Where’s the data on pediatric coronavirus vaccines?
Something seems to have gone awry in the trials. But it's not what anti-vaxxers hope.
Can you imagine the pressure that scientists and regulators responsible for the fate of coronavirus vaccines for children under 12 years old in the United States must feel right now? It’s almost unimaginable. And yet, halfway through September, there are no data available to the public. Is this a mistake or some dereliction of duty?
No. Given the urgency of the situation, we cannot rationally conclude that these committed professionals are dragging their feet unnecessarily, while lives hang in the balance; to think otherwise would be to fundamentally misunderstand the shared goal of every single person involved with the project of delivering a safe and effective vaccine to our youngest.
A far more straightforward explanation exists.
Either something has gone wrong in the clinical trials assessing the safety and efficacy of coronavirus vaccines in children under 12 or, to the untrained eye the data make it appear as though something has gone wrong.
I’m convinced it is the latter. If Pfizer, Moderna, or federal regulators were to release data that even hinted at casting pediatric coronavirus vaccines in a negative light—even if that that perception were ultimately incorrect—the fallout would be a disaster. Nobody would bother to read the explanations. They would just remember one damaging headline.
Through this lens, the delay in the public release of actionable information makes perfect sense. No other explanation I am aware of connects the dots.
Either something has gone wrong in the clinical trials assessing the safety and efficacy of coronavirus vaccines in children under 12 or, to the untrained eye the data make it appear as though something has gone wrong.
•••
What’s happening behind the scenes? I believe the main reason data on coronavirus vaccines in children under 12 remain under embargo is that we became victims of our own success. For the first three months that the trials testing coronavirus vaccines in children under 12 were live (starting in late March and April), there simply were not enough coronavirus infections for the vaccines’ benefits to be detectable.* The early months of these trials precisely coincided with the lowest rates of coronavirus recorded at any time in the US since the pandemic began—a fact that was likely in no small part due to the success of the adult vaccination campaign here. By June, daily case counts had dropped 95% from the January peak.
This means that only the occasional harms of the vaccine were being detected early in the trial. Imagine if the authorities reported that this summer? My belief is that rather than risk the consequences of having to explain all of this to a risk-averse public, it was determined that the trials needed to be expanded so as to pick up more infections in the placebo group. That way, when the vaccinated and placebo groups were compared, the benefits of the vaccine would actually be apparent.
This is almost what was reported in the media, but not quite. In July, we were told that the American Academy of Pediatrics (AAP) and the US Food and Drug Administration (FDA) asked the vaccine companies to expand their trials to include more children in order to “detect rare side effects.” That just does not make any sense.
At the time, Moderna had planned to enroll 6,750 children and Pfizer had planned to enroll 4,500 children into their respective studies. Centers for Disease Control and Prevention data from teens this spring suggested that around 1 in 16,000 children had developed myocarditis (albeit almost all mild bouts) in the 21 days following the 2nd dose one of the mRNA vaccines. Since then, Moderna increased its trial to 13,275 children. That’s not enough to detect any statistically meaningful harms, unless the rate of myocarditis among children under 12 is far, far worse than that in teens. That seems unlikely (though the rates may indeed be notably higher) because if that were the case, the investigators would have been obligated to halt the trial. In fact, even if the companies had quadrupled the number of kids recruited into the trials (some of whom would go on to receive placebo), only a few cases of myocarditis in each trial would be expected (again, unless the myocarditis rates were off-the-charts higher than in teens). But even if that were so, expanding the trials still wouldn’t accomplish the stated goal of detecting more rare adverse events unless those expansions were much greater.
What’s seems far more plausible to me is that word got out: There had been cases of myocarditis reported early in the trials among just the few thousand children under 12 who had been vaccinated. The rates might appear to be even higher than anticipated if, as I suspect, the trials have quietly been tracking not just clinical myocarditis, but blood tests that detect small amounts of subclinical damage to the heart—some of which may actually be meaningless. If my conjectures here are correct, and that information was known to the AAP and the FDA, the request that the companies expand the trials was not an effort to detect more cases of myocarditis. In fact, it was not a demand for improved safety monitoring at all. It was a lifeline in response to a distress call. It was a generous way of “asking” the companies to wait for more cases of Covid-19 to occur in the placebo group. That way, the benefits of the vaccine would be able to stand in judgment against the detected harms. If the tea leaves are as they seem, the results will please those eager to have their younger children protected, and will be disappointing to cynical anti-vaxxers looking for any win.
Prior to the trial expansions in July, however, it seems likely that only harms had been detected among vaccine recipients—and, yes, possibly at higher rates than seen in teens; The rates of myocarditis seem to go up, the younger we look. The trials were enlarged in July, right as the Delta variant began wreaking havoc. This development, unfortunate though it was, provided trial investigators with an influx of valuable information. Suddenly, with coronavirus finding its way into children more quickly than ever, the trials were now poised to show the protection that the vaccines were designed to impart.
•••
Is all of this fair? Yes and no. The Pfizer trial was designed to assess Covid-19 clinical outcomes if at least 22 cases had occurred in the trial. But with 9 different dosing strategies, and likely a small number of Covid-19 cases, the benefit of the vaccine had likely been diluted by a combination of its original design and low case counts before Delta emerged. By trying to do the right thing (i.e. testing a number of different doses to find the safest and most effective strategy), Pfizer had backed itself into a corner. So the AAP’s and FDA’s requests to expand the trials were likely welcome by the companies. You didn’t hear Pfizer complain about being slowed down by this request, did you? Me neither. That’s because they know what they’re doing.
Suddenly, with coronavirus finding its way into children more quickly than at any time during the pandemic, the trials were now poised to show the protection that the vaccines were designed to impart.
What if what happened this spring and early summer happens sustainably in the future—that at some point we have very low rates of infection? If that happens, we indeed will have put a small number of children at risk of myocarditis without much of an upside. However, with Delta hospitalizing more and more children, and the always-looming threat of other emerging variants, ending the pandemic by taking children off the table as unprotected targets makes sense. The data that will be released in the coming days or weeks will likely show just that. But we won’t know for certain until the numbers are released.
Meanwhile the FDA and the CDC has specifically instructed pediatricians not to vaccinate children under 12, a practice which many considered an option after Pfizer’s vaccine earned full FDA approval this summer. (Full FDA approval generally means that off-label prescribing at the discretion of a physician is permitted). There’s likely a reasonable explanation for the almost threatening language that has come out of the CDC on this topic. One is that, as mentioned, the Pfizer trial has been studying up to 9 different doses. If relatively high doses are found to cause harms that are not ultimately accompanied by markedly increased protection over that offered by lower doses, avoidable harms will have been caused by renegades guessing on what doses to inject to into children.
•••
This summer, my colleagues and I wrote in the New York Times that the best way to assess the risks and benefits of pediatric coronavirus vaccines was to war-game out what would happen if every child eventually contracted Covid-19 or if every child instead were to be vaccinated against it. That exercise demonstrated that even if scary Covid-19 outcomes were extremely unusual among kids and even if we worried that the vaccines caused far more important adverse events than they actually seem to, the scales still tipped strongly in favor of vaccinating. The “bad luck” of initiating the trials for children under 12 this spring meant that a very different scenario was momentarily playing out—all vaccine, no disease. That could happen again. But Covid-19 could also rage back at any time. We can't gamble with our kids. As this summer has shown, Covid-19 is like a casino. Unless you stack the deck, the house eventually wins.
Would you vaccinate your child without trial data? Do you trust the process that has brought us safe vaccines so far? Leave a comment below.
Want more Inside Medicine? Get it in your inbox first thing and stay ahead! Click here to subscribe.
•••
*These studies were primarily designed to assess two things: First, whether the vaccines stimulate enough of an antibody response that scientists can infer (based on data from older people) that vaccinated children generate an adequate immune response; Second, to assess how frequently adverse events occur, and how serious they are. The number of children who develop symptomatic or asymptomatic coronavirus infections are indeed being tracked, but as a “secondary outcome measure,” meaning that the studies have not been statistically designed to assess this. However, everyone will notice the results of these secondary outcomes because, to parents, that’s the kind of information that really matters.