Breast cancers contain many different cell types with different patterns of gene expression, but a new study provides reassurance that this variability should not be a barrier to using gene expression tests to help tailor cancer treatments to individual patients.

The findings were reported at the 5th IMPAKT Breast Cancer Conference in Brussels, Belgium. The IMPAKT meeting presents cutting edge, 'translational' breast cancer research that is beginning to have an impact for patients.

In recent years it has become clear that breast cancers contain a variety of different cell types. An important result of this heterogeneity is that different biopsy specimens from a single breast cancer tumour can exhibit significant variability in genes expression.

This is a major concern for doctors seeking to understand which patients are likely to benefit from drugs designed to be effective against tumour cells with particular genetic characteristics. A number of studies at this year's IMPAKT conference consider this issue.

In one study, Dr Michał Jarząb and colleagues from the Maria Skłodowska-Curie Memorial Cancer Center and Institute of Oncology, Poland took a total of 78 different biopsies from 26 individual tumours to assess the degree of genomic variation, and its impact on a set of 32 different prognostic and predictive multi-gene signatures.

"Some genomic tests have proven very useful in breast cancer, but in other important areas we have not achieved optimal results," Dr Jarząb explains. "One of these areas where we haven't done so well is in deciding whether a particular patient would benefit from certain type of chemotherapy or not, based on the material from pre-surgical needle biopsies. We hypothesized that some genomic tests may be prone to the heterogeneity of starting material and provide not reliable results."

The researchers performed gene expression profiling on their 78 samples using oligonucleotide microarrays. Overall, they found that the gene expression profiles of the cores were variable, and in at least 5 patients this heterogeneity was substantial.

However, when they analysed a number of multi-gene signatures selected from previous studies, this heterogeneity was considerably less significant.

The gene sets differed in their variance between biopsies, the authors found. The most pronounced heterogeneity was observed in immune response-related genes, while the least heterogeneous were the classifiers based on genes selected by advanced bioinformatical methods from both cell culture experiments and patient tissues.

"Overall, the heterogeneity among the potentially predictive genes was small enough and we conclude that this factor should not prohibit their effective use in clinical practice," Dr Jarząb says.

"Our study confirms that it is possible to address tumour heterogeneity when carrying out routine diagnostic procedures in patients. Our results may help to introduce the better tailoring of preoperative treatment."

Commenting on the results, Dr. Angelo Di Leo, Head of the Sandro Pitigliani Medical Oncology Unit and Chair of the Oncology Department at the Hospital of Prato, Istituto Toscano Tumori, Italy, said: "If other studies report similar results to this one, then it could become common practice to evaluate prognostic or predictive breast cancer markers from more than one primary tumour area."

"The study is innovative because it is one of the first to address the question of intratumour heterogeneity. It means that not all the cells from the same tumour have the same characteristics, and if we want to have a clear picture of the tumour biology we should not limit the evaluation of tumour markers to one area of the tumour itself," Dr. Di Leo said.