Abstract 7822: High-sensitivity cfDNA detection of KRAS amplification subtypes to inform targeted therapy response and resistance mechanisms.

Abstract Background: KRAS is a key oncogenic driver gene, and both mutations and amplifications serve as important therapeutic targets and resistance biomarkers. Point mutations, amplifications of the wildtype allele, or amplifications of the mutant allele (sometimes via mutant allele-specific imbalance (MASI)) in KRAS all may impact cancer therapy greatly. Detection of KRAS amplifications in cell-free DNA (cfDNA) samples is particularly challenging because KRAS spans a relatively small 45kb region within a highly variable segment on chromosome 12. Achieving sensitive and accurate detection of KRAS amplifications in cfDNA and distinguishing the amplification types remains a pressing need in tumor drug development. Methods: We developed a proprietary algorithm to enhance tumor fraction estimation and CNV calling of large genomic regions. Using 200 plasma samples , we developed a new model for the relationship among tumor fraction, observed copy number and variant allele frequencies. This model enables more comprehensive investigation of KRAS amplification events, including differentiation between focal amplification and large genomic fragment level amplification of KRAS, and between wild type allele amplification and mutant allele amplification in samples with KRAS driver mutations. The refined model also improves sensitivity of KRAS amplification calling when the amplification types can be assigned. Results: Application of the new model allowed accurate determination of KRAS amplification types in most of the evaluated samples. In an expanded cohort of 800 cfDNA samples representing diverse solid tumors, KRAS amplification positive cases increased by 10% with the use of the refined model. Amplification types (wild-type allele or mutant allele amplification; focal or large genomic level) could be resolved in more than 80% of the cases. KRAS ctDNA dynamics were observed in both baseline and on-treatment samples. Conclusions: This study provides an in-depth analysis of KRAS amplification profiles in cfDNA samples and introduces an improved model for KRAS copy number calling. The enhanced method increases detection sensitivity and yields richer biological insights by distinguishing amplification subtypes. These advancements strengthen the clinical utility of cfDNA-based KRAS assessment for therapeutic decision-making and drug development. Citation Format: Yong Huang, Lisha Zhu, Binggang Xiang, Pan Du.. High-sensitivity cfDNA detection of KRAS amplification subtypes to inform targeted therapy response and resistance mechanisms [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 7822.