Recursion reported mixed data Tuesday from its first Phase 2 study, in a disappointing start to an expected frenzy of clinical results.
The Salt Lake City-based biotech has forecast a stretch of about 10 readouts in 18 months, when including the pending acquisition of Exscientia. Venture capitalists and investors framed these readouts as “a moment of truth” for the buzzy biotech to prove its R&D ideas can actually deliver better medicines. In clearing that bar, Recursion’s first readout leaves something to be desired. The company shared few specifics and a muddled description of potential signals of efficacy.
Recursion tested a repurposed molecule called REC-994 in about 60 patients with symptomatic cerebral cavernous malformation, a rare neurological disorder that can cause strokes, seizures and partial paralysis.
The company said the drug succeeded in its primary goal of safety and tolerability, with a similar profile to a placebo group. But secondary endpoints looking for efficacy signals were less clear-cut. Neither the low dose nor high dose improved patient- and physician-reported scores after a year. Recursion said the high-dose group “showed a trend towards” fewer and smaller lesions in the brain, as measured by MRIs.
What was left out of the press release raises more questions. The release didn’t disclose specific data for the safety and efficacy measures. And while it described a trend for the high-dose group for fewer and smaller brain lesions, it didn’t mention how the low-dose group fared on that metric.
The release also made no mention of another secondary endpoint in the number of MRI-confirmed brain-bleeding events. A company spokesperson said they plan to eventually present and publish detailed results.
Recursion also said it “plans to advance development” of the compound, with R&D head Najat Khan calling the readout “an early proof-of-platform milestone.” It’s unclear if the company would advance to Phase 3 or run another Phase 2 trial. Khan added the results would help design the next study, particularly in study duration, dose level and number of patients.
Jan-Karl Burkhardt, a University of Pennsylvania neurosurgeon and principal investigator of the trial, said in the press release the readout “strongly supports the need for a future study, with a longer duration and a larger patient cohort.”
Wall Street was less enthused by the results, sending shares down about 10% Tuesday morning. Recursion’s stock price $RXRX has fallen nearly 80% since its 2021 initial public offering.
Chris GibsonWhile Recursion is a bellwether for the AI-focused biotech space, its lead drug programs have no connection to the latest AI technologies that have invigorated the field. REC-994, for instance, is a decades-old compound called tempol. As part of his PhD research, Recursion CEO Chris Gibson identified tempol using publicly available machine learning tools to search through compounds.
Recursion’s next two Phase 2 readouts are also testing in-licensed molecules rather than AI-designed drugs. The company expects results for one drug licensed from The Ohio State University in the fourth quarter of 2024 and another molecule licensed from Takeda in the first half of 2025.
Recursion’s first internally designed drug, REC-3964, is expected to start a Phase 2 study in C. diff infections by year’s end.