As a Government-commissioned interdisciplinary working group consisting of healthcare professionals and data scientists, we have been closely monitoring the safety of covid-19 vaccines in Hong Kong using a territory-wide electronic public healthcare database. This blog post tells the story of our discovery of the risk of myocarditis and the downstream impact on public policy and importantly, how this series of studies has prompted a previously unaddressed research question that we addressed in our new paper in Nature Communications.

Use of big data to identify myocarditis risk associated with BNT162b2

Shortly after the extension of the licensure of BNT162b2 to adolescents in April 2021, our colleagues from the Department of Paediatrics and Adolescent Medicine first informed us of an unusually increased frequency of myocarditis cases admitted to their wards following BNT162b2 vaccination, especially in young and male patients after the 2nd dose. This observation was consistent with case reports from other countries that used the vaccine, although at the time, the evidence was largely descriptive. Our team considered an urgent need to conduct both a retrospective cohort study (to compare the incidence rate with historical background rates) and a territory-wide case-control study to provide the first real-world analytic evidence on the association. The finding showed an evidently elevated risk of myocarditis after vaccination with BNT162b2, which was about threefold overall, but with a very rare absolute risk nonetheless. We informed the Hong Kong Government of the results and published them as research articles in Clinical Infectious Diseases and the Annals of Internal Medicine. In a prompt response to such risks, the Hong Kong Government suspended the 2^nd dose for adolescents in September 2021 and later recommended an 8-week interval between the two priming doses. An immediate decline of myocarditis cases was recorded and our team reported this observation in a research letter in JAMA Pediatrics. We then compared the 180-day prognosis of post-vaccination myocarditis cases with historical cases and the findings showed a reassuring milder prognosis for the former. Results were published in the Journal of the American College of Cardiology. 

How about vaccine effectiveness against Covid-19? Altered or not? 

It was apparent that there was a lower risk of myocarditis when the dosing interval was extended from 3 to 8 weeks, with accruing international epidemiologic evidence suggesting the same finding. It was hence, sensible to delay the 2nd dose to 8 weeks after the 1st dose. However, our team began receiving queries from clinicians and regulators about vaccine effectiveness resulting from this extension. It is a legitimate concern. Not only was there a wider time interval in which adolescents and children were only one-dose vaccinated, but the stimulation of the immune system by the vaccine might be affected with the change in timing of the 2nd dose. This question has indeed been posed before, although it was motivated by an entirely different rationale. In order to vaccinate the greatest number of people with at least one dose in the face of rapid surges in Covid-19 cases earlier during the pandemic, the United Kingdom Government took the approach of a delayed second dose in the general population, and there has since been an abundance of scientific evidence supporting this policy. Particularly, immunogenicity studies have been conducted to investigate the humoral and cellular responses to vaccination with varying dosing intervals and mostly found an enhanced response associated with dosing interval extension. A large Canadian real-world study also provided evidence of significantly better effectiveness. However, since the question was not motivated by reducing the risk of myocarditis in adolescents and children, no real-world studies had been conducted to investigate the potentially variable effectiveness in this vulnerable population.

First real-world study in the adolescent and children population

The importance of representation of adolescents and children in studies investigating vaccine effectiveness after dosing interval extension is profound, because we believe that scientific evidence is warranted to support any public policies or clinical recommendations to have population-wide and global impact. Our team was determined to provide this evidence. We started looking into this association in late 2022 and decided that a territory-wide nested case-control design would be the most suitable for the research question. Because of the retrospective perspective in ascertaining the exposure, i.e., extended dosing interval versus regular intervals, there would be no concern of immortal time bias that is typically entailed in a cohort study with similar exposures of interest. As we proceeded to analyze the data and cross-checked our results, we were excited to identify clear evidence that there was a lower risk of SARS-CoV-2 infection associated with extended dosing intervals, because this suggests that not only was the extension of the dosing interval associated with decreased risk of myocarditis, but the extended dosing interval was also associated with a significantly stronger vaccine effectiveness.

Prospect of the mRNA vaccine 

Covid-19 continues to exist, albeit far less deadly than previous stages of the pandemic because of better-informed preventive measures and high vaccination coverage. There continues to be children eligible for mRNA vaccines. We anticipate that the use of the vaccine will not be halted any time soon in absence of a safer or even more effective alternative. It is, therefore, important to use sound judgment in our decision to vaccinate children when considering the risks (including myocarditis) versus benefit. We believe that scientific evidence has shown that extending the dosing intervals of priming doses from 3 to 8 weeks was one of the essential steps to be taken and a change in the product insert to reflect this recommendation should be advocated.