Abstract 634: SOX11 is a potential clinical marker for hormone receptor negative ductal carcinoma in situ

Abstract

Abstract Background: Ductal carcinoma in situ (DCIS) comprises 20-25% of screen-detected breast cancer and, like invasive ductal carcinoma (IDC), is heterogeneous in terms of underlying biology, presentation and outcome. There are limited potential biomarkers of outcome for DCIS, although estrogen receptor (ER)-positive (pos), progesterone receptor (PgR)-pos, and HER2-negative (neg) DCIS appears to have a better outlook than DCIS neg for ER and PgR but pos for HER2. The aim of this study was to identify markers for DCIS to develop a panel to stratify DCIS into DCIS at low or high risk of further DCIS. Methods and results: To identify genes driving DCIS evolution followed by the progression to IDC, we first used transcriptional data sets (GSE788, GSE16873) which had data from both normal mammary glands (NMG) and DCIS and performed class comparison (NMG vs DCIS). For both GSE7882 and GSE16883, the number of over- and under-expressed genes was 297 and 187. Over-expressed genes in DCIS represented a mitotic/proliferative feature annotated as mitotic spindle and condensed chromosome. Under-expressed genes represented loss of epithelial features annotated as epithelial cell differentiation and development. A total of 484 differentially expressed genes in DCIS were further correlated with recurrence events of IDC using Kessler's breast cancer data set to identify genes contributing aggressive feature across DCIS and IDC. Genes correlating to recurrence events were selected. From 484 genes, 99 were significantly associated with recurrence events of IDC (with P<0.003). Among these 99 genes, proliferation-related genes were depleted but component genes of the Oncotype DCIS score and genes reported as relevant to DCIS biology, such as SOX11, were included for Nanostring transcriptional analysis. The final number of genes-of-interest was 58 including 5 housekeeping genes. 40 DCIS and 8 NMG lesions were macro-dissected from formalin-fix paraffin-embedded blocks (FFPE) and extracted transcripts were subjected for Nanostring analysis. Gene expression data was clustered in an unsupervised manner using R software. Two sample clusters were identified: an ER/PgR-neg cluster and an ER/PgR-pos cluster. Over-expression of SOX11 and HER2 were exclusively seen in the ER/PgR-neg cluster within which samples were further subcategorized into HER2low/SOX11+ and HER2high/SOX11+. These RNA expression findings are undergoing confirmation by immunohistochemistry of FFPE sections. A further independent series of DCIS which has recurred as DCIS or IDC is currently under investigation for validation of an ER/PgR/HER2/SOX11 signature of poor prognosis DCIS. Conclusion: SOX11 is exclusively overexpressed in ER/PgR-neg DCIS and is a candidate clinical marker for recurrence of DCIS or progression of ER/PgR-neg DCIS to IDC. Citation Format: Kazuharu Kai, Wei Lu, Fei Yang, Ximing Tang, Ignacio I. Wistuba, Subrata Sen, Alastair Thompson. SOX11 is a potential clinical marker for hormone receptor negative ductal carcinoma in situ [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 634.