Abstract PD6-02: The metastatic breast cancer project - Expanding the clinical, genomic, and transcriptomic landscape of metastatic breast cancer through patient-partnered research

Abstract

Abstract Background: The Metastatic Breast Cancer Project (MBCproject) is an ongoing research study that directly engages patients (pts) through social media and advocacy groups, and empowers them to share their biospecimens, clinical history, and experiences. The goal is to create a publicly available dataset of linked clinicogenomic and pt-reported data to enable research. From 2015-10-20 to 2020-3-31, 3,245 MBC pts who received treatment at >1,700 institutions, consented to share medical records, pt-reported data, and tumor/saliva/blood samples and to have genomic analysis performed. Here, we describe an analysis of the clinical and genomic features in this initial MBCproject cohort. Methods: We performed whole exome sequencing (WES) on 379 tumors (with matched germline) from 301 pts, 377 germline samples and RNA sequencing (RNA-seq) from 200 tumors from 141 pts. In 14 pts, we characterized 2 or more serial tumor biopsies. WES data was analyzed for mutations and copy number variants. RNA-seq data was used to call fusions, research-grade PAM50, and gene set enrichment scores. Medical records and pt-reported data were abstracted to create detailed clinical record for each pt. Results: WES of 249 metastatic tumors identified 34 cancer genes (e.g., TP53, PIK3CA, CDH1, PTEN, AKT1, NF1, ESR1) that were significantly recurrently altered. Potential clinically actionable alterations were identified in 39% of metastatic tumors. 45 tumors (22.5%) had fusions with known functional effects and 24 (12%) had in-frame fusions in key cancer genes like FANCD2 (3), FGFR3 (2), ESR1 (1), BRAF (1), and NCOR1 (1). PAM50 classification suggested depletion of Luminal A subtype in this cohort compared to TCGA breast cancer cohort (p-value <0.05). Of 29 pts with paired RNA-seq biopsies, 10 pts showed a PAM50 subtype switch [LumB to Basal = 1, Her2 to Basal =1, Basal to LumA = 1, Her2 to LumA = 2, LumA/B to Her2 = 3, LumA to LumB = 2]. Germline analysis showed that 30.2% (114/377) of pts had at least one pathogenic variant in a cancer predisposition gene, including BRCA1/BRCA2 (5%), NF1 (5.8%), ATM (1.06%) and PALB2 (1.59%). In 10.9% pts (33/301), pathogenic variants in a cancer predisposition gene was accompanied by an apparent somatic loss of function event in the same gene. There was ~ 90% concordance between abstracted features (e.g., receptor status, histology, diagnosis dates, metastatic sites) and pt-reported data, enabling the use of both data types for integrated analyses. The utility of integrated analyses is illustrated by a case study of a pt who had an initial diagnosis of ductal carcinoma in situ and developed metastatic disease 5.5 yrs later. Analysis of 3 metastatic tumors (WES and RNA-seq) and a circulating tumor DNA sample (WES only) from this pt collected over the course of therapy revealed evidence of tumor evolution and multiple mechanisms of resistance to endocrine therapies and CDK4/6 inhibitors, including 2 distinct acquired ESR1 mutations, an activating BRAF fusion, and estrogen receptor (ER) loss by immunohistochemistry with concomitant development of ESR1-PLEKHG1 in-frame fusion. RNA-seq data also showed a PAM50 luminal phenotype and a persistent ER. signature throughout despite the apparent ER loss, suggesting compensation by the ESR1 fusion. Analyses of 13 additional pts with serial biopsies will be presented. Integrated analyses of several additional cohorts of interest, such as de novo MBC (91 pts), resistance to CDK4/6 inhibitors (39 pts), and pts diagnosed with breast cancer <40 yrs, will also be presented. Conclusion: This clinicogenomic dataset generated by partnering directly with pts was uniquely built with pt-reported data, medical record data, and multi-omic characterization, and serves as a powerful tool for researchers to harness to accelerate discoveries in MBC. Citation Format: Esha Jain, Dewey Kim, Jorge Gomez Tejeda Zanudo, Sara Balch, Mary McGillicuddy, Brett N. Tomson, Tania G. Hernandez, Beena S. Thomas, Daniel L. Abravanel, Shahrayz Shah, Rafael Ramos, Delia Sosa, Ilan Small, Lauren Sterlin, Sarah Winnicki, Colleen Nguyen, Micheal Dunphy, Elana Anastasio, Todd R. Golub, Corrie A. Painter, Nikhil Wagle. The metastatic breast cancer project - Expanding the clinical, genomic, and transcriptomic landscape of metastatic breast cancer through patient-partnered research [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD6-02.