NeuroRx chief lines up Hail Mary for once-rejected Covid-19 drug
NeuroRx’s Covid-19 treatment was already shot down by the FDA once, and it failed the primary endpoint in a Phase IIb/III readout — but that won’t stop CEO Jonathan Javitt from lining up a Hail Mary pass to regulators.
Javitt gave Endpoints News a preview of Phase IIb/III data based on an initial 28-day endpoint on Monday afternoon. At 28 days, critically ill Covid-19 patients given NeuroRx and Relief Therapeutics’ aviptadil were 35% more likely to recover and be discharged from the hospital compared to those on a placebo, Javitt said. However, the p-value was 0.08 — falling well below the 0.05 p-value needed to demonstrate statistical significance, meaning there’s a greater chance that the results were random and not tied to the treatment.
“There’s 8 chances in 100 that the difference we observed could have been a random event versus 5 chances in 100,” Javitt said at the time, arguing that the “really strong signal” was median time to recovery and discharge.
At 28 days, the median time to get better and leave the hospital was 10 days shorter for the aviptadil patients, with a p-value of 0.006, Javitt said.
But early Tuesday morning, Javitt reached out with changes to the story. While the Phase IIb/III was initially OK’d as a 28-day study, NeuroRx added a 60-day endpoint in December. The 28 days was “not appropriate for critically ill patients with Covid-19, who are frequently maintained in ICU with advanced technologies well beyond this time point,” the company said in a statement.
Javitt said the FDA released new Covid-19 guidance on Monday changing the required time for measuring the endpoint of “alive and free of respiratory failure” to 60 days, making the Phase IIb/III readout an interim readout.
“The 60-day endpoint makes sense because too many people are still in the ICU at day 28 to know the true outcome, although the ZYESAMI patients are 38% more likely to have recovered and gone home. That means the curves have a good chance of diverging further between day 28 and day 60,” Javitt wrote in an email to Endpoints.
Rather than stating that those in the aviptadil arm were 35% more likely to recover and be discharged from the hospital, NeuroRx is now saying aviptadil patients had a “35% higher likelihood of recovery from respiratory failure with continued survival.”
“Should these trends continue through day 60, they have the potential to reach statistical significance,” the company said.
Updated guidance posted to the FDA’s website on Monday says the pre-specified endpoint should be measured after at least 28 days for hospitalized non-critically ill patients, and 60 days for critically ill patients. The FDA responded Wednesday confirming that the previous guidance “referred to ‘at least 28 days’ for severe and/or critically ill patients.”
This is the second time NeuroRx has changed their primary endpoint. Initially, the primary endpoints were blood oxygenation and mortality, according to a filing on clinicaltrials.gov.
In a subgroup of patients in tertiary hospitals, those in the aviptadil arm were 46% more likely to recover and return home before day 28, with a p-value of 0.058, also above the bar. When asked why the big difference in the tertiary subgroup, Javitt speculated it could have something to do with hospital capacity.
“It may well be that the difference we saw between the tertiary and the regional hospitals really could have been a difference between hospitals that were running below capacity and hospitals that were running 200% of capacity,” he said.
The only side effect observed was mild to moderate diarrhea in 30% of patients, Javitt said.
“In that context, to say ‘Well, you know, you shouldn’t give it because there’s 3 more chances in 100 that that measure could have been seen by chance, doesn’t seem very reasonable,” he initially said of the preview, before making the changes. “EUA is based on a standard of may be effective.”
Reasonable or not, the FDA has already turned down aviptadil once. NeuroRx and Relief submitted for an EUA in September based on a case-control study with 21 patients in the treatment arm and another group who received maximal standard-of-care in the same ICU.
The FDA denied their bid last month, causing Relief’s $RLFTF stock to plunge from about 40 cents on Dec. 29 to 31 cents on Dec. 30. On Monday, shares closed at 37 cents apiece.
Javitt says the first EUA submission was a “very narrow” one, for patients who would otherwise qualify under an expanded access protocol but didn’t have a way to get the drug. He said the company was running into situations where hospitals didn’t have the infrastructure to participate in an EAP.
After being denied an EUA, NeuroRx ended up treating some of those patients under Right to Try laws, he said. To date, the drug has been administered to about 520 patients total.
NeuroRx came under fire earlier in September after Politico reported it tapped a Republican congressman with close ties to its CEO and a history of mask skepticism to their Data and Safety Monitoring Board. Javitt confirmed that Rep. Andy Harris was hired to the DSMB, but denied that he is a mask skeptic.
Harris said at a House Appropriations subcommittee hearing back in June that there was a “cult of masks,” and questioned former CDC director Robert Redfield about the agency’s guidance to wear cloth masks, as opposed to surgical masks or face shields.
Javitt said on Monday that the last patient just finished 60 days of treatment, and that he expected to file for another EUA within a couple weeks. As part of the Tuesday announcement, he said they will file “should the above trends continue through day 60.”
Feb 23: A previous version of this article stated upon second reference that those in the placebo arm were 35% more likely to recover and be discharged in the hospital. A correction has been made to reflect that it was the aviptadil arm.
Feb 26: An update has been made to reflect that the FDA has confirmed the changes to its guidance.
For a look at all Endpoints News coronavirus stories, check out our special news channel.