Objectives: Determine the diagnostic performance of thallium-201 myocardial scintigraphy using dipyridamole injection for the detection of coronary heart disease (myocardial ischemia and/or necrosis). Determine for each coronary artery the degree of angiographic stenosis for optimal diagnostic performance.
Patients and methods: The study included 309 patients who underwent coronarography within 6 months of the scintigraphy examination. None of the patients experienced a coronary event during this interval. Diagnostic performance of the scintigraphic exploration was compared with angiographic findings (stenosis 70%) used as the gold standard. The degree of angiographic stenosis for optimal scintigraphic performance was determined from the receiver operating characteristic (ROC) curves.
Results: The sensitivity of scintigraphy to detect angiographically demonstrated coronary disease was 84% with a specificity of 72%. The positive and negative predictive values were 87% and 66% respectively. Test accuracy was 80%. Sensitivity was better for detecting lesions of the anterior interventricular coronary than for the right coronary or circumflex. In addition, sensitivity varied with the number of vessels involved: 76% for single-vessel disease versus 90% for two- or three-vessel disease. The data analysis also suggested that an angiographic stenosis threshold of 50% provided optimal predictive value for scintigraphy for each of the three vessel territories. An analysis based on maximal stenosis in each patient, notwithstanding the congruency between lesion localization and diseased vessel territories, was found to provide less diagnostic precision.
Conclusion: Thallium-201 dipyridamole myocardial scintigraphy offers diagnostic performance comparable to that established with thallium-201 scintigraphy performed after exercise alone. Scintigraphic detection of a perfusion defect generally corresponds to an angiographic stenosis of 50%.