Abstract
OBJECTIVE: We investigated the diagnostic performance of semi-quantitative hyperintensity on T2-weighted short-tau-inversion-recovery black-blood (T2W-STIR-BB) images in identifying active cardiac sarcoidosis (CS) in patients, and compared it with that of (18)F-fluoro-2-deoxyglucose positron emission tomography (FDG-PET). METHODS: This retrospective study included 40 steroid-naive patients (age 63.1±12.9 years, 20 men) diagnosed with CS who underwent both cardiac MRI and FDG-PET imaging. Active CS cases were defined as satisfying at least one of the following criteria for conventional indices: exacerbation of ventricular arrhythmia, newly identified advanced atrioventricular block, greater than 5% decrease in left ventricular ejection fraction on echocardiography, positive finding on gallium-scintigraphy or elevated levels of sarcoidosis-related serum biomarkers. T2W-STIR-BB images were semi-quantitatively analysed using a myocardium-to-spleen ratio (MSR). The diagnostic performance of T2W-STIR-BB and FDG-PET imaging for detecting active CS was investigated. RESULTS: Thirty-three patients satisfied at least one criterion and were considered as having active CS. Thirty patients (75%) tested positive with T2W-STIR-BB imaging, and 25 patients (63%) tested positive with FDG-PET. The sensitivity, specificity, accuracy, and positive and negative predictive values for identifying active CS by semi-quantitative MSR on T2W-STIR-BB images were 79%, 43%, 73%, 87% and 30%, respectively. These results were statistically comparable to those of FDG-PET (70%, 71%, 70%, 92% and 33%, respectively). CONCLUSIONS: When using conventional diagnostic indices for active CS as the gold standard, T2W-STIR-BB imaging demonstrated comparable diagnostic performance to that of FDG-PET. The semi-quantitative analysis of high signal intensity on T2W-STIR-BB images using MSR was useful for detection of active CS.