Cuiping Zhou1, Xiaohua Ban2, Jianxun Lv1, Lin Cheng1, Jianmin Xu3, Xinping Shen1. 1. Department of Radiology, The University of Hong Kong-Shenzhen Hospital, Shenzhen, China. 2. Department of Medical Imaging Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Sun Yat-sen University Cancer Center, Guangzhou, China. 3. Department of Radiology, Second Clinical Medical College of Jinan University, Shenzhen People's Hospital, Shenzhen, China.
Abstract
Background: Chromophobe renal cell carcinoma (chRCC) is often confused with oncocytoma and angiomyolipoma without visible fat (AML.wovf). The aim of this study was to determine computed tomography (CT) features predictive of chRCC to distinguish it from oncocytoma and AML.wovf. Methods: This multicenter study enrolled 38 patients with chRCC, 32 with oncocytoma, and 43 with AML.wovf of the kidney. The clinical and imaging features of all cases were reviewed retrospectively, and associations between the features and histopathology were analyzed using univariate analysis, followed by multinomial logistic regression analysis. Receiver operating characteristic (ROC) curve analysis was used to evaluate logistic regression models and determine optimal cut-off values for numeric data. Results: Univariate analysis revealed significant differences between chRCC and oncocytoma in tumor ratios of lesion to renal cortex net enhancement (RLRCNE) on both corticomedullary and nephrographic phase images (P<0.001 for both) and calcification (P=0.035). On multinomial logistic regression analysis, only corticomedullary RLRCNE remained an independent predictor for the differential diagnosis of chRCC from oncocytoma (P<0.001), with an optimal cut-off value of 0.53. Comparing chRCC and AML.wovf, univariate analysis revealed significant differences in age (P=0.003), segmental enhancement inversion (SEI) (P=0.006), corticomedullary RLRCNE (P<0.001), unenhanced ratio of lesion to renal cortex attenuation (RLRCA; P<0.001), size (P<0.001), enhancement pattern over time (P=0.017), angle (P=0.014), and central scar (P<0.001). Only unenhanced RLRCA (P<0.001), size (P=0.003), and enhancement pattern over time (P=0.002) remained as independent predictors on multinomial logistic regression analysis, with optimal cut-off values of 1.13 and 30.9 mm for RLRCA and size, respectively. On ROC curve analysis of the logistic regression models, the areas under curve (AUC) were 0.888 and 0.963 for chRCC versus oncocytoma and AML.wovf, respectively. Conclusions: Corticomedullary RLRCNE on CT images was an independent predictor for the differential diagnosis of chRCC from oncocytoma. Unenhanced RLRCA, size, and enhancement pattern over time on CT had predictive value for discriminating chRCC from AML.wovf. 2022 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Background: Chromophobe renal cell carcinoma (chRCC) is often confused with oncocytoma and angiomyolipoma without visible fat (AML.wovf). The aim of this study was to determine computed tomography (CT) features predictive of chRCC to distinguish it from oncocytoma and AML.wovf. Methods: This multicenter study enrolled 38 patients with chRCC, 32 with oncocytoma, and 43 with AML.wovf of the kidney. The clinical and imaging features of all cases were reviewed retrospectively, and associations between the features and histopathology were analyzed using univariate analysis, followed by multinomial logistic regression analysis. Receiver operating characteristic (ROC) curve analysis was used to evaluate logistic regression models and determine optimal cut-off values for numeric data. Results: Univariate analysis revealed significant differences between chRCC and oncocytoma in tumor ratios of lesion to renal cortex net enhancement (RLRCNE) on both corticomedullary and nephrographic phase images (P<0.001 for both) and calcification (P=0.035). On multinomial logistic regression analysis, only corticomedullary RLRCNE remained an independent predictor for the differential diagnosis of chRCC from oncocytoma (P<0.001), with an optimal cut-off value of 0.53. Comparing chRCC and AML.wovf, univariate analysis revealed significant differences in age (P=0.003), segmental enhancement inversion (SEI) (P=0.006), corticomedullary RLRCNE (P<0.001), unenhanced ratio of lesion to renal cortex attenuation (RLRCA; P<0.001), size (P<0.001), enhancement pattern over time (P=0.017), angle (P=0.014), and central scar (P<0.001). Only unenhanced RLRCA (P<0.001), size (P=0.003), and enhancement pattern over time (P=0.002) remained as independent predictors on multinomial logistic regression analysis, with optimal cut-off values of 1.13 and 30.9 mm for RLRCA and size, respectively. On ROC curve analysis of the logistic regression models, the areas under curve (AUC) were 0.888 and 0.963 for chRCC versus oncocytoma and AML.wovf, respectively. Conclusions: Corticomedullary RLRCNE on CT images was an independent predictor for the differential diagnosis of chRCC from oncocytoma. Unenhanced RLRCA, size, and enhancement pattern over time on CT had predictive value for discriminating chRCC from AML.wovf. 2022 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Entities:
Keywords:
Chromophobe renal cell carcinoma (chRCC); angiomyolipoma without visible fat (AML.wovf); computed tomography (CT); diagnosis; oncocytoma; predictive features
Authors: Brian Shuch; Ali Amin; Andrew J Armstrong; John N Eble; Vincenzo Ficarra; Antonio Lopez-Beltran; Guido Martignoni; Brian I Rini; Alexander Kutikov Journal: Eur Urol Date: 2014-05-21 Impact factor: 20.096
Authors: Aaron M Potretzke; Theodora A Potretzke; Tyler M Bauman; B Alexander Knight; Alyssa M Park; Jonathan M Mobley; Robert Sherburne Figenshau; Cary Lynn Siegel Journal: J Endourol Date: 2017-01-03 Impact factor: 2.942
Authors: Jonathan R Young; Daniel Margolis; Steven Sauk; Allan J Pantuck; James Sayre; Steven S Raman Journal: Radiology Date: 2013-02-04 Impact factor: 11.105
Authors: Siva P Raman; Pamela T Johnson; Mohamad E Allaf; George Netto; Elliot K Fishman Journal: AJR Am J Roentgenol Date: 2013-12 Impact factor: 3.959
Authors: Kousei Ishigami; Aaron R Jones; Laila Dahmoush; Leandro V Leite; Marius G Pakalniskis; Thomas J Barloon Journal: Insights Imaging Date: 2014-12-14