Characteristics of Chest CT Images in Patients With COVID-19 Pneumonia in London, UK

Background and objective Novel coronavirus 2019 (COVID-19) outbreak was first reported in Wuhan, Hubei Province in China in December 2019; it has then spread quickly and exponentially beyond the Chinese borders and is now regarded as a global pandemic. We aimed to evaluate the chest CT radiological characteristics and lesion distribution patterns in patients of COVID-19 pneumonia in London, UK. Methods We performed a retrospective study and reviewed data of patients with clinically suspected COVID-19 who underwent chest CT between February 1 and May 5, 2020. All patients underwent the reverse transcription-polymerase chain reaction (RT-PCR) test. Lung lesion characteristics and distribution patterns were evaluated by two radiologists. Fisher’s exact test was used for statistical analysis, and a p-value of <0.05 was considered statistically significant. Results A total of 18 patients (nine men and nine women) were analyzed. All of them had bilateral patchy lesions in the chest CT images. There was no correlation between the severity score and mortality (p=0.790). The distinctive CT features included ground-glass opacity (GGO) and consolidative patchy amorphous lesions, bilateral posterior and peripheral multi-lobar lung involvement, pleural effusions, subpleural fibrotic lines, subpleural sparing, vascular engorgement, occasional crazy paving, occasional mediastinal lymphadenopathy, pleural thickening, lack of cavitation, and absence of reverse halo (atoll) signs. Conclusion CT can facilitate the diagnosis of COVID-19 pneumonia. Our UK cohort showed slight variations compared with previously reported Asian and continental European cases with respect to chest CT images.


Introduction
Novel coronavirus disease 2019 (COVID-19) outbreak emerged in Wuhan in Hubei province, 1 2 3 which the text body contains the words "No COVID." Ultimately, a total of 18 patients (median age: 57 years) who were considered positive for COVID-19 viral pneumonia on CT imaging were selected for the study (Figure 1).

CT protocols
High-resolution CT (HRCT), CT of the chest, abdomen, and pelvis (CT-CAP), and/or CT pulmonary angiography (CTPA) were used for detecting pulmonary lesions. All images were obtained with a CT scan (GE Frontier; GE Healthcare, Chicago, IL) with patients in the supine position. Scans were performed with the following technical parameters: HRCT and CT-CAP: 120 kV, auto modulated mA, 0.625 slice thickness, 512×512 matrix; CTPA: 100 kV, auto modulated mA, 0.625 slice thickness, 512×512 matrix. Reconstructed images were also obtained and used for the current study.

Image analysis
Two board-certified radiologists (E.P. and U.M., with 25 and five years of experience, respectively) 2020  reviewed the chest CT images on a picture archiving and communication system (PACS, Carestream Health, Inc, Rochester, NY). Chest CT images were evaluated with both mediastinal (width: 350 HU, level: 40 HU) and lung (width: 1500 HU, level: -500 HU) window level settings. The two radiologists identified pulmonary lesions based on their density, shape, and margin. The locations of the lung lesions were recorded as lobar, axial, anterior, and posterior. Axial locations were categorized as central (inner two-thirds of the lung) or peripheral (outer one-third of the lung). Each lung was divided by an axial line into anterior and posterior halves (Figures 2, 7). The densities of the patchy-confluent lesions were classified as pure GGO, pure consolidation, or mixed. Nodular lesion densities were classified as pure GGO, solid, or partly solid. Margins were classified as well-defined or ill-defined. As most of the lesions were patchy, confluent, and illdefined margins, lesion sizes were not assessed. Instead, a visual severity score, which was slightly modified from Pan et al., ranging from 0 to 4 for each individual lobes was calculated. Each of the five lung lobes was visually scored on a scale of 0 to 4 as follows: 0: no involvement; 1: less than 25% involvement; 2: 25-50% involvement; 3: 50-75% involvement; and 4: 75-100% involvement.
The total CT score was the sum of all individual five lobar scores ranging from 0 (no involvement) to 20 (maximum involvement).

FIGURE 2: Division of lung fields in the axial plane
Dividing lung fields equally into two axial planes on CT: blue represents the anterior half; brown represents the posterior half CT: computed tomography

Statistical analysis
The distributions of the lesion pattern and severity score proportions were analyzed. Fisher's exact test (SPSS Statistics version 25; IBM, Armonk, NY) was applied to analyze the relationship between the severity score and death. A p-value of <0.05 was considered statistically significant.

Results
A total of 18 patients (median age: 57) who were considered positive for COVID-19 viral pneumonia on CT imaging were selected. Of them, four had negative RT-PCR results and 14 had positive RT-PCR results ( Figure 1). Patient demographics are summarized in Table 1. No gender predilection was encountered in our cohort (nine men and nine women). In a majority of cases (15 out of 18), symptoms were chest-related. The remaining cases presented with abdominal symptoms. One patient was diagnosed with small-bowel obstruction.

Chest image results
Of the 18 cases, 11 had positive X-ray examination results, four had equivocal bilateral ill-defined subtle X-ray changes, and one had negative X-ray results but with CT images consistent with COVID-19 pneumonia.
In the CT images, GGO plus consolidation was observed bilaterally in all 18 (100%) patients. No pure GGO or pure consolidation was encountered in our cohort. The predominant shape was amorphous (72%). Lesions showed no specific lobar predilection, but most lesions were localized posteriorly and peripherally.

Laboratory tests
Of the 18 patients with epidemiological, clinical, and radiological findings consistent with COVID-19 pneumonia, 14 had positive and four had negative RT-PCR test results. Film arrays were negative for the vast majority of cases, particularly for patients with negative RT-PCR test results, except for one case (film array was not performed for that particular patient). White blood cell counts were normal in 15 out of 18 patients and low in three out of 18 patients. Lymphocyte counts were low in 10 out of 18 cases. CRP levels were elevated in 17 of 18 cases. Of 18 patients, four had positive D-dimer test results, but no pulmonary embolus (PE) was encountered on CTPA.
The distinctive CT features in our cohort are summarized in Table 3, including GGO and consolidative patchy mostly amorphous (72%) lesions, bilateral posterior and peripheral multilobar lung involvement, pleural effusions, subpleural fibrotic lines, subpleural sparing, vascular engorgement, occasional crazy paving, occasional mediastinal lymphadenopathy, pleural thickening, lack of cavitation, and absence of reverse halo (atoll) signs.
Subpleural sparing, which was observed in this patient cohort, has never been reported before, whereas mediastinal lymphadenopathy has been reported occasionally in the literature in English [11]. Focal vascular engorgement, septal thickening, and a subpleural fibrotic line were seen frequently in our cohort, with frequencies of 83%, 72%, and 61%, respectively.
Patients' CT features are presented in detail in Figures 3,4,5,6,7,8. A majority of cases (56%) had moderate severity scores. There was no correlation between the severity score and mortality, p=0.790 ( Table 4).

Predominant density Number of cases Percentage
Ground-glass opacity 8 44%

FIGURE 4: Axial CT through lung bases of a male patient (lung parenchymal window level settings)
A 45-year-old male patient presented with cough, fever, and progressive shortness of breath while selfisolating. His respiratory function deteriorated and he was admitted to the hospital. A chest CT was performed, which showed evidence of COVID-19 and subsequent RT-PCR was positive. An axial contrast-enhanced CT image of the chest showed bilateral patchy ground-glass and consolidative changes with areas of confluence. There was a linear air bronchogram seen in the right upper lobe (black arrows). This was surrounded by peri-bronchial consolidation and ground-glass changes

FIGURE 8: A 67-year-old female patient with positive RT-PCR test for COVID-19
The patient's CT images showed a subpleural band, a common finding in our cohort of patients. An axial contrast-enhanced CT image demonstrated a thin curvilinear subpleural band seen posteriorly in both lower lobes (black arrows). In addition, there were patches of peripheral ground-glass opacification (red boxes) and thickened interlobular septa (white arrows) COVID-19: coronavirus disease 2019; CT: computed tomography; RT-PCR: reverse transcriptionpolymerase chain reaction

Discussion
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the pathogen for COVID-19, is an animal-to-human and human-to-human transmissible and potentially lethal virus that belongs to the coronavirus viridine family. Its clinic-radiological lung manifestations strongly differ between Chinese, Korean, and Italian populations. To the best of our knowledge, this is the first preliminary report describing the chest CT features of COVID-19 in a cohort from London, UK.
In our patients, the initial symptoms were similar to those described in the existing literature, including cough, fever, and shortness of breath. However, two out of 18 patients (11%) in our cohort presented with abdominal pain and one (6%) with vomiting. CRP levels were high in all cases, which is consistent with previous studies [12][13][14][15]. In the early stage of COVID-19 pneumonia, CRP levels could be positively correlated with lung lesion sizes. CRP levels could reflect disease severity and could be used as a key indicator of disease prognosis [15]. D-dimer levels were measured in five cases in our cohort [16]. Four out of five tests were positive. All five cases underwent CTPA with negative results for PE. PE rates have previously been reported to be around 20% [17][18], which contrasted with our findings with no PE encountered. Lymphopenia is one of the prominent features of COVID-19. Lymphocyte counts may be a useful biomarker in predicting the severity of the disease [19]. Lymphocyte counts were low (56%) or normal (44%) in all cases in our patient cohort.
Lung manifestations of COVID-19 are similar to those of SARS and MERS. Radiological studies have shown cardinal similarities, including bilateral patchy GGO and consolidations, subpleural involvement, and predominant posterior and peripheral lesion distributions [2,[4][5][6][8][9][10][11][12][13]20,21]. The CT findings of COVID-19 pneumonia in the UK have been generally consistent with those in the Far East [20,22]. However, the proportion of predominantly consolidative lesions in a Chinese cohort was found to be approximately 30-60% [22,23]. In Korean patients, no predominantly consolidative lesions were observed, which is consistent with our study findings [2]. In addition, the proportion of chest radiographic abnormalities was 60% in Chinese COVID-19 patients [24], but 33% in Korean patients [2]. Considering these radiologic observations and five deaths from COVID-19 in the UK, UK patients seem to experience a harsher disease course than Korean patients. We did not observe the reversed halo signs in our cohort, but a few cases with the reversed halo sign were observed in some other recent reports [2,[24][25][26]. Septal thickening, a subpleural fibrotic line, and focal vascular engorgements were seen frequently in our study.
This study was performed at one of the epicenters of the COVID-19 pandemic, in London, UK. Our hospital has received a high number of COVID-19 patients, but CT scanning has been reserved for equivocal chest X-Ray findings, excluding complications like PE and preoperative patients scanning to exclude COVID-19 before the RT-PCR test result. Therefore, the initial recommendation in our institution, as in several other UK hospitals, was to perform CT-CAP in all preoperative patients. However, this recommendation was changed in May 2020. Currently, only limited sections through both lung bases appear sufficient to detect COVID-19 lung pneumonia for patients who undergo semi-elective or urgent surgery [27].
According to a statement that was recently published by the British Society of Thoracic Imaging, the assumed role of CT in the diagnosis, triage, and prognosis of patients with COVID-19 infection continues to be refined, and is reserved for clinically and laboratory-suspected COVID-19 patients who are seriously ill with oxygen saturation of less than 94% or a National Early Warning Score (NEWS) equal to or greater than 3. Normal or equivocal chest X-ray and CT examinations can be carried out with or without contrast as in our study [28]. Furthermore, in line with the British Society of Thoracic Imaging guidelines, patients with persistent significant abnormalities on the second chest X-ray, abnormal pulmonary function tests, and/or significant unexplained breathlessness may require further investigations, which can include a pre-contrast high-resolution volumetric CT and CTPA to assess the presence of interstitial lung disease and PE, respectively [29].
There are several limitations to our current study. Firstly, it was a retrospective single-center study with a limited patient cohort without follow-up CT examinations. Secondly, CT examinations were carried out with and without contrast enhancement, but we did not face great difficulties to see the lung lesions, and we clearly observed vascular engorgement signs on post-IV contrastenhanced CT examinations. Thirdly, the RT-PCR test was only performed twice in our institute. And lastly, the time interval between initial symptom onset and CT was not precisely defined, but RT-PCR and chest X-ray were performed on the patient admission date, and subsequent CT scans were performed either on the same day or within a week's time, with or without RT-PCR test results.

Conclusions
Chest CT scanning was not a first-line investigation modality for suspected COVID-19 patients; all patients had initial chest X-ray at the time of admission along with an RT-PCR nasopharyngeal swab test. Mixed GGO-consolidation, amorphous morphology, and posterior and peripheral lung distributions of the COVID-19 pneumonia lesions were the cardinal findings in our limited but diverse UK patient cohort. Subpleural sparing, bilateral pleural effusion, presence of a halo sign, subpleural fibrotic lines, septal thickening, focal vascular engorgements, mediastinal lymphadenopathy, absence of a reverse halo sign, and absence of cavitation were observed in our study. Subpleural sparing, which has never been reported in the literature in English, was observed in our cohort. A majority of the UK cases presented with cardinal symptoms including fever, shortness of breath, and cough. All patients' CRP levels were elevated. Lymphopenia and normal lymphocyte counts were observed in our patient cohort. All D-dimer-positive cases did not show any PE, which has never been reported in the literature in English.
The findings from our limited cohort are generally in line with CT imaging findings from the published literature worldwide with a few exceptions, such as the presence of pleural effusion, mediastinal lymphadenopathy, and the lack of reverse halo sign.