Can AI replace animal testing?
The scientific community has often debated the ethics of using animals for research. Currently, there is no precise information about the number and types of animals sacrificed worldwide for the purpose of drug testing and experiments.
With a better perception of how human biology functions, scientists are realizing the limitations of testing on animals. Plus, the emergence of treatments based on human DNA, cells, and other non-conventional therapeutics have increased the demand for animal testing alternatives.
In an interview, the CEO of Quris, Isaac Bentwich, explained that, at the moment, the development of medications is not very effective. Even 89% of expensive clinical trials, which tend to be rather expensive, fail. One-third of them are safety failures.
The inability to forecast which drug would be safe for human use before the trial even starts is a considerable challenge which hasn’t yet been addressed.
Quris is an Israeli-US biotech company founded in 2020 that uses AI for predicting which drugs are safe for human use.
The history of animal testing
Animal testing originates from ancient Greece. The first tests were recorded during the 4th and 2nd century BC – even the famous philosopher Aristotle was aware of it.
This practice continued in Rome, where Galen, a famous Greek physician, surgeon and philosopher with Roman citizenship became was named “the father of vivisection.” He got this title for performing dissections on pigs and goats.
Over time, animal testing became increasingly important in the field of scientific research. In the late 19th century, Louis Pasteur illustrated the concept of germ theory by deliberately infecting live sheep with anthrax.
Fast forward to the 1970s, experimental vaccines for leprosy were being developed. Armadillos were intentionally infected with the disease in order to test the efficacy of these vaccines.
Despite animal testing having a long history, there have been individuals opposed to the practice for centuries. As early as the 17th century, scientists began to question the ethical implications of using animals for experimentation and whether the benefits outweighed the pain inflicted on the animals.
However, it wasn’t until the late 19th century that the first laws were established to regulate animal testing. These laws were a response to growing public concern about the treatment of animals in research and aimed to ensure that animal testing was conducted in a humane and ethical manner.
Shifting from animals to AI
Multiple nations and medical companies are gradually shifting away from animal testing by utilizing AI. This is particularly due to the high rate of failures in such testing. In fact, the United States Environmental Protection Agency (EPA) plans to completely stop funding and conducting studies involving mammals by 2035. This move sets the EPA apart as the first government entity to establish a definite timeline for eliminating animal research.
The U.S. Food and Drug Administration (FDA) estimates that approximately 92% of drugs that have passed preclinical tests, which includes crucial animal testing, are ultimately considered unsuitable and unsafe for human use. Newer research shows that, in spite of efforts to make animal testing more predictable, the failure rate has increased and is now closer to 96 percent. The primary causes of failure include safety concerns that weren’t anticipated and a lack of efficacy in comparison to human trials.
Less than 10% of what were regarded to be exceptionally promising therapeutic discoveries are put into widespread clinical use over a period of about 20 years.
According to PETA, a U.K. non-profit organization devoted to establishing and defending the rights of all animals, the failure rate of innovative treatments employed in animal testing exceeds 95%. This was especially evident in the case of cancer. The study suggests that 97% of the time a new medication for a s is tested particular type of cancer never made it to the market.
In the period of two decades, less than 10% of what was believed to be a promising medication is widely used. Hence, it can be concluded that many drugs that are effective in mice don’t work well in humans and vice versa. The industry’s low productivity rate, measured in terms of drugs that are tested but never reach the market, illustrates the sporadic reliability of animal testing.
How does AI drug testing work?
As Bentwich explained, Quris implements so-called Bio-AI merged with mighty machine learning (ML) and miniature “patients-on-a-chip” biology. Roche is another household name in the pharmaceutical industry that works on reducing animal testing.
The company claims to be developing state-of-the-art diagnoses and therapies for serious and potentially fatal illnesses. According to Roche, it is investing heavily in the development of methods for producing medicines without involving animals, such as computer simulations or strategies like organ on a chip or organoids.
These are systems with artificial or real tiny tissues developed inside microfluidic chips. In order to better resemble human physiology those chips are made to regulate cell microenvironments and preserve tissue-specific functionalities.
Matthias Lutolf, scientific director at Roche’s Institute for Translational Bioengineering (ITB), explained that these “mini organs,”created in the laboratory from the stem cells of unique individuals, are miniature replicas of the bodily organs such as livers and lungs. They help researchers perceive how human organs work and test particular medicine effects outside the body.
Roche implements cutting-edge techniques like organ-on-a-chip, machine learning (ML), and artificial intelligence (AI) for analysis. These techniques have the potential to replace animal testing in the medium future as well as improve treatment candidates or produce custom medications.
“Patient-on-a-chip” is a step up from a so-called “organ-on-a-chip.” It describes the arrangement of numerous separate tiny, three-dimensional organs (liver, brain, etc.) on a chip that are connected via a blood-like circulation. It might be a small brain, a liver, or other miniature organs that are only a third of a millimeter in size. This makes it possible to evaluate the security of drugs on these networked “patients on a chip,” or miniature human organs.
Bentwich explained that thousands of known medications are tested on various patients-on-a-chip. The generated data is then implemented to train AI to determine which drug is and which isn’t safe for humans. According to Bentwich, the technology proved to be fairly predictive when it comes to drug safety.
Can AI replace animal testing? Takeaway
Without a doubt, science believes that AI-supported technologies have a big potential to dramatically transform the way new medications are created.
The technologies covered here are only a few instances of how science is moving away from the use of lab animals and allowing for the development of precise or personalized medications or their clinical use in a way that maximizes patient benefit while obviating the necessity for animal testing.