Last Stop Biotech

President Joe Biden, in late December 2022, signed legislation that changes the requirement of testing in animals before submitting IND to the FDA to begin human clinical trials. This is a landmark change that will take the place of the requirement since 1938 for drugs to be tested in animals for assessing safety and efficacy. To gain FDA approval in the United States, drug candidates must undergo toxicity testing in rodent species (e.g., mice or rats) and non-rodent species (such as dogs or monkeys). This process involves the use of tens of thousands of animals annually. However, over 90% of drugs that progress to human clinical trials fail due to safety or efficacy concerns—raising critical questions about the predictive value, cost-effectiveness, and ethical implications of animal testing in drug development.

No Animal Testing Outlook

The proponents expressed two views on removing the requirement for animal testing of INDs before entering human clinical trials. The first outlook is that animal models are often wrong. Generally, the FDA will require you to do a toxicity study and a proof of mechanism of action study in rodents and one nonrodent species. Still, at the minimum, nine in 10 drugs fail in human clinical trials due to being unsafe or ineffective. I have noticed firsthand how difficult it is to model human disease in animal models. Even off-the-shelf animal models such as the popular 5XFAD (used in around 10% of all AD studies) have the disadvantage that they do not exhibit any tau pathology. Current mouse models of Alzheimer’s disease (AD) typically replicate only specific pathological features—such as amyloid-β (Aβ) plaques or neurofibrillary tangles (NFTs)—and rarely exhibit the full spectrum of AD-related symptoms in a single model. No mouse model can comprehensively replicate the symptomatology of human AD.

Continuing with the Alzheimer’s Disease example, there have been a plethora of experimental mouse models that mimic certain aspects of the disease’s neurodegeneration. While transgenic mouse models have been critical for expanding our understanding of disease pathology, they are also restricted by their ability to only assess certain pathological features of the disease—such as they can focus only on the tau pathology. This restricts their ability to deliver concrete data on the drug’s impact and safety.

The second outlook that proponents of removing the requirement for animal testing of INDs express is the cost of R&D for testing in animal models. These mice are not cheap and can range from ~$24 to more than $300 per mouse, depending on the strain and age of the mice. On top of that, a biotech or pharmaceutical company has to spend money on personnel who are experienced in processing tissue and animal models and developing experiments with these models. Moreover, special, sterile facilities are needed to house the mice and run these experiments, which adds to the cost. Coupling the difficulties of modeling the disease holistically, the cost of running these studies, and the high failure rate in human clinical trials, the return on investment (ROI) for biotech and pharma companies is relatively low.

Emerging New Approach Methodologies

Due to the inherent limitations of animal testing, there has been growing interest in recent years in creating faster, more cost-effective, and more informative alternatives for toxicology assessment. These emerging strategies, intended to replace traditional animal-based methods, are known as new approach methods (NAMs). Companies such as Emulate are developing technologies that can replace these animal models. Emulate is developing organ chip technology. These chips are usually made of silicone-based polymers and are roughly the size of a USB drive, with microchannels lined with living cells and tissues from organs like the brain, liver, lung, and kidney. Fluids are circulated through the channels to simulate the movement of blood and other biological fluids, replicating how substances flow through human organs. Since the liver plays a key role in metabolizing drugs, liver-on-a-chip systems are particularly useful in detecting toxicity—if an experimental drug harms the liver cells within the chip, it can signal potential liver damage before testing in humans.

_{A liver chip made by Emulate}

Other examples of NAMs are organoids—three-dimensional, hollow structures grown from stem cells that resemble specific human tissues. These models have shown potential in predicting liver and heart toxicities. There is also the promise of digital, artificial neural networks and machine learning models that can rapidly assess and predict the toxic effects of drug candidates.

FUTURE OUTLOOK

Though these changes are exciting and new technologies are gripping, how quickly they will be accepted and implemented remains to be seen. Wendy Jarret, a CEO at Understanding Animal Research, is apprehensive about these NAMs. In an interview with Science, Wendy mentioned, “We can drop a new [candidate drug] onto a bunch of liver cells. And we can see that it does not damage them,” she says. “But what we do not know is whether it is going to make the person cough, whether it’s going to damage their intestines or their brain.”

Though there is bipartisan support for nonanimal methods and removing the requirement of animal testing, it is still unclear how the FDA will pass judgment on new INDs submitted for human clinical trials without animal studies. The key change in this law is that the FDA no longer requires drugs to undergo animal testing before being approved for human trials. This opens the door for pharmaceutical companies and the FDA to actively explore and consider alternative testing methods.