What is the Role of Chest CT Scan for Pediatric Patient Diagnosis of COVID-19?

Objectives To retrospectively analyze the CT imaging features and patterns at baseline and as they evolve with time as the disease progresses or resolves in a cohort of pediatric patients affected with 2019 corona virus disease (COVID-19) pneumonia from three different cities in China. Methods We evaluated 29 pediatric patients with COVID-19 in the authors’ institution from Jan 1 to Feb 20, 2020. Cases were conrmed by laboratory test and were analyzed for epidemiological, demographic, clinical, radiological features and patterns. Results 29 initial scans and 23 follow-up scans were obtained from 29 patients. 15(52%) patients had been to Hubei Province and 26 (90%) of them had close contact with the COVID-19 positive patients in 14 days prior to the illness onset. The peak severity time was 5-8 days after symptom onset. A signicant difference between the number of involved segments at different time points was indicated (p=0.019). Half (52%) of the laboratory conrmed patients had no CT positive ndings. Nine (31%) of the laboratory conrmed patients had no symptoms. Six (21%) had no CT positive ndings nor symptoms. All the patients of one center(n=6) whose fecal samples remained positive after the respiratory samples became negative. Conclusion The common positive CT ndings included ground-glass opacities (50%), ground-glass opacities mixed with consolidation (36%), peribronchial thickening (21%), and consolidations (14%). We recommend for pediatric patients CT should not be used as a rst-line test to diagnose COVID-19.

testing by real-time reverse-transcriptase-polymerase-chain-reaction (rRT-PCR) with specimens obtained from the respiratory tract for assay as de ned by the World Health Organization (WHO) 3 . In children with COVID-19 the symptoms are usually less severe than adults and present mainly with cough and fever [4][5] .
According to the report by the Chinese Center for Disease Control and Prevention (CDC) in the con rmed cohort of 44,672 patients, 965 patients (2.2%) aged 0-19 years of age had case fatality rate of 0.1% (one case) 4 while aged ≥ 20 years had case fatality rate of 2.3% 4 . Most of the con rmed cases had CT scan in China. Considering the in uence of the dose in uence of X-ray radiation exposure get from CT scan, the role of CT scan for the pediatric patients diagnosis of COVID-19 needs to explore. We retrospectively evaluated a cohort of children under 18 years old with laboratory con rmation of COVID- 19

CT Scan and Image Analysis
Patients were scanned using High resolution CT technique with X-ray tube voltage of 100-120 kV and tube current of 210-250 mA. The images were reconstructed using high spatial resolution algorithm and at 1 mm or 2 mm slice thickness. The entire thorax was scanned in supine position during full suspended inspiration. Scans were interpreted on the dedicated PACS workstations using lung windows (window width, 1200 ~ 1500 H; level, -700~-1000 H) and mediastinal windows (window width, 400 H; level,40 H).
High-resolution CT (HRCT) scans were assessed for the presence and zonal distribution of ground-glass opaci cations, consolidation, interlobular and intralobular interstitial septal thickening and bronchiectasis. The CT ndings in the outer one third of the lung were de ned as peripheral, and ndings in the inner two thirds of the lung were de ned as central. Consolidation was de ned as an area of opaci cation that obscured the underlying vessels. Ground glass opacity (GGO) was de ned as a hazy increase in pulmonary parenchymal attenuation without obscuring the underlying vessels. The presence of additional ndings such as pleural effusions, lung brosis and emphysema were also evaluated.
The CT scans were reviewed by two radiologists with professional training in thoracic radiology who evaluated the pattern and distribution of the ndings by consensus. The images were also evaluated for response to therapy on follow-up scans.

Patients Characteristics
A total of 29 pediatric patients with suspected SARS-CoV-2 (11 females and 18 males) were tested and con rmed with two sequential positive respiratory tract samples collected every 1-2 days. 8 patients with suspected CT image ndings but the PCR resulted negative for SARS CoV-RNA. The age range was from 2 months to 17 years (mean, 7.7 years; median, 7 years). 20 patients (69%) were 12 years old, 9 (31%) were 12-17 years old. Cough (52%) and Fever (45%) were the most frequent symptoms. All 6 patients'(from Fifth A liated Hospital of Sun Yat-sen University)fecal samples were positive for SARS CoV-RNA. The fecal samples remained positive in all of these patients after the respiratory samples became negative (6/6, 100%). The demographics and baseline clinical features are listed in Table 1. Nine (31%) of the laboratory con rmed patients had no symptoms. Six (21%) had no CT positive ndings nor symptoms. Lymphocyte count(× 10 9 /L) 3.07 White blood cell count(× 10 9 /L) 7.14 Neutrophil count(× 10 9 /L) 3.62 According to the WHO COVID-19 Interim Guidance 3 and the Sixth Edition Guideline of 2019-nCoV published by National Health Commission of China 6 , we divided the con rmed cases into 2 groups according to the clinical symptoms and imaging ndings: 15 patients in the 'imaging negative group'. In this group we included the patients whose CT imaging had no positive ndings but had laboratory con rmed positive for COVID-19. The remainder of 14 cases were categorized as mild group. There were no patients in the severe and critical groups.
There were 5 (17%) patients who had been to Hubei Province and 26 (90%) of them had close contact with the COVID-19 positive patients. All patients denied the history of visiting the Huanan Seafood Wholesale Market in Wuhan in January. Some patients had history of preexisting conditions such as asthma (n = 1), congenital heart disease(n = 1), hernia after surgery (n = 2). There were no deaths among these patients.

CT imaging Features
All of the 29 patients had the initial scan at admission as the baseline. The average days from illness onset till the initial scan were 3.2 days (range: 0-12 days).
The initial high-resolution CT scan was performed 1.8 ± 1.3 days after admission to the hospital.
In the Mild Group (n = 14),the lesions were predominantly peripheral in 71% of patients and both peripheral and central in 29%. 92% of patients had lower lobe involvement while only 1 patient had single upper lobe lesion.
The details of the CT ndings are summarized in Table 2. Ground-glass opacity (GGO) (n = 7) was the most common CT nding in the initial scan after admission followed by GGO mixed with consolidation (n = 5), peribronchial thickening (n = 3), consolidation (n = 2), air bronchogram (n = 1), interlobular and intralobular septal thickening (n = 1) (Fig. 1). 15 patients were categorized as imaging negative with the CT scan showing no pulmonary parenchymal abnormalities. We classi ed the CT imaging patterns of the COVID-19 infection as follows.
Type A. Sub-pleural patchy consolidation/consolidation mixed with GGO/only GGO (9, 31%) in single lobe or both lungs (Fig. 2) Type B. Peribronchial wall thickening with or without GGO (3, %) ( Fig. 3) Type C. Large area of consolidation with or without GGO (2, %) ( Fig. 4) The CT pattern of the patients between 12-18 years of age are belonged to the type A (n = 5, 45%), 1 belonged to the type C (9%), and 45% belonged to the imaging negative group. Among the children between 0 to 12 years old (n = 20), 3 belonged to the type B, 6 belonged to the type A, 1 belonged to the type C, 10 belonged to the imaging negative group. Comparing with Teenage patients, younger children presented with larger variation of image patterns.
Follow-up second and third CT scans were obtained after admission depending on the clinical condition of the patients. 18 patients had a second scan, and 5 patients had a third scan. On the follow-up CT scans, the patchy GGO or GGO mixed with patchy consolidation or centrilobular nodules with bronchial wall thickening was mostly resolved or completely resolved on the second scan; while the Type C consolidation is relatively slower be resolved such as: an 14-year old female which had a 16*15 mm subpleural consolidation in the left lower lobe. 16 days later the follow-up CT scan showed the lesion partly resolved and turned to be GGO.

Time Course of CT image Changes
From symptom onset the number of involved segments of bilateral lungs are illustrated in Fig. 5. The highest number of segments recorded was 6. The number of involved segments changed over time (Fig. 5, p = 0.019). Peak severity was reached at 5-8 days, at which the involved segments was ranged from 1-6, and the average was 2.8.

CT Correlation with RT-PCR
The RT-PCR results were used as golden standard. 15 patients had no CT image abnormalities. One patient was normal in the initial scan and had ground-glass opacity in the second scan. The baseline CT scan sensitivity was 45% (13/29).

Discussion
In this study we evaluated a cohort of 29 pediatric patients with laboratory con rmed COVID-19 infection. All cases had the history of travel to or residence in Wuhan or other cities of Hubei Province or close contact history with con rmed patients. The most common CT features are (1) Ground-glass opacity; (2) Ground-glass opacity mixed with consolidation; (3) Centrilobular nodules with bronchial wall thickening. And the most common CT pattern in the entire cohort is Type A-Sub-pleural patchy consolidation/consolidation mixed with GGO/only GGO. Peak severity which re ected by number of involved segments on CT was at 5-8 days after symptom onset. And the peak severity of adults was reported at 10-12 days on CXR and 6-11 days on CT [7][8][9][10] We noted the two distinct imaging patterns among the children we studied. Teenage(≥10 years)patients mainly presented with Type A patchy consolidation/consolidation mixed with GGO/GGO in single or both lungs or normal, whereas the younger children presented mainly with all the 3 types including Type B centrilobular nodules with bronchial wall thickening and Type C large area of consolidation. The clinical course of patients with Types A, B was much milder and shorter compared with the patients with C. The younger children with Type C pattern who had large area of consolidation, the clinical course was more severe and the lesions resolved more slowly. This is different from the previous reports about children with SARS. In that report from Hong Kong 11 44 younger children (≤ 12 years) had much milder and shorter clinical course compared to the teenagers (12-18ys). There are individual differences in the course of the disease. Looking at the temporal change in the imaging patterns from the time of admission till discharge, we recommend the follow-up CT scans to be obtained about 10-15 days after the initial scan in Types A and B patients. In Type C patients, we suggest the follow-up scan could be obtained at least 1 month later or even longer. Imaging negative group occupied about 52% of the included patients. These pediatric patients with normal CT ndings may do not need follow-up scan any more if the symptomatic treatment is effective.
The CT ndings of COVID-19 pneumonia are similar as RSV pneumonia, HPIV pneumonia, and HMPV pneumonia 12 . These viral infections usually appear as multifocal patchy consolidation with GGO. Centrilobular nodules with bronchial wall thickening are also noted. Large area of consolidation with or without GGO are uncommon but more extensive when present. Different from adult patients with COVID-19 infection, the interlobular and intralobular septal thickening (25-53%) [8][9][10]13 is less common in children (our study 8%) and the percentage of imaging negative patients is relatively much higher (our study 52% vs 19% 10 ).
Radiologists should familiarize themselves with the CT appearance of COVID-19 infection in order to be able to identify ndings consistent with infection in patients imaged for other reasons. The analysis of CT patterns/features could help the clinicians to evaluate the outcome of the infection and select the suitable follow-up scan time. In our laboratory con rmed cohort, 9 patients were asymptomatic. 16 of these patients had no positive CT imaging ndings. 6 patients were asymptomatic but had positive CT imaging ndings. A normal chest CT does not mean a person does not have COVID-19 infection -and an abnormal CT is not speci c for COVID-19 diagnosis 14 . We suggest laboratory con rmed cases should be isolated regardless of being asymptomatic. We also recommend for pediatric patients CT should not be used as a rst-line test to diagnose COVID-19 as the ACR Recommendations 14 .
Besides  [17][18][19] . In our study, we agree with obtaining posterior oropharyngeal saliva samples and stool samples for testing with real-time RT-PCR.
This is highly recommended after the clearance of viral RNA in a patient's respiratory samples, especially in the pediatric patients 15 . This may potentially help in controlling oral-fecal transmission in con ned premises such as hostels, dormitories, trains, buses, and cruise ships.
There are few limitations in our study. Firstly, the number of cases in this study is small. It is important to note that the entire range of imaging features may not have been fully demonstrated during the time course since the emergence of COVID-19 outbreak. Secondly, only in one medical center the fecal samples were tested for SARS-CoV-2 RNA. Future studies should involve a larger patient population. Finally, we do not have the histopathologic evidence to re ect the imaging changes throughout the entire course of infection.

Conclusion
In summary, the most common CT imaging features of COVID-19 infection in pediatric patients are predominantly: 1. Ground-glass opacity. 2. Ground-glass opacity mixed with consolidation. 3.
Peribronchial thickening. About half (52%) of the laboratory con rmed patients had no CT positive ndings. We recommend for pediatric patients CT should not be used as a rst-line test to diagnose COVID-19.

Declarations
Ethics approval and consent to participate This retrospective study was approved by the Institutional Review Boards of The Fifth A liated Hospital of Sun Yat-sen University. Written consent was waived Change in COVID-19 non-contrast CT involved segments with duration since symptom onset. The number of the involvement by consolidation or GGO ranged of 0-6 was assigned to bilateral lungs. Kruskal-Wallis test (p=0.019) indicated a signi cant difference between the number of involved segments at different time points.

Figure 2
Chest image of a COVID-19 pneumonia patient with Type C CT image pattern A 7-year-old female presented with a 1-day history of fever 37.7℃) and 3-day history of cough. She had the close contract history with laboratory con rmed patient. Non-contrast CT showed large area of consolidation in the right middle lobe and ground-glass opacity in the basal segment of right lower lobe.  Chest image of a COVID-19 pneumonia patient with Type A CT image pattern A 11-year-old male without any symptoms which had the travel history to Wuhan on Jan 20 was diagnosed with COVID-19 on Feb 14. The non-contrast CT images showed tiny subpleural patchy ground-glass opaci cation was located in the right upper lobe.