CDC's All-or-Nothing Approach to Teen COVID Vaccination Is All Wrong— The agency should revisit its latest guidance to maximize benefits and minimize risks
by Vinay Prasad, MD, MPH, Ramin Farzaneh-Far, MD, Wes Pegden, PhD, Venk Murthy, MD, PhD, Amy Beck, MD, MPH June 29, 2021
Last week, the CDC's Advisory Committee on Immunization Practices (ACIP) met to discuss the safety signal of myocarditis among young people who receive mRNA vaccination against COVID-19. This dialogue has been months in the making. Ultimately, the panel continued to endorse a two-dose mRNA strategy for all ages. We are concerned with this recommendation and offer five alternative considerations. But first, let's review how we got to this moment in order to make sense of vaccine-induced myocarditis.
A Recent History of Vaccine-Induced Myocarditis
The potential risk for vaccine-induced myocarditis was first raised on February 1 in the Jerusalem Post, which reported the hospitalization and intensive-care admission of a healthy 19-year old male 5 days after receiving his second dose of the Pfizer vaccine. This was followed by a nationwide report in the Times of Israel on April 23, later picked up by Reuters on April 25. These news reports suggested that Israel had seen elevated rates of this event after young men were vaccinated with the Pfizer vaccine, almost always after the second dose (56 out of 62 cases or 90%).
The European Medicines Agency announced an investigation on May 7, which was the same day several of us cautioned against FDA's use of emergency use authorization (EUA) to expedite the availability of COVID-19 vaccines to U.S. kids ages 12 to 15.
Although they were aware of this safety signal, the FDA issued the EUA on May 10 for Pfizer's mRNA vaccine in kids ages 12 to 15. Despite the fact that the vaccine was already widely in use in people ages 16 and above under the existing EUA, specific rates of myocarditis in the U.S. for 'near age' vaccine recipients (kids ages 16 to 18) were not made publicly available. In other words, no data on myocarditis events in kids close in age to the group receiving the new EUA (those ages 12 to 15) were leveraged in the process for granting this EUA. This is unfortunate, as these data would have had the greatest relevance and implications for the adjacent-age group.
Over the last 2 months, several news reports on clusters of cases of myocarditis after mRNA vaccination -- particularly in young men -- have been reported in the U.S. Revised estimates from Israel found the rate of myocarditis to be to one in 3,000 to one in 6,000 among males ages 16 to 24. On May 26, the Times of Israel reported that Israel's health ministry would consider just one dose in teens to balance getting most of the benefit of viral protection against mitigating much of the risk of myocarditis.
Israel now recommends vaccinating kids 12 to 15, but other nations have been more cautious. The U.K. advisers have decided not to support vaccination for kids under 18. The Germany standing vaccination commission advised that only children with pre-existing conditions receive the vaccine. The Netherlands (Dutch) health counsel advised only kids with pre-existing conditions or those living in a household with a family member who cannot be vaccinated receive the vaccine.
On May 22, the CDC announced they had received reports of myocarditis, and asked healthcare providers to file additional and missing reports into the Vaccine Adverse Event Reporting System (VAERS).
Last week, on June 23, ACIP met to discuss the findings. To date, CDC has documented myocarditis in at least 323 cases age 29 or under (of whom 96% were hospitalized), while 148 remain under review. The CDC acknowledged more cases in young people than older people, more cases in young males than young females, and higher incidence after dose two than dose one. The absolute risk of myocarditis after the second dose based on the number of CDC confirmed cases would be one in 15,000 to 20,000 for boys ages 12 to 24. There is a smaller but still excess risk in women age 24 and younger.
At its meeting, ACIP considered figures and data, which claimed to weigh the benefits versus harms of "dose two" of mRNA vaccines in this age group. However, in reality, the scenarios presented by the CDC compared the risks versus benefits to young people of no vaccination at all versus a scenario in which they received both shots.
The CDC did not consider the harms versus benefits of one versus two doses, but only the harms versus benefits of vaccination itself. But the CDC went beyond this. They also used base rates of infection from the past, rather than current rates of SARS-CoV-2 spread, which are substantially lower. They did not differentiate between healthy kids -- who are at risk of idiosyncratic events, such as myocarditis -- and kids with pre-existing medical conditions that place them at high risk of severe outcomes from COVID-19, including hospitalization.
This insistence on an all-or-nothing, one-size-fits-all binary approach -- treating healthy kids who have recovered from confirmed prior infection as equivalent to infection-naive kids with comorbidities -- is at the heart of the fallacy underpinning ACIP's decision.
While we acknowledge the CDC and ACIP had to act based on short-term studies and limited and variable data, vaccines must be used in a way that maximizes benefit and minimizes risk.
Ultimately, the CDC's recommendations came out so unequivocally in favor of vaccination that the following is true: If a 15-year-old recovers from COVID-19 and has high antibody levels, and this 15-year-old then receives one dose of mRNA vaccine causing hospitalization from myocarditis, the CDC would still contemplate proceeding with dose two once the "heart has recovered."
These events raise several points of concern:
VAERS is a suboptimal system. While the VAERS system was well-positioned to detect a rare and entirely unprecedented safety event (e.g., vaccine induced thrombocytopenia and thrombosis in the cerebral vessels), the system is suboptimal for elevations in naturally occurring health outcomes. Voluntary reporting requires a provider to make a mental link between vaccination and the outcome, and the mere fact that the CDC received more cases after coverage in the New York Times is evidence that VAERS failed to capture these events without prompting. This indicates cases may still be underreported: U.S. rates are likely a floor and not a ceiling. The meticulous tracking in Israel is likely closer to the real figure. Facing a factor-of-5 discrepancy between rates reported by Israel and the U.S., it is not prudent to simply assume that Israel is overcounting myocarditis, rather than the other way around.
If you change even one assumption, the CDC's model tips. Using the CDC's own framework of risk and benefit, key differences tip the calculus. First, the comparison doesn't have to be either two doses or no doses. We can also consider just a single dose. Dose two is associated with greater rates of myocarditis, and one dose of an mRNA vaccine has strong protection (over 90% for severe outcomes) -- even against novel variants such as Delta. If you do this, the calculus tips. Second, building on this model, if one assumes rates of myocarditis documented in Israel, accepting the hypothesis that VAERS underestimates risk, it gets even worse. One of us (Wes Pegden, PhD) re-ran the CDC's analysis accounting for this, which shows that second dose vaccination is unfavorable at young ages. Finally, the CDC's analysis uses SARS-CoV-2 rates from the past -- when fewer adults were vaccinated. Rates might rise in the fall, but that's unclear.
image
Figure by Wes Pegden, PhD
The CDC did not consider alternative strategies. The decision facing the CDC is not whether or not COVID-19 vaccination in children is generally a good idea. Most immediately, it is whether kids ages 12 to 15 should continue to receive second doses. A range of vaccination strategies are possible in children. Believing that COVID-19 vaccines can be valuable even for healthy children is different from thinking we cannot afford to proceed cautiously. Above all, it does not mean CDC should feel a need to stay the course with second doses whose marginal risks in teens appear likely to exceed their marginal benefits. Manufacturers could also reconsider the dose given to young people under 25 years old. Children's vaccination trials currently underway use lower doses than the adult studies; perhaps a lower or intermediate dose of vaccine could preserve most of the anti-COVID-19 benefit while avoiding the myocarditis risk. The CDC did not explore this option. And notably, dose optimization is an area of drug development for which there is a lot of room for improvement.
The CDC is not accounting for COVID-19 risk factors. Vaccination strategies for young people should be responsive to risk factors that place children at elevated risk of severe COVID-19 disease. While it is true that some cases of multisystem inflammatory syndrome in children (MIS-C) are idiosyncratic -- occur even in healthy kids -- the bulk of adolescent hospitalizations are among individuals with pre-existing risk factors. In contrast, the risk of myocarditis is entirely idiosyncratic and can strike anyone, including healthy adolescents at very low risk of severe disease. Vaccinating those at high COVID-19 risk, but not all young people, is a strategy that must be considered when balancing tradeoffs, as the harms versus benefits to healthy children are different than for kids with risk factors.
The CDC is not factoring in natural immunity. It is hard to believe that the risk benefit balance favors a 15-year-old young man who has recovered from COVID-19, and who has detectable antibodies, getting two doses of an mRNA vaccine. Such an individual is accepting a non-negligible risk of myocarditis, with limited upside in terms of decreased risk of severe infection. If the CDC recommends vaccination for these children, it is imperative they weigh benefits versus harms in precisely this population, which, to date, they have not presented.
Immediately after the ACIP meeting, various agencies and professional societies released a joint statement arguing that the benefit of vaccination far outweighs the risk in all age groups and demographics. Yet, our analysis suggests this is a premature conclusion. It relies on models that use outdated COVID-19 risk rates -- the on-the-ground rates in the moment are far lower, shifting the harm/benefit calculus to harm. It assumes two doses or none at all are the only options. It does not tailor recommendations by sex, natural immunity, or even comorbidities. We acknowledge there are individual and community level benefits to vaccination that extend beyond preventing hospitalizations and are an important part of the discussion. But these omissions from ACIP/CDC are problematic.
The stakes of this decision are no small thing. Even ACIP acknowledged there is a lot we still do not know about myocarditis after vaccination. There are additional cases being adjudicated, including serious ones, and there are no long-term follow-up studies yet to determine, for example, whether documented evidence of myocardial scar may portend an increased risk of arrhythmias. The ACIP and CDC discussion around vaccinating young teenagers, specifically boys, left out reasonably middle ground positions.
True vaccine proponents -- as all of us are -- understand the best thing we can do for vaccines is deploy them in a way that maximizes benefit and minimizes risk. This is crucial to protect health and also to ensure public confidence in the safety of vaccination. The current CDC guidance is not that. It needs to be revisited.
Vinay Prasad, MD, MPH, is an associate professor in the Department of Epidemiology & Biostatistics at the University of California San Francisco. Ramin Farzaneh-Far, MD, is a cardiologist and drug-developer based in Boston. Wes Pegden, PhD, is a mathematician at Carnegie Mellon University. Venk Murthy, MD, PhD, is a cardiologist and associate professor of Medicine at the University of Michigan. Amy Beck, MD, MPH, is a pediatrician and associate professor at the University of California San Francisco.