Aggregated analysis of 1,000 patients with cancer to assess the benefits of integrated whole exome and whole transcriptome sequencing.

person Konstantin Chernyshov BostonGene, Waltham, MA info_outline Konstantin Chernyshov, Anna Love, Viacheslav Senichkin, Georgy Sagaradze, Konstantin Rumyantsev, Vladimir Kushnarev, Ekaterina Postovalova, Nikita Kotlov, Anna Ogloblina, Alexander Bagaev, Nathan Fowler

Authors person Konstantin Chernyshov BostonGene, Waltham, MA info_outline Konstantin Chernyshov, Anna Love, Viacheslav Senichkin, Georgy Sagaradze, Konstantin Rumyantsev, Vladimir Kushnarev, Ekaterina Postovalova, Nikita Kotlov, Anna Ogloblina, Alexander Bagaev, Nathan Fowler Organizations BostonGene, Waltham, MA, BostonGene, Corp., Waltham, MA Abstract Disclosures Research Funding No funding received None. Background: As the field of precision medicine rapidly expands, clinical oncologists are turning to integrated whole transcriptome (WTS) and whole exome (WES) sequencing to identify the most effective treatment options. We performed a retrospective cohort analysis of WES and WTS data from 1,000 patients with 15 major cancer types to define meaningful relationships between transcriptomic-based tumor microenvironment (TME) subtypes and pan-cancer genomic biomarkers, such as tumor mutational burden (TMB) and microsatellite instability (MSI) status, that are prognostic for outcome and predictive for checkpoint inhibitor response. Methods: We retrospectively analyzed WES/WTS data from cancer patients for biomarkers using our internal bioinformatics workflow. Using transcriptome-based methods described by Bagaev et al., samples were classified into one of four TME subtypes prognostic of patient outcome ( Cancer Cell , 2021). Statistical analysis consisted of Spearman’s rank correlation. Results: Comprehensive WES and WTS analysis provided by the BostonGene Tumor Portrait Test identified biomarkers based on accepted guidelines specific to the patient's cancer type in 22.5% of cases. In addition, we revealed a correlation between WTS and IHC for cancer-relevant genes, including CDX2 (R = 0.68), KRT7 (R = 0.85), KRT20 (R = 0.86), NECTIN4 (R = 0.81), and TACSTD2 (R = 0.81), indicating WTS may also be beneficial for estimating protein expression levels for cancer-specific biomarkers. Analysis of the TME showed that Fibrotic (F) and Immune Desert (D) were the most common subtypes. The F and D subtypes were linked to poor prognosis and were especially prevalent in liver and pancreatic cancers, respectively. TMB-high and MSI status were more prevalent in the immune-enriched subtypes, Immune-Enriched, Fibrotic (IE/F) and Immune-Enriched, Non-fibrotic (IE), over the less favorable F and D subtypes. The frequency of TMB-high and MSI status also differed in both IE subtypes associated with favorable prognosis, suggesting that TME subtyping provides additional prognostic power independent of TMB-high and MSI status. Conclusions: A deep interrogation of our integrated approach to genomic and transcriptomic profiling confirmed the advantages of a unified molecular analysis of tumor traits. We identified clinically significant biomarkers based on molecular features, demonstrated additional capabilities of TME analysis, and illustrated significant correlations between WTS and IHC data. These findings indicate that comprehensive WTS and WES analysis can assist clinicians in optimizing treatment plans as the landscape of personalized therapy continues to grow. Frequency of pan-cancer biomarkers across conserved TME subtypes. TME subtype F, 34.0% D, 30.7% IE/F, 18.0% IE, 17.3% Total TMB-high, % 5.8 8.4 7.5 12.1 8.0 MSI, % 0.7 1.6 0.7 4.3 1.6