Abstract 5743: Comprehensive next generation sequencing profiling in combination with transcriptomic-based tumor molecular subtyping and harmonized TMB calculation using paired specimens from late-stage CRC patients

Abstract

Abstract Background: Colorectal Cancer (CRC) can be classified into transcriptomics subtypes such as, stromal or immunogenic. A previously demonstrated consensus molecular subtype (CMS) classification method was utilized on CRC samples. We explored the potential clinical utility of combining a transcriptome derived subtypes approach with WES by demonstrating the association among subtypes and tumor characteristics such as whole exome based microsatellite instability (MSI), tumor mutational burden (TMB) and karyotype. Methods: Tumor tissue and paired blood samples were collected from 19 late-stage, treatment-naïve colorectal cancer (CRC) patients. gDNA and RNA were extracted and analyzed by the Personalis® ImmunoID NeXT Platform®. RNA-seq results were normalized with the R DESeq2 package, with CRC CMS classification performed by the R package CMScaller. Somatic variants, copy number variants and MSI were evaluated using paired tumor/normal (T/N) samples. Copy number (CN) was characterized both genome-wide as a ploidy estimate and focally as the number of amplified regions. TMB was computed in alignment with the Friends of Cancer Research phase I guidelines. Results: The most commonly mutated genes included TP53 (19/19), APC (16/19), and KRAS (9/19). CMScaller identified molecular subtypes (CMS1 immune/MSI n=1; CMS2 canonical/epithelial n=6; CMS3 metabolic/epithelial n=2; CMS4 mesenchyme/stromal n=8; unspecified n=2) in 17/19 samples. CMS2 and CMS4, the two most common molecular subtypes in this cohort, differed in terms of exome-wide MSI percent (p=0.03, Student’s t) and purity (p<0.007, Student’s t). TMB, in silico ploidy estimates, and focal copy number changes were assessed for associations with tumor purity, with a significant association only between purity and ploidy (p=0.014; Bonferroni q=0.42; Student’s t) at the highest and lowest purity quartiles. Exome-wide percent unstable MSI was significantly associated with tumor purity (p=0.0053, lowest vs. highest quartile purity), with high purity tumors possessing more MSI. No association between TMB and Union for International Cancer Control stage was identified. Conclusion: It has been demonstrated that the ImmunoID NeXT Platform’s transcriptomics capabilities enable CMS classification in most samples (17/19, 89%), an overall performance superior or comparable to previously reported results. An association between tumor purity and WES MSI using a broader exome-wide measurement was identified with transcriptomic based molecular subtyping. It has been corroborated that future comprehensive molecular classifiers can expand on transcriptomics based classification by leveraging DNA-based measurements to further delineate subtypes and eventually lead to biomarker driven precision oncology focused patient selection. Citation Format: Danyi Wang, Lee McDaniel, Sean Boyle, Juergen Scheuenpflug, Zheng Feng. Comprehensive next generation sequencing profiling in combination with transcriptomic-based tumor molecular subtyping and harmonized TMB calculation using paired specimens from late-stage CRC patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5743.