A quiet revolution in drug testing is underway. Scientists and regulators are moving away from traditional animal models and embracing human-relevant technologies that could dramatically cut the number of experimental drugs that fail in clinical trials. The shift, experts say, hinges not just on adopting these tools but on using them wisely.
At a recent industry conference in London, Merck KGaA’s chief science and technology officer Laura Matz announced the company’s ambition to replace the majority of its animal testing protocols over the next decade. The goal reflects a broader regulatory push toward “human-centric” approaches that better predict how a therapy will perform in people. Central to this change are new approach methodologies (NAMs), a category that includes organoids, organs-on-a-chip, and advanced computational models.
These tools offer a powerful advantage: they use human cells or tissues, making them far more predictive than animal tests. David Apelion, CEO of British biotech Theolytics, described how his team uses fresh human tumours with their intact microenvironment to test immunotherapies for ovarian cancer. “That kind of relevant component in preclinical selection will reduce the failure rate dramatically,” he said. Experts noted that NAMs are already having the greatest impact in fields like oncology and cardiovascular disease, where they are used to model drug efficacy.
But the technology is not a silver bullet. Chris Floyd, head of neuroscience at pharmaceutical consultancy tranScrip, warned that developers must carefully define what question they are trying to answer before running a model. “If you’re not learning something from the model that’ll move the programme forward, or you’re not able to interpret the results, it’s not a very valuable use of your time or resources,” he said. Harel Kotler, clinical AI lead at Merck KGaA Healthcare, added that experiments must be designed around key objectives, as the industry can now generate vast amounts of data from a single experiment.
Floyd also cautioned against swapping one unknown for another. “We must be careful not to swap an expensive model we understand for an expensive one we don’t,” he said. Developers need to understand how a model will influence decision making before they invest in it.
Looking ahead, the promise of NAMs is clear: faster, cheaper, and more accurate drug development that puts human biology at the center. As regulators encourage adoption and companies refine their strategies, patients stand to benefit from therapies that are more likely to work the first time. The road to replacing animal testing will require discipline, but the destination is a more efficient and humane system of medicine.