TCGA molecular classification in endometriosis-associated ovarian carcinomas: Novel data on clear cell carcinoma.

BACKGROUND: Clear cell and endometrioid ovarian carcinomas (OCC and OEC, respectively) have a presumed origin in endometriosis and share molecular alterations with each other and with their endometrial counterparts. The Cancer Genome Atlas (TCGA)-based molecular classification stratifies endometrial carcinomas into four categories: POLE mutated (POLEmut), mismatch repair deficient (MMRd), p53 abnormal (p53abn) and no specific molecular profile (NSMP) with divergent prognoses. The subsequent studies are indicative that this TCGA classification has some value in OEC, but the knowledge related to OCC is limited. METHODS: Endometrial carcinoma molecular classification was evaluated and compared in a large series of OCCs (n = 115) and OECs (n = 158). POLE mutation analysis and tissue microarray-based immunohistochemistry for mismatch repair and p53 proteins were performed. RESULTS: The distribution to the molecular groups was as follows: POLEmut 0.9%/3.2%, MMRd 3.5%/6.3%, p53abn 20%/30%, and NSMP 76%/60% in OCCs/OECs, respectively. The proportion of NSMP tumors was the largest in both histological types and significantly higher in OCC than OEC (p = 0.009). The molecular classification correlated significantly with DSS in both OCCs and OECs (p < 0.001 and p = 0.009, respectively), and with DFS in OCCs (p = 0.001). POLEmut and MMRd OCCs carried excellent prognosis, whereas MMRd OECs presented with poorer outcome. The p53abn group was associated with the poorest prognosis in both tumor types, particularly emphasized in OCCs. CONCLUSIONS: TCGA molecular classification associated with patient outcome in both histotypes, and the difference in prognosis between the molecular groups was even more marked in OCC. The large amount of NSMP tumors highlights the need for further studies.

Clinical factors and biomarker profiles associated with patient outcome in endometrioid ovarian carcinoma - Emphasis on tumor grade.

OBJECTIVE: The role of clinicopathological factors and molecular markers in prognostic classification of endometrioid ovarian carcinoma (EnOC) is not established. Tumor grade is used in risk assessment, but the role of current 3-tier grading system has been challenged. METHODS: Clinicopathological factors and 12 immunohistochemical biomarkers (PR, ER, ß-catenin, vimentin, ARID1A, HNF1-ß, p53, p16, MIB-1, E-cadherin, c-erb-B2 and L1CAM) were analyzed as regards patient outcome in 215 contemporarily classified EnOCs. RESULTS: Of clinical parameters, grade and stage appeared as strong independent prognostic factors both for disease-free and disease-specific overall survival. Grades 1-3 distinguished clearly from each other in the survival analysis, whereas stages I-II and stages III-IV clustered with each other. PR, ER, nuclear ß-catenin and vimentin positivity were associated with favorable overall outcome and clinical parameters, whereas abnormal expression of p53, overexpression of p16 and L1CAM positivity were associated with aggressive disease characteristics and poor survival. The frequency of good-prognosis markers PR and ß-catenin gradually decreased and poor-prognosis markers p53, p16 and L1CAM gradually increased from grade 1-3. However, vimentin and ER were expressed at similar frequencies across different grades and presented with independent prognostic significance. CONCLUSIONS: We found histological grade and disease stage, but not residual tumor, to be independent clinical prognostic factors in EnOC. A set of good-prognosis markers (PR, ER, ß-catenin and vimentin) and poor-prognosis markers (p53, p16 and L1CAM) were identified. Our findings support continuation of the use of the 3-tier grading system for EnOC and provide clinically feasible IHC biomarkers for prognostic profiling.

L1CAM expression associates with poor outcome in endometrioid, but not in clear cell ovarian carcinoma.

OBJECTIVE: Our aim was to study the expression of L1CAM in endometrioid and clear cell ovarian carcinomas and to evaluate its correlation with clinical parameters and patient prognosis. METHODS: Tissue microarray -based immunohistochemical analysis of L1CAM expression was performed in 249 endometrioid and 140 clear cell ovarian carcinomas. Concurrent endometrial carcinoma was found in 57 of these patients. RESULTS: L1CAM expression was found in 15% of endometrioid and 23% of clear cell ovarian carcinomas. L1CAM expression was strongly associated with poor disease-specific overall survival and poor disease-free survival in endometrioid (p<0.0001, p=0.0005), but not in clear cell ovarian carcinomas. Significant association of L1CAM expression with poor overall survival was observed in grade 1-2 carcinomas (p<0.0001), but not in grade 3 tumors. In endometrioid ovarian carcinomas, L1CAM expression was associated with aggressive tumor characteristics, such as higher grade and stage, and incomplete response to primary therapy. However, L1CAM expression was not an independent prognostic factor for overall or disease-free survival. Of the 57 patients with concurrent endometrial carcinoma L1CAM positivity was found in 4 cases both in the ovarian and endometrial tumors, and in 3 cases only in the endometrial tumor. All these seven patients with L1CAM positive tumors had poor outcome. CONCLUSIONS: L1CAM expression could serve as a biomarker for predicting clinical outcome and response to therapy in patients with endometrioid ovarian carcinoma, but not in clear cell carcinomas. L1CAM positivity also predicts poor outcome in patients with concurrent endometrioid ovarian and endometrial carcinomas.