Early detection of pancreatic cancer using 5-hydroxymethylation profiles in plasma-derived cell-free DNA.

• DOI: 10.1200/jco.2023.41.4_suppl.672
• Publication Date: 2023-01-24T21:05:07Z
• Authors: Anna Bergamaschi; David Haan; Micah Collins; Gulfem Guler; Melissa Peters; Lauren Gigliotti; Shimul Chowdhury; Wayne Volkmuth; Samuel Levy
• Keywords: Cell-free fetal DNA; Epigenomics

672 Background: Pancreatic cancer is one of the deadliest cancers, with approximately 15-20% patients who present at diagnosis a resectable disease. The major barrier to better outcomes lack early-detection molecular tools enable timely intervention. We have developed test that enables detection pancreatic from simple blood draw. incorporates novel, genome-wide sequencing-based epigenomics method enriches for DNA loci undergo active de-methylation. measurement 5-hydroxymethylcytosine (5hmC) provides unique and stable biomarker early including cancer. Methods: Whole-blood was obtained training cohort 660 individuals (consisting 132 cancers (PaCa) 528 non-cancers) validation 2,150 102 PaCa 2,048 non-cancers). Cell-free (cfDNA) isolated plasma which 5hmC whole-genome libraries were generated sequenced. Logistic regression algorithms employed using feature sets combined physical characteristics fragments optimally partition non-cancer samples. Results: Cross model yielded an overall sensitivity 65.9%,(95% CI, 57.2%–73.9%), early-stage (stage I-II) 57.1% (95% 44%–69.5%) specificity 98%. further validated in separate, non-overlapping set blinded independently processed samples 68.3% 51.9%–81.9%) 96.9% 96.0%–97.6%). Conclusions: Our results demonstrate plasma-derived cfDNA profiles accurate PaCa, providing valuable non-invasive tool especially those high risk disease, genetic predisposition newly diagnosed type 2 diabetes. A larger clinical study (NODMED - NCT05188586) ongoing will provide this deadly