The study, led by scientists at the Cancer Research UK Cambridge Institute and published in Nature, examined the patterns of genetic changes within tumours from nearly 2,000 women with breast cancer and followed their progress over 20 years, including whether their cancer returned.

The researchers looked for patterns that could be used to classify breast cancers with greater precision, by combining IntClust (a classification of 10 different subgroups with distinct molecular profiles) with factors known to influence breast cancer survival such as age, tumour grade and size and the number of lymph nodes affected. They then used this information to create a statistical tool to predict if, and when a women’s breast cancer could come back.

The new model produced results that matched those given by the established NHS tool PREDICT and provided further information on patient prognosis compared to existing methods. The study also defined a new subgroup of ‘triple negative' breast cancers that rarely come back after five years, as well as a subgroup that remains at risk of relapse – a finding that the authors hope could in future be important in treatment planning and disease monitoring.

Dr Simon Vincent, Director of Research at Breast Cancer Now, said:

It’s really promising that this molecular analysis could help us better predict the likelihood of a patient’s breast cancer returning and spreading, and how this risk may change over time. But much more research will be needed to develop this approach into a useful test that could be made available to patients in the clinic.

While more women are surviving the disease than ever before, some breast cancers can unfortunately come back many years later. The relapse and spread of the disease remains one of the greatest challenges facing researchers and clinicians, and we urgently need to find ways to identify who is most likely to see their breast cancer return.

These exciting findings could offer another step forward in the way we classify and treat breast cancers. We hope these 11 molecular subtypes could in future help us better predict long-term outcomes and lead to more personalised treatment plans for patients based on the properties of their tumour. For example, if we could identify patients with hormone-positive breast cancer at greater risk of recurrence twenty years after their treatment, we may be able to offer them prolonged hormone therapy or other long-term interventions to reduce this risk.

Crucially, this major study could help us tease apart the diverse set of hard-to-treat breast cancers currently grouped together as ‘triple negative’, and identify certain patients that are unlikely to experience a recurrence more than five years after their treatment. 

We now look forward to further studies to better understand these subgroups and identify how in-depth molecular analysis like this might be used in the clinic. In the meantime, we’d encourage any breast cancer patients who are concerned about their risk of recurrence or about any unusual changes to their health to speak to their GP.