Using AI to understand why some ER-positive breast cancers return after many years

Most breast cancers don’t come back after treatment. But some oestrogen receptor positive (ER-positive) cancers do – many years or even decades later. And without warning. Late relapses are devastating. When cancer comes back in other parts of the body, it can be treated, but it can’t be cured. And symptoms can be very similar to those of other conditions, so the link to the initial breast cancer from many years ago isn’t always recognised. That means diagnosis is delayed, and precious time is lost.

Some breast cancer cells can quietly survive in the body for decades. They stay inactive, or dormant, hiding from the immune system and escaping treatments. Until one day they wake up and start growing again.

Currently, we don’t know who’s most at risk of late relapse. Or how to stop these inactive breast cancer cells before they become dangerous.

That’s why we’re researching these questions now. So we’re not caught off guard by recurrence. And so no one dies of breast cancer in the future.

Maria and her team are working in collaboration with Professor Luca Magnani’s lab at the Breast Cancer Now Toby Robins Research Centre at The Institute of Cancer Research, London. They want to understand how ER-positive breast cancer cells enter, stay in, and exist in the inactive, or dormant, state. And they want to know what molecular features these dormant cells have, how they interact with their surroundings and what triggers them to wake up again.

To answer these questions, they’re combining their unique expertise in biology and AI to study existing data - data showing what genes are active in individual cells and where they are within tissues.

The researchers aim to map molecular changes that allow breast cancer cells to become inactive. They’re developing advanced AI models trained on large amounts of data from cancer cells as they undergo treatment. These models can learn the patterns and “rules” that cells follow – including how they switch into a dormant state and what causes them to become active again. By recognising these patterns across millions of cells, the team can begin to predict how and when cancer might return.

These cancer cells don’t exist in isolation. They’re influenced by their neighbouring immune cells, blood vessels and connective tissue. All of which can play a role in supporting these cells to stay in hibernation, or waking them up. And so the team is also studying how the surroundings can help or hinder these inactive breast cancer cells.

Using what they’ve learned, they want to identify when the dormant cancer cells are likely to wake up, so they can predict which breast cancers may come back.

Living with the fear of breast cancer returning can be all-consuming.

We need to know if it’s likely to happen, so that we can stop cancer cells in their tracks. And so people can face the future with confidence after breast cancer treatment

This project can help us get there. It can uncover the signs and conditions that lead to late relapse. Helping to develop tools that tell people their risk of ER-positive breast cancer coming back many years later.

It can also bring new knowledge and open the door to new treatments. Treatments that can stop inactive breast cancer cells from waking up, or get rid of them before they become dangerous. Meaning fewer lives are lost to breast cancer.

Around 80% of all breast cancers are ER-positive. That’s 44,000 people diagnosed in the UK every year.

The findings from this project may also help us understand other types of cancer, such as prostate, ovarian, and some types of leukaemia, that can also come back many years later.

We would like to thank Art For Cure for their generous support of this project.