In labs all across the country, researchers are using refined versions of an old research tool to provide critical new information on the biology of many cancers and determine which new drugs are the most promising to test in clinical trials. Their tools: laboratory mice. These next-generation mouse models differ from the xenograft models - mice, often with disabled immune systems, transplanted with human tumors - that cancer researchers have relied upon for decades. Instead, through various and often sophisticated techniques, these mice are programmed to develop specific types of cancer, and to do it in a way that substantially mimics how those same cancers arise in humans.

"In many cases, the biology of these models looks very good," says Dr. Cheryl Marks, who administers the NCI Mouse Models of Human Cancers Consortium (NCI-MMHCC). What many of the consortium's investigators are trying to determine now, she continues, "is whether the clinical course in the mice also mimics humans."

Ironically, she says, "The good news is that we don't readily cure these mice... However, we know now that when we test in these models the drugs and combination of drugs presently used in the clinic, we show that many of them would have been reasonably predictive of what actually happened in the clinic."

Engineered models of many cancers have been developed. And although there are many potential roles for these genetically engineered mouse models in cancer research, preclinical testing of new drugs is seen by many in the field as one of the most promising.

"But there is still much science we have to do," Dr. Marks cautions, before it's known whether these engineered mice routinely can be used in this way.

There is at least one successful example to date, however, of this very scenario. Results from preclinical testing of a new drug combination in engineered mouse models of acute promyelocytic leukemia (APL) led to a human trial of the drug combo that has had remarkable results.

APL is often successfully treated with one of two unconventional agents, retinoic acid or arsenic trioxide, to get patients into remission (at which point they then undergo chemotherapy). "But there were conflicting data as to whether treating patients with both agents simultaneously might be antagonistic or cooperative," explains Dr. Scott Kogan, associate professor-in-residence at the University of California, San Francisco. Results from tests of the combination therapy in engineered APL mouse models, he says, made it "very clear that they were cooperative."

Based on those studies, researchers in China initiated a randomized clinical trial to test the drugs in combination up front in newly diagnosed APL patients. With more than 3 years of follow-up in 61 patients given both drugs, there have been only 2 relapses.

At Memorial Sloan-Kettering Cancer Center, Dr. Eric Holland says human clinical trials are being pursued based on the results of tests in genetically engineered mouse models of the brain cancers medulloblastoma and glioblastoma developed in his laboratory.

"For some of the small molecular inhibitors that were identified as good choices based on mouse models, there are [human] trials that are being written and going through IRBs," he says. "This is happening now."

The models closely mimic the response in humans to standard therapies, he adds, so the models also are being used to improve existing therapies.

One of the models developed in Dr. Holland's lab reflects the increasing importance of advanced imaging technologies in mouse model research, which is allowing investigators to closely analyze the biologic and molecular events associated with the disease and its treatment. In the model, dubbed Efluc, cancerous cells express a built-in imaging agent, luciferase - the enzyme in fireflies that causes their tails to light up.

"In a noninvasive way, we can have a readout of a specific cellular behavior," Dr. Holland says. "It can tell you about the activity of certain signaling pathways involved in tumor development and what the therapeutic response is in those specific pathways."

In the laboratory of Dr. Jonathan Kurie at the University of Texas M.D. Anderson Cancer Center, imaging is also an important component of their mouse model studies. The lab has generated some significant results testing chemopreventive agents in an early-stage lung cancer model developed by Dr. Tyler Jacks and colleagues at MIT.

"We do CT scans on all of the mice," he says. "We also do a lot of biochemistry, trying to understand what pathways are being modulated and what cell types are being modulated."

The next step in this process is important: directly comparing the results from the mouse model investigations to similar analyses in human tumors. By looking at measures such as the activation of intracellular signaling pathways in early lesions as well as more advanced lesions, they can determine just how closely the model resembles the real deal.

"We're lucky enough," Dr. Kurie says, "to be at the cusp of a revolution in the mouse model field."

By Carmen Phillips