Abstract 3322: Comprehensive genomic profiling (CGP)-informed circulating tumor DNA analysis for postoperative monitoring of non-small cell lung cancer: BEACON study

Abstract Background: Detection of circulating tumor DNA (ctDNA) after surgery in early-stage non-small cell lung cancer (NSCLC) is associated with poorer outcomes. The previous MEDAL study (PMID: 37683638) demonstrated that PROPHET, a whole-exome sequencing (WES)-based, individualized ctDNA detection assay, outperformed fixed-panel assays in a direct comparison for NSCLC. However, WES requires high-quality tissue samples and is associated with high costs. Here, we further aim to validate CGP sequencing-based individualized ctDNA assay for minimal residual disease (MRD) detection in NSCLC. Methods: The BEACON study is a multicenter, prospective cohort study that collected plasma samples at pre-surgery, 30 days post-surgery, and at longitudinal time points from patients with early-stage NSCLC between February 2022 and September 2024. A 520-gene panel sequencing-based personalized tumor-informed technology known as PROPHET_lite was used, which employs deep sequencing up to 25 patient-specific variants. PROPHET_lite was developed to detect MRD with a limit of detection (LoD) of 0.01% and a sample-level specificity exceeding 99%, as established during analytical validation. Results: In our cohort, excluding two patients with no detectable mutations, 130 patients successfully underwent ctDNA assays, with a median of five tumor-specific tracking variants. Of these, 77.7% had stage I NSCLC, and 86.2% had adenocarcinoma. ctDNA was detected pre-surgery in 129 patients, with a 31% positivity rate. Among patients with stage I NSCLC, this rate was comparable to the PROPHET assay in the MEDAL study. After a median follow-up of 445 days (95% CI: 376-536), 18 patients experienced distant recurrence. Landmark MRD status was significantly associated with DFS (HR 3.36, 95% CI: 1.3-8.71, p = 0.008); the negative predictive value (NPV) and positive predictive value (PPV) for recurrence were 88.3% and 53.8%, respectively. Longitudinal MRD status was also significantly associated with DFS (HR 15.1, 95% CI: 4.33-52.61, p < 0.001); the NPV and PPV for recurrence were 95.8% and 66.7%, respectively. Among the 18 patients with recurrence, MRD was positive in 14 cases at or before recurrence, with a median lead time of 295 days before CT detection. The remaining four MRD-negative cases were associated with brain metastases or prolonged intervals without ctDNA detection. Multivariate Cox analysis further revealed that longitudinal MRD status (HR 52.28, 95% CI: 7.37-370.8, p < 0.001) was an independent predictor of recurrence. Conclusion: Our study demonstrated that the MRD approach, with its high clinical accessibility, broad applicability, and good sensitivity, enables prognostic predictions and supports disease monitoring during surveillance in NSCLC patients. Citation Format: Guanggui Ding, Jinquan Yuan, Peikun Ding, Xiaoxiang Huang, Junjun Li, Hongsheng Xie, Yuxin Wen, Jiaochun Shi, Shihua Dou, Wohua Han, Xuning Fan, Yuan Sun, Zhihong Zhang, Bing Li, Zhou Zhang, Jianping Zhou, Lin Yang. Comprehensive genomic profiling (CGP)-informed circulating tumor DNA analysis for postoperative monitoring of non-small cell lung cancer: BEACON study [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 3322.