Introduction: Indeterminate pulmonary lesions (IPL) detected by CT pose a significant clinical challenge, frequently necessitating long-term surveillance or biopsy for diagnosis. In this pilot investigation, we performed whole exome sequencing (WES) of plasma cell free (cfDNA) and matched germline DNA in patients with CT-detected pulmonary lesions to determine the feasibility of somatic cfDNA mutations to differentiate benign from malignant pulmonary nodules. Methods: 33 patients with a CT-detected pulmonary lesions were retrospectively enrolled (n = 16 with a benign nodule, n = 17 with a malignant nodule). Following isolation and amplification of plasma cfDNA and matched peripheral blood mononuclear cells (PBMC) from patient blood samples, WES of cfDNA and PBMC DNA was performed. After genomic alignment and filtering, we looked for lung-cancer associated driver mutations and next identified high-confidence somatic variants in both groups. Results: Somatic cfDNA mutations were observed in both groups, with the cancer group demonstrating more variants than the benign group (1083 ± 476 versus 553 ± 519, p < 0.0046). By selecting variants present in >2 cancer patients and not the benign group, we accurately identified 82% (14/17) of cancer patients. Conclusions: This study suggests a potential role for cfDNA for the early identification of lung cancer in patients with CT-detected pulmonary lesions. Importantly, a substantial number of somatic variants in healthy patients with benign pulmonary nodules were also found. Such "benign" variants, while largely unexplored to date, have widespread relevance to all liquid biopsies if cfDNA is to be used accurately for cancer detection.
Keywords: cancer detection; cell free DNA; lung cancer—diagnosis; pulmonary nodule; whole exome sequencing.