This finding, published recently in the Journal of Clinical Oncology, changes the current definition of these tumours, guaranteeing improved diagnosis by minimizing classification errors and hence improving the therapeutic approach and prognosis of patients with this breast cancer subtype.

Gene expression: the "digital fingerprint" of tumours

In any breast cancer case, it is essential to establish its molecular subtype. In recent years, gene expression has identified two major groups of hormone-sensitive tumours known as Luminal A and Luminal B. Treatment is based on this classification and therefore needs to be extremely precise. "Luminal A tumours have a good prognosis with hormone treatment, while Luminal B tumours have a less optimistic prognosis and in most cases require chemotherapy. The problem lies in the fact that the current method for defining these groups by histopathological data is not perfect, and we therefore need an additional parameter to help us yet further fine-tune the histopathological diagnosis of Luminal A and B tumours," explains Dr Aleix Prat, leader of the study.

At present, the way of defining luminal or hormone-sensitive tumours is by testing positive to hormone receptors and negative to HER2 expression. The difference between a Luminal A and Luminal B tumour is the amount of Ki-67 protein detected by the pathologist in the tumour cells. If the tumour has fewer than 14% Ki-67-positive cells, it is classified as Luminal A. The problem with this definition of subtype Luminal A is that ~30% of these tumours are actually shown as Luminal B by gene expression and hence have an unfavourable prognosis with hormone treatment alone.

A study offering diagnostic improvements

This study checked thousands of tumours from three independent groups from Spain (GEICAM cooperative group, Dr. Martín), Canada (University of British Columbia, Dr. Nielsen) and USA (University of North Carolina, Dr. Perou), to compare whether histopathological data matches the profiles of gene expression, and detected some major discrepancies. Once these differences had been identified, the team sought a new parameter that could be measured in the laboratory to fine-tune the definition. "The genomic study we conducted has shown that the amount of progesterone receptor is important in distinguishing the two biological entities. We can use this new variable as a marker to improve the current histopathological definition of Luminal A tumours," explains Dr Prat. The higher the number of tumour cells that are progesterone-receptor positive, the more likely it is that the diagnosis will be a Luminal A tumour.

The new histopathological definition of Luminal A tumours proposed by this study is "hormone receptor positive, negative HER2 expression, Ki-67-positive lower than 14% and progesterone-receptor positive higher than 20%," specifies Dr Prat.

Gene expression remains key

The method that unquestionably provides the most accurate and comprehensive information on each tumour are gene expression techniques. "Compared with histopathological data, tests based on gene expression techniques provide us with a more accurate molecular profile of each tumour and a more detailed estimate of the probability of relapse. The reason for this is that instead of determining three or four markers, we are assessing thousands of markers in one go. Having said that, these genomic techniques are not readily available in routine clinical practice," says Dr Aleix Prat, who was recently awarded the International Award for Breast Cancer Research for his work in bringing gene expression techniques closer to routine clinical practice. "In this study we evaluated a genomic test known as PAM50 which, in contrast to the other tests available, allow us to better characterise breast cancer at a biological level," continues Dr Prat. The VHIO is hoping to make this test available in 2013.