Abstract 4565: CANARy-TF: an improved copy number and tumor fraction estimator for liquid biopsy applications

Abstract Background: Accurate detection of copy number alterations (CNAs) and reliable estimation of tumor fraction from liquid biopsies are crucial for effectively triaging patient samples in large clinical trials. This is particularly vital in scenarios where variant allele frequency (VAF) data is unavailable, yet an estimation of tumor fraction remains necessary. We hypothesize that specific characteristics of circulating free DNA (cfDNA) can enhance the performance of algorithms aimed at detecting CNAs and estimating tumor fraction. Methods: We introduce CANARy-TF (Copy number ANomaly Assessment and RecoverY Tumor Fraction), a novel algorithm designed to accurately identify CNAs and estimate tumor fraction from low-pass whole-genome and off-target sequencing. CANARy-TF utilizes sequencing coverage data to pinpoint amplifications and deletions by comparing samples against a panel of normals. Additionally, it incorporates unique cfDNA features, such as fragment size and shifts in the distribution of wildtype/mutant fragments, to improve tumor fraction estimation. Results: The performance of CANARy-TF was evaluated on a cohort of 23 cfDNA samples from patients with non-small cell lung cancer (NSCLC), alongside 149 cfDNA samples from healthy individuals as a background population. For all samples, tumor-informed VAF data were available. Analysis revealed a strong correlation between true VAF and the estimated tumor fraction by CANARy-TF, with a Spearman’s rho of 0.86 (p-value = 1.38e-07). In silico mixtures representing tumor prevalence from 0.001% to 5% were generated, demonstrating a limit of detection for the algorithm at around 1%. Conclusion: CANARy-TF effectively estimates CNAs and tumor fraction from plasma cfDNA samples. Notably, this algorithm is robust when applied to low-pass whole-genome sequencing, offering an efficient solution for triaging samples in both clinical settings and liquid biopsy studies. Citation Format: Diego Almanza, Takeshi Sugio, Emily Hamilton, David Kurtz, Maximilian Diehn, Mohammad Shahrokh Esfahani, Ash A. Alizadeh. CANARy-TF: an improved copy number and tumor fraction estimator for liquid biopsy applications [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4565.