Non-destructive three-dimensional structural assessment of transbronchial cryobiopsy specimens using phase-contrast X-ray CT: a proof-of-concept study

Introduction

Histopathological examination remains the gold standard for diagnosing lung cancer, and tumor classification is fundamentally based on morphological assessments supplemented with immunohistochemistry and molecular testing (1,2). In current clinical practice, tissue specimens obtained via biopsy or surgical procedures are subjected to formalin fixation, paraffin embedding, trimming, sectioning, and staining prior to microscopic evaluation. Although this workflow enables detailed cytological and histological assessments, it is inherently destructive and reduces three-dimensional tissue structures to a limited number of two-dimensional sections. Importantly, the spatial distribution of tumor components within a biopsy specimen cannot be fully appreciated before sectioning, and once the tissue is cut, structural information outside the selected planes is irreversibly lost. This limitation has become increasingly important as tissue acquisition strategies have evolved.

In the era of precision oncology, small bronchoscopic biopsy specimens are frequently used not only for histological diagnosis but also for molecular testing, including genomic profiling and immune biomarker evaluation (1,3,4). In advanced non-small cell lung carcinoma, adequate tumor cellularity is critical for reliable molecular analysis and therapeutic decision making (4,5). However, biopsy samples from peripheral pulmonary lesions (PPLs) may contain limited or non-representative tumor components, particularly when small or crushed biopsy specimens are obtained (6,7). Consequently, suboptimal sectioning or tissue utilization may lead to non-diagnostic histology, insufficient tumor content, or molecular testing failure (8,9). If the three-dimensional structural context of a biopsy specimen can be visualized before destructive processing, such information may help generate strategies for more efficient tissue use, particularly in settings where limited biopsy material must support both diagnosis and downstream ancillary testing.

Although histopathology provides microscopic details, three-dimensional structural assessment in clinical practice is most commonly performed using computed tomography (CT). CT provides volumetric information at the organ and lesion levels; however, its spatial resolution and contrast are limited in delineating fine tissue architecture at the microscopic level within biopsy specimens, particularly in soft tissues with subtle density differences (10-12). Conventional X-ray imaging based on absorption contrast has limited sensitivity when imaging soft tissues with a low intrinsic contrast, restricting its ability to visualize the internal structures of formalin-fixed specimens. In contrast, phase-contrast X-ray imaging detects phase shifts induced by X-ray propagation and provides markedly higher sensitivity to density variations in soft tissues (11,13). Experimental studies using synchrotron radiation have demonstrated that phase-contrast techniques can be used to visualize fine structures in biological specimens without the need for exogenous contrast agents (14,15).

Despite these advances, the feasibility and clinical relevance of phase-contrast X-ray imaging of human biopsy specimens obtained in routine clinical settings remain unclear. Most prior studies have focused on proof-of-concept applications using postmortem tissues or small, carefully prepared samples under controlled conditions, often aiming to achieve a very high spatial resolution (16). Whether phase-contrast imaging can provide clinically meaningful three-dimensional structural information on intact bronchoscopic biopsy specimens has not yet been systematically investigated.

In the present study, we performed synchrotron-based phase-contrast X-ray CT imaging of human peripheral lung biopsy specimens obtained during bronchoscopy. We aimed to evaluate whether this technique could be used to non-destructively visualize three-dimensional tissue architecture in clinically obtained bronchoscopic biopsy specimens and to assess the reproducibility of image-based structural evaluation among observers. Through this proof-of-concept study, we aimed to clarify the potential role of phase-contrast X-ray imaging as a complementary tool for non-destructive structural assessment of lung biopsy specimens. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0422/rc).

Methods

Study design and subjects

This exploratory imaging study was conducted to evaluate the three-dimensional tissue architecture of lung biopsy specimens using synchrotron-based phase-contrast X-ray CT. Additional tissue specimens were obtained during bronchoscopic procedures for research purposes after obtaining written informed consent from all participants.

This study was conducted as a collaborative research project between the National Cancer Center Hospital (Tokyo, Japan) and Hitachi, Ltd. (Tokyo, Japan). The study protocol was approved by the National Cancer Center Institutional Review Board (No. 2022-012); the study procedures were additionally reviewed and approved by the Institutional Review Board of Hitachi, Ltd. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Participants were prospectively enrolled between November and December 2022 according to predefined eligibility criteria: (I) chest high-resolution CT (HRCT) was performed within 1 month before enrollment, (II) HRCT demonstrated a PPL suspected to be malignant, (III) bronchoscopic biopsy of the target lesion was planned, (IV) the patient was 18 years or older, and (V) written informed consent for participation in this study was obtained. Patients were excluded if they received prior treatment for the target lesion, including systemic therapy or radiotherapy. No formal sample size calculation was performed owing to the exploratory nature of this study.

Bronchoscopic biopsy procedures

Bronchoscopic procedures were performed under moderate-to-deep sedation using fentanyl combined with midazolam or propofol under topical airway anesthesia. All patients were intubated with an 8.0-mm endotracheal tube before the procedure. A thin bronchoscope (BF-P260F or BF-P290; Olympus) was used to approach the PPL. The target bronchus was identified using virtual bronchoscopic navigation based on the preprocedural HRCT images. Radial endobronchial ultrasound (UM-S20-17S; Olympus, Tokyo, Japan) was used to confirm the location of the target lesion, and bronchoscopic access was obtained under fluoroscopic guidance.

Diagnostic sampling for clinical purposes was performed according to the standard institutional procedures. In most cases, conventional sampling techniques were performed first, followed by cryobiopsy, which was the primary sampling technique, depending on procedural considerations. Rapid on-site cytological evaluation was performed during the procedure in all cases. After completion of clinical diagnostic sampling, a single additional biopsy specimen was obtained for research purposes using cryobiopsy with a 1.9-mm reusable cryoprobe (Erbe Elektromedizin GmbH, Tübingen, Germany), as previously described (17). The obtained specimens were immediately placed in a container containing 10% neutral-buffered formalin and stored until subsequent imaging analyses.

Synchrotron-based phase-contrast X-ray CT imaging

Phase-contrast X-ray CT was performed using a crystal X-ray interferometer-based imaging system (18) at the BL-14C beamline of the Photon Factory of the High-Energy Accelerator Research Organization (Tsukuba, Japan). Monochromatic synchrotron X-rays with an energy of 17.8 keV were used for imaging. To prevent specimen deformation during rotational CT scanning, each biopsy specimen was placed in a standard laboratory plastic pipette tip and immersed in 10% neutral-buffered formalin in a custom sample holder during image acquisition. The projection images were acquired over a full 360° rotation. The images were recorded using a scientific Complementary Metal-Oxide-Semiconductor detector (Zyla 5.5 HF; Andor Technology Ltd., Belfast, UK) with an effective pixel size of 6.5 µm. The field of view was approximately 16 mm × 13 mm. The acquisition time for each specimen was several hours with 500 projections. Tomographic reconstruction was performed to generate axial cross-sectional images, which were subsequently used for qualitative and semi-quantitative evaluation of tissue architecture.

Image assessment and statistical analysis

Phase-contrast CT images were evaluated to assess the three-dimensional tissue architecture of the biopsy specimens. Image assessment focused on the following structural features: (I) degree of alveolar-like airspace preservation, (II) overall tissue density, (III) presence of glandular structures, and (IV) presence of solid nest-like structures. These evaluation features and their corresponding scoring criteria were newly developed for this exploratory study by one author (Y.M.) based on visual comparisons of phase-contrast CT images with corresponding histological findings. No independent training or pilot datasets were available, and the scoring system was therefore applied to the same cohort in which it was developed. Before the observer-based image assessments, the scoring criteria were fixed and shared with the three observers.

Each feature was scored using a four-point ordinal scale [0–3] according to the following predefined criteria. For alveolar-like airspace preservation, the scores ranged from 0 (absent) to 3 (extensive and well-preserved alveolar architecture). The tissue density scores ranged from 0 (very low density) to 3 (highly dense and solid). For glandular and solid nest-like structures, the scores ranged from 0 (absent) to 3 (extensive and clearly identifiable). Image evaluation was independently performed by three physicians (H.F., T.I., and T.T.) experienced in bronchoscopic diagnosis and thoracic imaging. The observers were blinded to the pathological diagnoses and other clinical information to minimize assessment bias. The three observers were not involved in phase-contrast CT image acquisition or tomographic reconstruction. For each specimen, the scores assigned by each observer were recorded. The interobserver agreement for each evaluation feature was assessed using quadratic weighted kappa statistics for pairwise comparisons (H.F.-T.I., H.F.-T.T., and T.I.-T.T.). In addition, the overall agreement among the three raters was evaluated using intraclass correlation coefficients (ICCs) based on a two-way random-effects model for absolute agreement [ICC(A,1)]. All statistical analyses were performed using EZR (version 1.70), a graphical user interface for R designed for medical statistics (19), and the irr package.

Exploratory analyses: serial histological sectioning and slow-freezing micro-CT imaging

To validate the structural findings observed on phase-contrast CT and explore the potential for higher-resolution visualization, additional exploratory analyses were performed using serial histological sectioning and slow-freezing micro-CT imaging. For histological validation, serial histological sectioning was performed on three representative specimens selected to reflect distinct structural patterns observed on phase-contrast CT, including preserved alveolar-like architecture, glandular/luminal structures, and predominantly dense solid architecture. Serial sections were prepared at a 10-µm thickness using standard pathological procedures and stained with hematoxylin and eosin. These sections were used for qualitative comparisons with corresponding phase-contrast CT findings.

For slow-freezing micro-CT imaging, five representative specimens with different histopathological patterns were selected to explore whether higher-resolution absorption-based imaging could provide finer structural information. Biopsy specimens were first divided into smaller fragments using a surgical scalpel because the original specimens exceeded the size limits of the micro-CT measurement system when mounted on pipette tips. Because this procedure required physical division of the specimen, this workflow was not non-destructive at the whole-specimen level and was not intended for direct comparison with the main phase-contrast CT protocol. The resulting tissue fragments were placed in microtubes with inner diameters of approximately 2 mm. Controlled slow freezing was performed using a cold nitrogen stream to induce gradual ice crystal formation while preserving the overall tissue architecture (20).

Cryo-micro-CT imaging was performed using synchrotron radiation at the BL07 beamline of Saga Light Source (Saga, Japan). Monochromatic X-rays with energy of 10 keV were used. Images were acquired using a micro-X-ray imager (Kenvy 2; Nikon Metrology NV, Leuven, Belgium) (21), which yielded an effective pixel size of 1.3 µm. Projection images were obtained during rotational scanning and reconstructed to generate three-dimensional datasets, which were then used for qualitative assessment of fine structural features.

Results

Fifteen patients were included in the study (Table 1). The median tumor size was 45.7 mm (range, 17.3–88.3 mm). Most lesions were located in the upper lobes, and bronchus signs were present in all but one case. Histologically, adenocarcinoma was the most common type, followed by squamous cell carcinoma, non-small cell lung carcinoma, not otherwise specified.

Representative phase-contrast CT images from three cases are shown in Figure 1, including both long- and short-axis views relative to the cryoprobe insertion direction. In the specimen obtained from a lesion that yielded non-tumorous lung tissue (case H04; Figure 1A,1B), preserved alveolar-like air spaces were clearly visualized. A dense structure suggestive of a thickened bronchial wall was also observed, indicating that the specimen contained normal bronchial wall components rather than tumor tissue. In contrast, a specimen diagnosed as adenocarcinoma (case H02; Figure 1C,1D) demonstrated the collapse of alveolar-like airspaces and an overall increase in tissue density. Within the dense tissue, multiple small luminal structures suggestive of a glandular architecture were identified. In another specimen diagnosed as poorly differentiated adenocarcinoma (case H14; Figure 1E,1F), alveolar-like airspaces were not clearly identified. Compared with case H02, the gland-like luminal structures were much less conspicuous, and the tissue appeared uniformly dense, suggesting a solid nest-like growth pattern. Additionally, punctate high-density foci suggestive of calcification were observed. A central cavity corresponding to the cryoprobe tract was consistently observed in all specimens and was regarded as a sampling-related artifact; therefore, it was not used as a scoring feature.

Figure 1 Representative phase-contrast CT images of bronchoscopic biopsy specimens. Representative long-axis (A,C,E) and short-axis (B,D,F) views relative to the direction of cryoprobe insertion are shown. In a specimen obtained from a lesion that yielded nontumorous lung tissue (case H04; A,B), preserved alveolar-like air spaces are clearly visualized. A dense structure suggestive of a thickened bronchial wall (arrows) is also observed, indicating the presence of normal bronchial wall components. In contrast, a specimen diagnosed as adenocarcinoma (case H02; C,D) demonstrates collapse of alveolar-like airspaces with increased tissue density. Within the dense tissue, multiple small luminal structures suggestive of glandular architecture (arrowheads) are identified. In another specimen diagnosed as poorly differentiated adenocarcinoma (case H14; E,F), alveolar-like airspaces are not clearly identified. Compared to case H02, the gland-like luminal structures are much less conspicuous, and the tissue appears uniformly dense, suggesting a solid nest-like growth pattern. Punctate high-density foci suggestive of calcification are also observed. The central cavity, visible in all specimens, corresponds to the space originally occupied by the cryoprobe and did not reflect the intrinsic tissue architecture. CT, computed tomography.

Serial histological sectioning was performed on three representative specimens, and slow-freezing micro-CT imaging was performed on five specimens. To interpret the structural features observed in phase-contrast CT images, serial histological sections were examined in representative cases (Figure 2). Histological examination of the specimen containing non-tumorous lung tissue demonstrated preserved alveolar architecture and a thickened bronchial wall (Figure 2A), corresponding to the structures observed on CT (Figure 2B). In the adenocarcinoma specimen, histology revealed tumor cells proliferating in an acinar pattern within the background of the collapsed lung parenchyma (Figure 2C), corresponding to the luminal structures on CT (Figure 2D). In the poorly differentiated adenocarcinoma specimen, histology showed that tumor cells proliferated predominantly in compact solid nests with only focal glandular differentiation (Figure 2E), consistent with the dense solid structures on CT (Figure 2F). In addition, punctate high-density foci on the CT images corresponded to calcified structures within the specimen. Overall, the structural features observed on phase-contrast CT generally corresponded with the underlying histological architecture.

Figure 2 Comparison between phase-contrast CT images and corresponding histological sections in representative cases. Representative paired histological (A,C,E) and phase-contrast CT images (B,D,F), which are the magnified views of Figure 1B, 1D, and 1F, respectively, are shown. In a specimen containing non-tumorous lung tissue (case H04), histology demonstrates preserved alveolar architecture and a thickened bronchial wall (A), corresponding to the structures observed on CT (B). In an adenocarcinoma specimen (case H02), tumor cells proliferated in an acinar pattern within the background of collapsed lung parenchyma (C), corresponding to multiple small luminal structures on CT (D). In a poorly differentiated adenocarcinoma specimen (case H14), the tumor cells show predominantly solid growth with focal glandular differentiation (E), consistent with the dense solid architecture observed on CT (F). Punctate high-density foci on CT correspond to calcified structures within the specimen. H&E stain; whole-slide images scanned using a 40× objective lens and displayed at 5×. CT, computed tomography; H&E, hematoxylin and eosin.

The interobserver agreement results are summarized in Table 2. For alveolar-like airspace preservation and tissue density, interobserver agreement was moderate to substantial, with weighted kappa values ranging from 0.49 to 0.76 across rater pairs and ICC(A,1) values of 0.63 and 0.66, respectively. In contrast, agreement for glandular and solid nest-like structures was low, with weighted kappa values close to zero or negative in some rater pairs and ICC(A,1) values of 0.08 and 0.22, respectively. These findings indicated that global structural features can be assessed more reproducibly than finer morphological patterns.

Representative cryo-micro-CT images and the corresponding histological sections are shown in Figure 3. This analysis was exploratory and was performed under different technical conditions from the main phase-contrast CT imaging, including a different beamline, lower X-ray energy, smaller effective pixel size, and physically divided specimens. In case H11 (squamous cell carcinoma), micro-CT imaging based on conventional absorption contrast demonstrated a predominantly dense structure without identifiable alveolar airspaces, consistent with the histological findings. At a low magnification, the overall architecture of the specimen appeared to be relatively homogeneous and compact. At higher magnifications, a spatial correspondence between micro-CT and histology was observed, with solid tumor nests on histology corresponding to dense regions on the micro-CT images. Compared with phase-contrast CT, micro-CT enables partial visualization of finer microstructural details, including subtle clustered structures suggestive of cellular components. However, image blurring and structural distortion were also observed, and a clear delineation of the tissue architecture comparable to the histological evaluation was not achieved. These findings indicate that while higher-resolution absorption-based imaging may reveal additional microstructural information, substantial limitations remain in resolving tissue architecture at a level comparable to that of histology.

Figure 3 Correlation between cryo micro-CT and histological findings in a squamous cell carcinoma specimen (case H11). Representative paired cryo-micro-CT images (A,C) and corresponding histological images (B,D) are shown, including low-magnification overview images (A,B) and higher-magnification views (C,D). In this absorption-based micro-CT approach, the images reveal a predominantly dense tissue structure without identifiable alveolar airspaces (A), consistent with the solid growth pattern observed on histology (B). At higher magnification, subtle clustered structures with relatively lower density on micro-CT (C) show spatial correspondence with solid tumor nests observed on histology (D), suggesting the visualization of tissue features approaching the cellular scale, although the precise delineation of individual cells remains limited. H&E stain; whole-slide image scanned using a 20× objective lens and displayed at 5× (B) and at 20× (D). CT, computed tomography; H&E, hematoxylin and eosin.

Discussion

This exploratory study demonstrated that synchrotron-based phase-contrast CT could be used to non-destructively visualize three-dimensional tissue architecture in clinically obtained human bronchoscopic biopsy specimens. Although the spatial resolution was insufficient for detailed cellular-level histological interpretation, key structural features, including alveolar-like airspace preservation, overall tissue density, and major growth patterns, were identifiable and showed a general correspondence with histological findings. Interobserver agreement was moderate for features reflecting global tissue architecture, whereas it was lower for finer structural patterns, indicating both the potential and current limitations of this approach.

Recent advances in X-ray phase-contrast imaging have enabled so-called “virtual histology”, in which tissue microarchitecture can be visualized in three dimensions at high spatial resolution under optimized experimental conditions (14,16). These approaches typically rely on multiscale imaging strategies using small specimens or region-of-interest scanning to achieve near-cellular resolutions (14,22,23). In contrast, the present study focused on intact bronchoscopic biopsy specimens measuring several millimeters without trimming or selective imaging. This clinically oriented approach inevitably involves trade-offs between specimen size and achievable spatial resolution. The resulting difference in spatial resolution compared with previous high-resolution virtual histology studies is therefore not unexpected and reflects fundamental differences in imaging strategy and study design.

Recent studies have demonstrated the feasibility of using synchrotron-based X-ray phase-contrast imaging with human biopsy specimens. Li et al. showed that phase-contrast imaging of formalin-fixed paraffin-embedded blocks did not compromise subsequent routine histology, immunohistochemistry, or molecular analyses, supporting the complementary nature of this approach to conventional pathology workflows (11). However, their study used formalin-fixed paraffin-embedded blocks and a multiscale propagation-based imaging strategy with smaller effective voxel sizes, whereas the present study focused on intact formalin-fixed cryobiopsy specimens imaged without trimming or paraffin embedding. The greater specimen thickness and the need to preserve the entire clinically obtained biopsy inevitably required a wider field of view, thereby constraining the achievable spatial resolution under the current whole-specimen imaging protocol. Planinc et al. reported substantial agreement between X-ray phase-contrast virtual histology and classical histopathology for acute cellular rejection in endomyocardial biopsies (24). Their study used high-resolution propagation-based phase-contrast imaging with an effective pixel size of 0.65 µm and evaluated a histopathologically well-defined target by an experienced pathologist. In contrast, the present study evaluated lung tumor architectural patterns on clinically oriented, lower-resolution whole-specimen phase-contrast CT images using newly defined criteria and multiple blinded observers. These differences in specimen type, target histological features, image resolution, and reader expertise likely explain why the reproducibility for fine tumor growth patterns was lower in the present study.

Several methodological factors likely contributed to the relatively low apparent resolution observed in this study. First, the crystal X-ray interferometer-based phase-contrast imaging system used was designed to prioritize high-density sensitivity and wide-field structural assessment over maximum spatial resolution (18,25). Second, imaging was performed on relatively large specimens immersed in formalin, which limited the achievable magnification and may have introduced minor image degradation due to specimen movement or deformation during prolonged acquisition. Third, unlike multiscale approaches, no additional high-resolution imaging of the selected regions was performed. Together, these factors emphasize that the present study prioritized the preservation of the overall specimen architecture over maximizing spatial resolution. Importantly, this study was designed with a primary objective different from that of prior virtual histology studies. Rather than reproducing histological images at near-cellular resolution, we sought to determine whether the three-dimensional architecture of clinically obtained lung biopsy specimens could be assessed non-destructively before conventional histological processing. From this perspective, the ability to identify major structural features, such as preserved alveolar architecture in non-tumorous tissue, increased tissue density associated with tumor infiltration, and differences in growth patterns between gland-forming and predominantly solid tumors, represents a meaningful advancement. In particular, volumetric visualization of intact specimens may provide information complementary to conventional two-dimensional histology, including spatial relationships and structural heterogeneity that cannot be fully appreciated in thin sections (11,22).

Several other imaging approaches can also provide microscopic or mesoscopic information during sampling or specimen evaluation, but they address different parts of the diagnostic workflow. Probe-based confocal laser endomicroscopy and optical coherence tomography can provide rapid optical information during bronchoscopy or ex vivo specimen assessment and may therefore be useful for sampling guidance or surface-level tissue assessment (26-28). However, these optical techniques generally offer only a limited field of view and imaging depth and do not provide non-destructive volumetric information of the entire biopsy specimen. Ex vivo micro-CT can provide three-dimensional specimen imaging and is compatible with laboratory-based systems. However, conventional absorption-based imaging of soft tissues often requires contrast agents or additional tissue preparation to compensate for limited intrinsic soft-tissue contrast, which may introduce tissue distortion or interfere with downstream histologic workflows (10,11,29). Digital pathology-based three-dimensional reconstruction from serial histological sections can provide high-resolution spatial information, but it remains destructive and may exhaust small biopsy specimens (11,30). In contrast, the potential role of phase-contrast CT in the present study is not in real-time sampling guidance but in non-destructive visualization of the internal architecture of an intact clinically obtained cryobiopsy specimen before conventional sectioning.

Another important aspect of this study was the deliberate use of cryobiopsy specimens for imaging purposes. In contemporary thoracic oncology, cryobiopsy has gained increasing importance because it enables the acquisition of larger tissue samples with better preservation of tissue architecture, potentially improving diagnostic performance compared to conventional forceps biopsy (7,31,32). This characteristic is particularly advantageous for histological and molecular analyses as it increases the likelihood of obtaining diagnostically adequate materials. However, the circumferential sampling of cryobiopsy introduces unique technical and interpretative challenges (17,33). Because the tissue is obtained in a three-dimensional, radial manner surrounding the cryoprobe tip, diagnostically relevant tumor components may be unevenly distributed within the specimen rather than uniformly represented. Consequently, standard sectioning may fail to capture these components, potentially leading to a discordance between cytological and histological evaluations (31,34). This limitation highlights the importance of understanding the three-dimensional spatial distribution of tissue components within cryobiopsy specimens.

Therefore, the most plausible potential translational niche of this approach may be tissue stewardship rather than replacement of conventional histopathology. In lung cancer, small biopsy specimens are increasingly required to support histological diagnosis, immunohistochemistry, and comprehensive molecular profiling (4,8,9). Because tumor-rich areas may be unevenly distributed within cryobiopsy specimens, volumetric pre-sectioning information could, in principle, help identify regions where sectioning may yield higher tumor cellularity. Similar principles may also be applicable to other small-specimen settings in which architecture or spatial context is important, such as lymphoma workup or interstitial lung disease; however, these applications remain speculative and require dedicated validation. The present study should therefore be regarded as a feasibility step toward such future tissue-stewardship applications and not as evidence that phase-contrast CT improves diagnostic yield or molecular testing success.

Observer-based analysis demonstrated moderate interobserver agreement for features reflecting global tissue architecture, such as alveolar-like airspace preservation and tissue density. In contrast, agreement remained poor for glandular and solid nest-like structures, likely reflecting both technical and interpretative limitations. Unlike for conventional histopathology, standardized criteria for interpreting tumor growth patterns using phase-contrast CT images of lung biopsy specimens have not yet been established. In addition, the observers evaluated the images without a separate training set or validated reference atlas because the number of available research specimens was limited. Therefore, the low agreement should not be interpreted simply as observer variability but rather as evidence that the current imaging protocol and scoring system are insufficient for reproducible assessment of diagnostically relevant fine tumor architectural patterns. Further accumulation of paired phase-contrast CT-histology datasets, development of standardized imaging criteria, reader training, and external validation will be required before such features can be used for image-guided tissue-processing applications.

To explore potential strategies for improving the spatial resolution, we performed additional exploratory imaging using a slow-freezing micro-CT technique based on absorption contrast (20). This approach enabled partial visualization of finer microstructural features, including subtle clustered structures suggestive of cellular components. However, image blurring and structural distortion limit the ability to delineate tissue architecture with clarity comparable to that of histological evaluation. These findings suggest that, while higher-resolution imaging may be achievable through alternative approaches such as reduced sample size, optimized imaging geometry, or advanced contrast mechanisms, substantial challenges remain in balancing spatial resolution with specimen size and structural preservation when applying these techniques to clinically obtained biopsy specimens (18,23).

This study had several limitations. First, the sample size was small, which reflects the exploratory nature of the study. Second, the imaging system was not optimized for multiscale high-resolution imaging, which limits direct comparison with prior virtual histology studies. Third, additional analyses, including serial histological sectioning and slow-freezing micro-CT imaging, were performed on selected representative specimens, which may have introduced a selection bias. Fourth, the evaluation criteria used for image assessment were newly defined in this study using the same limited dataset that was subsequently evaluated. No independent pilot or training datasets were available, and standardized criteria for phase-contrast CT interpretation of lung tumor specimens have not yet been established. Finally, several technical barriers must be overcome before this approach can be translated beyond a synchrotron-based research setting. The current acquisition required a large-scale synchrotron facility and several hours per specimen, which is incompatible with routine pathology workflows. The spatial resolution achieved in this whole-specimen protocol was insufficient for reliable assessment of fine tumor growth patterns or cellular-level diagnosis. Increasing spatial resolution would require trade-offs in field of view, acquisition time, image noise, and specimen size. Future development should therefore focus on laboratory-compatible phase-contrast CT systems, shorter acquisition and reconstruction times, optimized specimen mounting to reduce motion and deformation, and validated endpoints that demonstrate added value over standard sectioning.

Conclusions

Synchrotron-based phase-contrast CT enabled non-destructive, three-dimensional visualization of tissue architecture in clinically obtained human lung biopsy specimens. Although the current spatial resolution remains insufficient for detailed histological interpretation, the ability to assess overall tissue architecture represents a promising step toward volumetric pathology. Further technical advances, standardized image interpretation criteria, and prospective validation using clinically meaningful endpoints are required before this approach can be translated into tissue-processing or diagnostic workflows.

Acknowledgments

We would like to thank Editage (www.editage.jp) for the English language editing.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0422/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0422/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0422/prf

Funding: This work was supported by a collaborative research project between the National Cancer Center Hospital and Hitachi, Ltd. (Nos. C2021-168, C2023-137, and C2024-195).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0422/coif). Y.M. reports research funding from Hitachi, Ltd. related to the present study and grants from JSPS KAKENHI. He also reports consulting/advisory fees from Intuitive Surgical, Inc. and Cook Medical, and lecture honoraria from Olympus, AstraZeneca, Novartis, AMCO Incorporated, Thermo Fisher Scientific, Erbe Elektromedizin GmbH, Fujifilm, Chugai, Merck, Yuan Yu Industry Co., Ltd., Takeda, Ethicon, DNA Chip Research Inc., and Kaneka Medix Corporation. R.B. is an employee of Hitachi, Ltd. H.F. reports grants from Hitachi High-Tech Corporation, DNA Chip Research Inc., and JSPS KAKENHI; and lecture honoraria from DNA Chip Research Inc., Erbe Elektromedizin GmbH, AstraZeneca, Olympus, and Kaneka Medix Corporation. T.I. reports grants from Hitachi, Ltd. and Hitachi High-Tech Corporation; and lecture honoraria from Olympus and AMCO Incorporated. S.Y. reports grants from MSD. T.T. reports grants from JSPS KAKENHI; honoraria from Nippon Medical School Foundation; and participation on a Data Safety Monitoring Board at Hamamatsu University School of Medicine. The other author has no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the National Cancer Center Institutional Review Board (No. 2022-012), and the study procedures were reviewed and approved by the Institutional Review Board of Hitachi, Ltd. Written informed consent was obtained from all the participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Lindeman NI, Cagle PT, Aisner DL, et al. Updated Molecular Testing Guideline for the Selection of Lung Cancer Patients for Treatment With Targeted Tyrosine Kinase Inhibitors: Guideline From the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. Arch Pathol Lab Med 2018;142:321-46. [Crossref] [PubMed]
2. Nicholson AG, Tsao MS, Beasley MB, et al. The 2021 WHO Classification of Lung Tumors: Impact of Advances Since 2015. J Thorac Oncol 2022;17:362-87. [Crossref] [PubMed]
3. Lantuejoul S, Sound-Tsao M, Cooper WA, et al. PD-L1 Testing for Lung Cancer in 2019: Perspective From the IASLC Pathology Committee. J Thorac Oncol 2020;15:499-519. [Crossref] [PubMed]
4. Faber E, Grosu H, Sabir S, et al. Adequacy of small biopsy and cytology specimens for comprehensive genomic profiling of patients with non-small-cell lung cancer to determine eligibility for immune checkpoint inhibitor and targeted therapy. J Clin Pathol 2022;75:612-9. [Crossref] [PubMed]
5. Roy-Chowdhuri S, Dacic S, Ghofrani M, et al. Collection and Handling of Thoracic Small Biopsy and Cytology Specimens for Ancillary Studies: Guideline From the College of American Pathologists in Collaboration With the American College of Chest Physicians, Association for Molecular Pathology, American Society of Cytopathology, American Thoracic Society, Pulmonary Pathology Society, Papanicolaou Society of Cytopathology, Society of Interventional Radiology, and Society of Thoracic Radiology. Arch Pathol Lab Med 2020; Epub ahead of print. [Crossref]
6. Huang CT, Tsai YJ, Ho CC, et al. Atypical cells in pathology of endobronchial ultrasound-guided transbronchial biopsy of peripheral pulmonary lesions: incidence and clinical significance. Surg Endosc 2019;33:1783-8. [Crossref] [PubMed]
7. Tang Y, Tian S, Chen H, et al. Transbronchial lung cryobiopsy for peripheral pulmonary lesions. A narrative review. Pulmonology 2024;30:475-84.
8. Kerr KM, Bubendorf L, Lopez-Rios F, et al. Optimizing tissue stewardship in non-small cell lung cancer to support molecular characterization and treatment selection: statement from a working group of thoracic pathologists. Histopathology 2024;84:429-39. [Crossref] [PubMed]
9. Hofman P, Berezowska S, Kazdal D, et al. Current challenges and practical aspects of molecular pathology for non-small cell lung cancers. Virchows Arch 2024;484:233-46. [Crossref] [PubMed]
10. Clark DP, Badea CT. Micro-CT of rodents: state-of-the-art and future perspectives. Phys Med 2014;30:619-34. [Crossref] [PubMed]
11. Li KYC, Dejea H, De Winne K, et al. Feasibility and safety of synchrotron-based X-ray phase contrast imaging as a technique complementary to histopathology analysis. Histochem Cell Biol 2023;160:377-89. [Crossref] [PubMed]
12. Costello L, Donnelley M, Nesterets Y, et al. Evaluating the feasibility of region-of-interest X-ray phase contrast imaging for lung cancer diagnostics. Sci Rep 2025;15:19881. [Crossref] [PubMed]
13. Celestre R, Quénot L, Ninham C, et al. Review and experimental comparison of speckle-tracking algorithms for X-ray phase contrast imaging. J Synchrotron Radiat 2025;32:180-99. [Crossref] [PubMed]
14. Eckermann M, Frohn J, Reichardt M, et al. 3D virtual pathohistology of lung tissue from Covid-19 patients based on phase contrast X-ray tomography. Elife 2020;9:e60408. [Crossref] [PubMed]
15. Albers J, Svetlove A, Duke E. Synchrotron X-ray imaging of soft biological tissues - principles, applications and future prospects. J Cell Sci 2024;137:jcs261953. [Crossref] [PubMed]
16. Reichmann J, Verleden SE, Kühnel M, et al. Human lung virtual histology by multi-scale x-ray phase-contrast computed tomography. Phys Med Biol 2023;
17. Matsumoto Y, Nakai T, Tanaka M, et al. Diagnostic Outcomes and Safety of Cryobiopsy Added to Conventional Sampling Methods: An Observational Study. Chest 2021;160:1890-901. [Crossref] [PubMed]
18. Yoneyama A, Takamatsu D, Lwin T-T, et al. Crystal-Based X-ray Interferometry and Its Application to Phase-Contrast X-ray Imaging, Zeff Imaging, and X-ray Thermography. Applied Sciences 2023;13:5424. Available online: https://doi.org/10.3390/app13095424
19. Kanda Y. Investigation of the freely available easy-to-use software 'EZR' for medical statistics. Bone Marrow Transplant 2013;48:452-8. [Crossref] [PubMed]
20. Yoneyama A, Kawamoto M, Yasuda M, et al. Fine visualization of biological cells using X-ray micro-CT with the slow freezing contrast improved method. Sci Rep 2025;15:27641. [Crossref] [PubMed]
21. Yoneyama A, Baba R, Kawamoto M. Quantitative analysis of the physical properties of CsI, GAGG, LuAG, CWO, YAG, BGO, and GOS scintillators using 10-, 20- and 34-keV monochromated synchrotron radiation. Opt Mater Express 2021;11:398-411.
22. Donato S, Arana Peña LM, Arfelli F, et al. Integrating X-ray phase-contrast imaging and histology for comparative evaluation of breast tissue malignancies in virtual histology analysis. Sci Rep 2024;14:5831. [Crossref] [PubMed]
23. Allan H, Doherty A, Navarrete-León C, et al. Multi-contrast X-ray microtomography of human lung specimens with an extended field-of-view. Med Phys 2026;53:e70335. [Crossref] [PubMed]
24. Planinc I, Ilic I, Dejea H, et al. A Novel Three-Dimensional Approach Towards Evaluating Endomyocardial Biopsies for Follow-Up After Heart Transplantation: X-Ray Phase Contrast Imaging and Its Agreement With Classical Histopathology. Transpl Int 2023;36:11046. [Crossref] [PubMed]
25. Yoneyama A, Nambu A, Ueda K, et al. Phase-contrast X-ray imaging system with sub-mg/cm3 density resolution. J Phys Conf Ser 2013;425:192007.
26. Tian S, Huang H, Zhang Y, et al. The role of confocal laser endomicroscopy in pulmonary medicine. Eur Respir Rev 2023;32:220185. [Crossref] [PubMed]
27. Goorsenberg A, Kalverda KA, Annema J, et al. Advances in Optical Coherence Tomography and Confocal Laser Endomicroscopy in Pulmonary Diseases. Respiration 2020;99:190-205. [Crossref] [PubMed]
28. Kramer T, Wijsman PC, Kalverda KA, et al. Advances in bronchoscopic optical coherence tomography and confocal laser endomicroscopy in pulmonary diseases. Curr Opin Pulm Med 2023;29:11-20. [Crossref] [PubMed]
29. Katsamenis OL, Olding M, Warner JA, et al. X-ray Micro-Computed Tomography for Nondestructive Three-Dimensional (3D) X-ray Histology. Am J Pathol 2019;189:1608-20. [Crossref] [PubMed]
30. Pichat J, Iglesias JE, Yousry T, et al. A Survey of Methods for 3D Histology Reconstruction. Med Image Anal 2018;46:73-105. [Crossref] [PubMed]
31. Nishimatsu K, Matsumoto Y, Kashima J, et al. Concordance between cryobiopsy and forceps biopsy specimens in assessment of immunohistochemistry staining for non-small cell lung carcinoma. Transl Lung Cancer Res 2023;12:1245-55. [Crossref] [PubMed]
32. Kosuge A, Matsumoto Y, Furuse H, et al. Diagnostic utility of cryobiopsy for invasive mucinous lung adenocarcinoma. Transl Lung Cancer Res 2025;14:4514-26. [Crossref] [PubMed]
33. Kho SS, Chan SK, Yong MC, et al. Performance of transbronchial cryobiopsy in eccentrically and adjacently orientated radial endobronchial ultrasound lesions. ERJ Open Res 2019;5:00135-2019. [Crossref] [PubMed]
34. Imabayashi T, Uchino J, Yoshimura A, et al. Safety and Usefulness of Cryobiopsy and Stamp Cytology for the Diagnosis of Peripheral Pulmonary Lesions. Cancers (Basel) 2019;11:410. [Crossref] [PubMed]