Abstract 3684: Comprehensive error suppressing approach allowing enhanced minimal residual disease detection in lung cancer patients

Abstract Objective: Identifying minimal residual disease (MRD) in postoperative lung cancer patients is crucial for timely cancer intervention. However, current fixed panel methods suffer from low sensitivity, whereas personalized panels face challenges as long yield time and require patient tissue samples. Our objective was to develop an accurate and efficient method for detecting circulating-tumor DNA (ctDNA) in post-surgery lung cancer patients. Methods: The study retrospectively enrolled 39 lung cancer patients (14 relapse, 25 non-relapse) with 63 post-surgical plasma samples. The mutational profiles of cell-free DNA (cfDNA) were obtained using SHIELDINGTM ULTRA Panel by Geneseeq, covering hotspots in 2365 cancer-associated genes. A comprehensive error suppressing algorithm, which combines intra-run sampling adaptive noise control, cfDNA fragment profiling, normal pool, and white blood cell background, was utilized to eliminate the background artifacts. The plasma ctDNA status were identified under both tumor-informed and tumor-naïve scenarios. Results: Our method yielded results on par with personalized panels, achieving a sensitivity of 85.7% at the specificity of 92.0%, with a negative predictive value (NPV) of 92.0% in longitudinal analysis. The ctDNA positive patients had a 21-fold increased risk (Hazard Ratio [HR] = 20.7, p = 9.2 × 10^(-5)) compared to the ctDNA negative patients. Remarkably, the landmark analysis using 39 samples (⇐ 7 days post-surgical) showed a 57.1% sensitivity at 100.0% specificity. Patients with a landmark (⇐ 7 days post-surgical) ctDNA positive status showed an increased risk of 16.2 times (HR = 16.2, p = 3.5 × 10^(-6)). Furthermore, we found that our approach could maintain its level of performance under tumor-naïve scenario, showing HRs of 12.7 (p = 1.3 × 10^(-4)) for longitudinal ctDNA positive patients and 12.3 (p = 3.7 × 10^(-5)) for landmark ctDNA positive patients. The ctDNA status determined by the tumor-informed and tumor-naïve approaches showed a substantial agreement, yielding a concordance of 90.5% (Cohen’s Kappa = 0.738). Conclusion: Herein, we report a comprehensive error suppressing approach for removing background artifacts by fixed MRD panel, allowing an enhanced MRD detection in post-surgical lung cancer patients. Our approach, which accurately identified ctDNA status at a cost-effective rate, outperformed other fixed panels while showing comparable performance with personalized panels. Furthermore, our method can be utilized under tumor naïve scenarios when personalized panels are unfeasible in the absence of tumor information. Finally, the faster turnout time of fixed panel makes our approach more suitable for clinical landmark MRD detection than a personalized panel. Citation Format: Hua Bao, Fanchen Meng, Zhijun Xia, Qinhong Sun, Siwei Wang, Wanxiangfu Tang, Ningyou Li, Min Wu, Jinfeng Zhang, Baihan Zhu, Dongqin Zhu, Xue Wu, Rong Yin, Yang Shao. Comprehensive error suppressing approach allowing enhanced minimal residual disease detection in lung cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3684.