1924: Reading the Waves: Identifying Distinct Phenotypes of Multisystem Inflammatory Syndrome in Children During the 2020-2021 COVID-19 Pandemic

Thomas Renson¹, Nils Forkert¹, Kimberly Amador¹, Paivi Miettunen¹, Simon Parsons¹, Muhammed Dhalla¹, Nicole Johnson², Nadia Luca², Heinrike Schmeling¹, Rebeka Stevenson¹, Marinka Twilt¹, Lorraine Hamiwka¹ and Susanne Benseler¹, ¹Cumming School of Medicine, University of Calgary, Calgary, AB, Canada, ²University of Calgary, Calgary, AB, Canada

Background/Purpose: The COVID-19-associated multisystem inflammatory syndrome in children (MIS-C) is characterized by Kawasaki disease (KD)-like mucocutaneous features. The clinical course is often unpredictable. The goals of the current study were (a) to compare MIS-C phenotypes across the different waves of the COVID-19 pandemic, and (b) to identify variables associated with pediatric intensive care unit (PICU) admission.

Methods: Youth aged 0-18 years, fulfilling the World Health Organization MIS-C case definition, and admitted to the Alberta Children's Hospital during the first four waves of the COVID-19 pandemic (May '20-December '21) were included in this observational cohort study. Captured data included KD-like manifestations, clinical signs of shock/hypotension, peak C-reactive protein (CRP) and ferritin, platelet count nadir, peak NT-proBNP and troponin, liver enzyme abnormalities, sodium and albumin nadir, echocardiogram findings, PICU admission, and use of biologic agents. MIS-C features were compared across the different waves. Variables associated with PICU admission were established using univariate analyses and multivariate logistics regression models.

Results: Fifty-seven MIS-C patients (median age 6 years; 72% males) were included. Fifty-four per cent required PICU admission. All received immunoglobulins, 77% received corticosteroids, and 14% were treated with biologic agents. Patients presenting during the third (April-July '21; Alpha variant) or fourth wave (August-December '21; Delta variant) presented with higher peak ferritin (p< 0.001) and NT-proBNP (p=0.012) levels, and a higher prevalence of liver enzyme abnormalities (p=0.008), hypoalbuminemia (p=0.027), and thrombocytopenia (p=0.017). An unsupervised clustering model correctly classified 49.1% of the patients in the correct wave without having access to this information. In univariate analyses, PICU admission was associated with the presence of shock/hypotension (p=0.002), higher CRP (p=0.008), ferritin (p=0.002), and NT-proBNP levels (p< 0.001), lower albumin levels (p=0.014), and ventricular dysfunction on echocardiogram (p< 0.001). A multivariate logistic regression model combining peak NT-proBNP, troponin, and ferritin levels as covariates explained 70% (Nagelkerke R²) of the variance in PICU admission and correctly classified 91% of the cases. NT-proBNP was the sole significant contributor in this model (p=0.017).

Conclusion: MIS-C patients presenting in a later stage of the pandemic displayed a different phenotype including a predominance of features associated with macrophage activation syndrome, such as higher ferritin levels, liver enzyme abnormalities, and thrombocytopenia. This may reflect the impact of distinct SARS-CoV-2 variants. NT-proBNP emerged as the most crucial MIS-C feature associated with PICU admission, underscoring the value of monitoring.
Table 1: Upper panel: Significant differences in MIS-C features of patients presenting during the first phase of the COVID-19 pandemic (first and second wave combined) compared to those presenting during the second phase (third and fourth wave combined) (*one missing value). Lower panel: Significant differences regarding MIS-C features between patients admitted to the intensive care unit compared to those managed on the general ward.

Figure 1: Violin plots depicting differences in key laboratory MIS-C features between the four waves (panel A) and both phases (phase one is wave one and two combined, phase two is wave three and four combined; panel B) of the COVID-19 pandemic: peak ferritin values, platelet counts nadir, albumin nadir values, and peak NT-proBNP values.

Figure 2: Three-dimensional plot of the first three principal components (PC1, PC2, PC3) resulting from principal component analysis explaining about 30% of the variance in the data. A clustering of MIS-C patients from the first and second wave, and from the third and fourth wave can be observed.
Disclosures: T. Renson, None; N. Forkert, None; K. Amador, None; P. Miettunen, None; S. Parsons, None; M. Dhalla, None; N. Johnson, None; N. Luca, None; H. Schmeling, None; R. Stevenson, None; M. Twilt, None; L. Hamiwka, None; S. Benseler, None.