CD10(+) and CK7/RON(-) immunophenotype distinguishes renal cell carcinoma, conventional type with eosinophilic morphology from its mimickers

BACKGROUND:: The distinction between renal cell carcinoma conventional (clear cell) type with eosinophilic morphology (ccRCC), chromophobe renal cell carcinoma eosinophilic variant (chRCC), and renal oncocytoma (RO) is a common diagnostic dilemma. We aimed to identify an immunohistochemical panel to discriminate ccRCC from its morphologic mimics. MATERIALS AND METHODS:: Fifty-three renal neoplasms (19 ccRCC, 18 chRCC, and 16 RO) were selected. Immunohistochemical stains for CD10, cytokeratin 7 (CK7), c-Kit, E-cadherin, N-cadherin, kidney-specific cadherin (Ksp-cadherin), and Recepteur d'origine nantais (RON) were performed. RESULTS:: Ten (53%) of 19 ccRCC were positive for CD10, 11 (58%) for E-cadherin, 8 (42%) for N-cadherin, 5 (26%) for Ksp-cadherin, 9 (47%) for RON, 6 (32%) for CK7, and 5 (26%) for c-Kit. In chRCC/RO group, 5 of 34 (15%) were positive for CD10, 32 (94%) for E-cadherin, 2 (6%) for N-cadherin, 1 (3%) for Ksp-cadherin, 22 (65%) for RON, 14 (41%) for CK7, and 25 (25/32, 76%) for c-kit. Univariately, negative c-Kit [odds ratio (OR)=8.75, P=0.001, area under the receiver operating characteristic curve (AUC)=0.747], negative E-cadherin (OR=11.64, P=0.005, AUC=0.681), positive N-cadherin (OR=11.64, P=0.005, AUC=0.681), positive Ksp-cadherin (OR=11.79, P=0.031, AUC=0.617), and positive CD10 (OR=6.44, P=0.005, AUC=0.690) detects ccRCC versus chRCC/RO. Multivariate analysis showed significant association between CD10 positivity and ccRCC (OR=16.90, P=0.007) and between RON negativity and ccRCC (OR=7.17, P=0.047) when CK7 is negative. CONCLUSIONS:: The best single predictors for ccRCC are negative c-Kit, negative E-cadherin, positive N-cadherin, positive Ksp-cadherin, and positive CD10. However, considering the studied markers, a combination of positive CD10 and negative CK7 and RON is considered the best immunohistochemical panel in distinguishing ccRCC from chRCC/RO.