Scientists Find 'Stem Cells' In Human Breast Cancer; Discovery Could Explain Current Treatment Failures And Lead To More Effective Therapies For Many Cancers

"These tumor-inducing cells have many of the properties of stem cells," says Michael Clarke, M.D., a U-M professor of internal medicine, who directed the study. "They make copies of themselves – a process called self-renewal – and produce all the other kinds of cells in the original tumor."

Although similar cells have been identified in human leukemia, these are the first to be found in solid tumors, Clarke adds. The cells were isolated from primary or metastatic breast cancers removed from nine women treated for cancer at the U-M's Cancer Center.

The discovery – reported this week in the online early edition of the Proceedings of the National Academy of Sciences – also may explain why current treatments for metastatic breast cancer often fail, says Max S. Wicha, M.D., an oncologist and director of the U-M Comprehensive Cancer Center.

"The goal of all our existing therapies has been to kill as many cells within the tumor as possible," Wicha says. "This study suggests that the current model may not be getting us anywhere, because we have been targeting the wrong cells with the wrong treatments. Instead, we need to develop drugs targeted at the tumor's stem cells. If we are to have any real cures in advanced breast cancer, it will be absolutely necessary to eliminate these cells.

"What this means for women with cancer is that, for the first time, we can define what we believe are the important cells – the cells which determine whether the cancer will come back or be cured," Wicha adds. "Before this, we didn't even know there were such cells."

All cancer cells have a unique pattern of proteins, similar to a fingerprint, on their surface membranes, explains Muhammad Al-Hajj, Ph.D., a U-M post-doctoral fellow and first author of the PNAS paper. " We used specific antibodies and flow cytometry technology to segregate the phenotypically heterogenous cancer cells within a tumor into isolated populations based on their surface protein markers," Al-Hajj says.

These isolated cell populations were then individually injected into immune-deficient mice and the mice were examined for tumor growth every week for up to six months.

Al-Hajj found a small group of cells, with a phenotype common to all but one of the human tumors in the study, could form new cancers in mice. These cells all expressed a protein marker called CD44, in addition to having either very low levels or no levels of a marker called CD24.

"As few as 100 to 200 of these tumor-inducing cells, isolated from eight of nine tumors in the study, easily formed tumors in mice, while tens of thousands of the other cancer cells from the original tumor failed to do so," Clarke says.

The fact that tumor-inducing stem cells from eight of nine women showed a common surface marker pattern is significant, Wicha explains. "Even though it's only nine patients, it shows that the markers identifying these stem cells were expressed in the majority of breast cancer patients in the study. This may not be the only expression pattern on every patient's stem cells, but it demonstrates the validity of the cancer stem cell model."

To test the stem cell's ability to regenerate the original tumor, U-M scientists repeated the experiment up to four times. First, 200 cells with the unique two-marker surface pattern were isolated from the original human tumor. When these cells produced a breast tumor in a mouse, Al-Hajj removed the mouse tumor and used flow cytometry to isolate 200 stem cells from it. These cells were then injected into another mouse to produce another tumor. Once again, the tumor was harvested, stem cells were separated, and injected into another mouse. Each procedure is called a passage.

"Tumor cells with this particular surface marker pattern produced a new tumor in the next generation of mouse every time," Clarke explains. "When we examined the tumors after each passage, we found their cell diversity to be the same as the original tumor."

Given that tumor-inducing cells now have been identified in breast and blood cancers, Wicha and Clarke believe it is likely that similar cells drive the development of other types of cancer, as well. The U-M Comprehensive Cancer Center is establishing a new research program to identify stem cells in other cancers and develop new therapies to destroy them.

"What we are working on now is finding out what makes these tumor stem cells different from the other cells in a tumor," Wicha says. "Now that we can actually identify them, we can start developing treatments to specifically target and hopefully eliminate them."

"This is not a cure for cancer," Clarke emphasizes. "But it is a very promising lead, which will focus our efforts to try to find a cure for cancer."

In addition to Al-Hajj, Wicha and Clarke, Sean J. Morrison, Ph.D., a Howard Hughes Medical Institute assistant investigator and U-M assistant professor of internal medicine, is a collaborator in the research study.

The U-M study was funded by the National Cancer Institute. The U-M has applied for a patent on the identity and function of tumor stem cells.