Answers are grounded in this paper and cite their sources.
Develops a plasma-based classifier for lung nodules, suggesting improved malignancy assessment.
Introduction Low-dose computed tomography (LDCT) is currently the recommended screening modality; however, its high false positive rate and reliance on radiologists’ subjective judgment limit its widespread clinical application. Biomarker-based approaches utilizing DNA methylation signatures may provide a more accurate method for lung nodule classification and enhance early lung cancer detection. This study aimed to develop and validate LunaCAM Plus, a non-invasive plasma-based classifier leveraging cfDNA methylation analysis to improve lung nodule malignancy assessment and enhance early-stage lung cancer detection. Methods A total of 96 patients were enrolled (37 malignant, 29 benign, 30 healthy). Lung cancer and adjacent normal tissues were collected, and epithelial cells were isolated via fluorescence-activated cell sorting (FACS) to minimize interference from immune cells. Genomic DNA was subjected to reduced representation bisulfite sequencing (RRBS), and a set of 539 differentially methylated regions (DMRs) was identified by integrating data from Panseer and LunaCAM databases. The LunaCAM Plus classifier was developed using Methylation Encoding Score (MES) and Normalized Count of Haplotype (nCOH). Model training was conducted using 5-fold cross-validation on a training cohort (164 plasma samples: 83 malignant, 81 benign) and a internal test cohort (100 plasma samples: 50 malignant, 50 benign), followed by validation on an independent test cohort (114 plasma samples: 64 malignant, 37 benign). Results In the training cohort, LunaCAM Plus achieved AUC = 0.911, specificity = 0.94, sensitivity = 0.56. In the independent test cohort, it maintained strong classification performance (AUC = 0.851, specificity = 0.93, sensitivity = 0.50). Conclusion LunaCAM-Plus has been successfully implemented in multicenter studies and real-world clinical settings, demonstrating significant potential for widespread adoption.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: