A new blood test may detect precursors to liver disease, which may also be a precursor to cancer. It is hoped that this test will be able to prevent liver cancer before it occurs.
This test uses machine learning models to analyze airborne genetic material circulating in the blood. In the new study, researchers used this to detect DNA fragments that indicate early stages of liver scarring, or fibrosis. If left untreated, this early scarring can develop into severe liver scarring called cirrhosis, which can eventually lead to cancer.
you may like
“The best way to intervene in liver cancer is not to detect liver cancer early, but to detect liver disease early,” Belculescu told Live Science.
Once fibrosis is detected, it can be reversed through anti-fibrotic medication, lifestyle changes and other treatments, he added. In contrast, cirrhosis is largely irreversible.
Signs of disease in the blood
Millions of Americans have liver fibrosis and don’t know it. Risk factors for this scarring include liver inflammation (hepatitis), diabetes, high blood pressure, and obesity. If detected early, liver fibrosis is reversible.
But currently, traditional clinical evaluations such as the Fibrosis-4 (FIB-4) blood test, which uses age, liver enzymes and platelet blood counts to estimate the level of liver scarring, are unable to detect early-stage liver disease, Belculescu said.
We’re trying to find changes that can occur in diseases that can occur across the genome.
Akshaya Annapragada, MD/PhD student at Johns Hopkins University Kimmel Cancer Center;
In the new study, published March 4 in the journal Science Translational Medicine, Velculescu and colleagues first examined blood samples from 423 people with and without liver disease. By analyzing tens of millions of cell-free DNA fragments in the blood, they discovered markers that can distinguish patients with early liver scarring from those with milder liver disease.
Cell-free DNA, also called free-floating DNA, contains small pieces of genetic material that are released into the blood when cells reproduce and die. Instead of looking for specific mutations or changes in DNA letters, the researchers used a computer model that looked for larger genome-wide patterns across all the free-floating DNA released by cells.
“We’re trying to find changes that can occur in diseases that can occur across the entire genome,” Akshaya Annapragada, lead author of the study and an MD/PhD student in the Belculescu lab, told Live Science. “So you have a better chance of finding something.”
What to read next
They identified several factors that were collectively associated with early liver disease. These include the length of DNA fragments and the frequency with which cells release repetitive sequences of DNA. They also discovered important epigenetic changes, or marks on the genome, that alter gene activity without changing the underlying DNA code.
Using these elements, they built a test that looks for these patterns in the blood.
The team then evaluated the blood tests in 221 additional participants to assess how well they worked. Of these, 30 had early liver disease, 85 had advanced liver disease, and 106 had no liver disease. The test detected 50% of early liver disease cases and about 78% of advanced cases.
It correctly identified people without the disease in 83% of cases. This means it falsely flags liver disease 17 out of 100 times.
By using machine learning to identify patterns across the genome, the new study allowed the team to analyze billions of fragments at once, said Alain Thierry, professor and research director at INSERM, France’s National Institute for Health and Medical Research, who was not involved in the study.
This is an advantage over previous blood tests that required sequencing the genome thousands of times to obtain enough DNA to search and interpret only specific mutations or disease markers, Annapragada said. In contrast, “this test is much cheaper and more efficient because it only sequences the genome once or twice.”
The next step is a large-scale clinical trial to validate machine learning models that can detect liver fibrosis, Belculescu said.
Researchers said they hope that tests like this one will eventually pave the way for non-invasive ways to screen for many diseases with a single blood test, allowing for early diagnosis and treatment before diseases become chronic and irreversible.
This article is for informational purposes only and does not provide medical advice.
Akshaya V. Annapragada et al., Cell-free DNA fragmentome for non-invasive detection of liver cirrhosis and other diseases. Science Translational Medicine.18, eadw2603(2026). DOI:10.1126/scitranslmed.adw2603
Source link
