Over the years, many cancer-fighting compounds have looked very promising in early-stage experiments. Unfortunately, fewer than one in 10 of them actually make it to the clinic, after years of development and clinical trials.

Why are safe, effective cancer drugs so hard to develop? Two key reasons illustrate the challenges faced by cancer researchers. First, cancer describes a huge number of different diseases. Each is characterized by uncontrolled cellular growth and division, but the origins, genetics and molecular profiles of cancers vary widely. Killing different cancers often requires very different methods.

Second, cancer growth is greatly affected by environment. The same tumor may be very different when grown in Petri dishes, different species, different people, even in different places within the same animal or person. This has made standard biomedical research using human cancer cell lines ineffective, and cancer research findings from animal models are frustratingly difficult to translate to human medicine, even with the recent advances in research tools and techniques such as genomic sequencing.

The Jackson Laboratory is spearheading a collaborative effort to address the problem. In 2009 it launched a National Tumor Consortium to build a library of primary human tumors from patient volunteers. Early support for the consortium came from the UC Davis Cancer Center, the Swedish Neuroscience Institute in Seattle, the University of Florida and the Scripps Translational Science Institute, and now there are 10 clinical partners donating solid human tumor samples for characterization and research purposes.

Getting cancer samples straight from patients is one part of the research puzzle. The other part is to have an effective way to grow and study the human tumors outside of patients. An amazing new mouse, called NSG, allows scientists to do just that. Using these special mice, which have almost no immune system of their own, researchers are able to grow and study human tumors in a living system, characterize them at a molecular level and identify driver mutations, and test new therapies. The translation from patients to lab and back to patients again is much faster, and the process promises to be more effective in developing useful therapies as well.

In addition to The Jackson Laboratory’s own research efforts, these resources are also being made available to external cancer researchers. The first two mouse models are now ready for breast cancer research, and many more are being developed for a variety of cancer subtypes. The hope is that these tools will provide a feasible way to significantly improve that crucial first step in cancer research, in which clinically useful target pathways and drugs are quickly identified. And, just as importantly, so are the many dead ends.