Science fiction is becoming reality at the Washington Biomedical Research Center. We’re at the forefront of innovating New Approach Methodologies, or NAMs. Two examples of these advanced tools include artificial neural networks and three-dimensional printed organs which test medical treatments before they reach a living body. But as powerful as they are, these models can’t advance cures alone. Nonhuman primates are still essential for proving whether a treatment actually works.
Think of these new technologies like a computer crash simulation for building a new car. Simulations can predict many outcomes quickly and more efficiently than crashing a bunch of cars. But manufacturers still validate the predictions before putting cars on the road
In the lab of Dr. Megan O’Connor, an investigator focused on viral infections, that simulation takes the form of a three-dimensional printed artificial airway linked to complex computer programming. Her collaborative team uses it to track exactly how respiratory drugs travel through the lungs, testing different flow rates and medicine formulations to ensure therapies reach deep into the respiratory system.
Across campus, neuroscience researcher Dr. Amy Orsborn uses a different kind of simulator: artificial neural networks. She is designing brain-computer interfaces, which are smart software systems that can intercept brain signals and reroute them around a spinal injury to help paralyzed patients move things like a robotic arm. A future use of these algorithms might be creating some prosthetic or tool that helps a paralyzed person learn to walk again. By testing her math algorithms in these virtual networks first, she bypasses repetitive trial-and-error, drastically reducing the number of animals required for her research.
But software has its limits. As O’Connor points out, “A computer model is simply not going to tell us whether or not an animal is going to have a cough reflex, or what the mucus production is actually going to look like, or how a microbe is going to spread from the upper respiratory tract to the lower respiratory tract.” To make sure her airway models are accurate, she has to validate them using real-world anatomical data and breathing measurements from nonhuman primates.
Orsborn hits the same wall with artificial intelligence. “Every neural network is only as good as the data it is trained on, and if you train it on bad data, it’s going to make bad predictions,” she said. Because monkeys have complex motor systems and precise hand control uniquely similar to humans, they provide the essential data that computational models require to succeed, data you cannot get from a mouse or a screen.
If primate research were to stop tomorrow, the risks to human and animal health would be severe. NAMs can’t replicate the complex feedback loops of a living body or a complete immune system. Moving a new therapy or a brain implant straight from a computer simulation into human clinical trials would create huge risks, leaving doctors completely blind to dangerous, unpredictable side effects.
What’s more, the losses to medicine would be devastating. Life-changing therapies for paralysis and neurological disorders could be delayed for decades. Orsborn warns that without the real-world data gained from primate studies, medical advancement stalls out completely: “If we don’t do this work, then we don’t learn how the brain works, and we don’t build devices that can help people.”
Furthermore, vulnerable populations would be left behind. O’Connor’s research specifically focuses on tailoring respiratory treatments for high-risk, underrepresented groups like infants, the elderly, and the immunocompromised. Without nonhuman primates to validate these models, safely developing specialized treatments for individuals with weakened immune systems just wouldn’t work, or worse, might lead to dangerous outcomes. “We want to be confident in what we’re putting out there. No model is perfect. Even humans are not perfect because there’s so much variation in genetics, age, sex, or the health status of humans,” she said. “That’s why you’re trying to get the best data that you can. NAMs can really help to complement that.”
The ultimate goal of modern science is to reach a day when animal models are no longer necessary, but that is still many years away. For now, keeping people safe requires a balanced approach that combines the innovative power of NAMs with the irreplaceable predictive strength of living systems.