Comparison of Contrast-Enhanced Multiphase Renal Protocol CT Versus MRI for Diagnosis of Papillary Renal Cell Carcinoma

OBJECTIVE. The objective of this study was to compare contrast-enhanced (CE) CT with MRI for the diagnosis of papillary renal cell carcinoma (pRCC). MATERIALS AND METHODS. Between 2006 and 2013, a total of 27 pRCCs were assessed using CECT or CE-MRI. A blinded radiologist placed ROIs that measured attenuation on unenhanced CT; corticomedullary and nephrographic phase CECT images, with an attenuation difference of 20 HU or more denoting enhancing lesions, 10-19 HU indicating indeterminate findings, and less than 10 HU denoting nonenhancing lesions. MRI enhancement ratios were calculated as follows: (signal intensity on gadolinium-enhanced image minus signal intensity) / (signal intensity on unenhanced image x 100) for phase 1 (acquired at 30 s), phase 2 (acquired at 70 s), and phase 3 (acquired at 180 s), where a difference of 15% or more denoted enhancement. Two additional blinded radiologists qualitatively assessed tumor margin, homogeneity, and calcification with the use of CT, and they also assessed enhancement with the use of subtraction MRI. A fourth radiologist established consensus. Twenty consecutive hemorrhagic/proteinaceous cysts served as a control group. Statistical analyses were performed using a chi-square test and multivariate regression. RESULTS. There was no statistically significant difference in patient age (p = 0.22), patient sex (p = 0.36), or tumor size (p = 0.29), when pRCCs were compared with hemorrhagic/proteinaceous cysts. On unenhanced CT, attenuation of pRCCs (mean +/- SD, 35.7 +/- 12.9 HU; range, 19-66 HU) was similar to that of hemorrhagic/proteinaceous cysts (mean, 38.9 +/- 16.9; range, 8-71 HU) (p = 0.48). A total of 51.9% of pRCCs (14/27) had either absent or indeterminate enhancement on corticomedullary phase CECT images (mean attenuation difference, 23.2 +/- 20.3 HU; range, 6-105 HU), and 14.8% of pRCCs (4/27) had indeterminate enhancement on nephrographic phase CECT images (mean attenuation difference, 36.4 +/- 24.9; range, 10-128 HU). No pRCC was nonenhancing on nephrographic phase CECT. Qualitatively, pRCCs were more heterogeneous (80% vs 45%; p = 0.02; kappa = 0.24), irregular (50% vs 5%; p < 0.001; kappa = 0.21), and calcified (25% vs 0%; p = 0.004; kappa = 0.67), with overlap existing between hemorrhagic/proteinaceous cysts. On CE-MRI, all pRCCs were quantitatively enhanced by phase 2 (95.4 +/- 83.1; percentage change in signal intensity ratio, 16-450%) and qualitatively enhanced after consensus review. No hemorrhagic/proteinaceous cyst enhanced on MRI when quantitative or subjective analysis was performed. CONCLUSION. A small number of pRCCs have indeterminate enhancement when renal protocol CT is used. Heterogeneity, irregular margins, and calcification are suggestive diagnostic features; however, quantitative and qualitative CE-MRI can accurately differentiate hemorrhagic/proteinaceous cysts from pRCC.