New tool predicts women's outcome in breast cancer
WASHINGTON (Reuters) - Evaluating how various proteins interact in tumors can help predict a woman's chances of surviving breast cancer, allowing doctors to better tailor treatment, Canadian researchers said on Sunday.
Knowing from the outset that a particular woman's prognosis is bad could allow doctors to give her aggressive treatment right away, but often it is difficult to know which breast cancer patients will do well and which will not.
The researchers analyzed networks of proteins -- chemical compounds vital in cellular processes -- in breast cancer tissue from about 350 women in the United States and Europe.
They found that women who survived the disease had a different organization of the network of proteins within the cancer cells than those who died.
Writing in the journal Nature Biotechnology, they said tracking these protein interactions enabled them to accurately predict in 82 percent of patients whether their breast cancer would kill them or not.
"We approached cancer as a problem in how proteins communicate with each other -- or how proteins interact with each other in networks," Jeff Wrana of Mount Sinai Hospital in Toronto, who led the study, said in a telephone interview.
"It could help to direct the appropriate therapies for individual patients."
The researchers observed 30,000 protein interactions involving about 8,000 proteins, then identified a core group of about 250 proteins most important in forecasting patient survival. Many of them regulate the actions of other proteins.
If a newly diagnosed patient has protein interactions that suggest a bad outcome, a doctor could give more aggressive treatment through surgery, chemotherapy and radiation.
Mount Sinai Hospital has a patent on the process and the researchers have formed a Toronto-based company called DyNeMo Biosystems to explore commercial applications.
Breast cancer is the leading cause of cancer death among women worldwide, with about 465,000 dying annually.
(Editing by John O'Callaghan)