Accurate and affordable multi-cancer early detection and localization via plasma cfDNA multi-omic profiling
Background: Early detection of major cancers, including HCC, lung cancer, colotectal cancer and etc., can improve prognosis, but has not yet been clinically implemented due to lack of accurate and affordable means. In recent years, DNA methylation has been proved by numerous studies to be ideal biom...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-02-01
|
| Series: | The Lancet Regional Health. Western Pacific |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666606524004334 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background: Early detection of major cancers, including HCC, lung cancer, colotectal cancer and etc., can improve prognosis, but has not yet been clinically implemented due to lack of accurate and affordable means. In recent years, DNA methylation has been proved by numerous studies to be ideal biomarkers for Multi-Cancer Early Detection (MCED), owing to their early occurrence, high accuracy, and tissue specificity. However, the gold standard for high throughput methylation analysis, the whole-genome bisulfite sequencing (WGBS), is highly molecular damaging and labour intensive, making high throughput cell-free DNA (cfDNA) methylation analysis challenging and unaffordable for widespread application. Plasma cfDNA contains numerous molecular features useful for MCED, including cfDNA fragmentation profile, which can be obtained via whole-genome sequencing (WGS). Findings: In this study, we performed both WGBS and WGS on 17 healthy individuals, 26 HCC patients, 32 lung cancer patients, and 15 colorectal patients to prove the feasibility of inferring cfDNA methylation patterns using cfDNA fragmentation profile. By combining cfDNA cleavage profile of CpG sites with machine learning algorithms, we have identified specific CpG cleavage profile as biomarkers to predict the methylation status of individual CpG sites, based on which we built in silico classifiers for prediction of each of the four groups previously mentioned, achieving considerable performance of AUC ranging from 0.8896 to 0.959. Therefore, methylation profile, a widely used epigenetic biomarker, can be obtained from a simple WGS assay for MCED in a cost-efficient manner. Interpretation: To further explore the potential of our method for MCED, we are recruiting more than 2000 plasma samples from a retrospective cohort covering 12 common cancers in China, including cancers in lung, liver, colon, gastric, pancreas, esophagus, breast, cervix, kidney, nasopharyngeal, bile duct, and gallbladder. Besides methylation patterns inferred from cfDNA fragmentation profile, we also combined other tissue specific cfDNA features including molecular end motif, CNV and transcription factor coverage to trace the organ origin of cancerous signals for cancer typing. |
|---|---|
| ISSN: | 2666-6065 |