Advertisement

Subpleural sparing: Clinical, physiological, and radiological implications

Published:November 22, 2022DOI:https://doi.org/10.1016/j.amjms.2022.11.002

      Abstract

      The term “subpleural sparing” refers to computed tomography (CT) images that indicate that there is limited disease/infiltrate in the immediate subpleural location. This observation is often associated with nonspecific interstitial pneumonitis and is a characteristic that distinguishes this pathology from usual interstitial pneumonitis (idiopathic pulmonary fibrosis). Subpleural sparing can also occur in acute respiratory disorders, including pulmonary contusion in children, acute lung disease associated with electronic cigarettes (vaping), and aspiration of exogenous lipids. Potential explanations for this observation include nonuniform distribution of lung injury/inflammation, nonuniform clearing/resolution of injury, and variations in CT image acquisition and presentation. The subpleural region contains lymphatic structures on the interior surface of the visceral pleura and in interlobular septa. The density of subpleural lymphatics decreases in more interior zones of the lung that largely contain alveolar-capillary units. These lymphatics transfer fluid and other inflammatory mediators from the peripheral lung into central lymphatics and veins. Consequently, the density and distribution of lymphatics could explain preferential clearing of the subpleural regions during acute injury. The acquisition of CT images also depends on the configuration of detectors, slice thickness, and the energy of the electron beam. Clinicians should carefully consider the disease process, lymphatic function and other clearance mechanisms, and the vagaries in CT image acquisition when they evaluate patients with subpleural sparing.

      Keywords

      Introduction

      The observation that some patients with parenchymal lung disease have subpleural sparing on computed tomography (CT) may provide important information needed for developing a differential diagnosis.
      • Chong WH
      • Saha BK
      • Austin A
      • Chopra A.
      The significance of subpleural sparing in ct chest: a state-of-the-art review.
      For example, patients with vaping-induced diffuse lung disease often have subpleural sparing. This diagnosis is relatively uncommon, and important historical information may be overlooked. However, the presence of subpleural sparing may lead the radiologist and clinician to consider this possible diagnosis. In addition, the presence of subpleural sparing in a patient with chronic interstitial lung disease would support the diagnosis of nonspecific interstitial pneumonitis and might reduce the need for an open lung biopsy.
      • Chong WH
      • Saha BK
      • Austin A
      • Chopra A.
      The significance of subpleural sparing in ct chest: a state-of-the-art review.
      Finally, the presence of subpleural sparing raises several important questions regarding pathogenesis. Does this observation reflect differences in lung injury in various regions, differences in lung clearance mechanisms in various regions, differences in the three-dimensional structure at the edge of the lung, or differences in image acquisition or presentation? In addition, does subpleural sparing have any implications with regard to lung pathology or patient prognosis? These considerations require review of lung anatomy in peripheral zones and of the effect of the respiratory cycle on lymphatic and venous blood flow from the peripheral regions of the lung into the central chest.

      Definition

      The phrase “subpleural sparing” is a descriptive term used to characterize thoracic images of patients who have parenchymal lung disease in which there is more disease deep within the lung tissue away from the thoracic wall and less disease close to the thoracic wall. There is no definition or criteria for this term from authoritative sources. This term was not discussed in the Fleischner Society publication Glossary of Terms for Thoracic Imaging and was not considered in the recent publication on the interpretation diffuse lung disease on chest CT imaging with a subtitle A theory of almost everything.
      • Hansell DM
      • Bankier AA
      • MacMahon H
      • et al.
      Fleischner Society: glossary of terms for thoracic imaging.
      ,
      • Gruden JF
      • Naidich DP
      • Machnicki SC
      • et al.
      An algorithmic approach to the interpretation of diffuse lung disease on chest ct imaging: a theory of almost everything.
      Pertinent considerations include location of the spared region, width of the spared region, length of the spared region, and vertical height of the spared region. These considerations should provide an estimate of the volume of the spared region. In addition, this image appearance should be evident on axial images, sagittal images, and coronal images, depending on the location.

      Diagnoses

      Acute lung diseases

      Panse et al. analyzed the radiological and pathologic findings in 24 patients with e-cigarette or vaping product use associated lung injury.
      • Panse PM
      • Feller FF
      • Butt YM
      • et al.
      Radiologic and pathologic correlation in EVALI.
      The most common CT finding was ground glass opacities (23 patients, 96%) followed by consolidation (10 patients, 42%). The distribution was multifocal in 13 patients (54%); 9 patients (45%) had relative sparing of the lung immediately adjacent to the pleura. The authors did not discuss possible explanations for subpleural sparing in these cases. Possibilities could include nonuniform distribution of inhaled gas into different regions of the lung, resulting in nonuniform injury. Also, relative clearing of one region compared to others might suggest that those regions have more efficient clearance processes related to ongoing lung injury.
      Donnelly and Klosterman reviewed the CT findings in children who had trauma with lung contusion.
      • Donnelly LF
      • Klosterman LA.
      Subpleural sparing: a CT finding of lung contusion in children.
      Subpleural sparing was present in 38 of 40 lung contusions. They suggested that decreased vascularity in the peripheral regions of the lung could reduce hemorrhage into these regions following contusion and that compression of the lung against the chest wall during injury could “squeeze” extravasated blood and fluid into more central zones of the lung. In addition, compression of the peripheral zone of the lung could cause atelectasis in this region. This process could persist for an unknown length of time, especially if the child has chest pain in that region. This, in turn, might reduce blood entry into that zone, which then appears normal following reexpansion.
      Gondouin et al. retrospectively reviewed 44 cases of exogenous lipid pneumonia in France.
      • Gondouin A
      • Manzoni P
      • Ranfaing E
      • et al.
      Exogenous lipid pneumonia: a retrospective multicentre study of 44 cases in France.
      Thirty of these cases were related to aspiration of liquid paraffin used for the treatment of constipation. Radiographic studies revealed alveolar consolidation, ground glass opacities, and alveolar nodules. Most of the opacities were bilateral and predominantly in posterior and lower zones. On 16 CT scans (out of 31 scans), there was relative sparing of the peripheral regions of the lung. The possible explanations for these radiographic findings were not discussed.
      Imai et al. analyzed the CT scans of 144 patients with ARDS. The mean age was 72; the in-hospital mortality was 42%.
      • Imai R
      • Nishimura N
      • Takahashi O
      • Tamura T.
      High-resolution computed tomography for the prediction of mortality in acute respiratory distress syndrome: a retrospective cohort study.
      The CT scans were classified by the patterns found in 6 slices of the lung taken at an upper level of the thorax, at a mid-level, and a basal level. A diffuse infiltrate was most frequent pattern and was found in 3.1 ± 2.1 slices in survivors and 4.4 ± 1.7 slices in non-survivors. Subpleural sparing patterns occurred in 1.5 ± 0 .8 slices in survivors and 0.5 ± 1.1 slices in non-survivors. The diffuse pattern was associated with in-hospital mortality. The time period between the development of ARDS and the CT study was not reported in this study. This study demonstrates that subpleural sparing is not a characteristic radiographic pattern in patients with ARDS. These authors suggested subpleural sparing may occur in patients with edema formation that preferentially moves out of the subpleural zone.
      Demirci et al. reviewed the CT scans of 1563 patients hospitalized with COVID-19 infection.
      • Demirci NY DA
      • Tasci C
      • et al.
      Relationship between chest computed tomography findings and clinical conditions of coronavirus disease (COVID-19): a multicentre experience.
      These scans were done the day of the initial evaluation of the patient. The majority of patients (1134) patients had CT abnormalities consistent with viral infection.; 981 patients (83.6%) had areas of consolidation and 909 patients (76.2%) had areas of subpleural sparing. Based on multivariable analysis, the presence of the consolidation but not subpleural sparing was associated with severe COVID-19 infection defined by abnormal respiratory parameters, including tachypnea (>30 breaths per minute), O2 saturation ≤ 93% on room air, PO2/FiO2 ratio less than 300, or requirements for mechanical ventilation.
      These five studies in patients with acute lung injury demonstrate that subpleural sparing occurs in situations that have clearly different mechanisms of injury. This radiographic pattern developed following inhalation of smoke with particulates, direct trauma to the chest, aspiration, and viral infection but not diffuse lung injury associated with ARDS. There is no obvious reason that these diverse injury events would spare the subpleural region of the lung. Since the clinical consequences and radiographic representation of lung injury reflect the location and type of injury and the lung response to injury, a possible explanation for subpleural sparing in these syndromes could involve faster or more complete tissue repair in lung regions near the pleural surface.

      Chronic lung disease

      Silva et al. retrospectively reviewed the CT features of 66 patients to determine if these features differentiated non-specific interstitial pneumonia (NSIP), idiopathic pulmonary fibrosis (IPF), and chronic hypersensitivity pneumonitis (HP).
      • Silva CI
      • Müller NL
      • Hansell DM
      • et al.
      Nonspecific interstitial pneumonia and idiopathic pulmonary fibrosis: changes in pattern and distribution of disease over time.
      The features that best differentiated NSIP (n=25) were relative subpleural sparing (seen in 16 patients), absence of lobular areas with decreased attenuation, and lack of honeycombing. Features differentiating IPF were basal predominance of honeycombing, the absence of relative subpleural sparing (seen only in 1 patient), and the absence of centrilobular nodules. Chronic HP was characterized by presence of lobular areas with decreased attenuation and vascularity, centrilobular nodules, and absence of lower zone predominance of abnormalities.
      Ebina et al. studied surgical and autopsy specimens from 18 patients with IPF, 6 patients with organizing pneumonia, and 6 patients with cellular NSIP.
      • Ebina M
      • Shibata N
      • Ohta H
      • et al.
      The disappearance of subpleural and interlobular lymphatics in idiopathic pulmonary fibrosis.
      In the lung tissues of patients with IPF, the lymphatics in the subpleural and paraseptal parenchyma were significantly reduced and fragmented when compared to the lymphatics in control lungs, cellular NSIP, or organizing pneumonia. In contrast, tissues of patients with NSIP and organizing pneumonia contained dilated lymphatics in the intralobular septa and subpleural tissues with no significant differences when compared to control lungs. In addition, multiple new lymphatics were extending from the alveolar lesions to existing lymphatics in the interlobular or subpleural tissues in both NSIP and organizing pneumonia. This particular finding was infrequent in lungs of patients with IPF.
      In contrast, Parra et al. investigated the distribution of lymphatics in different remodeling stages of interstitial pneumonias.
      • Parra ER
      • Araujo CA
      • Lombardi JG
      • et al.
      Lymphatic fluctuation in the parenchymal remodeling stage of acute interstitial pneumonia, organizing pneumonia, nonspecific interstitial pneumonia and idiopathic pulmonary fibrosis.
      In organizing pneumonia (OP), few lymphatics were observed. They were tortuous with minimal dilatation. In the NSIP pattern, also a few lymphatics were observed, but they were more dilated when compared to OP. In the UIP/IPF pattern, lymphatic vessels presented a large luminal area, especially in the periphery and around fibroblastic foci.
      The last two studies report obvious differences in the lymphatics in lung tissue with chronic lung disease. This is hard to explain, but most studies do not have uniform diagnostic criteria for different chronic interstitial pneumonias and a limited number of specimens are used in each study. Ebina et al. used autopsy specimens and surgical specimens; Parra et al. used only surgical specimens. Subpleural sparing is a common finding in some chronic lung diseases, especially NSIP, and lymphatic clearance and distribution probably contributes to this phenomenon.
      • Johkoh T.
      Nonspecific interstitial pneumonia and usual interstitial pneumonia: is differentiation possible by high-resolution computed tomography?.
      However, the different results with regard to lymphatic distribution and lymphangiogenesis makes this hard to prove.
      Chong et al. have published a comprehensive review of the clinical disorders and radiographic patterns in patients with subpleural sparing on CT scans.
      • Chong WH
      • Saha BK
      • Austin A
      • Chopra A.
      The significance of subpleural sparing in ct chest: a state-of-the-art review.

      Mechanisms

      The development of subpleural sparing on chest radiographs should reflect differential injury to regions of the lung, different repair responses in regions of the lung, or a combination of these two factors. The mechanisms of lung injury and repair become critically important in understanding whether or not the injury event contributes to differential involvement of the lung parenchyma.

      Ventilation

      Ventilation should uniformly distribute toxic gases, such as the chemicals associated with vaping-induced acute lung injury, throughout the lung in proportion to the distribution of the tidal volume. Delivery of gases to the very peripheral zone to the lung largely reflects diffusion in alveolar spaces. There is no obvious reason to think that there is less diffusion in zones near the pleural surface unless the inspiratory phase of the respiratory cycle is extremely short and does not allow enough time for adequate diffusion. One potential explanation might involve the absorption of the toxic compound to lung surfaces during the inspiratory phase and consequently a steady decrease in concentration as the tidal volume moves into the lung and alveolar spaces. This possibility would require animal studies to confirm this. Ventilation also creates cyclical changes in intrathoracic pressure, which influences both lymphatic and venous flow.
      • Soni N
      • Williams P.
      Positive pressure ventilation: what is the real cost?.
      This will be discussed in a later paragraph.

      Blood flow

      Blood flow through the lung in the pulmonary arterial system could contribute to the development of lung contusion but should not contribute to the other types of injury discussed above. The resolution and removal of inflammatory processes, e.g., cellular elements and cytokines, in the lung parenchyma through the pulmonary venous system depends on the distribution of veins in subpleural tissue and in inter-alveolar septa.
      • O'Hagan LA
      • Windsor JA
      • Itkin M
      • et al.
      The lymphovenous junction of the thoracic duct: a systematic review of its structural and functional anatomy.
      ,
      • Ratnayake CBB
      • Escott ABJ
      • et al.
      The anatomy and physiology of the terminal thoracic duct and ostial valve in health and disease: potential implications for intervention.
      The pressure gradients resulting in venous blood flow in the lung depend on cyclical changes in intrapleural pressure, which results in reduced pressure in the central pulmonary veins and blood return to the left atrium. Consequently, larger pressure swings secondary to respiratory distress, abnormal lung mechanics, or exercise should increase venous return into the systemic circulation. Depending on the density of veins in the peripheral region of the lung, this could enhance clearance and create subpleural sparing

      Lymphatics

      Pulmonary lymphatics in the bronchovascular bundles accompany pulmonary arteries and bronchioles into the peripheral lung.
      • Kambouchner M
      • Bernaudin JF.
      Intralobular pulmonary lymphatic distribution in normal human lung using D2-40 antipodoplanin immunostaining.
      These structures have relatively large dimensions and remove fluid from the bronchovascular bundles. The lymphatics in these bronchovascular bundles are usually located adjacent to the pulmonary arteries in the bundles. Pulmonary lymphatics are also present in the visceral pleura and in interlobular septa.
      • Robinson SK
      • Ramsden JJ
      • Warner J
      Correlative 3D imaging and microfluidic modelling of human pulmonary lymphatics using immunohistochemistry and high-resolution μCT.
      These structures have a complex anatomy, which includes prelymphatic tissue, lymphatic reservoirs, and lymphatic conduits.
      • Weber E
      • Sozio F
      • Borghini A
      • et al.
      Pulmonary lymphatic vessel morphology: a review.
      In experimental models, the size of these lymphatics increases in conditions associated with edema formation.
      • Aharinejad S
      • Nourani F
      • Lametschwandtner A
      • et al.
      Pulmonary lymphatic filling is increased in spontaneously hypertensive rats.
      • Schraufnagel DE
      • Basterra JL
      • Hainis K
      • et al.
      Lung lymphatics increase after hyperoxic injury. An ultrastructural study of casts.
      • Schraufnagel DE
      • Agaram NP
      • Faruqui A
      • et al.
      Pulmonary lymphatics and edema accumulation after brief lung injury.
      The conduits are located on the interior surface of the visceral pleura and in the interlobular septa and drain of interstitial fluid, dendritic cells with antigens, proteins, and inflammatory mediators into larger lymphatic vessels, lymph nodes, and central veins. These structures have muscular cells that can contract and propel lymph, and they have valves to prevent backflow. The density of the lymphatics in the lung periphery decreases as the distance from the pleura increases. Interlobular lymphatics in the right lung empty into the right thoracic duct; interlobular lymphatics in the left lung injury empty into the left thoracic duct. Lymph production and lymph volume influence terminal thoracic duct pressures, and these pressures are higher than venous pressures at all times.
      • Ratnayake CBB
      • Escott ABJ
      • et al.
      The anatomy and physiology of the terminal thoracic duct and ostial valve in health and disease: potential implications for intervention.
      The ostial valve opens at the onset of inspiration and flow through the valve increases. Bicuspid semilunar valves in the terminal thoracic ducts prevent backflow into the lymphatic system. Increased fluid movement into the peripheral lymphatics reduces the accumulation of fluid and cellular elements in these zones and could cause disproportionate clearing of these regions during acute injury associated with the clinical conditions discussed above. In addition, in some patients with respiratory distress, larger intrathoracic pressure swings could increase cellular and fluid movement from the peripheral lung into central lymphatics and veins and increase the resolution of fluid collection in the subpleural region.
      • Soni N
      • Williams P.
      Positive pressure ventilation: what is the real cost?.

      Pressure gradients

      Pressure gradients move lymphatic fluid from the peripheral zone of the lung into central lymphatics of the lung and into central blood vessels.
      • Soni N
      • Williams P.
      Positive pressure ventilation: what is the real cost?.
      Pressure gradients should also move the venous blood from the peripheral lung into the central veins. The pressure gradients in the thorax vary throughout the respiratory cycle. During inspiration the intrapleural pressure decreases; during expiration the intrapleural pressure increases. These changes will occur during each respiratory cycle and will depend in part on the timing of the respiratory cycle, the mechanical properties of the lung, and respiratory muscle function. In the peripheral lung the combination of an increased density of lymphatic vessels and increased flow from peripheral to central veins should promote alveolar clearance in these regions and could produce subpleural sparing.

      Other considerations–prior disease

      Patients may have prior pulmonary disorders that prevent the development of subpleural sparing. For example, patients with emphysema have loss of lung parenchyma including both vessels and alveoli. Tissue that predominantly consist of airspaces might not have the same injury patterns as normal lung parenchyma. Therefore, the absence of subpleural sparing may not provide important information. Its presence does lead to a more focused list of possible diseases and considerations about the pathogenesis.

      Radiology

      The images obtained in radiographic studies of the lung reflect the disease process, normal lung anatomy, and radiographic techniques. The latter considerations include electron beam energy, detector location, detector size, contrast between tissues in region of interest, pixel size, partial volume effect, and noise. Pixel dimensions in millimeters depend on the display field-of-view and the matrix. Voxel dimensions depend on the pixel size and the slice thickness. The partial volume effect occurs when different tissues are present in the same voxel. Thoracic CT scans are obtained in the supine position and do not provide any information about possible effects of position on radiographic abnormality.
      Subpleural sparing is a radiographic term meaning there is more disease deep within the lung tissue away from the thoracic wall and less disease close to the thoracic wall. It is possible to have local subpleural sparing; one can see disease and sub-pleural sparing near the left and right lateral chest wall without seeing it near the anterior and posterior chest wall. However, the appearance of subpleural sparing near the right lateral chest wall should be apparent on both the axial slices AND the coronal slices. Similarly, disease with subpleural sparing seen near the posterior chest wall can be seen without findings near the anterior or lateral chest walls. However, sub-pleural sparing seen near the posterior chest wall should be apparent on both the axial slices AND the sagittal slices (Figure 1).
      Fig 1
      Fig. 1CT planes. Created by David Richfield and Mikael Häggström, MD. Wikimedia Commons. CC-BY-SA-3.0.
      Figures 2 and 3 demonstrate that true subpleural sparing can be confirmed by seeing the same degree of sparing on two different orthogonal views. In this case, both the axial and the coronal slices through the same region demonstrate the same degree of sub-pleural sparing. However, this is not always the case, and in some patients this subpleural sparing is visible only on one plane of the chest.
      Fig 2
      Fig. 2Axial CT slice. The orange arrow points to a region of subpleural sparing in the right postero-lateral lung. The yellow horizontal line shows the level of the coronal slice shown below in .
      Fig 3
      Fig. 3Coronal slice from the same patient as . The orange arrow points to a region of sub-pleural sparing in the right postero-lateral lung. The yellow horizontal line shows the level of the previous axial slice shown above in Figure 2.
      How can subpleural sparing be seen in one plane but not the necessary orthogonal plane? One possibility is the mapping of tissue density to grayscale on the CT image. This mapping is nonlinear with compression of densities (low contrast) not of interest and expansion of densities (high contrast) that are the subject of interest. The mapping is called the “window,” and, for example, much different settings for brightness and contrast mapping are used to highlight bone than to highlight lung tissue. Altering the brightness and contrast settings (window) might very well produce the appearance of a dark region next to ribs (Figures 4 and 5).
      Fig 4
      Fig. 4CT image with normal lung window settings for contrast and brightness. Note the peripheral sparing in the right postero-lateral lung (orange arrow). Note there is no peripheral sparing of the right anterior lung (yellow arrow).
      Fig 5
      Fig. 5Same CT image with decreased brightness and decreased contrast from normal lung window settings. The real peripheral sparing in the right postero-lateral lung remains (orange arrow). However, there is now the illusion of peripheral sparing in the right anterior lung (yellow arrow).

      CT slice thickness and image quality

      A computerized tomography (CT) image consists of a series of slices. As discussed above, these slices can be along three orthogonal axes: axial, coronal, and sagittal. The image quality depends on two concepts: the slice thickness and the slice interval.22,23 The slice interval or distance interval between slices is limited by the slice thickness and the number of slices in the image. Assuming that the slices are contiguous (the end of one slice is spatially coincident with the beginning of the next slice), a thinner image slice requires more slices and a smaller detector size. Smaller detector sizes permit thinner image slices. Thinner image slices permit greater resolution in the image, but they result in greater noise in the image.
      For a given radiation exposure, smaller detectors will have fewer photons strike them. This reduces the ability to distinguish white from gray and gray from black. For a given radiation exposure, smaller detectors have decreased contrast of the image (Figure 6). As discussed above, decreases in contrast can create an illusion of “sparing” of white or gray in a region of the image. So, thin slice images are better able to resolve two dots from a single blob, but they are less able to distinguish a small white structure adjacent to a small black structure from a larger gray structure. All other things being equal, a thin slice CT image will require greater radiation exposure to achieve equal image contrast.
      Fig 6
      Fig. 6CT slice resolution. Smaller detectors permit the increased resolution required to distinguish two small structures from one larger blob.

      Conclusions

      Subpleural sparing can develop in both acute and chronic lung diseases. It can reflect both the distribution of disease and the vagaries of CT imaging. Nonuniform involvement of lung tissue during an acute injury process could reflect either a differential distribution of the factors causing injury or nonuniform clearance and repair processes. The lymphatic tissue in the visceral pleura and interlobular septa have the capacity to remove fluid into the central region of the lung through lymphatics and venous connections. In chronic diseases, damage to the peripheral lymphatic system alters the repair process and helps explain the pathologic changes associated with idiopathic pulmonary fibrosis. Clinicians should review cases with subpleural sparing to determine whether or not the CT imaging shows this finding in at least two planes and to consider the pathogenetic implications of this finding in patient management.

      References

        • Chong WH
        • Saha BK
        • Austin A
        • Chopra A.
        The significance of subpleural sparing in ct chest: a state-of-the-art review.
        Am J Med Sci. 2021; 361: 427-435https://doi.org/10.1016/j.amjms.2021.01.008
        • Hansell DM
        • Bankier AA
        • MacMahon H
        • et al.
        Fleischner Society: glossary of terms for thoracic imaging.
        Radiology. 2008; 246: 697-722https://doi.org/10.1148/radiol.2462070712
        • Gruden JF
        • Naidich DP
        • Machnicki SC
        • et al.
        An algorithmic approach to the interpretation of diffuse lung disease on chest ct imaging: a theory of almost everything.
        Chest. 2020; 157: 612-635https://doi.org/10.1016/j.chest.2019.10.017
        • Panse PM
        • Feller FF
        • Butt YM
        • et al.
        Radiologic and pathologic correlation in EVALI.
        AJR Am J Roentgenol. 2020; 215: 1057-1064https://doi.org/10.2214/AJR.20.22836
        • Donnelly LF
        • Klosterman LA.
        Subpleural sparing: a CT finding of lung contusion in children.
        Radiology. 1997; 204 (In eng): 385-387https://doi.org/10.1148/radiology.204.2.9240524
        • Gondouin A
        • Manzoni P
        • Ranfaing E
        • et al.
        Exogenous lipid pneumonia: a retrospective multicentre study of 44 cases in France.
        Eur Respir J. 1996; 9: 1463-1469https://doi.org/10.1183/09031936.96.09071463
        • Imai R
        • Nishimura N
        • Takahashi O
        • Tamura T.
        High-resolution computed tomography for the prediction of mortality in acute respiratory distress syndrome: a retrospective cohort study.
        Health Sci Rep. 2021; 4: e418https://doi.org/10.1002/hsr2.418
        • Demirci NY DA
        • Tasci C
        • et al.
        Relationship between chest computed tomography findings and clinical conditions of coronavirus disease (COVID-19): a multicentre experience.
        International J of Clin Practice. 2021; 75: e14459
        • Silva CI
        • Müller NL
        • Hansell DM
        • et al.
        Nonspecific interstitial pneumonia and idiopathic pulmonary fibrosis: changes in pattern and distribution of disease over time.
        Radiology. 2008; 247: 251-259https://doi.org/10.1148/radiol.2471070369
        • Ebina M
        • Shibata N
        • Ohta H
        • et al.
        The disappearance of subpleural and interlobular lymphatics in idiopathic pulmonary fibrosis.
        Lymphat Res Biol. 2010; 8: 199-207https://doi.org/10.1089/lrb.2010.0008
        • Parra ER
        • Araujo CA
        • Lombardi JG
        • et al.
        Lymphatic fluctuation in the parenchymal remodeling stage of acute interstitial pneumonia, organizing pneumonia, nonspecific interstitial pneumonia and idiopathic pulmonary fibrosis.
        Braz J Med Biol Res. 2012; 45: 466-472https://doi.org/10.1590/s0100-879x2012007500055
        • Johkoh T.
        Nonspecific interstitial pneumonia and usual interstitial pneumonia: is differentiation possible by high-resolution computed tomography?.
        Semin Ultrasound CT MR. 2014; 35: 24-28https://doi.org/10.1053/j.sult.2013.10.004
        • Soni N
        • Williams P.
        Positive pressure ventilation: what is the real cost?.
        Br J Anaesth. 2008; 101: 446-457https://doi.org/10.1093/bja/aen240
        • O'Hagan LA
        • Windsor JA
        • Itkin M
        • et al.
        The lymphovenous junction of the thoracic duct: a systematic review of its structural and functional anatomy.
        Lymphat Res Biol. 2020; https://doi.org/10.1089/lrb.2020.0010
        • Ratnayake CBB
        • Escott ABJ
        • et al.
        The anatomy and physiology of the terminal thoracic duct and ostial valve in health and disease: potential implications for intervention.
        J Anat. 2018; 233: 1-14https://doi.org/10.1111/joa.12811
        • Kambouchner M
        • Bernaudin JF.
        Intralobular pulmonary lymphatic distribution in normal human lung using D2-40 antipodoplanin immunostaining.
        J Histochem Cytochem. 2009; 57: 643-648https://doi.org/10.1369/jhc.2009.953067
        • Robinson SK
        • Ramsden JJ
        • Warner J
        Correlative 3D imaging and microfluidic modelling of human pulmonary lymphatics using immunohistochemistry and high-resolution μCT.
        Sci Rep. 2019; 9: 6415https://doi.org/10.1038/s41598-019-42794-7
        • Weber E
        • Sozio F
        • Borghini A
        • et al.
        Pulmonary lymphatic vessel morphology: a review.
        Ann Anat. 2018; 218: 110-117https://doi.org/10.1016/j.aanat.2018.02.011
        • Aharinejad S
        • Nourani F
        • Lametschwandtner A
        • et al.
        Pulmonary lymphatic filling is increased in spontaneously hypertensive rats.
        Anat Embryol (Berl). 1999; 200: 425-432https://doi.org/10.1007/s004290050292
        • Schraufnagel DE
        • Basterra JL
        • Hainis K
        • et al.
        Lung lymphatics increase after hyperoxic injury. An ultrastructural study of casts.
        Am J Pathol. 1994; 144: 1393-1402
        • Schraufnagel DE
        • Agaram NP
        • Faruqui A
        • et al.
        Pulmonary lymphatics and edema accumulation after brief lung injury.
        Am J Physiol Lung Cell Mol Physiol. 2003; 284: L891-L897https://doi.org/10.1152/ajplung.00333.2002