Ocean provided the platform and expert services required to define the approach, review the analysis results, and revise the endpoint.

Ocean’s platform produced complete feature extraction, classifier generation, QC, and comprehensive reporting on 55 samples (FASTQ) and patient metadata in 19 hours, providing a 24-hour turn-around time with partner.

Novel target discovery in pembrolizumab-resistant gastric cancer using a comprehensive RNA-seq analysis pipeline.

https://meetinglibrary.asco.org/record/186530/abstract

DOI:10.1200/JCO.2020.38.15_suppl.e16541

Authors:

Jeeyun Lee, Seung Tae Kim, Kyoung-Mee Kim, Eric Schultz, Stan Skrzypczak, Rob Patro, Carl Kingsford; Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea; Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea; Ocean Genomics, Inc., Pittsburgh, PA Abstract Disclosures

Research Funding:

None

Background: Immune checkpoint inhibition (ICI) has made significant breakthroughs in several tumor types including gastric cancer (GC) in recent years. We recently showed that single agent pembrolizumab demonstrated remarkable and durable response in MSI and EBV GC. However, the response to ICI remains low in MSS and most patients progress after initial response. We explore novel targets in ICI-resistant GC patients by analyzing pre- and post-resistant expression. Methods: Of the 61 patients who were enrolled onto our previously reported phase II pembrolizumab trial (NCT#02589496), whole transcriptome RNA-seq analysis of 10 paired freshly collected tissue samples (all from primary gastric tumors) was performed using TruSeq. All biopsies were performed at progression following stable disease (SD) or partial response (PR) to pembrolizumab. All patients had a MSI status of MSS and EBV negative. Molecular features were extracted using the validated Ocean Genomics, Inc. gene expression analysis pipeline, which trims reads, computes transcript- and gene-level expression, predicts structural variants, assembles novel isoforms, and computes per-sample quality control metrics, among other analyses. Samples that passed quality control, with mapping rates > 88%, were selected for analysis. Differentially expressed (DE) genes between resistant and pre-resistant samples were identified using a statistical test with a study design that accounted for the pairing of samples for each patient. Results: 16 genes (GENCODE v31) had absolute log2-fold expression changes (L2FC) > 2, P-value < 10−5 and FDR-adjusted P-value < 0.05. Because sex was only partially controlled for, we excluded genes on the X and Y chromosomes. We also excluded non-protein encoding genes and pseudogenes. The 7 remaining genes are in the table. PDL-1 (CD274) was not identified as significantly DE (FDR-adjusted P-value > 0.9999). Conclusions: This is the first study to identify novel targets in pembrolizumab-resistant GC using RNA-seq algorithms beyond PDL-1.