Abstract
Background/Aim: The pathological diagnosis of organizing pneumonia (OP) relies on conventional traditional histopathological analysis, which involves examining stained thin slices of tissue. However, this method often results in suboptimal diagnostic objectivity due to low tissue sampling rates. This study aimed to assess the efficacy of tissue-clearing and infiltration-enhanced 3D spatial imaging techniques for elucidating the tissue architecture of OP. Materials and Methods: H&E staining, 3D imaging technology, and AI-assisted analysis were employed to facilitate the construction of a multidimensional tissue architecture using six OP patient specimens procured from Taichung Veterans General Hospital, enabling a comprehensive morphological assessment. Results: Specimens underwent H&E staining and exhibited Masson bodies and varying degrees of interstitial fibrosis. Furthermore, we conducted a comprehensive study of 3D images of the pulmonary histology reconstructed through an in-depth pathology analysis, and uncovered heterogenous distributions of fibrosis and Masson bodies across different depths of the OP specimens. Conclusion: Integrating 3D imaging for OP with AI-assisted analysis permits a substantially enhanced visualization and delineation of complex histological pulmonary disorders such as OP. The synergistic application of conventional histopathology with novel 3D imaging elucidated the sophisticated spatial configuration of OP, revealing the presence of Masson bodies and interstitial fibrosis. This methodology transcends conventional pathology constraints and paves the way for advanced algorithmic approaches to enhance precision in the detection, classification, and clinical management of lung pathologies.
Organizing pneumonia (OP), formerly known as bronchiolitis obliterans organizing pneumonia (BOOP), presents a unique clinical and pathological entity in pulmonary disorders that has been a subject of interest since its initial description over a century ago (1, 2). The histopathology of OP is distinguished by the presence of intraluminal plugs of granulation tissue, termed Masson bodies, within the alveolar ducts, alveoli, and bronchioles. This reflects the process of organizing connective tissue replacing the exudate typically seen in pneumonia (3). These Masson bodies are composed of loose connective tissue, fibroblasts, and myofibroblasts, and are a hallmark of the disease, indicating an ongoing repair process that becomes pathological when it becomes excessive and obstructive (3, 4). The pathogenesis of OP is complex and multifactorial, involving alveolar epithelial injury that leads to cell death and disruption of the basal lamina, triggered by a range of etiologies from viral infections and microaspiration to subclinical connective tissue diseases (CTD) (5, 6). The disease process of OP is thus considered to represent a common response pattern to lung injury.
Radiographically, OP is characterized by bilateral and diffuse alveolar opacities with preserved lung volumes (7). High-resolution computed tomography (HRCT) scans are clinical tools utilized to examine peripheral or peribronchial areas of consolidation and ground-glass opacities that can be indications of migratory patterns or responsiveness to corticosteroid treatment (7, 8). Furthermore, OP is diagnostically discerned through a combination of clinical, radiological, and histopathological criteria. While a definitive diagnosis often relies on tissue biopsy, less invasive procedures such as bronchoalveolar lavage (BAL) can provide evidence supporting clinical and imaging findings (9, 10). Typical treatments for OP, such as corticosteroids or clarithromycin, can yield a rapid and dramatic response albeit a subset of patients may experience remission without pharmacologic intervention (11). Despite its responsiveness to steroid and antibiotic treatments, OP can be a relapsing condition, which often necessitates prolonged therapy and carries higher risk of drug side effects (12). It is thus crucial to differentiate OP from other idiopathic interstitial pneumonias (IIPs), as therapeutic approaches and prognoses differ significantly (13).
As the damaged lung undergoes remodeling through re-epithelialization and repair, histological assessments often reflect the regenerative capacity of the lung tissue. The regenerative process, nevertheless, can be complicated by the development of fibrosis, particularly in cases where there is an overlap with nonspecific interstitial pneumonia (NSIP) that is often associated with a higher risk of disease progression (14). The potential for OP to progress to fibrosing lung disease or to acute respiratory failure hence highlights the heterogeneity of the disease progression, underscoring the need for a comprehensive approach to diagnosis and management.
Individuals afflicted with OP commonly exhibit a spectrum of non-specific respiratory symptoms that encompass cough, dyspnea, febrile episodes, and general malaise (14). In a subset of patients, thoracic discomfort or pain may exacerbate the clinical presentation due to the considerable heterogeneity in clinical manifestations among OP patients. This presents diagnostic challenges for clinicians and necessitates a discerning approach to differentiate OP from other pulmonary disorders such as NSIP with similar symptomatology (15). The symptomatic overlaps between OP and NSIP identified by histologic and radiologic findings are associated with unfavorable disease progression (14, 15). Thus, the precision and promptness of OP diagnosis are of paramount importance, constituting the central focus of the current study. The aim is to effectively diagnose and manage OP clinical conditions and ultimately facilitate the initiation of appropriate therapeutic interventions for optimizing patient outcomes.
Materials and Methods
OP specimen sampling and histological analysis. Examinations of OP specimens collected in this research were conducted in accordance with formal approval from the Institutional Review Board (IRB) of Taichung Veterans General Hospital (IRB approval number: CE23165A), ensuring strict adherence to ethical guidelines and the safeguarding of patient privacy. A total of six distinct patient specimens were acquired, each representing a unique clinical presentation, radiology, lung function, comorbidities, and other possible risk factors. Microscopic examinations of the specimens resulted in a consensus diagnosis of OP and were reviewed by pathologists YH. and YHT. Histological complexities of OP were first elucidated by performing hematoxylin and eosin (H&E) staining, which enabled 2D microscopic visualization on the intricate patterns, cellular arrangements, and pathological alterations in the architecture of OP-affected tissues across different patient specimens.
3D imaging and construction. Sections of lung tissue specimens, ranging from 100 to 150 μm in thickness, were prepared from formalin-fixed paraffin-embedded (FFPE) samples. These sections were de-waxed and rehydrated before undergoing staining. Briefly, the specimens were first treated with Triton X-100 (2%) and the lipophilic tracer DiD (20 μg/ml, cat, D307, ThermoFisher Scientific, Waltham, MA, USA) for 8 h, followed by an overnight incubation in a proprietary clearing solution (JelloX Biotech, Zhubei, Taiwan, ROC) with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI) (5 μg/ml, Cat No. D9542, Sigma-Aldrich, Burlington, MA, USA) at room temperature. These slides were then gently transferred to chambered coverslips for 3D image acquisition under confocal microscopy. Each image was generated through CLSM Z-stack scanning, incorporating a pinhole size of 235 μm and a Z interval of 1 μm between consecutive layers (Figure 1).
Schematic workflow of organizing pneumonia tissue processing for 3D histological analysis. This schematic workflow illustrates the sequential stages in preparing clinical tissue samples for microscopic examination and analysis. Initially, tissue specimens were embedded prior to sectioning into thin slices conducive to light transmission. The slices were subsequently labeled as described in Materials and Methods, followed by a clearing process to strip off any obfuscating media and labeling compounds. The cleared samples were then ready for 3D image reconstruction, which was followed by the application of AI algorithms for analyses on derived quantitative data and morphological features (AI analysis). Each of the steps is pivotal in ensuring the clarity and precision of the resulting histological data.
AI-assisted image analysis. The exported images were analyzed using AI software MetaLite developed by JelloX Biotech Inc, Hsinchu City, Taiwan (16, 17). The histological status of fibrosis and Masson bodies in lung tissues was assessed using two models within MetaLite, namely the fibrosis recognition model and the Masson body recognition model. Individual cells within the tissues were determined based on nuclear SYTO-16 staining and the DiD membranous staining, which allowed calculation of distribution for each image on the X-Y plane.
Three-dimensional reconstruction. The methodological advancement in our study was marked by the incorporation of state-of-the-art 3D imaging technology. This innovative approach facilitated the meticulous construction of comprehensive 3D representations depicting tissue structures affected by OP. Through the capture of spatial relationships, depth, and topographical features inherent to the tissue, our methodology surpassed the constraints associated with conventional 2D histology. The resultant 3D reconstruction provided a holistic perspective on the intricate three-dimensional organization of tissues in OP, thereby imparting invaluable insights into the pathogenesis of the disease.
Results
Demographics and health variables in the OP study cohort. In an effort to enhance our current understanding of OP, this investigation involved a cohort of six patients diagnosed with this intriguing pulmonary disorder. The inclusion of patients from diverse demographics, lung function, and other health backgrounds was intended to derive a more comprehensive understanding of OP in our analysis (Table I). Our study cohort was curated to achieve a balanced distribution comprised of an equal number of cancer/non-cancer and smoker/non-smoker patients. This balance was aimed to mitigate potential biases introduced by cancer- and smoking-attributed factors, ensuring robustness and applicability of our findings.
The clinicopathological features of the patients.
Histopathological variability in OP. Following the acquisition of tissue specimens from our patient cohort, H&E staining was conducted to analyze the patterns of Masson body formation within the OP-affected tissues. The tissue sections under microscopic examination revealed the distinctive presence of Masson bodies as indicated by the intraluminal plugs of fibroblast proliferation without collagenous fibrotic granulation tissue, in the peripheral alveoli and bronchioles. This observation was noted in three tissue sections (Figure 2A-C), whereas the rest appeared to elicit collagenous fibrosis, shrunken or without Masson bodies within the alveolar spaces (Figure 2D-F). This spectrum of fibrotic involvement suggests diverse progression and treatment responsiveness among individuals with OP.
2D histological staining results elicit spatial divergence of organizing pneumonia (OP). Representative H&E staining of the OP cohort showed areas with (A, B, C) or without Masson body formation (D, E, F). (H&E sections, original magnification: A, D: 100×; B, C, E, F: 400×).
AI-oriented analysis of tissue morphology variations. The 2D histological analysis in Figure 2 laid the groundwork for our further in-depth analysis of histological features associated with OP. Recently, 3D histology has emerged as a novel approach, wherein the tissue clearing process alters the physicochemical properties of tissues to achieve optical transparency, enabling researchers to visualize histological properties in depth (18). First, we successfully accomplished tissue clearing for the specimens from six OP patients. We next employed lipophilic tracer DiD with markedly red-shifted fluorescence excitation and emission spectra to avoid autofluorescence and facilitate two-color labeling on our optically transparent tissue sections. The reconstructed 3D images were then utilized for histological assessments including defining the regions of fibrosis and Masson bodies, followed by AI-assisted analysis on the images. This pioneering exploration into 3D pulmonary pathology of OP navigated the complex 3D terrain of tissues affected by OP (Figure 3).
3D reconstructed histological images demonstrate marked spatial divergence of organizing pneumonia (OP). Representative histological 3D image from each six cases of OP patients were reconstructed as described in Materials and Methods. Marked spatial divergence was observed, with red regions indicating the location of nuclei, and green indicating membranes.
Diverse distributions unveiled through the implementation of 3D imaging techniques. The 3D imaging techniques unveiled a plethora of insights within the six OP tissue specimens. As shown in Figure 4A, cases 2 and 4 demonstrated significantly more areas undergoing fibrosis, whereas cases 5 and 6 showed varying fibrosis areas at different depths. Meanwhile, cases 1 and 3 elicited minimal areas of fibrosis. In terms of Masson body distribution, we observed at least five Masson bodies in cases 3 and 6 at depths between 10 and 70 μm (Figure 4B). Since OP can potentially progress to fibrosing lung or acute respiratory diseases, we next investigated whether the observed Masson body distribution overlapped with the area of fibrosis. In two of our six cases, case 3 elicited at least two Masson bodies in the fibrosis area from 10 to 130 μm of Z-depth, and case 6 showed nearly one Masson body from 10 to 70 μm (Figure 4C). This nuanced comprehension of the spatial arrangement of pathological features offers a better understanding of the disease’s architectural intricacies, enabling us to discern how OP might manifest in distinct lung regions in clinical disease progression among patients (Figure 4).
Spatial divergence analysis on interstitial fibrosis and Masson from 3D images. Varying depth levels of interstitial fibrosis distribution (A), Masson body (B) and Masson bodies in the fibrosis area (C), ranging from 10 to 130 μm Z-depth, were analyzed for all six cases of OP patients.
Discussion
AI-assisted image analysis has emerged as a prominent approach across a variety of medical specialties. For instance, the accuracy of AI-assisted detection of esophageal cancer and neoplasms on endoscopic images provides valuable insights into the capacities and limitations of AI-assisted image analysis software (19). Stewart et al. also mapped ethical and legal principles for the use of AI in gastroenterology, emphasizing the integration of AI not only in image analysis during endoscopy and colonoscopy but also in genomic and epigenetic data analysis for cancer classifications (20).
The exhaustive investigation into OP presented in this study using MetaLite software offers profound insights into the intricate nature of this pulmonary disorder. As human tissues are naturally opaque and heterogeneous, which prevent efficient light penetration, tissue clearing techniques employed for OP tissues rendered them transparent. This improved light penetration and signal detection, facilitating universal fluorescence staining for AI-assisted image analysis by MetaLite.
The study cohort comprised six patients diagnosed with OP and was meticulously designed to ensure a sex-balanced distribution, with an equal representation of males and females (Table I). The histological analysis commenced with H&E staining, providing a foundational insight into the cellular structures and pathological features associated with OP (Figure 2). Traditional histopathological techniques have played a crucial role in the diagnosis of OP, yet they are not without limitations, especially concerning the sampling rate. To address these constraints, our study pioneers an innovative approach by introducing tissue-clearing, infiltration-based high-resolution 3D imaging that facilitated a thorough analysis of OP-affected tissue structures in three dimensions (Figure 3), marking a significant advancement in tissue analysis methodologies.
One of the major challenges associated with managing OP pertains to its intricate clinical and radiological features, which often closely resemble those of various other respiratory conditions, thereby presenting a diagnostic conundrum for healthcare providers. HRCT as the presently most sensitive diagnostic tool for interstitial lung disease (ILD) and COP diagnosis, still heavily relies on collaborative assessments form clinical evaluation, radiological imaging, and histopathological examination (21). Notably, Masson bodies, indicative of OP lung injury alongside varying degrees of interstitial fibrosis were identified in 3D reconstructed images of these tissue samples. Our AI-assisted image analysis by MetaLite for the six OP samples successfully identified 3D distributions of both interstitial fibrosis and Masson bodies (Figure 4), adding depth to our understanding of the disease’s complexity and its potential implications for prognosis. This pivotal discovery in the diverse distribution patterns of fibrosis and Masson bodies within OP tissues, particularly in terms of depth, emphasizes spatial correlation of the nuanced OP manifestation to different regions of the lung. In fact, recent advancements in AI tools have shown significant promise in transforming lung disease diagnosis by improving the accuracy and efficiency of medical imaging analysis. Diagnosis on a diverse range of lung conditions, such as cancer, tuberculosis, idiopathic pulmonary fibrosis, and COVID-19 has been systemically evaluated (22), highlighting the versatility of AI in lung disease diagnosis and management. More recently, common algorithms like U-Net, VGG16, and AlexNet utilized in AI systems have been employed for diagnosing various lung conditions (23).
In conclusion, this study adds a significant advancement to the understanding of OP and the role of AI-driven tools in revolutionizing lung disease diagnosis by providing advanced image analysis capacities. The integration of traditional histopathology and cutting-edge 3D imaging techniques in this research unveils previously obscured intricacies in tissue structures and their spatial distribution. Future incorporation of more sophisticated algorithms in the analysis of medical images is hence anticipated to further improve precise detection, classification, and management of pulmonary abnormalities.
Acknowledgements
This study was supported by TCVGH-1115801D from Taichung Veterans General Hospital Taiwan. The Authors thank all participants, Margaret Dar-Tsyr Chang, Sky Chung, and Alex Chi-yu Lin of Jellox Biotech Inc. (Taiwan), for their assistance in 3D pathology and immunofluorescence staining.
Footnotes
Authors’ Contributions
Conceptualization: YH and CJC; methodology: YH and YHT; formal analysis: YH, YHT and PKF; investigation: YH, YHT, PKF and CJC; data curation: YH and CJC; writing the original draft preparation: CJC; writing, reviewing, and editing: YH, YHT and CJC. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no conflicts of interest in relation to this study.
- Received March 27, 2024.
- Revision received April 12, 2024.
- Accepted April 15, 2024.
- Copyright © 2024, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.