Outcomes with Revascularization vs. Medical Therapy According to Plaque Burden from Coronary Computed Tomography Angiography

Abstract

Aims: We aimed to investigate, whether plaque burden from coronary computed tomography angiography (CCTA) could be used to identify patients potentially benefitting from revascularization.

Methods and results: We assessed consecutive patients undergoing CCTA and selective 15O-water perfusion positron emission tomography for evaluation of coronary artery disease (CAD) at two tertiary care centers in Finland and the Netherlands. Per-patient percent atheroma volume (PAV) and maximum per-vessel PAV in each patient was quantified by artificial intelligence-guided quantitative computed tomography (AI-QCT). We constructed a Cox regression for death, myocardial infarction (MI), or unstable angina pectoris (uAP) including continuous PAV, revascularization, and their interaction, adjusted for calcium score, ischemia, cardiovascular risk factors, symptoms, and medication in a subcohort of 2233 patients (206 events;median follow-up 6.8 years). There was significant interaction between revascularization and continuous PAV on patient-level (p-interaction=0.042) and vessel-level (p-interaction=0.026). Revascularization was associated with a significantly lower event rate at per-patient PAV 22% (HR 0.70, 95% CI 0.43-0.98) and per-vessel PAV 22% (HR 0.64, 95% CI 0.29-0.99) or higher. In subgroup analyses, after adjustment for age, sex, cardiovascular risk factors, ischemia, antiplatelet and lipid-lowering drugs, revascularization in patients with per-vessel PAV ≥22% was associated with a significantly reduced event rate (HR 0.50, 95% CI 0.27-0.91, p=0.024) (p-interaction=0.016), whereas patient-level results remained non-significant (HR 0.62, 95% CI 0.35-1.10, p=0.104) (p-interaction<0.001).

Conclusion: In this cohort study of patients referred for CCTA, revascularization on top of medical therapy was associated with a lower rate of long-term death, MI, or uAP from per-vessel PAV of 22% upwards.

Keywords: artificial intelligence; coronary computed tomography angiography; plaque burden; revascularization.