Distinguishing Enhancing From Nonenhancing Renal Lesions With Fast Kilovoltage-Switching Dual-Energy CT

Abstract

The purpose of this article is to evaluate the accuracy of dual-energy CT in distinguishing enhancing from nonenhancing or equivocally enhancing renal lesions. We retrospectively reviewed fast kilovoltage-switching dual-energy renal mass CT performed in 39 patients. On the contrast-enhanced dual-energy CT scans, renal lesions were graded subjectively for enhancement using iodine density images and iodine overlay images. Lesion iodine density was measured to identify an optimal threshold for detection of enhancement. Lesion attenuation measurements on unenhanced and contrast-enhanced scans were performed to identify enhancing (increase of > 20 HU) lesions, which were used as the reference standard. Sensitivity, specificity, and accuracy for detection of enhancement were calculated for the different dual-energy CT techniques. Eighty-three renal lesions were evaluated. On the basis of attenuation measurements, there were 20 enhancing and 63 nonenhancing lesions. The sensitivity, specificity, and accuracy for the detection of enhancement according to the lesion appearance were 70%, 98.4%, and 91.6%, respectively, on iodine density images and were 85%, 90.5%, and 89.2%, respectively, on iodine overlay images generated from contrast-enhanced dual-energy CT scans. Of the various thresholds of measured lesion iodine density (1-3 mg/cm(3)), a threshold of 2 mg/cm(3) showed the highest accuracy for the detection of enhancement, with sensitivity, specificity, and accuracy of 90%, 93.7%, and 92.8%, respectively. Fast kilovoltage-switching dual-energy CT is highly specific in excluding enhancement and moderately to highly sensitive in detecting enhancement of renal lesions. Of the available dual-energy CT techniques, iodine density measurement using a threshold of 2 mg/cm(3) is most accurate in distinguishing enhancing from nonenhancing renal lesions.