Weizmann Institute and Hebrew University researchers decode the epigenetic signatures of breast tumours, including those linked to hidden populations of aggressive cells.

One of the great hopes of cancer medicine over the past decade has been a simple blood test that would detect cancer at its earliest stages, help physicians choose the right treatment and even assess the risk of recurrence. Known as liquid biopsies, these tests work by identifying traces of cancer cells in the bloodstream. Although they are already used to tailor treatment for some patients, they still struggle to detect cancer early in the general population or to predict whether the disease will return after treatment. One reason is that, in the early stages, genetic traces of cancer in the blood are extremely scarce and hard to detect without triggering false alarms – results suggesting cancer when none is present.

Now researchers from the laboratories of Professor Efrat Shema at the Weizmann Institute of Science and Professor Guy Ron at the Hebrew University of Jerusalem have developed a blood test that is able to distinguish with a high degree of accuracy between women with early-stage breast cancer and healthy women. The test also outperforms conventional tissue biopsies in assessing the risk of recurrence after treatment.

The findings, published today in Nature, grew out of a collaboration with physicians at Kaplan Medical Center in Rehovot. Together, the researchers decoded the distinctive epigenetic signatures of breast tumours and identified a particularly aggressive population of cells lurking within them.

If DNA is the cookbook that every cell in the body keeps on its shelf, epigenetic tags are the sticky notes and bookmarks telling the cell which recipes to follow and how often to use them. In cancer cells, things can go wrong both with the recipes themselves, through genetic mutations, and with the instructions for using them. Although breast tumours are already classified and treated according to their patterns of gene expression, the epigenetic tags that help control these patterns have yet to be fully mapped. In the new study, Dr Noa Furth and Dr Christina Riemenschneider of Shema’s lab, together with Yishai Arieli from Ron’s group, set out to decipher the epigenetic signatures of breast tumours.

“We recruited 12 women who were scheduled to have their tumours surgically removed,” Furth recalls. “Once a tumour is removed, there’s only a short window of opportunity before its epigenetic tags begin to deteriorate. So as each surgery began, the doctors would alert me. I would rush the tissue sample from the hospital to our lab on ice, and we’d start analysing it that very day.”

Evidence is mounting that cancer cells exploit their epigenetic flexibility to change their behaviour, become more aggressive and evade treatment

Having mapped the epigenetic signatures of breast tumours from several women, the scientists looked for a way to detect those signatures in blood samples, to enable early diagnosis. They built on a technology developed by Shema that takes advantage of a natural process: When cells die, they release nucleosomes – segments of DNA wrapped around proteins – into the bloodstream. Cancer cells are no exception.

Using single-molecule microscopy and fluorescent labels that recognize dozens of different epigenetic tags, the technology can map epigenetic changes in millions of nucleosomes. In fact, in a 2022 study, it had already shown promise for detecting early-stage colorectal and pancreatic cancers through a blood test.

Importantly, collecting all these molecular measurements was only half the challenge. The researchers also needed to find the tell-tale patterns hidden in this wealth of data. That’s where Shema’s expertise in epigenetics and single-molecule technologies met the quantitative approach of Ron’s physics group. The latter tackled the challenge as an experimental physics problem, using computational analysis to identify patterns that could help detect cancer and assess the risk of its return.

The scientists tested their new method on blood samples from 77 breast cancer patients who had been followed for at least five years at the Mayo Clinic in Florida, along with samples from 30 healthy women. They trained artificial intelligence models to distinguish between the two groups, achieving 95 percent accuracy in detecting early-stage breast cancer, when the tumour was still confined to the breast and measured no more than two centimetres.

“Breast cancer screening saves lives,” Shema says. “Regular mammograms reduce the risk of dying from the disease by 22 percent, but the denser the breast tissue, the easier it is for a tumour to ‘hide’ in the image. Our blood test showed strong diagnostic performance regardless of breast density. It performed just as well across different types of breast cancer, and also when we expanded the number of patient blood samples to 188 with the help of Sheba and Kaplan medical centres.”

Taming the aggressive cells
Even when breast cancer is caught early and treatment appears successful, the disease returns in 20 to 40 percent of cases. One possible reason is that not all tumour cells are alike. A treatment may wipe out most of them while leaving behind a handful capable of causing the cancer to return.

During the study, the researchers discovered that even breast tumours not considered particularly aggressive can harbor a rare population of highly aggressive, invasive cells with a distinct epigenetic signature. This raised a fundamental question: How can cells with such different characteristics coexist within the same tumour? Do the aggressive cells carry genetic mutations that set them apart from their neighbours, or can ordinary tumour cells acquire aggressive traits through epigenetic changes alone?

“The differences appear to be mainly epigenetic, when we grew breast cancer models in the lab starting with nonaggressive cells, an aggressive population always emerged eventually,” Shema explained.

The scientists suspected that these aggressive cells might be among the main culprits behind cancer recurrence.

“When we looked for their epigenetic signature in blood samples, we found that the more prevalent it was, the greater the risk that the disease would come back,” Ron said.

“Using this signature and additional epigenetic markers, we were able to distinguish with 81 percent accuracy between patients whose cancer returned during follow-up and those who didn’t experience a recurrence. The test even flagged a risk of recurrence in patients considered low risk by conventional measures.”

Shema added: “Evidence is mounting that cancer cells exploit their epigenetic flexibility to change their behaviour, become more aggressive and evade treatment.

“The methods we’ve developed allow us to track aggressive cell populations in different types of cancer. In the future, they could help us not only predict recurrence but also search for ways to prevent it. We’re now looking for treatments that could deprive cancer cells of their epigenetic flexibility – and perhaps even coax cells that have already turned aggressive into adopting less aggressive traits,” she said.

Also participating in the study were Matan Avni, Michael Yakov Povolotski, Hadas Yung and Dr Nir Erez of Weizmann’s Immunology and Regenerative Biology Department; Shiri Karagach, Ofir Atrakchi and Professor Tamar Geiger of Weizmann’s Molecular Cell Biology Department; Dr Tomer-Meir Salame, Dr Bareket Dassa and Ofra Golani of Weizmann’s Life Sciences Core Facilities Department; Dr Roberta Noberini and Professor Tiziana Bonaldi of the European Institute of Oncology in Milan, Italy; Adeline Durand, Grégoire Jouault and Dr Celine Vallot of PSL University in Paris, France; Mirie Zerbib and Dr Roni Oren of Weizmann’s Veterinary Resources Department; Dr Barak Bar-Zakai, Dr Yael Mavor, Dr Orli Friedman-Eldar, Dr Hadas Kadar-Sfarad, Dr Shani Gilron, Dr Osnat Givon-Madhala, Dr Naama Ronel, Dr Faisal Abu Rmilah, Dr Itai Moshe, Dr Shahar Ish-Shalom and Dr Ady Yosepovich of Kaplan Medical Center in Rehovot, Israel; Dr Miriam Kaplan, Dr Michal Bakalenik, Dr Maya Dadiani and Dr Einav Nili Gal-Yam of Sheba Medical Center in Tel Hashomer, Israel; Rakel Tryggvadóttir, Gabriel Shatkin and Professor Andrew P. Feinberg of Johns Hopkins University in Baltimore, Maryland; Dr Mariam Hakobyan, Alexander Kovacsovics, Dr Sebastian D. Mackowiak and Professor Alexander Meissner of the Max Planck Institute for Molecular Genetics in Berlin, Germany; Amanda Arnold and Professor Sarah McLaughlin of the Mayo Clinic in Jacksonville, Florida; and Dr Ekaterina Andreishcheva, Abhijeet Shinde, Dr Jiaren Zhang and Dr Daniel Jones of Epivision Genetics in Boston, Massachusetts.