Clinicopathologic, Immunohistochemical And MicroRNA Profile Investigation On Renal Epithelial Eosinophilic Tumors

Usually, pathologists have difficulties in differentiating oncocytoma, papillary carcinoma, eosinophilic variant subtype of chromophobe cell carcinoma and clear cell carcinoma. Accurate pathological classification determines different prognostic estimate and corresponding treatment choice. Immunohistochemical detection and colloid iron stain can be used in differential diagnosis, but immunohistochemical stain often leads to false positive and false negative result. Due to the relatively low incidence of the tumors and only few cases studies reported, positive result of colloid iron stain is not only limited to chromophobe cell carcinoma, but also difficult to read dyeing results accurately.MicroRNA (miRNA) is a class of non-coding small RNA with about 18-25 nucleotides (nt). The expression of microRNA is highly conservation, time sequence and tissue specificity, of which the cell or tissue specificity is the main feature of microRNA expression. Demonstrated abnormal expression patterns of microRNAs in human disease tissues highlight their potential use as diagnostic and prognostic biomarkers, especially in the case of cancer. Its clinical tests may become promising and applicative new index in disease diagnosis and prognosis prediction.Currently, more than 10 microRNA studies based on microarrays in renal tumors have been published, including comparative studies of renal tumors and matched normal tissue, also different subtypes. From the independent studies, we can make the following summary:1). Many research groups are focused on the comparison of clear cell carcinoma with normal tissue.2). The expression of hsa-miR-210 and hsa-miR-155 is dramatically up-regulated in renal clear cell carcinoma, while expression of hsa-miR-141 and hsa-miR-200C is significantly down-regulated in renal clear cell carcinoma.3). The expression of hsa-miR-126 is higher in clear cell carcinoma than that in papillary carcinoma.4) There are limited studies on the tumor differential diagnosis, especially comparative studies of oncocytoma and chromophobe cell carcinoma.5). The published results were mainly generated from microarray technologies but lack of clinical validation.This study had collected 147 renal epithelial eosinophilic tumors, including 20 cases of clear cell carcinoma eosinophilic variant subtype,36 cases of papillary carcinoma,47 cases of oncocytoma, and 44 cases of chromophobe cell carcinoma eosinophilic variant subtype. Their clinical features, morphological characteristics, immunohistochemical expressing features were studied. To explore clinical application of six candidate microRNAs in renal tumors, quantitative RT-PCR was performed on 147 cases of renal eosinophilic tumors. Through ROC curves analysis, calculating the area under the curve (AUC), establishing boundary point value (cut-off point), verifying the established model in the validation set and evaluating the model’s diagnostic value in the identification of different pathological subtypes.Part 1:Cliniocpathological features of renal eosinophilic tumorsObjective:Summarize clinical pathology characteristics of the 147 cases of kidney eosinophilic tumor.Methods:20 cases of clear cell carcinoma eosinophilic variant subtype; 36 cases of papillary carcinoma,44 cases of chromophobe cell carcinoma eosinophilic variant subtype,47 cases of the oncocytoma were reviewed independently by three experienced pathologists who were expert at Urological Pathology.4 cases of oncocytoma,1 case of chromophobe cell carcinoma and 1 case of clear cell carcinoma eosinophilic variant subtype were fixed with 2.5% glutaraldehyde, and observed with transmission electron microscopy (JEOL12001X). The patients concerned had been followed up by phone.Results:73 cases of the four eosinophilic tumors were consistent with its original diagnosis among the three pathologists. Electron microscopy found rich mitochondria in oncocytoma and chromophobe cell carcinoma. But the mitochondria’s ridge was acino tubular and lamellar crest separating in chromophobe cell carcinoma and oncocytoma; the number of mitochondria in one case of clear cell carcinoma eosinophilic variant was small. In clinical follow-up, death cases resulting directly from tumor were all clear cell carcinoma eosinophilic variant and papillary carcinoma.Conclusion:The pathological variants of typical kidney eosinophilic tumors were described in 2004 edition WHO classification. However, there were some cases with non typical pathological morphology, which easily lead to misdiagnosis.Part 2:Investigation of immunohistochemistry in the diagnosis of renal eosinophilic tumorObjective:to study the significance of immunohistochemistry in differential diagnosis in renal eosinophilic tumor.Methods:147 cases of renal eosinophilic tumors were made into organization chip using tissue microarray technology, and these chips were detected with immunohistochemistry.147 cases were divided into 2 groups, one group include 14 cases of clear cell carcinoma eosinophilic variant,30 cases of papillary carcinoma,15 cases of chromophobe cell carcinoma eosinophilic variant,14 cases of oncocytoma; the other group include 6 cases of clear cell carcinoma eosinophilic variant,6 cases of papillary carcinoma,29 cases of chromophobe carcinoma eosinophil variant,33 cases of OA oncocytoma. The decision tree was made based on the immunohistochemical results of 73 cases; comparing the result between the decision tree’s diagnosis and former pathological diagnosis in 74 cases.Results:The expression of VIM, RCC, P504S, CD117 had significant differences in the papillary carcinoma/clear cell carcinoma eosinophilic variant and chromophobe cell carcinoma eosinophilic variant/oncocytoma; while the expression of EpCam, MOC31, BerEP4, CK7 had significant differences in oncocytoma and chromophobe carcinoma, CK7, P504S, MOC31, CK19 had significant differences in the papillary carcinoma and clear cell carcinoma eosinophilic forms. The decision tree’s diagnostic coincidence rate was 79.73% in 74 cases compared with pathological diagnosis.Conclusion:The antibodies panel above could be used in most differential diagnosis of the four renal eosinophilic tumors, but it was difficult to complete the differential diagnosis only depending on a single antibody.Part 3:Clinical evaluation of candidate microRNA biomarkers in the differential diagnosis of renal eosinophilic tumorsPurpose:To compare the expression differences of the candidate microRNAs in four subtypes of renal epithelial eosinophilic tumor in the training set, establish boundary point value (cut-off point), verify the cut-off point in the validation set and evaluate the diagnostic value of candidate microRNA in the identification of different pathological subtypes.Methods:Six candidate micro RNAs were evaluated in the 147 cases of paraffin embedded tissue samples using quantitative RT-PCR (Taqman). In training group, independent sample t-test, ROC curves and AUC were applied, and calculated the cut off point for the diagnosis of four renal eosinophilic tumors subtypes. In the validation group, the cut off point were validated the model’s diagnosis value in differentiating the four pathological subtypes.Results:Five microRNAs except hsa-miR139-5p had significant differential expression in four subtypes. Using five microRNAs to precede the differential diagnosis for four eosinophilic tumors in two steps. First step:hsa-miR-210 and hsa-miR-221 divided the four kinds of tumor into two groups. The second step: hsa-miR-200c distinguished oncocytoma from chromophobe cell carcinoma; hsa-miR-31 and hsa-miR-126 distinguished clear cell carcinoma eosinophilic subtypes from papillary carcinoma. In training group, the area under the curve (AUC) was near 0.9 for 4 microRNAs to diagnose the subtypes of renal tumor, and near 0.8 for one microRNA. The similar result could be seen in the validation group.Conclusion:Five microRNA markers could identify the four renal epithelial eosinophilic cancer subtypes in paraffin embedding tissue, which further indicated that microRNAs play an important role in the occurence and development of different pathological subtypes. MicroRNA-139-5p had no clinical value in the differential diagnosis of renal eosinophilic tumors.