UCLA blood test detects multiple cancers and diseases from single sample

Researchers at UCLA have developed a blood test that can detect multiple cancers and diseases from a single sample by reading chemical markers on DNA fragments already circulating in your bloodstream. The test, called MethylScan, analyzes patterns of DNA methylation—chemical tags that sit on genetic material and change when cells become diseased or cancerous. The work, published in the Proceedings of the National Academy of Sciences, represents a shift in how doctors might screen for serious illness: instead of looking for rare tumor mutations buried in a sea of normal blood cells, the test reads the health status written into the DNA itself.

Every day, between 50 and 70 billion cells in the human body die. Their DNA doesn't vanish; it spills into the bloodstream as cell-free DNA, carrying molecular signals from every organ. This is the insight that drives MethylScan. Dr. Jasmine Zhou, the study's senior author and a professor of pathology and laboratory medicine at UCLA's Jonsson Comprehensive Cancer Center, emphasizes that early detection transforms outcomes. "Survival rates are far higher when cancers are caught before they spread," she said. "If you detect cancer at stage one, outcomes are dramatically better than at stage four."

The technical challenge is substantial. About 80 to 90 percent of cell-free DNA in the blood comes from normal blood cells, creating noise that makes it expensive and difficult to spot the rare fragments signaling disease. The UCLA team solved this by developing a technique to strip away most of this background noise before sequencing. Using specialized enzymes, they selectively remove unmethylated DNA fragments—the ones mostly from healthy blood cells—and enrich for methylated DNA from solid organs, including potentially diseased ones. This approach dramatically reduces the amount of sequencing needed. Achieving the necessary depth of analysis requires only 5 gigabytes of data per sample, bringing the cost to less than $20 if sequencing prices hold at under $4 per gigabase.

To test MethylScan's accuracy, researchers analyzed blood samples from 1,061 people: patients with liver, lung, ovarian and stomach cancers; individuals with various liver diseases including hepatitis B, hepatitis C, alcohol-related liver disease, and metabolic-associated liver disease; people with benign lung nodules; and healthy controls. Machine learning algorithms processed the complex methylation data. The results were promising. At a specificity of 98 percent—meaning very few false alarms—the test detected about 63 percent of cancers across all stages and roughly 55 percent of early-stage cancers. For liver cancer surveillance among high-risk patients, including those with cirrhosis or hepatitis B virus infection, the test detected nearly 80 percent of cases with a false positive rate just over 10 percent.

Beyond simply flagging disease, MethylScan can pinpoint where in the body a signal originates. This matters because a positive blood test must be followed by imaging or other procedures directed at the right organ. Dr. Wenyuan Li, a co-corresponding author, notes that methylation patterns act like a health radar, allowing the test to detect when specific organs are under stress or damaged. The test also distinguished between different types of liver disease, correctly classifying about 85 percent of patients—a result that could reduce the need for invasive liver biopsies.

The work remains preliminary. Larger prospective trials will be needed before MethylScan can be deployed for real-world screening. But Zhou frames the research as a crucial step toward a single, affordable blood test that detects a broad spectrum of diseases earlier and more comprehensively than current methods allow. The vision is not incremental: a universal assay that could transform how medicine approaches prevention and early detection, replacing some invasive procedures and catching illness before it advances.