To enhance reproducibility and rigor in research, NIH is creating a working group to advise the agency on possible changes to animal studies. NIH Director Francis Collins announced the launch of the working group at the June 13 meeting of the Advisory Committee to the Director (ACD). In other action at the meeting, the ACD also heard recommendations to bolster research against interference or influence by foreign entities (“NIH: Examples of Foreign Influence in Research Include Employment Agreement, Secrecy Pledge,” RRC 16, no. 7).
Collins called the new working group an outgrowth of a “long-term focus” during the past five years and one “we believe is highly justified by the need for us to pay attention to whether the money that the tax payers are providing you’re asking scientists to use for good purposes are used in a way that produces” reproducible data from studies that are “rigorously designed.”
The working group also responds to a request in the 2016 21st Century Cures Act calling for NIH to convene such a panel to issue “recommendations for a formal policy” on rigor and reproducibility generally.
NIH reviewed the Cures Act requirement carefully and “looked at what we’ve already done, and we tried to identify areas where we made good progress and others where we think there is still work that needs to be done,” deciding that animal studies are “where we think there are still needs for some deep thinking about this.”
Questions concern whether animal models are “actually [a] good representation of the human illness? Are we, in fact, studying something that we can extrapolate and translate into clinical utility?” asked Collins. Secondly, “are those animal models being studied in a way where the design of the experiments is appropriately rigorous” and “not confounded by all sorts of misadventures in terms of drawing conclusions” that later cannot be confirmed?
Study design “is critical,” as are appropriate techniques for blinding, using enough animals to sufficiently power the study,” among other issues, said Collins.
Financial Implications to Be Addressed
The new working group was to begin its efforts “shortly after this meeting” of the ACD to address the following:
“Identify the gaps and opportunities to improve rigor, reproducibility, translational validity and transparency of studies involving animal models.
“Evaluate how animal models of human disease are currently developed, validated, and accepted into routine use, and how this process could be improved.”
“Assess the current state of science for validating alternative models to animal research,” which Collins said “includes such things as tissue chips.”
“Consider the benefits and burdens of registering animal studies that aim to lead to the first in human trials (e.g., preregistration of the experimental plan).” According to Collins, “we do this for clinical trials, and if you’re designing an animal study that you think [might be] the definitive way to show that an intervention works, why not register the study in advance of initiating it so it’s absolutely clear how it’s designed and what the end point is, the number of animals in the power calculations…just to get that out there?” Collins acknowledged “there will be objections” to this possible requirement “because it sounds like it might be a bit of a burden.”
“Model the financial implications of potential changes in the average costs of grants using animal models, the number of studies funded or the need to develop multi-lab consortia to achieve appropriate statistical power.” Said Collins: “If we are insisting on power, sometimes that means you need to [study] more animals instead of fewer. Yet, at the same time, there’s a lot of arguments for trying to limit the number of animals in a study.” However, this can handicap a study. As Collins said, “If you limited it to the point where you don’t have statistical power, why do it at all?”
“Consider how rigor in animal research is incorporated into training (e.g., statistical training, experimental design).”
ACD Offers Support
The working group will be chaired by Barbara Wold, professor of microbiology at CalTech and a new ACD member, and Larry Tabak, NIH principal deputy director; those to serve had not been appointed at the time of the ACD meeting. Collins said there was a “tentative roster” and asked for suggestions of potential members.
Representatives of industry will be included, Collins said, as “they have a lot of perspective here in terms of why what we have been doing hasn’t necessarily worked out all that well in terms of things that they can then translate to human trials.”
Linda Griffith, director of the Center for Gynepathology Research at Massachusetts Institute of Technology, called the working group an “incredibly wonderful and welcome initiative.” She said it could help the maturation of “micro-physiological systems programs” in terms of producing data to “illustrate with numbers how much more cost effective it may be to put a little more resources to get the tissue chips up to speed.”
Anne Churchland, associate professor of neuroscience at the University of California, San Francisco, asked for clarity regarding what investigators are already required to explain to institutional animal care and use committees reviewing their proposals and the working group’s charge.
“I think that sort of falls upon the investigator to say, ‘Is there right now an alternative I could use instead of an animal experiment to answer this question?’ I think from our perspective it could be what should NIH be doing to enhance the development of alternatives to animal models that we don’t have yet,” said Collins. Investigators may determine they have no other options but to conduct a study using animals, he added.
Churchland also suggested that NIH consider having someone on the working group who studies “genetic disease models and has thought about how those models will inform human disease even if it involves a complex disorder that’s being studied in a simpler animal.”