular arrhythmia (VA) prediction in ACM patients.

Demographic, clinical, and outcome data were retrieved from all ACM patients encountered at our institution. Patients were classified according to disease phenotype (R-ACM; Bi-ACM; LD-ACM). Overall and by phenotype long-term survival were calculated; the novel Cadrin-Tourigny et al. algorithm was used to calculate the a priori predicted VA risk, and it was compared with the observed outcome to test its reliability. One hundred and one patients were enrolled; three subgroups were defined (R-ACM, n = 68; Bi-ACM, n = 14; LD-ACM, n = 19). Over a median of 5.41 (2.59–8.37) years, the non-classical form cohort experienced higher rates of VAs than the classical form [5-year freedom from VAs: 0.58 (0.43–0.78) vs. 0.76 (0.66–0.89), P = 0.04]. The Cadrin-Tourigny et al. predictive model adequately described the overall cohort risk [mean observed-predicted risk difference (O-PRD): +6.7 (−4.3, +17.7) %, P = 0.19]; strafing by subgroup, excellent goodness-of-fit was demonstrated for the R-ACM subgroup (mean O-PRD, P = 0.99), while in the Bi-ACM and LD-ACM ones the real observed risk appeared to be underestimated [mean O-PRD: −20.0 (−1.1, −38.9) %, P < 0.0001; −22.6 (−7.8, −37.5) %, P < 0.0001, respectively].

Non-classical ACM forms appear more prone to VAs than classical forms. The novel prediction model effectively predicted arrhythmic risk in the classical R-ACM cohort, but seemed to underestimate it in non-classical forms.