Recursion Reports New REC-4881 Results in FAP, Targets 2027 Pivotal Trial

Recursion Pharmaceuticals reported new results on October 15, 2026, for REC-4881, its experimental oral drug for familial adenomatous polyposis (FAP). The company said 13 of 15 adults in its Phase 1b/2 TUPELO trial had less total polyp burden after 12 weeks on the drug. It put the median reduction at 47%. The figures came in a business update that also announced a $200 million stock sale.

FAP is an inherited condition that causes many precancerous growths, called polyps, in the colon. It has no approved drug treatment, Recursion says. The new figures are company-reported topline results from a small trial with no comparison group. Recursion aims to start a registrational trial, meant to support an approval application, in the second half of 2027. That depends on ongoing talks with the U.S. Food and Drug Administration (FDA).

What did Recursion report for REC-4881?

Patients took 4 mg of REC-4881 once a day for 12 weeks, then stopped for 12 weeks. The update includes adults as young as 18, after the trial lowered its age limit from 55. The data cutoff was September 10, 2026. Recursion's SEC filing defines polyp burden as the sum of the diameters of all polyps, checked by endoscopy at the start and at weeks 13 and 25.

For the whole digestive tract, the company reported:

• 13 of 15 patients (87%) had less polyp burden after 12 weeks of treatment.
• The median reduction was 47%, and 5 of 15 patients had a cut of 65% or more.
• At week 25, after 12 weeks off the drug, 11 of 13 (85%) were still below their starting level.

The filing counts 16 efficacy-evaluable patients, meaning they had measurable polyps, took at least 75% of the drug and had a later endoscopy. Recursion's investor presentation says one missed the week 13 endoscopy, leaving 15. Only 13 had week 25 results by the cutoff.

For the upper digestive tract, where two of the 16 had no polyps at the start, it reported:

• 92% of 13 evaluable patients had less polyp burden after 12 weeks, with a median cut of 55%.
• 6 of 13 had a cut of 65% or more, and two had cuts of 98% or more.
• At week 13, 6 of 12 improved by at least one stage on the Spigelman scale, which grades polyps in the duodenum.
• At week 25, 11 of 12 (92%) were still below their starting level.

Recursion says the 47% median is consistent with the 43% it reported in December 2025 from 12 patients.

What side effects did patients have?

Side effects are graded by severity, and higher grades are worse. Of 26 patients assessed for safety, 21 (80.8%) had drug-related side effects of Grade 2 or lower, the company said. These included acne-like rash, diarrhea and raised levels of the enzyme creatine phosphokinase (CPK). Some had drops in LVEF, a measure of heart function. Recursion says these caused no symptoms and were reversible, though one patient was lost to follow-up.

Four patients had Grade 3 side effects, two involving skin rash and two involving raised CPK. None of those four stopped treatment. Four patients did stop because of adverse events, but none because of a Grade 3 drug-related event, the company says. No Grade 4 or 5 drug-related events were reported. Recursion says the profile is consistent with blocking MEK1 and MEK2, the proteins the drug targets.

What is FAP, and why do upper GI polyps matter?

FAP is caused by changes in the APC gene, according to MedlinePlus Genetics from the U.S. National Library of Medicine. People with the classic form can have hundreds to thousands of colon polyps. At least 90% develop colorectal cancer by age 50 unless the colon is removed first. MedlinePlus puts the incidence at 1 in 8,500 people.

Removing the colon does not stop new polyps from forming, Recursion notes. It says about 90% of people with FAP develop adenomas, a type of polyp, in the duodenum, part of the small intestine. It estimates that more than 50,000 people in the U.S. and five large European countries have FAP. It calls upper GI disease the greatest unmet need, citing risks of bleeding and perforation when polyps are removed.

TUPELO enrolls adults who have had all or part of the colon removed, its ClinicalTrials.gov record shows. As of October 2026, the registry lists it as recruiting, with an estimated 67 participants. Data collection for its main goals is expected to end in June 2027. Every group in the trial's second part gets REC-4881, with no placebo group.

How did AI point Recursion to this drug?

The molecule came from Takeda, which had tested it in solid tumors, Recursion said in December 2025. Recursion licensed it after screening thousands of compounds on human cells lacking working APC. Machine learning compared more than a thousand visual features of diseased and healthy cells. REC-4881 was among the strongest at pushing the cells back toward a healthy state, the company says.

So the AI claim concerns a new use for an existing molecule, not a new design. For background, see how AI drug discovery works. The AI-discovered drugs tracker follows REC-4881 and similar candidates.

What else did Recursion announce on October 15?

The same update included early results for REC-1245, an experimental cancer drug designed to break down a protein called RBM39. Recursion says its platform found that target. In the dose-finding part of the Phase 1/2 DAHLIA trial, three patients had partial responses, a standard measure of tumor shrinkage. The data cutoff was August 21, 2026. Two of the responses were confirmed. All came at daily doses of 45 mg or more, in heavily pretreated patients with endometrial or liver cancer. Recursion chose 45 mg for the next stage. At 65 mg, three of four patients assessed had dose-limiting side effects.

Sanofi also picked the first drug candidate from its Recursion partnership, an oral drug for inflammatory disease. The choice triggered a $10 million payment.

Recursion priced 61,540,000 shares at $3.25 each, for gross proceeds of about $200 million, its pricing release says.

What is still unknown, and what comes next?

Recursion's own release warns that interim data from a small single-arm study may not predict final results. Reviewing the December 2025 data, the trade publication Drug Discovery & Development called them hypothesis-generating rather than definitive. It asked how much evidence of lower cancer risk regulators will require for a preventive drug.

REC-4881 has Orphan Drug and Fast Track designations from the FDA. Fast Track is a process meant to speed the development and review of drugs for serious conditions. Neither designation is an approval.

The next dates are:

• November 2, 2026: Lisa Boardman, M.D., of Mayo Clinic is set to present fuller TUPELO data at the CGA-IGC annual meeting in Denver, in a Presidential Plenary session.
• Later in 2026: Recursion says it will share more on the registrational trial. It also expects more REC-1245 Phase 1 data.
• Second half of 2027: Recursion's target for starting the registrational trial, pending the FDA talks. Recursion also expects early REC-1245 data in endometrial cancer then.