Improved Major Arrhythmic Event Prediction in Non-Ischemic Cardiomyopathy

By Michael H. Crawford, MD, Editor

Synopsis: An international multicenter registry study of patients with non-ischemic cardiomyopathy without a history of major arrhythmic adverse cardiac events (MAACE) that compared a new cardiac magnetic resonance imaging score for predicting MAACE and consequently determining the need for a primary prevention implantable cardioverter defibrillator has found that this new score is superior to the standard left ventricular ejection fraction criterion.

Source: Guaricci AI, Carrabba N, Romano SM, et al. Redefining the risk of major arrhythmic events in non-ischaemic cardiomyopathy: Insights from the DERIVATE-NICM study. Eur Heart J Cardiovasc Imaging. 2025;26(10):1609-1619.

The current criterion for implantable cardioverter defibrillator (ICD) placement in patients with non-ischemic cardiomyopathy is a left ventricular (LV) ejection fraction (EF < 35%. The potential of improving this decision using tissue characterization by cardiac magnetic resonance (CMR) imaging was investigated in a large international registry study. A multiparametric CMR score was derived (DERIVATE 1.0), which included male sex (two points), LV end-diastolic volume index > 121 mL/m (three points), and > 3 LV segments with mid-wall late gadolinium enhancement (LGE; two points) for a total of seven points maximum.

In the DERIVATE 1.0 study, major adverse arrhythmic cardiac events (MAACE) were better predicted compared to LVEF. However, it was observed that the location of the LV segments with LGE might be a better measure than just the presence of LGE. Thus, DERIVATE 2.0 was developed, which included mid-wall LGE location in the septum, inferior wall, and inferolateral wall (three points for a total of eight maximum points).

DERIVATE 2.0 was a prospective observational registry study of consecutive patients with heart failure (HF) caused by non-ischemic cardiomyopathy without a history of MAACE referred to 21 sites in Europe or the United States to determine the need for primary prevention ICD. The aim of DERIVATE 2.0 was to compare 2.0 to 1.0 and LVEF for predicting MAACE. Included were patients with chronic HF and EF < 50%. Excluded were patients with ischemic cardiomyopathy, decompensated HF, acute myocarditis within the last three months, severe valve disease, congenital heart disease, and primary or secondary heart disease (e.g., hypertrophic, amyloid).

All patients had a CMR and transthoracic echocardiogram within three months of enrollment, and the CMRs were evaluated blinded by a core lab. Clinical follow-up was conducted at the individual centers. Subjects without ICDs had Holter monitoring to detect arrhythmias. The primary endpoint was MAACE, defined as sudden cardiac death, aborted sudden cardiac death, or sustained ventricular tachycardia (VT; > 30 sec or hemodynamic instability).

The centers enrolled 1,384 patients, mean age 56 years, and 68% men. The median follow-up was 959 days. MAACE occurred in 128 patients (9%). Multivariable analysis showed that MAACE was independently associated with male sex (hazard ratio [HR], 1.61; 95% confidence interval [CI], 1.05 to 2.45; P = 0.03), LVEF per point (HR, 0.98; 95% CI, 0.96 to 0.99; P = 0.005), and mid-wall LGE in the septum, inferior, and inferolateral wall (HR, 1.07; 95% CI, 1.05 to 1.09; P < 0.001). Also, DERIVATE 2.0 provided incremental prognostic value over 1.0 with a net reclassification index of 55% (P < 0.001) and over LVEF with a net reclassification index of 35% (P < 0.001). An independent validation cohort of 244 patients who met the inclusion criteria for the study was studied and showed a similar rate of MAACE (8%) and a net reclassification index of 38% vs. the EF criterion.

The authors concluded that the improved predictive value of DERIVATE 2.0 is largely the result of the inclusion, with LVEF and male sex, of the location of mid-wall myocardial LGE by CMR. Thus, randomized controlled trials of this CMR-guided strategy for ICD placement in non-ischemic cardiomyopathy patients compared to the standard LVEF criterion now are needed.

Commentary

It is well known that the presence of a myocardial scar can be a nidus for arrhythmias. This study shows that an LV scar detected by LGE on CMR is a strong predictor of future MAACE, especially if it is in certain locations, and that its predictive value is superior to LVEF by echocardiography. Also, a scar is rather permanent, whereas EF can change over time with therapy and it has a wide standard deviation on repeated studies. Thus, the authors believe that the use of LVEF alone to determine which patients with non-ischemic cardiomyopathy will benefit from a primary prevention ICD no longer is appropriate. In addition, they have developed a simple clinical score using LGE, EF, and sex to estimate the risk of MAACE in such patients.

There are limitations to DERIVATE 2.0. First, there is a referral bias of selecting only patients from centers with well-developed CMR programs. Second, the MAACE rate was low (< 10%) in part because of the exclusion of higher-risk patients, such as those with hypertrophic cardiomyopathy and arrhythmogenic right ventricular cardiomyopathy. Third, none of the patients were taking sodium-glucose cotransporter 2 inhibitors or sacubitril/valsartan, which might have further lowered the MAACE rate. Fourth, genetics and biomarkers were not determined, which also may affect the risk of MAACE. However, these results strongly suggest that the use of CMR for risk prediction in cardiomyopathy patients has arrived. All that is needed now is a randomized controlled outcomes trial to prove the superiority of this CMR strategy compared to using EF alone for determining who needs an ICD.

Michael H. Crawford, MD, is Professor Emeritus of Medicine and Consulting Cardiologist, UCSF Health, San Francisco.