Quantifiable renal lesion diagnostic MDCT enhancement criteria for pathology prediction and clinical correlation.

Abstract

e16502 Background: Contrast-enhanced MDCT phases are proven to help differentiate renal solid tumors. However, obtaining four sequences for all patients in the golden time frame has always been challenging for technicians. Therefore, confirmatory parameters per each enhancement phase have the potential to predict pathology in daily practice reports. Our comprehensive kidney database provided a ground for multidisciplinary informed decision-making. The purpose is to evaluate confirmatory imaging biomarkers aligned with RCC lesions’ cytology features. Methods: Patients with solid renal lesions were consecutively extracted from our institutional pathology database who obtained pre-procedural MDCT up to four enhancement phases (unenhanced, corticomedullary, nephrographic, and excretory). The mean and maximum signal intensity of 323 clear cell RCCs, 133 papillary RCCs, 131 oncocytomas, 26 angiomyolipoma (fat poor and rich combined), and 58 chromophobe RCCs were evaluated for mean and maximum lesion signal intensity. The Brown–Forsythe test statistic and multiple comparisons with Bonferroni correction were used in the analysis. Results: In this ongoing study we evaluated RCC mean signal intensity in lesion, aorta and cortex per each enhancement phase and compared absolute and relative ratios to detect differentiating patterns. According to our ongoing evaluation, the corticomedullary enhancement phase is promising to differentiate clear cells (lesion mean attenuation/healthy kidney cortex) by a ratio of higher than > 0.9 (p < 0.001). Also, papillary cell type showed a considerably low ratio (lesion mean attenuation/aorta) of less than <0.4 (p <0.001). For other subtypes, qualitative criteria had better accuracy (89%) and specificity (74%) among three radiology readers. Conclusions: Our study shows the first enhancement phase (40-90 seconds) at multiphasic multidetector CT, if prospectively labeled, may assist in the discrimination of almost 85% of malignant renal lesions due to neovascularization and RCC subtypes cytology characters. These quantitative markers may also be used with computer-assisted designs and would benefit both innovative research categorization and learning purposes.