Ex vivo fluorescence confocal microscopy for assessment of ureteroscopic biopsy adequacy in upper tract urothelial carcinoma

Abstract

Objectives

To evaluate the feasibility of ex vivo fluorescence confocal microscopy (FCM) for real‐time assessment of ureteroscopic biopsy specimen adequacy in patients with suspected upper tract urothelial carcinoma (UTUC).

Patients and Methods

In this prospective, single‐centre feasibility study, four ureteroscopic biopsy specimens obtained from two patients undergoing evaluation for suspected UTUC were analysed using ex vivo FCM. Specimens were briefly stained with acridine orange, scanned within 50–60 s, and images were stored for deferred review. A board‐certified genitourinary pathologist reviewed the FCM images in a blinded fashion and classified specimens as adequate or non‐representative based on a pre‐defined criterion of identifiable urothelial tissue. All specimens subsequently underwent routine haematoxylin and eosin (H&E) histopathological analysis for confirmation.

Results

Ex vivo FCM was successfully performed on all specimens without workflow interruption, with total preparation and image acquisition completed within approximately 5 min/specimen. Blinded FCM review identified urothelial tissue in three of four biopsy specimens, which were classified as adequate, while one specimen was composed predominantly of blood clot with only minimal, non‐diagnostic cellular material and was therefore classified as non‐representative. FCM‐based adequacy assessments were concordant with corresponding H&E sections. Final histopathology demonstrated non‐invasive papillary UC, high grade in one patient and low grade in the other; the non‐representative specimen showed only focal atypical papillary urothelial proliferation admixed within the blood clot.

Conclusion

Ex vivo FCM was feasible for immediate imaging of freshly obtained ureteroscopic biopsy specimens and allowed identification of urothelial‐containing fragments as a marker of specimen adequacy while preserving tissue for standard histopathological evaluation. These preliminary findings support further investigation in larger cohorts to evaluate reproducibility and to clarify the potential role of this approach for intraoperative or real‐time assessment of biopsy adequacy in UTUC, including correlation with final histopathological grade and stage.

Abbreviations

CLE

confocal laser endomicroscopy

FCM

fluorescence confocal microscopy

H&E

haematoxylin and eosin

UTUC

upper tract urothelial carcinoma

Introduction

Upper tract urothelial carcinoma (UTUC) remains diagnostically challenging because preoperative risk stratification relies heavily on ureteroscopic visual assessment, selective urine cytology, and small‐volume endoscopic biopsies, each of which is constrained by interpretive or sampling‐related limitations. Early clinical series comparing ureteroscopic biopsy with nephroureterectomy specimens demonstrated imperfect reliability for predicting definitive tumour characteristics, with invasive disease under recognised in a subset of high‐grade tumours [1]. Subsequent surgical correlation studies have continued to report preoperative misclassification, reflecting ongoing challenges related to tissue representativeness despite advances in endoscopic technique and standardised pathological grading [2, 3]. More recent multicentre data indicate that, although ureteroscopic biopsies are often sufficient to establish a histological diagnosis, limitations in staging accuracy persist in contemporary practice [4]. Together, these observations highlight the need for techniques capable of providing immediate feedback on specimen adequacy at the time of biopsy.

Given these diagnostic constraints, adjunctive optical imaging modalities have been explored to provide real‐time microarchitectural assessment during upper‐tract evaluation, aiming to improve endoluminal characterisation of urothelial lesions beyond conventional visual inspection. Early studies reported that in vivo ureteroscopic confocal laser endomicroscopy (CLE) can depict papillary and flat urothelial patterns with qualitative correspondence to histopathology [5]. Subsequent investigations reported that confocal imaging features correlate with histological grade, suggesting potential utility for lesion characterisation under selected imaging conditions [6, 7]. Collectively, these studies support the ability of CLE to visualise urothelial morphology in vivo, providing a biological rationale for extending confocal imaging beyond endoluminal assessment.

Ex vivo fluorescence confocal microscopy (FCM) extends these imaging principles to freshly obtained tissue specimens, enabling rapid microarchitectural assessment while preserving material for conventional histopathological processing. Initial clinical experience has focused on prostatic tissue, where ex vivo FCM has been used for intraoperative evaluation of surgical margins and biopsy specimens without compromising downstream pathology [8].

Broader evaluations of ex vivo confocal platforms suggest that diagnostically interpretable images can be obtained within minutes of tissue acquisition, supporting interest in their potential role for real‐time specimen assessment [9].

In this context, the present study examined one of the earliest applications of ex vivo FCM to freshly obtained ureteroscopic biopsy specimens from patients undergoing evaluation for suspected UTUC. The objective of this preliminary study was to determine whether urothelial tissue can be identified on ex vivo FCM of freshly obtained biopsy specimens as a minimal, pathology‐relevant criterion for assessing specimen adequacy during ureteroscopic biopsy.

Patients and Methods

Study Design

This prospective, single‐centre feasibility study evaluated ex vivo FCM for the assessment of ureteroscopic biopsy specimens in patients with suspected UTUC. A total of four biopsy specimens obtained from two patients, including repeat biopsies in one patient, were scanned immediately after acquisition using a confocal microscopy platform (Histolog® Scanner, SamanTree Medical SA, Lausanne, Switzerland), and the resulting images were stored for deferred interpretation. All specimens were subsequently submitted for routine formalin‐fixed haematoxylin and eosin (H&E) histopathological processing.

Specimen Handling and Confocal Imaging Workflow

Ex vivo FCM was performed using a large‐field confocal scanning system (Histolog Scanner). The platform integrates a wide‐area confocal laser‐scanning microscope with an onboard display interface, enabling rapid visualisation of fresh, unfixed tissue. Imaging was performed using a 488‐nm excitation laser with detection of emitted fluorescence above 500 nm, generating en face optical sections at an approximate resolution of 2 μm/pixel. The system incorporates a single rectangular scanning window measuring 48 × 36 mm, and each acquisition was completed within 50–60 s.

Immediately after retrieval, each biopsy fragment was placed on a sterile non‐adherent dressing and transferred to a glass coverslip using fine forceps. An acridine orange solution (Histolog® Dip, SamanTree Medical SA) was applied directly to the specimen for ~5 s to provide nuclear contrast. Excess dye was gently blotted with gauze prior to positioning the specimen within the scanning area. A piece of gauze was placed loosely over the imaging window to reduce ambient light during acquisition.

The en face confocal images were captured and stored digitally. Images were automatically pseudo‐coloured by the system software to enhance contrast, without altering underlying structural information, and optional greyscale or brightness adjustments were available during later review. Following imaging, all specimens were submitted for routine histopathological processing, with no observed impact on downstream histopathological evaluation.

Pathology Review and Adequacy Assessment

Digitised FCM images were reviewed in a deferred, blinded fashion by a board‐certified genitourinary pathologist. The reviewer first assessed each image for the presence or absence of identifiable urothelial tissue based on confocal features consistent with urothelial epithelium, including the presence of epithelial cells with discernible nuclear detail and architectural organisation distinct from non‐diagnostic material. The presence of non‐diagnostic material such as blood clot, necrotic debris, or acellular fragments was also documented.

For this preliminary study, specimen adequacy was intentionally defined using a minimal criterion: the presence of identifiable urothelial tissue on FCM imaging. Based on this determination, each specimen was classified as adequate or non‐representative for diagnostic interpretation. After FCM image review, corresponding H&E‐stained sections were examined to confirm tissue identity and provide context for the FCM findings.

Outcome Measures

The outcomes of interest were defined a priori to align with the exploratory objectives of this study. No assessment of tumour grade, stage, or histological classification concordance was planned. Feasibility was defined as the ability to successfully perform ex vivo FCM imaging on each biopsy specimen and to obtain images considered interpretable by the reviewing pathologist. Technical limitations, including motion, overstaining, fluid artefacts, or blood‐clot obscuration, were recorded when present.

Sample adequacy was evaluated based on identification of urothelial tissue on FCM imaging. For each specimen, the pathologist indicated whether urothelial tissue could be reliably recognised and whether the biopsy could be classified as adequate or non‐representative for diagnostic interpretation. Final H&E‐stained sections served as the reference standard for confirming tissue identity and specimen representativeness.

Results

Two patients undergoing diagnostic ureteroscopy with biopsy for suspected UTUC were included in this feasibility study, yielding a total of four biopsy specimens. One patient contributed a single specimen, while the second patient contributed three specimens. Ex vivo FCM imaging was performed successfully on all specimens immediately after retrieval. There were no technical failures, workflow interruptions, or delays in specimen processing, and all specimen preparation and FCM image acquisitions were completed within ~5 min/specimen.

Blinded review of the FCM images revealed identifiable urothelial tissue with discernible cellularity in three of the four biopsy specimens. These specimens exhibited epithelial features distinct from blood clot or acellular debris and were classified as adequate for diagnostic interpretation according to the pre‐defined adequacy criterion. One biopsy specimen lacked identifiable urothelial architecture and consisted predominantly of blood clot with minimal cellular material and was therefore classified as non‐representative on FCM imaging. The reviewing pathologist reported high confidence in adequacy classification for all specimens. From a technical standpoint, all FCM image acquisitions were rated as having good overall interpretability, with no motion, staining, fluid, or clot‐related artefacts significantly impacting image assessment.

Review of corresponding H&E‐stained sections confirmed the presence or absence of urothelial tissue in all specimens and was concordant with FCM‐based adequacy assessments. Final histopathological evaluation demonstrated non‐invasive papillary UC, high grade, in specimens classified as adequate from one patient, and non‐invasive papillary UC, low grade, in the specimen from the second patient. The non‐representative specimen consisted predominantly of blood clot, with only focal atypical papillary urothelial proliferation admixed within the clot. Representative examples of adequate and non‐representative biopsy specimens on FCM imaging, with corresponding histopathological sections, are shown in Fig. 1.

Fig. 1.

Ex vivo FCM assessment of ureteroscopic biopsy adequacy. Representative ex vivo FCM images (Histolog) of freshly obtained ureteroscopic biopsy specimens with corresponding H&E‐stained histopathological sections. (A) FCM image of a biopsy specimen classified as non‐representative, demonstrating predominantly blood clot and non‐diagnostic material without identifiable urothelial architecture. (B) Corresponding H&E section confirming limited, non‐representative tissue admixed with blood clot. (C) FCM image of a biopsy specimen classified as adequate, demonstrating identifiable urothelial tissue with preserved epithelial architecture. (D) Corresponding H&E section confirming the presence of diagnostic urothelial tissue.

Discussion

In this feasibility study, ex vivo FCM was successfully applied to all freshly obtained ureteroscopic biopsy specimens without technical limitations or disruption of specimen handling. FCM imaging enabled the reviewing pathologist to identify urothelial tissue and to classify biopsy specimens as adequate or non‐representative with high confidence. Specimen preparation and image acquisition were completed within a short time interval, with image quality sufficient for interpretative assessment in all cases. Taken together, these findings support the technical feasibility of ex vivo FCM for the assessment of ureteroscopic biopsy specimens and justify further evaluation of its potential role in real‐time biopsy adequacy assessment in the upper tract.

These findings should be interpreted within the broader diagnostic limitations of ureteroscopic tissue sampling for UTUC. Biopsies may contain limited or partially preserved urothelial tissue, a feature that has been associated with variable diagnostic accuracy and occasional misclassification in both historical and contemporary series [1, 2, 3, 4]. At present, there is no widely adopted or validated method to determine during ureteroscopy whether a retrieved fragment contains sufficient tissue for meaningful histological assessment, and specimen inadequacy typically becomes evident only after permanent section review. A technique capable of providing immediate feedback on specimen adequacy at the time of biopsy could represent a potential approach to address a recognised gap in UTUC diagnostic workflows.

The biological plausibility of confocal microscopy for assessing urothelial tissue is supported by prior in vivo studies showing that CLE reliably visualises urothelial microarchitecture. Initial feasibility studies established that in vivo confocal imaging depicts papillary structures, fibrovascular stalks, and cellular features of UTUC, with qualitative correspondence to histopathology [5]. Subsequent research has shown that confocal features related to cellular organisation and architectural patterns are reproducible when image quality is sufficient [6, 7]. Additionally, bladder confocal studies have described reproducible urothelial microarchitectural features and demonstrated consistent visualisation of epithelial patterns across urothelial tissues [10, 11]. Together, these data indicate that confocal imaging is capable of visualising key urothelial architectural features, supporting its extension from in vivo endoluminal imaging to specimen‐based applications focused on tissue identification.

Building on this rationale, ex vivo FCM has been investigated across multiple urological settings as a tool for rapid assessment of fresh tissue specimens. Most reported experience has focused on prostate cancer, where ex vivo FCM imaging has enabled real‐time evaluation of surgical margins and biopsy specimens while preserving tissue for conventional histopathological processing [12]. Exploratory applications have also been described in other urological contexts, including real‐time confirmation of muscularis propria in transurethral bladder tumour resection specimens and qualitative assessment of fresh surgical tissues without interference with downstream pathology [13, 14]. Prior confocal microscopy studies relevant to UTUC and urothelial specimen adequacy are summarised in Table 1 [5, 6, 7, 14, 15].

Table 1.

Selected confocal microscopy studies relevant to UTUC and urothelial specimen adequacy.

References	Imaging modality	Clinical setting	Specimen/target	Study focus	Key contribution
Bui et al. (2015) [5]	In vivo CLE	Ureteroscopy	UTUC lesions	Feasibility	Initial clinical feasibility study of confocal imaging in UTUC
Breda et al. (2018) [6]	In vivo CLE	Ureteroscopy	UTUC lesions	Tumour grading	Confocal features correlated with histological grade
Freund et al. (2019) [7]	In vivo CLE	Ureteroscopy	UTUC lesions	Validation/grading	Proposed scoring system for real‐time grading
Afferi et al. (2025) [15]	Ex vivo FCM	Ureteroscopy	UTUC tissue	Diagnostic characterisation/feasibility	Initial clinical experience with ex vivo confocal microscopy in UTUC
Uleri et al. (2024) [14]	Ex vivo FCM	TURBT	Bladder tumour specimens	Specimen adequacy	Real‐time confirmation of detrusor muscle

More recently, ex vivo FCM has been applied to UTUC within diagnostic paradigms aimed at tumour detection and grading during ureteroscopic evaluation [15]. In contrast, the present study adopts a deliberately narrower and complementary objective by focusing specifically on specimen adequacy rather than diagnostic classification. Within this framework, our findings suggest that ex vivo FCM imaging can support the identification of urothelial tissue in freshly obtained ureteroscopic biopsy specimens, providing a practical foundation for real‐time assessment of biopsy representativeness without precluding standard histopathological evaluation.

Rapid confirmation of urothelial tissue during specimen handling may have potential implications for ureteroscopic biopsy practice. Currently, biopsy adequacy is assessed only after permanent histopathological processing, which may necessitate repeat sampling if tissue is non‐representative. An adjunctive technique that provides immediate feedback on specimen adequacy may increase procedural confidence by confirming the presence of interpretable urothelium. Ex vivo FCM is designed to complement, not replace, conventional histopathological evaluation by enabling real‐time assessment of specimen representativeness while preserving tissue for definitive diagnosis. If validated in larger studies, this approach may help reduce non‐diagnostic biopsies and could streamline diagnostic pathways for patients evaluated for suspected UTUC.

Several limitations should be acknowledged in light of the exploratory design. The small sample size and single‐centre setting restrict generalisability and preclude assessment of diagnostic accuracy or clinical outcomes. FCM image interpretation was conducted by a single experienced genitourinary pathologist, reflecting early clinical adoption of this technology rather than a standardised multi‐reader approach. As protocols for specimen preparation and FCM image interpretation continue to evolve, these results represent an initial clinical experience. Future studies with larger cohorts will be required to assess reproducibility, facilitate workflow integration, and determine whether ex vivo FCM could eventually support diagnostic characterisation and grading once specimen adequacy is established. The present study was not designed to assess diagnostic accuracy or histological concordance, and the findings should be interpreted within the context of this feasibility cohort.

Conclusions

Ex vivo FCM enabled identification of urothelial‐containing biopsy fragments during ureteroscopic specimen processing while preserving tissue for standard histopathological evaluation. This approach may help address a practical limitation of UTUC biopsy, in which specimen adequacy is typically determined only after permanent section review. Further studies are warranted to refine adequacy criteria, assess reproducibility, and evaluate the potential role of ex vivo FCM in diagnostic characterisation in larger cohorts.

Author Contributions

Karim Daher conceived the manuscript and drafted the initial version. Nicolas A. Soputro, Mohammad El Hussein, Salim K. Younis, Lin Wang, Abdulrahman Al‐Bayati, Rui Bernardino and Mohamad Watfa contributed to manuscript review and editing. Ruben Olivares, Nima Almassi, Christopher J. Weight, Zeyad R. Schwen, Jihad Kaouk, Jane K. Nguyen, and Riccardo Autorino provided senior oversight, expert review, and critical appraisal of the manuscript.

Disclosure of Interests

Jihad Kaouk is a Consultant for Intuitive Surgical, Vascular Technology Inc., and EndoQuest robotics; Riccardo Autorino serves as a Proctor for Intuitive Surgical. The remaining authors have no disclosures.

Funding

This study received no specific funding.

Ethics Statement

This study was approved by the Institutional Review Board of the Cleveland Clinic.