AKI in COVID-19–Associated Multisystem Inflammatory Syndrome in Children (MIS-C)

Abstract

Visual Abstract

Background 

Multisystem inflammatory syndrome in children (MIS-C) is a recently identified entity in association with COVID-19. AKI has been widely reported in patients with primary COVID-19 infection. However, there is a paucity of literature regarding renal injury in MIS-C. We aim to characterize AKI in MIS-C in this cohort identified at a major children's hospital in New York City during the COVID-19 pandemic.

Methods 

We conducted a retrospective cohort study of children 0–20 years old admitted to Morgan Stanley Children's Hospital (MSCH) between April 18th and September 23rd, 2020. Patients were included if they met criteria for MIS-C on the basis of CDC guidelines. All patients were evaluated for the presence of AKI, and AKI was staged according to KDIGO criteria.

Results 

Of the 57 children who met inclusion criteria, 46% (26 of 57) were found to have AKI. The majority of patients (58%; 15 of 26) were classified as KDIGO stage 1. AKI was present upon admission in 70% of those identified. All patients had resolution of AKI at discharge, with 61% achieving recovery by day 2. One patient required dialysis. When compared with those without renal injury, the AKI cohort was older (P<0.001) and had higher median peak values of CRP (P<0.001), IL-6 (P=0.02), ferritin (P<0.001), and procalcitonin (P=0.02). More patients with AKI had left ventricular systolic dysfunction (P<0.001) and lymphopenia (P=0.01) when compared with those without AKI. No differences in body mass index or sex were found.

Conclusions 

Although children with MIS-C may develop AKI, our study suggests that most experience mild disease, swift resolution, and promising outcome. Older age, increased inflammation, and left ventricular systolic dysfunction may be risk factors. Our study highlights the substantial differences in epidemiology and outcomes between AKI associated with pediatric MIS-C versus primary COVID-19 infection.

In Brief

Key Points

• AKI was a common finding among pediatric patients hospitalized with COVID-19–associated MIS-C.
• Older age, increased inflammation, and left ventricular systolic dysfunction may be risk factors for AKI in children with MIS-C.
• Although children with MIS-C may develop AKI, our study suggests that most experience mild disease, swift resolution, and promising outcome.

Introduction

As the coronavirus disease 2019 (COVID-19) pandemic continues around the world, manifestations of infection by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus in children continue to be an active area of research. A syndrome currently named “multisystem inflammatory syndrome in children” (MIS-C) was first described in late April 2020 by clinicians in the United Kingdom, who recognized previously healthy children presenting with a severe inflammatory syndrome after testing positive for concurrent or recent infection of COVID-19 (¹). The syndrome was soon thereafter recognized in the United States, and as of December 4, 2020, 1288 American patients had been identified by the Centers for Disease Control and Prevention (CDC) (²).

CDC characterizes MIS-C by persistent fever and laboratory markers of inflammation, with evidence of severe illness requiring hospitalization and multiorgan involvement (e.g., cardiac, gastrointestinal, renal, hematologic, dermatologic, and neurologic) (²). Common manifestations include diminished left ventricular systolic function with or without coronary artery changes (^3–5). The affected individual must be <21 years old and have had an exposure to a confirmed or suspected patient with COVID-19. Patients may exhibit some or all features of Kawasaki disease (^1–3).

AKI has been widely reported in patients with primary COVID-19 infection (⁶,⁷). Recently, a multicenter study reported on the prevalence of AKI among critically ill children with primary COVID-19 infection (⁸). However, there have been limited studies describing the incidence and characteristics of renal complications in MIS-C (²,³,⁹). Previous experience taking care of adult patients with COVID-19 in our center has given us the opportunity to be able to highlight the substantial differences in epidemiology and outcomes in patients with AKI in MIS-C in contrast to AKI from primary COVID-19 infection (¹⁰). Here, we report the characteristics of AKI in children diagnosed with MIS-C in a large, tertiary, free-standing children's hospital in New York City, one of the early epicenters of the COVID-19 pandemic in the United States. To our knowledge, this is the first study to date specifically focused on describing the characteristics of pediatric AKI in MIS-C.

Materials and Methods

Study Population and Data Collection

A retrospective chart review was performed to identify children <21 years of age who were admitted to Columbia University Irving Medical Center/Morgan Stanley Children's Hospital of the NewYork-Presbyterian System with symptoms and clinical findings consistent with MIS-C between April 18 and September 23, 2020. Chart review was carried out with approval of the institutional review board. Patients were included if they met criteria for MIS-C, as defined according to CDC guidelines (²). Diagnosis criteria included fever and laboratory evidence of inflammation, with multisystem organ involvement, as detailed in Table 1. Patients either tested positive for current or recent SARS-CoV-2 viral infection by RT-PCR, serology, or antigen test or had exposure to a suspected or confirmed patient with COVID-19 within the 4 weeks prior to the onset of symptoms. Patients with positive PCR testing were distinguished from primary COVID-19 infection by lack of pulmonary involvement or otherwise did not fit the clinical diagnosis of primary infection.

Table 1. - Centers for Disease Control and Prevention criteria for multisystem inflammatory syndrome in children

CDC Criteria for MIS-C
Age <21 yr
Clinical presentation consistent with MIS-C, including all of the following
Fever
Documented fever >38.0°C (100.4°F) for ≥24 h or report of subjective fever lasting ≥24 h
Laboratory evidence of inflammation including, but not limited to, any of the following
Elevated CRP
Elevated ESR
Elevated fibrinogen
Elevated procalcitonin
Elevated d -dimer
Elevated ferritin
Elevated LDH
Elevated IL-6 level
Neutrophilia
Lymphocytopenia
Hypoalbuminemia
Severe illness requiring hospitalization
Multisystem involvement—two or more organ systems involved
Cardiovascular (e.g., shock, elevated troponin, elevated BNP, abnormal echocardiogram, arrhythmia)
Respiratory (e.g., pneumonia, ARDS, pulmonary embolism)
Renal (e.g., AKI, renal failure)
Neurologic (e.g., seizure, stroke, aseptic meningitis)
Hematologic (e.g., coagulopathy)
Gastrointestinal (e.g., elevated liver enzymes, diarrhea, ileus, gastrointestinal bleeding)
Dermatologic (e.g., erythroderma, mucositis, other rash)
No alternative plausible diagnoses
Recent or current SARS-CoV-2 infection or exposure

MIS-C, multisystem inflammatory syndrome in children; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; LDH, lactate dehydrogenase; BNP, B-type natriuretic peptide; ARDS, acute respiratory distress syndrome; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Demographics. including age and sex. were recorded for all patients. As data review was limited to the electronic medical record, race was not consistently documented. Patients were, however, listed as Hispanic or non-Hispanic ethnicity. Serum creatinine values, as measured by the enzymatic method, were noted at both peak and nadir of hospitalization. Given lack of previous laboratory values for most children, hospitalization nadir was treated as patients’ baseline creatinine. Additional clinical criteria, including body mass index percentile, as well as other laboratory values, such as peak value of inflammatory markers, were reviewed for each patient.

AKI was defined according to the 2012 Kidney Disease Improving Global Outcomes (KDIGO) classification that is applicable to both adult and pediatric populations (¹¹,¹²). Given the inconsistencies in documentation regarding urine output, patients were classified on the basis of serum creatinine alone, as detailed in Table 2. Factors potentially contributing to AKI were recorded and included: exposure to nephrotoxic medications commonly used in our hospital (i.e., aminoglycosides, trimethoprim/sulfamethoxazole, vancomycin, acyclovir, nonsteroidal anti-inflammatory drugs, iodine-based contrast for imaging, and calcineurin inhibitors), hypotension requiring vasopressors, and echocardiographic findings of decreased left ventricular systolic function. Two time points were recorded: time to peak creatinine (day of hospitalization) as well as time to recovery (days) defined as return to baseline/nadir creatinine. The presence of significant proteinuria was defined as ≥2+ protein (100 mg/dl) on urinalysis. Patients for whom certain data were unavailable or missing were excluded from analysis regarding that factor. Data with values that were outside a laboratory reference range were analyzed as the closest value (minimum or maximum) within range.

Table 2. - Kidney Disease Improving Global Outcomes classification

Kidney Disease Improving Global Outcomes Stage	Serum Creatinine
1	1.5–1.9 times baseline or 0.3-mg/dl increase
2	2.0–2.9 times baseline
3	3.0 times baseline or increase in serum creatinine to 4.0 mg/dl or initiation of RRT or decrease in eGFR to 35 ml/min per 1.73 m2 (in patients over 18 yr)

Statistical Analyses

Statistical analyses were performed using Prism 7 (Graphpad Inc., San Diego, CA) and SPSS Statistics, Version 26.0. Continuous data were presented as medians and ranges. Continuous variables were compared using the Mann–Whitney test, whereas categorical variables were compared using the Fisher exact test. Cox proportional hazards model univariate analysis was performed. Significance was defined as P<0.05.

No covariables were found to be significant, and therefore, we did not proceed with multivariate analysis.

Results

Our cohort included 58 patients admitted during the defined period and meeting criteria for MIS-C. One patient was identified as meeting criteria for MIS-C and developed severe AKI while at an outside hospital that had resolved prior to transfer to our center. Data during that hospitalization were limited, and therefore, this patient was excluded from our group.

The majority of patients (74%) were negative for active SARS-CoV-2 infection by PCR at time of hospitalization; the remaining 26% had positive PCR results. Additionally, only 52 patients were tested for SARS-CoV-2 antibodies, 88% of which had positive serology. Five patients had negative serology but met CDC criteria for MIS-C with suspicion of exposure. Five patients were not tested for antibodies, and one patient had indeterminant serology; however, all met CDC criteria for MIS-C (Figure 1).

Figure 1.: Flow chart of enrollment of patients who met criteria for multisystem inflammatory syndrome in children (MIS-C), and results of SARS-CoV-2 testing, RT-PCR and serology. ED, emergency department.

AKI occurred in 26 of the 57 (46%) patients: stage 1 AKI (n=15, 58%), stage 2 AKI (n=7, 27%), and stage 3 AKI (n=4, 15%). Continuous RRT was initiated in only one patient, due to worsening uremia. The majority of patients with AKI (70%) presented with AKI at the time of admission, and an additional 27% developed AKI within 24–72 hours of hospitalization. All patients with AKI recovered renal function, with 19% recovering within 24 hours. Most patients (75%) with stage 3 AKI required >3 days to recover. Comparatively, only 20% and 43% of patients with stages 1 and 2 AKI, respectively, required a recovery time >3 days (Figure 2). The patient who required dialysis recovered after approximately 3 weeks.

Figure 2.: Time to AKI recovery by KDIGO stage. Kaplan–Meier curve outlining time to recovery of AKI by KDIGO stage.

Of the 57 eligible patients, 46% were girls, and 54% were boys. The median age of the cohort was 7 years (range, 8 months to 20 years old). There was no statistically significant difference between groups with regard to sex, Hispanic versus non-Hispanic ethnicity, or elevated body mass index (Table 3). The AKI group was older, with a median age of 10 years as compared with 4 years (P<0.001). The majority of patients with AKI (81%) were admitted to the pediatric intensive care unit.

Table 3. - Demographics and clinical characteristics of patients with multisystem inflammatory syndrome in children

Variables	AKI, n=26	Non-AKI, N=31	P Value
Age, yr
Median (range)	10 (2–20)	4 (0–17)	<0.001
Sex
Boys	14/26=54%	14/31=45%	0.60
Girls	12/26=46%	17/31=55%
Hispanic ethnicity	8/26=31%	9/31=29%	>0.99
BMI≥85th percentile	N=26; 15/26=58%	N=30; 13/30=43%	0.42
LV systolic dysfunction	N=26; 15/26=58%	N=28; 2/28=7%	<0.001
Lymphopenia	24/26=92%	18/31=26%	0.006
Defined as <1500/mm3
IL-6, pg/ml	N=20	N=15	0.02
Median (range)	274.5 (32.0–315.0)	46.7 (3.0–315.0)
Peak ferritin, ng/ml	N=26	N=29	<0.001
Median (range)	670.5 (284–100,000)	233.1 (19.9–2769)
Peak fibrinogen, mg/dl	N=26	N=24	0.10
Median (range)	624 (166–875)	501.5 (273–838)
Peak d -dimer, μg/ml FEU	N=26	N=28	0.35
Median (range)	3.9 (0.5–426)	2.9 (0.3–20.0)
Peak LDH, U/L	N=25	N=27	0.81
Median (range)	339 (207–4087)	356 (178–1295)
Peak CRP, mg/L	N=26	N=30	0.01
Median (range)	211 (29.4–300)	135.1 (0.21–300.0)
Peak ESR, mm/h	N=20	N=23	0.97
Median (range)	56.5 (23–130)	59.0 (0–130)
Peak procalcitonin, ng/ml	N=23	N=24	0.02
Median (range)	3.6 (0.2–126.9)	1.6 (0.1–42.1)
Significant proteinuria	N=25	N=29	0.58
Defined as ≥2+ on dipstick	11/25=44%	10/29=34%

Patients for whom certain data were unavailable or missing were excluded from analysis regarding that factor. Lymphopenia is defined as having an absolute lymphocyte count <1500/mm³ (or <2000/mm³ in children younger than 6 years of age). BMI, body mass index; LV, left ventricular; LDH, lactate dehydrogenase; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.

Within the AKI group with a documented urinalysis, we found that 44% had significant proteinuria ≥2+ (100 mg/dl), but there was no significant difference in proteinuria between those with AKI and those without (P=0.58). Half of the patients with AKI (13 of 26) had an abdominal and/or renal sonogram, of which only one sonogram revealed renal abnormalities with increased cortical echogenicity. Additional risk factors for AKI were evaluated: nephrotoxin exposure prior to AKI was identified in 54% of patients, and 69% of patients with AKI required vasopressor support (Table 4). All patients who required fluid resuscitation received normal saline crystalloid boluses in accordance with pediatric life support guidelines (¹³,¹⁴).

Table 4. - Clinical characteristics of the AKI cohort

Clinical Characteristics	AKI, n=26
Bed location
ICU	21 (81%)
General ward	5 (19%)
Nephrotoxic medications a	14 (54%)
Decreased left ventricular systolic function	15 (58%)
Vasopressors	18 (70%)
Mechanical ventilation	1 (4%)
Dialysis	1 (4%)
Treatment
Steroids	26 (100%)
IVIG	21 (81%)
Anakinra	5 (19%)

ICU, intensive care unit; IVIG, intravenous Ig.

^aNephrotoxic medications include aminoglycosides, trimethoprim/sulfamethoxazole, vancomycin, acyclovir, nonsteroidal anti-inflammatory drugs, iodine-based contrast for imaging, and calcineurin inhibitors.

Patients with AKI were more likely to have left ventricular systolic dysfunction (P<0.001). Peak inflammatory markers, including IL-6, ferritin, fibrinogen, d-dimer, lactate dehydrogenase, C-reactive protein, erythrocyte sedimentation rate, and procalcitonin, as well as lymphopenia, defined as having an absolute lymphocyte count <1500/mm³ or <2000/mm³ in children younger than 6 years of age (¹⁵), were compared between the two groups (Table 3). We found a statistically significant difference between the AKI and non-AKI groups when comparing peak inflammatory markers of C-reactive protein (P=0.01), IL-6 (P=0.02), procalcitonin (P=0.02), and ferritin (P<0.001). Similar to adult patients with AKI due to primary COVID-19 infection (¹⁶), lymphopenia occurred in patients with MIS-C and occurred more often in the AKI group (P=0.01). Fibrinogen, d-dimer, lactate dehydrogenase, and erythrocyte sedimentation rate were not significantly different across groups.

Discussion

Although AKI has been reported as a prominent feature in primary COVID-19 infection (⁶,⁷,¹⁰,^16–18), there have been limited studies describing AKI in MIS-C (²,³,⁹). To our knowledge, we present the first study to describe the characteristics of AKI in pediatric patients with MIS-C. We showed that, similar to adult patients with primary COVID-19 infection (⁷), AKI in MIS-C occurred early in hospitalization, with 97% of AKI occurring within 72 hours of admission. Importantly, resolution of AKI usually occurred rapidly, with 61% recovering within 3 days. The majority of patients had cases that were mild (KDIGO stage 1 AKI). More severe AKI took longer to resolve, but only one patient required dialysis (Figure 2). This swift improvement of AKI with medical management is in stark contrast to the clinical course of AKI from primary COVID-19 infection; need for RRT has been reported at 55% and incidence of mortality has been reported at 50% in adult patients with AKI and primary COVID-19 infection (⁶).

A limitation of our study is that we were unable to include values in analysis that occurred outside our laboratory's reference range, and therefore, this may have skewed significance. For example, the lowest creatinine value measured by our laboratory is <0.2 mg/dl, but a value of 0.2 mg/dl was used for these patients. Likewise, several patients had IL-6 levels >315 pg/ml but were analyzed using a value of 315 pg/ml. This limitation may have overestimated baseline creatinine and therefore, underestimated AKI incidence, as well as blunted significance in inflammation between the AKI and non-AKI groups.

Given lack of previous laboratory values for most children, hospitalization nadir was treated as patients’ baseline creatinine. However, unlike some adult patients with AKI in primary COVID-19 infection (⁶), none of the patients in our study had documented CKD prior to hospitalization. The predominance of mild AKI in MIS-C with rapid recovery of renal function suggests transient hypoperfusion, ischemia, or hypoxia. Although the presenting symptoms of patients with MIS-C varied, they all included persistent fever by clinical definition. This, along with vomiting in some patients, may have contributed to dehydration. Additionally, most patients required intensive care unit admission mostly due to hypotension, with 69% requiring vasopressor support. A majority of patients with AKI (58%) had decreased left ventricular systolic function at the time of AKI, another factor contributing to decreased effective renal perfusion. Although the majority of patients with AKI followed a clinical course most consistent with prerenal injury, other etiologies of AKI, such as nephrotoxin exposure and acute tubular necrosis, might have contributed as well. This does underscore the importance of long-term follow-up to ensure adequate renal function as these children grow into the adulthood.

The pathogenesis of AKI in children with MIS-C, which is considered a postviral inflammatory response, is not clearly understood at this time. Our data indicate that patients with MIS-C and AKI have a greater degree of inflammation compared with those who did not. These findings raise the concern whether inflammation- and cytokine-mediated hypotension leads to renal hypoperfusion, as other systemic inflammation, such as sepsis, has been associated with AKI (¹⁹). It is unknown whether intrarenal inflammation was present and contributed to AKI because none of our patients underwent a renal biopsy, either due to lack of indication or due to being clinically unstable to undergo biopsy. To treat the inflammatory condition, all patients with MIS-C and AKI received steroids, and 81% of them also received intravenous Ig for Kawasaki-like features. Nineteen percent of patients who had AKI were also given anakinra as a third-line agent. Effective anti-inflammatory treatment may facilitate prompt renal recovery.

Previous studies have reported younger age as a risk factor for pediatric multiple organ dysfunction syndrome, likely owing to the difference in physiology between neonates and older children (²⁰,²¹). No neonates were included in our population, with the youngest infant being 8 months of age, and the median age of our cohort was 7 years old. Our study found that the median age for children with AKI was significantly older than that of children without AKI during MIS-C. It is possible that younger children may have been brought to medical attention sooner than those of older age, allowing for timely medical interventions that prevented or reduced renal injury and facilitated renal recovery. Because our study only includes children who presented to the emergency department and/or were hospitalized and does not include those in ambulatory settings, older children with mild signs and symptoms resembling MIS-C may have been under-represented.

MIS-C is a new disease entity, and we captured 58 children with the diagnosis during a 6-month period, including the height of the pandemic in New York. The relatively small sample size did not permit statistical analysis designed for data of normal distribution. Additionally, the sequence between clinical presentation and laboratory finding with the occurrence and resolution of AKI was not examined. For example, only peak but not the trend of inflammatory markers was recorded, and timing of administering fluid, vasopressors, and immunomodulatory agents as related to AKI was not studied. However, we did observe that shock and AKI usually occurred early in the hospitalization, which supports renal hypoperfusion as a major contributing factor in causing AKI.

In conclusion, we were able to characterize AKI as it occurs in MIS-C and highlight substantial differences between AKI associated with MIS-C versus those with primary COVID-19 infection. We found that AKI was a common finding among pediatric patients hospitalized with MIS-C, with a significant difference in age and inflammatory markers, as well as cardiac function, between the MIS-C groups with and without AKI. AKI appears to be mild and has a short recovery period, which differs from AKI from primary COVID-19 infection. However, we have only had short follow-up since the resolution of AKI. Further clinical and translational studies are required to have a more complete understanding of this novel medical condition that involves multiple organs in children.

Disclosures

I. Batal reports research funding from the American Society of Transplantation outside the current work. F. Lin is a member on the JASN editorial board. All remaining authors have nothing to disclose.

Funding

None.