FDA’s shift from animal testing opens doors for organoid makers

The last few weeks have been busy for Jim Corbett.

Corbett, the CEO of organ-chip developer Emulate, has been fielding requests from potential clients, as well as from early investors who say they want to put more money into the start-up. The impetus: an announcement from the US Food and Drug Administration in early April indicating that the agency plans to phase out animal tests—a longtime cornerstone of preclinical drug testing—in favor of other technologies.

Among those technologies are Emulate’s organ chips. Originally designed by Don Ingber and his team at Harvard University’s Wyss Institute for Biologically Inspired Engineering, the chips are tiny, thumb-drive-sized devices with hollow channels lined with living, human cells, sourced from a variety of places, such as living patients or stem cells. Those channels are positioned such that they re-create the tissue interfaces that make an organ functional. Emulate’s liver chip, for instance, has hepatocytes on one side and capillary cells on the other. That chip was able to correctly identify 87% of hepatotoxic drugs that caused liver injury in patients in a 2022 study that the FDA cites in its announcement (Nat. Commun. Med., DOI: 10.1038/s43856-022-00209-1).

The FDA says that over the next 3–5 years it aims “to make animal studies the exception rather than the norm” for preclinical safety testing, starting with monoclonal antibodies and eventually encompassing all drugs. The agency might even fast-track reviews of studies that use alternatives to animals.

“This is a clear and deliberate shift,” Corbett says.

Emulate and other organ-chip manufacturers stand to gain from this shift. So do developers of organoids and computer simulations. Drug development has been inching away from animal testing for several years, spurred in part by the 2022 FDA Modernization Act 2.0. But executives say the FDA’s new road map is a clear signal that the agency is taking matters into its own hands.

“It was a much bigger step than we expected. It feels like a key watershed, historic moment,” says Tomasz Kostrzewski, chief scientific officer of another organ-chip developer, CN Bio. “It’s a real flag in the sand. This is the point where the FDA is saying, ‘We’re totally committed to move forward and away from animals in a 3-5-year window.’”

The science of safety

For decades, researchers have used animals to predict how safe and effective a drug candidate would be in humans. Doing studies with rats and mice, whose bodily systems are similar enough to ours, is standard practice before a compound ever makes its way into a human body. Scientists sometimes use larger animals, including dogs and nonhuman primates like rhesus macaques.

A whole industry exists around animals that are sacrificed to science. Estimates vary, but the global animal biotechnology market easily reaches tens of billions of dollars.

Animal testing has its pitfalls. Mice, rats, beagles, and monkeys are smelly and expensive to maintain, and testing with them consistently raises ethical concerns. (Animal rights groups were some of the loudest voices celebrating the FDA’s announcement.) And animal models aren’t great stand-ins for humans. Nearly 90% of drug candidates fail in clinical trials. The oversimplified explanation: just because a molecule successfully treats, say, a mouse version of Parkinson’s disease doesn’t mean that it will have that same effect on a human patient.

“The human brain is incredibly complex. It’s incredibly intricate. Animals just don’t have a brain that’s representative, quite frankly, of anything close to a human brain,” says Alif Saleh, CEO of AxoSim, which develops brain organoids. “The idea that a mouse brain or a rat brain or any other animal brain can predict how a human brain would react to a particular drug—it's not credible. Yet up until this point, this has been the only solution that pharma has had available to them.”

AxoSim essentially makes miniature brains by growing human neurological cells in hundreds upon hundreds of small plastic wells. Those cells naturally form a 3D organoid. In a number of days—Saleh doesn’t say how many—they form a lab-grown brain representative of the brain of a teenager. They keep growing and eventually become a decent mirror of a 35-year-old.

Researchers at Takeda Pharmaceuticals International recently tested the ability of AxoSim’s microbrains to correctly predict the neurotoxicity of 84 known drugs. The microbrains detected safety issues with high specificity and midrange sensitivity, which suggests that they “may be used as an effective early screen for potential drug candidates,” the scientists wrote (Altex 2022, DOI: 10.14573/altex.2112221).

Takeda is one of several pharmaceutical companies that have already begun to incorporate animal alternatives into preclinical practice. Emulate has partnered with Moderna, which uses its organ chips to screen lipid nanoparticles for safety. Organ-chip maker Hesperos has worked with Sanofi, AstraZeneca, Argenx, and Apellis Pharmaceuticals to screen drug candidates for safety and efficacy in neurodegenerative and other diseases. Several of those compounds are now making their way through clinical trials, says Hesperos chief scientist James Hickman.

It’s likely that, going forward, drugmakers will use a combination of tools to decide which of their compounds they want to move into human testing. Artificial intelligence and in silico models—computer simulations that show how a drug candidate might bind to a specific protein, for instance—will “have a place at the table, without a doubt,” Corbett says.

Tina Morrison, vice president of scientific strategy at the start-up EQTY and former senior executive and science adviser to the FDA’s chief scientist, says organ chips should exist alongside a computational surrogate.

Thomas Hartung, director of the Center for Alternatives to Animal Testing at Johns Hopkins University, suggests a similar approach with organoids. “Perhaps with AI and organoids together, we might be able to do some things like cognitive functions and behavior, exactly the type of things where you’d typically refer to nonhuman primates,” Hartung says.

Trusting a new technology

The biggest challenge in replacing animal tests with such technologies will be safety. An unnamed former FDA official told analysts at Leerink Partners that efforts to adopt these new tests “could be set back by any patient deaths that are associated with a drug tested using a prematurely adopted model,” the analysts wrote in an April 11 research note.

Hickman is blunter. “If you fail on safety, people get really upset,” he says. “That could set the field back years.”

Models like organoids and organ chips were originally designed to mimic one organ at a time. But side effects often happen in body systems other than the one the drug is meant to target. These off-target hits can trigger unwanted immune responses or damage the liver or gastrointestinal systems, for instance.

There are some ways around that. For example, immune cells could be added to one of the hollow channels in an organ chip. And any new drug candidates could be tested in a liver model in addition to a model of the target organ; since the liver acts as a sink for many drugs, it is a strong indicator of toxicity.

Ingber’s lab at the Wyss Institute has developed around 15 different organ chips and linked them together. He calls the system a “human body on a chip” and says it can quantitatively predict a drug’s pharmacokinetic parameters and how drug levels change over time in the body. That in turn can help “shortcut clinical trial design,” he says.

Hartung suggests a similar model that incorporates immune cells specifically. “For me, the biggest concern is that we are not including the immune system in many of these models,” he says. “There’s essentially no disease in which inflammation plays no role. There’s no toxicity without it. But if you know it, you can build it in as an engineering system.”

“The more trust we have, the more this becomes a reliable tool,” Hartung adds.

The cost of innovation

Ensuring safety is not an inexpensive endeavor. The former FDA official who spoke with Leerink said that success is likely to require funding from Congress along with private-public partnerships. A shift away from animal testing has historically enjoyed bipartisan congressional support; the FDA Modernization Act 2.0 aimed to clarify that the agency could use data from organ chips, computer models, cell-based assays, and other models to evaluate new drugs. But that bill did not come with any change to the FDA’s funding for such endeavors. Neither did the FDA Modernization Act 3.0, which was reintroduced at the beginning of this year in an effort to direct the agency to implement the 2022 law.

Meanwhile, funding cuts and freezes under President Donald J. Trump’s current administration have threatened work that would advance nonanimal models. Ingber says two of his organ-chip projects were recently hit with stop-work orders, which he considers collateral damage from Trump’s ongoing feud with Harvard University over its response to alleged antisemitism on campus.

Emulate has likewise experienced “pauses” on some of its government contracts, Corbett says. The funding issues have also affected academic clients who receive money from the federal government.

“There’s been some people on hold until they know further what to do,” Corbett says.

AxoSim CEO Saleh thinks the biotech sector will be the quickest to transition to nonanimal models for preclinical drug testing. Unlike large pharmaceutical companies, biotechs are usually focused on one modality or disease area at a time. And because they’re smaller, they can be nimbler.

Pharma firms can be slower to change their processes, but they also have deep pockets, and many have already begun to use organoids and other animal alternatives.

The elephant in the room is clinical research organizations, or CROs, according to multiple animal alternative experts. CROs contract with drugmakers to test new compounds in their own labs, often using animals. The experts suggest that CROs are among the least inclined to move away from animal testing because that’s where they’ve historically made the most money.

“Engaging with CROs has been quite hard,” says Kostrzewski at CN Bio. “They are entrenched in the standard ways of working. They just want to do the same assays over and over again, because that’s what they do really well, and they know how to do them really well, and they make good profit margins on those studies.”

Charles River Laboratories has long specialized in traditional animal trials. The firm’s stock price dropped more than 28% the day the FDA made its announcement as investors hurriedly sold off shares. Less than 2 weeks later, Charles River touted a pivot to “humanized platforms,” including tumoroids, cell assays, and AI software.

Other CROs are also beginning to expand their offerings beyond animal testing. That’s good business for organ-chip companies like CN Bio.

“CROs are hearing questions from their clients: ‘Do you offer organ-chip studies? Do you offer some of these other new, alternative methods?’ So they’re coming to us,” Kostrzewski says. “I think we will see more of that happening.”