Insilico Medicine reported Phase 2a results for its lead program, saying the study achieved its primary endpoint of safety and tolerability. But the efficacy data it released were much more vague.
In a 71-patient trial testing ISM001-055 in idiopathic pulmonary fibrosis (IPF), Insilico said Wednesday that the drug improved a measurement of breathing called forced vital capacity compared to placebo after 12 weeks, and the responses were dose-dependent. However, Insilico did not report the absolute scores of each trial arm and whether the difference was statistically significant.
In an interview Tuesday morning, Insilico CEO Alex Zhavoronkov declined to comment on the program’s statistical significance, saying he was not legally cleared to do so. He instead pointed to a trend identified by Insilico’s researchers across dosing arms as evidence the program works.
“It depends on what kind of statistics you use,” Zhavoronkov said of ISM001-055’s statistical significance. “Currently, we don’t have enough patients to claim that it is statistically significant for approval, for example. But if you look at the data, it becomes very clear that there is a trend and there is a response.”
If Insilico had run a longer trial specifically looking for efficacy, he said the dose response would have continued to increase. There’s no timeline yet for the program’s next steps, though Zhavoronkov said the company’s most immediate tasks are to submit the results to a medical journal and talk to regulators.
ISM001-055 is a small molecule targeting TNIK, short for Traf2- and Nck-interacting kinase. Insilico said it discovered TNIK through its AI software and that inhibiting the target can halt or reverse fibrotic processes.
Insilico is one of several biotechs aiming to speed up drug development through the use of artificial intelligence, and the readout comes a few weeks after Recursion — another AI player — also unveiled its first Phase 2 data. Some investors and VCs had framed the Recursion data drop as a “
moment of truth
” for the company and, by extension, the field after the first wave of AI-designed drugs
fell short in the clinic
.
But Recursion’s data were too short on specifics for the public markets, and the startup’s stock price closed 16% lower the day it announced the data. In a rare neurological disease called cerebral cavernous malformations (CCM), Recursion’s drug, REC-994, succeeded on safety and tolerability endpoints but provided mixed results on efficacy. Recursion plans to advance the program, though it’s not yet clear if it will run a Phase 3 study or another Phase 2 trial.
Zhavoronkov resisted comparing Insilico’s results to Recursion’s, citing both the differences in program design and diseases. He said IPF is a much more heterogeneous disease than CCM. While he said he respects Recursion and its CEO Chris Gibson, Zhavoronkov said people he spoke to were surprised the company didn’t terminate REC-994.
“I don’t want to even compare with Recursion because it’s a different story,” Zhavoronkov said. “It’s not apples to apples. It’s like apples to seeds.”