Back to 2017 Abstracts
Integration of CT and cell-free DNA as a biomarker for early lung cancer detection in patients with indeterminate pulmonary nodules
Tina Tailor, Edward Patz
Duke University Medical Center, Durham, NC
Purpose: Indeterminate pulmonary nodules (IPN) pose a significant challenge in chest medicine. Recent evidence suggests that cell-free plasma DNA (cfDNA) - that is, DNA released into circulation - may be useful for cancer diagnosis and staging. We propose to integrate CT with a blood test with the hypothesis that the combination of CT and cfDNA biomarkers can differentiate benign from malignant IPN.
Materials and Methods: An IRB-approved prospective pilot study was completed in 33 patients (n=17 lung cancer, n=16 benign nodules). Following consent, patients were enrolled based on CT with an IPN. All IPN were confirmed benign or malignant (by tissue diagnosis or ≥2 year CT stability). Blood samples were collected from all patients. cfDNA and DNA from peripheral blood mononuclear cells (PBMC, serving as matched germline control) was isolated and amplified. Whole exome sequencing of cfDNA and PBMC DNA was performed to a mean depth 49X. Following genomic alignment and rigorous filtering, a subset of high-confidence genetic variants was selected (with allele frequency >1.5% for cfDNA, <1% for PBMC DNA, p<0.03). Variants were compared between groups.
Results: There were a high number of variants in both cancer and benign groups. The cancer group had significantly more variants than the benign group (p=0.0046, 2-sample t-test). By selecting those variants present in >2 patients from the cancer group but not in the benign group, we were able to identify 14 of the 17 (82%) of the cancer patients, without false positives.
Conclusions:We show that cfDNA mutations are more prevalent in patients with lung cancer than those without cancer, suggesting certain cfDNA mutations may be specific for lung cancer. Our data also support emerging evidence that a fraction of mutations seen in cancers occur in normal cells. With growing interest in cancer biomarkers, knowledge of such “benign” variants is critical for optimizing specificity. In this regard, a larger follow-up study is underway utilizing deeper targeted sequencing (depth >1000x), which will allow the development of a robust classifier utilizing CT and cfDNA to effectively and expeditiously detect lung cancer in patients with IPN.
Back to 2017 Abstracts