In mid-April, vaccine scientists and regulators were alerted to six serious adverse events after administration of the Johnson & Johnson COVID-19 vaccine. These events occurred in women ages 18 to 48 who had received the vaccine under emergency use authorization (EUA), and were similar to reports from Europe regarding the other adenoviral-based vaccine made by AstraZeneca.
The syndrome is best described as vaccine-induced thrombocytopenia and thrombosis (VITT). These events are devastating. They involve cerebral venous thrombosis (CVT) or thrombosis of the mesenteric vessels, and are thought to occur in the setting of runaway platelet activation due to auto-antibody formation -- a pathophysiology most similar to heparin-induced thrombocytopenia and thrombosis. VITT can result in death, but also, serious and irreversible neurological impairment in an otherwise young healthy individual.
After the FDA and CDC were alerted to these six cases, the agencies called for a pause or moratorium to be placed on use of the J&J shot pending further review. A few days later, a panel of experts extended the pause by 1 week, during which time at least nine additional cases of the syndrome were identified. After a hearing to discuss the risk and benefit, the Advisory Committee of Immunology Practices (ACIP) voted to resume vaccination with J&J without any restrictions by age or gender, though they recommended placing a warning.
These events highlight several interlocked lessons that warrant discussion. Some lessons are counterintuitive, but necessary to consider. In my view, the only acceptable way to think about vaccines is as a scientist and not ideologue. Unfortunately, scientists are in short supply.
Responding to the Pause
The moment six cases were publicly reported and the pause was instituted, a large number of experts expressed anger on social media. They claimed the denominator was the seven million doses of the vaccine that had been administered, and this event rate -- less than one in a million -- was a trivial safety signal. Some were critical of the pause itself -- which I will discuss in the next section -- while others worked to immediately downplay concerns of the risk.
A popular series of memes compared the risk of CVT and VITT to the risk of thrombosis from other events. One compared VITT to the risk of clot with oral contraception and thrombosis after COVID-19. These went viral (no pun intended).
Unfortunately, these comparisons perpetuated at least five errors.
First, one in a million was almost surely the wrong number. The moment a novel safety signal is identified, preliminary estimates of frequency are utterly unreliable. Reporting of vaccine adverse events is still largely voluntary, and there are important gaps. There are likely more unreported cases, which will raise the numerator, and the denominator should not include all vaccinated individuals. The correct denominator is the fraction of vaccinated individuals in the demographic group experiencing the severe event -- in this case, women ages 18 to 48. I tweeted that I would not be surprised if the true incidence jumped one order of magnitude when we had more facts -- a prediction that has since been vindicated.
Second, comparing the risk of CVT in the setting of VITT to a garden variety venous thromboembolism is misleading. A deep vein thrombosis of the leg is not comparable to one in the cerebral vein in the setting of runaway platelet activation. I have long wondered how much anatomy should be taught in medical school, but I can now confidently say we should definitely clarify the difference between veins that drain the brain and those that drain the leg.
Third, comparing the risk of CVT and VITT after vaccination to the risk of clot after COVID-19 is inappropriate. Getting COVID-19 and getting vaccinated are different. A vaccine's risk cannot be changed, while the risk of COVID-19 can be altered or modified by behavior. In fact, the bulk of this past year has been making behavior changes to modify SARS-CoV-2 risk.
Fourth, some numbers are simply wrong. A 16.5% risk of clot among someone with COVID-19 is an inflated figure that, as far as I can tell, comes from meta-analyses of ICU patients in the first wave. These analyses likely suffer from selection, surveillance, and ascertainment bias, among other problems.
Fifth, comparing the risk of CVT and VITT to oral contraception was objectionable to several women I know. Oral contraception is taken by women to take control of their lives. SARS-CoV-2 is best avoided, but if a 20-year-old woman becomes infected, there is more than a 99% chance that she will recover and her life will be the same. Having a child at the age of 20 ensures a pretty certain chance that life is never the same. Moreover, information about the risk of clot is available before women start taking the pill and they can balance this against their other desires. With CVT and VITT, women did not know the precise risk, still don't know exactly, and these folks were essentially arguing against providing women that information.
Why the Immediate Reaction to Downplay?
The first crux of the public communication failure is to ask why, when we hear of a novel adverse event, is the reaction of so many experts to downplay or trivialize the risk? Why construct minimizing memes when you have not even gathered all the relevant facts?
The answer to this question warrants reflection, but I will offer a hypothesis. In 2021, there is clearly a small, but vocal minority of individuals opposed to nearly all vaccinations. They invent preposterous reasons to justify their opposition, and no evidence can persuade them of their folly. I share the view that these individuals are a threat to public health, but I am not sure if hatred, anger, and demonization or rather education, dialog, and empathy is the best way to deal with them. Yet, the fact is these individuals exist.
In response, there is a group of individuals on the other extreme. To them, either one must embrace all vaccines for all indications for all ages, or one can be lumped with the other extreme. They favor universal child vaccination of SARS-CoV-2 via an EUA, even before they have the data for that claim. They were quick to embrace vaccination for pregnant woman prior to appropriate trials establishing safety. Suppressing critical thinking to extol vaccines is also wrong and may backfire, but I believe this explains why it occurs. It is, to some degree, a counter-movement against the anti-vaxxers, which can go too far.
A Good Call to Pause?
All of this discussion is not a referendum on the pause, which must be considered on its own merits. No matter how one feels about the prior issues, whether the FDA and CDC should have issued a total pause for J&J vaccine or a selective pause (just in women under 50 or 60) or no pause at all is a complex empirical question. It cannot be settled by rhetoric. It requires careful studies. Pauses have complex downstream effects. Yes, they may poison vaccine acceptance. On the other hand, inaction, while the tally of women with CVT and VITT rises, is also a dangerous game. A few anecdotes about a young 30-year-old who herniates her brain is also not ideal for vaccine acceptance.
As a scientist and policy expert, I withhold final judgement about the pause until better data maps the question, but if you forced me to make a call, I would have just paused J&J for women less than 65. That judgment is based on my second order principle of policy: when you cannot think through all the intangibles, do what makes sense.
Should J&J Be Given? If So, for Whom?
On April 23rd, the ACIP met again, and at least nine more cases had been identified. The denominator was also clarified to be the approximately two million women in the demographic group. Thus, as I predicted, the event rate became one order of magnitude more common. It is now approximately one in 200,000 overall, though in the subgroup of women ages 18 to 39 it is closer to one in 100,000. This too is best thought of as the floor rate, as more data may change it slightly.
This risk of CVT and VITT must be weighed against the benefit of vaccination. That is a tricky proposition. Technically, the risk of getting a J&J shot now must be weighed against the risk of getting an mRNA shot later (Pfizer and Moderna), and the fact that the full course requires a second dose. The mRNA shots have been given to many more Americans and are entirely free of this event. How much later does a woman have to wait? At the time of the ACIP's decision and today the wait may be miniscule in most places, but perhaps a couple weeks in others.
Waiting carries an increased risk of SARS-CoV-2 in that time, but that risk is modifiable -- based on what the person does. Moreover, that risk is proportional to population rates of viral spread, which are plummeting in the U.S. in most (but not all) states, as the rate of adults receiving at least one dose crosses the halfway mark.
In the unfortunate event that a person gets SARS-CoV-2 in these interim weeks, the risk of a bad outcome is related to her physical health and age. That risk is lower in May 2021 than it was in May of 2020. This is in contrast with CVT and VITT, which appears to strike even the youngest in this group (18- to 40-year-olds).
Although models were presented at the ACIP meeting, those models did not account for what would happen if more mRNA vaccines were made available -- a devastating shortcoming. I personally crunched the numbers several times, trying to weigh the risk of VITT in this age group against the risk of COVID-19 death during a waiting period, using varying guesses for local SARS-CoV-2 rates and varying wait times, and I wasn't sure which is a better choice: get J&J now or mRNA (possibly) later.
If you do not know that J&J is surely preferable to waiting for an mRNA vaccine, placing J&J back on the market is the wrong call. You have to be confident, and overwhelmingly so, that the net benefits outweigh the harms. Moreover, once you release J&J on the market you cannot predict which 18- to 50-year-old women will get the shot. Are these women at average risk of getting SARS-CoV-2 or are they at reduced risk? These are additional and important complexities.
What would I have done? I would have let J&J return to the U.S. market, but not for women less than 65 years old. These folks should be encouraged to seek the mRNA shot, which is far safer for them. This is similar to Canadian guidelines, which say mRNA is preferred. I would have felt differently if the U.S. was where India is today, but the reality is we are not facing the same situation. We have alternative choices, and we are passing 50% vaccinated with at least one dose. Combine that with pre-existing immunity, and the situation is as good as one can hope for right now. It is not worth taking a J&J gamble. Yet, ACIP voted to release the J&J vaccine to all people over 18.
Opposing Viewpoints
Some object to my position and ask what message it would send to the world if the U.S. were to curtail use of J&J. That message might hamper efforts in countries where adenoviral vector vaccines are desperately needed. I am sympathetic to this view; however, it must be dismissed entirely. U.S. drug and vaccine regulation will only focus on the best options for people in the U.S. We cannot make decisions considering the globe. Just as a doctor must advocate for her patient; our vaccine regulators can only advocate for our inhabitants.
Another objection is that J&J is necessary for rural places. The mRNA vaccines have to be kept cold -- really cold, and this does not work for vulnerable populations and rural settings. I am also sympathetic to this argument, but the burden is on those who think the vaccine is beneficial to prove that is true beyond a doubt. I need to be convinced that giving these populations J&J is more beneficial to them than improving mRNA transit and outreach to get them an mRNA shot.
Another objection is autonomy. All that matters is quantifying risks, and then the choice should be up to the person getting the vaccine. This is simply not true. The fact that we have a regulatory system, and the EUA, shows that society gives people choices, but does not permit unfettered choice. When it comes to vaccines, the role of regulation is to ensure people have access to safe and effective vaccines. That balance depends on many factors including the extent of the problem at hand. In this case, the standard for vaccines is permitting choices that are certain to be of benefit -- a "close call" is not acceptable.
Where Have All the Scientists Gone?
A small faction of people vigorously opposed to all vaccination have done damage in several ways. First, they have discouraged individuals from getting necessary childhood immunization. Second, they have cast aspersions around the mRNA vaccines, which offer amazing efficacy. Third, they have distorted scientific dialog around vaccines, and we struggle to approach them impartially. Now, as a reaction, many confuse cheerleading with science.
Tweeting silly memes that trivialize the risk of CVT in the setting of VITT is not science. Claiming a risk is one in a million with incomplete and preliminary data is not science. Instead, science means being able to say that mRNA vaccines are terrific; their benefit to Americans in massive. The J&J vaccine also has an important role, but that role is uncertain in women under 65, and for that subgroup the EUA may still be rescinded. A true scientist navigates these troubled waters and does not take reflexive extremes. Sadly, there are few scientists left.
Vinay Prasad, MD, MPH, is a hematologist-oncologist and associate professor of medicine at the University of California San Francisco, and author of Malignant: How Bad Policy and Bad Evidence Harm People With Cancer.