Artificial intelligence (AI)-based drug developer Recursion has dosed the first patient in its Phase II ALDER trial (NCT06536465) assessing its first new chemical entity to be developed through its RecurionOS Operating System, an oral, non-antibiotic small molecule for recurrent Clostridioides difficile (C. diff) infection.
REC-3964 is designed to treat C. diff by selectively inhibiting the glucosyltransferase activity of toxin B produced by the bacterium in the gastrointestinal tract, offering a different mechanism of action from antibiotics. Unlike antibiotics, which disrupt the gut microbiome, REC-3964 precisely targets the bacterial toxin while sparing healthy tissue, an approach that Recursion reasons could potentially minimize adverse events.
That’s a departure from previous antibiotics-based treatment approaches toward C. diff, a toxin producing bacterium that causes diarrhea and colitis, and can be life threatening. According to the U.S. Centers for Disease Control and Prevention (CDC), each year nearly 500,000 cases of C. diff infection are estimated to occur in the U.S., with the bacterium responsible for an estimated 29,300 deaths.
“While antibiotics can successfully clear the infection, they simultaneously pose an increased risk of developing C. diff. They are at once a treatment and a cause,” Chris Gibson, PhD, Recursion’s co-founder and CEO, told GEN Edge.
C. diff toxin B disrupts the tight junctions in colonic cells and increases vascular permeability, leading to a leaky gut. The standard treatment for C. diff. infections is antibiotics that disturb the gut microbiome due to their non-selective nature. Despite initial success, antibiotics fail to prevent recurrence in 20–30% of primary cases—a risk that rises to 40% after the first case and 45–65% after two or more cases, according to a 2019 study.
Citing research indicating that up to 60% of C. diff cases are treated with antibiotics within the four months preceding the infection, Gibson said, Recursion saw a need for new therapies and an emphasis on antimicrobial stewardship as articulated by the Infectious Disease Society of America (IDSA) and other organizations.
“This is where we believe an oral small molecule drug like REC-3964 could initially intervene,” Gibson said. “By specifically targeting the primary mediator of the symptomatic infection, there is a potential for REC-3964 to intervene during the maintenance stage, without the increased risk of developing infections associated with antibiotics.”
“We expect to share an update on the study in terms of enrollment, preliminary efficacy (rate of recurrence), and safety in the high-risk rCDI [recurrent C. diff infection] population in Q4 2025,” Gibson said.
Dennis Ding, equity analyst with Jefferies, wrote October 21 in a research note that his firm is awaiting results from trials for REC-3964 and another Recursion candidate expected to show data next year, MEK1/MEK2 inhibitor REC-4881 for familial adenomatous polyposis: “In terms of data, we’re looking to RXRX’s MEK1/2 inhibitor in H1:25 and from REC-3964’s Phase II C-diff in 2025 to help validate the platform.”
Crowded field
REC-3964 joins an increasingly crowded field of drugs and vaccines in development against C. diff. Among developments in recent weeks, researchers at the Perelman School of Medicine at the University of Pennsylvania and the Children’s Hospital of Philadelphia on October 3 published a study in the journal Science reporting positive preclinical results from animal models showing that their messenger RNA (mRNA)-lipid nanoparticle (LNP) C. diff vaccine was found to protect against first-time infections and relapsing infections by inducing a robust immune response.
The first-of-its-kind vaccine for C. diff also promoted clearance of existing bacteria from the gut, and overcame deficits in host immunity to protect animals after infection. The research was funded in part by BioNTech, to which Penn has optioned and licensed intellectual property related to the C. diff mRNA vaccine program, while Penn’s Perelman School of Medicine receives funding for research and development from BioNTech related to the work.
Crestone Pharmaceuticals reported positive topline results last month from its Phase II trial (NCT04781387) assessing its small-molecule protein synthesis inhibitor CRS3123. Among 43 patients in the primary intent-to-treat analysis population, C. diff recurrence rates at day 40 were 4% for CRS3123-treated patients, compared with 23% for patients treated with comparator drug vancomycin, the glycopeptide antibiotic marketed in different forms as Vancocin® by ANI Pharmaceuticals, and as Firvanq® by Azurity Pharmaceuticals.
However, Pfizer’s C. diff vaccine PF-06425090 failed the Phase III CLOVER trial (NCT03090191), missing its primary endpoint by failing to show significant reduced incidence vs. placebo of C. diff infection (CDI) in at-risk adults after being given two or three doses. According to results published August 24 in Clinical Infectious Diseases, vaccine efficiency was just 31% as 17 participants receiving all three doses of PF-06425090, and 25 randomized to placebo, had a primary CDI episode 14 or more days after the third dose. Among those who received two doses, 24 PF-06425090 and 34 placebo recipients had a first CDI episode 14 or more days post dose two, for a vaccine efficiency of 28.6%.
In early work on C. diff, Gibson said, Recursion researchers identified numerous compounds which mitigated the effect of C. diff toxin B on human cells, and which were likely to be working via a variety of different host or toxin-specific targets. Based on later orthogonal screens, precursors to REC-3964 emerged as the most promising substrate for additional work.
“Said another way, we didn’t go into this field with a goal of identifying an anti-toxin molecule, but instead leveraged the RecursionOS to identify the non-antibiotic modality that the data supported as being most promising,” Gibson added.
RecursionOS is an AI-enabled platform allowing the company to map and navigate trillions of biological and chemical relationships within one of the world’s largest proprietary biological and chemical datasets—more than 50 petabytes across multiple data layers—including phenomics, transcriptomics, and patient and animal data.
“Cycles of virtuous learning”
By closely integrating our wet-lab experimentation and dry-lab machine learning in an iterative manner, we create cycles of virtuous learning, where large fit-for-purpose wet-lab datasets support better in silico model generation and enable more focused future wet-lab experiments,” Gibson said. “On average, we are 3x as fast and half the cost to IND enabling studies compared to industry.”
Because RecursionOS goes through cycles of learning, he added, the company expects increasing efficiency and ability to gain novel insights and create new chemical entities.
What savings in time or costs come as a result?
“Using the power of our massive connected datasets to surface novel opportunities, we’ve developed a small set of standardized and industrialized experiments to rapidly prioritize and confirm our in silico insights for tens of thousands of dollars each and in just weeks— as opposed to spending years and multiple millions of dollars to research one specific target,” Gibson replied.
At the 6th Edition of World Congress on Infectious Diseases, held in June in Paris and online, preclinical studies showed REC-3964 to be superior over bezlotoxumab (a C. diff-binding monoclonal antibody marketed by Merck & Co. as Zinplava®) in a human disease-relevant C. diff. hamster model. Additionally, Phase I studies in healthy volunteers showed REC-3964 was well tolerated with no serious adverse events.
The Phase II ALDER trial is a multi-center randomized study designed to investigate the safety, tolerability, pharmacokinetics (PK) and efficacy of REC-3964 at doses of either 250 mg or 500 mg for the reduction of C. diff. and will include an observation only arm.
Approximately 80 patients will ultimately be enrolled in the study across the U.S. and Europe. Recursion says it is working with leading academic researchers in the C. diff space to build awareness around the trial. As part of that effort, Andrew Skinner, MD, assistant professor of infectious disease at the University of Utah, will discuss the REC-3964 program at the Peggy Lillis foundation’s first scientific symposium on November 15.
