Inflammatory bowel disease (IBD) impacts 1.6 million people in the U.S. — and a new artificial intelligence-generated drug could help alleviate symptoms.
Insilico Medicine, an AI-driven biotech company based in Hong Kong and in New York City, recently announced that its new AI-designed IBD drug — ISM5411 — has entered Phase I clinical trials.
This is Insilico’s fifth AI-designed drug to enter the pipeline.
WHAT IS ARTIFICIAL INTELLIGENCE (AI)?
If approved, it would be the first medication to treat IBD by blocking prolyl hydroxylase domain (PHD), a protein that regulates the body’s gut barrier protection genes, according to Alex Zhavoronkov, PhD, founder and CEO of Insilico Medicine.
“ISM5411 is a potentially first-in-class candidate for the treatment of IBD – this means that it is designed to treat the disease in an entirely new way,” Zhavoronkov told in an interview.
IBD, which includes Crohn’s disease and ulcerative colitis, causes debilitating symptoms that include abdominal pain, diarrhea, fatigue, weight loss and rectal bleeding, according to Mayo Clinic.
“This condition is on the rise across the industrialized world, but currently there are no cures and few treatments,” said Zhavoronkov.
NEW AI-GENERATED COVID DRUG ENTERS PHASE I CLINICAL TRIALS: ‘EFFECTIVE AGAINST ALL VARIANTS’
Many current drugs for IBD are anti-inflammatory and rely on immunosuppression to work.
“This carries its own set of risks, as suppressing the immune system can give rise to chronic infection and tumor diseases,” Zhavoronkov said.
Other biologic drugs for IBD are administered by IV or self-injection.
“We knew there was a significant need in this space, and we wanted to find a new way to treat IBD that focused on healing and rebuilding the intestinal lining to lead to long-term improvement of the condition rather than just treating the symptoms,” said Zhavoronkov.
IBD patients produce significantly more of the PDH protein than those without the condition.
Insilico’s research and development team used Chemistry42, the company’s generative AI chemistry engine, to design a molecule to block the PDH and produce genes that protect the gut barrier.
FIRST AI-GENERATED DRUG ENTERS HUMAN CLINICAL TRIALS, TARGETING CHRONIC LUNG DISEASE PATIENTS
“You can think of Chemistry42 as a ChatGPT for new molecules,” said Zhavoronkov.
“But instead of providing text commands and having an AI model generate a new essay or poem based on existing data, our platform draws from biological and chemical data to design an entirely new molecule using our instructions.”
“Our AI platform designed a number of possible molecules to meet our criteria, and ranked them, and our R&D team synthesized and tested a number of these before selecting ISM5411 as the most promising candidate,” said Zhavoronkov.
An oral medication, ISM5411, is described as “intestinal restrictive,” which means it only works on the gut without impacting other parts of the body, he noted.
“Very importantly, it promotes mucosal repair,” said Zhavoronkov.
“In other words, we designed this drug to help restore the normal functioning of the gut.”
Previous research has shown that mucosal repair is effective in reducing hospitalization and improving long-term prognosis for IBD patients.
Insilico’s IBD drug has officially begun trials in 76 healthy volunteers in Australia to evaluate its safety and tolerability in gradually increasing doses, Zhavoronkov said.
After completing the Phase 1a trial, Insilico plans to launch an international multi-center clinical trial in the U.S., China and other locations, with three treatment groups and one placebo group.
“Our clinical team is working hard to design and execute the clinical trials in order to develop it further so we can begin helping patients in need,” said Zhavoronkov.
“As we continue to advance this drug, we also look forward to connecting with companies with relevant experience and resources who are interested in co-developing this asset with our team.”
STUDENTS USE AI TECHNOLOGY TO FIND NEW BRAIN TUMOR THERAPY TARGETS — WITH A GOAL OF FIGHTING DISEASE FASTER
ISM5411 is one of more than 30 drugs that have been designed by Insilico’s AI platform, including many for cancer — as well as fibrosis, central nervous system diseases and COVID-19.
The company’s lead drug, the first AI-discovered and AI-generated drug to reach Phase II trials with patients, is for the treatment of idiopathic pulmonary fibrosis, a progressive lung condition with an extremely poor prognosis.
Dr. Supriya Rao, a Massachusetts-based gastroenterologist, was not involved in Insilico’s drug development but offered her input on ISM5411.
“I think it’s very promising,” Rao told . “Obviously, there will be hurdles to jump through, and data is very novel at this stage of development.”
“Longer-term safety and efficacy would be concerns, but we’ve only just begun to embark on a journey of AI and health care,” she added.
All treatments have their pros and cons, Rao noted, particularly with “very individualized diseases” like IBD.
BE WELL: IMPROVE YOUR GUT HEALTH WITH 5 TOP TIPS FROM A NUTRITIONAL THERAPIST
“Could this speed up the recovery process and offer a more targeted treatment plan with minimal side effects? The possibility is exciting,” she said. “Our goal as health care professionals is always to achieve better patient outcomes.”
AI expert and emergency medicine physician Dr. Harvey Castro from Dallas, Texas, who was also not involved in the development of ISM5411, noted that while promising, these types of AI-generated drugs can carry inherent risks.
“The primary concern is the uncertainty surrounding their long-term effects, as AI-designed drugs are a relatively new medical frontier,” Castro told.
“There’s also the challenge of ensuring that the AI algorithms are trained on comprehensive and unbiased data sets, as misinterpretations or gaps in data could lead to ineffective or harmful drugs.”
Castro noted that existing regulatory frameworks may not adequately address the nuances of AI-driven drug development.
“Public awareness is key,” he said.
“As these drugs move from research labs to clinical settings, the public needs to understand the potential and the limitations,” Castro went on.
“Educating patients and health care providers about how AI-generated drugs are developed and their potential impact on treatment paradigms is vital.”