Prognostic value of stress perfusion CMR: a real-world single-center experience in portuguese patients with known coronary artery disease

Background and aim

Adenosine stress perfusion cardiovascular magnetic resonance (CMR) has high diagnostic accuracy for detecting obstructive coronary artery disease (CAD), but its prognostic value remains less well established than other imaging modalities. This study evaluated the prognostic utility of stress CMR in a contemporary cohort of patients with known CAD.

Methods

This single-center retrospective study included patients with documented CAD who underwent stress CMR (1.5-T) between 2019 and 2023. Patients were followed for major adverse events: all-cause mortality, non-fatal myocardial infarction (MI), cardiovascular hospitalization, and late coronary revascularization (>12 months post-CMR). The primary endpoint was the composite of all-cause death and non-fatal MI. Survival analysis assessed the prognostic value of inducible ischemia (≥2 segments) and ischemic late gadolinium enhancement (LGE).

Results

A total of 485 patients (75%male; mean age 68 ±11years) were included; 62% had prior MI, 63% prior percutaneous coronary intervention (PCI), and 25% prior coronary artery bypass grafting (CABG). Inducible ischemia was present in 147 patients (30%) and ischemic LGE in 338 (70%); both were present in 93 (19%), and non-ischemic LGE in 89 (18%).

During a median follow-up of 2.9 years,126 events (26%) were recorded, including 92 primary outcomes (60 deaths, 32 non-fatal MIs), 70 cardiovascular hospitalizations, and 16 late revascularizations.

Survival analysis showed that the presence of both inducible ischemia or LGE was associated with a significantly higher risk of the primary composite endpoint (log-rank p<0.001), consistent across individual outcomes (all p<0.01). Ischemic LGE predicted the primary endpoint overall (p=0.001) and when combined with ischemia (p=0.04), whereas ischemia alone was not (p=0.24) but was associated with higher rate of late revascularization(p=0.01).

In multivariable analysis, with age (HR1.03, 95% CI 1.01–1.06; p=0.006), left ventricular ejection fraction (HR0.97; 0.96–0.99; p<0.001) and chronic kidney disease (HR1.47, 0.93–2.03; p=0.10), LGE or ischemia independently predicted the primary endpoint (HR2.52, 1.08–5.89; p=0.03).

In other models, ischemic LGE with concomitant ischemia (HR4.44, 1.62–12.21;p=0.004), was also an independent preditor, while ischemic LGE (p=0.07) was not. Inducible ischemia independently predicted late revascularization (HR 2.79, 1.01–7.66; p=0.047).

Conclusion

In patients with known CAD, stress CMR findings of ischemic LGE and/or inducible ischemia were associated with increased risk and were independent preditors of the primary endpoint while inducible ischemia was not, however it was associated with a higher rate of late revascularization. These findings highlight the prognostic value of stress perfusion CMR in risk stratification and support the routine assessment of myocardial scar, suggesting that inducible ischemia alone may have limited prognostic value.Freedom from primary outcome in years