RESUMEN
OBJECTIVES: The aim of this analysis is to examine the incremental prognostic value of coronary artery calcium (CAC) score and myocardial flow reserve (MFR) in patients with suspected coronary artery disease (CAD) undergoing positron emission tomography (PET) myocardial perfusion imaging (MPI). BACKGROUND: Advances in cardiac PET and computed tomography imaging enabled the simultaneous acquisition of anatomic and physiological data for patients suspected of CAD. METHODS: Consecutive patients who underwent PET MPI and CAC score calculation at King Abdulaziz Cardiac Center, Riyadh, Saudi Arabia, between May 2011 and May 2018 were included in the study. MPI and CAC images were obtained in the same setting. The primary endpoint of the study was a composite of cardiac death and nonfatal myocardial infarction. Cox proportional hazard regression was used to assess the incremental prognostic value of CAC and MFR by sequentially adding the variables to a model that included clinical and PET variables. RESULTS: A total of 4,008 patients (mean age 59.7 ± 11.6 years, 55% women) were included in the analysis. Risk factors were prevalent (77.6% hypertension, 58.1% diabetes). In total, 35.9% of the cohort had CAC of 0, 16.5% had CAC ≥400, and 43.9% had MFR <2. Over a median follow up of 1.9 years, 130 (3.2%) patients had cardiac death/nonfatal myocardial infarction. CAC and MFR score added incremental prognostic value over clinical and perfusion variables (base model: c-index 0.8137; Akaike information criterion [AIC]: 1,865.877; p = 0.0011; CAC model: c-index = 0.8330; AIC: 1,850.810; p = 0.045 vs. base model; MFR model: c-index = 0.8279; AIC: 1,859.235; p = 0.024). Combining CAC and MFR did not enhance risk prediction (c-index = 0.8435; AIC: 1,846.334; p = 0.074 vs. MFR model; p = 0.21 vs. CAC model.) CONCLUSIONS: In this large cohort of patients referred for PET MPI, both CAC and MFR independently added incremental prognostic value over clinical and MPI variables. Although combining both may have synergetic prognostic effect, this relation was not shown in multivariable model of this analysis.