RNA‐Seq and Recurrence Risk Testing for Breast Cancer: Implications for Patient Education

Abstract

Rapid advances in RNA­seq technology and bioinformatic analysis of gene expression has normalized the use of precision diagnostics and recurrence risk testing for breast cancer patients. While these diagnostic approaches have numerous benefits, patients often lack the scientific literacy to evaluate data models ‐ and their uncertainties ‐ for enhancing their health decision‐making. To increase access to scientific knowledge for stakeholders, we designed an educational module that evaluates the validity of RNA‐seq gene expression as a tool for subtyping and determining the prognostic signature of a human breast tumor. Using a crowdsourced approach in the context of an undergraduate course, we recreated the workflow of prominent precision oncology companies ‐ we prepared a whole transcriptome library from a single breast cancer patient, and used multiplex sequencing on twelve uniquely barcoded Illumina preparations. Datasets obtained from the sequencing were subjected to quality control analysis using the Green Line of DNA Subway (FAST QC, FAST X, Kallisto, Sleuth algorithms). Log­ transformed gene expression ratios were used to determine the molecular subtype of the tumor. To calculate the recurrence risk, the Oncotype DX model was used, including 16 cancer genes and 5 reference genes. Results from the proof‐of‐concept experiment indicated significant overexpression of ER and PR genes (p<0.01) and concomitant underexpression of HER2 and Ki67 genes (p<0.01), consistent with a Luminal A tumor subtype. The Oncotype DX score was >25, suggesting high risk for 10 year distant recurrence (95% CI). Further exploration of the model indicated that normalization to aberrantly expressed reference genes can greatly skew the recurrence risk calculation, highlighting limits to the predictive power of this tool. Taken together, our findings illustrate the validity of RNA‐seq‐based tumor subtyping through an educational module, and underscores the importance of equitable access to genomics education as a form of patient self‐empowerment and decision‐making. We discuss the future implications of this work for education and policy.