Lung cancer is the leading cause of cancer-related death, with the majority of cases being diagnosed at an advanced stage. However, early detection and treatment can greatly reduce mortality. Lung cancer can be detected in its early stages if small growths or lesions in the lung, known as lung nodules, are deemed to be malignant. Yet, when lung nodules are discovered incidentally or through screening, they are often difficult to evaluate. The high false positive rate (96.4% according to the National Lung Screening Trial) associated with screening for suspicious nodules via CT scan may lead to unnecessary follow-up, which may include repeat imaging or tissue biopsy. Therefore, a non-invasive biomarker that could differentiate malignant from benign nodules would be of significant clinical utility.

DNA circulating in the blood, called circulating cell-free DNA (ccfDNA), is a mixture of DNA originating from multiple cell types and is found even in healthy individuals. In cancer patients, a portion of this ccfDNA may have been shed by the tumor. Evaluation of ccfDNA and/or circulating tumor DNA (ctDNA) represents a potential non-invasive method of cancer detection. Changes in methylation, an epigenetic modification generally associated with gene silencing, are commonly associated with cancer. To evaluate the potential of ccfDNA methylation to predict lung nodule status, 10 patients with benign nodules and 14 patients with malignant nodules were included in this discovery set. The patients were consented and plasma ccfDNA samples were profiled using the Infinium Methylation EPIC array to assay 850,000 methylation sites.

159 differentially methylated regions (DMRs, unadjusted p-value > 0.05 and Δβ > 0.15) were identified. Supervised clustering using these DMRs performed well at separating the malignant from the benign cases, indicating that the two subsets had distinct methylation patterns at these loci. These data suggest that a panel of DMRs could be able to discern benign from malignant nodules, although these results will require validation in a larger cohort.

A method for determining the composition of ccfDNA, known as deconvolution, was also applied to the data. By referencing the unique methylation signatures of different types of cells, the percentage of ccfDNA derived from each of the assessed cell types could be estimated. Methylation data from lung cancer tissue was retrieved online from the Genomic Expression Omnibus and was used as one of the reference cell types. However, there were no statistically significant differences detected between benign and malignant cases for any cell type.