Can an AI-Discovered Pill Finally Spare FAP Patients the Scalpel?

DENVER, CO – November 02, 2026 — For decades, a diagnosis of familial adenomatous polyposis (FAP) has come with a grim, unavoidable guarantee: a lifetime of invasive surgeries. Patients with this rare genetic disease face a near-100 percent lifetime risk of colorectal cancer due to a mutation in the APC gene. The standard of care is as brutal as it is effective—removing the colon entirely. But colectomy does not stop the disease. Polyps continue to aggressively colonize the remaining upper and lower gastrointestinal tract, requiring relentless endoscopic surveillance and, inevitably, more organ removal.

Now, an oral therapy discovered through artificial intelligence is showing tangible promise in disrupting this surgical cycle. At the Collaborative Group of the Americas on Inherited Gastrointestinal Cancer (CGA-IGC) Annual Meeting, Recursion Pharmaceuticals presented updated Phase 2 data from its TUPELO trial for REC-4881. The results suggest that we may finally be inching toward a pharmacological intervention for a disease that currently has zero FDA-approved drugs.

But as we strip away the persistent hype surrounding AI in drug discovery, the real story lies in the clinical data, the nuances of chronic drug toxicity, and the steep regulatory hurdles that remain before this pill reaches the pharmacy counter.

Beyond the Algorithm: Clinical Efficacy in the Real World

Recursion, a clinical-stage biotechnology company valued at over $2.4 billion, built its reputation on the "Recursion OS"—a platform that merges automated wet-lab biology with machine learning to identify therapeutic targets. In the case of REC-4881, unbiased phenotypic screening revealed that inhibiting the MEK1 and MEK2 pathways could effectively reverse the cellular chaos caused by APC gene loss.

AI may have found the molecule, but human trials are the ultimate arbiter of truth. The updated TUPELO results, presented by Dr. Lisa Boardman of the Mayo Clinic, offer a compelling look at REC-4881’s efficacy.

In a cohort of 15 evaluable adult patients, 87 percent achieved a reduction in their total polyp burden after 12 weeks of treatment, with a median reduction of 47 percent. Even more striking is the drug's impact on larger, high-risk polyps—those greater than 5 millimeters, which are most likely to harbor high-grade dysplasia or cancer. Six of eight patients with these larger polyps saw their numbers fall, and two patients watched their counts plummet from more than 90 high-risk polyps to a single remaining lesion.

Perhaps the most critical finding is the drug's durability. Twelve weeks after stopping treatment, 85 percent of evaluable patients maintained a polyp burden below their original baseline.

“The durability of a reduction in polyp burden is an important metric to understand for any type of chemopreventive agent in FAP,” said Dr. Bryson Katona, Director of the GI Cancer Genetics Program at the University of Pennsylvania Perelman School of Medicine and a TUPELO investigator. “The dream chemopreventive agent for FAP is one that a patient doesn’t have to take lifelong and that potentially could be taken for a shorter amount of time or with an intermittent dosing pattern.”

The Chemoprevention Conundrum

The concept of intermittent dosing is not just a convenience; it is a clinical necessity when dealing with MEK inhibitors.

Historically, the graveyard of FAP chemoprevention is littered with failed therapies. Agents like celecoxib (a COX-2 inhibitor) and sulindac (an NSAID) demonstrated the ability to reduce polyp burden, but their long-term use was derailed by severe cardiovascular, gastrointestinal, and renal toxicities. The FDA’s Oncologic Drugs Advisory Committee has consistently struggled with the risk-benefit math of FAP treatments: How much chronic toxicity is acceptable to prevent cancer in a patient who is currently cancer-free?

REC-4881 is an MEK1/2 inhibitor, a class of drugs well-characterized in oncology for side effects like acneiform rash, diarrhea, creatine phosphokinase (CPK) elevation, and decreases in left ventricular ejection fraction. In the TUPELO trial's safety cohort of 26 patients, 81 percent experienced Grade 1 or 2 treatment-related adverse events. Four patients (15 percent) experienced Grade 3 events, and four patients discontinued the trial entirely due to low-grade toxicities.

While gastroenterologists observing the trial note that a Grade 1 or 2 rash might seem trivial compared to a cancer diagnosis, they point out that asking a patient to endure chronic, daily side effects for decades is a tall order.

This is why the durability data is the linchpin of Recursion’s strategy. If REC-4881 can be administered in a "12 weeks on, 12 weeks off" cycle, it fundamentally alters the quality-of-life equation, allowing patients periods of recovery from the drug's toxicities while keeping polyp growth suppressed.

Targeting the Upper GI: The Highest Unmet Need

For post-colectomy patients, the upper gastrointestinal tract is the primary battleground. Nearly all FAP patients develop adenomas in the duodenum. Because the duodenal wall is incredibly thin, removing polyps endoscopically carries a high risk of bleeding and catastrophic perforation. When endoscopy fails, the last resort is a Whipple procedure—a massive, life-altering surgery that removes parts of the pancreas, intestine, and stomach.

Recursion’s data shows particular strength in this precise area. In the upper GI tract, 92 percent of patients saw reductions, and half improved by at least one Spigelman stage, the clinical grading system used to determine when a patient requires surgery.

“REC-4881 is reducing the larger polyps that often drive much of the surgery for these patients, across the GI tract from a single oral therapy,” noted Najat Khan, President and CEO of Recursion. “This is a lifelong disease with no approved pharmacologic therapies, and REC-4881 is the clearest proof point yet of what we built Recursion to do.”

The Regulatory Reality Check

Despite the optimism, pragmatic observers must acknowledge the limitations of the data. The TUPELO efficacy findings are based on just 15 patients. Furthermore, it is a single-arm trial with no placebo control.

To contextualize the results, Recursion partnered with the Amsterdam University Medical Center to analyze a natural history registry of FAP patients, finding that matched untreated patients saw their upper GI polyp burden increase by 33 percent annually. While this historical comparison strongly suggests REC-4881 is actively shrinking polyps, the FDA considers historical controls to be notoriously vulnerable to bias. Changes in standard of care and patient selection can skew comparisons, meaning this Phase 2 data, while highly encouraging, is not definitive.

The true test will come in the second half of 2027, when Recursion plans to initiate a registrational trial. The company is currently in discussions with the FDA to align on a regulatory path. The primary sticking point will likely be the selection of a surrogate endpoint. The FDA will demand rigorous proof that reducing polyp size and number directly translates to a delay in major surgeries or a reduction in cancer incidence.

As AI continues to dominate headlines with promises of revolutionizing medicine, REC-4881 stands out as a concrete example of technology translating into biology. It is a reminder that while algorithms can identify the lock and design the key, it is the patient—navigating the delicate balance between a drug's side effects and the threat of the scalpel—who must ultimately walk through the door.