Original Article
Comparative study of CT appearances in renal oncocytoma and chromophobe renal cell carcinoma
Acta Radiologica 2016, Vol. 57(4) 500–506 ! The Foundation Acta Radiologica 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185115585035 acr.sagepub.com
Jingtao Wu*, Qingqiang Zhu*, Wenrong Zhu, Wenxin Chen and Shouan Wang
Abstract Background: Renal oncocytoma (RO) and chromophobe renal cell carcinoma (ChRCC) share histologic and some imaging features. Purpose: To investigate the multidetector computed tomography (MDCT) characteristics of these two tumor types. Material and Methods: Fifty-six patients with RO and 54 patients with ChRCC were studied retrospectively. MDCT was undertaken to investigate differences in tumor characteristics. Results: Calcifications were visible in 24 (42.8%) patients with RO and in 11 (20.4%) patients with ChRCC (P ¼ 0.011). 26 patients with RO had stellate scars as did 14 patients with ChRCC (P ¼ 0.025). Spoken-wheel-like enhancement was visible in 41 patients with RO and in 11 with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion (P < 0.001). Two patients with RO had retroperitoneal lymph node enlargement as did 13 patients with ChRCC (P ¼ 0.002). Combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features were found to have a sensitivity of 99.1% (106 of 107), a specificity of 100% (3 of 3), a positive predictive value of 100% (106 of 106), and a negative predictive value of 75% (3 of 4). The attenuation of RO tumors was greater than that of ChRCC tumors, normal renal parenchyma on unenhanced CT (P ¼ 0.031). Enhancement was higher with RO than with ChRCC tumors in all phases (P ¼ 0.021, < 0.001, and 0.007, respectively). Conclusion: CT imaging features such as stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion were more common in RO and they may help in differentiating RO from ChRCC.
Keywords Renal oncocytoma, chromophobe renal cell carcinoma, computed tomography (CT) Date received: 28 September 2013; accepted: 8 April 2015
Introduction Chromophobe renal cell carcinoma (ChRCC) is postulated to arise from the intercalated cells of the renal cortex. It is interesting that ChRCC and renal oncocytoma (RO) have much in common regarding morphologic, histologic, immunohistochemical, and ultrastructural aspects (1,2). ROs develop from type B intercalated cells of the cortical collecting duct and are indistinguishable from ChRCCs on imaging studies, sharing features such as central scar, and spoke wheel pattern of enhancement (3). Despite these common features, RO and ChRCC exhibit differences in prognosis (4). RO is considered benign in that it does not usually show renal vein invasion, postoperative recurrence, or
distant metastases. ChRCC is an RCC subtype with overall favorable prognosis compared with other RCCs, including slower growth and significantly lower rates of metastases, progression, and death (5,6). ChRCC is generally managed via partial Department of Medical Imaging, Subei People’s Hospital, Medical School of Yangzhou University, Yangzhou, PR China *Equal contributors. Corresponding author: Jingtao Wu, Department of Medical Imaging, Subei People’s Hospital, Medical School of Yangzhou University, Yangzhou, China. No 98 West Nantong Road, Yangzhou 225001, PR China. Email: [email protected]
Wu et al. nephrectomy when technically possible, and routine radical nephrectomy for this entity would be gross mismanagement, with needlessly exposing the patient to greater risk of reduced renal function and associated higher rate of cardiovascular events from the more extensive surgery (7). The purpose of the present study was to retrospectively characterize multidetector computed tomography (MDCT) characteristics which lead to the histological diagnosis of RO and ChRCC.
501
This study was approved by the Institutional Research Ethics Committee. A search of pathology records and PACS system identified 56 patients with RO and 54 patients with ChRCC who were hospitalized at our hospital between 2001 and 2013. Details of the patient’s gender, surgery or biopsy confirmation, and clinical symptoms were recorded.
communication system (PACS) workstation. The parameters evaluated included tumor calcification and attenuation on unenhanced CT scans. The degree of enhancement on the different phases of the enhanced CT scans were thereafter assessed. The attenuation of the tumor, normal renal cortex, and medulla were measured during all enhanced phases. The measured tumor area (the region of interest [ROI]) was at the center of the mass to avoid partial volume effects, however, intratumoral calcification and cystic components were excluded from the ROI if central necrosis or calcifications were considered. Normal renal cortex and medulla were measured using uninvolved homolateral renal cortex and medulla. Each 10 mm ROI was measured three times for each phase and the mean value was used. The enhancement pattern of the tumor was classified as homogeneous or heterogeneous. The degree of enhancement was based on dynamic CT imaging using Hounsfield unit (HU) attenuation of the tumor, renal cortex, and medulla.
CT technique
Statistical analysis
CT examinations were performed by using a 16 detector row scanner (CT Lightspeed Ultra, GE Healthcare, Milwaukee, WI, USA; [n ¼ 27]), or a 64 detector row scanner CT (Somatom Definition, Siemens Medical Solutions, Forchheim, Germany; [n ¼ 83]). Intravenous contrast material (iopromide, Ultravist 370; Bayer Schering Pharma AG, Berlin, Germany) was administered at a dose of 2 mL per kg of body weight and at a rate of 3 mL/s to a maximum of 150 mL by using a power injector, or at a slower rate (minimum of 2 mL/s) if required when venous access was suboptimal. All patients underwent unenhanced, arterial phase (cortical phase), cortico-medullary phase (medullary phase), and excretory phase (delayed phase). Unenhanced imaging was performed before administering the intravenous contrast agent. The contrast-enhanced CT scans began after 20 s for the arterial phase (cortical phase), 65 s for the cortico-medullary phase (medullary phase), and 300 s for the excretory phase (delayed phase). The scanning parameters varied for the different scanners but were for all phases in the following ranges: tube voltage, 120 kVp (constant for all); tube current, 150–250 mA (depending on patient size); pitch, 0.891–1.6; section thickness, 1.25–5.0 mm; reconstruction interval, 2.5– 5 mm, and rotation time, 0.5–0.75 s.
Statistical analysis was undertaken using SPSS version 17.0 statistical software (SPSS Inc, Chicago, IL, USA). Numeric data were expressed as means and standard deviations (SD) and categorical data were expressed as percentages. Evaluated characteristics were compared between the two tumor types using chi-square test, analysis of variance (ANOVA); post-hoc test (Tukey) was performed after ANOVA. Values of P < 0.05 were considered statistically significant. To assess the diagnostic performance of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion in terms of differentiating ROs from ChRCCs, the sensitivity, specificity, and positive and negative predictive value of these signs were calculated.
Material and Methods Patients
Imaging analysis Two radiologists with more than 10 years of experience each, blinded to the final diagnosis, reviewed the CT images in consensus at a picture archiving and
Results The study included 56 patients (26 women, 30 men) with RO and 54 patients (28 men, 26 women) with ChRCC. The presenting symptoms of RO and ChRCC included flank pain (25 and 28 patients), hematuria (34 and 26 patients), palpable mass (12 and 13 patients), and fever (22 and 34 patients) in each group, respectively. There was evidence of calcifications in 24 (42.8%) of the RO tumors whereas 11 (20.4%) patients with ChRCC had evidence of calcifications (P ¼ 0.011, Table 1). In 26 (46.4%) patients with RO (Figs. 1 and 2) and in 14 (25.9%) patients with ChRCC the tumors exhibited stellate scars (P ¼ 0.025). There was evidence
502
Acta Radiologica 57(4)
of spoken-wheel-like enhancement in 41 (73.2%) of the RO tumors compared to 11 (20.4%) patients with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion during the cortico-medullary phase and excretory phase, which significantly differentiated ROs from ChRCCs (P < 0.001, Table 1). For differentiating RO from ChRCC, combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features was Table 1. Comparative study of CT appearances in RO and ChRCC. Main CT findings Calcification Stellate scar Homogeneous enhancement Spoken-wheel-like enhancement Segmental enhancement inversion Lymph node enlargement/metastasis
RO (n ¼ 56)
ChRCC (n ¼ 54)
24 (42.8%) 26 (46.4%) 36 (64.3%)
11 (20.4%) 14 (25.9%) 21 (38.9%)
0.011 0.025 0.008
41 (73.2%)
11 (20.4%)
Comparative study of CT appearances in renal oncocytoma and chromophobe renal cell carcinoma
Acta Radiologica 2016, Vol. 57(4) 500–506 ! The Foundation Acta Radiologica 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185115585035 acr.sagepub.com
Jingtao Wu*, Qingqiang Zhu*, Wenrong Zhu, Wenxin Chen and Shouan Wang
Abstract Background: Renal oncocytoma (RO) and chromophobe renal cell carcinoma (ChRCC) share histologic and some imaging features. Purpose: To investigate the multidetector computed tomography (MDCT) characteristics of these two tumor types. Material and Methods: Fifty-six patients with RO and 54 patients with ChRCC were studied retrospectively. MDCT was undertaken to investigate differences in tumor characteristics. Results: Calcifications were visible in 24 (42.8%) patients with RO and in 11 (20.4%) patients with ChRCC (P ¼ 0.011). 26 patients with RO had stellate scars as did 14 patients with ChRCC (P ¼ 0.025). Spoken-wheel-like enhancement was visible in 41 patients with RO and in 11 with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion (P < 0.001). Two patients with RO had retroperitoneal lymph node enlargement as did 13 patients with ChRCC (P ¼ 0.002). Combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features were found to have a sensitivity of 99.1% (106 of 107), a specificity of 100% (3 of 3), a positive predictive value of 100% (106 of 106), and a negative predictive value of 75% (3 of 4). The attenuation of RO tumors was greater than that of ChRCC tumors, normal renal parenchyma on unenhanced CT (P ¼ 0.031). Enhancement was higher with RO than with ChRCC tumors in all phases (P ¼ 0.021, < 0.001, and 0.007, respectively). Conclusion: CT imaging features such as stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion were more common in RO and they may help in differentiating RO from ChRCC.
Keywords Renal oncocytoma, chromophobe renal cell carcinoma, computed tomography (CT) Date received: 28 September 2013; accepted: 8 April 2015
Introduction Chromophobe renal cell carcinoma (ChRCC) is postulated to arise from the intercalated cells of the renal cortex. It is interesting that ChRCC and renal oncocytoma (RO) have much in common regarding morphologic, histologic, immunohistochemical, and ultrastructural aspects (1,2). ROs develop from type B intercalated cells of the cortical collecting duct and are indistinguishable from ChRCCs on imaging studies, sharing features such as central scar, and spoke wheel pattern of enhancement (3). Despite these common features, RO and ChRCC exhibit differences in prognosis (4). RO is considered benign in that it does not usually show renal vein invasion, postoperative recurrence, or
distant metastases. ChRCC is an RCC subtype with overall favorable prognosis compared with other RCCs, including slower growth and significantly lower rates of metastases, progression, and death (5,6). ChRCC is generally managed via partial Department of Medical Imaging, Subei People’s Hospital, Medical School of Yangzhou University, Yangzhou, PR China *Equal contributors. Corresponding author: Jingtao Wu, Department of Medical Imaging, Subei People’s Hospital, Medical School of Yangzhou University, Yangzhou, China. No 98 West Nantong Road, Yangzhou 225001, PR China. Email: [email protected]
Wu et al. nephrectomy when technically possible, and routine radical nephrectomy for this entity would be gross mismanagement, with needlessly exposing the patient to greater risk of reduced renal function and associated higher rate of cardiovascular events from the more extensive surgery (7). The purpose of the present study was to retrospectively characterize multidetector computed tomography (MDCT) characteristics which lead to the histological diagnosis of RO and ChRCC.
501
This study was approved by the Institutional Research Ethics Committee. A search of pathology records and PACS system identified 56 patients with RO and 54 patients with ChRCC who were hospitalized at our hospital between 2001 and 2013. Details of the patient’s gender, surgery or biopsy confirmation, and clinical symptoms were recorded.
communication system (PACS) workstation. The parameters evaluated included tumor calcification and attenuation on unenhanced CT scans. The degree of enhancement on the different phases of the enhanced CT scans were thereafter assessed. The attenuation of the tumor, normal renal cortex, and medulla were measured during all enhanced phases. The measured tumor area (the region of interest [ROI]) was at the center of the mass to avoid partial volume effects, however, intratumoral calcification and cystic components were excluded from the ROI if central necrosis or calcifications were considered. Normal renal cortex and medulla were measured using uninvolved homolateral renal cortex and medulla. Each 10 mm ROI was measured three times for each phase and the mean value was used. The enhancement pattern of the tumor was classified as homogeneous or heterogeneous. The degree of enhancement was based on dynamic CT imaging using Hounsfield unit (HU) attenuation of the tumor, renal cortex, and medulla.
CT technique
Statistical analysis
CT examinations were performed by using a 16 detector row scanner (CT Lightspeed Ultra, GE Healthcare, Milwaukee, WI, USA; [n ¼ 27]), or a 64 detector row scanner CT (Somatom Definition, Siemens Medical Solutions, Forchheim, Germany; [n ¼ 83]). Intravenous contrast material (iopromide, Ultravist 370; Bayer Schering Pharma AG, Berlin, Germany) was administered at a dose of 2 mL per kg of body weight and at a rate of 3 mL/s to a maximum of 150 mL by using a power injector, or at a slower rate (minimum of 2 mL/s) if required when venous access was suboptimal. All patients underwent unenhanced, arterial phase (cortical phase), cortico-medullary phase (medullary phase), and excretory phase (delayed phase). Unenhanced imaging was performed before administering the intravenous contrast agent. The contrast-enhanced CT scans began after 20 s for the arterial phase (cortical phase), 65 s for the cortico-medullary phase (medullary phase), and 300 s for the excretory phase (delayed phase). The scanning parameters varied for the different scanners but were for all phases in the following ranges: tube voltage, 120 kVp (constant for all); tube current, 150–250 mA (depending on patient size); pitch, 0.891–1.6; section thickness, 1.25–5.0 mm; reconstruction interval, 2.5– 5 mm, and rotation time, 0.5–0.75 s.
Statistical analysis was undertaken using SPSS version 17.0 statistical software (SPSS Inc, Chicago, IL, USA). Numeric data were expressed as means and standard deviations (SD) and categorical data were expressed as percentages. Evaluated characteristics were compared between the two tumor types using chi-square test, analysis of variance (ANOVA); post-hoc test (Tukey) was performed after ANOVA. Values of P < 0.05 were considered statistically significant. To assess the diagnostic performance of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion in terms of differentiating ROs from ChRCCs, the sensitivity, specificity, and positive and negative predictive value of these signs were calculated.
Material and Methods Patients
Imaging analysis Two radiologists with more than 10 years of experience each, blinded to the final diagnosis, reviewed the CT images in consensus at a picture archiving and
Results The study included 56 patients (26 women, 30 men) with RO and 54 patients (28 men, 26 women) with ChRCC. The presenting symptoms of RO and ChRCC included flank pain (25 and 28 patients), hematuria (34 and 26 patients), palpable mass (12 and 13 patients), and fever (22 and 34 patients) in each group, respectively. There was evidence of calcifications in 24 (42.8%) of the RO tumors whereas 11 (20.4%) patients with ChRCC had evidence of calcifications (P ¼ 0.011, Table 1). In 26 (46.4%) patients with RO (Figs. 1 and 2) and in 14 (25.9%) patients with ChRCC the tumors exhibited stellate scars (P ¼ 0.025). There was evidence
502
Acta Radiologica 57(4)
of spoken-wheel-like enhancement in 41 (73.2%) of the RO tumors compared to 11 (20.4%) patients with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion during the cortico-medullary phase and excretory phase, which significantly differentiated ROs from ChRCCs (P < 0.001, Table 1). For differentiating RO from ChRCC, combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features was Table 1. Comparative study of CT appearances in RO and ChRCC. Main CT findings Calcification Stellate scar Homogeneous enhancement Spoken-wheel-like enhancement Segmental enhancement inversion Lymph node enlargement/metastasis
RO (n ¼ 56)
ChRCC (n ¼ 54)
24 (42.8%) 26 (46.4%) 36 (64.3%)
11 (20.4%) 14 (25.9%) 21 (38.9%)
0.011 0.025 0.008
41 (73.2%)
11 (20.4%)
