¹Department of Radiology, Government Medical College Nagpur, India
²Department of Medicine, Indira Gandhi Government Medical College, Nagpur, India
³Department of Neurology, NIMHANS, Bangalore, India

^*Corresponding author: Dr. Maherafsha Hundekari, Junior Resident, Department of Radiology,Government Medical College Nagpur,Hanuman Nagar-440024, Nagpur, India, Tel: 7798100399 E-mail: maherafsha@gmail. com

Copyright: © 2021 Parate R, et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Article Information: Submission: 22/02/2021; Accepted: 27/05/2021; Published: 31/05/2021

Objectives: To elucidate the chest HRCT imaging manifestations in coronavirus (COVID-19) infected Indian patients, study and compare the patterns identified during time course and among differently aged patients.

Materials and Methods: A retrospective observational study of 500 laboratory-confirmed SARS-CoV-2 infected patients was performed. The patients were categorized into 3 groups based on the time interval between onset of symptom and HRCT imaging and according to their age. The distribution, patterns, and extent of lung abnormalities were recorded.

Results: A predominant peripheral distribution of the abnormalities was observed as in 262/390 patients (67.2%). The Mean CT score of the right lower lobe (2.06 ± 1.13) and left lower lobe (2.01 ± 1.128) were significantly greater than other lung lobes (P-value < 0.001). Overall, most commonly observed pattern on imaging was crazy-paving pattern observed in 329/390 patients (84.4%). In the early phase of the disease (0-7 days), predominant patterns were crazy-paving (90.7%) and GGO with consolidation (50.4%). The consolidation pattern and the vacuolar sign showed a significant rise during advanced phase of the disease (8-14 days of symptom onset) and eventually decreased during the absorption phase (> 14 days after symptom onset). The reticular pattern, subpleural line sign were the dominant patterns observed during the absorption phase. The Mean total CT severity score (9.79 ± 4.79) was greatest for group C patients: age ≥ 45 years. (P-value < 0.001). GGO with consolidations was more common in group B (25-44 years) and C, whereas GGO plus reticulations and reticular pattern were more common in group C (P-value < 0.001).

Conclusion: HRCT features of COVID-19 pneumonia vary according to the disease course and the patient’s age. Crazy-paving pattern is dominant during the early phase and repairing signs during the absorption phase. The extent and pattern of involvement are more severe in the elderly population.

COVID- 19; Coronavirus disease 2019; SARS- CoV-2; Severe acute respiratory syndrome coronavirus 2; CT, Computed Tomography; HRCT, High resolution computed tomography; RT-PCR, Reversetranscriptase-polymerase chain reaction

The outbreak of COVID- 19 originated in Wuhan, China, began in December 2019, and has been continuing since then. As of 21st November 2020, more than 56,982,476 confirmed SARS- CoV-2 cases along with 1,361,847 deaths have been reported globally by the World Health Organization (WHO) [1]. Amongst these, more than 9 million cases have been reported in India [2]. At present confirmation of COVID -19 infection is done by a specific viral nucleic acid assay using reverse transcription-polymerase chain reaction (RT-PCR) on sputum, nasopharyngeal/oropharyngeal swabs and other specimens [3]. However, as per the recent literature, some patients with COVID- 19 infections might show initial negative nucleic acid assay results [4]. The various reasons for a negative test could be insufficient cellular material for detection and improper extraction of nucleic acid from clinical materials [5]. Therefore Chest HRCT has higher sensitivity for the diagnosis of COVID- 19 pneumonia and presently HRCT has been used as one of the most important tools for a comprehensive evaluation and follow-up especially in epidemic areas [6].

In this study, we analyzed the HRCT lung manifestations in 500 infected patients and compared the HRCT patterns identified during the disease course and among differently aged patients.

This was a retrospective observational study approved by the ethical committee of our institute and the requirement of written informed consent was waived. It was conducted between August 15 to October 15, 2020, in our dedicated Covid hospital. Symptomatic COVID-19 infected patients were screened using the following criteria.

a. At least one positive RT-PCR for SARS- CoV-2 obtained with nasopharyngeal/oropharyngeal swabs.
b. At least one HRCT scan done per patient as and when requisition received from the clinicians.

a. Patients with RT-PCR test negative results.
b. Asymptomatic patients.
c. Technical errors within the HRCT scan.
d. Antenatal patients.

Thus, about 500 patients were included in the study. 308/500 (61.6%) were male; mean age - 43.25 years, range: 17-87 years and 192/500 (38.4%) were female; mean age - 40.76 years; range: 11-91 years. Based on age distribution: Group A: < 25 years, n= 46 (9.2%); Group B: 25-44 years, n= 236 (47.2%); Group C: ≥45 years, n= 218 (43.6%). The clinical, demographic and imaging data of all the patients were recorded.

The patients were clustered into 3 groups based on the number of days between the onset of the symptoms and the HRCT scan (course of the disease):

a. The early phase: 0- 7 days after the symptom onset.
b. The advanced phase: 8-14 days after the symptom onset.
c. The absorption phase: > 14 days after the symptom onset.

HRCT acquisition protocol: All HRCT scans were performed on a 128 slice SIEMENS CT scanner. Patients were examined in supine position and image acquired during a single inspiratory breath-hold. The scanning range extended from the apex of lung up to the adrenal glands. Following were the scanning parameters: X-ray Tube parameters- 140 KVp; 234mAs; rotation time - 0.5 s; pitch - 1.0; section thickness- 5mm; intersection space- 5 mm; additional reconstruction using the B80f ultra-sharp kernel and a slice thickness of 1 mm. Lung window setting was with a window level of -600 Hounsfield units (HU) and window width of 1500 HU. Appropriate infection and control measures were implemented.

HRCT image analysis: Two senior radiologists with 10-15 years of experience in thoracic radiology evaluated the scanned images on the console using multiplanar reconstruction tools. The patterns observed were classified into 3 major categories: lung, bronchial and pleural changes. These categories were further classified into the following subcategories:

a. GGO- Ground glass opacities: increased attenuation without obscuration of the underlying vessels.
b. Consolidation: Homogenous increased intensity of lung parenchyma with obscuration of underlying vessels.
c. Crazy-paving pattern: GGO with an interlobular and intralobular septal thickening.
d. Reticular pattern: a disordered arrangement of coarse linear or curvilinear opacities.
e. GGO with Consolidation.
f. GGO plus Reticulations.
g. Vacuolar sign: A transparent vacuole like a shadow of < 5 mm in length observed within the lesion.
h. Halo sign: area of consolidation surrounded by GGO.
i. Reverse halo sign: area of GGO surrounded by consolidation
j. Microvascular dilation sign: dilated small vessels within the lesion.
k. Subpleural line: an arc-shaped linear shadow 2-5 cm in length appearing parallel to the chest wall.
l. Subpleural transparent line: a thin transparent line lying between the lesions and the visceral pleura.

a. Air bronchogram: an air-filled image of bronchus in lung lesions.
b. Bronchiectasis.

a. Pleural thickening. b. Pleural retraction: lesions pulling off the visceral pleura.
c. Pleural effusion.
d. Miscellaneous changes: Mediastinal lymphadenopathy, cavitation, calcifications, tree-in-bud appearance (centrilobular nodules with a linear branching pattern).

The horizontal distribution of pulmonary lesion was noted as peripheral: involving mainly the peripheral one-third of the lung or central plus peripheral. The area of involvement was categorized as predominant anterior or posterior involvement (the area before or after the vertical line of the midpoint of the diaphragm in the sagittal position respectively). The number of lesions was noted as single or multiple (>1) lesions, the size of the lesion was accounted as: < 1 cm, 1-3 cm, and >3 cm.

A semi-quantitative CT severity scoring was used to evaluate the extent of lung involvement and was recorded for each of the 5 lobes based on anatomic involvement, as follows: 0, no involvement; 1, < 5% involvement; 2, 5-25% involvement; 3, 26-50% involvement; 4, 51-75% involvement; and 5, > 75% involvement. The total CT score was the sum of each lobar score: 0 to 25 [8].

Statistical analyses were performed using SPSS version 19.0. Continuous variables were expressed as Mean ± Standard Deviation, Numerical data were expressed as a percentage (%) of the total. The incidence of various HRCT characteristics, lesion distribution, predominant lung area involved was expressed as frequency (%) and compared using the Pearson Chi-Square test between the early, advanced and absorption stage as well as between the age groups. We used one-way ANOVA analysis of variance (post hoc multiple comparisons) to do pairwise comparisons for CT severity score as well as for the patterns found among the age groups. The difference was statistically significant with a P-value < 0.05.

Table 1 summarizes demographic and clinical data

HRCT findings: Out of 500 patients, parenchymal abnormalities were observed in 390 (78%) cases, while 110 (22%) cases had a normal scan. A significantly predominant peripheral distribution of the abnormalities was observed in 262/390 patients (67.2%), while central plus peripheral involvement was seen in 128/390 cases (32.8%) (P-value <0.001). A higher frequency of predominant posterior involvement (311/390) of the lung area compared to the predominant anterior involvement (79/390) was noted (P-value<0.001) (Fig 1). The mean CT score of the right lung was 4.74 ± 2.89, higher than the left lung (3.55 ± 2.05) (P-value<0.001). Among the lobes, the mean CT score of right lower lobe and left lower lobe were 2.06 ± 1.13 and 2.01 ± 1.128, significantly greater than the other lobes (P-value < 0.001), however, not much difference was noted between the scores of right lower lobe and left lower lobe (P-value=0.154) (Table 2).

Table 3 shows the detailed evaluation of various HRCT characteristics. The most common parenchymal abnormality was the crazy-paving pattern observed in 329/390 cases (84.4%). Other common patterns observed were pure GGO (60/390;15.4%); Consolidation (84/390;21.5%); GGO with consolidation (135/390;34.6%); GGO plus reticulations (176/390;45.1%); Reticular pattern (137/390;35.1%) and micro vascular dilation sign (264/390;67.7%) (Figure 2 and 3). Other patterns seen were Subpleural line sign (96/390;24.6%), Vacuolar sign (35/390;8.9%); Halo sign (54/390;13.8%) and Subpleural transparent line (57/390;14.6%). The patterns observed here are similar to those mentioned in previously conducted studies [9,10]. Some characteristics observed in our study were Reverse halo sign in 2/390 (<1%), cavitation in 4/390 (1%) and tree in bud appearance noted in 2/390 (<1%). Regarding bronchial changes, 64 patients (16.4%) showed air bronchogram sign and 37 patients (9.5%) showed bronchiectasis. The pleural changes noted were pleural thickening (78/390;20%), pleural retraction sign (60/390;15.3%) and pleural effusion (4/390;1%).

As per a preliminary COVID-9 pneumonia imaging diagnostic guide in China [11], we categorized the HRCT images into an early stage (≤ 7 days), advanced stage (8-14 days) and absorption stage (> 14 days) according to the disease course and compared each of the density characters among each phase (Table 4 and Figure 4). The predominant patterns in the early phase of the disease (≤ 7 days) were crazy-paving (187/206; 90.7%) and GGO with consolidation patterns (104/206; 50.4%) which were more common when compared to the absorption and advanced phases (P value<0.001). The consolidation pattern showed a significant increase during the advanced phase (57/109; 52.3%) as compared to the early phase (20/206;9.7%) (P-value < 0.001), furthermore, it decreased during the absorption phase (7/75;9.3%) (P-value < 0.001), however, no significant difference was seen between the early and absorption phase concerning this pattern (P-value=0.011). The vacuolar sign showed a significant rise during the advanced phase with a frequency of 27/82 (24.7%) when compared to the early and absorption phases (P-value<0.001). The GGO plus reticulations showed an increasing trend during the advanced and absorption phase as compared to early phase (P-value=<0.001 and 0.002 respectively). The reticular pattern, subpleural line sign and bronchiectasis showed increasing trends, highest frequency observed in the absorption phase (P-value<0.001). Not much difference was seen regarding the distribution of the following patterns: pure GGO (P=0.363), pleural thickening (P=0.615) and pleural retraction sign (P=0.742).

Furthermore, we compared the various parameters among the age groups as illustrated in Table 5. The positivity rate for abnormal scan for group B (165/236;69.9%) and C cases (201/218;92.2%) was significantly greater than the group A cases (24/46;52.1%) (P-value<0.001) (Figure 5). Overall, the mean CT score of each of the lung lobe was greatest in group C (P-value<0.001) (Figure 6). The total mean CT score was greater for group B and group C when compared to group A (F=22.550; P<0.001).

No significant difference was observed for the pure GGO and crazy-paving patterns among the age groups (P=0.483; P=0.126). The predominant patterns observed in group C were GGO plus reticulations and reticular pattern, greater than the group A and group B cases (P<0.001). GGO with consolidation and consolidation patterns were more common in group B and C than in group A (P<0.001), without much difference, observed between group B and C (P>0.005). The relative representation of the following patterns: subpleural line sign, subpleural transparent line and vacuolar sign were more common in groups C as compared to group A and B (P<0.001). Further, no significant difference was observed for microvascular dilation sign (P=0.585), air bronchogram sign (P=0.384), bronchiectasis (P=0.617), pleural thickening (P=0.192), and the pleural retraction sign (P=0.106) among the age groups. Table 6 compares various HRCT patterns among different age groups.

The COVID-19 pandemic has hit hard on the globe. Just as clinicians are evaluating more patients suspected of having the infection, radiologists are similarly interpreting more chest scans in those suspected of having COVID-19 pneumonia. It is also known that asymptomatic patients can have a positive chest HRCT with the converse also being true [12]. Moreover, considering the limited number of RT-PCR kits in some centres of the developing countries, with the added possibility of getting large false negative RT-PCR results, solely relying on the laboratory tests for confirmation of the diagnosis could not be helpful. According to A rapid advance guide - Use of chest imaging in COVID-19, published by World Health Organization [WHO] [13], Chest imaging could be included as a part of the diagnostic workup of patients with suspected/probable COVID-19 infection in settings where the laboratory testing is not available or results are delayed or are initially negative in the presence of symptoms attributable to COVID-19. Moreover, the sensitivity of HRCT imaging is more as compared to the chest radiography as evaluated by some studies. In another literature review, the value and roles of different imaging modalities for the diagnosis and management of COVID-19, Chest radiography and HRCT scan were considered as key diagnostic modalities in suspected cases. With chest radiography having limited sensitivity for COVID-19 in the early stage of the disease, HRCT scan could be more accurate and sensitive in identifying pneumonia, especially during the early stages [14,15].

We conducted this retrospective study to study the spectrum of imaging characteristics. A positive HRCT rate was observed in a high proportion, 390/500 (78%) of RT-PCR confirmed symptomatic patients.

In a study conducted by Fang. Y et.al describing the CT image visual quantitative evaluation and clinical classification of COVID-19, 71.8 % symptomatic confirmed cases had CT evidence of pneumonia Zhan. J et.al reviewed CT scans of 110 patients describing CT pattern of evolution of COVID-19 pneumonia and reported an overall rate of 8.1 % negative scans [16,17]. Similarly, Liu. X et al. found a pooled positive CT rate of 89.7% cases among 2378 COVID-19 cases in a meta-analysis which included a total of 13 studies [18]. Thus, a comparable positive rate in this study could reflect the possible similar course of disease in Indian population as well as the fact that this study included only symptomatic positive patients; also 43.6% of cases belong to the age group ≥ 45 years. Some studies conducted in past have shown that the high proportion of elder age group (≥ 45 years) contributes towards a high positive rate [19].

We found higher mean CT score for right lung (4.74±2.89) than the left lung (3.55±2.05), with right lower lobe having highest mean score than the other lobes, a finding consistent with studies conducted previously [20]. The basis for this is, the bronchus of right lower lobe is more straighter, steep and is in continuation with the trachea. About 67.7% cases had a distribution of the lesions in the peripheral two-thirds of the lung and in 79.7% cases; posterior area of lungs was involved. The predominant patterns of lung opacities were crazy paving pattern (84.4%), GGO with consolidation (34.6%) and GGO plus reticulations (45.1%), findings consistent with several studies as in a meta-analysis conducted by Bao et al. [18]. The microvascular dilation sign was accounted in 67.69 % cases. The various theories put forward regarding its etiology are disordered vasoregulation [21], pro-inflammatory cytokines induced vasodilation, alveolar-capillary microthrombi and infection-induced pulmonary vasculitis [22]. Lang et, al. even proposed that this disordered vasoregulation could be an early marker of ARDS onset even before the clinical symptoms and abnormal appearance on imaging [21].

When we compared the lung opacities among different phases, we observed that during the early phase, the dominant patterns were crazy paving and GGO with consolidation. In early stages, the virus is likely to attack peripheral vessels and bronchi to cause an increase in the intraductal pressure which results in exudation, manifested as subpleural pure GGO. Over time, crazy-paving sign is formed due to thickening of the interlobular septum and increased exudation of the alveolus. If the disease continues to progress, the thickened lobular septum limits the absorption of the alveolar exudation, resulting in the alveolar consolidation formation. We observed that the consolidation, GGO plus reticulations and vacuolar sign were the predominant patterns in the advanced phase. Thus, the advanced phase could be described as a phase manifesting the combination of aggravation of the ongoing abnormalities along with initiation of repair changes within the parenchyma. This inhomogeneous involvement is evident by the vacuolar sign which is predominant during this phase. The absorption phase showed a predominance of the reticular pattern, subpleural line sign and subpleural transparent line sign which could be regarded as the repairing signs [10,23]. All the observed lesions could be eventually seen absorbing into irregular linear streaky opacities.

Furthermore, we noticed diversified patterns among different age groups of patients. Younger population < 25 years of age had a lesser extent of involvement than those above 45 years of age. This was supported by the fact that the mean total CT score of group C cases was significantly greater than group A cases. The patterns more commonly seen in elderly population included the GGO plus reticulations, reticular pattern, bronchiectasis and subpleural line sign. Deterioration of the lung structure and function along with a weaker immune system of the geriatrics could be the responsible factor for these patterns. Pleural thickening was mainly caused by the direct spread of the lesion, lymphatic reflux, and pleural inflammation. This may indicate that elderly patients show a more severe inflammatory response. At the same time, in elderly patients, whole lung lobes were more likely to be involved. This may indicate that in the same disease course, disease progression in elderly patients was faster than that in young patients [24,25].

Thus, our results fairly corroborate the distribution and types of pulmonary opacities reported in COVID-19 pneumonia worldwide. Also, the study findings are likely to help in understanding and evaluating the imaging abnormalities of cases during time course as well as depending upon the age and pattern of the patients. This could be a valuable tool in determining the treatment strategies as well as predicting prognosis of the disease. Therefore, it is essential to make full use of chest HRCT to achieve a comprehensive diagnosis and guide treatment for patients with COVID-19 infection.

But at the same time, it is pertinent to restrict the use of CT during the pandemic. Owing to the greater radiation dose of CT and the increased frequency of scans done presently, there is a growing concern regarding its long-term effects on the human body such as cancer, sterility, hematologic depression, etc. [26], thus it is of great importance to limit the number of scans done per person. Studies have shown that after the initial CT which is usually done after 5 days of symptom onset, appropriate follow-up time for subsequent chest scans should be during second week (approximately 12-14 days after symptoms onset) to look for disease progression and during fourth fifth week to see for disease resolution [27]. Also, additional chest scans can be done when the patient shows acute exacerbation of the symptoms or deterioration and when the chest radiograph obtained is non-conclusive.

This study had many limitations. Firstly, only RT-PCR positive symptomatic cases were included in the study, this may have overruled the fact that many asymptomatic and RT-PCR negative yet infected cases were not evaluated which could have resulted in bias. Also, patterns like reticulations and bronchiectasis need further follow-up to note whether the fibrosis seen in COVID-19 is reversible or irreversible. Finally, no lung biopsies were performed to assess the correlation between radiological and histopathologic findings.

In summary, HRCT features of COVID-19 pneumonia vary according to the disease course and the patient’s age. Crazy-paving pattern is dominant during the early phase and repairing signs during the absorption phase. The extent and pattern of involvement are more severe among the elderly population.

We express our gratitude to the Department of Radiology, Government Medical College Nagpur, Dr Ashwini Lahoti, Dr Pooja Raut and Dr Shivani Muppidwar for their support in providing necessary assistance in the recording of clinical and demographic data, and Mr Raghavendra Rao C. for helping in statistical analyses.