Prevalence of Angiographic Coronary Artery Disease in Systolic Heart Failure Without Infarct-Pattern Late Gadolinium Enhancement: A Systematic Review and Meta-analysis
Ciaran Grafton-Clarke, Maryum Farooq, Alexander Gall, Nicolas Sawh, Cristina Dinescu, Colin Berry, Mark C Petrie, John P Greenwood, Pankaj Garg, Peter P SwobodaAbstract
Aims
Rest late gadolinium enhancement cardiovascular magnetic resonance (LGE-CMR) identifies infarct-pattern scar but does not assess coronary anatomy. We estimated the prevalence of angiographic coronary artery disease (CAD) among adults with heart failure and reduced ejection fraction (HFrEF) or otherwise unexplained left ventricular systolic dysfunction without infarct-pattern LGE.
Methods and results
MEDLINE, Embase, CENTRAL, and ClinicalTrials.gov were searched to 27 July 2026. Eligible studies provided separable scar-negative denominators and corresponding angiographic CAD counts from invasive coronary angiography (ICA). Scar-negative was defined as the absence of subendocardial or transmural LGE in a coronary distribution. Random-effects logit meta-analysis used restricted maximum likelihood and Hartung-Knapp confidence intervals. Twenty-two reports representing 20 studies included 2,236 paired LGE-CMR and ICA. 818 patients had infarct-pattern scar, and 1,418 were scar-negative. Eight studies contributed 44 CAD cases among 468 scar-negative patients for ≥50% stenosis, giving a pooled prevalence of 10.8% (95% confidence interval [CI] 7.4–15.4). Five studies contributed 19 CAD cases among 271 scar-negative patients for ≥70% stenosis, giving a pooled prevalence of 7.6% (95% CI 4.4–12.6). Seven studies contributed 33 CAD cases among 679 scar-negative patients for high-risk-pattern CAD, giving a pooled prevalence of 4.9% (95% CI 2.2–10.6) and a 95% prediction interval of 0.8–24.8%. These strata represented different study subsets and were not nested.
Conclusion
Among selected adults with HFrEF or otherwise unexplained LVSD who underwent paired LGE-CMR and ICA, absence of infarct-pattern LGE did not exclude angiographic CAD. Further coronary assessment should be considered when resolving that uncertainty would alter diagnosis or management.