A new analytical technique could greatly improve the accuracy of breast cancer diagnosis, say scientists.
Breast cancer is a killer
British and Swedish researchers used the technique - called proteomics - to successfully classify breast tissue as healthy, benign or malignant.
They believe it could be a big improvement on current diagnosis - which relies on a microscope and the naked eye.
Proteomics analyses the protein molecules produced by our genes.
Often it is better to analyse proteins than genes, because while a complete set of genes may be present in a cell, only those that produce proteins will affect its behaviour.
The technique could also be used to identify new targets for anti-cancer drugs or to help doctors choose the most appropriate treatment for each patient.
Researchers from the University of Westminster and the Karolinska Institute in Sweden used proteomics to analyse 32 samples of breast tissue to see if a standard technique could be developed to pick out breast cancers from benign lumps.
They measured the levels of around 350 types of protein for each sample of breast tissue, in order to build up a molecular signature for each one.
By analysing the results with statistical software, they were able to separate breast cancer specimens from healthy and benign tissue.
Further analysis also enable them to separate out benign from healthy tissue.
The researchers hope that looking at much larger numbers of samples will allow them to refine the technique, so they can more precisely separate different kinds of tumour.
Lead researcher Dr Miriam Dwek said: "Proteins are 'doing' molecules and by measuring their levels directly, you can get a very accurate impression of what kinds of proteins the tissue is producing and therefore what kind of tissue it is.
"In the past we've only been able to look at a handful of proteins at the same time, but using proteomics, we can examine a tissue sample for hundreds, possibly even thousands, of proteins at once.
"The technique still needs to be further developed and validated in larger studies, but it has the potential to be rolled out to pathology labs throughout the country and to significantly improve cancer diagnosis."
Current diagnosis relies on pathologists studying samples of breast tissue under the microscope to look for abnormal cells.
It depends on the expertise and sharp eye of individual scientists, and only gives rather limited information about the nature of the tumour.
Knowing the profile of protein molecules in each breast sample may not only allow breast samples to be classified as healthy, benign or malignant, but could also help scientists to search for particular proteins specific only to cancerous tissue.
And in future, doctors could use the technique to examine the molecular profile of an individual patient's tumour - in order to make decisions about which treatments are most likely to work.
The research is published in the British Journal of Cancer.