SAN FRANCISCO (The New York Times News Service) -- In a move that could alter the way that breast cancers are treated, researchers have redefined the disease into four main classes and determined that one type of breast cancer has more in common with an aggressive form of ovarian cancer than other breast cancers.

The finding that a form of breast cancer may be genetically similar to a type of ovarian cancer underscores a new way thinking about cancer that moves away from defining cancers by the organ of origin.

The findings are the result of the largest and most comprehensive study of the genetics of breast cancer to date and could offer new hope to cancer patients.

"We're going to move farther and farther from the practice of classifying cancers by where they arise and more and more by what their molecular composition and wiring is all about," said Dr. Christopher Benz, an oncologist at the University of California at San Francisco and co-principal investigator of a partnership between the Buck Institute for Research on Aging in Novato and UC Santa Cruz.

The new research, published online in the journal Nature on Sunday, is the fifth study to come out of the Cancer Genome Atlas, a project funded by the National Institutes of Health to examine the key genomic changes in at least 20 different cancer types.

The project's researchers have previously published reports on glioblastoma, an aggressive type of brain cancer, and on a form of ovarian cancer as well as colon cancer and, earlier this month, squamous cell lung cancer, a type of non-small cell lung cancer.

Breast cancer, the most common type among women, is responsible for 1.3 million new cases and 450,000 deaths annually worldwide.

The UCSF-Buck Institute collaboration is one of seven Genome Data Analyses Centers in the country -- the only one in California -- combining a collective 200 researchers charged with developing new roadmaps to understanding the disease that will hopefully lead to new ways to combat it.

This latest research on breast cancer involved the genomic analysis of tissue samples from 825 breast cancer patients and may change the way doctors have long viewed the many subtypes of the disease.

Benz said doctors have typically lumped breast cancers into three primary groups for treatment purposes:

-- Hormone receptor-positive disease, or tumors that have receptors for hormones on the surface of their cells, meaning those tumors can be treated with hormone-targeted therapy.

-- HER2-positive breast cancer, referring to the over-expression of the gene HER2, a type of protein that can also be targeted.

-- "Triple-negative" disease, which lacks the hormone receptors -- estrogen and progesterone -- as well as the HER2 genes that are targeted by some of the newest, most successful treatments.

The problem with those classifications, Benz explained, was that some patients with HER2-positive breast cancer tended not respond to HER2-targeted therapies. Also, some patients with triple-negative breast cancers responded to therapies while others didn't.

"We have been lumping things together that shouldn't be lumped together," Benz said.

The new classifications -- four types called HER2 "enriched," Luminal A, Luminal B and basal-like -- categorize breast cancers by their genomic structure using a dizzying array of data points not previously available. These data points have identified new pathways for the cancer to do its damage making it possible for researchers to identify new places to target the disease.

In what could be seen as the most promising development of this breast cancer study, researchers determined that some basal-like cancers had more in common with an aggressive form of ovarian cancer known as "serous" than with other types of breast cancer. Researchers suggested these cancers may potentially be targeted more effectively with therapies used in ovarian cancer, or with a combination of breast and ovarian cancer treatments.

Josh Stuart, a co-principal investigator and a UC Santa Cruz associate professor in biomolecular engineering, helped lead the effort to analyze the billions of data points, identify genetic mutations and then runs the sequences numerous times to make sure the anomalies weren't random.

Stuart said finding connections between seemingly unrelated tumors is going to prompt researchers to start making new connections across an array of cancer types. "It's a more holistic view of tumor samples than we've ever had before," he said.

The Buck Institute's Benz now has a particularly personal stake in his research: His wife, Connie, was diagnosed late last year with breast cancer, a form that he believes is Luminal A. She underwent a mastectomy, and current testing methods did not recommend chemotherapy.

Benz acknowledged changing the way clinicians view cancer will be challenging.

"It's sort of like charting the world when people thought the Earth was flat," Benz said. "You can go on new journeys if you have a better map."  

Copyright 2012 The New York Times News Service. All rights reserved.