REC-4881

REC-4881 is Recursion Pharmaceuticals' oral MEK1/2 inhibitor in Phase 1b/2 testing (4 mg QD) for Familial Adenomatous Polyposis (FAP), a genetic disorder where patients carry an APC gene mutation and grow hundreds of pre-cancerous polyps in their colon by their teens. Without surgery, nearly all develop colorectal cancer. The molecule itself is not new — it was originally TAK-733 from Takeda, which ran a Phase 1 dose-escalation study in 51 patients with advanced solid tumors and saw only one partial response (a BRAF-mutant melanoma patient) before being shelved [1]. Recursion in-licensed it and ran it through their phenotypic screening platform (testing compounds on living cells and watching what happens, rather than designing drugs to hit a pre-specified protein target), which flagged FAP polyposis as a hit. The TUPELO trial (NCT05552755) reported in December 2025 that 75% of 12 efficacy-evaluable patients showed polyp burden reduction at 12 weeks (median 43% drop), with the median deepening to 53% at Week 25 — 12 weeks after stopping therapy — in 11 patients with follow-up [3]. REC-4881 carries FDA Orphan Drug Designation (granted September 2021) and FDA Fast Track Designation (granted April 2022) for FAP, plus EU Orphan Drug Designation [7][8]. Recursion plans to discuss a registration path with FDA in H1 2026 [3]. The company ended Q4 2025 with $753.9M in cash and equivalents and guided to runway into early 2028 [4], which removes the near-term financing-risk overhang. The stakes: a chronic oral therapy that prevents polyps could replace or delay prophylactic colectomy in thousands of FAP patients globally. The catch: MEK inhibitors have a chronic toxicity profile (rash, diarrhea, retinal and cardiac signals) that has never been validated for years-long preventive dosing.

Not a novel compound. REC-4881 is TAK-733, originally developed by Takeda for BRAF/RAS-mutant solid tumors. The Phase 1 dose-escalation in 51 patients produced one partial response (a BRAF-mutant melanoma) and Takeda discontinued development for inadequate exposure at tolerated doses [1]. Recursion in-licensed it and ran it through their cellular imaging platform to find new indications; they landed on APC-mutant polyposis. The lead trial, TUPELO (NCT05552755), is registered as Phase 1 with an expansion cohort that Recursion communicates as Phase 1b/2 [2][3]. Enrollment target is 67 FAP patients, recruiting [2]. The primary registry endpoint is pharmacokinetic — Cmax dose-finding (Cmax is the peak drug concentration in blood, a standard PK measure of how much drug is absorbed) — with polyp burden change at 12 weeks as a secondary clinical readout. Recursion presents the polyp data as the headline result [3]; that mismatch between the registered primary and the publicly communicated headline matters for any registration discussion. The December 2025 update reported the 75% / 43% polyp signal at Week 13 in 12 efficacy-evaluable patients on 4 mg QD, with a 53% median reduction maintained at Week 25 in 11 patients (12 weeks after stopping therapy) — a durability finding [3]. Management has guided to FDA registration-path discussions in H1 2026 [3]. REC-4881 holds FDA Orphan Drug Designation (granted September 29, 2021), FDA Fast Track Designation (granted April 2022), and EU Orphan Drug Designation, all for FAP [7][8] — earlier draft language saying these designations were unconfirmed was wrong. Worth flagging: calling REC-4881 itself 'AI-designed' overstates the case. The molecule predates Recursion's platform; the FAP indication is what their phenotypic screening surfaced.

FAP starts with a broken brake. The APC gene normally tags β-catenin for destruction, keeping the Wnt growth-signal pathway quiet. When APC is mutated (every FAP patient inherits one bad copy and loses the second somatically in cells that become polyps), β-catenin builds up, turns on growth genes, and intestinal cells start dividing into adenomas. Why MEK? MEK1 and MEK2 are kinases in the MAPK growth pathway, sitting downstream of RAS and RAF — the same pathway activated by KRAS mutations in most colorectal cancers (canonical chain: RAS → RAF → MEK → ERK). Standard textbook biology says APC loss is a Wnt problem, not a MAPK problem, so blocking MEK shouldn't obviously help. Recursion's case is that in their phenotypic screen of APC-mutant cells, MEK inhibition rescued the abnormal cell behavior, a finding consistent with newer literature showing crosstalk between Wnt and MAPK in intestinal stem cells. That's a thinner mechanistic story than, say, BRAF/MEK combos in melanoma, where the genetics scream the pathway. The 75% / 43% polyp reduction signal from TUPELO is the strongest validation we have [3]. The flip side: MEK inhibitors as a class are well-studied. Trametinib, cobimetinib, binimetinib, and selumetinib are all approved, all hit the same node, and all have predictable toxicity. The question is not whether MEK inhibition does something in polyps (early data say it does) but whether you can dose it safely for years in patients who would otherwise just have surgery.

TUPELO (NCT05552755) is a single-arm, open-label dose-escalation study in adults with classical FAP, sponsored by Recursion [2]. Target enrollment is 67. Per the registry, the primary endpoint is pharmacokinetic (Cmax — peak plasma concentration, used here for dose selection), with polyp burden change at 12 weeks as a secondary clinical readout. Recursion presents the polyp data as the headline [3]. No comparator arm. Eligibility includes patients with retained colon or rectum/pouch, which broadens the recruitment pool but complicates polyp counting (rectal-pouch polyps behave differently from colonic polyps). The 12-week endpoint is short for a prevention indication; FDA has historically wanted longer durability data and recurrence assessments for FAP drugs, although the Week 25 off-treatment follow-up showing maintained 53% reduction begins to address that [3]. Context on the FAP treatment landscape: there is currently no FDA-approved medical therapy for FAP polyp burden. Sulindac, a non-selective NSAID, has decades of off-label use for polyp suppression in FAP patients with residual colon, with roughly 40–50% polyp burden reduction in small trials but no FDA approval and chronic GI/renal toxicity at the doses required. The DFMO/eflornithine + sulindac combination has been studied for chemoprevention. Celecoxib was the only drug ever FDA-approved for FAP polyposis (1999), and Pfizer withdrew that indication in 2011 after cardiovascular signals emerged in colorectal cancer prevention trials [5]. This regulatory vacuum is both the opportunity (no approved medical therapy for ~50,000 FAP patients across US and EU5) and the source of uncertainty about what evidence FDA will require — there is no recent precedent for what counts as registrational evidence in FAP. A single-arm design is workable for a rare disease, but the FDA discussion will likely turn on whether they accept polyp burden reduction as a surrogate without a placebo-controlled confirmatory trial. Expanded cohort details and any randomized confirmatory plans have not been disclosed publicly [3][4].

Our model estimates a 7% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 13%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by a non-randomized design and its light or open-label blinding; it is held back by the sponsor's thin or weak approval record and weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Safety is the dominant risk. MEK inhibitors share a predictable toxicity bundle: acneiform rash, diarrhea, peripheral edema, decreased left ventricular ejection fraction, and retinal vein occlusion / serous retinopathy. In melanoma, patients accept these because the alternative is a fatal disease and treatment is typically months to a couple of years. FAP patients are largely healthy young adults whose alternative is surgery. The bar for chronic tolerability is therefore much higher than for any approved MEK indication. TUPELO safety reported to date is mostly Grade 1–2 with 15.8% Grade 3 events and no Grade ≥4, consistent with MEK1/2 inhibition class effects — but n is small (12 patients dosed at 4 mg QD), follow-up is short, and chronic-dosing retinal and cardiac signals typically emerge with cumulative exposure [3]. Efficacy risk is moderate. The 75% / 43% polyp signal at Week 13 and 53% median reduction at Week 25 (12 weeks post-treatment) is real but rests on 11–12 patients. Polyp burden can fluctuate, and short-term reduction does not guarantee maintained reduction at 12 months of continuous dosing — celecoxib showed an early polyp signal before its FAP label was withdrawn for unrelated CV reasons [5]. Regulatory risk is high and binary. FDA's H1 2026 feedback determines whether this is a 2027–2028 approval path or a 2030+ randomized-trial slog. Orphan Drug and Fast Track designations are in hand [7][8], but neither guarantees a single-arm registrational design. Commercial risk: FAP affects roughly 1 in 10,000 people; >50,000 prevalent patients across US and EU5. Payers will scrutinize chronic MEK inhibitor pricing against the one-time cost of colectomy. Recursion will need to make the case on quality of life, downstream surveillance savings, and rectal-pouch polyp control post-colectomy. Execution / financing risk is materially lower than earlier drafts suggested. Recursion reported $753.9M cash and equivalents as of December 31, 2025, with 2026 non-GAAP cash operating expenses guided below $390M and management projecting runway into early 2028 [4]. That covers the FDA interaction, the next data readout, and the start of any confirmatory work without forcing a dilutive raise. A Phase 3 program, if required, would likely still need a partner or non-dilutive financing.

Worth watching, with one specific trigger. The signal that matters in 2026 is the H1 FDA registration-path discussion (likely a Type B or Type C meeting — a formal FDA consultation where the agency gives binding feedback on the trial design needed for approval), followed by an updated safety and durability readout from TUPELO. If FDA accepts polyp burden reduction with the current dataset and Recursion holds a clean chronic safety profile through six months, this becomes a real Phase 3 candidate and Recursion gets a clinical-stage anchor to argue their AI platform delivered something tangible. If FDA demands a randomized confirmatory trial or any of the known MEK toxicities surface at chronic doses, the program either gets partnered out or pushed years right. This is also a referendum on Recursion's broader thesis. Post-Exscientia merger, RXRX has been heavily discounted by the market. A clean FAP path validates the indication-finding side of their platform (even if the molecule itself isn't AI-designed). The cash runway extension to early 2028 buys time for that story to play out [4]. Check back at: (1) any 8-K from Recursion in Q1/Q2 2026 mentioning FDA Type B/C meeting outcomes, (2) DDW 2026 (Digestive Disease Week, the major GI medicine conference) or ASCO 2026 abstracts with longer-term TUPELO data, (3) next 10-Q updates on cash runway and any FAP partnership conversations.