Background: The diagnostic accuracy of coronary computed tomographic angiography (CTA) is high with few reports noting its ability to stratify risk. The quantity and quality of prognostic evidence with myocardial perfusion single-photon emission computed tomography (SPECT) (MPS) is diverse, with little comparative evidence between methods. The aim of this report was to compare all-cause death rates for 7 CTA subsets, using the Duke prognostic index, compared with percentage of ischemic myocardium by MPS.
Methods: We performed a matched cohort comparison of patients with suspected coronary artery disease (CAD) referred for evaluation of new onset chest pain with 693 and 3067 patients undergoing CTA and MPS. The primary endpoint was time to all-cause death estimated with univariable and multivariable (controlling for pretest CAD likelihood and cardiac risk factors) Cox proportional hazards models. Patients undergoing MPS were matched, using a propensity scoring technique, to the CTA cohort, yielding 16%, 60%, and 24% of the patients with low, intermediate, and high pretest CAD likelihood (P = 0.39).
Results: Two-year mortality was similar for CTA and MPS at 3.2% (P = 0.71). For CTA, the Duke prognostic index was independently predictive of death in risk-adjusted models controlling for risk factors and pretest likelihood of CAD (P < 0.0001). Patients with <50% stenosis had the highest survival at 99.7%. Survival worsened from 96% for patients with 2 moderate stenoses or 1 >or=70% stenosis (P = 0.013) to 85% survival for patients with >or=50% left main stenosis (P < 0.0001). For MPS, the percentage of ischemic myocardium was independently predictive of death (P < 0.0001). For patients with no MPS ischemia, 100% survival was observed. Survival worsened from 94.0% to 83.0% for patients with 5% to >or=20% ischemic myocardium (P < 0.0001). In the comparative analysis of CTA to MPS, annual mortality rates were similar with the Duke CAD index compared with the percentage of ischemic myocardium (P = 0.53). Annual mortality rates ranged from 0.1% to 11.7% by the extent and severity of abnormalities noted on CTA and MPS (P = 0.53).
Conclusion: A directly proportional relation was observed between the extent and severity of MPS ischemia and angiographic CAD. High-risk ischemia is more often associated with extensive CAD and high mortality risk. The results from this matched, observational study require additional validation for longer-term predictive models that include major adverse cardiovascular events and diverse patient subsets.