Kalle Saksela

Background: While COVID-19 lockdowns have been systematically associated with increased incidence rate of type 1 diabetes, less is known about the disease rates after lockdowns. We compared the incidence of type 1 diabetes after the COVID-19 pandemic-related closure of society (i) to the time when lockdown was implemented and to (ii) the time before the pandemic in Finland. Methods and findings: The incidence of type 1 diabetes after the lockdown (09/2021–12/2022) was compared to the incidence during the lockdown (03/2020–08/2021) and before the lockdown (three 18-month terms 2014–2019) in Finnish children under the age of 15 years. Previous exposure to SARS-CoV-2 was identified by combining information on COVID-19 vaccination with data on SARS-CoV-2 antibodies analyzed in blood samples. The temporal cohort included 3,467 pediatric type 1 diabetes cases, distributed across post-lockdown ( n = 586), lockdown ( n = 785), and pre-pandemic ( n = 2,096) periods. The incidence of type 1 diabetes after the lockdown was significantly lower than the incidence during the lockdown [incidence rate ratio (IRR = 0.85; 95% Cl [0.77,0.95]; p = 0.004)]. The rate of type 1 diabetes after the lockdown did not significantly differ from the time before lockdown (IRR = 1.00; 95% Cl [0.91,1.09]; p = 0.932). The frequency of children testing positive for infection-associated SARS-CoV-2 antibodies increased from 0.9% during the lockdown to more than 40% after the lockdown. Study limitations include a relatively short follow-up period after the lockdown and the lack of explicit modeling of the underlying long-term secular trend in type 1 diabetes incidence. Conclusions: The lockdown represents a peak in childhood type 1 diabetes incidence in Finland compared to both the pre-pandemic and post-lockdown periods. However, the reversion toward historical levels after the lockdown does not necessarily establish a return to the expected historical trajectory, nor can it fully distinguish a transient change from temporal compensation superimposed on the background secular trend. The contrasting patterns between type 1 diabetes incidence and SARS-CoV-2 antibody prevalence during and after the lockdown suggest an indirect rather than a direct role of the virus in triggering the observed changes in incidence.

Figure S8. In vitro and in vivo activity of low a7inity scFv-guided CAR (KD = 3910 nM).

Figure S13. SKOV3 ovarian tumor volume measurements of individual mice treated with T cells expressing constitutive or inducible Sb1205 anti-HER2 CAR (related to Figure 5B and C).

Figure S3. Antigen expression and HER2 knock-down in a panel of cancer cell lines and additional Jurkat T cell reporter activation (NFAT-GFP) assays in co-culture with 293T* (HER2 KD) cells transfected with tumor associated antigens (related to Figure 2C, 2D.

Figure S5. Cytotoxic activity of Sb CARs expressed in human primary CD8+ T cells (related to Figure 2E)

Figure S9. Tumor volume measurement of individual mice treated with T cells expressing Sb1205 anti-HER2 CAR (related to Figure 4C and E).

Nucleotide sequences of constructs used in the study

Figure S4. NFAT-EGFP reporter activation in Jurkat T cells exposed to phorbol myristateacetate (PMA) and ionomycin or co-cultured with target cells expressing the cognate tumor associated antigen.

Figure S12. Statistical analysis of fraction of CD62L+CD45RA+ in T cells engineered with constitutive ScFv and SbCAR and inducible SbCAR (related to Figure 5A).

Figure S14. HCC1569 breast tumor volume measurements of individual mice treated with T cells expressing constitutive or inducible Sb1205 anti-HER2 CAR (related to Figure 5B).