As I reported back in my November 11th column in Taki’s Magazine, Pfizer shut down lab processing of its world-historical vaccine clinical trial from late October until the day after the election to avoid having to follow its published protocol on when to assess its results. Now, what I gleaned from a close reading of StatNews has been confirmed by a New York Times article:
The furious race to develop a coronavirus vaccine played out against a presidential election, between a pharmaceutical giant and a biotech upstart, with the stakes as high as they could get.
By Sharon LaFraniere, Katie Thomas, Noah Weiland, David Gelles, Sheryl Gay Stolberg and Denise Grady
Published Nov. 21, 2020
Updated Nov. 30, 2020
… In a double-blind clinical trial, the gold standard for testing new medicines or vaccines, neither the company nor the participants know who receives vaccines and who gets placebos. Only an independent review board has access to that information.
The protocols for the trials lay out under what conditions the board can look at the results. Pfizer’s trial protocol was the most aggressive of all six vaccine candidates, allowing for a check of interim results once 32 participants developed Covid-19.
Pfizer’s published protocol called for interim analyses when the number of cases across the two arms reached 32, 62, 92, and 120, with a final analysis after 164 cases (see pp. 102-104.)
Pfizer wound up blowing off its first two called-for analyses without disclosing their change of plans to investors, causing some stock analysts to issue cautionary reports.
It was a lower benchmark than Moderna and the other companies had adopted and F.D.A. regulators warned Pfizer they were highly unlikely to issue an emergency use authorization for a coronavirus vaccine based on such a small data set.
But that’s conflating Pfizer’s announcement of efficacy vs. the FDA’s approval of safety, which are two different issues. As it turned out, Pfizer announced the spectacular efficacy of its vaccine on the morning of Monday, November 9th, even though it did not yet have the required 2 months of safety data to officially apply for FDA approval until the second half of November (and the FDA meeting won’t be until Thursday, December 10).
Outside experts criticized Pfizer for allowing itself to peek at the data too early and too often.
But Pfizer laid out the precise math it would follow in declaring whether the vaccine was likely to meet the FDA’s 50% efficacy hurdle or should be declared futile:
Pfizer decided it should drop that first benchmark and asked the F.D.A. to approve a new protocol on Oct. 29, effectively dashing Mr. Trump’s hopes of an announcement before Election Day.
The company also stopped processing test results from trial participants while it worked with the F.D.A.
That’s the smoking gun.
Protocol changes are discouraged once data is available, and Pfizer did not want to cross the benchmark before it got verbal approval to drop it.
In other words, Pfizer crossed its published benchmark before Election Day, but didn’t want to have to announce its results, so it shut down its lab work on the world’s most important project.
That approval came on Nov. 3, Election Day, or the day after.
The delay later enraged Mr. Trump, who claimed it was part of a conspiracy to damage his chance at re-election. Dr. Bourla said he turned a blind eye to the Election Day deadline.
“Before, people were saying it’s too soon,” he said. After the election, “people are saying, ‘Oh, it’s too late.’”
Once Pfizer resumed processing test swabs on Nov. 4, it quickly became apparent that the infection rate had skyrocketed, as it had nationwide and in other countries.
With 94 Covid-19 cases, the company asked the data monitoring board to reveal the results.
In other words, during its lab shutdown, Pfizer blew past both its first three disclosed interim analyses (32, 62, 92).
On Sunday, November 8th, the independent board called the Pfizer CEO with the results of the blind trial:
Then everyone was ushered from the room except Dr. Bourla and Pfizer’s general counsel, Doug Lankler, so the two men could hear a breakdown of the data that showed the vaccine was more than 95 percent effective. Of 94 people who had gotten sick, they were told, 90 were in the placebo group and only four were in the vaccine group.
With only 4 COVID cases in the vaccine arm out of the first 94 cases blowing away the third interim analysis’s threshold of no more than 25 cases in the vaccine arm out of the first 92 cases overall, we know that Pfizer’s vaccine would have been declared efficacious in either the second interim analysis (62 cases overall with no more than 15 in the vaccine arm) or the first interim analysis (32 cases overall with no more than 6 in the vaccine arm). Even if all 4 vaccine arm cases were among the first 32, that still would be better than Pfizer’s published threshold of efficacy.
So, almost certainly Pfizer should have announced the efficacy of its vaccine by, at latest, early on Monday November 2, using either the first (32 case) or, possibly, the second (62 case) interim analysis.
That evening, Pfizer officials informed a key F.D.A. official of its news, along with a short list of others. Biden’s team was alerted to the development that night. But in a sign of the suspicions that mark the Trump administration, the president’s top health officials did not learn of the news until the next morning, when it became public.
Mr. Trump’s anger about the timing has not abated. In a news conference on Friday, he suggested Pfizer and other drugmakers had taken revenge on him for pushing drug price controls.
“They were going to come out in October, but they decided to delay it because of what I’m doing,” he grumbled. “They waited and waited and waited.”
Obviously, Trump is right.
On the other hand, if Trump really were the authoritarian strongman his haters claim he is and his fanboys hope he is, he would have done something about this, such as, at minimum, dispatch his SEC to warn Pfizer that if they don’t disclose results according to their published protocol, they will be sued.
But that’s not who Trump is.
The NYT also implies that the similar Moderna vaccine, which was announced to be extremely efficacious on November 16, 13 days after the election, would also have been announced to be efficacious in late October or, at least, the day before the election if the Trump Administration weren’t so politically correct:
The call was tense, the message discouraging. Moncef Slaoui, the head of the Trump administration’s effort to quickly produce a vaccine for the coronavirus, was on the phone at 6 p.m. on Aug. 25 to tell the upstart biotech firm Moderna that it had to slow the final stage of testing its vaccine in humans.
Moderna’s chief executive, Stéphane Bancel, a French biochemical engineer, recognized the implication. In the race to quell the pandemic, he said, “every day mattered.” Now his company, which had yet to bring a single product to market, faced a delay of up to three weeks. Pfizer, the global pharmaceutical giant that was busy testing a similar vaccine candidate and promising initial results by October, would take the obvious lead.
“It was the hardest decision I made this year,” Mr. Bancel said.
Moderna’s problem seemed fitting for late summer 2020, when the United States was reeling from not just a pandemic but unrest over racial injustice. Dr. Slaoui informed Mr. Bancel that Moderna had not recruited enough minority candidates into its vaccine trials.
White people are more likely to volunteer for scientific research. I was in the consumer packaged goods marketing research business from 1982-2000. We put together a remarkably scientific panel of consumers by relying on white women’s urge to help out.
If it could not prove its vaccine worked well for Black and Hispanic Americans, who have been disproportionately affected by the pandemic, it would not make it over the finish line.
Nonwhites tend to have a perfectly rational but socially dysfunctional “What’s in it for me?” attitude toward scientific research.
Both companies ultimately completed the crucial stages of their human trials this month and reported spectacular initial results, vaccines that appear to be about 95 percent effective against a virus that has killed 1.3 million people, a quarter million of them in the United States.
“A delay of up to three weeks”: If Moderna had been allowed to announce three weeks earlier that would have been October 26. Two weeks earlier would have been the Monday before Election Tuesday.
Would it have made a difference if the stock market boomed the day before the election because of the great news about the vaccine?
If the news had flipped 1 out of 300 votes, that would have been enough to flip Georgia (0.24% Biden Advantage: 16 Electoral Votes), Arizona (0.31% : 11), and Wisconsin (0.63%: 10). That would have caused a 269-269 tie in the Electoral College. The addition of Washington D.C.’s 3 Electoral Votes due to the XXIII Amendment caused the Electoral College to have an even number of votes, which means you can have ties.
A tie in the Electoral College then goes to the House of Representatives, where each state in the new Congress votes as a delegation. Because the GOP is stronger in small states, Trump likely would have won in the House.
So, yeah, Trump probably would have been re-elected if he’d made Pfizer follow its published protocol or let Moderna carry out its clinical trial on the kind of people who want to volunteer for clinical trials.
But Trump failed at those tasks.