Impact of incomplete revascularization of coronary artery disease on long-term cardiac outcomes. Retrospective comparison of angiographic and myocardial perfusion imaging criteria for completeness

Coronary revascularization in patients with coronary artery disease may be guided by coronary angiography (CA) or alternatively by ischemia on stress myocardial perfusion imaging (MPI). Which strategy leads to optimal cardiac outcomes is uncertain. We performed a retrospective analysis of 170 patients with MPI ischemia and percutaneous coronary intervention. The primary endpoint was all-cause mortality at a mean follow-up of 47 ± 21 months; the secondary end point was the composite of deaths, nonfatal myocardial infarction, and repeat coronary revascularization (MACE). The coronary revascularization was defined as complete (CCR) or incomplete (ICR) as judged by CA criteria and by MPI ischemia matched with CA criteria. Nighty-two patients (54%) had ICR by CA criteria (ICR-CA) and 84 (49%) had ICR by MPI criteria (ICR-MPI). Mortality and MACE were lower in patients with CCR-MPI than with ICR-MPI (P = .048, and P = .025). Survival of patients with CCR-CA and ICR-CA was not different (P = .081). Patients with both ICR-MPI and ICR-CA had the worst survival, whereas patients with CCR-MPI and CCR-CA had the best survival (P = .047). By multivariate analysis, ICR-MPI + ICR-CA was an independent predictor of death (P = .025). Patients with ICR by MPI were at higher risk than those with CCR. Patients with both ICR by MPI and CA were at the highest risk, while patients with CCR by both MPI and CA had the best long-term event-free survival.