Detailed Abstract
[Poster Presentation 1 - Biliary & Pancreas (Pancreas Disease/Surgery)]
[BP PP 1-S3] Fragmentomics of Cell-free DNA As a Sensitive Biomarker for Early Detection of Pancreatic Cancer
Lingdi YIN 1, Cheng CAO 1, Kai ZHANG 1, Feng GUO 1, Jianmin CHEN 1, Min TU 1, Jishu WEI 1, Junli WU 1, Wentao GAO 1, Yi MIAO 1, Zipeng LU 1, Kuirong JIANG 1
1 Pancreas Center, The First Affiliated Hospital of Nanjing Medical University, CHINA
Background : Pancreatic Ductal Adenocarcinoma (PDAC) is one for the most lethal types of cancer. Nearly 80% of the PDAC patients are diagnosed at advanced unresectable stages. Therefore, an accurate early diagnosis assay is urgently needed for reducing PDAC related mortalities. In this study, we aim to utilize cfDNA fragmentomics and machine learning techniques for sensitively detecting PDAC patients.
Methods : In this study , we recruited 278 PDAC patients and 278 healthy controls. These 556 participants were assigned into a training cohort (166 PDAC patients, 167 healthy controls) and a validation cohort (112 PDAC patients, 111 healthy controls) in a 2:1 ratio. Plasma sample was collected from each participants for shallow WGS (~5X). Four fragmentomics profiles, including copy number variation, fragment size, mutation signatures, and FRAGmentomics-based Methylation Analysis (FRAGMA) were utilized by a machine learning algorithm for developing a predictive model.
Results : The predictive model showed an exceptional performance to distinguish PDAC patients from healthy controls, yielding Area Under the Curve (AUC) of 0.993 in the training cohort (5-fold cross validation) and 0.990 in the validation cohort. In the training cohort, our model was capable of detecting PDAC patients at 97.0% sensitivity and 95.5% specificity while using 0.47 as cutoff. The validation cohort achieved 99.1% sensitivity at 89.2% specificity while applying the same cutoff.
Conclusions : Our model was able to accurately detect PDAC at early stages, by incorporating fragmentomics features through machine learning. It can potentially be used for PDAC early screening, and therefore reducing PDAC related mortalities.
Methods : In this study , we recruited 278 PDAC patients and 278 healthy controls. These 556 participants were assigned into a training cohort (166 PDAC patients, 167 healthy controls) and a validation cohort (112 PDAC patients, 111 healthy controls) in a 2:1 ratio. Plasma sample was collected from each participants for shallow WGS (~5X). Four fragmentomics profiles, including copy number variation, fragment size, mutation signatures, and FRAGmentomics-based Methylation Analysis (FRAGMA) were utilized by a machine learning algorithm for developing a predictive model.
Results : The predictive model showed an exceptional performance to distinguish PDAC patients from healthy controls, yielding Area Under the Curve (AUC) of 0.993 in the training cohort (5-fold cross validation) and 0.990 in the validation cohort. In the training cohort, our model was capable of detecting PDAC patients at 97.0% sensitivity and 95.5% specificity while using 0.47 as cutoff. The validation cohort achieved 99.1% sensitivity at 89.2% specificity while applying the same cutoff.
Conclusions : Our model was able to accurately detect PDAC at early stages, by incorporating fragmentomics features through machine learning. It can potentially be used for PDAC early screening, and therefore reducing PDAC related mortalities.
SESSION
Poster Presentation 1
Zone A 3/21/2024 2:50 PM - 3:30 PM