AstraZeneca on Wednesday revealed new results from a large U.S. trial of its coronavirus vaccine, responding to a rare rebuke from U.S. officials who publicly questioned initial findings the British drugmaker reported earlier this week.
The updated data, delivered in a press release late Wednesday evening, confirm the shot's protective benefit and should be enough to support clearance for emergency use in the U.S. According to AstraZeneca, the more recent results show the vaccine to be 76% effective against symptomatic COVID-19, and strongly protective against severe disease.
But the controversy over AstraZeneca's earlier statement, which National Institutes of Allergy and Infectious Diseases director Anthony Fauci called an "unforced error," was the latest in a series of communication missteps that have damaged AstraZeneca's credibility and perception of its shot.
"I think the vaccine is better than what it's getting credit for," said Taison Bell, an assistant professor of medicine at UVA Health, and "some of that is due to the bungled rollout when it comes to the press releases [from AstraZeneca]."
On Monday, AstraZeneca said its shot was 79% effective at preventing symptomatic COVID-19 in the study of 32,449 volunteers. Vaccination was equally effective regardless of age or ethnicity, the company said, and protected against severe disease and hospitalization. No serious safety issues were reported, an important finding given reports of rare but serious blood clotting events that caused several European countries to suspend use of the shot.
The results seemed a much needed win for AstraZeneca, an apparently clear-cut result from a rigorous trial after confusing findings from earlier tests that showed the vaccine, on average, 70% effective against COVID-19.
But AstraZeneca didn't specify how many cases occurred in the vaccine and placebo groups, nor when the results were from in its announcement. Those omissions proved important when, later that evening, the data safety monitoring board — a group of 11 independent experts appointed by the National Institutes of Health to oversee the trial — sent what Fauci described in a television interview as a "rather harsh note" to the company.
The note itself wasn't made public, but the gist of it was revealed in a highly unusual statement from the NIAID, which said AstraZeneca used "outdated information" in its announcement. The Washington Post, citing the DSMB's letter, reported the committee wanted AstraZeneca to report a result in the range of 69% to 75% to account for COVID-19 cases that had yet to be confirmed.
The DSMB was concerned the data "was incomplete," said Larry Corey, a virologist and former president of the Fred Hutchinson Cancer Research Center, in an e-mailed statement on Tuesday.
Corey, who co-leads the COVID-19 Prevention Trials Network, a network of studies that includes AstraZeneca's trial, noted the DMSB's response "illustrates the importance of [its] essential role in providing independent evaluation of the process of how data is released and interpreted."
AstraZeneca responded with a statement clarifying its numbers were based on an interim analysis as of Feb. 17. The fuller results released Wednesday are from a primary analysis with more recent follow-up.
But the episode has seeded further doubt about a shot that's suffered as much from mistakes in messaging as anything else.
"Clearly AstraZeneca and the NIH were not on the same page. It's just not helpful," said Walid Gellad, an associate professor at the University of Pittsburgh and a director of the Center for Pharmaceutical Policy and Prescribing there, in an interview.
For a time, AstraZeneca's vaccine was viewed as a frontrunner and likely to be the first to reach the public, leading the U.S. to preorder 300 million doses last summer. But delays over safety concerns, along with uncertainty introduced by confusing data readouts, have held up the shot's progress.
Days after the U.S. trial first started, for instance, testing was temporarily paused for a safety review when a participant in a U.K. study developed an unexplained neurological illness. Though testing around the world restarted quickly, AstraZeneca provided few details publicly and was reportedly slow to follow up with the FDA. The pause dragged on in the U.S. for almost two months before the FDA allowed the trial to proceed, causing AstraZeneca to fall behind rivals Pfizer, Moderna and Johnson & Johnson. All three have since had their vaccines cleared for use in the U.S.
In December, AstraZeneca reported perplexing data pooled from a group of late-stage studies in the U.K. and elsewhere. The company claimed the shot was more effective in a small group of participants who had mistakenly received a mix of lower and higher doses. But the explanation shifted in the following months, with protection apparently superior in those who had received two full doses three months apart.
Over that time, questions emerged about the shot's effectiveness against concerning coronavirus variants and whether the vaccine was causing rare but serious blood clots in some immunized patients in Europe. (European and U.K. drug regulators have said the shot's benefits outweigh its risks, and vaccinations have since restarted in most countries.)
The U.S. trial, AstraZeneca's largest, was viewed as an opportunity to clear up some of those questions, and largely did so, showing the vaccine to be safe and protective in younger as well as older adults. But the unusual back-and-forth between the company and NIAID has injected new uncertainty and showed the partners to be jarringly out of sync.
The findings AstraZeneca delivered late Wednesday were reassuringly consistent, making it unclear why the DSMB took such a strong stance in criticizing the initial disclosure.
"NIH looks forward to the review by the U.S. Food and Drug Administration," an NIH spokesperson responded in an e-mailed statement to Biopharma Dive.
AstraZeneca said the primary analysis was based on 190 symptomatic cases of COVID-19, 49 more than it included in Monday's interim result. The drugmaker still hasn't disclosed how many of those cases occurred in the vaccine and placebo groups, however.
There were eight cases of severe COVID-19 among volunteers given placebo, compared to zero in the vaccine group. And while the updated numbers reduced overall efficacy from 79% to 76%, estimated protection in the elderly was boosted from 80% to 85%.
AstraZeneca said the numbers may still "fluctuate slightly" because of 14 "possible or probable" COVID-19 cases that need to be evaluated. The confidence interval — a statistical measure to assess the range of likely outcomes in a dataset — around the 76% figure reported by AstraZeneca Wednesday was relatively wide, from 68% to 82%.
AstraZeneca plans to submit this data to the FDA for emergency use authorization in the "coming weeks," potentially setting up a decision in the early spring. As part of its review, the agency will independently analyze and report on AstraZeneca's data, before convening a group of outside experts for additional evaluation.
"In the end, it seems like a decent vaccine that's going to save lives," said Pittsburgh's Gellad.
Ned Pagliarulo contributed reporting. Note: This story has been updated with comment from the National Institutes of Health.