Abstract LB-253: Novel prognostic gene profiles in tumors for breast cancer survival

Abstract

Abstract Recent development in molecular phenotyping of breast cancer to guide treatment strategy has improved breast cancer survival. While several prognostic panels, such as Oncotype DX (21 genes), MammaPrint (70 genes), ProSigna (50 genes), have been approved for clinical use with demonstrated utility, there is still room for improvement. Using a novel bioinformatics approach, we previously demonstrated that higher “activity” (instead of level of expression) of miR-500a-5p in tumor was associated with shorter survival of ER+ breast cancer patients. However, the activity of miR-500a cannot be directly measured and can only be inferred by the expression of its associated genes. Using NanoString platform, we profiled 162 miR-500a-related genes as well as PAM50 genes (22 genes overlapped between PAM50 set and miR-500a-related set) from tumors collected as part of the Long Island Breast Cancer Study Project, where a population-based sample of 1508 women, predominantly Caucasians aged 25-98 (median age 58), newly diagnosed with first primary breast cancer in 1996-1997, were followed up for 15+ years using the National Death Index. We were able to obtain sufficient RNA for profiling from paraffin-embedded tumors tissues for 609 cases, among which 119 breast cancer deaths were reported at the end of follow up. Survival analyses were carried out using Cox proportional hazards models adjusting for known prognostic factors. Consistent with our a priori hypothesis, most of the selected miR-500a related genes (81 out of 162) significantly associated with breast cancer-specific mortality. Co-expression analysis revealed two major clusters of genes with opposite associations with breast cancer-specific mortality. For example, expression of SUSD3 from the first cluster showed inverse associations with breast cancer-specific mortality [high vs low expression: Hazard Ratio (HR) 0.33, 95% CI: 0.23-0.47] while TPX2 from the second cluster showed the opposite [HR 2.64, CI: 1.84-3.80]. These associations were also confirmed in independent combined datasets (described in “KM-Plotter”, kmplot.com). Most importantly, these associations remained significant after adjusting for known prognostic factors including tumor hormonal receptor status and PAM50-derived risk of recurrence category, implicating they are independent markers for breast cancer survival. Our study identified novel predictors that may improve prognostic efficiency of current molecular phenotyping. In addition, these results shed light on molecular mechanism of breast cancer progression and may point to the targets for treatment of the disease. Citation Format: Vasily N. Aushev, Marilie D. Gammon, Humberto Parada, Davide Degli Esposti, Hector Hernandez-Vargas, Zdenko Herceg, Susan Teitelbaum, Jun Zhu, Eunjee Lee, Jia Chen. Novel prognostic gene profiles in tumors for breast cancer survival [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr LB-253. doi:10.1158/1538-7445.AM2017-LB-253