The problem with modeling metastatic cancers: Is it the mouse’s fault?

While the mouse is widely used to model many diseases, including cancer, the results we gain from it rarely translate well to people. But whose fault is it? (Wednesday Elf - Mountainside Crochet/Flickr)

The humble house mouse (or Mus musculus) is probably the most widely used model animal in biomedical research (beating out my favorite, the zebrafish, by a long mile). Millions are studied around the world every year, helping us understand the genetics of health and disease as well as the biology of cancer, diabetes and a host of other conditions. Mouse modeling is also often a major step in developing and getting FDA approval for new drugs.

But the mouse sometimes gets a bad rap in the research world. While it can be an effective and affordable model, and 95 percent of its genes are similar to ours, it is less than ideal for some of the diseases we study with it.

Take cancer, for instance. It’s relatively easy to cure cancer in a mouse; we’ve done it millions of times over. (The late Judah Folkman, MD, founding father of Boston Children’s Hospital’s Vascular Biology Program (VBP) and of the field of angiogenic research, famously said, “If you’re a mouse and you have cancer, we can take good care of you.”) Mouse and human tumor cells are fundamentally different in many ways. And the way that tumors behave in mouse models doesn’t necessarily reflect the way they behave in their natural environment (that is, in us)—a major consideration, especially when it comes to looking for new treatments for cancer that has spread (aka metastasized). More often than not, drugs that are successful in mouse models fail in the clinic.

But is it the mouse’s fault? Or is the problem the way we develop our models and run our experiments?

“What we typically do is implant a tumor in a lab mouse’s back, and start treating the mouse that same day with the drug that we want to study,” the VBP’s Bruce Zetter, PhD, explained recently at the 4th Judah Folkman Lecture, an annual event commemorating Folkman’s life and work. “If the drug works, we take the results to the clinic and give the drug to very sick patients with late-stage, metastatic disease. And most of the time, it doesn’t work in those patients.”

“Essentially we put the tumor in an environment it isn’t normally found in, don’t take the time for it to establish itself, and don’t give it a chance to metastasize,” continued Zetter, who has spent the last 30 years working on the problem of cancer metastasis. ” The model doesn’t reflect the clinical realities of metastatic disease.”

“We can get better answers from the mouse. We just have to ask better questions.”

To really make headway against cancer, we need to tackle metastatic tumors. As Zetter noted in his talk, they cause 90 percent of cancer deaths, but 90 percent of therapeutic cancer research focuses on how tumors spread, not what to do once they have spread. “Traditional chemotherapy is not successful against metastatic disease,” he noted, “and dissemination isn’t a good therapeutic target. If we want patients to survive, we have to work on the metastatic tumors themselves.”

To better reflect what metastatic tumors actually do, Zetter has taken a new tack on modeling them in the mouse. The first step is to direct tumor cells known to be metastatic directly to the organs or other sites where they’d wind up in patients. So in the case of prostate cancer, instead of implanting a tumor in a mouse’s back, you’d inject cells into the bones; for breast cancer, you’d inject them into the lungs or the brain. Or you could inject them into the bloodstream so the cells can get to where they want to go on their own.

The second step is to wait: Let the injected metastatic cells get settled in their new home, then start administering the drug or compound in question, and watch for any effects both on the metastatic tumors themselves and on overall survival of the mice.

Zetter has put his new modeling schema to work in his own research on metastatic prostate cancer, with good results. By modeling metastases in this more realistic way, he has already found that a group of compounds initially developed to combat parasitic infections could have strong tumor-fighting capabilities as well.

“We can get better answers from the mouse,” said Zetter. “We just have to ask better questions.”