Increasing value and reducing waste in research has to start at the pre-clinical stage, with high quality animal studies and well-characterised mouse models of disease. Steve Perrin provides recommendations for researchers to “make mouse studies work” in a comment in Nature1.
Steve Perrin studies amyotrophic lateral sclerosis (ALS), a progressive, fatal degenerative disease, also called Lou Gehrig’s disease or motor neuron disease. In his role as the chief scientific officer of the ALS Therapy Development Institute in Cambridge, Massachusetts, he recently re-examined 100 compounds in an ALS mouse model that had been identified as candidates for ALS therapy and found that most of these failed to slow the disease in the animals. Importantly, the researchers could not reproduce the results for 8 drugs that had looked promising in previous mouse studies (using the same animal model) and proceeded to human trials, which ultimately failed. Perrin argues that this discrepancy most likely means that the original mouse studies were poorly conducted, and that those compounds should never have made it to clinical trials1,2.
This is a separate issue to the fact that many mouse models differ significantly from the human disease, which Perrin also highlights in his comment. In one ALS model, mice die from bowel obstruction whereas humans with the disease often die from an inability to breathe as a result of muscle wasting1,2.
These problems with mouse studies are not unique to the field of ALS. 150 drugs tested in human trials of sepsis failed because the corresponding mouse models poorly mimic the human disease3, and animal studies in cancer have similar reproducibility issues4.
In his freely available comment, Perrin provides recommendations to reduce waste in preclinical research and ultimately harm to patients, who may only have one shot at participating in a clinical trial. Similar to human trials, animal studies should be carefully prepared, designed and controlled2. This includes regular genotyping and backcrossing of mouse strains to maintain disease-causing genes, mathematical simulations to determine the number of animals required per experimental group and “unglamorous” research to characterise mouse models.
What Perrin fails to mention are some of the key recommendations from the Lancet series on waste in research that apply just as well to animal studies as to human clinical trials. A systematic review of previously reported studies should identify the need and help define the research question; all studies should be registered; protocols, analysis plans and raw data should be made publicly available; negative results should be reported; and results should be reproducible.
- Perrin S. (2014) Make mouse studies work. Nature 507; 423-425.
- Hayden EC. Misleading mouse studies waste medical resources. Nature News 26 March 2014.
- Seok J et al.(2013) Genomic responses in mouse models poorly mimic human inflammatory diseases. PNAS 110(9):3507-12.
- Begley CG & Ellis LM. (2012) Drug development: raise standards for pre-clinical cancer research. Nature 483, 531-533.