Expanding the versatility of cardiac PET/CT: Feasibility of delayed contrast enhancement CT for infarct detection in a porcine model

It has recently been suggested that, similar to MRI, CT can be used to detect infarcts at high resolution by delayed myocardial contrast enhancement. In cardiac PET/CT, this ability to detect infarcts may increase the versatility and integrative potential of PET and CT study components. We sought to determine the feasibility of delayed CT-enhancement in the PET/CT environment and compared it with PET-defined rest perfusion for the measurement of infarct size. Methods: Experimental myocardial infarction was induced in 10 young farm pigs by occlusion and reperfusion of the left anterior descending coronary artery. After 4-6 wk, the animals underwent 64-slice PET/CT. Rest perfusion was measured by ¹³N-ammonia PET. Then, 120 mL of contrast were injected, and retrospectively gated helical CT was performed for angiography and after 1.5-, 5-, 10-, and 15-min delays. Two days later, 6 pigs again underwent contrast-enhanced CT, using a low-radiation-dose approach (prospective gating and thicker slices as used for clinical calcium scoring) and the same delay times. Polar maps of PET perfusion and CT myocardial enhancement were created for further analysis. Results: CT Hounsfield units (HUs) in the infarct area started to exceed those of arterial blood at 5-10 min after contrast injection, and the ratios of infarcted myocardium to remote myocardium and of infarcted myocardium to blood plateaued at around 1.9 and 1.2 between 10 and 15 min. Excellent agreement between high- and low-dose CT acquisitions (R = 0.87, P < 0.001) was demonstrated. At 10 min, CT infarct size (area with HU > 3.5 SDs from remote) was 30% ± 8% of the left ventricle, using the low-dose approach. The PET perfusion defect size (area with uptake < 60% of the left ventricular maximum) was comparable at 31% ± 8% of the left ventricle (range, 17%-44%). Using a 16-segment myocardial model, we showed an excellent inverse relationship between regional ammonia retention and contrast enhancement (R = - 0.93, P < 0.001). Conclusion: In our animal model, infarct size can be measured accurately and reproducibly using cardiac PET/CT with delayed CT-enhancement. For measurement, a low-dose, prospectively gated acquisition was comparable to higher-dose spiral CT. These results provide a rationale for further clinical work to explore whether delayed CT-enhancement can improve the accuracy of myocardial viability assessment, substitute for rest studies in perfusion imaging, or improve localization of PET-derived molecular signals.