The reason why many women with breast cancer see the disease return after apparently successful treatment has been discovered in a breakthrough that could lead to improved treatments.
The new insight, which challenges current understanding of cancer, could pave the way to more effective strategies to fight the disease, which affects around 45,000 women annually in Britain alone.
Breast cancer survivors have a substantial risk of recurrence: of the women who were cancer-free at five years, one fifth or so can develop the disease during the following decade.
The new results published in Science by Dr Katrina Podsypanina and colleagues at Memorial Sloan-Kettering Cancer Centre in New York, suggest that disease spread might also arise from the relatively normal cells that have spread and remain dormant at the new organ site until cancer genes are switched on.
In this way, the spread of cancer to new sites in the body - the process that is ultimately responsible for most cancer deaths - may happen earlier because it seems to be linked to normal cells migrating around the body and then mutating into a cancerous form.
This discovery suggests an explanation for why some breast cancers, for example, appear to spread throughout the body long after the initial tumour has been treated.
The team says that the next generation of treatments should target more normal breast cells as well as those that are cancerous to cut the risk of relapse.
"If follow-up research suggested that dissemination of normal breast cells can account for metastatic relapse in breast cancer patients (a hypothetical situation, at present), we would argue that treatment strategies should aim to compromise viability and/or proliferation of normal breast cells, and not just breast tumour cells."
One approach, adds Dr Podsypanina, could be to use drugs with a similar hormone action to tamoxifen, though she stressed more work needs to be done first.
Because the spread of cancer cells involves a series of steps to change the makeup of normal cells - the cells must be equipped to survive the trip in the bloodstream and initiate malignant growth in their new environment - researchers have traditionally considered it to be a late event in cancer development that occurs after tumour cells in the primary site, where cancer first develops, have racked up a series of genetic alterations that switch on cancer genes.
To investigate, team injected mice with normal mammary cells from donor mice that had been experimentally manipulated in a way that allowed the researchers to turn on in breast cells certain cancer genes, known as "oncogenes," at various times after injection.
The team, including the Nobel prizewinner Prof Harold Varmus found that the normal mammary cells were capable of traveling in the bloodstream to the lungs and surviving there for up to 16 weeks without using any oncogenes.
The donor cells did not begin growing aggressively in the lungs until the oncogenes had been turned on, by giving the recipient mice a specially treated feed.
Liz Baker, senior science information officer at Cancer Research UK, said: "Learning more about the spread of cancer - or metastasis - is essential because it is harder to treat the disease once it has spread. These are important but early results in mice - it will be interesting to see whether this can one day help the outcome of people affected by cancer."



