Recursion's FAP Day: The Clinical Data Arrives, the FDA Question Remains

A patient-heavy Q3 call shifts Recursion's keyword map from AI platform to polyp burden and Whipple procedures, but the registrational path is still the same open question.

RXRX · Earnings Call · 2026-11-03

The earnings call that became a FAP day

Recursion's Q3 call was not a normal quarterly update. Management hosted a mini clinical symposium inside it: two FAP specialists, two patients and a caregiver, and a fresh cut of the TUPELO data. The company's keyword map flipped accordingly. Polyp burden is now the No.1 company keyword, with Upper GI, Whipple procedure, Endometrial cancer and even patient names crowding the top 30. In the prior quarter, the top keywords were Roche Genentech, PI3 kinase, product engine and data factory. That is a genuine narrative pivot: from "look at our AI platform" to "look at the patients." Najat Khan framed it as the next proof point: “The higher bar we set for ourselves is repeatability. Not one target, not one molecule, or one clinical result...” — Najat Khan, Executive / Investor Relations · 2026-11-03 The stock is not yet voting for it. Recursion is a $2.2 billion market-cap biotech, still down roughly 80% since its 2021 listing and 91% from its peak; the last 90 days are flat, and the shares sit 23% below the October high. A clinical narrative is arriving into a badly broken tape.

Clinically meaningful, regulatorily unresolved

The FAP update is the strongest data Recursion has shown. In TUPELO, “After 12 weeks of treatment, 87% of evaluable patients had a reduction in total polyp burden.” — Najat Khan, Executive / Investor Relations · 2026-11-03 About one in three had a 65% or deeper reduction; roughly 85% of patients remained below baseline 12 weeks after stopping drug. Upper GI reduction was seen in 92% of evaluable patients, with median reductions around 55% upper and 50% lower; larger high-risk polyps also fell, and six of 12 evaluable patients downstaged on Spigelman score. For a disease with no approved drug, that is real clinical signal. But the call's own KOL supplied the cold water:

...unfortunately, that polyp burden is not a regulatory endpoint that's typically used. But from the clinician's standpoint, it's incredibly meaningful.
Bryson Katona, Doctor / Clinical Expert · 2026-11-03

That is the central tension. Management is still "in ongoing discussions with the FDA" and now targets registrational study initiation in the second half of 2027. The same uncertainty was on the prior two calls. In August, Vicki Goodman said “there's very limited regulatory precedent in FAP” — Vicki Goodman, Chief Medical Officer · 2026-08-05 and flagged the primary-endpoint question; in May, Najat Khan listed the open items as “patient population, endpoint that has clinically meaningful benefit, and then, of course, dose and dose escalation” — Najat Khan, CEO · 2026-05-06. The new data improves the odds; it does not yet answer the FDA question. REC-1245 is earlier still: three partial responses in 12 response-evaluable patients, two confirmed in endometrial cancer and one unconfirmed in HCC, a 91.7% disease-control rate at eight weeks, and a 45 mg dose selected for expansion. Vicki Goodman was careful: “RBM39 is novel biology. We're potentially first in class.” — Vicki Goodman, Moderator / Panel Host · 2026-11-03 Expansion data are guided to the second half of 2027. This is a 2027-2028 story, not a 2026 catalyst.

The cash runway is the real product

The financial update is cleaner than the clinical one. Recursion raised $200 million in an oversubscribed October offering; Ben Taylor said “Over 90% of the financing came from healthcare-dedicated specialists and mutual funds that did not own Recursion prior to the financing.” — Ben Taylor, Executive / Finance · 2026-11-03 Pro forma cash was $749 million at September 30, extending runway into 2029. Management reiterated 2026 cash operating expense below $375 million and guided 2027 below $360 million. That is real discipline. But the P&L still looks like a science project: total revenue was just $8 million in the July quarter, down 60% year over year, while operating income was -$135 million and R&D was $90 million, down 30% year over year. The balance sheet is the asset that matters: effective net cash was $552 million at July quarter-end, before the October raise. The equity still trades at 35.7x trailing revenue, even after an 87% de-rating in that multiple over the past year. The prior call promised the same $375 million number while doing “the same amount or more with less” — Ben Taylor, Chief Financial Officer · 2026-08-05; the new call adds a 2027 step-down. Partnerships are now the non-dilutive oxygen: more than $535 million of cash inflows, nine milestones and over $100 million in roughly two years, including Sanofi's $10 million option payment. Prior management had described milestone guidance as probability-weighted rather than certain — “we actually look at all of the programs that we know and we probability weight them” — Ben R. Taylor, Chief Financial Officer · 2025-08-05 — and that caution still applies.

What the tape and keyword map are saying

The global tape is not helping the "AI drug discovery" narrative. The market's AI enthusiasm is concentrated in infrastructure, and even that complex is wobbling: AI data center names are down about 21% over the last 30 days in the tape, with 55 decliners versus 10 advancers. Recursion is not trading with that group. Its own keyword map is now dominated by clinical and patient terms, while the prior platform/partnership vocabulary has receded. That is the right pivot if the data hold: a company that was valued as an AI story is trying to become a clinical-stage rare-disease and oncology story with a cash runway long enough to reach the readouts. The appointment of Fred Hassan as chair — a former Schering-Plough and Pharmacia CEO with launch experience — reinforces that the next chapter is supposed to be approval and commercialization, not another platform demo. The risk is equally clear: if the FDA does not accept a polyp-burden-based path, or if the REC-1245 expansion data are thin, Recursion has bought time but not a re-rating. The stock's 91% drawdown says the market has already taken the platform premium out; the Q3 call is the first credible attempt to replace it with clinical evidence.