November 30, 2022 – Artificial intelligence is designed to make clinical trials and drug development faster, cheaper and more efficient. Part of this strategy is the event of “synthetic control arms” that use data to create “simulants” or computer-generated “patients” in a study.
This way, researchers should enroll fewer real people and may recruit enough participants in half the time.
Both patients and drug corporations may benefit, experts say. One advantage for patients, for instance, is that simulants receive the usual treatment or a placebo treatment, meaning all trial participants ultimately receive the experimental treatment. For drug corporations unsure of which of their drug candidates are most promising, AI and machine learning can narrow down the prospects.
“Until now, machine learning has been effective primarily in optimizing efficiency – not in developing better drugs, but in optimizing screening efficiency. AI uses the lessons learned from the past to make drug discovery more effective and efficient,” says Dr. Angeli Moeller, Head of Data and Integrations for Insights on the pharmaceutical company Roche in Berlin and Vice Chair of the Board of the Alliance for Artificial Intelligence in Healthcare.
“Let me give you an example. You might have a thousand small molecules and you want to find out which one binds to a receptor associated with a disease. With AI, you don't have to screen thousands of candidates. Maybe you can screen just a hundred,” she says.
“Synthetic” study participants
The first clinical trials using data-based comparisons of patients moderately than control patients matched by age, gender or other characteristics have already begun. For example, Imunon Inc., a biotechnology company developing next-generation chemotherapies and immunotherapies, used Phase 1B study an lively ingredient that's added to preoperative chemotherapy for ovarian cancer.
This early study showed the researchers that it will be worthwhile to further evaluate the brand new drug in a Phase 2 study.
Using an artificial control arm is “extremely cool,” says Sastry Chilukuri, co-CEO of Medidata, the corporate that provided the information for the Phase 1B trial, and founder and president of Acorn AI.
“We have the first FDA and EMA approval for a synthetic control arm, where the entire control arm is replaced with synthetic control patients. These are patients derived from historical clinical trial data,” he explains.
A wave of AI-powered research?
The role of AI in research is predicted to extend. To date, most AI-driven drug research has focused on neurology and oncology. The start in these specialties is “likely due to the high unmet medical need and many well-characterized targets,” in line with a March 2022 study. News and analysis within the Journal Nature.
It has been speculated that this use of AI is just the start of a “coming wave”.
“There is increasing interest in the use of synthetic control methods [that is, using external data to create controls]”, such a Review article In Natural medicine in September.
It was said that the FDA had already a drug approved in 2017 for a form of a rare neurological disorder in children called Batten disease, based on a study of historical control participants.
One example in oncology where a synthetic control group could make a difference is glioblastoma research, Chilukuri says. This brain tumor is extremely difficult to treat, and patients typically drop out of trials because they want the experimental treatment and don't want to stay in the control group with the standard treatment, he says. Plus, “it's very difficult to complete a trial just given the life expectancy.”
Using a synthetic control arm can speed up research and improve the chances of completing a glioblastoma trial, Chilukuri says. “And patients actually receive the experimental treatment.”
It is still early
AI could also help limit the number of “non-responders” in research.
Clinical trials “are really difficult, they are time-consuming and they are extremely expensive,” says Naheed Kurji, chair of the board of the Alliance for Artificial Intelligence in Healthcare and president and CEO of Cyclica Inc, a data-driven drug discovery company based in Toronto.
“Companies are working very hard to find more efficient ways to bring AI into clinical trials so they can achieve results faster, at lower cost, and with higher quality.”
Many clinical trials fail, not because the molecule is not effective, but because among the patients who participated in a trial there were many patients who did not respond to the therapy. They simply cancel out the data from the patients who did respond to the therapy,” says Kurji.
“You've heard a lot of people talk about how we're going to make more progress in the next decade than we did in the last century,” says Chilukuri. “And that's simply because of the availability of high-resolution data that allows us to understand what's happening at the individual level.”
“This will lead to an explosion in precision medicine,” he predicts.
In some ways, AI in clinical research continues to be in its infancy. Kurji says: “There is still a lot of work to be done, but I think you can point to many examples and companies that have made really great progress.”