Multisystem Inflammatory Syndrome Related to COVID-19 in Previously Healthy Children and Adolescents in New York City

Abstract

This case series describes clinical characteristics, treatment, and outcomes of 17 previously healthy SARS-CoV-2–infected children and adolescents with an inflammatory phenotype overlapping with but distinct from Kawasaki disease and toxic shock syndrome admitted to a New York City hospital in late April and early May 2020.

Severe coronavirus disease 2019 (COVID-19) has been reported rarely in children.^1,2 International data suggest the development of a proinflammatory syndrome with features of Kawasaki disease (KD) or toxic shock syndrome (TSS) in children, possibly related to COVID-19.³

Methods

Patients were included if they (1) were 21 years or younger; (2) were hospitalized at Columbia University Irving Medical Center/New York-Presbyterian Morgan Stanley Children’s Hospital in New York City between April 18 and May 5, 2020; (3) presented with a clinical syndrome characterized by prolonged fever, systemic inflammation, shock, end-organ dysfunction, or symptoms reminiscent of KD or TSS; and (4) had evidence of recent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. SARS-CoV-2 infection was determined by reverse transcriptase–polymerase chain reaction (RT-PCR) of nasopharyngeal swabs or positive serology. Serology testing was done using a New York State Department of Health–approved combined assay for IgM and IgG antibodies against SARS-CoV-2 spike trimer or nucleocapsid protein (96% specificity, 93% sensitivity). Admission testing included hematologic parameters, chemistries, co-infections, and inflammatory markers along with assessments of cardiac function (electrocardiograms, transthoracic echocardiograms). The Columbia University ethics committee approved the study with a waiver of informed consent.

Results

Among 17 patients (8 male; median age, 8 years [range, 1.8-16 years]) (Table 1), most were white (n = 12) and previously healthy (mild asthma in 3). All patients had fever (median duration, 5 days). Fourteen had gastrointestinal symptoms, with 1 showing acute ileocolitis on imaging. Mucocutaneous findings were common (rash [n = 12], conjunctivitis [n = 11], and lip redness/swelling [n = 9]). Three patients were hypoxic at presentation, and 13 had shock. Fourteen had abnormal chest radiograph findings, most commonly bilateral, interstitial opacities. Eight met criteria for KD and 5 for incomplete KD.⁴

Table 1. Demographics, Clinical Characteristics, Treatment, and Outcomes of Patients With COVID-19–Related Multisystem Inflammatory Syndrome in Children (N = 17).

	No. (%)a
Age, median (range), y	8 (1.8-16)
Male	8 (47)
Race/ethnicity/ancestryb
Ashkenazi Jewish	6 (35)
White
Non-Hispanic	2 (12)
Hispanic	4 (24)
Black	4 (24)
Asian	1 (6)
Presenting symptoms
Median days of fever (range)	5 (1-12)
Gastrointestinal (abdominal pain, vomiting, and/or diarrhea)	15 (88)
Shock at presentation	13 (76)
Rash	12 (71)
Conjunctivitis	11 (65)
Lip redness/swelling	9 (53)
Neurologic (headache, stiff neck, vision change)	8 (47)
Respiratory (cough, dyspnea)	7 (41)
Myalgia	6 (35)
Cervical lymphadenopathy	6 (35)
Skin desquamation	3 (18)
Hypoxia at presentation	3 (18)
History of COVID-19 sick contact	11 (65)
Admission to ICU	15 (88)
Treatment of shock in ICU	10 (59)
Hypoxia in ICU	9 (53)
Mechanical ventilation	0
Treatmentc
Methylprednisolone	12 (71)
Hydrocortisone	3 (21)
Intravenous immunoglobulin	13 (76)
Anticoagulation
Enoxaparin
Prophylaxis	10 (59)
Treatment	1 (6)
Aspirin	4 (24)
Disposition
Length of stay, mean (range), d
ICU	6.4 (3-12)
Hospital	7.1 (3-18)

Abbreviations: COVID-19, coronavirus disease 2019; ICU, intensive care unit.

^a Unless otherwise indicated.

^b Race/ethnicity was classified by parent or self-reporting as documented in the electronic health record. Race/ethnicity was assessed to evaluate population at risk for disease and to contextualize to other similar diseases with known race/ethnic predilections.

^c Patients may have received more than 1 of the treatments.

Eight patients tested positive for SARS-CoV-2 by RT-PCR and the other 9 by serology. Levels of inflammatory markers were elevated in all patients, and most had lymphopenia (n = 12), bandemia (n = 11), elevated troponin T level (n = 14), and elevated N-terminal pro–brain-type natriuretic peptide (NT-proBNP) level (n = 15). Studies suggesting co-infections were infrequent (n = 3) (Table 2). Serum IL-6 level was elevated in 16. Cytokine profiling in 8 patients showed elevated IL-2R, IL-18, and CXCL 9 levels in all and mildly increased IFN-γ (n = 3) and IL-8 (n = 2) levels in some. TNF-α, IL-1b, Il-2, IL-4, IL-5, and IL-13 levels were normal. Fifteen patients required pediatric intensive care unit admission; vasoactive support was required in 10. Of 9 patients with hypoxia, none required mechanical ventilation.

Table 2. Laboratory, Echocardiogram, and Imaging Characteristics of COVID-19–Related Multisystem Inflammatory Syndrome in Children (N = 17).

Characteristic	Value
Echocardiogram findings, No. (%)
Left ventricular function by echocardiogram at admission, No. (%)
Normal	6 (35)
Mildly decreased	5 (29)
Mild-moderately decreased	4 (24)
Moderate-severely decreased	2 (12)
Pericardial effusion on admission echocardiograma	8 (47)
Positive microbiologic testing, No. (%)
EBV IgG/IgM	2 (11.7)
Parvovirus PCR	1 (5.8)
Throat culture for group A Streptococcus	1 (5.8)
Laboratory studies on admission, mean (range)
White blood cell count, ×103/μL	14.0 (4-35.9)
Neutrophils, %	76.2 (65-95)
Lymphocytes, %	9.6 (1-31)
Bands, %	6.8 (0-24)
Absolute lymphocyte count, /μL (reference, <1500/μL)	1212.1 (115.5-6444.9)
Hemoglobin, g/dL	11.2 (7.9-12.9)
Hematocrit, %	32.7 (23.2-36.9)
Platelets, ×103/μL	237 (69-892)
Serum sodium, mEq/L (reference, 137-145 mEq/L)	133.1 (125-141)
Serum bicarbonate, mEq/L (reference, 19-27 mEq/L)	18.9 (13-25)
Serum creatinine, mg/dL (reference, 0.6-1.0 mg/dL)	0.7 (0.2-3.6)
Aminotransferases, U/L
Aspartate (reference, 10-37 U/L)	51.5 (18-151)
Alanine (reference, 9-50 U/L)	49.6 (11-167)
Prothrombin time, s (reference, <12.5 s)	16.3 (13.8-19.6)
Activated partial thromboplastin time, s (reference, <36.6 s)	37.9 (27.5-85)
Lactate dehydrogenase, U/L (reference, 120-260 U/L)	362.8 (195-851)
hs-CRP, mg/L (reference, <10)	200 (17-300)
Procalcitonin, ng/mL (reference, ≤0.08 ng/mL)	21.7 (0.8-127)
Ferritin, ng/mL (reference, ≤150 ng/mL)	647.9 (83-1828)
D-dimer, mg/mL (reference, ≤0.5 μg/mL)	4 (0.9-11)
Interleukin 6, pg/mL (reference, ≤5 pg/mL)	226.3 (3.1-315)
Troponin T, high sensitivity, ng/L (reference, <22 ng/L)	56.8 (6-278)
NT-proBNP, pg/mL (reference, <207 pg/mL)	15 833 (631-59 291)

Abbreviations: COVID-19, coronavirus disease 2019; EBV, Epstein-Barr virus; hs-CRP, high-sensitivity C-reactive protein; NT-proBNP, N-terminal pro–brain-type natriuretic peptide; PCR, polymerase chain reaction.

SI conversion factors: To convert creatinine values to μmol/L, multiply by 88.4; D-dimer values to nmol/L, multiply by 5.476; aspartate and alanine aminotransferase values to μkat/L, multiply by 0.0167; lactate dehydrogenase values to μkat/L, multiply by 0.0167.

^a All pericardial effusions were trivial to small in size. None required intervention.

Fourteen patients received steroid treatment, either with methylprednisone (dose range, 2-30 mg/kg per day) or hydrocortisone (dose, 2 mg/kg per day); 1 received prednisone. Thirteen patients received intravenous immunoglobulins (dose range, 2-4 g/kg), including 3 patients who did not receive steroids and 8 who met criteria for KD. One patient received tocilizumab.

Electrocardiograms of 16 patients showed nonspecific ST/T-wave abnormalities in 10 and attenuated QRS voltage in 1. Dysrhythmias were noted in 3 (premature ventricular contractions, nonsustained ventricular tachycardia, and sinus bradycardia). Admission echocardiograms showed normal to mildly decreased left ventricular function (n = 11) or moderate or more ventricular dysfunction (n = 6). All patients had normal coronary arteries by measurement, though coronary arteries were described as prominent or echogenic in 7. Most patients had improved function on follow-up echocardiogram (range, 2-18 days from admission)—12 with normal and 1 with mildly decreased function. One patient (aged 4 years) had a medium-sized aneurysm (z score, 5.2) of the left anterior descending coronary artery. This patient presented with fever, diarrhea, and shock, with no additional features of KD. Admission testing revealed lymphopenia (absolute lymphocyte count, 540/μL); elevated levels of proBNP (44 677 pg/mL), ferritin (1195.0 ng/mL), and D-dimer (1.39 μg/mL [7.61 nmol/L]); normal troponin T level (19 ng/L); and thrombocytopenia (105 × 10³/μL). Thirteen days later, thrombocytosis developed (maximum, 671).

By May 20, after a mean total length of hospital stay of 7.1 (range, 3-18) days, all patients had been discharged home with no fatalities.

Discussion

This study describes 17 previously healthy children and adolescents who developed an inflammatory phenotype related to COVID-19. Features overlapped with, but were distinct from, those of KD and TSS. The observed pattern of cytokine expression suggests an interferon signaling component, along with IL-6 and IL-10 production, seen in KD⁵ and acute pulmonary COVID-19 infection. The lack of elevated TNF-α or IL-13 levels may differ from acute pulmonary COVID infections.⁶ The occurrence of abnormal cardiac findings suggests the need for long-term surveillance. Limitations include the small number of patients, short follow-up period, and the inability to establish causality.

Section Editor: Jody W. Zylke, MD, Deputy Editor.