A decade-long trial answers a question doctors couldn't settle

Current guidelines for implantable cardioverter-defibrillators, the small devices that monitor heart rhythm and deliver corrective electrical pulses to prevent sudden cardiac death, draw a fairly hard line: patients qualify only if their left ventricular ejection fraction, a measure of how much blood the heart pumps with each beat, has dropped to 35% or lower despite optimal medication. That threshold leaves out a substantial group of heart failure patients with milder dysfunction, roughly 36% to 50% ejection fraction, even though most sudden cardiac deaths actually occur in exactly this broader, currently ineligible population. The CMR GUIDE trial, presented in a Hot Line session at ESC Congress 2026 in Munich and published simultaneously in JAMA, set out to test whether that guideline boundary is drawn in the right place, and the results suggest it may need to move, at least for younger patients.

Principal investigator Professor Joseph Selvanayagam of Flinders University and Flinders Medical Centre in Adelaide led the trial across 18 centers in Australia, Germany, and the United Kingdom, following patients for more than a decade to get a genuinely clear answer to a question the field has debated for years.

A trial built around a specific risk marker

Rather than testing ICDs across every patient with mild-to-moderate heart dysfunction, the CMR GUIDE trial used a specific enrollment criterion to identify which patients in that broader group actually carry meaningful sudden death risk: myocardial scarring detectable on cardiovascular magnetic resonance imaging. Scar tissue in heart muscle is a known substrate for the dangerous arrhythmias that ICDs are designed to catch and correct, and using CMR imaging to identify patients with confirmed scarring, rather than enrolling based on ejection fraction alone, let researchers target the subgroup most likely to actually benefit from device therapy.

The trial enrolled 353 patients with ischemic or non-ischemic cardiomyopathy, ejection fraction between 36% and 50% on optimal heart failure medication, and confirmed myocardial scarring on CMR imaging. Participants were randomized to receive either a full ICD or an implantable loop recorder, a device that monitors heart rhythm continuously but cannot deliver any corrective therapy, functioning as the trial's comparison arm. Most participants, 68%, were over age 70, and 18% were women.

Why the overall result looked disappointing at first glance

Measured against its primary endpoint, a composite of sudden cardiac death or hemodynamically significant ventricular arrhythmias causing loss of consciousness or a significant blood pressure drop, the trial did not find a statistically significant difference between the ICD and loop recorder groups across the full study population. On its own, that result might have been read as a disappointing outcome for a trial that took more than a decade to complete, seemingly confirming that the current 35% ejection fraction cutoff is roughly where it should be.

But averaging results across the entire study population obscured a genuinely important pattern hiding underneath it, one that only became visible once researchers examined how the intervention performed across different age groups.

The age-based finding that changes the picture

In a prespecified subgroup analysis, ICD implantation was associated with a 72% reduction in the primary endpoint among patients younger than 70, a substantial and statistically significant effect. Among patients aged 70 or older, no such benefit was observed. That's a striking split, and it explains why the trial's overall, age-blended result looked neutral: a large protective effect concentrated in younger patients was diluted into statistical insignificance once averaged against a much larger group of older patients who saw no benefit at all, especially given that 68% of the trial's participants fell into that older, non-benefiting category.

The secondary endpoint of sudden cardiac death specifically was reduced across the full study group, adding further support that the device therapy is doing genuine protective work, just concentrated in a particular age range rather than distributed evenly across the whole population.

Why age might matter this much

The researchers didn't fully explain the underlying biological mechanism behind the age-based split, but the pattern itself fits a broader theme showing up across recent cardiology research this year: age-related differences in how much benefit a given intervention actually delivers, even when the underlying disease process looks similar on paper. It's a dynamic that echoed elsewhere at this same ESC Congress, where a separate trial found statins delivered a genuine 30% cardiovascular event reduction in adults over 70 without translating into improved disability-free survival, a reminder that age-stratified analysis increasingly reveals genuinely different risk-benefit calculations across a patient's lifespan rather than a single answer applying uniformly to every age group.

For ICDs specifically, older patients with mild-to-moderate ejection fraction reduction likely face a different balance of competing mortality risks, including non-cardiac causes of death that a defibrillator does nothing to prevent, potentially diluting any arrhythmia-specific benefit the device might otherwise provide. Younger patients, with fewer competing health risks and a longer expected lifespan during which a dangerous arrhythmia could eventually occur, may simply have more to gain from a device designed to prevent one specific, catastrophic failure mode.

What this means for actual clinical practice

Selvanayagam was explicit about how this finding should translate into the exam room, describing it as evidence supporting a shared decision-making conversation rather than an automatic new eligibility rule. For a younger patient with mild-to-moderate ejection fraction reduction and confirmed myocardial scarring on cardiac MRI, this trial gives cardiologists genuine evidence to discuss ICD implantation as a real option, a conversation that current guidelines don't currently support having at all, since these patients fall outside the standard 35% ejection fraction threshold entirely.

That's a meaningfully different clinical reality than existed before this trial published. A decade-long, multi-country randomized trial finding a 72% relative risk reduction in a prespecified subgroup is a considerably stronger evidentiary basis for expanding treatment conversations than anecdotal cardiologist judgment or smaller observational studies could provide, even though the result falls short of an outright guideline change on its own.

What still needs to happen before guidelines actually shift

A single trial, even a well-designed one spanning over a decade and multiple countries, typically isn't sufficient on its own to trigger an immediate formal guideline revision, particularly when the headline primary endpoint result was neutral and the meaningful finding emerged from a subgroup analysis rather than the trial's main pre-specified comparison. Cardiology societies will likely want to see this finding replicated, or at minimum discussed extensively within the broader body of evidence on CMR-guided risk stratification, before formally lowering the ejection fraction threshold for ICD eligibility in younger patients. That kind of careful, incremental integration of new evidence into clinical guidelines is a similar pattern to how this year's ESC also introduced new combined screening guidance connecting cardiovascular and kidney disease management, reflecting a broader trend toward more individualized, risk-stratified cardiology practice built on layered evidence rather than a single trial's result alone. For now, the CMR GUIDE trial gives younger patients with mild-to-moderate heart dysfunction and confirmed scarring a genuine, evidence-backed reason to have a conversation with their cardiologist that current guidelines simply didn't support having before.