A strategy of primary prevention ICD implantation guided by cardiac magnetic resonance imaging: effect on appropriate therapies during follow-up

Background

Cardiac magnetic resonance (CRM) imaging with late gadolinium enhancement (LGE) provides superior tissue characterization and can refine arrhythmic risk stratification in patients who are candidates for implantable cardioverter defibrillator (ICD) therapy. However, its impact on clinical decision-making and outcomes when guiding ICD implantation for primary prevention remains uncertain.

Purpose

Our objective was to evaluate whether CMR-guided ICD implantation and myocardial fibrosis (in the form of LGE) extent were associated with appropriate device therapies during follow-up.

Methods

We retrospectively analyzed consecutive patients who underwent ICD implantation in a single-center Hospital, from March 2016 to February 2024, in primary prevention with CMR prior to the procedure. Patients were divided according to whether the ICD implantation was guided by CMR findings or not. Baseline clinical and imaging characteristics were compared between groups using appropriate statistical tests. The primary endpoint was antitachycardia pacing (ATP) or appropriate shock. Logistic regression models including only variables derived from CMR were used to identify independent imaging predictors. Event-free survival was compared using Kaplan-Meier and log-rank analyses.

Results

A total of 100 patients were included (59 CMR-guided and 41 non-CMR-guided). Baseline characteristics were similar between groups (Table 1). The CMR-guided group showed a trend toward greater transmural LGE (45.6% vs. 26.3%, p = 0.092). No significant differences in left ventricular ejection fraction (LVEF), total LGE extent, or clinical outcomes were found between the groups.

During a median follow-up of approximately 600 days, ATP/appropriate shock occurred in 25.4% of CMR-guided patients compared with 29.3% of non-CMR-guided patients (p = 0.49). Kaplan-Meier analysis confirmed no significant differences in arrhythmia-free survival (log-rank p = 0.486).

Among CMR parameters, no single variable (including LVEF, indexed left ventricular mass, LGE extent, transmural pattern, or right ventricular function) independently predicted arrhythmic events. However, patients with extensive LGE (≥3 segments) had a two-fold risk of appropriate ICD therapies (OR 2.3, 95% CI 1.0-5.4, p = 0.05). After adjustment for LVEF, this association remained borderline (OR 1.9, p = 0.08), whereas LVEF itself was an independent predictor (p = 0.02).

Conclusions

In patients undergoing ICD implantation for primary prevention, CMR-guided implantation does not appear to affect arrhythmic event rates. However, extensive myocardial fibrosis (≥3 LGE segments) identifies individuals at higher arrhythmic risk, partly independent of LVEF. Although the current findings may reflect limited statistical power, larger multicenter studies are needed to clarify the prognostic and decision-making role of CMR before ICD therapy.Table 1