Introduction: The magnitude of perfusion abnormality on myocardial perfusion SPECT as measured by summed stress score (SSS) has suboptimal sensitivity in detecting severe coronary artery disease (CAD). SSS alone tends to underestimate severe CAD (presence of 3-vessel disease or left main involvement) due to global balanced ischemia and the relative nature of perfusion defect analysis. The aim of the study was to evaluate the incremental value of left ventricular ejection fraction (LVEF) reserve in predicting severe CAD.
Methods: A cross-sectional study of 62 consecutive patients with normal or abnormal scintigraphy who underwent thallium-201 dipyridamole myocardial perfusion scan with coronary angiogram done within 6 months of the perfusion study, with no intervening cardiac event. Magnitude of perfusion abnormality was calculated using a 17-segment model and a 5-point scoring system. LVEF reserve was computed as stress LVEF minus rest LVEF, or the absolute percent difference in the ejection fraction.
Results: On receiver-operating characteristics (ROC) curve analysis, summed stress score (SSS) of greater than equal or equal to 8 was the optimal cut-point with are under the curve (AUC) of 0.7 to detect severe CAD. This threshold had sensitivity and specificity of 75% and 63%, respectively with a positive likelihood ratio of 2.05 (ĸ=0.385 ± 0.126, p=0.001). While for LVEF reserve, on ROC curve analysis, the best cut-point in predicting severe CAD was less than 5% (AUC of 0.6). The validity of LVEF reserve <5% in predicting severe disease was lower compared to SSS ≥8 with the sensitivity and specificity of 63% and 50%, respectively and positive likelihood ratio of 1.25 ((ĸ=0.125 ± 0.126, p=1.61). To enhance diagnostic utility of SSS ≥8 or LVEF reserve <5% (LVEF reserve + SSS ≥8) + [(SSS ≥8_ or (LVEF reserve <5%)] in predicting severe CAD. The sensitivity improved to 91% with its negative predictive value increased from 70% to 77% compared with SSS ≥8 alone.
Conclusion: LVEF reserve of <5% increased the sensitivity of myocardial perfusion scintigraphy in detecting severe CAD compared to the magnitude of perfusion abnormality alone.