Case Series: Serum <scp>CXCL9</scp> and <scp>sTNF</scp> ‐ <scp>RII</scp> Levels as Useful Disease Activity Markers in Patients With <scp>COVID</scp> ‐19‐Associated Multisystem Inflammatory Syndrome in Children

Multisystem inflammatory syndrome in children (MIS-C) is a life-threatening condition that often occurs several weeks after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) exposure. MIS-C is characterized by systemic hyperinflammation with multiorgan failure. Therefore, appropriate immunosuppressive treatment is required to prevent the development of severe conditions in MIS-C [1]. The symptoms of MIS-C are similar to those of Kawasaki disease (KD) which is common in Japan, and differentiating MIS-C from KD can be challenging. Laboratory findings, including decreased platelets and increased C-reactive protein (CRP), D-dimer, troponin, ferritin, and IL-6 levels, have been reported to be diagnostic and disease severity markers. However, these markers are nonspecific and their association with MIS-C pathogenesis is indirect. In immunological analysis, dendritic cells and activated CD8+ T cells play a crucial role in patients with MIS-C [2]. Additionally, the expansion of T-cell receptor (TCR) Vβ21.3+CD4+ and CD8+ T lymphocytes have been observed in patients with MIS-C [3, 4]. MIS-C is also hyper-cytokine condition, and serum chemokine (C-X-C motif) ligand 9 (CXCL9) levels are elevated in patients with MIS-C5. In Japan, we previously reported the usefulness of serum cytokine analysis, including CXCL9, for differentiating between MIS-C and KD-mimicking diseases [5]. Another study demonstrated that serum levels of soluble tumor necrosis factor receptor II (sTNF-RII), which is one of the TNF-α receptors and correlates well with TNF-α production, are elevated in severe patients with severe KD compared to those with milder cases [6]. Furthermore, serum sTNF-RII levels have been indicated useful marker for disease activity in patients with KD7. Infliximab, a mouse–human IgG1 chimeric anti-TNF monoclonal antibody, has been demonstrated to efficacy in the treatment of KD patients [7]. In the context of MIS-C, the blockade of TNF-α has also been reported to be efficacious [8, 9], with TNF-α playing an important role in the development of a severe form of MIS-C. These findings suggest that CXCL9 and sTNF-RII may be potential disease activity markers. However, the clinical usefulness of these inflammatory proteins for assessing disease severity remains unclear. This report presents cases of MIS-C in Japan and evaluates the cytokine features associated with disease activity in these patients. MIS-C cases were diagnosed based on the criteria established by the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC). The disease severity was assessed using MIS-C severity score (MSS), defined as a 5-point ordinal scale ranging from 0 to 4, with higher scores indicating increased severity [10]. Serum levels of CXCL9 and sTNF-RII were measured using enzyme-linked immune sorbent assay (CXCL6, sTNF-RII, and IL-6; R&D Systems Inc., Minneapolis, MN, USA). Statistical analyses were performed using GraphPad Prism 9 software (GraphPad Software, San Diego, CA, USA). Correlations are expressed using Spearman's rank correlation coefficients. Statistical significance was set at p < 0.05. Three patients with MIS-C are described in this report. Patient characteristics, including initial laboratory findings and MSS, are presented in Table 1. Patient 1 was a 5-year-old Japanese boy who presented with fever, bilateral conjunctival hyperemia, mildly red lips, bilateral cervical lymphadenopathy, abdominal pain, frequent vomiting, diarrhea, and redness around the BCG scar in the upper left arm for 5 days. Abdominal ultrasound revealed mild splenomegaly and a minor amount of ascites. Echocardiography results were normal and exhaustive viral PCR and culture tests were negative. The patient had COVID-19 26 days prior to admission, and a SARS-Cov-2 antibody test was positive on the day of hospitalization. The patient's clinical symptoms rapidly improved after intravenous high-dose immunoglobulin therapy (IVIg) (2 g/kg) administration on Day 5. On the 14th day of follow-up, there were no symptom flare-ups, and blood tests including serum CRP, ferritin, and platelet counts were within normal limits (Figure S1A). Patient 2 was a 9-year-old boy who presented with fever, conjunctival hyperemia, abdominal pain, diarrhea, strawberry tongue, bilateral cervical lymphadenopathy, eyelid and lower extremity rashes, erythema, and indurated edema of the palms and soles. Peripheral cyanosis and coldness were noted, and the systolic blood pressure was 74 mmHg, requiring intravenous administration of extracellular fluids. Echocardiography revealed mild tricuspid regurgitation and trivial mitral regurgitation with an ejection fraction of 62%. The patient's medical history included autism spectrum disorder and COVID-19 infection 1 month prior to presentation. The patient presented with severe symptoms, including hypotension and hyponatremia, and was treated with IVIg (2 g/kg) and a high dose of prednisolone (PSL) (2 mg/kg/day) in the intensive care unit on Day 7. The patient's symptoms gradually improved, and inflammatory markers such as CRP and ferritin also showed improvement after treatment. Echocardiographic findings were unremarkable, and the patient was discharged on Day 14 (Figure S1B). Patient 3 was an otherwise healthy 7-year-old girl who presented with fever, conjunctival hyperemia, and swelling on the left side of her neck. Echocardiography results were normal. IVIg (2 g/kg) and oral aspirin were initiated on Day 6. The patient's symptoms gradually improved from Day 10 onward (Figure S1C). The patient tested positive for SARS-Cov-2 antibody. We sequentially measured serum cytokine and chemokine levels in these patients of varying disease severity, using blood samples collected before and after treatment (Table 1, Figure S1). The timing of blood sampling differed among the three cases as it was determined based on the treating physician's clinical judgment; however, in all cases, blood samples had been collected repeatedly from each patient both before and after treatment. The peak serum concentrations of cytokines and chemokines in the respective cases were as follows: In Patient 1, the serum CXCL9 level was 7829 pg/mL, serum sTNF-RII level was 16 055 pg/mL, and serum IL-6 level was 60 pg/mL on Day 6. In Patient 2, serum CXCL9 level was 30 155 pg/mL, serum sTNF-RII level was 39 999 pg/mL, and serum IL-6 level was 421 pg/mL on Day 7. For Patient 3, the serum CXCL9 level was 9257 pg/mL, serum sTNF-RII level was 28 201 pg/mL, and serum IL-6 level was 102 pg/mL on Day 8. Serum CXCL9 and sTNF-RII levels in Patient 2, who had the most severe symptoms, were higher than those in the other patients. In addition, we analyzed cytokine and chemokine kinetics in these patients. In all three cases, serum levels of CXCL9 and sTNF-RII were elevated in the active phase, when symptoms were most severe, and decreased, respectively, in the inactive phase, when symptoms had completely resolved after anti-inflammatory treatment (Figure 1A,B). Next, we assessed the correlation between cytokine levels and laboratory data concerning disease activity in patients with MIS-C. Serum CXCL9 levels were significantly correlated with serum ferritin (p = 0.0029, ρ = 0.7972), aspartate aminotransferase (AST) (p = 0.0016, ρ = 0.7989), lactate dehydrogenase (LDH) (p = 0.0016, ρ = 0.7989), and platelet counts (p = 0.0029, ρ = −0.7088) in patients with MIS-C (Figure 1C–F). Moreover, serum sTNF-RII levels were significantly correlated with serum ferritin (p = 0.0101, ρ = 0.6978), AST (p = 0.0006, ρ = 0.8172), LDH (p = 0.0006, ρ = 0.8172), and platelet counts (p = 0.0142, ρ = −0.5907) in patients with MIS-C (Figure 1G–J). Patients with MIS-C often develop severe conditions, including myocardial involvement and shock. Therefore, an accurate diagnosis and immunosuppressive treatment are required during the early stages. Serum biomarkers, including CXCL9, have been previously reported to be useful for diagnosing MIS-C [5, 11]. A previous report showed that serum CXCL9 and sTNF-RII levels in MIS-C patients are significantly higher than those in KD patients [5]. In this report, we demonstrate that serum CXCL9 and sTNF-RII levels are elevated in patients with MIS-C in the active phase and are higher in the severe case, indicating that these could represent novel disease activity markers. CXCL9 is an inflammatory protein that is mainly induced by IFN-γ. This chemokine has been reported to be elevated in macrophage activation syndrome (MAS), a secondary hemophagocytic lymphohistiocytosis (HLH) complicating systemic juvenile idiopathic arthritis (s-JIA), and rheumatic diseases [12]. MAS is also one of the severe cytokine release syndromes requiring adequate immunosuppressive treatment. Previous reports have indicated that serum CXCL9 levels are associated with disease severity in patients with s-JIA complicating MAS [13, 14]. Anti-IFN-γ monoclonal antibodies have been reported to be clinically effective in patients with MAS [15]. Therefore, IFN-γ plays a pivotal role in MAS, and the amount of IFN-γ production may be associated with a more severe condition in MAS. In this study, we showed that serum CXCL9 levels increased during the active phase and decreased after immunosuppressive treatment. Furthermore, Patient 2, with a severe condition, has much higher serum CXCL9 levels than the others. Another report indicated that inflammatory protein signatures in the blood are similar to MIS-C, MAS, and thrombotic microangiopathy, and INF-γ responses are dysregulated in patients with MIS-C [16]. Therefore, CXCL9 may be a potential marker of disease severity in MIS-C patients. However, our findings cannot be generalized because of the limited sample size. Further large-scale analyses are required to elucidate the pathophysiology of MIS-C. sTNF-RII is one of the soluble forms of TNF-α receptor, and these levels are known to be correlated with TNF-α overproduction. In previous reports, plasma TNF-α levels in MIS-C are elevated compared to those of febrile patients [11]. Another report demonstrated that sTNF-RII levels in MIS-C are significantly higher than those in KD cases; however, the sensitivity and specificity for sTNF-RII are not high according to receiver operating characteristic analysis [5]. In this study, serum sTNF-RII levels were also elevated in MIS-C patients during the active phase and were higher in Patient 2 with severe MIS-C than in those with other conditions. Some cases of MIS-C have been reported to require treatment with anti-TNF blocker [8, 9]. In the cases reported here, additional anti-inflammatory treatment such as anti-TNF blocker was not needed, but high levels of sTNF-RII correlated with TNF-α production could represent disease severity marker in patients with MIS-C. In conclusion, serum CXCL9 and sTNF-RII levels in patients with severe MIS-C are extremely high during the active phase. These inflammatory proteins may serve as potential markers of disease activity in patients with MIS-C. S.H. and M.M. contributed to the design of the study, collected data, and drafted the manuscript. T.T., Y.I., and S.M. contributed to the acquisition of data. S.K., A.S., and S.M. contributed to analyzing serum cytokine and chemokine. Y.N. supervised and revised the final version of the manuscript. All authors agreed to the final version of the manuscript. This work was supported by grants from Medical and Biological Laboratories Co. Ltd., Japan. Consent for publication of this report was obtained from the parents of the patients. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. Figure S1. Clinical course in (A) patient 1, (B) patient 2, and (C) patient 3 with MIS-C. IVIG: intravenous injection of immunoglobulin, ASA: acetylsalicylic acid, PSL: prednisolone, CXCL9: Chemokine (C-X-C motif) ligand 9, sTNF-RII: soluble tumor necrosis factor receptor II. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.