Animal models in pharma have become obsolete

The Food and Drug Administration’s recent decision to phase out animal use in preclinical testing of monoclonal antibodies — and eventually other drugs —with more advanced, human-relevant methods marks a transformative shift with profound implications for drug development. The FDA has now officially recognized that advancing science and technology have pushed us across a threshold: Animal models have become obsolete.

We are now moving away from animal testing toward surrogate systems based on human biology. No longer will our medicines be gated through animals, but rather through entirely human systems — leading to breakthrough options for the humans that need them.

Animal models are often flawed and offer incomplete representations of human disease. Yet we have long relied on their data to establish the risk/benefit for experimental drugs. Despite a decades-long rise in research and development costs for new drugs, the pharmaceutical industry has had relatively little to show for it. A nearly 10-fold increase in industry investment these past 40 years has enabled only about a doubling of new drug approvals.

Many industries would view this as a losing business strategy. But for pharma, there has been a certain acceptance of a diminishing return on investment in exchange for the knowledge that at least the process is routinized and predictable.

So predictable, in fact, that it is commonly understood that a staggering 90% of drug candidates that reach clinical trials will fail to advance to approval. Each drug that fails — usually because it is too toxic or simply doesn’t work — is first tested in animals. These animal tests have been required since the middle of the last century as proof of concept that the drugs will help and not hurt people.

But the requirement was never based on a scientific rationale that animals would provide predictive models for drug interactions with human physiology. In practice, this has meant significant wasted effort on drug candidates that were never going to work. Worse, the process has eliminated candidates that could have been game-changers for patients. The screening mechanism is faulty.

FDA Commissioner Marty Makary has released a roadmap that may mark the beginning of the end of this practice. This effort culminates years of consideration on the part of the FDA, the National Institutes of Health, Congress, and other experts. The roadmap outlines a vision that begins with reducing toxicity testing in animals over the next three years to replace them with human-relevant experimental systems tied to in silico tools like AI. The FDA will be working to validate novel methods, develop new guidance and standards, and monitor outcomes.

This may become a new era in drug development. There are three main areas of challenge to getting this done: policy, social, and technical.

The policy space is suddenly wide open. The political will is there in both branches of government and across parties. The FDA, NIH, and ARPA-H have been laying the groundwork for the last few years to promote new and improved methods to assess drugs before they enter human trials. As it advances its roadmap, the FDA should first bring its regulations into compliance with a 2022 law, the FDA Modernization Act 2.0, that removed the statutory mandate for preclinical animal testing. Congress is not interested in delay and is working to force the FDA’s hand with additional legislation. Meanwhile, HHS is gearing up with a new NIH office that will prioritize human-based technologies.

Socially, the challenge will be convincing pharmaceutical companies and their investors that not only can you make this shift, but that it will be better in every way. The FDA will need to be very specific in its guidance to industry and in the verbiage of any revised rules that define what kind of data it will accept — and what it really wants to see — in a new drug application. The 12-page roadmap is only the beginning. The FDA will need to work closely with biotechnology and pharmaceutical companies to ensure a kind of co-development of technologies that will meet and exceed the standards that have been the norm for decades.

The technical challenge will be decried as too hard. Yet technical roadblocks should not be viewed as reasons to quit the course, but rather as opportunities to advance the science. The Manhattan Project, Apollo Program, Stockpile Stewardship Program, Human Genome Project, and Operation Warp Speed all encountered what at first blush may have appeared to be impossible-to-overcome challenges blocking the path to goal. The Stockpile Stewardship Program’s efforts to eliminate nuclear warhead testing, for instance, necessitated simulations of  nuclear weapon explosions that required supercomputers that ran at speeds many logs faster than were available or technically possible at the time. But we persevered, and the science advanced.

Offer bioengineers a prize, they’ll give you a great prototype; offer them a genuine regulatory and financial incentive, they’ll get you solutions that obviate the perceived need for animals and create far more successful shots on goal. From organs-on-a-chip to digital twins to machine learning algorithms and other technologies not yet invented, the future lies with more predictive methods and models and the data that power them. We can measure all sorts of things that can feed into models, and they don’t even need to be clean: Computer science is learning to sort out these challenges, and fill the gaps that variability creates. Treasure troves of federally funded data sets can be moved from servers to a whole new class of contract research organization dedicated to supporting predictive technologies and efforts. Their outputs will then give us the confidence to take a candidate into a subtherapeutic “Phase 0” trial — a tiny dose in humans to evaluate toxicity.

Public sentiment is turning against animal testing, too. Polls show a majority and non-partisan objection to the practice. There is a parallel here with nuclear testing. Modernizing and maintaining the safety of the nuclear weapons program rode on public sentiment and ultimately political will in the executive branch and Congress. Importantly, it was accompanied by an established timeframe for change, provision of funding to advance viable alternatives, and agency accountability to deliver the change despite daunting technical challenges.

It’s time to disrupt a baked-in system that hasn’t been serving the people who need it most. We all deserve a system that pushes out much-needed, first-in-class innovations that could define a new era of therapeutic solutions for infectious disease and chronic health challenges. It’s time to retire the beleaguered animals and put predictive technologies to work. We’re excited to see what’s next.

Ellen P. Carlin is a vice president at Pathway Policy Group and a veterinarian. Jason Paragas is the CEO and founder of DVLP Medicines and a virologist.