Advertisement

Chest high-resolution computed tomography is associated to short-time progression to severe disease in patients with COVID-19 pneumonia

      Highlights

      • Chest computed tomography predicts progression to severe COVID-19 pneumonia
      • Chest computed tomography predicts clinical evolution in patients at lower risk
      • Extensive lung involvement and organizing pneumonia pattern predict severe disease

      Abstract

      Objective

      In patients with mild COVID-19 pneumonia, chest high-resolution computed tomography (HRCT) is advised when risk factors for severe disease (i.e., age > 65 years and/or comorbidities) are present, and can influence management strategy. The objective was to assess whether HRCT is associated to short-time development of severe disease in patients with COVID-19 pneumonia.

      Methods

      Seventy-seven consecutive patients (mean age, 64 ± 15 years) with mild COVID-19 pneumonia (no or mild respiratory failure) that underwent HRCT were retrospectively identified. Fifty-two on 77 patients had reported risk factors for severe disease. A chest-imaging devoted radiologist recorded, on a per-examination basis, the following HRCT features: ground-glass opacity, crazy-paving pattern, consolidation, organizing pneumonia (OP) pattern, mosaic attenuation, and nodules. The extent of each feature (total feature score, TFS) was semi-quantitatively assessed. Total lung involvement (TLI) was defined as the sum of all TFSs. The study outcome was defined as the occurrence of severe disease (moderate-to-severe respiratory failure) within 15 days from HRCT. Logistic regression analysis was performed to assess if age, comorbidities, and HRCT features were associated to severe disease.

      Results

      On univariable analysis, severe disease was significantly associated with age > 59 years (29/47 patients, 61.7%) (p = 0.013), and not significantly associated with having comorbidities (22/44 patients, 50.0%). On multivariable analysis, TLI >15 and OP pattern >5 were independently associated to severe disease, with odds ratio of 8.380 (p = 0.003), and of 4.685 (p = 0.035), respectively.

      Conclusion

      Short-time onset of severe COVID-19 was associated to TLI >15 and OP pattern score > 5. Severe disease was not associated to comorbidities.

      Keywords

      1. Introduction

      Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been identified as the agent causing the SARS-CoV-2 disease (COVID-19) ongoing pandemic, responsible of a constantly growing number of infections and deaths worldwide [
      Coronavirus disease 2019 (COVID-19) in the EU/EEA and the UK – ninth update, 23 April.
      ]. Clinical presentation of COVID-19 is variable. While some patients are asymptomatic, the disease can manifest with symptoms such as fever, dry cough, and dyspnea, possibly progressing to respiratory failure requiring admission to intensive care unit (ICU) and death [
      • Guan W.
      • Ni Z.
      • Hu Y.
      • et al.
      Clinical characteristics of coronavirus disease 2019 in China.
      ].
      To date, real-time reverse transcription-polymerase chain reaction (RT-PCR) of viral nucleic acid is regarded as the reference standard for COVID-19 diagnosis [
      • Corman V.M.
      • Landt O.
      • Kaiser M.
      • et al.
      Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR.
      ]. The peculiar COVID-19 pandemic contingency we are facing, along with possible constraints in having fast RT-PCR testing, has made high-resolution computed tomography (HRCT) a potentially valuable tool for helping referring physicians in the clinical decision making process. Radiologic societies worldwide released consensus statements and advice documents, in order to clarify the role of imaging in diagnosis and management of patients with suspected or diagnosed COVID-19 [
      ACR recommendations for the use of chest radiography and computed tomography (CT) for suspected COVID-19 infection.
      ,
      RCR position on the role of CT in patients suspected with COVID-19 infection.
      ,
      • Rubin G.D.
      • Ryerson C.J.
      • Haramati L.B.
      • et al.
      The role of chest imaging in patient management during the COVID-19 pandemic: a multinational consensus statement from the Fleischner society.
      ,
      • Revel M.P.
      • Parkar A.P.
      • Prosch H.
      • et al.
      COVID-19 patients and the radiology department - advice from the European Society of Radiology (ESR) and the European Society of Thoracic Imaging (ESTI).
      ]. In particular, in patients who show mild respiratory disease and RT-PCR positivity or moderate-to-high pre-test probability of COVID-19 in the absence of RT-PCR test, imaging is advised for age > 65 years and comorbidities such as cardiovascular disease, diabetes mellitus, chronic respiratory disease, arterial hypertension, and immunocompromise [
      • Rubin G.D.
      • Ryerson C.J.
      • Haramati L.B.
      • et al.
      The role of chest imaging in patient management during the COVID-19 pandemic: a multinational consensus statement from the Fleischner society.
      ]. In this scenario, HRCT can provide a baseline examination for future comparison and assessment of comorbidity-related abnormalities, thus influencing the management and follow-up strategy.
      In the last months, many studies focused on the diagnosis of COVID-19 pneumonia, investigating main and ancillary HRCT features [
      • Li Y.
      • Xia L.
      Coronavirus disease 2019 (COVID-19): role of chest CT in diagnosis and management.
      ,
      • Zhou S.
      • Wang Y.
      • Zhu T.
      • Xia L.
      CT features of coronavirus disease 2019 (COVID-19) pneumonia in 62 patients in Wuhan, China.
      ,
      • Salehi S.
      • Abedi A.
      • Balakrishnan S.
      • Gholamrezanezhad A.
      Coronavirus disease 2019 (COVID-19): a systematic review of imaging findings in 919 patients.
      ], as well as typical and atypical imaging patterns [
      • Simpson S.
      • Kay F.U.
      • Abbara S.
      • et al.
      Radiological Society of North America expert consensus statement on reporting chest CT findings related to COVID-19. Endorsed by the Society of Thoracic Radiology, the American College of Radiology, and RSNA.
      ,
      • Ye Z.
      • Zhang Y.
      • Wang Y.
      • Huang Z.
      • Song B.
      Chest CT manifestations of new coronavirus disease 2019 (COVID-19): a pictorial review.
      ]. As far as we know, a few studies [
      • Tabatabaei S.M.H.
      • Talari H.
      • Moghaddas F.
      • Rajebi H.
      Computed tomographic features and short-term prognosis of coronavirus disease 2019 (COVID-19) pneumonia: a single-center study from Kashan, Iran.
      ,
      • Colombi D.
      • Bodini F.C.
      • Petrini M.
      • et al.
      Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
      ,
      • Yuan M.
      • Yin W.
      • Tao Z.
      • Tan W.
      • Hu Y.
      Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China.
      ,
      • Feng Z.
      • Yu Q.
      • Yao S.
      • et al.
      Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
      ] evaluated whether HRCT can predict the clinical course of disease by means of individual findings (e.g., consolidation, air bronchogram, central lung involvement, pleural effusion, and percentage of well aerated lung parenchyma) or combined findings (e.g., computed tomography [CT] severity score). We therefore aimed to contribute to the evolving knowledge on disease by evaluating whether HRCT has a prognostic role in the early stage of disease.
      The purpose of the study was to assess whether HRCT is associated to short-time development of severe respiratory failure in patients with COVID-19 pneumonia.

      2. Material and methods

      2.1 Study population

      By performing a computerized search, we identified all the consecutive adult patients with suspected COVID-19 who underwent chest HRCT examination in our COVID-19-center between March 08, 2020, and April 16, 2020. The disease severity was classified according to the following criteria from the Italian Society of Emergency Medicine (SIMEU) [
      • Paglia S.
      • Storti E.
      First line Covid-19. Emergency department organizational management within epidemic or pre-epidemic outbreak areas.
      ]: (i) category I disease, including fever without respiratory failure and normal chest X-ray; (ii) category II disease, including fever with chest X-ray and arterial blood gas test indicating lung focus and/or mild respiratory failure (partial pressure of arterial blood oxygen [PaO2] > 60 mmHg); (iii) category III disease, including fever with moderate-severe respiratory failure (PaO2 < 60 mmHg in room air); (iv) category IV disease, including respiratory failure with suspected initial acute respiratory distress syndrome (ARDS) or complicated pneumonia; and (v) category V disease, consisting of ARDS. Treatments included oxygen therapy and/or continuous positive airway pressure (CPAP) ventilation in patients with category III-IV disease, and orotracheal intubation with invasive ventilation in patients with category IV-V disease, respectively [
      • Paglia S.
      • Storti E.
      First line Covid-19. Emergency department organizational management within epidemic or pre-epidemic outbreak areas.
      ].
      Of 192 eligible subjects we excluded 104 patients with negative RT-PCR test for SARS-CoV-2, and 11 patients with SIMEU category III-V at the time of HRCT. Final study population included 77 patients (40 men and 37 women; mean age, 64 ± 15 years) showing confirmed positive result for SARS-CoV-2 from RT-PCR test and SIMEU category I-II disease at the time of HRCT. Flowchart of patient selection is presented in Fig. 1. In the case of multiple HRCTs, only the baseline examination was included in the analysis.
      Fig. 1
      Fig. 1Flowchart of patient selection.
      HRCT: high-resolution computed tomography; SIMEU: Italian Society of Emergency Medicine; RT-PCR: real-time reverse transcription-polymerase chain reaction; SARS-CoV-2: Severe acute respiratory syndrome coronavirus 2.

      2.2 HRCT examinations

      Examinations were performed on a 64-row scanner (LightSpeed, General Electric, Milwaukee, Wisconsin, USA), with the patient in the supine position. The whole thorax was scanned volumetrically at suspended full inspiration, with acquisition parameters as follows: tube potential, 120 kV; tube current modulation range, 100–350 mA; gantry revolution time, 0.6 s; detector configuration, 64 mm × 0.625 mm; reconstructed section thickness and reconstructed interval, 1.25 mm. Iodinated contrast administration (iomeprol 350 mgI/mL, [Iomeron, Bracco Imaging, Milan, Italy]) was performed in 4/77 patients (5.2%).
      Two sets of images were reconstructed and displayed, namely a first set with high-spatial-frequency algorithm and pulmonary parenchyma windowing (level, −500 HU; width, 1700 HU), and a second set with soft tissue algorithm and windowing (level, 50 HU; width, 350 HU).

      2.3 Image analysis

      A radiologist with 10 years of experience in thoracic imaging reviewed all the chest HRCT examinations on a picture archiving and communication system workstation (Suitestensa Ebit srl, Esaote Group Company, Genoa, Italy), blinded to patients' clinical and laboratory data.
      For each patient, the reader recorded the following six main HRCT pulmonary features: (i) ground-glass opacity (GGO); (ii) crazy-paving pattern; (iii) consolidation; (iv) parenchymal findings of organizing pneumonia (OP) (i.e., GGO or consolidation triangular or polygonal in shape, perilobular pattern, bronchial dilatation, reverse halo sign, linear and band-like opacities, and signs of fibrosis) (18); (v) mosaic attenuation; and (vi) nodules.
      On a per-examination basis, each lung was divided into three zones, as follows: upper zone (above the carina), middle zone (from the carina to the inferior pulmonary veins), and lower zone (below the inferior pulmonary veins), resulting in a total of 6 zones (3 per lung). The reader assessed the zonal extent of each of the over-mentioned six HRCT pulmonary features, scoring it semi-quantitatively as follows: score 0 if there was no involvement; score 1 in the case of <25% involvement; score 2 for a ≥ 25% to <50% involvement; score 3 for a ≥ 50% to <75% involvement; score 4 for ≥75% involvement [
      • Yuan M.
      • Yin W.
      • Tao Z.
      • Tan W.
      • Hu Y.
      Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China.
      ]. Therefore, the total score for each pulmonary feature (total feature score [TFS]) ranged 0–24. Total lung involvement (TLI) was defined as the sum of all TFSs. Presence of significant pleural effusion (>3 cm) was also evaluated.

      2.4 Clinical data analysis

      For all patients we recorded age, sex, presence of comorbidities (number and type), time from symptoms onset and HRCT examination, and SIMEU category as assigned by the referring physician (both at the time of HRCT, as well as the worst one observed in the 15 days following the examination). When not clearly reported by the referring physician, the SIMEU category was derived by the study coordinator, who was not involved in image analysis, and was unblinded to clinical data. This occurred in 26/77 patients at the time of HRCT, and 17/77 patients in the 15 days following HRCT.
      For the purpose of analysis, we dichotomized SIMEU categories into two groups, namely: (i) mild disease group, including patients with no or mild respiratory failure (SIMEU category I or category II); (ii) severe disease group, including patients with moderate-to-severe respiratory failure or ARDS (SIMEU category III to V).

      2.5 Statistical analysis

      After checking for data normality with the Shapiro–Wilk test, we presented clinical data and HRCT features with mean ± standard deviation or median with the interquartile range (IQR). Proportions were coupled with 95% confidence intervals (95%CI) when relevant.
      First, we performed a receiver-operating characteristic (ROC) analysis, using the Youden index to calculate the cut-off of TFS, TLI, and age better balancing sensitivity and specificity in assessing the study outcome. The latter was defined as the occurrence of severe disease in the 15 days following the examination, i.e. the shift from SIMEU category I-II at the time of HRCT to SIMEU category III-V. The area under the curve (AUC) was calculated as well.
      Second, we run a logistic regression analysis with the stepwise approach to assess which of the clinical variables and HRCT features was associated as an independent predictor of the study outcome occurrence. The model included age, sex, number of comorbidities and the HRCT features showing the strongest association with the outcome at the preliminary univariable analysis, performed with the chi-square test. TFS, TLI and age values were dichotomized using the operative cut-offs obtained with ROC analysis before entering them in the model.
      Statistical analysis was performed using a commercially available software (MedCalc Software bvba, version 18.11.6, Ostend, Belgium). The alpha level was set to 0.05.
      Our referring Ethical Committee approved the study. The acquisition of informed consent was waived, due to the retrospective design.

      3. Results

      3.1 Study population

      Clinical characteristics of the study population are detailed in Table 1. The median time between symptoms onset and HRCT examination was 5 days (IQR, 2–9 days). Forty-four over 77 patients (57%) had ≥1 comorbidities, while 20/77 (26%) patients had ≥2 comorbidities. Cardiovascular diseases were the most frequent ones (32/77, 42%). In the 15-day period following HRCT examination, 38/77 patients (49%) developed severe disease. The median time between HRCT and severe disease onset was 1 day (IQR, 1–2 days).
      Table 1Patient-related clinical variables.
      Clinical variablesAll patients (n = 77)
      Age (mean ± standard deviation)64 ± 15 years
      Sex, n (%, 95%CI)
      Female37 (48, 34–66)
      Male40 (52, 37–71)
      Comorbidities, n (%, 95%CI)
      ≥1 comorbidity (all types)44 (57, 42–77)
      ≥2 comorbidities (all types)20 (26, 16–40)
      Cardiovascular32 (42, 28–59)
      Respiratory9 (12, 5–22)
      Chronic renal failure3 (4, 1–11)
      Tumors10 (13, 6–24)
      Obesity4 (5, 1–13)
      Diabetes mellitus6 (8, 3–17)
      Immunocompromise3 (4, 1–11)
      The results of ROC analysis defining the operative cut-off values for the analysis are reported in Table 2.
      Table 2Results of the receiver operating characteristic (ROC) analysis defining operative cut-off values.
      VariablesCut-offAUC
      Clinical features
      Age (years)>590.663
      HRCT features (score)
      GGO>40.809
      OP pattern>50.851
      Consolidations>10.732
      Crazy-paving pattern>10.532
      Mosaic attenuation>00.660
      Nodules>10.565
      TLI>150.886
      AUC: area under the curve; HRCT: high-resolution computed tomography; GGO: ground-glass opacities; OP: organizing pneumonia; TLI: total lung involvement (see the text for details).

      3.2 HRCT features

      HRCT findings are listed in Table 3. The most frequent HRCT features were GGO and OP pattern, reported in 66/77 (86%) and in 71/77 (92%) patients, respectively. The other features occurred with the following frequency: consolidation in 50/77 cases (65%), crazy-paving pattern in 43/77 cases (56%), mosaic attenuation in 32/77 cases (42%), and nodules in 20/77 cases (26%). Three patients (4%) had no visible lung involvement. Pleural effusion >3 cm in thickness was found in 1 patient.
      Table 3HRCT findings in the study population. Proportions are calculated over 77 patients.
      HRCT findingsNumber of patients with findingsTFS median (IQR)
      HRCT features, n (%, 95%CI)
      GGO66 (86, 66–100)6 (2–8)
      OP pattern71 (92, 72–100)5 (2–8)
      Consolidations50 (65, 48–86)1 (0–3)
      Crazy-paving pattern43 (56, 40–75)1 (0–2)
      Mosaic attenuation32 (42, 28–59)0 (0–2)
      Nodules20 (26, 16–40)0 (0–1)
      Pleural effusion >3 cm1 (1, 0–7)
      HRCT: high-resolution computed tomography; GGO: ground-glass opacities; OP: organizing pneumonia; TFS: total feature score (see the text for details).
      Lung involvement was bilateral in 70/77 patients (91%), and unilateral in 4/77 (5%). Upper, middle, and lower lung zones were affected in 64 (83%), 72 (94%), and 73 (95%) of 77 patients, respectively. TFS values are reported in Table 3. Median TLI was 16 (IQR 8.75–22).

      3.3 Association with severe disease

      Results from logistic regression analysis are shown in Table 4. On univariable analysis, severe disease occurred at a significantly higher extent (p = 0.013) in patients older than >59 years (29/47 [61.7%; 95%CI 41.3–88.6]) than in ≤59 year-old patients (9/30 [30.0%; 95%CI 13.7–57.0]. On the contrary, the occurrence of severe disease was not significantly associated with having ≥1 comorbidities (22/44 [50.0%; 95%CI 31.3–75.7]) rather than <1 comorbidities (16/33 [48.5%; 95%CI 27.7–78.7]) (p = 0.921).
      Table 4Results from the logistic regression model (endpoint: development of severe disease)
      Prevalence of outcomeUnivariable analysisMultivariable analysis
      Variablesn (%, 95%CI)p (Chi-square test)OR (95%CI), p
      Age > 59 years29/47 (61.7, 41.3–88.6)0.013
      Sex (M)23/40 (57.5, 36.5–86.3)0.208
      Comorbidities ≥122/44 (50, 31.3–75.7)0.921
      GGO >433/47 (70.2, 48.3–98.6)<0.0001
      OP pattern >527/32 (84.4, 55.6–100)<0.00014.685 (1.111–19.755), 0.035
      Consolidation >126/37 (70.3, 45.9–100)0.001
      TLI >1532/40 (80, 54.7–100)<0.00018.380 (2.087–33.647), 0.003
      M: male; GGO: ground-glass opacities; OP: organizing pneumonia; OR: odds ratio; TLI: total lung involvement (see the text for details).
      On multivariable analysis, TLI > 15 and OP pattern >5 were independently associated with the development of severe disease. Overall, the latter occurred in 33 over 42 patients with TLI > 15 and/or OP pattern >5 (78.6%; 95%CI 54.1–100), and in 26 over 30 patients with TLI > 15 and OP pattern >5 (86.7%; 95%CI 56.6–100). We excluded from the multivariable model the variables with the lowest association with the outcome at univariable analysis, i.e. crazy-paving pattern >1 (prevalence in patients with severe disease 14/38 [37%; p = 0.572]), mosaic attenuation >0 (22/38 [58%; p = 0.008]), and nodules >1 (7/38 [18%; p = 0.144]).
      Example cases are illustrated in Fig. 2, Fig. 3.
      Fig. 2
      Fig. 256-year-old man with confirmed severe acute respiratory syndrome coronavirus 2 disease (COVID-19) pneumonia and no comorbidities. High-resolution computed tomography (HRCT) was performed on day of hospital admission.
      A–B, HRCT images on the axial plane (A) and on coronal reformation (B) depicted bilateral, peripheral ground-glass opacities (GGO) with crazy-paving pattern. Total lung involvement (TLI) score was 16, above the cut-off value of 15. After 5 days the patient developed respiratory failure (Italian Society of Emergency Medicine [SIMEU] category IV disease).
      Fig. 3
      Fig. 382-year-old man with confirmed severe acute respiratory syndrome coronavirus 2 disease (COVID-19) pneumonia and no comorbidities. High-resolution computed tomography (HRCT) was performed on day of hospital admission.
      A–B, HRCT images on the axial plane (A) and on sagittal reformation (B) showed ground-glass opacities (GGO) and band-like opacities with a perilobular distribution, resembling an organizing pneumonia (OP) pattern. OP pattern score was 8 (above the cut-off value of 5). After 3 days the patient developed respiratory failure (Italian Society of Emergency Medicine [SIMEU] category III disease).

      4. Discussion

      There is intense debate on the use of HRCT in COVID-19 pneumonia, given the variability in locally available resources, and need for avoiding secondary exposure of patients and healthcare professional in the Radiology Department [
      • Sverzellati N.
      • Milanese G.
      • Milone F.
      • Balbi M.
      • Ledda R.E.
      • Silva M.
      Integrated radiologic algorithm for COVID-19 pandemic.
      ,
      • Hope M.D.
      • Raptis C.A.
      • Shah A.
      • Hammer M.M.
      • Henry T.S.
      • six signatories
      A role for CT in COVID-19? What data really tell us so far.
      ]. We found that short-term clinical evolution, i.e. a shift from mild disease (SIMEU category I-II) to severe disease (SIMEU category III-V) within 15 days from baseline HRCT, was significantly associated with the presence of an OP pattern score > 5, and a TLI score > 15. This association was also found in univariable analysis for patients aged >59 years.
      In our series, about two third of patients mirrored the “clinical scenario 1” recently prompted by a Fleischner Society consensus document [
      • Rubin G.D.
      • Ryerson C.J.
      • Haramati L.B.
      • et al.
      The role of chest imaging in patient management during the COVID-19 pandemic: a multinational consensus statement from the Fleischner society.
      ], which recommends chest imaging when mild features consistent with COVID-19 pneumonia coexist with risk factors for severe disease (age > 65 years or at least one comorbidity). Our results reasonably validate that indication to HRCT in order to provide a baseline examination for future comparison, and intensify the monitoring of patients with imaging findings predictive for clinical worsening.
      However, in accordance with Wei et al. [
      • Wei Y.Y.
      • Wang R.R.
      • Zhang D.W.
      • et al.
      Risk factors for severe COVID-19: evidence from 167 hospitalized patients in Anhui, China.
      ], we found no relationship between comorbidities and severe disease at multivariable analysis. This result is in contrast with previous Authors, who found comorbidities to predict several adverse outcomes [
      • Feng Z.
      • Yu Q.
      • Yao S.
      • et al.
      Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
      ,
      • Li K.
      • Wu J.
      • Wu F.
      • et al.
      The clinical and chest CT features associated with severe and critical COVID-19 pneumonia.
      ], including admission to ICU and death [
      • Colombi D.
      • Bodini F.C.
      • Petrini M.
      • et al.
      Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
      ]. The discrepancy might be related to confounding factors such as smoking status, and under-reporting of pre-existing pathologies impacting on COVID-19 outcomes [
      • Jordan R.E.
      • Adab P.
      • Cheng K.K.
      Covid-19: risk factors for severe disease and death.
      ]. This might also explain why about 50% of patients (16/33) with no comorbidities progressed to severe disease in our series. Our results suggest that, in the current scenario in which it is difficult to assess the presence or type of comorbidities representing risk factors [
      • Jordan R.E.
      • Adab P.
      • Cheng K.K.
      Covid-19: risk factors for severe disease and death.
      ], HRCT findings might help in identifying patients at risk of clinical progression by imaging features alone. Whether this can lead refining current indications to CT should be the matter for further research. Of note, our model included comorbidities as a whole rather than individually. This choice was related to the low prevalence of each comorbidity in our series, except for cardiovascular ones (42%). We believe that further studies on larger populations should address whether different comorbidities can affect prognosis at a different degree.
      Our results on TLI are in line with previous studies showing that the extent of lung involvement, defined as CT score [
      • Yuan M.
      • Yin W.
      • Tao Z.
      • Tan W.
      • Hu Y.
      Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China.
      ], CT severity score [
      • Feng Z.
      • Yu Q.
      • Yao S.
      • et al.
      Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
      ], or, conversely, well aerated lung parenchyma [
      • Colombi D.
      • Bodini F.C.
      • Petrini M.
      • et al.
      Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
      ], is predictive of mortality, progression to severe disease, and ICU admission or death, respectively. This is reasonably related to the extensive diffuse alveolar damage (DAD) as the distinctive pathophysiological characteristic of the disease [
      • Tian S.
      • Xiong Y.
      • Liu H.
      • et al.
      Pathological study of the 2019 novel coronavirus disease (COVID-19) through postmortem core biopsies.
      ,
      • Kligerman S.J.
      • Franks T.J.
      • Galvin J.R.
      From the radiologic pathology archives: organization and fibrosis as a response to lung injury in diffuse alveolar damage, organizing pneumonia, and acute fibrinous and organizing pneumonia.
      ]. As expected from still limited knowledge on COVID-19 pneumonia, our results are difficult to compare with previous ones, e.g. in terms of definition of severe disease, method for assessing lung involvement, and amount of involved lung parenchyma that predicts clinical progression. While software-based methods such as the one proposed by Colombi et al. [
      • Colombi D.
      • Bodini F.C.
      • Petrini M.
      • et al.
      Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
      ] can expectedly provide reliable and repeatable quantitative assessment, TLI can reasonably be of help when software-based evaluation is unavailable, or provide more reproducible results than those of different software-based methods. Of note, differently from Feng et al. [
      • Feng Z.
      • Yu Q.
      • Yao S.
      • et al.
      Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
      ], who assessed lung involvement using opacification and consolidation, we calculated TLI as the sum of different TFSs, i.e. as one single index combining the amount of affected lung and the type of HRCT features (e.g., GGO, consolidation, crazy-paving pattern, and OP pattern). We believe this can better account for the spectrum of DAD-related findings with which COVID-19 pneumonia can present.
      To the best of our knowledge, no previous studies assessed the role for OP pattern in assessing COVID-19 pneumonia. This feature has been recognized as a typical imaging marker in the later course of the disease [
      • Simpson S.
      • Kay F.U.
      • Abbara S.
      • et al.
      Radiological Society of North America expert consensus statement on reporting chest CT findings related to COVID-19. Endorsed by the Society of Thoracic Radiology, the American College of Radiology, and RSNA.
      ], as well as in severe acute respiratory syndrome (SARS), and Middle East respiratory syndrome (MERS) [
      • Tse G.M.
      • To K.F.
      • Chan P.K.
      • et al.
      Pulmonary pathological features in coronavirus associated severe acute respiratory syndrome (SARS).
      ,
      • Kim I.
      • Lee J.E.
      • Kim K.H.
      • Lee S.
      • Lee K.
      • Mok J.H.
      Successful treatment of suspected organizing pneumonia in a patient with Middle East respiratory syndrome coronavirus infection: a case report.
      ]. OP pattern consists of bilateral, peripheral or peribronchial GGO or consolidation that are triangular or polygonal in shape, along with perilobular pattern, bronchial dilatation, linear and band-like opacities, and signs of fibrosis [
      • Polverosi R.
      • Maffesanti M.
      • Dalpiaz G.
      Organizing pneumonia: typical and atypical HRCT patterns.
      ]. Characteristic histopathological presentation includes intra-alveolar organizing fibrous plugs, loose interstitial fibrosis, and chronic inflammatory infiltrates [
      • Tian S.
      • Xiong Y.
      • Liu H.
      • et al.
      Pathological study of the 2019 novel coronavirus disease (COVID-19) through postmortem core biopsies.
      ,
      • Kligerman S.J.
      • Franks T.J.
      • Galvin J.R.
      From the radiologic pathology archives: organization and fibrosis as a response to lung injury in diffuse alveolar damage, organizing pneumonia, and acute fibrinous and organizing pneumonia.
      ,
      • Zhang H.
      • Zhou P.
      • Wei Y.
      • et al.
      Histopathologic changes and SARS-CoV-2 immunostaining in the lung of a patient with COVID-19.
      ,
      • Pernazza A.
      • Mancini M.
      • Rullo E.
      • et al.
      Early histologic findings of pulmonary SARS-CoV-2 infection detected in a surgical specimen.
      ]. We found that an OP pattern score > 5, corresponding to about 20% of lung involvement, was predictive of short-time progression to severe disease. In accordance with the pathogenic pathway proposed by Siddiqi et al. [
      • Siddiqi H.K.
      • Mehra M.R.
      COVID-19 illness in native and immunosuppressed states: a clinical-therapeutic staging proposal.
      ], we hypothesize that the occurrence of an OP pattern in a patient with mild COVID-19 pneumonia can be an imaging marker of the ongoing host inflammatory response causing progression to severe disease. If confirmed by further studies, our result might be of help in identifying patients at risk of clinical evolution, or that can benefit from therapies targeted to the systemic hyperinflammation phase of the disease [
      • Siddiqi H.K.
      • Mehra M.R.
      COVID-19 illness in native and immunosuppressed states: a clinical-therapeutic staging proposal.
      ].
      Several study limitations warrant mention. First, the study results have been observed in a small sample size, on a retrospective, single-center basis. Since the ongoing pandemic makes difficult designing and organizing prospective multicenter trials, we believe that our results can be of interest, as testified by the fact they are in line with previous ones in suggesting a potential prognostic role for HRCT [
      • Tabatabaei S.M.H.
      • Talari H.
      • Moghaddas F.
      • Rajebi H.
      Computed tomographic features and short-term prognosis of coronavirus disease 2019 (COVID-19) pneumonia: a single-center study from Kashan, Iran.
      ,
      • Colombi D.
      • Bodini F.C.
      • Petrini M.
      • et al.
      Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
      ,
      • Yuan M.
      • Yin W.
      • Tao Z.
      • Tan W.
      • Hu Y.
      Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China.
      ,
      • Feng Z.
      • Yu Q.
      • Yao S.
      • et al.
      Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
      ]. Second, having involved one single reader made impossible evaluating the inter-reader agreement. Though results can be reasonably considered robust by having been obtained by a chest-imaging devoted radiologist, we acknowledge that further studies should be performed to assess the reliability of assessing TLI and OP pattern. Finally, we investigated progression to severe disease only, excluding long-term prognosis outcomes. However, the reported median time from clinical onset to ARDS development or ICU admission is 12 days according to Zhou et al. [
      • Zhou F.
      • Yu T.
      • Du R.
      • et al.
      Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study [published correction appears in lancet 2020 mar 28;395(10229):1038].
      ]. Thus, it is reasonably unlikely that progression from mild to severe disease occurs later than the 15-day period we selected. Studies on longer periods of time should assess clinical and imaging sequelae from COVID-19 pneumonia.
      In conclusion, we observed that short-time onset of severe COVID-19 disease in patients who underwent HRCT was independently associated with the extent of lung involvement (TLI > 15) and OP pattern (score > 5). Severe disease was not associated to comorbidities in our series. Indeed, we observed clinical progression in about 50% of cases with no reported comorbidities, suggesting that HRCT has the potential to identify patients at risk of progression even beyond currently accepted and/or known risk factors. Further studies with prospective design should validate our results on a larger scale.

      Declaration of competing interest

      None.
      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      References

      1. Coronavirus disease 2019 (COVID-19) in the EU/EEA and the UK – ninth update, 23 April.
        • Guan W.
        • Ni Z.
        • Hu Y.
        • et al.
        Clinical characteristics of coronavirus disease 2019 in China.
        N Engl J Med. 2020; 382: 1708-1720
        • Corman V.M.
        • Landt O.
        • Kaiser M.
        • et al.
        Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR.
        Euro Surveill. 2020; 25: 2000045
      2. ACR recommendations for the use of chest radiography and computed tomography (CT) for suspected COVID-19 infection.
      3. RCR position on the role of CT in patients suspected with COVID-19 infection.
        • Rubin G.D.
        • Ryerson C.J.
        • Haramati L.B.
        • et al.
        The role of chest imaging in patient management during the COVID-19 pandemic: a multinational consensus statement from the Fleischner society.
        Chest. 2020 Apr 7; 158: 106-116
        • Revel M.P.
        • Parkar A.P.
        • Prosch H.
        • et al.
        COVID-19 patients and the radiology department - advice from the European Society of Radiology (ESR) and the European Society of Thoracic Imaging (ESTI).
        Eur Radiol. 2020; 1-7 (Apr. [Epub ahead of print]): 20
        • Li Y.
        • Xia L.
        Coronavirus disease 2019 (COVID-19): role of chest CT in diagnosis and management.
        AJR Am J Roentgenol. 2020; 4 (Mar. [Epub ahead of print]): 1-7
        • Zhou S.
        • Wang Y.
        • Zhu T.
        • Xia L.
        CT features of coronavirus disease 2019 (COVID-19) pneumonia in 62 patients in Wuhan, China.
        AJR Am J Roentgenol. 2020; 5 (Mar. [Epub ahead of print]): 1-8
        • Salehi S.
        • Abedi A.
        • Balakrishnan S.
        • Gholamrezanezhad A.
        Coronavirus disease 2019 (COVID-19): a systematic review of imaging findings in 919 patients.
        AJR Am J Roentgenol. 2020; 14 (Mar. [Epub ahead of print]): 1-7
        • Simpson S.
        • Kay F.U.
        • Abbara S.
        • et al.
        Radiological Society of North America expert consensus statement on reporting chest CT findings related to COVID-19. Endorsed by the Society of Thoracic Radiology, the American College of Radiology, and RSNA.
        J Thorac Imaging. 2020 Apr 28; ([Epub ahead of print])https://doi.org/10.1097/RTI.0000000000000524
        • Ye Z.
        • Zhang Y.
        • Wang Y.
        • Huang Z.
        • Song B.
        Chest CT manifestations of new coronavirus disease 2019 (COVID-19): a pictorial review.
        Eur Radiol. 2020; 19 (Mar. [Epub ahead of print]): 1-9
        • Tabatabaei S.M.H.
        • Talari H.
        • Moghaddas F.
        • Rajebi H.
        Computed tomographic features and short-term prognosis of coronavirus disease 2019 (COVID-19) pneumonia: a single-center study from Kashan, Iran.
        Radiology: Cardiothoracic Imaging. 2020; 20 (Apr. [Epub ahead of print]): 2(2)
        • Colombi D.
        • Bodini F.C.
        • Petrini M.
        • et al.
        Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia.
        Radiology. 2020 Apr 17; ([Epub ahead of print]): 201433
        • Yuan M.
        • Yin W.
        • Tao Z.
        • Tan W.
        • Hu Y.
        Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China.
        PLoS One. 2020; 15: e0230548
        • Feng Z.
        • Yu Q.
        • Yao S.
        • et al.
        Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics.
        (medRxiv)2020 Feb 23 ([2020.02.19.20025296 [Epub ahead of print])
        • Paglia S.
        • Storti E.
        First line Covid-19. Emergency department organizational management within epidemic or pre-epidemic outbreak areas.
        https://www.simeu.it/w/articoli/leggiArticolo/4015/leggi;
        Date: 2020
        Date accessed: June 10, 2020
        • Polverosi R.
        • Maffesanti M.
        • Dalpiaz G.
        Organizing pneumonia: typical and atypical HRCT patterns.
        Radiol Med. 2006; 111: 202-212
        • Sverzellati N.
        • Milanese G.
        • Milone F.
        • Balbi M.
        • Ledda R.E.
        • Silva M.
        Integrated radiologic algorithm for COVID-19 pandemic.
        J Thorac Imaging. 2020; ([Epub ahead of print])https://doi.org/10.1097/RTI.0000000000000516
        • Hope M.D.
        • Raptis C.A.
        • Shah A.
        • Hammer M.M.
        • Henry T.S.
        • six signatories
        A role for CT in COVID-19? What data really tell us so far.
        Lancet. 2020; 395: 1189-1190
        • Wei Y.Y.
        • Wang R.R.
        • Zhang D.W.
        • et al.
        Risk factors for severe COVID-19: evidence from 167 hospitalized patients in Anhui, China.
        J Infect. 2020; 81: e89-e92
        • Li K.
        • Wu J.
        • Wu F.
        • et al.
        The clinical and chest CT features associated with severe and critical COVID-19 pneumonia.
        Invest Radiol. 2020; 55: 327-331
        • Jordan R.E.
        • Adab P.
        • Cheng K.K.
        Covid-19: risk factors for severe disease and death.
        BMJ. 2020; 368 ([Epub ahead of print]): m1198
        • Tian S.
        • Xiong Y.
        • Liu H.
        • et al.
        Pathological study of the 2019 novel coronavirus disease (COVID-19) through postmortem core biopsies.
        Mod Pathol. 2020 Apr 14; ([Epub ahead of print])https://doi.org/10.1038/s41379-020-0536-x
        • Kligerman S.J.
        • Franks T.J.
        • Galvin J.R.
        From the radiologic pathology archives: organization and fibrosis as a response to lung injury in diffuse alveolar damage, organizing pneumonia, and acute fibrinous and organizing pneumonia.
        Radiographics. 2013; 33: 1951-1975
        • Tse G.M.
        • To K.F.
        • Chan P.K.
        • et al.
        Pulmonary pathological features in coronavirus associated severe acute respiratory syndrome (SARS).
        J Clin Pathol. 2004; 57: 260-265
        • Kim I.
        • Lee J.E.
        • Kim K.H.
        • Lee S.
        • Lee K.
        • Mok J.H.
        Successful treatment of suspected organizing pneumonia in a patient with Middle East respiratory syndrome coronavirus infection: a case report.
        J Thorac Dis. 2016; 8: E1190-E1194
        • Zhang H.
        • Zhou P.
        • Wei Y.
        • et al.
        Histopathologic changes and SARS-CoV-2 immunostaining in the lung of a patient with COVID-19.
        Ann Intern Med. 2020 Mar 12; ([Epub ahead of print]): M20-0533
        • Pernazza A.
        • Mancini M.
        • Rullo E.
        • et al.
        Early histologic findings of pulmonary SARS-CoV-2 infection detected in a surgical specimen.
        Virchows Arch. 2020; 1-6 (Apr): 30
        • Siddiqi H.K.
        • Mehra M.R.
        COVID-19 illness in native and immunosuppressed states: a clinical-therapeutic staging proposal.
        J Heart Lung Transplant. 2020; 39: 405-407
        • Zhou F.
        • Yu T.
        • Du R.
        • et al.
        Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study [published correction appears in lancet 2020 mar 28;395(10229):1038].
        Lancet. 2020; 395: 1054-1062