A non-invasive methylation-based NGS assay for sensitive detection of pancreatic cancer.
e16452 Background: Non-invasive tools with high sensitivity to distinguish pancreatic cancer from benign pancreatic lesions are limited. Cell-free DNA (cfDNA) offers promise for early detection but is constrained by low DNA input and separated workflows. We developed SPIRAL (Single-Portion Input Resourceful Assay for Liquid Biopsy), which enables simultaneous genomic and epigenetic library construction from a single cfDNA sample. We evaluated a methylation-based assay using SPIRAL platform in patients with pancreatic lesions. Methods: In the prospective DAYBREAK Study (NCT05495685), peripheral blood samples were collected from patients with pancreatic cancer (n = 67) and benign pancreatic lesions (n = 31). cfDNA methylation was analyzed using the STELLA algorithm. Sensitivity and specificity were assessed overall and by disease stage and histology. Results: Overall sensitivity was 70% (48/67) with a specificity of 87% (27/31). Sensitivity increased by stage: 63% in stage I (12/19), 70% in stage II (23/33), 80% in stage III (8/10), and 100% in stage IV (5/5). Sensitivity was 79% (41/52) for pancreatic ductal adenocarcinoma and 30% (3/10) for pancreatic neuroendocrine tumors. Conclusions: A methylation-based cfDNA assay using the SPIRAL platform distinguishes malignant from benign pancreatic lesions. Ongoing studies integrating mutation and methylation features and expanding sample size may improve the performance. Clinical trial information: NCT05495685 .
Authors
- Peiru Liu (ORCID: https://orcid.org/0000-0002-2202-2994)
- Jinyu Yang
- Lingyun Gu
- X H Shi
- S W Guo
- Lei Wang
- Chaoran Zheng
- Zhihong Zhang
- Bo Yang
- Shuailai Wu
- Gang Jin
- Huan Wang
Institutions
- Second Military Medical University (CN)
- Burning Rock Biotech (China) (CN)
- Changhai Hospital (CN)
Publication Details
- Journal
- Journal of Clinical Oncology
- Published
- 2026-05-28
- DOI
- https://doi.org/10.1200/jco.2026.44.16_suppl.e16452
- Primary Topic
- Cancer Genomics and Diagnostics
- Type
- article
- Field-Weighted Citation Impact
- 0.00