Abstract 4622: Deciphering gene subsets associated with clinical outcomes of early stage breast carcinoma patients treated with Tamoxifen

Abstract

Abstract NSABP Protocol B-09 reported primary breast carcinomas expressing estrogen (ER) and progestin receptors (PR) exhibited increased likelihood of response to Tamoxifen. However, some patients with early stage lesions relapse within 10 years. We identified a 5-gene model from 14 candidate expression profiles that improved differentiation of these patients (Kerr & Wittliff, Horm Canc 2:261-71, 2011). Goals are to examine relationships of clinical outcomes of patients whose carcinomas express individual genes and subsets of the models and externally validate their clinical utility. Procedures: De-identified primary breast carcinomas from 342 patients treated with Tamoxifen were used to assess clinical relevance of 14 candidate genes. Laser capture microdissection (LCM) was used previously to isolate carcinoma cells and assess 22,000 genes by microarray. qPCR validated expression of 14 candidate genes. Univariable Cox regression analyses were performed and Kaplan-Meier plots were constructed to assess relationships between either progression-free (PFS) or overall survival (OS) and expression of each gene using R Studio 1.0.143. Relationships of gene expression were segregated by ER/PR (quantified by EIA or radio-ligand binding) with clinical outcomes. SurvExpress evaluated performance externally of individual gene expression and signatures. Results: Univariable Cox regression of microarray results for 14 candidate genes from LCM-procured cancer cells of 247 biopsies without regard to ER/PR status indicated that RERG, ERRB4 and ESR1 were related to PFS while these 3 genes and EGFR and CAXII expression were associated with OS (adjusted p value < 0.30). Validation of individual gene expression of 274 intact tissue sections by qPCR revealed that 4 of these genes (exception of EGFR) as well as PGR, BCL2, SLC39A6, EDG-1, and CD34 were associated with PFS. Expression levels of each of these 9 candidate genes as well as PTGDS, SDF and NQO-1 were independently associated with OS. Kaplan-Meier plots assessed ability of each gene expression level to predict clinical outcomes (PFS/OS). Multivariate Cox regression, performed with backward conditional selection using either microarray data or qPCR results, revealed clinically relevant genes subsets for PFS and OS. SurvExpress, an online tool, was used to externally validate molecular signatures. Conclusions: We externally validated the 5-gene model composed of PGR, BCL2, ERBB4, RERG and CD34 that identified a subset of early stage, ER+ /PR+ breast cancers patients treated with Tamoxifen that exhibited high recurrence rates. ER, PR and HER2 proteins with gene expression were correlated with prediction of clinical outcomes. Clinically relevant molecular signatures identified gene candidates for development of companion diagnostics to their protein products to improve breast cancer management and prediction of risk of recurrence. Citation Format: Zohair Riaz Hameed, Michael W. Daniels, D. Alan Kerr II, James L. Wittliff. Deciphering gene subsets associated with clinical outcomes of early stage breast carcinoma patients treated with Tamoxifen [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 4622.