Abstract 3398: Analysis of whole transcriptome RNA-seq of large numbers of clinical FFPE samples

Abstract

Background: We evaluate the feasibility, quality, and analytical potential of whole-transcriptome RNA-seq on over one thousand clinical FFPE tumor samples. Transcriptional profiling of RNA is used for clinical decision making in many tumor types. Clinical analysis of RNA is complicated by the common use of formalin-fixed paraffin-embedded (FFPE) tissue storage, which can cause low yield and RNA degradation. Methods: RNA was extracted from FFPE material using commercially available kits using an RNAseH based ribodeplete. Multiple libraries per sample are sequenced using standard Illumina sequencing. Bowtie2, RSEM, and custom software are used for alignment, transcript quantification, fusion detection, and variant expression analysis. Results: We observe a >85% success rate on whole transcriptome RNA-seq on our cohort of more than 1000 samples. We find reliable transcript quantification upon successful sequencing, and on a subset of samples quantify differences between FFPE and fresh-frozen material when using ribodeplete. We also compare transcriptional profiles of clinical FFPE samples to an independent set of fresh frozen, poly-A capture samples from The Cancer Genome Atlas (TCGA) and show differences between poly-A capture and ribodeplete RNA isolation methods as well as FFPE vs. FF effects. We present a robust mapping methodology for comparison of public FF poly-A dataset to our FFPE ribodeplete samples. We demonstrate that we are able to utilize this joint gene expression space to perform site of origin prediction on FFPE samples. This is especially important for clinical application in analyzing Cancer of Unknown Primary (CUP) samples as well as detecting outlier samples whose molecular features may suggest additional therapeutic avenues. Finally, we show reliable detection of fusion transcripts from FFPE RNA-seq material from whole-transcriptome analysis and expression of somatic tumor variants detected from DNA sequencing. Conclusions: Large scale sequencing of RNA from clinical FFPE materials provides reliable transcriptomic results comparable to existing public databases of RNA, enabling research on cohorts of tumors that are FFPE-banked and unavailable as frozen tissue. Citation Format: Yulia Newton, Justin Golovato, Mark Johnson, Shahrooz Rabizadeh, Zack Sanborn, Steve Benz, Charles J. Vaske. Analysis of whole transcriptome RNA-seq of large numbers of clinical FFPE samples [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 3398.