Diagnostic accuracy of lung subtraction iodine mapping CT for the evaluation of pulmonary perfusion in patients with chronic thromboembolic pulmonary hypertension: Correlation with perfusion SPECT/CT

Background For treatment of chronic thromboembolic pulmonary hypertension (CTEPH), the evaluation of segmental pulmonary perfusion is important. There are no previous reports about lung subtraction iodine mapping (LSIM) computed tomography (CT) for evaluation of segmental pulmonary perfusion in patients with CTEPH, using lung perfusion SPECT/CT (LPS) as the reference. Methods 50 patients (age, 60.7 ± 16.7 years) with known or suspected CTEPH were enrolled in this study. Non-contrast chest CT and CT pulmonary angiography (CTPA) were performed on a 320-detector row CT system. Then, based on a non-rigid registration followed by subtraction of non-contrast images from contrast-enhanced images, color-coded LSIM images were generated. LPS was performed using a SPECT/CT system within a period of 2 months, and served as the reference standard. LSIM and CTPA images were evaluated in a blinded manner for the detection of pulmonary perfusion defects on a segment-by-segment basis. Results The sensitivity, specificity, accuracy, and positive and negative predictive values of LSIM for the detection of segmental perfusion defects were 95% (734/773), 84% (107/127), 93% (841/900), 97% (734/754) and 73% (107/146), respectively, while the corresponding values for CTPA were 65% (505/773), 61% (78/127), 65% (583/900), 91% (505/554) and 23% (78/346). Generalized estimating equations analyses revealed a significantly better performance of LSIM than that of CTPA regarding the sensitivity, accuracy, and positive and negative predictive values (all P < 0.0001). Conclusions LSIM is a feasible technique for segment-based evaluation of pulmonary perfusion in patients with CTEPH, and it provides a significantly higher diagnostic accuracy compared with CTPA.