Primary prevention ICD implantation did not reduce sudden cardiac death or hemodynamically significant ventricular arrhythmia compared to an implantable loop recorder.
14 (7.8%) vs 16 (9.2%) at Median 6.3 years (IQR 4.8-7.6)
Key points
Context & question:Primary-prevention ICDs are restricted to ejection fractions of 35% or below, yet most sudden cardiac deaths occur in patients with higher ejection fractions. CMR GUIDE asked whether CMR-detected myocardial scar identifies patients with mild-moderate systolic dysfunction (LVEF 36-50%) who benefit from a defibrillator.
Finding & limits:An ICD did not reduce the primary composite of sudden cardiac death or hemodynamically significant ventricular arrhythmias versus an implantable loop recorder in patients with LVEF 36-50% and CMR-detected scar, though sudden cardiac death alone fell and patients under 70 saw a 72% reduction in the primary endpoint. Statistical certainty: only 30 primary events occurred; the wide CI (0.37-1.58) allows meaningful benefit or harm.
“Overall, the primary endpoint was neutral. However, younger patients appeared to derive clinical benefit from ICDs. We suggest that data from the CMR GUIDE trial are discussed with younger patients in whom ICDs may be an option in a shared decision-making process.”
Endpoint critique·Electrophysiology · Baptist Health Louisville
“I am almost done with my column on CMR-GUIDE. I realize trials are hard and costly but that does not mean we should over-interpret this decidedly null trial result There is neither evidence for ICD benefit in this group nor is there support for the "scar" theory of selecting pts”
“CMR GUIDE is key trial for the EP field - gives the first randomized data we have for the LVEF 36-50% population with myocardial scar. The positive signal in younger pts, reductions in SCD rates are important and nuanced discussions to have with out pts about ICD therapy.”
Endpoint critique·Heart Failure / Cardiomyopathy · University Hospital Southampton NHS Foundation Trust
“@josephselvacmr CMR-GUIDE did not meet its primary endpoint in patients with LVEF 36–50% and myocardial scar: ICD vs ILR HR 0.76 (95% CI 0.37–1.58). But the components are thought-provoking: SCD 3 vs 10 (HR 0.26), while haemodynamically significant VA was numerically higher with ICD”
“@drjohnm CMR GUIDE is informative. It may be underpowered but is reassuring that LVEF 36-50% with scar do NOT show mortality with ICDs & reports Cx of & impact on Mini living with HF of ICDs. Also reinforces low event rates in the current era with GDMT”
“MR GUIDE found no significant reduction in sudden cardiac death or haemodynamically significant ventricular arrhythmia with a scar-guided ICD strategy in patients with LVEF 36–50% and CMR-defined scar. HR 0.76, 95% CI 0.37–1.58, P=0.47.”
Open questions from this trial that would benefit from expert input.
Ask about this paper
Answers are grounded in this paper and cite their sources.
Full evidence
Why the trial, tables, limitations, PICO, abstract, media & sources
Why the trial?
Primary-prevention ICDs are restricted to ejection fractions of 35% or below, yet most sudden cardiac deaths occur in patients with higher ejection fractions. CMR GUIDE asked whether CMR-detected myocardial scar identifies patients with mild-moderate systolic dysfunction (LVEF 36-50%) who benefit from a defibrillator.
Does a primary prevention implantable cardioverter-defibrillator (ICD) reduce the composite of sudden cardiac death or hemodynamically significant ventricular arrhythmia in adults with cardiomyopathy, LVEF 36-50%, and CMR-defined myocardial scar?
View moreView less
Evidence details
What drove the result?
Outcome
ICD
ILR
Sudden cardiac death or HSVA HR 0.76 (95% CI 0.37-1.58) - not significant
ICD 14 (7.8%)
ILR 16 (9.2%)
Sudden cardiac death Also a secondary outcome - lower with ICD (HR 0.26, 0.07-0.95)
ICD 3 (1.7%)
ILR 10 (5.8%)
Significant ventricular arrhythmia (HSVA) Numerically higher with ICD (HR 1.77, 0.65-4.81); components moved in opposite directions
ICD 12 (6.7%)
ILR 6 (3.5%)
Primary outcome, age <70 years Prespecified subgroup - HR 0.28 (0.09-0.89); reversed at >=70 (HR 2.33); P=.01 for interaction
ICD 3.3%
ILR 10.0%
Limitations & tradeoffs
Statistical certainty
only 30 primary events occurred; the wide CI (0.37-1.58) allows meaningful benefit or harm.
Representation
participants were 82% men and 72% had an ischemic etiology.
Design limitations
the trial was open-label.
Structured PICO
Does a primary prevention implantable cardioverter-defibrillator (ICD) reduce the composite of sudden cardiac death or hemodynamically significant ventricular arrhythmia in adults with cardiomyopathy, LVEF 36-50%, and CMR-defined myocardial scar?
PPopulation
353 adults with ischemic or nonischemic cardiomyopathy, LVEF 36% to 50%, and CMR-defined myocardial scar, followed for a median of 6.3 years.
Composite of sudden cardiac death (SCD) or hemodynamically significant ventricular arrhythmia (HSVA)composite endpoint
Main Result
Hazard Ratio0.76 (95% CI 0.37–1.58)
Absolute Event Rate7.8% vs 9.2%
In patients with an LVEF of 36% to 50% and CMR-defined myocardial scar, prophylactic ICD implantation did not significantly reduce the composite of sudden cardiac death or hemodynamically significant ventricular arrhythmia compared to an implantable loop recorder.
Where experts stand
Cardiac MRI evidence still forming
Experts read CMR GUIDE as a null trial that does not make scar on cardiac imaging a new reason to implant a defibrillator, though some think its sudden-death and younger-patient signals deserve discussion.
Most clinicians agree the trial missed its main goal and does not support defibrillators for patients selected by scar alone when heart function is only mildly reduced. They split on how much weight the lower sudden-death rate and the apparent benefit in younger patients should carry. The live question is whether the trial disproves the scar-selection idea or simply could not show a benefit because sudden death was rare.
Agreement
Several clinicians agree the primary endpoint was negative and that scar on imaging alone does not justify a primary-prevention ICD in patients with LVEF 36-50%.
3 clinicians say this directly
Main debate
Null result that refutes scar selection, or a hard trial with signals worth acting on?
Null, do not over-read2
These posts say the trial shows no ICD benefit and no support for scar-based selection.
vs
Signals still matter2
These posts say the sudden-death and younger-patient findings deserve weight and that low event rates may hide a real effect.
Not placed3 reporting, watching, or context only
What they’re arguing about
supportiveneutralcautiouscritical
Counts are expert takes we classified by axis. Tap a row to see the takes behind its count.
Still unclear
Experts question whether the benefit in patients under 70 is real, since subgroups were not adjusted for multiplicity. It is also open whether low absolute sudden-death risk, rather than the scar hypothesis itself, explains the null result. More device-related events with ICDs and more recorded arrhythmias also remain to be weighed.
Implication
Does not support primary prevention ICD in LVEF 36-50% with scar; challenges CMR late gadolinium enhancement alone for defibrillator selection.
Abstract
Importance: Current guidelines do not recommend primary prevention implantable cardioverter-defibrillators (ICDs) unless a patient's left ventricular ejection fraction (LVEF) is 35% or less. Many sudden cardiac deaths (SCD) occur when LVEF is 36% to 50%. Myocardial scar (a key arrhythmic substrate) can be assessed by late gadolinium enhancement on cardiovascular magnetic resonance (CMR), but robust evidence is lacking regarding a scar-based approach to ICD insertion. Objective: To determine whether implantation of ICDs reduces SCD or hemodynamically significant ventricular arrhythmia (HSVA) in patients with an LVEF of 36% to 50% and myocardial scar. Design, Setting, and Participants: An open-label randomized clinical trial enrolled adults between 2015 and 2022 who had ischemic or nonischemic cardiomyopathy, an LVEF of 36% to 50%, CMR-defined myocardial scar, and were receiving guideline-directed medical therapy at 18 sites in Australia, Germany, and the UK. Follow-up assessments were completed in 2026. Interventions: A primary prevention ICD (n = 180) vs an implantable loop recorder (ILR) (n = 173). Main Outcomes and Measures: The primary composite outcome was SCD or HSVA. Five secondary outcomes were evaluated: SCD, HSVA, heart failure-related hospitalization, cardiovascular mortality, and all-cause mortality. Results: Of 353 patients randomized (median age, 65 years [IQR, 57-61 years]; 18% female; and 72% had an ischemic etiology), 70% had an LVEF of 40% or greater. The median follow-up was 6.3 years (IQR, 4.8-7.6 years). The primary composite outcome occurred in 14 patients (7.8%) in the ICD group compared with 16 patients (9.2%) in the ILR group (hazard ratio [HR], 0.76 [95% CI, 0.37-1.58]). For the individual components of the primary composite outcome, SCD occurred in 3 patients (1.7%) vs 10 patients (5.8%) in the ILR group (HR, 0.26 [95% CI, 0.07-0.95]) and HSVA occurred in 12 patients (6.7%) vs 6 patients (3.5%), respectively (HR, 1.77 [95% CI, 0.65-4.81]). The rates for all-cause mortality, cardiovascular mortality, and heart failure-related hospitalization were similar between groups. In a prespecified analysis of 6 subgroups, the primary outcome occurred less often in patients younger than 70 years in the ICD group (3.3%) vs patients in the ILR group (10.0%) (HR, 0.28 [95% CI, 0.09-0.89]) but not in those aged 70 years or older (16.9% vs 7.5%, respectively) (HR, 2.33 [95% CI, 0.75-7.26]; P = .01 for interaction). Conclusions and Relevance: Implantation of an ICD did not reduce the composite outcome of SCD or HSVA in patients with an LVEF of 36% to 50% and myocardial scar. Trial Registration: ClinicalTrials.gov Identifier: NCT01918215.
Cite This Study
Selvanayagam et al. (2026) conducted an RCT in Ischemic or nonischemic cardiomyopathy with LVEF 36% to 50% and myocardial scar (n=353). Primary prevention implantable cardioverter-defibrillator (ICD) vs. Implantable loop recorder (ILR) was evaluated on Composite of sudden cardiac death (SCD) or hemodynamically significant ventricular arrhythmia (HSVA) (HR 0.76, 95% CI 0.37-1.58). Primary prevention ICD implantation did not reduce sudden cardiac death or hemodynamically significant ventricular arrhythmia compared to an implantable loop recorder.