In a Chinese study reported in the Journal of Clinical Oncology, Yin et al attempted to develop a cell-free DNA (cfDNA) fragmentomics–based model for early detection of pancreatic ductal adenocarcinoma.
Study Details
The study consisted of training and validation cohorts from several Chinese centers, including patients with pancreatic ductal adenocarcinoma and healthy subjects. Plasma was collected from participants and subjected to shallow whole-genome sequencing. The model was developed using cfDNA fragmentomics profiling that integrated copy-number variations, fragment size, mutational signatures, and methylation patterns analyzed by advanced machine learning.
Key Findings
Among 166 patients diagnosed with pancreatic ductal adenocarcinoma and 167 healthy participants in the training cohort, the model showed high accuracy in distinguishing patients with pancreatic ductal adenocarcinoma from controls, with an area under the curve (AUC) of 0.992. At a model cutoff value of 0.52 derived in the training cohort, the model had a sensitivity of 93.4% and a specificity of 95.2%.
In a validation cohort of 112 patients with pancreatic ductal adenocarcinoma and 111 healthy individuals, the model showed high accuracy, with an AUC of 0.987. At the model cutoff of 0.52, sensitivity was 97.3% and specificity was 92.8%.
In two external validation cohorts with 198 patients with pancreatic ductal adenocarcinoma (including 108 with early-stage disease) and 200 healthy individuals, the model had a sensitivity of 90.91% and a specificity of 94.5%.
The investigators concluded: “This study underscores the effectiveness of using cfDNA fragmentomics and machine learning for early detection of [pancreatic ductal adenocarcinoma]. Our approach promises significant potential in reducing [pancreatic ductal adenocarcinoma] mortalities through early intervention and could serve as a breakthrough in oncologic diagnostics.”
Kuirong Jiang, MD, PhD, of the Pancreas Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, is the corresponding author for the Journal of Clinical Oncology article.
Disclosure: The study was supported by the National Natural Science Foundation of China, National Science Foundation for Young Scholars of China, and others. For full disclosures of all study authors, visit the Journal of Clinical Oncology.
