Diagnosis of cancer from a milliliter of blood

Sep 10, 2022

(News from Nanowerk) Blood tests – simple, non-invasive and economically feasible – promise to become the next major step in cancer diagnosis. However, most of these tests, called liquid biopsies, are currently not reliable enough for widespread use. A new multiparameter approach developed at the Weizmann Institute of Science could lead to a blood test that will diagnose cancer with unprecedented accuracy.

This research is published in Natural biotechnology (“Single-molecule multiplexed epigenetic analysis of isolated nucleosomes in plasma for the diagnosis of cancer”).

Different patterns of epigenetic markers revealed by EPINUC on blood nucleosomes (bright red dots) of a healthy person (left) and a patient with colorectal cancer (right). (Image: Weizmann Institute of Science)

“Most of the conventional methods currently available in the clinic for detecting and diagnosing cancer are invasive and inconvenient,” says Dr. Efrat Shema of Weizmann’s Department of Immunology and Regenerative Biology, who led the research team.

Obtaining biopsy specimens by needle, endoscopy, or surgery can be painful and sometimes risky, and imaging methods, such as MRI or PET, require expensive and bulky equipment that is not universally available. . Effective blood tests for screening or diagnosing cancer could be an interesting alternative.

“The elimination of discomfort means that people would be less likely to avoid getting tested – and more likely to have their cancer detected earlier,” says Vadim Fedyuk, who led the study with fellow graduate student Nir Erez.

The idea of ​​diagnosing cancer using liquid biopsies arose from the fact that blood contains floating DNA and proteins released from dead blood cells in healthy people – and in cancer patients, for example. dead tumor cells as well. “Some of the byproducts of cell destruction, including cancer DNA and proteins, are dumped into the bloodstream, and we know how to collect and analyze them,” Shema says.

A number of blood tests for cancer are already in advanced stages of development, but most have drawbacks that can limit their use. When the first tests of this type were developed, they looked for genetic signs of cancer, i.e. mutations, but these could be difficult to pin down because the mutated segments represent only a small fraction of DNA in free circulation.

Moreover, these mutations do not always lead to cancer and can also be present in healthy people. More recently, liquid biopsy approaches have begun to rely on epigenetics, or changes to the cell’s genome that do not involve mutations in the DNA – for example, chemical tags that attach to the DNA molecule, altering gene expression.

These approaches have also run into problems, either because they require excessive amounts of blood or because they look for a single epigenetic change that cannot yield sufficiently reliable results.

In the new study, Shema set out to rethink this epigenetic analysis, aiming to develop one that would rely on a small blood sample to assess multiple epigenetic parameters. She relied on a method of imaging individual molecules that she had developed during her postdoctoral research at Harvard Medical School and the Broad Institute.

The method allows accurate epigenetic mapping with only a very small amount of raw material, using a fluorescence microscope. It can be used, for example, to visualize epigenetic marks on nucleosomes, pieces of DNA wrapped around “coils” of proteins. These can be thrown back into the bloodstream like pieces of shipwreck when the cells are destroyed, so Shema reasoned that the millions of nucleosomes found in the blood could be analyzed for cancer.

Using Shema’s single-molecule imaging method, Fedyuk and Erez, along with their colleagues, compared nucleosomes in the blood of 30 healthy people with those of 60 colorectal cancer patients at different stages. They found that the nucleosomes of the two groups were characterized by very different epigenetic marking patterns. This analysis covered six different epigenetic changes linked to cancer, as well as a variety of other indicators of cancer, including protein segments from dead tumors, which are undetectable by conventional technologies.

Then, working with Professor Guy Ron of the Hebrew University of Jerusalem’s Racah Institute of Physics, the scientists combined what they had revealed about the molecular biology of cancer with artificial intelligence algorithms, applying the machine learning to large datasets obtained from both groups. The analysis was performed not only on all these cancer markers, but also on combinations and relationships between them. To ensure that their findings are not limited to colorectal cancer, the scientists also applied their technology to compare blood nucleosomes from healthy volunteers with those from 10 patients with pancreatic cancer.

“Our algorithm could differentiate between healthy groups and patient groups at an all-time high of certainty for studies of this type – with 92% accuracy,” Shema says. The scientists call the new technology EPINUC, an acronym for “epigenetics of isolated nucleosomes in plasma.”

If supported by studies involving larger numbers of patients, these findings could lead to a multiparameter blood test to detect and diagnose cancer using less than 1ml of blood. In the future, due to the level of detail revealed in the analysis, the results of this blood test could also advance personalized medicine by suggesting the best treatments for each patient.

Shema summarizes: “We have successfully completed a proof of concept for our method, which now needs to be confirmed by clinical trials. In the future, our multiparameter approach could be used to diagnose not only various cancers, but also other diseases that leave traces in the blood, such as autoimmune diseases or heart disease.