Arrhythmic risk stratification in nonischemic dilated cardiomyopathy: moving beyond the ejection fraction criterion

The diagnosis of inherited or acquired cardiomyopathies is traditionally based on morphological and/or functional criteria. Nonischemic dilated cardiomyopathy (DCM) is diagnosed in the presence of a dilated left ventricle with decreased systolic function in the absence of significant coronary artery disease. DCM is associated with significant morbidity and mortality and may lead to heart failure (HF) development or ventricular arrhythmias (VA) and sudden cardiac death (SCD).

Currently the decision for the prophylactic use of an implantable cardioverter-defibrillator (ICD) is guided by a single criterion: that of LV ejection fraction (LVEF). Current guidelines recommend ICD implantation for all HF patients with LVEF≤35% to reduce total mortality and SCD risk, although more recent randomized clinical trials have challenged this approach. In the DANISH trial¹, the prophylactic ICD use in DCM with LVEF≤35% reduced SCD risk but not total mortality. Whilst LVEF is a strong predictor of mortality in DCM, the relationship becomes weaker in patients with severe systolic impairment. Moreover, although SCD is less frequent in patients with LVEF>35%, the proportion of SCD to all-cause mortality is higher in this subgroup.² Therefore, a more tailored approach to DCM treatment is needed, considering patient’s age and the lifetime exposure to the risk of SCD vs. the competing risks of HF death and non-cardiac mortality.³

Late gadolinium enhancement (LGE) by cardiac magnetic resonance (CMR) imaging can identify myocardial fibrosis, the substrate for VAs. The independent prognostic value of LGE has been confirmed in several, albeit relatively small, cohorts.³ Recently Di Marco et al. presented the results of one of the largest CMR cohorts in 1,165 patients with DCM who were followed up for a median of 36 months for the composite of appropriate ICD therapies, sustained monomorphic or polymorphic ventricular tachycardia, aborted cardiac arrest, and SCD. LGE positive patients (42%) had a 14-fold higher risk of VA events compared to LGE negative patients. Importantly, LGE retained its predictive value across all LVEF strata (≤20%, 21% to 35%, >35%), and the relationship was even stronger in patients with LVEF>35%. Furthermore, the authors have reiterated previous evidence confirming that certain patterns of fibrosis (such as septal or free-wall LGE) carry a higher arrhythmic risk. Using the combination of LVEF, LGE presence and distribution (“high-risk” LGE patterns) a risk stratification algorithm for SCD risk was developed that performed better than the LVEF criterion.

The study of Di Marco et al. is significant as it validates previous observations on the relationship between LGE and SCD risk in one of largest to date CMR cohorts of DCM patients and proposes a simple algorithm for the tailored approach to SCD risk. This is important as the risk of VA may be present even in forms of myocardial disease such as hypokinetic non-dilated cardiomyopathy or non-hypokinetic dilated LV which lie within the broader DCM spectrum, but typically do not fulfil the traditional criteria for ICD. In addition, certain forms of genetic cardiomyopathies due to mutations in genes encoding desmosomal proteins or FLNC are known to be associated with a higher arrhythmic risk, and the identification of such high-risk LGE phenotypes before the development of overt LV dysfunction, may contribute to the timely device implantation and help save lives.⁴

Given that the algorithm of Di Marco et al. is derived from a retrospective observational study, it is debatable whether it can change current clinical recommendations on ICD use in DCM. However, it is certainly moving in the right direction, beyond the simplistic LVEF cut-off and towards the personalised treatment of the DCM superfamily.