European Radiology https://doi.org/10.1007/s00330-022-08580-2 CHEST Air trapping in COVID-19 patients following hospital discharge: retrospective evaluation with paired inspiratory/expiratory thin-section CT Tomás Franquet 1,2 & Ana Giménez 1,2 & Loren Ketai 3 & Sandra Mazzini 1,2 & Andrea Rial 4 & Virginia Pomar 2,5,6 & Pere Domingo 2,5,6 Received: 18 August 2021 / Revised: 14 January 2022 / Accepted: 15 January 2022 # The Author(s) 2022 Abstract Objectives The study reports our experience with paired inspiration/expiration thin-section computed tomographic (CT) scans in the follow-up of COVID-19 patients with persistent respiratory symptoms. Methods From August 13, 2020, to May 31, 2021, 48 long-COVID patients with respiratory symptoms (27 men and 21 women; median age, 62.0 years; interquartile range: 54.0–69.0 years) underwent follow-up paired inspiration-expiration thin-section CT scans. Patient demographics, length of hospital stay, intensive care unit admission rate, and clinical and laboratory features of acute infection were also included. The scans were obtained on a median of 72.5 days after onset of symptoms (interquartile range: 58.5–86.5) and at least 30 days after hospital discharge. Thin-section CT findings included ground-glass opacity, mosaic attenuation pattern, consolidation, traction bronchiectasis, reticulation, parenchy- mal bands, bronchial wall thickening, and air trapping. We used a quantitative score to determine the degree of air trapping in the expiratory scans. Results Parenchymal abnormality was found in 50% (24/48) of patients and included air trapping (37/48, 77%), ground-glass opacities (19/48, 40%), reticulation (18/48, 38%), parenchymal bands (15/48, 31%), traction bronchiectasis (9/48, 19%), mosaic attenuation pattern (9/48, 19%), bronchial wall thickening (6/48, 13%), and consolidation (2/48, 4%). The absence of air trapping was observed in 11/48 (23%), mild air trapping in 20/48 (42%), moderate in 13/48 (27%), and severe in 4/48 (8%). Independent predictors of air trapping were, in decreasing order of importance, gender (p = 0.0085), and age (p = 0.0182). Conclusions Our results, in a limited number of patients, suggest that follow-up with paired inspiratory/expiratory CT in long- COVID patients with persistent respiratory symptoms commonly displays air trapping. Key Points • Our experience indicates that paired inspiratory/expiratory CT in long-COVID patients with persistent respiratory symptoms commonly displays air trapping. • Iterative reconstruction and dose-reduction options are recommended for demonstrating air trapping in long-COVID patients. Keywords Air-trapping . Long-COVID . Inspiratory * Tomás Franquet 3 Department of Radiology, University of New Mexico Health Science tfranquet@santpau.cat Center, Albuquerque, NM, USA * Ana Giménez 4 Department of Pneumology, Hospital de la Santa Creu i Sant Pau, agimenez@santpau.cat Barcelona, Spain 5 Department of Infectious Diseases, Hospital de la Santa Creu i Sant 1 Department of Radiology, Hospital de la Santa Creu i Sant Pau, Pau, Barcelona, Spain Barcelona, Spain 6 Institut de Recerca Biomèdica del Hospital de Sant Pau, 2 Department of Medicine, Universitat Autónoma de Barcelona, Barcelona, Spain Barcelona, Spain Eur Radiol Abbreviations Expiratory CT has established itself as an essential adjunct AST Aspartate aminotransferase to conventional CT, in the demonstration of air trapping in COVID-19 Coronavirus disease-19 patients with suspected obstructive small airway disease CPR C-reactive protein [12]. In our institution, paired inspiratory/expiratory CT ex- DLCO Diffusing capacity for carbon monoxide amination is obtained routinely to evaluate patients in whom ICU Intensive care unit airway disease is suspected. IL-6 Interleukin-6 The purpose of our study was to evaluate the usefulness of IQR Interquartile range paired inspiratory/expiratory CT in the detection of air trap- KL-6 Krebs von den Lungen glycoprotein-6 ping in the follow-up of long-COVID patients presenting with LDCT Low-dose CT respiratory complaints. LDH Lactate dehydrogenase LFTs Lung function tests NETs Neutrophil extracellular traps Materials and methods RT-PCR Reverse transcription polymerase chain reaction Patient population SAD Small airway disease SARS-CoV-2 Severe acute respiratory Between August 13, 2020, and May 31, 2021, 1722 proved syndrome-coronavirus-2 COVID-19 patients were admitted to the Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 105 died, whereas 1617 were discharged. Among them, 162 patients fulfilled the clin- Introduction ical criteria for diagnosis of long COVID. After excluding 102 patients because of smoking or prior respiratory disease, and The coronavirus disease 2019 (COVID-19) pandemic caused 12 who declined or were unable to participate, 48 previously by the severe acute respiratory syndrome coronavirus 2 healthy patients with respiratory symptoms were enrolled in (SARS-CoV-2) arose in late December 2019 in Wuhan, the present study. Hubei Province, China [1–3]. SARS-CoV-2 has resulted in a These patients were recruited during regularly scheduled pandemic with more than 180 million confirmed cases world- follow-up visits. All the patients had persistent cough and/or wide and a death toll of more than 3.9 million (Johns Hopkins dyspnea. Coronavirus Resource Center. https://coronavirus.jhu.edu/ All 48 patients underwent paired inspiratory/expiratory CT map.html). after clinical recovery and were released from quarantine, be- Clinically, COVID-19 infection can range from asymp- tween 22 and 96 days (mean, 51.8 days ± 20.2), after the onset tomatic or mild illness (e.g., fever with or without cough) to of symptoms. In 9 out of the 48 patients, a previous conven- severe respiratory distress, multiorgan failure, and death. tional inspiratory CT in the acute phase of the infection was The term “long-COVID” describes those patients with available. COVID-19 who experience persistent symptoms regardless Patients were divided into two groups: those with severe of initial disease severity or age. Respiratory symptoms in- respiratory disease (i.e., requiring ICU admission because of clude dyspnea, shortness of breath, and repetitive cough, and ARDS) and those with mild to moderate disease (not admitted it might take weeks or months to eventually recovery [2–5]. to the ICU). Adult respiratory distress syndrome (ARDS), Typical and atypical imaging findings of COVID-19 infec- defined according to the American-European consensus con- tion have been extensively reported [6–8]. A meta-analysis ference [10], occurred in 14 (29.2%) of 48 patients. including twenty-eight studies and 3,466 patients has been Demographic, clinical, and laboratory data related to the recently published to optimize the diagnostic interpretation severity of the respiratory disease were recorded. of chest CT scanning for COVID-19 [9]. In patients without Demographics include age, sex, and comorbidities (arterial severe respiratory disease, the major pulmonary CT findings hypertension, diabetes mellitus, dyslipidemia, epilepsia, and of COVID-19 are ground-glass opacities (GGOs) with a typ- gout), clinical and radiologic data, and treatment modalities, ical bilateral and multilobar distribution and middle-lower such as corticosteroids, tocilizumab, or other immunomodu- lung predominance, mainly with a subpleural distribution lating agents. Laboratory data included peak serum C-reactive [10, 11]. In more severely affected patients, consolidation re- protein (CPR) levels, lactate dehydrogenase (LDH), aspartate places GGO, and the abnormalities extend to the upper lobes aminotransferase (AST), D-dimer, ferritin, interleukin-6 (IL- or become bilateral or both [11]. 6), Krebs von den Lungen glycoprotein (KL-6), and leukocyte However, little is known about sequential CT findings dur- differential counts. ing the subsequent course of COVID-19, especially about This study was conducted with institutional review board sequelae that may occur during convalescence. approval and was performed during the period of Eur Radiol convalescence, approximately 3 months after the patient’s The major CT findings were described based on the rec- clinical recovery from COVID-19. Written informed consent ommendations of the Nomenclature Committee of the was obtained from all patients before inclusion. The study was Fleischner Society [13]. Thin-section CT findings, including approved by the Ethics Committee of the Hospital de la Santa ground-glass opacity, mosaic-attenuation pattern, consolida- Creu I Sant Pau. tion, traction bronchiectasis, reticulation, parenchymal bands, bronchial wall thickening, and air trapping, were recorded. Laboratory methods Ground-glass opacity involved an increase in lung parenchy- mal opacification without obscuration of the underlying ves- The diagnosis of SARS-CoV-2 infection was microbiologi- sels, and consolidation involved an increase in parenchymal cally proved in all the patients. opacification with obscuration of the underlying vessels. Nasopharyngeal swabs were obtained according to a stan- Parenchymal bands were defined as non-tapering linear opac- dardized hospital protocol. Detection of SARS-CoV-2 was ities a few millimeters thick and several centimeters long. Air done through RT-PCR (Xpert® Xpress SARS-CoV-2). trapping was defined as an area of low attenuation in contrast with the background attenuation of lung parenchyma on im- ages obtained during expiration. CT protocol Satisfactory CT examinations were based on a qualitative visual assessment of lung parenchyma and the decreased Before CT scanning, patients were coached in the breathing cross-sectional area between maximum inspiration and maxi- technique by a radiology technologist to hold breaths at full mum expiration on CT. inspiration and full expiration. Other than verbal coaching, we Areas of air trapping were present when lung regions failed did not attempt to control the patient’s respiratory status dur- to increase in attenuation, as compared with findings on inspi- ing scanning. After carefully instructing breathing, the pa- ratory scan at or near the same level [13, 14]. All paired tients were scanned at the suspended end-inspiratory and inspiratory/expiratory CT scans were reviewed directly from end-expiratory volume from the lung apices through the lung the workstation by three thoracic radiologists aware that all bases. No IV contrast medium was injected. patients had proven COVID-19 infection but were otherwise Images at full inspiration and end-expiration without a blinded to other clinical information. spirometer-controlled imaging technique were acquired using All thin-section CT scans were reviewed at a window a multi-detector 16-slice scanner (Philips Brilliance CT 16 width and level of 1000–1500 HU and −500 to −650 HU, slice) with the following parameters: X-ray voltage, 120 respectively, for lung parenchyma, and 300–350 HU and kVp; tube current, auto exposure control (100–200 mAs); col- 20–50 HU, respectively, for soft tissue and mediastinum, by limation, 1 mm; high-speed mode; and pitch equivalent, 1.5. using image data that complied with the Digital Imaging and The scanning field of view ranged from 28 to 44 cm, based on Communications in Medicine standard. the subject’s body habitus. Exposure time was 0.53 s, and the The degree of air trapping within each lung zone was eval- matrix size was 768 × 768 pixels. Images were reconstructed uated by scoring the CT images using a 4-point scale based on with a 1.25-mm slice thickness (with 0.625 mm overlapping), visual assessment (Table 1). Each lung was divided into three using the “Bone” algorithm. For this sequence, patients were lung zones: upper (above the carina), middle (below the carina instructed to take a deep breath and hold it. After end- up to the inferior pulmonary vein), and lower (below the in- inspiratory scanning, the patients were coached with instruc- ferior pulmonary vein) zones. Each lung zone (total of six lung tions for the dynamic expiratory component of the study. For zones) was assigned a score as follows: score 0 (no air this sequence, patients were instructed to take a deep breath and blow it out during the image acquisition, which was co- ordinated with, to begin with, the patient’s forced expiratory effort. A reduced-dose technique was employed to minimize Table 1 System for scoring air trapping in expiratory thin-section CT radiation exposure for dynamic expiratory sequences. scans. Each lung was divided in three lung zones: upper, middle, and lower. The maximum score for each lung was 12 points. Both lungs had a global maximum score of 24 points Image review and scoring Score Definition The CT images in the 48 patients in the study cohort were 0 None reviewed, and findings were scored by three chest radiologists 1 1 – 25% of air trapping in each lung zone (T.F., A.G. and S.M., who had 20, 16, and 5 years of experi- 2 26 – 50% of air trapping in each lung zone ence in thoracic radiology, respectively) in consensus. 3 51 – 75% of air trapping in each lung zone Because the three observers did not independently review 4 > 75% of air trapping in each lung zone each case, interobserver agreement was not evaluated. Eur Radiol Table 2 Demographic, clinical, and laboratory data in patients Characteristic With ARDS (N = 14) Without ARDS (N = 34) p value with persistent dyspnea and cough after acute COVID-19 Age, median, year 66.0 (62.0 – 75.0) 55.5 (47.0 – 65.0) 0.0027 Male, n (%) 9 (64.3) 18 (52.9) 0.5361 Charlson comorbidity index 0.0 (0.0 – 0.0) 0.0 (0.0 – 0.0) 0.4631 Comorbidities, % 9 (64.3) 18 (52.9) 0.5361 Arterial hypertension, n (%) 6 (42.8) 10 (29.4) 0.5026 Diabetes mellitus, n (%) 0 (0) 3 (8.8) 0.5459 Other*, n (%) 3 (21.4) 3 (8.8) 0.4714 Interval onset-admission, days 7.5 (6.0 – 10.0) 7.0 (4.0 – 7.0) 0.1306 Symptoms during acute COVID-19 Fever, n (%) 14 (100) 34 (100) – Dyspnea, n (%) 5 (35.7) 25 (73.5) 0.0219 Anosmia/ageusia, n (%) 4 (28,6) 9 (26.5) 0.8349 Cough, n (%) 4 (28,6) 27 (79,4) 0.0019 Fatigue, n (%) 2 (14.3) 12 (35.3) 0.1812 Diarrhea, n (%) 2 (14.3) 13 (38.2) 0.0759 Arthromyalgia, n (%) 3 (21.4) 5 (14.7) 0.8870 Headache, n (%) 2 (14.3) 7 (20.6) 0.9184 Rash, n (%) 0 (0) 3 (8.8) – Abdominal pain, n (%) 2 (14.3) 0 (0) – O2 saturation, % 92.0 (92.0 – 93.0) 94.5 (92.0 – 98.0) 0.0151 PaO2/FiO2 ratio 173.0 (69.0 – 385.0) 304.5 (261.0 – 348.0) 0.0656 C-reactive protein, mg/L 226.5 (201.0 – 256.0) 97.7 (63.1 – 148.0) < 0.0001 AST, IU/l 54.0 (31.0 – 79.0) 51.5 (28.0 – 100.0) 0.5243 ALT, IU/l 66.0 (54.0 – 79.0) 53.0 (31.0 – 87.0) 0.3506 LDH, IU/l 541.0 (367.0 – 654.0) 299.0 (179.0 – 504.0) 0.0008 D dimer, ng/mL 633.0 (558.0 – 665.0) 560.0 (243.0 – 862.0) 0.2857 KL6, U/L 782.0 (138.0 – 941.7) 184.0 (144.7 – 331.0) 0.5715 IL-6, pg/mL 2094.0 (73.6 – 5573.0) 42.0 (18.0 – 116.0) 0.0103 Ferritin, ng/mL 2852.5 (1343.0 – 3771.0) 747.0 (336.0 – 1691.0) 0.0006 Neutrophil/lymphocyte ratio 7.95 (5.95 – 17.76) 2.9 (1.9 – 5.0) < 0.0001 Therapeutics Dexamethasone, n (%) 14 (100) 26 (76.5) – Tocilizumab, n (%) 10 (71.4) 15 (44.1) 0.1603 Other**, n (%) 13 (92.8) 18 (52.9) 0.0216 Air trapping score 12.5 (12.0 – 14.0) 10.0 (8.0 – 12.0) 0.0028 All values expressed as median and interquartile range (IQR) unless otherwise specified. * Includes dyslipidemia [4], epilepsia [1], and gout [1] ** Includes antibiotics [29], hemodialysis [2], and ECMO [2] ARDS adult respiratory distress syndrome, O2 oxygen, PaO2 partial pressure of oxygen in arterial blood, FiO2 fraction of inspired oxygen, AST aspartate aminotransferase, ALT alanine aminotransferase, LDH lactate dehy- drogenase, KL6 Krebs von den Lungen glycoprotein, IL-6 interleukin 6 Table 3 Comparison of thin- section CT parenchymal abnor- Findings* First CT (acute phase) Second CT (post-COVID) malities and air trapping at two n=9 n = 48 serial examinations in patients with proven COVID19 infection. Ground-glass opacities 7 (78) 19 (40) A patient may have more than one Mosaic attenuation pattern 1 (11) 9 (19) imaging finding Consolidation 6 (67) 2 (4) Traction bronchectasis 6 (67) 9 (19) Reticulation 0 (0) 18 (38) Parenchymal bands 3 (33) 15 (31) Bronchial wall thickening 3 (33) 6 (13) Air trapping 1✝ 37 (77) * Data are numbers of patients. Numbers in parentheses are percentages ✝ Only one patient had expiratory thin-section CT to assess air trapping Eur Radiol ƒFigure 1 A, B Transverse CT scans in a 47-year-old nonsmoking man who had COVID-19 obtained during inspiration at follow-up examina- tion, 70 days after dyspnea and shortness of breath, show minimal pe- ripheral ground-glass opacities (arrows); C, D expiratory transverse CT images at the same levels show linear consolidation in the subpleural region of the left lower lobe (arrowheads) and bilateral multifocal areas of low attenuation due to air trapping (arrows) that takes up the entire segments. Note that the same levels appear normal on the inspiratory images trapping score was obtained by summing six scores. The scores were added to provide a total CT severity score cumu- lative score ranging from 0 to 24. Thus, for each lung, the maximum possible score was 12 (three levels times four points at each level), and for both lungs, 24. Statistical analyses All the data were analyzed with statistical software (SPSS, version 11.0; SPSS). The Mann-Whitney test was used to analyze differences in ground-glass opacity, and air trap- ping scores between patient subgroups based on binary variables (male vs female sex, with vs. without ARDS and steroid therapy). Spearman rank correlation was per- formed to analyze the relationship between continuous variables (age, peak levels of CRP, LDH, and AST) and categoric variables (thin-section CT scores). All values were expressed as the median and interquartile range (IQR) unless otherwise indicated. A p value of less than .05 was considered to indicate a statistically significant difference for all test results. A multivariable logistic re- gression model was used to identify factors independently associated with air trapping. Any variable tested in uni- variate analysis with a p < 0.25 and all known clinical significance variables was selected as the first multivariate model candidates. We then followed the purposeful selec- tion of covariates method described by Hosmer et al [15]. Final parameter estimates are shown as odds ratios (ORs) with their corresponding 95% confidence intervals (CIs). Data were analyzed using IBM® SPSS®, statistics, ver- sion 26.0. Results Demographics, clinical, and laboratory data Forty-eight symptomatic patients were recruited during regularly scheduled follow-up visits. All the patients pre- sented with persistent cough and/or dyspnea or both at the trapping visible); score 1 (1–25% of the cross-sectional area of examination. There were 27 men and 21 women with a lung affected), score 2 (26–50% affected; score 3) (51–75% median age of 62 (IQR: 54–69) years (range: 31–79). affected), and score 4 (76–100% affected). The total air Twenty-seven patients (56.2%) had comorbidities Eur Radiol Figure 2 A 67-year-old nonsmoking man who had COVID-19. improve the degree of air trapping (arrows). D Inspiratory CT scan at Inspiratory CT and expiratory CT scans at follow-up examination, 50 the level of the lower lobes shows no abnormality. E Expiratory CT scan days after dyspnea and shortness of breath. A Inspiratory CT scan shows obtained at same level as D shows significant multifocal areas of low no abnormalities. B Expiratory CT scan obtained at same level as A attenuation due to air trapping in both lower lobes (arrows). F MiniP shows significant multifocal areas of low attenuation due to air trapping reformatted image also shows bilateral air trapping (arrows) in both upper lobes (arrows). C MiniP reformatted image is useful to predisposing to SARS-CoV-2 infection, the most common Factors associated with air trapping being arterial hypertension (16, 33.3%). The median Charlson index was 0.0 (IQR: 0.0–0.0). Fourteen patients Air trapping was gender-associated, total air trapping score in (29.2%) developed adult distress respiratory syndrome men 12.0 (10.0–140) vs. 10.0 (0.0.0–10.0) in women (p = (ARDS). Demographics, clinical, and laboratory data are 0.0026). It was unassociated with comorbidities. Acute summarized in Table 2. COVID-19 clinical and laboratory parameters associated with air trapping are summarized in Table 4. The interval between the onset of symptoms and admission (r = 0.41, p = 0.0040), CT findings cough (p = 0.0037), and ARDS (p = 0.0232) was also associ- ated with the degree of air trapping. The degree of air trapping Follow-up paired inspiratory/expiratory CT findings in the correlated with age (r = 0.62, p < 0.0001), admission PaO2/ 48 patients are listed in Table 3. Parenchymal abnormality FiO2 ratio (r = 0.39, p = 0.0054), peak CRP (r = 0.34, p = was found in 50% (24/48) of patients and included air 0.0171), and ferritin (r = 0.36, p = 0.0109). Charlson comor- trapping (37/48, 77%), ground-glass opacities (19/48, bidity index, initial oxygen saturation, D-dimer, LDH, AST, 40%), reticulation (18/48, 38%), parenchymal bands (15/ ALT, KL-6, IL-6, neutrophil and lymphocyte counts, and 48, 31%), traction bronchiectasis (9/48, 19%), mosaic at- neutrophil-to-lymphocyte ratio did not correlate with the de- tenuation pattern (9/48, 19%), bronchial wall thickening gree of air trapping. A multivariable logistic regression analysis (6/48, 13%), and consolidation (2/48, 4%). The absence was performed taking air trapping score as the dependent var- of air trapping was observed in 11/48 (23%), mild air iable and age, sex, the interval between symptom onset and trapping in 20/48 (42%), moderate in 13/48 (27%), and admission, cough, PaO2/FiO2 ratio, peak CRP, LDH, D-dimer, severe in 4/48 (8%) (Figs. 1, 2 and 3). The median air IL-6, ferritin, and neutrophil-to-lymphocyte ratio as indepen- trapping score was 11.0 (IQR: 10.0–13.0). Pleural thick- dent variables. Variables independently associated with the de- ening, nodules, round cystic changes, bronchiectasis, gree of air trapping were gender (OR = 14.92; 95% CI: 1.99– pleural effusion, and lymphadenopathy were not ob- 111.71, p = 0.0085), and age (OR = 1.14; 95% CI: 1.02–1.26, p served. Air trapping was associated with ground-glass = 0.0182). Recently, immune-endocrine processes have been opacities (p = 0.0223) and reticulation (p = 0.0034) but proposed for a better understanding in the variations of SARS- not with traction bronchiectasis (p=0.0946), parenchymal CoV-2 pathogenicity. In fact, females mount higher innate im- bands (p = 0.1359), mosaic attenuation pattern mune responses than males, resulting in a faster viral recogni- (p=0.0946), consolidation (p = 1.0), and bronchial wall tion and production of interferon and inflammatory cytokines, thickening (p = 0.3132). inducing more rapid viral clearance [16]. Eur Radiol ƒFigure 3 A 67-year-old nonsmoking man who had COVID-19. A Anteroposterior chest radiograph shows bilateral and multifocal areas of consolidation (arrows). B Inspiratory CT scan at follow-up examination, 65 days after dyspnea and shortness of breath shows no abnormalities. C Expiratory CT scan, obtained at same level as B, and D coronal multiplanar reformation image shows significant bilateral patchy areas of low attenuation due to air trapping (arrows) Discussion Our study shows a high prevalence of air trapping in post- COVID patients with respiratory symptoms. Besides, it sug- gests small-airway disease as its cause and highlights the use- fulness of paired inspiration/expiration CT when this condi- tion is entertained. Among patients who recovered from COVID-19 prolonged symptoms, particularly fatigue and dyspnea are common, and it may take weeks or months for their resolution and return to usual health [2–5]. Recognizing imaging patterns in the follow- up of long-COVID patients is paramount for understanding the pathophysiologic features and natural history of this infection. Long-COVID symptoms could be related to several results of lung damage including [1] micro-angiopathy, [2] fibrosis, [3] obstructive ventilatory defects, and [4] chronic local and systemic inflammation. Recent COVID-19 follow-up studies have evaluated the presence of microangiopathy and fibrosis [17, 18]. The findings of air trapping in our study suggest that obstructive ventilatory defects and small airway inflammation may play a role in long-COVID pathophysiology. COVID-19 is characterized by an exuberant inflammatory response caused by overwhelming cytokine release that may predispose patients to thrombotic disease because of excessive inflammation, platelet activation, endothelial dysfunction, and stasis. Among patients with COVID-19, Barnes et al identified the abundant neutrophils and neutrophil extracellular traps (NETs) in the airway and alveolar air spaces [19]. The most noteworthy finding from this study was the discovery of active NETosis in the infected lungs with the persistent release of NETs, observed in bronchiolar and alveolar regions from dam- aged areas of the lungs even after the virus had cleared [19]. NETs formed in the lungs infected with respiratory viruses represent potential COVID-19 inflammatory lung damage drivers, including small airways, interstitial lesions, and throm- bosis leading to fibrin deposition. Air trapping might be attrib- uted either to the capillary wall damage and thickening of small airways walls caused by pro-inflammatory factors [12, 20]. Evaluation of patients following novel coronaviruses, SARS [21], and recently in COVID survivors, have not shown physiologically limiting airflow obstruction. Instead, lung function test (LFT) abnormalities were mainly related to the pulmonary diffusing capacity for carbon monoxide [22, 23]. This is not surprising considering our previous work showing that the extent of air trapping did not correlate with LFT ab- normalities, possibly representing bronchiolar inflammation Eur Radiol Table 4 Demographic, clinical, and laboratory data in patients Characteristic With air trapping Without air trapping p value with and without air trapping (N = 37) (N = 11) Age, year 65.0 (56.0 – 74.0) 47.0 (38.0 – 56.2) 0.0005 Male, n (%) 31 (83.8) 6 (54.5) < 0.0001 Charlson comorbidity index 0.0 (0.0 – 1.0) 0.0 (0.0 – 0.0) 0.0328 Comorbidities, % 29 (78,4) 2 (18,2) 0.0005 Arterial hypertension, n (%) 20 (54.0) 0 (0.0) 0.0012 Diabetes mellitus, n (%) 4 (10.8) 0 (0.0) 0.5607 Other*, n (%) 3 (8.1) 2 (18.2) Interval onset-admission, days 6.0 (4.0 – 7.0) 7.0 (6.0 – 9.5) 0.0500 Symptoms during acute COVID-19 Fever, n (%) 37 (100) 11 (100) – Dyspnea, n (%) 22 (59.4) 7 (63.6) 0.9184 Anosmia/ageusia, n (%) 10 (27.0) 2 (18.2) 0.7054 Cough, n (%) 19 (51.3) 10 (91.8) 0.0324 Fatigue, n (%) 7 (18.9) 4 (36.3) 0.2457 Diarrhea, n (%) 11 (29.7) 1 (9.1) 0.2475 Arthromyalgia, n (%) 2 (5.4) 4 (36.4) 0.0193 Headache, n (%) 7 (18.9) 3 (27.3) 0.6753 Rash, n (%) 0 (0) 2 (18.2) – Abdominal pain, n (%) 2 (5.4) 1 (9.1) 0.5507 ARDS, n (%) 13 (35.1) 1 (9.1) 0.1388 O2 saturation, % 93.0 (92.0 – 96.0) 93.0 (92.0 – 97.0 0.9309 PaO2/FiO2 ratio 261.0 (176 – 317) 328.0 (285.0 – 414.0) 0.0305 C-reactive protein, mg/L 167.0 (124.3 – 229.0) 81.4(63.1 – 96.3) 0.0024 AST, IU/l 53.0 (31.0 – 118.0) 47.0 (19.7 – 55.0) 0.1046 ALT, IU/l 73.0 (31.0 – 126.0) 61.0 (32.0 – 65.0) 0.4385 LDH, IU/l 367.0 (282.0 – 471.7) 233.0 (148.7 – 504.0) 0.1073 D dimer, ng/mL 638.0 (310.0 – 1080.0) 560.0 (243.0 – 635.2) 0.0749 KL6, U/L 329.5 (141.0 – 548.0) 150.0 (143.0 – 184.0) 0.1461 IL-6, pg/mL 1041.0 (116.0 – 1966.0) 39.1 (19.0 – 112.0) 0.0673 Ferritin, ng/mL 1343.0 (686.0 – 3373.0) 476.0 (104.0 – 769.5) 0.0016 Neutrophil/lymphocyte ratio 5.03 (2.99 – 7.61) 2.21 (1.86 – 5.05) 0.0179 Therapeutics Dexamethasone, n (%) 31 (83.8) 9 (81.8) 0.7587 Tocilizumab, n (%) 15 (40.5) 5 (45,4) 0.9537 Other**, n (%) 27 (72.5) 5 (45.4) 0.1440 All values expressed as median and interquartile range (IQR) unless otherwise specified. * Includes dyslipidemia [3], epilepsia [1], and gout [1] ** Includes antibiotics [29], hemodialysis [2], and ECMO [2] ARDS adult respiratory distress syndrome, IQR interquartile range, O2 oxygen, PaO2 partial pressure of oxygen in arterial blood, FiO2 fraction of inspired oxygen, AST aspartate aminotransferase, ALT alanine aminotransferase, LDH lactate dehydrogenase, KL6 Krebs von den Lungen glycoprotein, IL-6 interleukin -6 preceding detectable LFT abnormalities [24]. Regional abnor- Notwithstanding we cannot thoroughly rule out previous malities in lung ventilation could, however, have a profound bronchiolar damage, the fact that the patients had not any effect on ventilation perfusion mismatch in the post COVID pulmonary disease prior to COVID-19 suggest that the bron- lung, particularly given the accumulating evidence for wide- chiolar involvement was of inflammatory origin. spread pulmonary vasculopathy in COVID-19 infections [17]. In general, underventilation manifesting as air trapping is Increased perfusion of underventilated lung may play a major an important early indicator of small-airway disease (SAD) role in driving COVID-19-related hypoxia [25]. [26] but limited histological data are available regarding Eur Radiol clinical long-COVID-19-related bronchiolitis [27–29]. The COVID19-PoC (to PD), Departament de Salut and BIOCET, Generalitat de Catalunya. P. D. is supported by a grant from the Programa de association of air trapping with inflammatory biomarkers sug- Intensificación de Investigadores (INT19/00036)-ISCIII. gests that SAD may be a manifestation of an inflammatory Role of funding sources: None state. For instance, ferritin, an important biomarker of macro- phage activation, positively correlates with air trapping, in- Declarations flammation, and lung immune-mediated injury in COVID- 19 [30]. Guarantor The scientific guarantor of this publication is Tomás We have observed air trapping using expiratory CT scans in Franquet MD, PhD. 77% of our long-COVID patients. Although air trapping on Conflict of interest The authors of this manuscript declare no relation- expiratory CT has been observed in long-COVID patients ships with any companies whose products or services may be related to [31–34], it has not been systematically studied in the setting of the subject matter of the article. long COVID with persistent respiratory complaints. Recently, a multicenter study of 108 patients with COVID-19 evaluated the Statistics and biometry One of the authors have significant statistical small-airway disease using inspiratory and expiratory chest expertise. high-resolution computed tomography. The study detected air Informed consent Written informed consent was obtained from all pa- trapping following COVID-19 infection but did not show that tients in this study its presence was associated with the interval since infection [32]. Considering the above, post-infectious SAD is a consider- Ethical approval Institutional Review Board approval was obtained. ation among patients with SARS-CoV-2 infection presenting with hypoxemia and mosaic perfusion. Although it is not pos- Methodology sible to exclude the possibility that some air trapping is attrib- • Prospective • Observational utable to chronic pulmonary embolism [35], we believe that the • Performed at one institution presence of significant air trapping in these patients probably reflects the presence of persistent bronchiolar inflammation or Open Access This article is licensed under a Creative Commons may correspond to a constrictive bronchiolitis as a sequelae. Attribution 4.0 International License, which permits use, sharing, adap- tation, distribution and reproduction in any medium or format, as long as In our group of patients, we did not observe enlargement of you give appropriate credit to the original author(s) and the source, pro- pulmonary trunk or right-side cardiac cavities. vide a link to the Creative Commons licence, and indicate if changes were Although it is not always feasible due to the unstable clin- made. The images or other third party material in this article are included ical conditions, especially in critically ill patients, the most in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's widely used imaging technique for assessing air trapping is Creative Commons licence and your intended use is not permitted by paired inspiratory and expiratory CT scans. A low-dose CT statutory regulation or exceeds the permitted use, you will need to obtain protocol with iterative reconstruction can be potentially rec- permission directly from the copyright holder. To view a copy of this ommended for the follow-up evaluation of symptomatic long- licence, visit http://creativecommons.org/licenses/by/4.0/. COVID patients. Our study had several limitations. Firstly, the present study was a retrospective study with small sample size, and future prospective studies with a larger number of patients would be References helpful to confirm our results. Secondly, we did not use 1. Chung M, Bernheim A, Mei X et al (2020) CT imaging features of spirometrically gated inspiratory and expiratory CT, and air 2019 novel coronavirus (2019-nCoV). Radiology. 295(1):202–207. trapping was evaluated by using subjective analysis. https://doi.org/10.1148/radiol.2020200230 Future automated quantification CT will be extremely useful 2. Carfi A, Bernabei R, Landi F, Gemelli Against COVID-19 Post- in evaluating an increasing population of post-COVID-19 lung Acute Care Study Group (2020) Persistent symptoms in patients after acute COVID-19. JAMA. 324(6):603–605. https://doi.org/10. parenchymal changes such as fibrosis or air trapping [36]. 1001/jama.2020.12603 In summary, our study suggest that air trapping is a com- 3. Mahase E (2020) Covid-19: what do we know about "long covid"? mon finding in long-COVID patients with respiratory symp- BMJ. 370:m2815. https://doi.org/10.1136/bmj.m2815 toms. The presence of air trapping on dynamic expiratory CT 4. Mendelson M, Nel J, Blumberg L et al (2020) Long-COVID: an evolving problem with an extensive impact. S Afr Med J 111(1): scans is highly suggestive of obstructive SAD and could re- 10–12. https://doi.org/10.7196/SAMJ.2020.v111i11.15433 flect the presence of persistent bronchiolar inflammation. 5. Tenforde MW, Kim SS, Lindsell CJ et al (2020) Symptom duration and risk factors for delayed return to usual health among outpatients Acknowledgements The authors thank Mónica Carrasco, radiologic with COVID-19 in a multistate health care systems network - technician. United States, March-June 2020. MMWR Morb Mortal Wkly Rep 69(30):993–998. https://doi.org/10.15585/mmwr.mm6930e1 Funding This study was partially supported by the grant COV20/00070 6. Ojha V, Mani A, Pandey NN, Sharma S, Kumar S (2020) CT in (to PD), Instituto de Salud Carlos III, Madrid, Spain, and by the grant coronavirus disease 2019 (COVID-19): a systematic review of Eur Radiol chest CT findings in 4410 adult patients. Eur Radiol 30(11):6129– 23. Zhao YM, Shang YM, Song WB et al (2020) Follow-up study of 6138. https://doi.org/10.1007/s00330-020-06975-7 the pulmonary function and related physiological characteristics of 7. Wan S, Li M, Ye Z et al (2020) CT manifestations and clinical COVID-19 survivors three months after recovery. characteristics of 1115 patients with coronavirus disease 2019 EClinicalMedicine 25:100463. https://doi.org/10.1016/j.eclinm. (COVID-19): a systematic review and meta-analysis. Acad Radiol 2020.100463 27(7):910–921. https://doi.org/10.1016/j.acra.2020.04.033 24. Franquet T, Diaz C, Domingo P, Gimenez A, Geli C (1999) Air 8. Franquet T, Jeong YJ, Lam HYS et al (2020) Imaging findings in trapping in primary Sjogren syndrome: correlation of expiratory CT coronavirus infections: SARS-CoV, MERS-CoV, and SARS-CoV- with pulmonary function tests. J Comput Assist Tomogr 23(2): 2. Br J Radiol 93(1112):20200515. https://doi.org/10.1259/bjr. 169–173. https://doi.org/10.1097/00004728-199903000-00002 20200515 25. Santamarina MG, Boisier D, Contreras R, Baque M, Volpacchio M, 9. Adams HJA, Kwee TC, Yakar D, Hope MD, Kwee RM (2020) Beddings I (2020) COVID-19: a hypothesis regarding the Chest CT imaging signature of coronavirus disease 2019 infection: ventilation-perfusion mismatch. Crit Care 24(1):395. https://doi. in pursuit of the scientific evidence. Chest. 158(5):1885–1895. org/10.1186/s13054-020-03125-9 https://doi.org/10.1016/j.chest.2020.06.025 26. Tanaka N, Matsumoto T, Miura G et al (2003) Air trapping at CT: 10. Goyal N, Chung M, Bernheim A et al (2020) Computed tomogra- high prevalence in asymptomatic subjects with normal pulmonary phy features of coronavirus disease 2019 (COVID-19): a review for function. Radiology. 227(3):776–785. https://doi.org/10.1148/ radiologists. J Thorac Imaging 35(4):211–218. https://doi.org/10. radiol.2273020352 1097/RTI.0000000000000527 27. De Michele S, Sun Y, Yilmaz MM et al (2020) Forty postmortem 11. Ng MY, Lee EYP, Yang J et al (2020) Imaging profile of the examinations in COVID-19 patients. Am J Clin Pathol 154(6):748– COVID-19 infection: radiologic findings and literature review. 760. https://doi.org/10.1093/ajcp/aqaa156 Radiol Cardiothorac Imaging 2(1):e200034. https://doi.org/10. 28. Li H, Liu L, Zhang D et al (2020) SARS-CoV-2 and viral sepsis: 1148/ryct.2020200034 observations and hypotheses. Lancet. 395(10235):1517–1520. 12. Arakawa H, Webb WR (1998) Air trapping on expiratory high- https://doi.org/10.1016/S0140-6736(20)30920-X resolution CT scans in the absence of inspiratory scan abnormali- ties: correlation with pulmonary function tests and differential di- 29. Vasquez-Bonilla WO, Orozco R, Argueta V et al (2020) A review agnosis. AJR Am J Roentgenol 170(5):1349–1353. https://doi.org/ of the main histopathological findings in coronavirus disease 2019. 10.2214/ajr.170.5.9574614 Hum Pathol 105:74–83. https://doi.org/10.1016/j.humpath.2020. 13. Hansell DM, Bankier AA, MacMahon H, McLoud TC, Muller NL, 07.023 Remy J (2008) Fleischner Society: glossary of terms for thoracic 30. Domingo P, Mur I, Pomar V, Corominas H, Casademont J, de Benito imaging. Radiology. 246(3):697–722. https://doi.org/10.1148/ N (2020) The four horsemen of a viral Apocalypse: the pathogenesis radiol.2462070712 of SARS-CoV-2 infection (COVID-19). EBioMedicine. 58:102887. 14. Stern EJ, Webb WR (1993) Dynamic imaging of lung morphology https://doi.org/10.1016/j.ebiom.2020.102887 with ultrafast high-resolution computed tomography. J Thorac 31. Alarcon-Rodriguez J, Fernandez-Velilla M, Urena-Vacas A et al Imaging 8(4):273–282. https://doi.org/10.1097/00005382- (2021) Radiological management and follow-up of post-COVID- 199323000-00003 19 patients. Radiologia (Engl Ed) 63(3):258–269. https://doi.org/ 15. Hosmer DW, Lemeshow S, Sturdivant R (2000) Model-building 10.1016/j.rx.2021.02.003 strategies methods for logistic regression. In: Hosmer DW, 32. Huang R, Zhu J, Zhou J et al (2021) Inspiratory and expiratory chest Lemeshow S, Sturdivant R (eds) Applied Logistic Regression high-resolution CT: small-airway disease evaluation in patients Hoboken. John Wiley & Sons, Inc. 2000 ed, NJ, USA, pp 89–152 with COVID-19. Curr Med Imaging. https://doi.org/10.2174/ 16. Wehbe Z, Hammoud SH, Yassine HM, Fardoun M, El-Yazbi AF, 1573405617999210112194621 Eid AH (2021) Molecular and biological mechanisms underlying 33. Guler SA, Ebner L, Aubry-Beigelman C et al (2021) Pulmonary gender differences in COVID-19 severity and mortality. Front function and radiological features 4 months after COVID-19: first Immunol 12:659339. https://doi.org/10.3389/fimmu.2021.659339 results from the national prospective observational Swiss COVID- 17. Han X, Fan Y, Alwalid O et al (2021) Six-month follow-up chest 19 lung study. Eur Respir J 57(4). https://doi.org/10.1183/ CT findings after severe COVID-19 pneumonia. Radiology. 13993003.03690-2020 299(1):E177–EE86. https://doi.org/10.1148/radiol.2021203153 34. Ebner L, Funke-Chambour M, von Garnier C, Ferretti G, Ghaye B, 18. Lang M, Som A, Carey D et al (2020) Pulmonary vascular mani- Beigelman-Aubry C (2020) Imaging in the aftermath of COVID- festations of COVID-19 pneumonia. Radiol Cardiothorac Imaging 19: what to expect. Eur Radiol. https://doi.org/10.1007/s00330- 2(3):e200277. https://doi.org/10.1148/ryct.2020200277 020-07465-6 19. Barnes BJ, Adrover JM, Baxter-Stoltzfus A et al (2020) Targeting 35. Arakawa H, Stern EJ, Nakamoto T, Fujioka M, Kaneko N, potential drivers of COVID-19: neutrophil extracellular traps. J Exp Harasawa H (2003) Chronic pulmonary thromboembolism. Air Med 217(6):e20200652. https://doi.org/10.1084/jem.20200652 trapping on computed tomography and correlation with pulmonary 20. Hansell DM (2010) Thin-section CT of the lungs: the Hinterland of function tests. J Comput Assist Tomogr 27(5):735–742. https://doi. normal. Radiology. 256(3):695–711. https://doi.org/10.1148/ org/10.1097/00004728-200309000-00010 radiol.10092307 36. Nagpal P, Guo J, Shin KM et al (2021) Quantitative CT imaging 21. Hui DS, Joynt GM, Wong KT et al (2005) Impact of severe acute and advanced visualization methods: potential application in novel respiratory syndrome (SARS) on pulmonary function, functional coronavirus disease 2019 (COVID-19) pneumonia. BJR Open 3(1): capacity and quality of life in a cohort of survivors. Thorax. 20200043. https://doi.org/10.1259/bjro.20200043 60(5):401–409. https://doi.org/10.1136/thx.2004.030205 22. Wu Q, Zhong L, Li H et al (2021) A follow-up study of lung function and chest computed tomography at 6 months after dis- Publisher’s note Springer Nature remains neutral with regard to jurisdic- charge in patients with coronavirus disease 2019. Can Respir J tional claims in published maps and institutional affiliations. 2021:6692409. https://doi.org/10.1155/2021/6692409