CT findings of patients infected with SARS-CoV-2

Since December 8th, 2019, several pneumonia cases with unknown causes were reported in Wuhan, China. The disease has rapidly spread around the whole country and even the world [1‐3]. In February 2020, an increased number of patients infected with this severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were reported around the world. Most clinical studies focused on clinical and epidemiological features of the patients infected with SARS-CoV-2 [4, 5]. The most common onset symptoms were fever, cough, and myalgia or fatigue; less common symptoms were sputum production, headache, hemoptysis, and diarrhea [4]. However, there are only a few reports about computed tomography (CT) findings of the patients infected with SARS-CoV-2.

Coronaviruses may produce specific features in chest CT images, such as severe acute respiratory syndrome (SARS) and middle east respiratory syndrome (MERS) [6‐9]. The typical CT findings in patients infected by SARS were ground glass opacities [9]. The SARS lesions can progress rapidly to focal, multifocal, or diffuse consolidation [9]. The most common CT findings in patients infected with MERS-CoV was subpleural and basilar airspace changes [8]. In our hospital, a chest CT scan was widely used in the diagnosis and follow-up evaluation of patients infected with SARS-CoV-2. The purpose of this study was to describe the chest CT findings of laboratory-confirmed SARS-CoV-2 cases.

As shown in Table 1, there were a total of 67 COVID-19 patients (31 males and 36 female patients) with an age range of 5–72 years old (median age, 44 years old). Most of the patients infected with SARS-CoV-2 have lesions in the chest CT images (54/67). Common symptoms at the onset of illness were fever (58/67), cough (52/67), and sputum production (35/67). Less common symptoms were chest tightness (12/67), Sore throat (8/67), diarrhea (5/67), dizziness (3/67), shortness of breath (4/67), nausea and vomiting (2/67), myalgia or fatigue (3/67), and headache (2/67). For the job type, 59 patients were self-employed, four patients were farmers, and four patients were company employees. For the type of infection source, 34 patients lived in Wuhan, three patients lived in Hubei province, 27 patients had a history of contact with COVID-19 patients, two patients had a history of contact with suspected patients, and only one patient had no history of contact with COVID-19 patients or suspected patients. For the smoking status, two patients were former smokers, three patients were current smokers, and the remaining sixty-two patients were non-smokers. There are sixteen patients with one or more of the following comorbidities: hypertension (11/16), diabetes (7/16), cancer (1/16), respiratory disease (1/16), cardiovascular disease (1/16), and connective tissue disease (1/16). After the treatment, 44 patients were cured, and the remaining 23 patients were still hospitalized.

All the patients underwent CT 1–11 days after admission (median, 4 days). There were 54 of 67 patients with lesions, and the remaining 13 patients were normal in chest CT images. As shown in Table 2, fifty of the fifty-four patients had lesions in the peripheral regions (Fig. 1), and four patients had lesions in the central areas. Twelve patients had abnormalities in the middle-upper lobe, and the remaining forty-two patients had abnormalities in the middle-lower lobe. There were eight patients with a single lesion (Fig. 2(a)) and 46 patients with multiple lesions (Fig. 1). Forty-two of the fifty-four patients had ground-glass opacities (Figs. 1, 2, 3, 4), and twelve patients had isolated consolidation (Fig. 2(a)). Interlobular septal thickening was identified in eleven patients. Reversed halo sign was noted in nine patients. There were eighteen patients with air bronchogram. Three patients had bronchial wall thickening. Tree-in-bud pattern was identified in one patient (Fig. 5(a)). Air cavities were present in four patients (Fig. 3). Pleural thickening or pleural effusion was noted in fourteen patients. Only two patients had intrathoracic lymph node enlargement (Fig. 4 (d)).

This study shows that multiple peripheral lesions on CT are common in patients hospitalized with SARS-CoV-2 infection. Ground-glass opacities were more extensive than consolidation in most patients. The average time interval between the onset of symptoms and the CT examination was 4.6 days. It suggested that the CT findings in our study were typical for COVID-19 patients in the early stage. Given patients with a history of exposure and the imaging features mentioned above, SARS-CoV-2 infection would be highly suspected. Further tests will be necessary. As the exposure history became complicated and unclear over time, the chest CT scans would play an essential role in the diagnose of COVID-19. However, there were still 13 COVID-19 patients with negative chest CT findings. As far as we know, both CT and nucleic acid detection are indispensable in the diagnose.

In our study, four patients had small cavities in the chest CT (Fig. 3), and their clinical symptoms were relatively acute. The short-term progress of chest CT imaging was somewhat evident (Fig. 3). Our study reported that a tree-bud pattern was identified in one case before the treatment and fully recovered after the treatment. Thus, the tree-bud pattern was not considered to be a typical CT feature caused bySARS-CoV-2 due to the short duration of the COVID-19 presence. Distinguished from previous reports of viral pneumonia, air cavity, bronchial wall thickening, and pleural effusion were found in the patients infected with SARS-CoV-2 (Table 2). In fifty-four positive CT finding cases, there are 5 cases with a single lesion, and the maximum of major axis length is 23 mm. To our knowledge, a single lesion located in the middle of the lung with the major axis length over 30 mm is possibly not a lesion of COVID-19.

After the treatment, thirty-nine of 67 patients infected with SARS-CoV-2 were cured and discharged. Repeated nucleic acid tests were performed to show SARS-CoV-2 clearance before hospital discharge. The chest CT scans before hospital discharge were also partially or fully recovered compared to the first ones after hospitalization. The partial recovery in lung lesions could be caused by the clinical lag of lung imaging. We also noticed that there were 13 cases with the progress of CT findings during the treatment, and ten of the 13 cases recovered in CT features after 10-day treatment.

There are two limitations in our study. First, the number of our patient cohort was small. Only a small amount of severe patients were presented. Recently, a large number of patients were confirmed to be infected with SARS-CoV-2. We will collect the clinical and CT features of a large patient population in future studies. Second, we did not perform a sequential CT study due to the urgency of sharing our CT findings of this new disease. In the next step, we will further evaluate the longitudinal change of CT features during the treatment. A predictive model of the clinical outcome will be developed based on CT imaging.

In summary, CT features of patients with SARS-CoV-2 were predominantly subpleural multiple ground glass opacities in two lungs, some with consolidations. There are also less common CT features, such as air bronchogram and reversed halo sign. It is rare to observe some features, such as bronchial wall thickening, tree-in-bud pattern, and pleural effusion. The CT features can play an important role in the early diagnosis and follow-up of COVID-19 patients.