Abstract 6280: Leveraging RNA and long-read DNA to improve genetic etiology identification in individuals with elevated cancer risk: A pilot study in individuals with Li-Fraumeni-like phenotype.

Abstract Individuals with clinical Li-Fraumeni Syndrome (LFS) are at hereditary risk of developing multiple cancers in their lifetime. Among those meeting classic LFS criteria, about 10% do not have a germline mutation identified in TP53. Long-read DNA sequencing identifies complex and structural variants (SVs) missed on short-read sequencing. Clinical germline testing labs also recently began using short-read RNA-seq to characterize alternative splicing events. This project aims to explore underlying genetic etiology of those with LFS and no identifiable germline TP53 variant using long-read DNA-seq and RNA-seq. We performed germline Oxford Nanopore long-read DNA-seq and Illumina short-read RNA-seq on blood samples from 21 individuals participating in the NCI’s Li-Fraumeni Syndrome Study. Selected individuals had a personal history of cancer and/or at least 2 1st degree relatives diagnosed with cancer and no known pathogenic germline TP53 variant. Insertion-deletion variants (indels) in the long-read DNA-seq were called using Clair, and SVs were called using Severus. Annotation was performed using AutoGVP (indels) and LRGenotate (SVs). Variants were excluded if they were designated as benign/likely benign or if they were common (1% for indels, 0.1% for SVs) in GnomAD populations. Variants were prioritized if they were designated as “pathogenic” or “likely pathogenic” or if they were predicted to have functional effects. We used RNA-seq to compare relative expression levels of candidate associated genes versus housekeeping genes as compared to control blood samples from the GTEx dataset. We also used rMATS to calculate a Percent Spliced In index (PSI) for splicing variants to parallel commercial labs. We identified 21 indels in 11 genes classified as pathogenic/likely pathogenic and confirmed by IGV. The genes most affected by these prioritized variants were ATXN3, ADCY2, KRT10, ITPKB, CCDC103, TBP, MUC5B, KRT81, MPO and GBE1. Two related individuals (one diagnosed with cancer, one not) had a TP53 variant of uncertain significance (c.845GA) and two unrelated individuals had variants in ATM: c.9023GA and c.8147TC. We identified 2,453 SVs, with one deletion within DHX32 and other SVs most often involving SOS2, DICER1, NTHL1, TSC2, RUNX1, WAS, and GPC3. We did not observe any significant difference in expression levels of TP53 compared to housekeeping genes when comparing our samples to GTEx. We did not observe any family-specific PSI patterns. Long-read DNA-seq and RNA-seq can help characterize the familial cancer syndromes without established genetic etiology. In this pilot study, we are actively evaluating potential causal genes related to clinical presentations of LFS with no identifiable pathogenic germline TP53 variant. Citation Format: Yewon Kim, Anvith Kakkera, Asher Bryant, Sharon A. Savage, Michael C. Dean, Payal Khincha, Misha Kolmogorov, Sheila Rajagopal. Leveraging RNA and long-read DNA to improve genetic etiology identification in individuals with elevated cancer risk: A pilot study in individuals with Li-Fraumeni-like phenotype abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6280.