New research in mice and five independent collections of human breast tumours has enabled National Cancer Institute (NCI) scientists to confirm that genes for factors contributing to susceptibility for breast cancer metastasis can be inherited. The new findings support earlier results from the same laboratory and appear in the January 1, 2009, issue of 'Cancer Research'.

The study results also show that gene activities in tumour cells and immune cells that infiltrate, or invade, tumours can contribute to the development of expression profiles, called gene signatures that are predictive of cancer progression. The analysis of normal mouse tissue as well as tumours transplanted into mice suggests that predictive, or prognostic, gene signatures that point to a tumour's potential for spreading throughout the body can be the result of both inherited and non-inherited factors, with inherited factors being more consistently predictive. The research team that reported these findings is from the Center for Cancer Research at NCI, which is part of the National Institutes of Health.

The researchers were able to perform their analyses by using advances in microarray technology, which allows scientists to scan vast amounts of genetic information and identify gene signatures that can be used to predict cancer outcomes. Many scientists had assumed that metastatic ability is primarily determined by somatic, or non-inherited, gene mutations in tumour tissue. "Our earlier studies clearly established that inherited factors also play an important role in metastatic progression and can help distinguish which tumours have a propensity to metastasize," said author Kent W Hunter, head of NCI's Metastasis Susceptibility Section in the Laboratory of Cancer Biology and Genetics. "Hopefully in the future we will be able to determine which women are more likely to have a tumour that would metastasize, and we could then tailor therapy specifically for them, avoiding the use of harsh treatments for those with a low probability of metastasis."

To determine whether mouse tumour gene expression profiles could be used to predict outcomes in human breast cancer, the investigators identified a gene expression signature that allowed them to distinguish between the tumours of mice that have a high or a low inherited susceptibility to tumour metastasis (a 20-fold difference). They then converted the mouse gene signature to the corresponding human gene signature and analyzed five pre-existing sets of human breast tumours. This signature successfully predicted outcomes (either relapse or disease-free survival) in four of the five sets of human breast tumours.

Because other studies have suggested that gene expression patterns in the nearby tissue, or stroma, are altered in tumours that are prone to metastasis, the investigators conducted transplant experiments by putting highly metastatic tumour cells into the mammary fat pads of mice that have either a high or a low susceptibility to tumour metastasis. These transplants resulted in tumours that had identical tumour cells but different stroma and immune cells that infiltrated the tumour. No significant differences were seen in tumour weight or metastasis to the lung in the two types of mice after 28 days, suggesting that metastatic differences between individual mice in this experiment were possibly due to genes in the outer layer of tissue that surrounds the tumour (the epithelium) rather than in the stroma. However, differences in gene signatures were still seen in mice with either high or low potential to develop metastases, and the corresponding human gene signatures were predictive of relapse or survival in patients. The researchers concluded that both the tumour epithelium and the stroma probably contributed to the development of the prognostic gene profiles.

"Our study provides additional evidence of the role of inherited genes in human breast cancer progression. Therefore our next step is to improve our current understanding of the role of the epithelium and stroma in tumour progression and develop more effective therapeutic strategies based on our new knowledge," said Hunter.