key: cord-0727473-9pk3duyg
authors: Nino, Gustavo; Molto, Jose; Aguilar, Hector; Zember, Jonathan; Sanchez‐Jacob, Ramon; Diez, Carlos T.; Tabrizi, Pooneh R.; Mohammed, Bilal; Weinstock, Jered; Xuchen, Xilei; Kahanowitch, Ryan; Arroyo, Maria; Linguraru, Marius G.
title: Chest X‐ray lung imaging features in pediatric COVID‐19 and comparison with viral lower respiratory infections in young children
date: 2021-09-15
journal: Pediatr Pulmonol
DOI: 10.1002/ppul.25661
sha: 139785ca42341c85dd92ed521e7e95639917b24f
doc_id: 727473
cord_uid: 9pk3duyg

RATIONALE: Chest radiography (CXR) is a noninvasive imaging approach commonly used to evaluate lower respiratory tract infections (LRTIs) in children. However, the specific imaging patterns of pediatric coronavirus disease 2019 (COVID‐19) on CXR, their relationship to clinical outcomes, and the possible differences from LRTIs caused by other viruses in children remain to be defined. METHODS: This is a cross‐sectional study of patients seen at a pediatric hospital with polymerase chain reaction (PCR)‐confirmed severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) (n = 95). Patients were subdivided in infants (0–2 years, n = 27), children (3–10 years, n = 27), and adolescents (11–19 years, n = 41). A sample of young children (0–2 years, n = 68) with other viral lower respiratory infections (LRTI) was included to compare their CXR features with the subset of infants (0–2 years) with COVID‐19. RESULTS: Forty‐five percent of pediatric patients with COVID‐19 were hospitalized and 20% required admission to intensive care unit (ICU). The most common abnormalities identified were ground‐glass opacifications (GGO)/consolidations (35%) and increased peribronchial markings/cuffing (33%). GGO/consolidations were more common in older individuals and perihilar markings were more common in younger subjects. Subjects requiring hospitalization or ICU admission had significantly more GGO/consolidations in CXR (p < .05). Typical CXR features of pediatric viral LRTI (e.g., hyperinflation) were more common in non‐COVID‐19 viral LRTI cases than in COVID‐19 cases (p < .05). CONCLUSIONS: CXR may be a complemental exam in the evaluation of moderate or severe pediatric COVID‐19 cases. The severity of GGO/consolidations seen in CXR is predictive of clinically relevant outcomes. Hyperinflation could potentially aid clinical assessment in distinguishing COVID‐19 from other types of viral LRTI in young children.

A pneumonia of unknown cause detected in the city of Wuhan in Hubei province (China) was first reported to the World health Organization (WHO) office in China on December 31, 2019. 1 The disease was then confirmed to be caused by a novel coronavirus (SARS-CoV-2) and later termed coronavirus disease 2019 . This potentially lethal disease spread quickly in the world and was declared a global pandemic in March 2020. Notably, COVID-19 has become the most lethal pandemic in the modern times with millions of deaths attributed to COVID-19 worldwide. 1 Initially, the risk of serious illness or mortality was thought to be of exclusive concern to adults and the elderly. However, new facts have made clear that children are also at risk for hospitalization and severe health complications. [2] [3] [4] [5] [6] According to the American Academy of Pediatrics, children make up to 10% of COVID-19 infections, but less than 2% of the literature on the virus has focused on children. 7 Although children with SARS-CoV-2 infections are often asymptomatic or have minimal clinical or lung imaging manifestations, [8] [9] [10] [11] [12] there is no doubt they get infected with this virus and can develop severe COVID-19 complications. [2] [3] [4] [5] [6] In a recent metaanalysis (n = 1026 children), we reported that COVID-19 lung disease is present in a significant portion of the pediatric population. 2 We found that 64% of lung CT images in PCR-confirmed pediatric COVID-19 cases had abnormalities, primarily characterized by focal ground-glass opacities (GGO) and consolidations. 2 In adult COVID-19 cases, CT scan-based algorithms have recently been developed for lung disease quantification and prediction of life-threatening complications. [13] [14] [15] [16] However, chest CT-based risk prediction approaches cannot be readily applied to infants and children due to concerns about sedation requirements, radiation exposure, and costs. 17, 18 An additional complicating factor for the clinical and lung imaging evaluation of pediatric COVID-19 cases is that viral lower respiratory tract infections (LRTI) in young children (often termed viral bronchiolitis) are the top cause of pediatric sick visits affecting more than 800,000 children each year in the United States or 20% of the annual birth cohort. 19 Thus, pediatric clinicians face an enormous challenge differentiating early stages of severe SARS-CoV-2 infection from thousands of common cases of viral LRTIs in infants and young children.

We have previously described predictive algorithms using chest X-ray (CXR) in children as a noninvasive approach of lung disease quantification in viral LRTIs. [19] [20] [21] [22] [23] However, the specific CXR abnormalities in pediatric COVID-19 and their relation to clinical outcomes remain to be defined. Addressing this gap is important given the age-related differences in the clinical and imaging features of COVID-19 2 and the challenge of differentiating SARS-CoV-2 infections from other types of viral LRTIs in young children. Accordingly, the goal of this study was to conduct an age-based comparison of CXR lung imaging features in pediatric cases of COVID-19 including infants, children, and adolescents. To define the imaging features that identify severe pediatric COVID-19 cases, we linked the presence of GGO/consolidations and other CXR features with clinical outcomes (e.g., hospitalization and critical care admission). We also performed a subanalysis in infants (0-2 years of age) focused on typical lung imaging findings of viral LRTI (e.g., hyperinflation or increased peribronchial markings/cuffing [19] [20] [21] [22] [23] [24] ) to examine whether these features could potentially help clinicians distinguish COVID-19 cases from common viral LRTIs caused by other respiratory viruses in infants and young children.

We conducted a single-center, cross-sectional study that included a retrospective collection of lung images and electronic health records 

CXR images were acquired in the posteroanterior or anteroposterior projection. Images were retrospectively reviewed by three fellowship-trained pediatric radiologists. All three radiologists assessed all CXRs independently and consensus was reached if disagreement. For scoring purposes, radiological features were assessed in terms of the type and the severity of abnormality. We included three main categories: GGOs/consolidations, hyperinflation, and increased peribronchial markings/cuffing using standard definitions for the radiographic findings of viral pneumonia. 25, 26 These features were scored as binary variables. To assess the severity of GGO/ consolidation, we quantified the number of lung zones compromised from 0 to 4 based on right/left and superior/inferior distribution as described by our team. 23 We also recorded additional features previously reported as rare in pediatric COVID-19, including airway bronchogram, pleural effusion, pleural thickening, bronchiectasis, and widening of the cardio-mediastinal contour. EHR of patients included were reviewed for the following demographic and clinical information: date of admission, age, sex, self-reported race/ethnicity, hospitalization, need for pediatric critical intensive care unit (PICU), need for supplemental oxygen, maximal temperature, wheezing, subcostal retractions, and the presence of multisystem inflammatory syndrome in children (MIS-C) according to published criteria and definition. 27 

Differences between groups on continuous variables were analyzed using the unpaired t test, the Mann-Whitney U test, or one-way analysis of variance for continuous variables, as appropriate. Associations between categorical variables were analyzed using the Fisher exact test or χ 2 test. Multivariate analysis (logistic regression) was used to examine the link between the number of lung areas affected and the respiratory outcomes adjusting by age, sex, and race. The data were analyzed with the Minitab Statistical Package V.19.

(Minitab, Inc.).

We screened a total of 422 patients who tested positive on PCR for COVID-19 during the study period at CNH. We enrolled all pediatric individuals (range: 0-19 years) with positive PCR test for COVID-19 and available CXR (n = 95) independently of clinical presentation or comorbidities. Given that our team and others have reported that pediatric COVID-19 radiological manifestations are affected by age, 2-6 we subdivided our study group according to age groups including infants (0-2 years, n = 27), children (3-10 years, n = 27), and adolescents (11-19 years, n = 41).

The mean age of the enrolled individuals was 9 years, 52% were males, and most were Hispanics or Black/African American (61% and 30%, respectively) ( 

A total of 49 (52%) of pediatric patients with COVID-19 had abnormalities observed in CXR ( Table 2) . Examples of pulmonary lesions in pediatric COVID-19 cases are shown in Figure 1 . The most common pulmonary abnormality identified was the presence of GGO/ consolidations (35% of all study subjects, Table 2 ). The severity of GGO/consolidations was influenced by age as only one infant (3.7%) had GGO/consolidations in multiple lung zones (Table 2) . Increased peribronchial markings/cuffing was also common (34% of all study subjects, Table 2 ). This finding was more common in young and school-age children compared with other age groups ( Table 2 ). All other radiological manifestations were rare in pediatric COVID-19, and we did not identify cases of pleural thickening, bronchiectasis or widening of the cardio-mediastinal contour ( Table 2 ). 

We next compared the CXR features of SARS-CoV-2 in young children exposure and sedation. 17, 18 One of the most clinically used diagnostic modalities to evaluate COVID-19 in the pediatric population is CXR. [26] [27] [28] In this study, we found that more than half of all pediatric COVID-19 patients had positive CXR results ( which CXRs were primarily characterized by perihilar bronchial wall thickening (58%) and/or airspace consolidation (35%). 30 In our pediatric study (n = 95), we also found that GGO/consolidations and peribronchial markings were the most common CXR findings in pediatric COVID-19 (Table 2 ). Furthermore, we identified age-related differences in lung imaging ( Adults and older children with COVID-19 appear to have alveolar involvement leading to a GGO pattern without bronchial lumen obstruction or air trapping. 2 In contrast, common viral respiratory pathogens in children, such as RSV, 18 are known to cause airway mucosal edema, mucosal plugging, bronchoconstriction, and bronchial lumen obstruction leading to increased perihilar markings and hyperinflation. 18 The latter respiratory syndrome is often referred to as "viral bronchiolitis" and is a leading cause of morbidity and mortality in infants and young children worldwide. 18 The definition of specific characteristics of CXR has become of vital importance in the pediatric population since the symptoms of COVID-19 in children can be confused with viral bronchiolitis. In this study, we found that, unlike viral bronchiolitis, COVID-19 rarely causes lung hyperinflation regardless of the age-group. Furthermore, we found that in young children, pulmonary hyperinflation is much more common in individuals with viral LRTI caused by RSV or other viruses than in cases of SARS-CoV-2 ( Figure 2 ). Although increased peribronchial markings are common in pediatric COVID-19, this feature also appears more frequently in viral bronchiolitis ( Figure 2 ). Taken together, our results indicate that the detection of typical lung imaging patterns of viral bronchiolitis (e.g., hyperinflation) could potentially be used to complement clinical evaluations in pediatric COVID-19 cases.

In studies of adults with COVID-19, lung imaging quantification has been successfully implemented to predict adverse outcomes and severe complications. [13] [14] [15] [16] We previously described CXR-based methods to quantify lung disease severity in pediatric viral

LRTIs. [19] [20] [21] [22] [23] In this study, we examined whether specific CXR abnormalities in pediatric COVID19 are associated with clinical severity.

We found that the quantification of GGO/consolidations was predictive of the need for supplemental oxygen during acute infection, the need for hospitalization, and the probability of PICU admission.

These data support the notion that CXR analysis can potentially be cases. New machine learning technology for the analysis of CXR [32] [33] [34] and lung ultrasound, which has become increasingly available to perform bedside monitoring of without radiological risk, 35 can potentially be implemented in children to enable an objective and informed decision on the severity of lung disease and the risk of complications from pediatric COVID-19, resulting in better outcomes and potentially life-saving benefits.

The study was funded by NIH (Grant HL145669 and HL141237 ).

WHO Coronavirus Disease (COVID-19) Dashboard | WHO Coronavirus Disease (COVID-19) Dashboard n.d

Pediatric lung imaging features of Covid-19: a systematic review and meta-analysis

Severe coronavirus disease-2019 in children and young adults in the Washington, DC, metropolitan region

Cardiac dysfunction and thrombocytopenia-associated multiple organ failure inflammation phenotype in a severe paediatric case of COVID-19

Racial and/or ethnic and socioeconomic disparities of SARS-CoV-2 infection among children

COVID-19 Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in children and adolescents: a systematic review of critically unwell children and the association with underlying comorbidities

National trends of cases of COVID-19 in children based on US State health department data

Screening and severity of coronavirus disease 2019 (COVID-19) in children in Madrid

COVID-19 virus and children: what do we know? Arch Pediatr

Why is COVID-19 so mild in children?

Chinese pediatric novel coronavirus study team. SARS-CoV-2 infection in children

Detection of Covid-19 in children in early

Artificial intelligence for the detection of COVID-19 pneumonia on chest CT using multinational datasets

Automated assessment of CO-RADS and chest CT severity scores in patients with suspected COVID-19 using artificial intelligence

Differentiating novel coronavirus pneumonia from general pneumonia based on machine learning

Prior-attention residual learning for more discriminative COVID-19 screening in CT images

The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk

The value of chest CT as a COVID-19 screening tool in children

Severity quantification of pediatric viral respiratory illnesses in chest X-ray images

A generic approach to lung field segmentation from chest radiographs using deep space and shape learning

Marginal shape deep learning: applications to pediatric lung field segmentation

Automatic tissue characterization of air trapping in chest radiographs using deep neural networks

Phenotypical sub-setting of the first episode of severe viral respiratory infection based on clinical assessment and underlying airway disease: a pilot study

Viral bronchiolitis in children

Radiographic and CT features of viral pneumonia

RadioGraphics update: radiographic and CT features of viral pneumonia

Overcoming COVID-19 investigators; CDC COVID-19 response team. Multisystem inflammatory syndrome in U.S. children and adolescents

COVID-19 in pediatric patients: a case series from the Bronx

Thoracic imaging of coronavirus disease (COVID-19) in children: a series of 91 cases

Pediatric chest x-ray in covid-19 infection

Clinical characteristics of pediatric SARS-CoV-2 infection and coronavirus disease 2019 (COVID-19) in Kuwait

Predicting mechanical ventilation requirement and mortality in COVID-19 using radiomics and deep learning on chest radiographs: a multi-institutional study

Deep learning COVID-19 features on CXR using limited training data sets

COVID-Net: a tailored deep convolutional neural network design for detection of COVID-19 cases from chest X-ray images

Applicability of lung ultrasound in COVID-19 diagnosis and evaluation of the disease progression: a systematic review

Chest X-ray lung imaging features in pediatric COVID-19 and comparison with viral lower respiratory infections in young children

The authors declare that there are no conflict of interests. 

http://orcid.org/0000-0001-8621-6109