Abstract P3-08-34: Development of distant recurrence risk model based on hormone receptor-positive breast cancer gene expression profiling in the Chinese population

Abstract

Abstract Background For early-stage hormone receptor (HR)-positive breast cancer (BC), one of the most clinical challenges is how to identify the patients with good prognosis and avoid unnecessary adjuvant chemotherapy. Several multi-gene expression marker assays were developed to predict distant recurrence. However, most of them were derived from Caucasian patient gene expression landscape, and lack of large-scale validation in Asian cohorts. Accumulating evidence has proved the differences in key molecules and pathways in BC among different populations. In this study, we used RNA-seq analysis on primary breast tumors of 300 patients to identify differentially expressed genes between patients with or without distant metastases, and established a preliminary model to predict recurrence risk. Materials and Methods We reviewed the clinical and follow-up information of 1,602 Chinese patients with axillary node-negative, HR-positive early-stage BC from 2007 to 2012. Patients older than 60 years, and who had a family history of breast or ovarian cancer were excluded. The paraffin-embedded specimens from patients with recurrences and age-matched patients without recurrences were subject to whole transcriptome sequencing. The subjects were then randomly divided into two sets, a training (70%) and a validation (30%) set. Based on the data generated from highly degraded FFPE-extracted RNA, we set up a novel bioinformatics pipeline for gene-expression measurement. A prediction model was developed via lasso regression to further identify genes relating to breast cancer recurrence. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of the model. Results Among the total 1,602 cases, 131 relapsed patients were identified, including eight with local recurrences, 92 with distant recurrences, and 31 with uncertain recurrences. Finally, RNA-seq data from 300 patients, including 97 with- (case) and 203 without- recurrences (control), were enrolled in the further analysis. Base on the differentially expressed genes, we developed a preliminary prognosis prediction model, including 117 genes. Patients in the validation set were divided into high-risk and low-risk group by the model. The area under the ROC curve (AUC) was 0.710, and the sensitivity and specificity were 68% and 75%, respectively. Consistent with previous studies, functional annotations and pathway analyses revealed that the majority of the genes (65/117, 55.6%) were components of major pathways involved in tumor metastasis, including cell proliferation (55/117, 47%), invasion (50/117, 42.7%), angiogenesis (25/117, 21.4%) and immune response (33/117, 28.2%). Top enriched specific pathways were Wnt signaling pathway and MAPK signaling pathway, including 29 and 22 genes, respectively. Conclusion Gene expression profiles has proved to be an effective tool to identify patients with high risk of recurrence of breast cancer in previous studies. This is the first model developed based on the data from the Chinese population, which may provide better prediction power for Asian patients. Larger validation and prospective studies are under way to modify and validate this model. Citation Format: Xin Wang, Li Ma, Xiangzhi Meng, Jiaqi Liu, Jingchao Chai, Jingyi Yang, Zhi Yang, Qixi Wu, Yueping Liu, Jianming Ying, Xiang Wang. Development of distant recurrence risk model based on hormone receptor-positive breast cancer gene expression profiling in the Chinese population [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-08-34.