In a major leap forward for early disease detection and affordable health monitoring, researchers at UCLA have developed an innovative blood test called MethylScan.
This simple, cost-effective assay analyzes tiny DNA fragments circulating in the blood to simultaneously screen for multiple cancers, liver conditions, and organ injury, and even predict ancestry—all from a single vial of blood. In early studies, supported in part by grants from the National Cancer Institute and published in the journal Proceedings of the National Academy of Sciences, MethylScan demonstrated robust performance and holds the promise to transform how we approach disease detection and health surveillance.[1]
Early and Broad Disease Detection
Early detection saves lives, especially in cancer, where survival rates plummet once the disease spreads beyond its origin.
“If you detect cancer at stage one, outcomes are dramatically better than at stage four,” explained Xianghong Jasmine Zhou, Ph.D., senior author of the study and professor of pathology and laboratory medicine at UCLA Health’s Jonsson Comprehensive Cancer Center. Yet most current screening tools focus on a single disease at a time, often require invasive procedures or costly sequencing, and may miss signals from other organs where disease is developing.
Cell-free DNA (cfDNA)—tiny fragments released into the bloodstream as cells die—offers a potential solution. Every day, billions of our cells die, shedding their DNA into the blood. cfDNA carries molecular ‘snapshots’ of what’s happening in tissues throughout the body. If we can decode these signals, we could potentially detect disease anywhere, long before symptoms appear.
The Science Behind MethylScan
The question remains: what is cfDNA, and why does it matter? When cells in our organs die, their DNA isn’t lost—it enters the bloodstream as cfDNA, a complex mixture from every tissue. In cancer or organ damage, more cfDNA from the affected tissue appears, often bearing molecular signatures unique to the disease. Traditionally, tests have sought rare tumor mutations in cfDNA, but these require ultra-deep, expensive sequencing because the signal is faint and swamped by cfDNA from healthy blood cells.
Zhou and her colleagues took a different approach. Instead of hunting for genetic mutations, MethylScan analyzes ‘methylation’—chemical tags on DNA that regulate gene activity. Methylation patterns are highly tissue-specific and change early in disease, making them powerful markers for both the presence and origin of illness.
“DNA methylation reflects the health status of a tissue,” says Wenyuan Li, Ph.D., adjunct professor, pathology and laboratory medicine, and co-author of the study.
“It’s a very informative signal,” Li added.
How MethylScan Works
A major challenge is that up to 90% of cfDNA in blood comes from normal blood cells, not from diseased organs. This background noise makes it hard and expensive to find the rare, disease-relevant DNA fragments. MethylScan overcomes this by using specialized enzymes to selectively cut away the vast majority of cfDNA that comes from blood cells, which typically have little methylation in certain regions. What remains is a concentrated pool of methylated cfDNA fragments, enriched for signals from solid organs and diseased tissues.[2]
Next, a custom genome-wide panel captures these fragments, focusing sequencing efforts only on regions likely to harbor disease signals. This strategy dramatically reduces the amount of sequencing required—achieving clinically useful depth with just 5 gigabases of data per sample, at a cost of less than $20 if sequencing is under $4 per gigabase. In other words, highly informative cfDNA analysis is brought within reach for large-scale screening.
Promising Results in Over 1,000 Individuals
To validate their method, the UCLA team analyzed blood samples from 1,061 people, including patients with liver, lung, ovarian, and stomach cancers; individuals with various liver diseases (like hepatitis B/C, alcohol-induced disease, and metabolic disorders); people with benign lung nodules; and healthy controls.
Machine learning algorithms were applied to the methylation data to classify diseases and pinpoint their tissue of origin.
The results showed:
- Accuracy: At a specificity of 98% (meaning very few false positives), MethylScan detected about 63% of cancers across all stages and 55% of early-stage cancers.
- Tissue Identification: Methylation patterns often indicated where in the body the cancer signal originated. This crucial feature allows clinicians to direct follow-up imaging or biopsies to the right organ.
- Performance by Cancer Type: Sensitivity was at least 40% for all four cancers (liver, lung, ovarian, stomach) across all stages, with an area under the Curve (AUC) of 0.938.
Impressive Results
Liver cancer is one of the fastest-rising causes of cancer death, and high-risk patients (such as those with hepatitis or cirrhosis) require regular surveillance. In these at-risk individuals, MethylScan detected nearly 80% of liver cancers, with a specificity of just over 90%. Even more impressively, the test could distinguish between different types of liver disease (viral, alcohol-related, metabolic), correctly classifying about 85% of cases—potentially reducing the need for invasive liver biopsies.
Because cfDNA methylation patterns are organ-specific, MethylScan could act as a ‘health radar.’ The test detected elevated cfDNA from the liver in patients with liver disease or cancer, and from the lung in people with benign or cancerous lung nodules. The degree of tissue damage, reflected in cfDNA levels, correlated with disease severity. This opens possibilities for monitoring organ stress or injury even before clinical symptoms arise.
Interestingly, MethylScan’s rich methylation data could also accurately predict ancestry (99% for White and 90% for Asian individuals in this study), highlighting the test’s ability to capture subtle biological signals.
Making cfDNA Methylome Sequencing Affordable
A major barrier to using cfDNA methylation for broad disease detection has been the prohibitive cost and complexity of sequencing. Tumor-derived or disease-specific DNA may comprise less than 0.1% of cfDNA, especially in early disease, requiring ultra-deep sequencing (often >1,000× coverage) to reliably detect. [3]
The researchers demonstrated that MethylScan addresses this challenge on multiple fronts [4]:
- Enzyme-Based Depletion: By using methylation-sensitive restriction enzymes, the vast majority of hypomethylated (normal) cfDNA is removed before sequencing, greatly enriching for disease-relevant fragments.
- Targeted Panel Design: The capture panel contains over 150,000 genomic regions consistently hypomethylated in healthy blood, thereby focusing the assay on potential disease markers while minimizing wasted sequencing.
- Efficient Protocol: Only undigested (hypermethylated) cfDNA fragments are sequenced, reducing cost and background noise.
- Machine Learning: Advanced algorithms analyze the complex methylation data, classify diseases, and predict the tissue of origin.
- The result: a sensitive, accurate, and scalable test that brings comprehensive blood-based screening within practical and economic reach.
Implications for the Future of Diagnostics
MethylScan’s versatility moves diagnostics beyond the ‘one test, one disease’ paradigm. By providing a panoramic view of health status across organs and diseases, this approach could transform early detection, risk assessment, and patient management. The researchers stressed the advantages of MethylScan: [5][6][7]
- Noninvasive: A single blood draw replaces the need for multiple targeted tests or invasive biopsies.
- Comprehensive: Detects multiple cancers, liver diseases, and organ injuries simultaneously.
- Early Detection: Sensitive enough to pick up early-stage cancers and subtle tissue damage.
- Affordable: Lowers the cost barrier to large-scale population screening.
- Personalized: Methylation patterns provide clues to disease type and tissue of origin, enabling tailored follow-up.
- Potential Applications: population-wide cancer screening, surveillance of high-risk patients (e.g., chronic liver disease, smokers), monitoring organ transplant health, assessing systemic stress or injury (e.g., after trauma or infection), and reducing the need for invasive procedures
Challenges and Next Steps
Despite its promise, MethylScan will require validation in larger, prospective studies before routine clinical adoption. Real-world screening must demonstrate that the test reliably detects cancers and other diseases early, minimizes false positives, and improves outcomes. The test’s ability to distinguish disease from inflammation or other benign conditions will also need refinement.
As with all machine-learning-based diagnostics, large and diverse training datasets are essential to avoid confounding factors (such as ancestry) and ensure consistent performance across populations.
The Broader Impact
The rapid progress in omics and AI-driven diagnostics is ushering in a new era of health monitoring. By unlocking the information encoded in cfDNA methylation, MethylScan exemplifies the shift from targeted, symptom-driven testing to broad, proactive health surveillance. As costs drop and technology advances, a future where a single blood test can screen for a spectrum of diseases at once—catching them before symptoms arise—now seems within reach.
Zhou and her team remain optimistic.
“This study demonstrates that blood-based methylation profiling can deliver clinically meaningful information across multiple diseases,” she noted.
“It’s an exciting advancement that brings us closer to realizing the dream of a single assay for universal disease detection,” she further added.
MethylScan represents a pioneering step in the evolution of noninvasive diagnostics, offering a glimpse into a future where comprehensive health monitoring is accessible, affordable, and actionable. By combining cutting-edge molecular biology with smart engineering and artificial intelligence, UCLA’s breakthrough blood test could help rewrite the rules of early disease detection—potentially saving countless lives and shifting the focus of medicine from late-stage treatment to prevention and early intervention.
Reference
[1] Zeng W, Liu CC, Li S, Zhou Y, Stackpole ML, Xiao Y, Hu R, Tang C, Liu Q, Zeng W, Yeh A, Melehy A, Tran B, Noor Z, Yokomizo M, Amara D, Gumate S, Ahuja P, Li DY, Zhao J, Rose I, Walker C, Malik S, Zhu Y, Tseng HR, Garon EB, French SW, Magyar CE, Dry SM, Lajonchere CM, Geschwind D, Choi G, Saab S, Shetty A, Wong CR, King KG, Lu DS, Raman SS, Xiang X, Shetty K, Mishra L, Memarzadeh S, Liu Y, Albe F, Hsu W, Krysan K, Dubinett SM, Aberle DR, Agopian V, Han SB, Wong WH, Ni X, Li W, Zhou XJ. Toward the simultaneous detection of multiple diseases with a highly cost-effective cell-free DNA methylome test. Proc Natl Acad Sci U S A. 2026 Apr 14;123(15):e2518347123. doi: 10.1073/pnas.2518347123. Epub 2026 Apr 6. PMID: 41941615.
[2] Zhang L, Li J. Unlocking the secrets: the power of methylation-based cfDNA detection of tissue damage in organ systems. Clin Epigenetics. 2023 Oct 19;15(1):168. doi: 10.1186/s13148-023-01585-8. PMID: 37858233; PMCID: PMC10588141.
[3] Stackpole ML, Zeng W, Li S, Liu CC, Zhou Y, He S, Yeh A, Wang Z, Sun F, Li Q, Yuan Z, Yildirim A, Chen PJ, Winograd P, Tran B, Lee YT, Li PS, Noor Z, Yokomizo M, Ahuja P, Zhu Y, Tseng HR, Tomlinson JS, Garon E, French S, Magyar CE, Dry S, Lajonchere C, Geschwind D, Choi G, Saab S, Alber F, Wong WH, Dubinett SM, Aberle DR, Agopian V, Han SB, Ni X, Li W, Zhou XJ. Cost-effective methylome sequencing of cell-free DNA for accurately detecting and locating cancer. Nat Commun. 2022 Sep 29;13(1):5566. doi: 10.1038/s41467-022-32995-6. Erratum in: Nat Commun. 2024 May 1;15(1):3693. doi: 10.1038/s41467-024-48018-5. PMID: 36175411; PMCID: PMC9522828.
[4] Miller RH, Pollard CA, Brogaard KR, Olson AC, Barney RC, Lipshultz LI, Johnstone EB, Ibrahim YO, Hotaling JM, Schisterman EF, Mumford SL, Aston KI, Jenkins TG. Tissue-specific DNA methylation variability and its potential clinical value. Front Genet. 2023 Jul 19;14:1125967. doi: 10.3389/fgene.2023.1125967. PMID: 37538359; PMCID: PMC10394514.
[5] Li W, Li Q, Kang S, Same M, Zhou Y, Sun C, Liu CC, Matsuoka L, Sher L, Wong WH, Alber F, Zhou XJ. CancerDetector: ultrasensitive and non-invasive cancer detection at the resolution of individual reads using cell-free DNA methylation sequencing data. Nucleic Acids Res. 2018 Sep 6;46(15):e89. doi: 10.1093/nar/gky423. PMID: 29897492; PMCID: PMC6125664.
[6] Kang S, Li Q, Chen Q, Zhou Y, Park S, Lee G, Grimes B, Krysan K, Yu M, Wang W, Alber F, Sun F, Dubinett SM, Li W, Zhou XJ. CancerLocator: non-invasive cancer diagnosis and tissue-of-origin prediction using methylation profiles of cell-free DNA. Genome Biol. 2017 Mar 24;18(1):53. doi: 10.1186/s13059-017-1191-5. PMID: 28335812; PMCID: PMC5364586.
[7] Liang N, Li B, Jia Z, Wang C, Wu P, Zheng T, Wang Y, Qiu F, Wu Y, Su J, Xu J, Xu F, Chu H, Fang S, Yang X, Wu C, Cao Z, Cao L, Bing Z, Liu H, Li L, Huang C, Qin Y, Cui Y, Han-Zhang H, Xiang J, Liu H, Guo X, Li S, Zhao H, Zhang Z. Ultrasensitive detection of circulating tumour DNA via deep methylation sequencing aided by machine learning. Nat Biomed Eng. 2021 Jun;5(6):586-599. doi: 10.1038/s41551-021-00746-5. Epub 2021 Jun 15. Erratum in: Nat Biomed Eng. 2021 Nov;5(11):1402. doi: 10.1038/s41551-021-00818-6. PMID: 34131323.
Featured image © 2017 – 2026 Fotolia/Adobe. Used with permission.
DOI: 10.14229/onco.2026.04.06.005
