Abstract B014: Increasing the clinical utility of transcriptome analysis in high-risk childhood precision oncology

Abstract INTRODUCTION Utilizing whole genome (WGS) and transcriptome sequencing (RNA-seq) to identify the molecular features of individual cancers sits at the heart of pediatric cancer precision medicine. Whilst the utility of WGS for mutation detection is well established, most precision medicine programs utilize RNA-seq for fusion detection, which underutilizes the potential of RNA-seq information. In ZERO (Australia’s national precision medicine program for children with high-risk cancer), we showed that combining WGS with RNA-seq identified the molecular driver of their cancer in 94% of patients (n=247). We have sought to measure the added utility integrating a comprehensive RNA-seq pipeline provides in a precision medicine program (n=477). METHODS We developed a comprehensive RNA-seq pipeline (Carbonite) to identify expressed single nucleotide variants (SNV), small insertion and deletions (InDels) and structural variant (SV) events. SVs extend beyond fusions to include intragenic deletions, duplications, inversions and insertions, previously only identifiable through WGS. Fusion detection was expanded beyond gold-standard algorithms by incorporating a reference-free WGS guided approach that identifies cryptic SVs and intragenic alternate splicing events. The application of RNA-seq measured gene abundance was applied to identify enhancer hijacking events and biallelic deletions resulting from an SV. SNV and InDel analysis reports the zygosity in comparison to germline and tumor DNA. RESULTS A total of 218 SVs and 537 SNVs/InDels were identified as either pathogenic or likely pathogenic in our cohort of 477 patients. Carbonite identified 96% of SVs in comparison to 54% identified using gold-standard fusion detection algorithms only. In addition, Carbonite resolved the expressed chimeric transcript from complex chromosome shattering events, multi-hop complex rearrangements and SVs occurring in homologous and difficult to sequence regions in 33 patients. This information also aided the functional interpretation of SVs through accurate frame calling of potential activating oncogenic SV events, and disruption of key tumor suppressor genes. Of the driver SNV/InDels identified in WGS, Carbonite corroborated 92%. Those not validated resulted in either non-sense mediated decay or the gene was not expressed. Importantly, in 22% of somatic SNV/InDels, RNA-seq altered the assessment of pathogenicity, principally through resolving allele specific expression (ASE) and confirmation of the transcriptional consequence of splice site mutations. Germline SNVs/InDels exhibited ASE in the tumor RNA in 48%, aiding in interpretation of penetrance and oncogenicity. CONCLUSIONS Our results show that the enhanced RNA-seq pipeline has clinical applicability in precision medicine beyond fusion detection, providing functional insights derived from the transcriptional consequences of SVs and SNVs/InDels. We propose that RNA-seq is an essential tool in interpretation of the pathogenicity prediction of molecular drivers of cancer. Citation Format: Chelsea Mayoh, Paulette Barahona, Angela Lin, Lujing Cui, Pamela Ajuyah, Ann Altekoester, Loretta MS Lau, Dong-Anh Khuong-Quang, Vanessa Tyrrell, Michelle Haber, Marie Wong-Erasmus, Paul G. Ekert, Mark J. Cowley. Increasing the clinical utility of transcriptome analysis in high-risk childhood precision oncology abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pediatric Cancer Research; 2024 Sep 5-8; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl):Abstract nr B014.