Acute Kidney Injury in Pediatric Inflammatory Multisystem Syndrome Temporally Associated With Severe Acute Respiratory Syndrome Coronavirus-2 Pandemic

Experience From PICUs Across United Kingdom

Akash Deep, FRCPCH; Gaurang Upadhyay, MRCPCH; Pascale du Pré, MRCPCH; Jon Lillie, MRCPCH; Daniel Pan, MRCP; Nadeesha Mudalige, MRCPCH; Hari Krishnan Kanthimathinathan, MD; Mae Johnson, FRCA; Shelley Riphagen, MBChB; Buvana Dwarakanathan, FRCPCH; Dusan Raffaj, MD; Santosh Sundararajan, MD; Patrick Davies, MRCPCH; Zoha Mohammad, FRCPCH; Nayan Shetty, MRCPCH; Stephen Playfor, DM; Michelle Jardine, MSc; Oliver Ross, FRCA; Richard Levin, FRCP; Gareth Waters, FRCA; Ruchi Sinha, MRCPCH; Barnaby R. Scholefield, PhD; Elizabeth Boot, MBChB; Ashwani Koul, DNB(MD); Xabier Freire-Gomez, MBBS; Padmanabhan Ramnarayan, MD

Disclosures

Crit Care Med. 2020;48(12):1809-1818.

Abstract and Introduction

Abstract

Objectives: To study the prevalence, evolution, and clinical factors associated with acute kidney injury in children admitted to PICUs with pediatric inflammatory multisystem syndrome temporally associated with severe acute respiratory syndrome coronavirus-2.

Design: Multicenter observational study.

Setting: Fifteen PICUs across the United Kingdom.

Patients: Patients admitted to United Kingdom PICUs with pediatric inflammatory multisystem syndrome temporally associated with severe acute respiratory syndrome coronavirus-2 between March 14, 2020, and May 20, 2020.

Interventions: None.

Measurements and Main Results: Deidentified data collected as part of routine clinical care were analyzed. All children were diagnosed and staged for acute kidney injury based on the level of serum creatinine above the upper limit of reference interval values according to published guidance. Severe acute kidney injury was defined as stage 2/3 acute kidney injury. Uni- and multivariable analyses were performed to study the association between demographic data, clinical features, markers of inflammation and cardiac injury, and severe acute kidney injury. Over the study period, 116 patients with pediatric inflammatory multisystem syndrome temporally associated with severe acute respiratory syndrome coronavirus-2 were admitted to 15 United Kingdom PICUs. Any-stage acute kidney injury occurred in 48 of 116 patients (41.4%) and severe acute kidney injury in 32 of 116 (27.6%) patients, which was mostly evident at admission (24/32, 75%). In univariable analysis, body mass index, hyperferritinemia, high C-reactive protein, Pediatric Index of Mortality 3 score, vasoactive medication, and invasive mechanical ventilation were associated with severe acute kidney injury. In multivariable logistic regression, hyperferritinemia was associated with severe acute kidney injury (compared with nonsevere acute kidney injury; adjusted odds ratio 1.04; 95% CI, 1.01–1.08; p = 0.04). Severe acute kidney injury was associated with longer PICU stay (median 5 days [interquartile range, 4–7 d] vs 3 days [interquartile range, 1.5–5 d]; p < 0.001) and increased duration of invasive mechanical ventilation (median 4 days [interquartile range, 2–6 d] vs 2 days [interquartile range, 1–3 d]; p = 0.04).

Conclusions: Severe acute kidney injury occurred in just over a quarter of children admitted to United Kingdom PICUs with pediatric inflammatory multisystem syndrome temporally associated with severe acute respiratory syndrome coronavirus-2. Hyperferritinemia was significantly associated with severe acute kidney injury. Severe acute kidney injury was associated with increased duration of stay and ventilation. Although short-term outcomes for acute kidney injury in pediatric inflammatory multisystem syndrome temporally associated with severe acute respiratory syndrome coronavirus-2 appear good, long-term outcomes are unknown.

Introduction

Novel coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) was declared a global pandemic by World Health Organization (WHO) in March 2020, and by June 20, 2020, nearly 9 million people got affected and resulted in nearly half a million deaths.^[1] Initial reports from China, confirmed subsequently from Europe and North America, indicated that children appear to be affected less frequently and less severely by COVID-19.^[2] However, from March onward, clinicians in the United Kingdom (UK), Europe, and the United States (US) started reporting children with an unexplained inflammatory condition possibly associated with COVID-19. Case definitions for this condition, called pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS) pandemic in the UK and multisystem inflammatory syndrome in children (MIS-C), in the US, have now been published by the UK Royal College of Pediatrics and Child Health (RCPCH), the US Centers for Disease Control and Prevention, and the WHO.^[3–5] Diagnostic criteria common to all case definitions include presence of fever, inflammation, and multiple organ involvement, predominantly cardiac dysfunction and shock. Published reports of this inflammatory condition (referred hereafter as PIMS-TS) indicate that it shares features of, but is distinct from, other inflammatory conditions such as Kawasaki disease (KD), toxic shock syndrome, and KD shock syndrome.^[6–12]

Approximately 10% of all patients admitted to the PICUs develop acute kidney injury (AKI), the frequency of which increases with increasing severity of patient illness.^[13,14] Worsening severity of AKI has been associated with a stepwise increase in 28-day mortality.^[15] In the largest case series of PIMS-TS published so far (n = 58, 29 of whom required PICU admission), elevation of serum creatinine above upper limit for age was seen in 22% of cases, although further details regarding factors associated with AKI in this condition, or details of progression of AKI and its relationship with patient outcomes, were not reported.^[8] The etiology and pathogenesis of AKI may be multifactorial: it could develop in PIMS-TS as a part of multisystem involvement secondary to hypovolemia, low cardiac output state, vasculitis, or immune-mediated inflammation. AKI is also a known complication in KD and is reported in about one-third of these patients.^[16] In adults with typical features of acute COVID-19 infection, AKI has been reported in approximately 30% of patients.^[17–19] Since PIMS-TS is a postinfectious inflammatory response condition, complications may be substantially different to those seen in active SARS-CoV-2 infection. Factors associated with AKI in PIMS-TS, its course, and relationship with patient outcomes are currently unknown. In this report, we aim to describe the prevalence, evolution, and clinical factors associated with AKI in a cohort of children admitted to UK PICUs with PIMS-TS over a 9-week period from March 2020 to May 2020.

1 2 3 4 5

Next Section

Table 1. Characteristics of Study Participants Cross Tabulated by Acute Kidney Injury Stage During PICU Admission
Variables	AKI Stage 0/1 (n = 84)	AKI Stage 2/3 (n = 32)	p
Demographics
Median age, yr (IQR)	11 (7–14)	10 (8–14)	0.66
Male (%)	56 (67)	20 (63)	0.67
Ethnicity (%)
Afro-Caribbean	30 (37)	21 (66)	0.03^a
Asian	23 (28)	6 (19)
Caucasian	19 (23)	5 (16)
Other	9 (11)	0 (0)
Median body mass index, kg/m² (IQR)	19.6 (16.3–24.2)	23.2 (17.5–30.3)	0.02^a
SARS-CoV-2 status
SARS CoV-2 polymerase chain reaction (%)	13 (15)	6 (19)	0.67
SARS-CoV-2 immunoglobulin G
Positive (%)	39 (46)	17 (53)
Admission parameters
Median Paediatric Index of Mortality 3 score, % (IQR)	3.9 (1.4–4.8)	6.2 (1.6–10.8)	0.048^a
Median lactate, mmol/L(IQR)	1.5 (1.2–2.5)	2.1 (1.2–3.3)	0.14
Median ferritin, ug/L (IQR)	426 (209–958)	1,107 (393–2,668)	0.009^a
Median D-dimers, ng/mL (IQR)	3,745 (2,282–6,288)	5,000 (2,383–7,920)	0.32
Median C-reactive protein, mg/L (IQR)	190 (110–266)	250 (145–332)	0.05^a
Median lactate dehydrogenase, U/L (IQR)	415 (270–656)	600 (367–1,049)	0.03^a
Median troponin, ng/L (IQR)	60 (16–473)	436 (143–1,441)	0.001^a
Median eGFR, mL/min/1.73 m²	45 (14–79)	55 (46–68)	0.15
Median creatinine, umol/L	49 (32–66)	103 (78–150)	0.001
PICU management
Admission urine output at 24 hr, mL/kg/hr (IQR)	1.1 (0.7–1.8)	1.1 (0.66–2)	0.8^a
Admission fluid bolus (%)	51 (61)	23 (72)	0.26
Vasodilated shock (%)	39 (46)	18 (56)	0.34
Vasoconstricted shock (%)	8 (10)	6 (19)	0.17
Number of nephrotoxic agents (%)
0	61 (73)	19 (60)	0.08^a
1	14 (17)	3 (9)
2	8 (9)	6 (19)
3	1 (1)	4 (12)
Vasopressor (%)	39 (46)	24 (75)	0.006
Inotropic agent (%)	49 (58)	25 (78)	0.047
Invasive mechanical ventilation (%)	22 (26)	19 (59)	0.001
Median duration of invasive mechanical ventilation, d (IQR)	2 (1–3)	4 (2–6)	0.04^a
Median duration of PICU stay, d (IQR)	3 (1.5– 5)	5 (4–7)	< 0.001^a
Day 7 eGFR, mL/min/1.73 m²	134 (111–149)	126 (89–156)	0.5
Day 7 creatinine, umol/L	43 (36–46)	43 (30–67)	0.6

AKI = acute kidney injury, eGFR = estimated glomerular filtration rate, IQR = interquartile range, SARS-CoV-2 = severe acute respiratory syndrome coronavirus-2.
^aNonparametric hypothesis test used due to violation of normality assumption.

Table 2. Univariable Odds of Severe Acute Kidney Injury During PICU Admission As Estimated by Logistic Regression Modeling
Variables	OR (95% CI)	Wald z Statistic	p
Demographics
Age (per 1 yr increase)	1.02 (0.93–0.12)	0.45	0.66
Male	1.2 (0.51–2.80)	0.42	0.67
Ethnicity
Caucasian	Referent
Afro-Caribbean	2.66 (0.86–8.25)	1.69	0.09
Asian	0.99 (0.26–3.76)	–0.01	0.99
Body mass index (per 1-unit increase in kg/m ²)	1.02 (1.001–1.035)	20.5	0.04
SARS-CoV-2 status
SARS CoV-2 PCR	1.26 (0.43–3.66)	0.43	0.67
Admission parameters
Lactate, mmol/L(per 1-unit increase)	1.20 (0.93–1.59)	1.41	0.16
Ferritin, ug/L (per 100-unit increase)	1.04 (1.01–1.08)	2.75	0.006
D-dimers, ng/ml (per 1,000-unit increase)	–0.01 (–0.07 to 0.44)	–0.42	0.68
C-reactive protein, mg/L (per 100-unit increase)	1.55 (1.05–2.28)	2.22	0.03
Lactate dehydrogenase, U/L (per 100-unit increase)	1.04 (0.98–1.09)	1.35	0.18
Troponin, ng/L (per 100-unit increase)	1.01 (1–1.02)	1.16	0.25
PICU management
Paediatric Index of Mortality 3 score (per 10 % increase)	2.32 (1.04–5.19)	2.04	0.04
Median urine output at 24 hr, mL/kg/hr (per 1-unit increase)	1.09 (0.77–1.56)	0.49	0.62
Admission fluid bolus	1.65 (0.68–4.01)	1.11	0.27
Vasodilated shock	1.48 (0.65–3.37)	0.94	0.35
Vasoconstricted shock	2.19 (0.70–6.91)	1.34	0.18
Nephrotoxic agents	1.81 (0.77–4.26)	1.37	0.17
Vasopressor	3.46 (1.40–8.58)	2.68	0.007
Inotropic agent	2.55 (0.99–6.55)	1.94	0.05
Invasive mechanical ventilation	4.12 (1.75–9.70)	3.24	0.001
Duration of invasive mechanical ventilation (per one day increase)	1.24 (0.99–1.56)	1.84	0.06
Duration of PICU stay (per one day increase)	1.27 (1.09–1.59)	3.07	0.002

OR = odds ratio, SARS-CoV-2 = severe acute respiratory syndrome coronavirus-2.

Table 3. Multivariable Odds of Severe Acute Kidney Injury During PICU Admission As Estimated by Logistic Regression Modeling (Area Under the Curve = 0.74)
Outcome: Severe AKI (AKI Stage 2/3)	Multivariable Analysis (Base Model)
Outcome: Severe AKI (AKI Stage 2/3)	OR (95% CI)	Wald z Statistic	p
Body mass index (per 1-unit increase in kg/m²)	1.06 (0.98–1.13)	1.49	0.14
Ferritin, ug/L (per 100-unit increase)	1.04 (1.01–1.08)	2.09	0.04
Paediatric Index of Mortality 3 score (per 10% increase)	1.72 (0.58–5.13)	0.97	0.33

AKI = acute kidney injury, OR = odds ratio.

Authors and Disclosures

Akash Deep, FRCPCH^1,2, Gaurang Upadhyay, MRCPCH¹, Pascale du Pré, MRCPCH³, Jon Lillie, MRCPCH⁴, Daniel Pan, MRCP⁵, Nadeesha Mudalige, MRCPCH⁶, Hari Krishnan Kanthimathinathan, MD⁷, Mae Johnson, FRCA³, Shelley Riphagen, MBChB⁴, Buvana Dwarakanathan, FRCPCH⁸, Dusan Raffaj, MD⁹, Santosh Sundararajan, MD¹⁰, Patrick Davies, MRCPCH⁹, Zoha Mohammad, FRCPCH¹¹, Nayan Shetty, MRCPCH¹², Stephen Playfor, DM¹³, Michelle Jardine, MSc¹⁴, Oliver Ross, FRCA¹⁵, Richard Levin, FRCP¹⁶, Gareth Waters, FRCA⁴, Ruchi Sinha, MRCPCH¹⁷, Barnaby R. Scholefield, PhD^7,18, Elizabeth Boot, MBChB⁴, Ashwani Koul, DNB(MD)¹⁹, Xabier Freire-Gomez, MBBS⁴ and Padmanabhan Ramnarayan, MD^17,20

¹Pediatric Intensive Care Unit, King's College Hospital NHS Foundation Trust, London, United Kingdom.
²Department of Women and Children's Health, School of Life Course Sciences, King's College London, London, United Kingdom.
³Pediatric Intensive Care Unit, Great Ormond Street Hospital, London, United Kingdom.
⁴Pediatric Intensive Care Unit, Evelina Children's Hospital, London, United Kingdom.
⁵Department of Respiratory Sciences, University of Leicester, Leicester, United Kingdom.
⁶UCL Great Ormond Street Institute of Child Health, London, United Kingdom.
⁷Pediatric Intensive Care Unit, Birmingham Women's and Children's NHS Foundation Trust, Birmingham, United Kingdom.
⁸Pediatric Intensive Care Unit, St George's Hospital, London, United Kingdom.
⁹Pediatric Critical Care Unit, Nottingham Children's Hospital, Nottingham, United Kingdom.
¹⁰Pediatric Intensive Care Unit, Leeds Children's Hospital, Leeds, United Kingdom.
¹¹Pediatric Intensive Care Unit, Leicester Royal Infirmary, Leicester, United Kingdom.
¹²Pediatric Intensive Care Unit, Alder Hey Children's Hospital, Liverpool, United Kingdom.
¹³Pediatric Intensive Care Unit, Royal Manchester Children's Hospital, Manchester, United Kingdom.
¹⁴Pediatric Critical Care Unit, Children's Hospital for Wales, Cardiff, United Kingdom.
¹⁵Pediatric Intensive Care Unit, Southampton Children's Hospital, Southampton, United Kingdom.
¹⁶Pediatric Intensive Care Unit, Royal Hospital for Children, Glasgow, United Kingdom.
¹⁷Pediatric Intensive Care Unit, St Mary's Hospital, London, United Kingdom.
¹⁸Birmingham Acute Care Research Group, Institute of Inflammation and Ageing, University of Birmingham, Birmingham, United Kingdom.
¹⁹Pediatric Critical Care Unit, John Radcliffe Hospital, Oxford, United Kingdom.
²⁰Children's Acute Transport Service, Great Ormond Street Hospital NHS Foundation Trust and NIHR Biomedical Research Centre, London, United Kingdom.

Address requests for reprints to
Akash Deep, FRCPCH, Pediatric Intensive Care Unit, King's College Hospital NHS Foundation Trust, London SE5 9RS, United Kingdom. E-mail: akash.deep@nhs.net

Dr. Pan disclosed that he is funded by the National Institute for Health Research (NIHR) and he received support for article research from NIHR. The remaining authors have disclosed that they do not have any potential conflicts of interest.