If any researcher thinks that animals are the best models for testing cancer therapeutics, they are clearly mistaken. Clinical trials are the gold standard and in order to get these trials initiated, pre-clinical work with murines, canines and primates is often needed by CTEP but not always. There have been a few phase I trials where strong results from the bench top with clinical specimens, not including animal studies, were enough to kick-off a few clinical trials. I agree that animal studies can provide tremendous insight but their correlation to humans is limited at best.
There have been numerous reports where immuno-compromised mice have been cured of cancer with novel agent, "X," but fails miserably in patients

. There are simply too many differences between mice and humans for researchers to make concrete, objective conclusions that will predict patient outcomes. The differences in metabolism and immune responses are often the biggest players in the studies I've performed. You can get around the immune system but there is no way to convert drug dosages between species.
I really have no solution to animal testing but if researchers are to use live animals, they better have extensive bench work and solid results that supports their hypothesis.
OP: The video is very touching and those beagles look like they will be awesome companions.