A lung nodule is a rounded lesion which can measure up to 3 cm in size. Lung nodules can be hard to identify based on just morphology, which can lead to an indeterminate diagnosis. The current standard for molecular detection is tissue biopsies, however a biopsy is an invasive procedure and re-biopsies are not always possible. 

Cell-free DNA is found in the blood of both healthy patients and patients with cancer. By using the methylation signatures of a patient's cfDNA, it might be possible to distinguish malignant and benign cancers. Starting with patient blood samples through which we extracted plasma and then further cfDNA, across two separate years of methylation analysis.

Our goal with this project was to validate the data and results found in a previous study conducted in the lab. In the 2022 cohort, 24 patients cfDNA was analyzed, and in the 2023 cohort 35 patients cfDNA was analyzed. We extracted cfDNA from plasma of patients blood, then sent samples off for methylation profiling using a EPIC v2.0 array. A EPIC v.1.0 Array was used for the 2022 cohort. Then, we ran the CpG sites through a data pipeline to extract the differentially methylated regions.  

When supervised clustering was done separately for the 2022 cohort and the 2023 cohorts respectively, the separation of patients who had malignant versus benign lung nodules was very evident as demonstrated by the clear dark and light gray separation on figures C and D. However, the Differentially methylated regions from the 2023 cohort were not successful when used on the 2022 cohort and the 2022 differentially methylated regions were not able to clearly separate the malignant versus benign patients.

In the future, it could be beneficial to run methylation analysis on a combination of two cohorts to increase sample size.