The data strongly suggest a relation of the risk for PIMS-TS to the prevailing VOC with the highest risk related to Alpha_B.1.1.7 and the lowest with Omicron_BA. Uniformity of the decrease across age groups does not suggest a major role for vaccination on the risk of PIMS-TS in infected children. SARS-CoV-2 vaccination does not appear to increase or decrease the risk of PIMS-TS. This is consistent with the analysis from the UK [15] and Denmark [16], which found a lower risk of PIMS-TS during the period with Delta and Omicron as dominant variants.
The challenge of such studies is the denominator because the true number of SARS-CoV-2 infections is unclear. In this analysis as well as in the UK study, the total number of reported rtPCR confirmed infections was used. This is sensible as far as the relative rate of PIMS-TS during Delta and Omicron predominance is the focus of the analysis. Interestingly the rate ratios in UK and Germany are almost identical.
The Danish study has some strengths compared to our data. While our study is strictly ecological, the Danish data allowed to disentangle vaccinated and unvaccinated cohorts but has some limitations concerning small numbers and correction for unidentified cases in the incidence data provided. The Danish study found virtually the same risk for PIMS-TS in the wild virus period, as shown previously [1]. Unfortunately, the numbers are too small to calculate rate ratios, although eyeballing suggests a similar reduction to this analysis and the UK study during the Delta and Omicron periods. A further strength of the Danish study is breaking down the data by age groups, but the number of cases is so small that conclusive analysis is not possible. Our age group-specific data, however, are based on substantial higher case numbers and do not suggest a higher reduction in vaccinated age cohorts.
Despite 2-dose-vaccination coverage reached > 60% in Germany in 12 to 17-year-old children by more than 60% by April 2022, there was no indication of a higher reduction of the PIMS-TS rate in this age group. Interestingly vaccination coverage of more than 50% were already attained by December 2021, well before the emergence of Omicron in Germany. Since vaccination reduces the absolute risk of infection, there could be an effect of the absolute risk of PIMS-TS, which we were not able to detect in our data. With respect to the low and rapidly waining vaccine effectiveness in relation to mild disease courses after infection with Omicron in children [17, 18], this effect, however, must be presumed to be low.
The strength of our analysis is the use of population-based data and the inclusion of more than 5 million confirmed infections with a uniform, widely available testing strategy during the observation period – with more than 30% of seropositive cases identified as of March 2021 [19].
In addition, these epidemiological findings might shed some light on the pathophysiology of PIMS-TS. Since the different variants of SARS-CoV-2 mainly differ in mutations within the viral S-protein, one might suggest that the immunological response to the S-protein is responsible for the hyperinflammatory state of PIMS-TS. This hypothesis is also supported by the fact that there are some observations of true PIMS-TS following vaccination (without infection) with the mRNA-vaccine coding for the S-protein.
Limitations
Our study has some limitations. First, even with a consistent mandatory routine RAT screening strategy for all children during the time period of our analysis, a certain degree of ascertainment bias for the infection rate cannot be completely ruled out. However, given the degree of the variant-effect on the risk of PIMS-TS, it is unlikely caused by variations in conducted rtPCR-tests.
Second, our analysis cannot provide the absolute risks for PIMS-TS for different SARS-CoV-2 variants, given that even with a uniform routine screening strategy, there will be a certain degree of underreporting in the numbers of the statutory notification systems. We can, therefore, only provide relative changes. These, however, were substantial with a reduction of about 70–90%. Previous estimates of the absolute risk of PIMS-TS in Germany [1] and the US [15] yielded an absolute risk of about 1:4,000 for the original Wuhan and the Alpha variant. Thus, the absolute risk for PIMS-TS related to infections with Omicron appears to be between 1:40,000 to 1:80,000.
Additionally, the PIMS-registry is based on voluntary notification. Although we strongly believe that the vast majority of hospitals which in fact care for patients with PIMS-TS are reporting their cases to our registry, we cannot rule out a certain amount of underreporting. As there is no second reporting system for PIMS-TS cases in Germany, we were not able to validate the completeness of the reports. Nevertheless, it can be assumed that even in case of incompleteness, it has remained the same over time, thereby not influencing the relative trend.