p.63 抽出テキスト

p.Arg14del variant, variant-specific [https://plnriskcalculator.shinyapps. io/final_shiny])542 risk-prediction scores have been developed that consider genotype characteristics and additional phenotypic features. Where such scores are available, they should be used to guide primary prevention ICD implantation (Figure 17). As discussed in Section 7.1.5, the Task Force acknowledges the challenges associated with defining universal thresholds for acceptable risk across different cardiomyopathy phenotypes, but is of the opinion that a similar approach to that taken in risk stratification for HCM is reasonable. Although the 2022 ESC Guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death have suggested a higher threshold of 10% risk at 5 years to guide primary prevention ICD implantation in patients with DCM and LMNA variants,3 this Task Force recommends shared decision-making based on real-world data as well as individual preferences, beliefs, circumstances, and values. Gaps in evidence should be shared with patients, and competing risks related to the disease (heart failure, stroke), and to age and comorbidity, as well as device-related complications, should be discussed. In support of this, a recent study validating the LMNA-risk VTA calculator overestimated arrhythmic risk when using ≥7% predicted 5-year risk as threshold (specificity 26%, C-statistic 0.85), despite a high sensitivity.871 Importantly, there are also data to suggest that other genotypes (e.g. TTN truncating variants) are associated with recovery of LVEF with standard heart failure criteria from the 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure.185,867,872

In patients without a high-risk genotype and LVEF >35%, the presence and extent of myocardial scarring determined by LGE on CMR imaging can be helpful in risk stratification in patients with DCM.873,874 Late gadolinium enhancement is observed in 25–35% of patients with DCM and its presence is a strong risk marker for all-cause mortality and ventricular arrhythmias, both in retrospective and prospective studies. A recent prospective study of 1020 DCM patients with a median follow-up of 5.2 years showed that myocardial scar had a strong and incremental prognostic value for predicting SCD, while LVEF ≤35% was not associated with SCD.138 In another study, a risk calculator was developed that among others, incorporated the presence of LGE on CMR imaging540, although this has not yet been externally validated. There are at least two ongoing trials of ICD therapy according to the presence of scar on CMR imaging, including DCM patients (NCT04558723 and NCT03993730), but the Task Force’s opinion is that the existing level of evidence can support using LGE to guide ICD implantation in subgroups of patients with DCM (Figure 17). Additional risk factors, such as syncope or the presence of NSVT and burden of ventricular ectopy (VE), may also help guide ICD implantation. There are no data currently to support a specific threshold for VE burden, and this will depend on the underlying genotype and other

Cardiac arrest or VT with haemodynamic compromise

LVEF <35%

Gene specific risk assessmenta

High risk ICD (Class I)

ICD (Class IIa)

ICD (Class IIa/IIb)b

N

Additional risk factorsc

N

N

High-risk gene

N Patients with DCM/NDLVC

Y

Y

Y

Y

ICD (Class IIb)

Figure 17 Implantation of implantable cardioverter defibrillators in patients with dilated cardiomyopathy or non-dilated left ventricular cardiomyopathy flowchart. CMR, cardiac magnetic resonance; DCM, dilated cardiomyopathy; ICD, implantable cardioverter defibrillator; LGE, late gadolinium enhancement; LVEF, left ventricular ejection fraction; NDLVC, non-dilated left ventricular cardiomyopathy; VE, ventricular ectopic beats; VT, ventricular fibrillation. aSee Table 21. bStrength of recommendation depends on gene and context. cAdditional risk factors include syncope, LGE presence on CMR.

ESC Guidelines                                                                                                                                                                                          3565

Downloaded from academic.oup.com/eurheartj/article/44/37/3503/7246608 by guest on 07 October 2026