Landmark study classifies breast cancer into 10 genetic types

BMJ 2012; 344 doi: (Published 18 April 2012) Cite this as: BMJ 2012;344:e2829
  1. Ingrid Torjesen
  1. 1London

Scientists have reclassified breast cancer tumours into 10 categories according to genetic features that correlate with survival, in the largest global study of breast cancer tissue ever performed.

It is hoped that eventually doctors will be able to use these subtypes to predict more accurately a woman’s likely survival and to better tailor treatment to individual patients. The first tests to identify which group a woman belongs to are expected to be available in the NHS in three to five years.

Scientists at Cancer Research UK in Cambridge and the BC Cancer Agency in Vancouver, British Columbia, analysed DNA and RNA samples taken from tumours in 2000 women who had been given a diagnosis of breast cancer between five and 10 years ago. They identified genetic markers, determined which genes were switched on and off, and ascertained numbers of copies of particular genes the chromosomes contained. By studying this genetic fingerprint they found that breast cancer could be classified into at least 10 subtypes grouped by common genetic features that correlate with survival.

Carlos Caldas, senior group leader at Cancer Research UK’s Cambridge Research Institute, said that the study, published in Nature today (doi:10.1038/nature10983), made breast cancer an “umbrella term” for an even greater number of diseases and would mean that in the future patients with breast cancer could hope to receive treatment targeted to the genetic fingerprint of their tumour.

He said: “It is not going to change the management of women treated in the NHS tomorrow, but it is surely going to change the way in which we do clinical trials. We will be doing clinical trials that are much more targeted at each of these 10 subtypes.

“And finally, because we have discovered new breast cancer genes, this will give us new avenues to develop targeted treatments.”

The most attractive targets identified were genes for enzymes because they have sites on their surface that can be targeted by drugs. These enzymes include kinases, which add a phosphate molecule to proteins to make the majority more active and deactivate a minority; phosphatases, which remove a phosphate molecule; and chromatin modifiers, which act on chromatin, the complex of protein and nucleic acids.

Before this study breast cancer was basically classified into four types according to whether the tumour was positive for receptors for oestrogen (Er+), human epidermal growth factor receptor 2 (Her2+), and progesterone (Pr+). Er+ tumours may respond to tamoxifen or aromatase inhibitors, whereas Her2+ tumours may respond to trastuzumab (Herceptin). Tumours that are negative for all these receptors are known as triple negative: they are very aggressive, the only treatment is chemotherapy, and there is no targeted therapy.

Seventy per cent of women have breast cancer that is Er+ and Her2−, and the researchers have now subdivided this group into seven subtypes. The likelihood that a woman with a high proliferation Er+ and Her2− tumour would be alive after 10 years is 75%, but when these tumours are subdivided into the seven new types the researchers found that survival varied from less than 40% to more than 80% after 15 years.

Of the other three new subtypes, one “robustly identifies” the Her2+ tumours, Caldas said. “All previous molecular tests have failed to do that properly.” The second contains most of the triple negative tumours, which have a very poor prognosis.

The third and final subtype contains a subset of Er+ tumours, a subset of Er− tumours, and a subset of triple negative tumours. What this group have in common is significant infiltration of inflammatory cells and lymphocytes. “This subset of breast cancer is a very important one for us to have recognised,” said Caldas, “because it looks like in this subset the immune system is playing a very active role in improving the prognosis of these women.”

Subsequent studies will try to understand why in this group the immune system seems to be particularly engaged and fighting against the tumour, giving the woman a better prognosis, Caldas predicted. “The fact that we know that breast cancer is not one disease but 10 different types of disease will allow us to go forward and look within these groups and see if we can detect differential responses.”


Cite this as: BMJ 2012;344:e2829

View Abstract

Sign in

Log in through your institution

Free trial

Register for a free trial to to receive unlimited access to all content on for 14 days.
Sign up for a free trial