Frequency and characterization of ancillary chest CT findings in COVID-19 pneumonia

Objectives: Ground-glass opacity and consolidation are recognized typical features of Coronavirus disease-19 (COVID-19) pneumonia on Chest CT, yet ancillary findings have not been fully described. We aimed to describe ancillary findings of COVID-19 pneumonia on CT, to define their prevalence, and investigate their association with clinical data. Methods: We retrospectively reviewed our CT chest cases with coupled reverse transcriptase polymerase chain reaction (rt-PCR). Patients with negative rt-PCR or without admission chest CT were excluded. Ancillary findings included: vessel enlargement, subpleural curvilinear lines, dependent subpleural atelectasis, centrilobular solid nodules, pleural and/or pericardial effusions, enlarged mediastinal lymph nodes. Continuous data were expressed as median and 95% confidence interval (95% CI) and tested by Mann–Whitney U test. Results: Ancillary findings were represented by 106/252 (42.1%, 36.1 to 48.2) vessel enlargement, 50/252 (19.8%, 15.4 to 25.2) subpleural curvilinear lines, 26/252 (10.1%, 7.1 to 14.7) dependent subpleural atelectasis, 15/252 (5.9%, 3.6 to 9.6) pleural effusion, 15/252 (5.9%, 3.6 to 9.6) mediastinal lymph nodes enlargement, 13/252 (5.2%, 3 to 8.6) centrilobular solid nodules, and 6/252 (2.4%, 1.1 to 5.1) pericardial effusion. Air space disease was more extensive in patients with vessel enlargement or centrilobular solid nodules (p < 0.001). Vessel enlargement was associated with longer history of fever (p = 0.035) and lower admission oxygen saturation (p = 0.014); dependent subpleural atelectasis with lower oxygen saturation (p < 0.001) and higher respiratory rate (p < 0.001); mediastinal lymph nodes with shorter history of cough (p = 0.046); centrilobular solid nodules with lower prevalence of cough (p = 0.023), lower oxygen saturation (p < 0.001), and higher respiratory rate (p = 0.032), and pericardial effusion with shorter history of cough (p = 0.015). Ancillary findings associated with longer hospital stay were subpleural curvilinear lines (p = 0.02), whereas centrilobular solid nodules were associated with higher rate of intensive care unit admission (p = 0.01). Conclusion: Typical high-resolution CT findings of COVID-19 pneumonia are frequently associated with ancillary findings that variably associate with disease extent, clinical parameters, and disease severity. Advances in knowledge: Ancillary findings might reflect the broad range of heterogeneous mechanisms in severe acute respiratory syndrome from viral pneumonia, and potentially help disease phenotyping.


INTRODUCTION
The diagnosis of severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2) infection is confirmed by viral nucleic acid detection. 1,2 The infection from SARS-CoV-2 with clinical impairment is known as coronavirus disease 2019  and this is frequently represented by pulmonary damage up to acute respiratory distress syndrome (ARDS), which is termed COVID-19 pneumonia and recognized as a major cause of mortality. 3 The role of radiology in managing COVID-19 has been evaluated since the earliest Chinese outbreak of SARS-CoV-2 and recommendations were thereafter issued by International scientific societies. [4][5][6][7] The Fleischner Society recommended the use of chest imaging in patients with confirmed COVID-19 infection and worsening respiratory status, and in those with suspected infection presenting with moderate-to-severe clinical features in resource-constrained environments. They also highlighted Objectives: Ground-glass opacity and consolidation are recognized typical features of Coronavirus disease-19 (COVID-19) pneumonia on Chest CT, yet ancillary findings have not been fully described. We aimed to describe ancillary findings of COVID-19 pneumonia on CT, to define their prevalence, and investigate their association with clinical data. Methods: We retrospectively reviewed our CT chest cases with coupled reverse transcriptase polymerase chain reaction (rt-PCR). Patients with negative rt-PCR or without admission chest CT were excluded. Ancillary findings included: vessel enlargement, subpleural curvilinear lines, dependent subpleural atelectasis, centrilobular solid nodules, pleural and/or pericardial effusions, enlarged mediastinal lymph nodes. Continuous data were expressed as median and 95% confidence interval (95% CI) and tested by Mann-Whitney U test. Results: Ancillary findings were represented by 106 15/252 (5.9%, 3.6 to 9.6) pleural effusion, 15/252 (5.9%, 3.6 to 9.6) mediastinal lymph nodes enlargement, 13/252 (5.2%, 3 to 8.6) centrilobular solid nodules, and 6/252 (2.4%, 1.1 to 5.1) pericardial effusion. Air space disease was more extensive in patients with vessel enlargement or centrilobular solid nodules (p < 0.001). Vessel enlargement was associated with longer history of fever (p = 0.035) and lower admission oxygen saturation (p = 0.014); dependent subpleural atelectasis with lower oxygen saturation (p < 0.001) and higher respiratory rate (p < 0.001); mediastinal lymph nodes with shorter history of cough (p = 0.046); centrilobular solid nodules with lower prevalence of cough (p = 0.023), lower oxygen saturation (p < 0.001), and higher respiratory rate (p = 0.032), and pericardial effusion with shorter history of cough (p = 0.015). Ancillary findings associated with longer hospital stay were subpleural curvilinear lines (p = 0.02), whereas centrilobular solid nodules were associated with higher rate of intensive care unit admission (p = 0.01). Conclusion: Typical high-resolution CT findings of COVID-19 pneumonia are frequently associated with ancillary findings that variably associate with disease extent, clinical parameters, and disease severity. Advances in knowledge: Ancillary findings might reflect the broad range of heterogeneous mechanisms in severe acute respiratory syndrome from viral pneumonia, and potentially help disease phenotyping.
the higher sensitivity of chest CT over chest radiography for early parenchymal lung abnormalities, disease progression, and alternative diagnoses. 3 High-resolution CT (HRCT) has been largely performed to assist the clinical evaluation of patients with suspected COVID-19 pneumonia. [6][7][8][9] This widespread use of HRCT led to definition of typical HRCT findings of COVID-19 pneumonia, which are described as patchy peripheral groundglass opacity (GGO) with or without consolidation. 2,10-13 Moreover, a number of less frequent findings such as centrilobular solid nodules, intrapulmonary vessels enlargement, subpleural curvilinear lines, pleural or pericardial effusion have been reported in some series, however without full characterization. 14 A detailed characterization of these findings might contribute in integrated clinicoradiological stratification of disease severity.
The purpose of this study was to describe ancillary findings on HRCT of patients with laboratory confirmed COVID-19 pneumonia, to define their prevalence, and to test their association with clinical data and in-hospital clinical deterioration.

Ethics statement
This study was approved by the local Institutional Review Board and informed consent was waived.
Study population 378 consecutive patients referred to the respiratory triage between February 29 and March 11, 2020. Patients were screened for symptoms: temperature >37.5°C, oxygen saturation <95%, respiratory rate >25 breaths per minute, and/or history of recent cough. 15 Provided clinical evaluation of respiratory status, patients with moderate to severe pulmonary involvement underwent HRCT scan. Subjects with at least one positive rt-PCR and chest HRCT on admission were included in this study. Patients with negative rt-PCR or without admission HRCT were excluded ( Figure 1). Clinical data were obtained from medical records.

Imaging interpretation
A chest radiologist with 17 years of experience (NS) in chest imaging retrospectively reviewed the HRCT scans on the local Picture Archiving and Communication System (PACS) workstation (suite Estensa, Esaote, Genova, Italy). The study reader was aware that the study population included only patients with confirmed diagnosis of SARS-CoV-2 infection.
Typical findings: the extent of combined GGO and consolidation was visually scored at the nearest 5% on the whole lungs. The distribution was described as follows: (a) axial distribution: predominantly peripheral (within the outer third of the lung), predominantly central, or mixed; (b) craniocaudal distribution: predominantly upper (above the carina), middle (between the carina and the right inferior pulmonary vein) or lower (below the right inferior pulmonary vein) 16 ; (c) bilateral or unilateral involvement; (d) lobar involvement was accounted over six lobes (lingula was considered as a single lobe). Description of the pattern was also tabulated into the categories of our local COVID-19 protocol. 15 These categories aimed to define disease severity by encompassing both morphology and extent of parenchymal findings, as follows: (1) non-COVID-19 findings, (2) findings indeterminate for COVID-19, either because of differential or overlapping disease, (3) typical pattern of COVID-19, including different combinations of GGO and consolidations and their overall extent (Table 1 and Supplementary Figure  1). 15 In particular, category 2 included any HRCT with findings that did not suggest exclusive alternative diagnosis nor isolated COVID-19 pneumonia, and it was therefore meant either as "differential between diseases" or "overlapping diseases". For instance, category 2 as "differential between diseases" (COVID-19 or other disease) was assigned when GGO (with or without consolidation) were seen with a pattern compatible Figure 1. Flow-chart of patient selection according to rt-PCR availability and positive outcome. *rt-PCR outcome was collected on first swab for the present study. HRCT, highresolution CT; rt-PCR, reverse transcription polymerase chain reaction.
BJR COVID-19 pneumonia phenotype on CT with other diseases (e.g. pulmonary edema). Otherwise, category 2 as "overlapping diseases" (COVID-19 and other disease) was assigned when HRCT findings typical for other diseases (e.g. fibrotic lung disease, oncologic disease, bacterial infection) were seen in association with typical findings of COVID-19 pneumonia. Category 3 was defined by the appearance of HRCT pattern typical for COVID-19 (GGO with or without consolidation, with multifocal-multilobar distribution and peripheral predominance), and it was further described across a range of disease severity that included both morphology (exudative vs organized morphology) and extent by visual score.
The retrospective expert reading was compared with the prospective clinical reading from clinical practice (general radiologists with experience range 3-30 years), the interobserver agreement was tested for category and extent of disease.
Differential between "moderate" and "severe" is entirely subjective and will not impact on the decision about hospitalization. b This category was chosen in the presence of conspicuous organized consolidation, despite predominant pattern was still ground-glass.

BJR
Silva et al Enlarged vessel within the areas of GGO was qualitatively assessed on the axial plane, by comparing the "enlarged vessel" with homologous contralateral vascular structures passing through normal parenchyma ( Figure 2). When the contralateral area of lung parenchyma was also affected, the comparison was made with vascular structures at the same distance from the pleura surface passing through normal parenchyma, in the same slice. The "enlarged vessel" sign was classified as appearing in dependent and non-dependent regions of the lung or both. Pulmonary vessels were differentiated into pulmonary arteries or veins by tracking the vessel centrally to the mediastinum.
Subpleural curvilinear lines were defined as thin curvilinear opacity, 1-3 mm in thickness, lying less than 1 cm from and parallel to the pleural surface.
Dependent subpleural atelectasis was described as consolidation with dependent crescent shape ( Figure 3).
Pleural effusion was defined by pleural fluid thickness ≥5 mm measured on the axial plane with the mediastinal window at the level of maximum thickness of the pleural fluid; the side of effusion was also recorded.
Mediastinal lymph nodes with a short-axis diameter >10 mm were deemed significantly enlarged. The number of enlarged mediastinal lymph nodes was recorded along with the node station.
Centrilobular solid nodules were defined by size <10 mm (by measuring the largest recognizable nodule in the lungs for each patient) ( Figure 4) and shape (free text was allowed for the description). The adjacency to vessels was also recorded.
Pericardial effusion was measured at the maximal circumferential extent of the pericardium at four points around the circumference: anterior, posterior, left lateral, and right lateral pericardium. 17 The sum of these four measurements was then calculated.
All terms were defined in accordance with Fleischner Society glossary. 18 Statistical analysis Continuous data were expressed as median and its 95% confidence interval (95% CI) and tested by Mann-Whitney U test. Categorical data were expressed as absolute and relative distribution, with corresponding 95% CI using Wilson method, and were tested using the Fisher exact test. Interobserver agreement was tested by Cohen's K test with quadratic weights (k w ) and its 95% CI. A p-value < 0.05 was deemed statistically significant.

Ancillary findings and clinical data
Ancillary findings and symptoms are detailed in Table 4. Vessel enlargement was associated with longer history of fever (p = 0.035) and lower admission oxygen saturation (p = 0.014).
We did not observe association between ancillary findings and comorbidities (Table 5).
Ancillary findings associated with in-hospital disease severity were subpleural curvilinear lines (longer hospital stay, p = 0.02) and centrilobular solid nodules (higher rate of ICU admission, p = 0.01) ( Table 6).

DISCUSSION
We report ancillary findings in 42.1% of patients with COVID-19 pneumonia. Vessel enlargement and centrilobular solid nodules were found in patients with more extensive pulmonary involvement. The presence of vessel enlargement, centrilobular solid nodules and dependent subpleural atelectasis was associated with a worse clinical status on admission. Moreover, centrilobular solid nodules were associated with a higher rate of ICU admission.
Recognized typical findings of COVID-19 pneumonia are GGO and consolidation. 10,12,18,20 Such typical presentation can be variably associated with less common findings, which might be termed "ancillary findings". Some ancillary findings were previously mentioned in the literature, while others are still underdescribed.
We proposed a methodical approach to define enlarged intrapulmonary vessels, namely by side-by-side-comparison ( Figure 2). Enlarged intrapulmonary vessels were previously described in 45.2-82.4% of patients with COVID-19 pneumonia. 6,10,[20][21][22][23] Of note, the variability in reporting this sign might depend on its relatively subtle appearance, which is prone to interpretation. These HRCT features are observed both in dependent and non-dependent regions, which makes it different from the typically dependent vascular redistribution commonly observed in cardiac failure. 24 Because vessel enlargement was seen only in GGO areas, it could be hypothesized that it represents a local response to local damage. [24][25][26] Unfortunately, we did not have Consolidation reported in COVID-19 pneumonia included exudative morphology as well as signs of organized pneumonia (OP) with pulmonary distortion. 27 We are keen on describing a further type of consolidation in COVID-19 pneumonia, namely dependent subpleural atelectasis with specific morphology as opposed to clearly infectious pneumonia (Figure 3). Previous authors interpreted this finding in COVID-19 pneumonia as fibrosis or edema, without pathologic confirmation. 14 Dependent atelectasis is well known in ICU in patients with severe respiratory decay and need for advanced respiratory support. 28 On admission chest CT, we observed dependent subpleural atelectasis in association with lower oxygen saturation and increased respiratory rate, even in those patients with pneumonia extent ≤30%. This type of consolidation might represent dependent atelectasis deriving from damage alveolar epithelium and representing functional alveolar dead space, already described in ARDS. 29 Progressive recruitment of these alveolar units can be assisted by prone positioning, 30 which has been proved to be effective in COVID-19 patients, both conscious and treated with non-invasive ventilation, 31 and critically ill who required intubation and invasive ventilation. 32 Although it remains speculative, the specific finding of subpleural atelectasis on admission CT might contribute in treatment choice and serve as an indicator of a potentially worse clinical course.
Prior studies including more than 100 patients reported centrilobular solid nodules in 3% of patients with COVID-19 pneumonia, 31,33 whereas smaller series described a much higher prevalence (up to 27.3%) 9,22,32 ; however, there is insufficient description of specific features for differential with small airway disease. 32, 33 We described solid centrilobular nodules with polyhedral shape in 5.2% of patients with COVID-19 pneumonia (Figure 4). This morphology is different from the rounded or branching centrilobular nodules observed in small airway disease. 18 Such detailed characterization might help in the differential with other infections, however there are practical limits from subjective interpretation and from supervening bacterial co-infection 34 or other viral pneumonia. 35 34,36 We observed this finding always close to enlarged vessels within GGO, opening to speculation about associated vasculopathy. 37 In our series, this finding was associated with disease severity at admission and during hospital stay. Aware of the limited representation of this finding in our population, we believe that its detection ought to prevent any delay in medical treatment. We report subpleural curvilinear lines in 19.8% of patients and observed a positive association between their presence and longer hospitalization. Similar findings are typically described in subjects with exposure to asbestos. 38 In up to 50% of cases, it was found to be associated with areas of consolidations, which had been correlated with more severe outcomes in ARDS patients. 39 Finally, only a minority of patients presented with mediastinal lymph node enlargement, pleural or pericardial effusion, in line with prior investigations. 7,10,21,37-41 These relatively non-specific findings within a complex pathology (e.g. involvement of both pulmonary and cardiac failure) hampers the possibility to draw hypothesis on such a small sample size.
Our study has several limitations. First, the retrospective design, is prone to confounding factors such as selection of patients; however, we tried to limit this bias by selecting consecutive rt-PCR positive subjects. Second, survival data were not available nor definitely certified for a substantial proportion of this population, and thus not included in the analysis, limiting our possibilities to investigate their prognostic relevance. Third, clinical information was not available for all patients enrolled, affecting the significance of the attempted correlation between ancillary findings and clinical picture. Further investigation of these findings is warranted by comprehensive inclusion of clinical parameters as well as with comparison against subjects with negative rt-PCR. Fourth, the presence of a single reader limited the objectiveness of interpretation of ancillary findings and does not warrant on the repeatability of these data.
In conclusion, typical HRCT findings of COVID-19 pneumonia are quite frequently associated with ancillary findings that variably associate with disease extent, clinical parameters, and disease severity. Correlation with survival data and follow-up CTs are fostered for in-depth understanding of ancillary findings and their clinical relevance.