1260TiP Enhancing lung nodule discrimination with a novel cfDNA test: The cancer signature ensemble (CSE) approach

Lung cancer is the leading cause of cancer-related deaths worldwide, with over 40% of patients diagnosed at stage IV, mainly due to difficulties in early detection. Although low-dose CT (LDCT) screening assists in early diagnosis, its high false-positive rate demands more accurate and minimally invasive biomarkers. In LDCT screening, about 5% of cases exhibit lung nodules with a 5% or higher risk of malignancy (Lung-RADS category 4); however, most are benign, with actual lung cancers representing only 0.6% of the total screened population. A diagnostic test is essential to avoid unnecessary invasive examinations and identify individuals at higher lung cancer risk. Other liquid biopsy approaches using restricted range of biomarkers have diagnostic limitations for early-stage cancers. We are recruiting participants aged between 50 and 80 years old who have a smoking history of at least 20 pack-years and are either current smokers or former smokers who quit within the past 15 years. Eligible individuals must have lung nodules identified on chest CT scans as Lung-RADS category 4B or 4X and be scheduled for a pathological diagnosis for lung cancer with TNM stage IA. Five participating hospitals are collecting whole blood (15mL) samples and sending them to a central laboratory, IMBdx. IMBdx is extracting cfDNA, performing quality control (QC), and assessing [1] whole-genome methylation patterns, [2] fragmentomics, and [3] copy number variations. The results will be integrated into a machine learning-based algorithm to establish a cancer signature ensemble (CSE) index as a 5-tier categorical indicator for differentiating malignant and benign nodules. A training set of 150 cases will be used to develop the CSE index, and its diagnostic performance will be validated in a validation set of 150 cases. The CSE index for lung cancer discrimination is expected to exhibit a sensitivity and specificity of 75% and 95%, respectively. The study's outcomes aim to reduce unnecessary CT follow-ups or invasive tissue sampling/surgery. Future research will explore its applicability as a screening test for non-smokers or individuals at lower lung cancer risk. Clinical Research Information Service of Korea (CRiS), Registration ID: KCT0007484. The authors. National Cancer Center Korea (HA22C0140000022).