Impact of the Paris system for reporting urine cytopathology on predictive values of the equivocal diagnostic categories and interobserver agreement

Reclassification of Urinary Cytology according to the Paris System for Reporting Urinary Cytology Correlation with Histological Diagnosis

INTRODUCTION

The Paris System for Reporting Urinary Cytology (TPS) was designed to provide precise diagnostic criteria when evaluating urine cytology and standardize the terminology used in reporting. In our study, we have aimed to determine the effect of TPS on the diagnostic performance of urine cytology, its impact on establishing appropriate risk stratification, and its effectiveness in the diagnosis and follow-up of the patients.

METHODS

We reevaluated 200 liquid-based urine cytologies with available histological diagnoses reported between 2015 and 2021 according to TPS criteria and compared them with the original cytological diagnoses. Area under the curve, sensitivity, specificity, and diagnostic accuracy of both methods were calculated and statistically analyzed to determine the diagnostic performance of the original reporting and TPS.

RESULTS

The sensitivity, specificity, positive predictive, negative predictive, and diagnostic accuracy rates of TPS were 60%, 99.3%, 97.2%, 97.2%, 85.7%, and 87.2%, respectively. In TPS, the risk of malignancy for negative for high-grade urothelial carcinoma, atypical urothelial cells, suspicious for high-grade urothelial carcinoma, and high-grade urothelial carcinoma (HGUC) is 3.5%, 20.9%, 60.8%, 97.2%, respectively. In the original reporting, the corresponding risks were 13.4%, 15%, 52%, 100%, respectively. A statistically significant difference was observed between diagnostic criteria of original cytology and TPS (p = 0.001). When the original reporting was compared with the TPS, the discriminative power of TPS in the diagnosis of HGUC was significantly higher (p < 0.001).

CONCLUSIONS

The use of TPS provided a more accurate risk stratification of patients. The diagnostic performance of urine cytology was improved, especially for HGUC.

Digital and Computational Pathology Applications in Bladder Cancer: Novel Tools Addressing Clinically Pressing Needs

Bladder cancer (BC) remains a major disease burden in terms of incidence, morbidity, mortality, and economic cost. Deciphering the intrinsic molecular subtypes and identification of key drivers of BC has yielded successful novel therapeutic strategies. Advances in computational and digital pathology are reshaping the field of anatomical pathology. This review offers an update on the most relevant computational algorithms in digital pathology that have been proposed to enhance BC management. These tools promise to enhance diagnostics, staging, and grading accuracy and streamline efficiency while advancing practice consistency. Computational applications that enable intrinsic molecular classification, predict response to neoadjuvant therapy, and identify targets of therapy are also reviewed.

A review of the performance of urinary cytology with a focus on atypia, upper tract and updates on novel ancillary testing

The Paris System for Reporting Urine Cytology (TPS) is remarkable for its high predictive value in the detection of high-grade urothelial carcinoma, especially of the bladder. However, universal compliance with TPS-recommended threshold for atypical call rates (15%) and TPS performance in the rarer upper tract urothelial carcinomas (UTUC) are challenging. UTUC diagnosis is compounded by instrumentation artifacts, degenerative changes superimposed on an ambiguous cytology, difficult-to-access location, lack of specific standardized criteria, and a limited number of UTUC-focused studies. We reviewed TPS-applied studies published since 2022, noting up to 50%, exceeding the suggested 15% threshold for atypia. Our examination of ancillary tests for UTUC explored novel approaches including DNA methylation analysis, the detection of overexpressed tumor-linked messenger RNAs, and immunohistochemistry on markers such as CK17. Preliminary evidence from our review suggests that ancillary tests display superior performance over cytology, including in voided samples and low-grade urothelial carcinoma. Importantly, voided samples obviate the risks of ureterorenoscopy. Finally, we explored the future opportunities offered by artificial intelligence and machine learning for a more objective application of TPS criteria on urine samples.

Reproducibility of the Paris System in Urine Cytology by Third-Year Pathology Residents

INTRODUCTION

Standardized basic morphology and the algorithmic approach make the Paris System (TPS) for Reporting Urinary Cytology understandable and applicable. This study examined how well the TPS categories are understood by pathology residents and how well these criteria are enabling them reaching accurate diagnosis.

MATERIALS/METHODS

A hundred consecutive cases representing all categories were selected. Authors reevaluated slides using TPS regardless of their original diagnosis. In the next step, the TPS was explained to four residents and trained them by five optimal urine cytology samples from each category. Then they were asked to diagnose the selected slides according to the TPS. The diagnoses were compared to authors. The agreement was assessed using kappa. Discordant diagnoses were classified as high and low impact based on potential on clinical practice.

RESULTS

The sensitivity of authors was 62.8%, and residents' were 24-31.8%. The specificity of authors was 98.8%, and residents' were 82.3-92.8%. Reproducibility of TPS was 40-46%. Kappa values were below 0.40 except for one resident. The highest rate of concordance was for negative for high-grade urothelial carcinoma (NHGUC): authors assigned 38 NHGUC (35 biopsy-proven benign cases). Twenty to twenty-six of them were assigned as NHGUC by residents. While authors assigned 42 cases as suspicious for high-grade urothelial carcinoma (SHGUC) or high-grade urothelial carcinoma (HGUC) (35 biopsy-proven malignant cases), residents assigned 22-29 of them. Discordant diagnosis with high clinical implication was 56-63%.

CONCLUSION

Diagnostic accuracy rates of junior pathology residents using the TPS were unsatisfactory. The best agreement was observed in NHGUC and HGUC categories. Combining HGUC and SHGUC doubled the sensitivity of residents.

The Feasibility and Diagnostic Adequacy of PD-L1 Expression Analysis Using the Cytoinclusion Technique in Bladder Cancer: A Prospective Single-Center Study

Background: Programmed death-ligand 1 (PD-L1) expression has been recognized as a potential biomarker for various cancers, yet its diagnostic and prognostic significance in urothelial bladder cancer (BCa) requires further investigation. Methods: In this prospective single-center study, we aimed to assess the feasibility and diagnostic adequacy of PD-L1 expression analysis using cytoinclusion in BCa patients. We enrolled consecutive patients undergoing endoscopic transurethral resection of bladder tumor (TURBT), repeat TURBT, or robot-assisted radical cystectomy. Urinary and tissue specimens were collected from these patients for cytoinclusion and histopathological analysis to evaluate PD-L1 expression. Results: Out of 29 patients, PD-L1 expression was detected from cytoinclusion in 42.8% (3 out of 7), 10% (1 out of 10), and 66.8% (8 out of 12) of patients with negative/papilloma, low-grade, and high-grade tumors, respectively. Conversely, histopathological analysis identified PD-L1 expression in 57.2% (4 out of 7), 30% (3 out of 10), and 83.3% (10 out of 12) of patients with negative/papilloma, low-grade, and high-grade tumors, respectively. The diagnostic concordance between cytoinclusion and histopathology was 85.7%, 80%, and 83.3% in patients with negative/papilloma, low-grade, and high-grade tumors, respectively. Conclusions: Our study underscores the promise of cytoinclusion as a minimally invasive method for quantifying urinary PD-L1 percentages. This approach could serve as both a potential prognostic and diagnostic indicator, easily obtainable from urine samples. Standardizing this technique could facilitate its widespread use as a valuable tool.

Interobserver variability in cytopathology: How much do we agree?

Interobserver variability remains a major challenge for cytopathologists despite the development of standardized reporting and classification systems. Indeed, whereas moderate-to-good interobserver agreement is generally achievable when the differential diagnosis between benign and malignant entities is straightforward, high levels of variability make the diagnostic interpretation of atypical and suspicious samples not consistent. This review explores the landscape of interobserver agreement in cytopathology across different anatomical sites.

Development and validation of an artificial intelligence-based model for detecting urothelial carcinoma using urine cytology images: a multicentre, diagnostic study with prospective validation

Background

Urine cytology is an important non-invasive examination for urothelial carcinoma (UC) diagnosis and follow-up. We aimed to explore whether artificial intelligence (AI) can enhance the sensitivity of urine cytology and help avoid unnecessary endoscopy.

Methods

In this multicentre diagnostic study, consecutive patients who underwent liquid-based urine cytology examinations at four hospitals in China were included for model development and validation. Patients who declined surgery and lacked associated histopathology results, those diagnosed with rare subtype tumours of the urinary tract, or had low-quality images were excluded from the study. All liquid-based cytology slides were scanned into whole-slide images (WSIs) at 40 × magnification and the WSI-labels were derived from the corresponding histopathology results. The Precision Urine Cytology AI Solution (PUCAS) was composed of three distinct stages (patch extraction, features extraction, and classification diagnosis) and was trained to identify important WSI features associated with UC diagnosis. The diagnostic sensitivity was mainly used to validate the performance of PUCAS in retrospective and prospective validation cohorts. This study is registered with the ChiCTR, ChiCTR2300073192.

Findings

Between January 1, 2018 and October 31, 2022, 2641 patients were retrospectively recruited in the training cohort, and 2335 in retrospective validation cohorts; 400 eligible patients were enrolled in the prospective validation cohort between July 7, 2023 and September 15, 2023. The sensitivity of PUCAS ranged from 0.922 (95% CI: 0.811-0.978) to 1.000 (0.782-1.000) in retrospective validation cohorts, and was 0.896 (0.837-0.939) in prospective validation cohort. The PUCAS model also exhibited a good performance in detecting malignancy within atypical urothelial cells cases, with a sensitivity of over 0.84. In the recurrence detection scenario, PUCAS could reduce 57.5% of endoscopy use with a negative predictive value of 96.4%.

Interpretation

PUCAS may help to improve the sensitivity of urine cytology, reduce misdiagnoses of UC, avoid unnecessary endoscopy, and reduce the clinical burden in resource-limited areas. The further validation in other countries is needed.

Funding

National Natural Science Foundation of China; Key Program of the National Natural Science Foundation of China; the National Science Foundation for Distinguished Young Scholars; the Science and Technology Planning Project of Guangdong Province; the National Key Research and Development Programme of China; Guangdong Provincial Clinical Research Centre for Urological Diseases.

A Fully Automated Artificial Intelligence System to Assist Pathologists' Diagnosis to Predict Histologically High-grade Urothelial Carcinoma from Digitized Urine Cytology Slides Using Deep Learning

BACKGROUND

Urine cytology, although a useful screening method for urothelial carcinoma, lacks sensitivity. As an emerging technology, artificial intelligence (AI) improved image analysis accuracy significantly.

OBJECTIVE

To develop a fully automated AI system to assist pathologists in the histological prediction of high-grade urothelial carcinoma (HGUC) from digitized urine cytology slides.

DESIGN, SETTING, AND PARTICIPANTS

We digitized 535 consecutive urine cytology slides for AI use. Among these slides, 181 were used for AI development, 39 were used as AI test data to identify HGUC by cell-level classification, and 315 were used as AI test data for slide-level classification.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Out of the 315 slides, 171 were collected immediately prior to bladder biopsy or transurethral resection of bladder tumor, and then outcomes were compared with the histological presence of HGUC in the surgical specimen. The primary aim was to compare AI prediction of the histological presence of HGUC with the pathologist's histological diagnosis of HGUC. Secondary aims were to compare the time required for AI evaluation and concordance between the AI's classification and pathologist's cytology diagnosis.

RESULTS AND LIMITATIONS

The AI capability for predicting the histological presence of HGUC was 0.78 for the area under the curve. Comparing the AI predictive performance with pathologists' diagnosis, the AI sensitivity of 63% for histological HGUC prediction was superior to a pathologists' cytology sensitivity of 46% (p = 0.0037). On the contrary, there was no significant difference between the AI specificity of 83% and pathologists' specificity of 89% (p = 0.13), and AI accuracy of 74% and pathologists' accuracy of 68% (p = 0.08). The time required for AI evaluation was 139 s. With respect to the concordance between the AI prediction and pathologist's cytology diagnosis, the accuracy was 86%. Agreements with positive and negative findings were 92% and 84%, respectively.

CONCLUSIONS

We developed a fully automated AI system to assist pathologists' histological diagnosis of HGUC using digitized slides. This AI system showed significantly higher sensitivity than a board-certified cytopathologist and may assist pathologists in making urine cytology diagnoses, reducing their workload.

PATIENT SUMMARY

In this study, we present a deep learning-based artificial intelligence (AI) system that classifies urine cytology slides according to the Paris system. An automated AI system was developed and validated with 535 consecutive urine cytology slides. The AI predicted histological high-grade urothelial carcinoma from digitized urine cytology slides with superior sensitivity than pathologists, while maintaining comparable specificity and accuracy.

Utility of the Paris System in Urine Cytology for Improved Screening of High-Grade Urothelial Carcinoma in Bahrain

BACKGROUND

Urothelial carcinoma, a prevalent and aggressive urological malignancy, necessitates early detection for improved prognosis. Urine cytology serves as a cost-effective screening tool, but inconsistencies in reporting due to the lack of standardized criteria limit its efficacy. The Paris System for reporting urinary cytology (TPS) was introduced to address this issue, aiming to improve diagnostic accuracy. This retrospective study investigates the effectiveness of urine cytology in detecting high-grade urothelial carcinoma (HGUC) using TPS classification, specifically focusing on atypical urothelial cells (AUC) categorized as TPS-III and suspicious for high-grade urothelial carcinoma (SHGUC) categorized as TPS-IV.

METHODS

We reviewed 470 urine cytology samples collected over two years at a tertiary healthcare center in Bahrain. All samples were re-evaluated using TPS classification by two independent consultant cytopathologists blinded to the original cytology report. The analysis included only samples categorized as TPS-III or TPS-IV with corresponding histopathology reports from confirmatory biopsies performed within four months of urine collection. Biopsy results were categorized as either benign/low-grade urothelial carcinoma (non-HGUC) or malignant (HGUC). The positive predictive value (PPV) of urine cytology for HGUC detection was calculated for both TPS-III and TPS-IV categories. Statistical significance was assessed using Fisher's exact test.

RESULTS

Among the 470 urine cytology samples, 40 (8.5%) were classified as TPS-III or TPS-IV. Within this subset, 16 patients underwent confirmatory biopsies. Histopathological analysis revealed HGUC in 12 (75%) patients and non-HGUC (benign or low-grade) in 4 (25%) patients. The PPV of TPS-III for HGUC was 50%, while TPS-IV demonstrated a higher PPV of 90%. However, the difference between these values was not statistically significant (p = 0.25). This study explored the utility of TPS classification in urine cytology for HGUC detection. While SHGUC (TPS-IV) exhibited a numerically higher PPV compared to AUC (TPS-III), the lack of statistical significance necessitates further investigation. Our findings highlight the potential of TPS to improve the accuracy of urine cytology. TPS implementation has been shown to reduce the number of inconclusive "atypical" diagnoses, leading to more targeted investigations.

CONCLUSION

Our study suggests that SHGUC (TPS-IV) within TPS classification framework might hold promise as a more specific indicator for HGUC compared to AUC (TPS-III). However, further research with larger cohorts is necessary to definitively establish the clinical significance of this observation. This investigation paves the way for future studies exploring the potential of TPS, particularly the SHGUC category, as a reliable screening tool for HGUC, potentially leading to earlier diagnoses and improved patient outcomes.

International Society of Urological Pathology (ISUP) Consensus Conference on Current Issues in Bladder Cancer. Working Group 1: Comparison of Bladder Cancer Grading System Performance

Grade is a key prognostic factor in determining progression in nonmuscle invasive papillary urothelial carcinomas. The 2 most common grading methods in use worldwide are the World Health Organization (WHO) 2004 and 1973 schemes. The International Society of Urological Pathology (ISUP) organized the 2022 consensus conference in Basel, Switzerland on current issues in bladder cancer and tasked working group 1 to make recommendations for future iterations of bladder cancer grading. For this purpose, the ISUP developed in collaboration with the European Association of Urology a 10-question survey for their memberships to understand the current use of grading schemes by pathologists and urologists and to ascertain the areas of potential improvements. An additional survey was circulated to the ISUP membership for their opinion on interobserver variability in grading, reporting of urine cytology, and challenges encountered in grade assignment. Comprehensive literature reviews were performed on bladder cancer grading prognosis and interobserver variability along with The Paris System for urine cytology. There are notable differences in practice patterns between North American and European pathologists in terms of used grading scheme and diagnosis of papillary urothelial neoplasm of low malignant potential. Areas of common ground include difficulty in grade assignment, a desire to improve grading criteria, and a move towards subclassifying high-grade urothelial carcinomas. The surveys and in-person voting demonstrated a strong preference to refine current grading into a 3-tier scheme with the division of WHO 2004 high grade into clinically relevant categories. More variable opinions were voiced regarding the use of papillary urothelial carcinoma with low malignant potential.

Are we on track for diagnosing high-grade urothelial carcinoma with a minimum quantity of five malignant cells in lower tract specimens? Critical analysis of The Paris System Quantitation Criteria

BACKGROUND

The Paris System for Reporting Urinary Cytology (TPS) has gained universal acceptance as the standard for reporting urine cytology requiring at least 5-10 malignant cells to diagnose high-grade urothelial carcinoma (HGUC) in lower and upper urinary tract specimens, respectively. These quantitation criteria are still subject to discussion, and this study specifically aims to validate the quantitation criterion of HGUC in lower urinary tract.

DESIGN

The authors reviewed two cohorts of lower urinary tract cases. The first cohort consisted of 100 liquid-based ThinPrep slides with the diagnosis of HGUC having positive histology on concurrent or follow-up biopsies within 3 months. The second cohort was 36 HGUC cases with negative histology on concurrent biopsies and within 3 months. The number of high-grade cells (HGCs) meeting the TPS qualitative criteria were counted under the light microscope driven in a grid-like manner.

RESULTS

The first 100 urine samples showed five cases (5.0%) with three HGCs, three cases (3.0%) had four HGCs, five cases (5.0%) showed five HGCs, and 25 cases (25.0%) had between 6-10 HGCs. The risk of high-grade malignancy (ROHM) in cases with five or more HGCs was 100%, whereas those with three HGCs was 60.0%. The second cohort of HGUC was considered "positive" despite a negative histology.

CONCLUSION

This study confirms that quantitation is an essential key to diagnose HGUC. The current TPS criterion of a minimum of five malignant cells in lower tract is robust with a ROHM of 100%. Diagnosing HGUC with less than five HGCs runs the risk of lowering the ROHM.

The Role of Checkpoint Inhibitor Expression Directly on Exfoliated Cells from Bladder Cancer: A Narrative Review

Bladder cancer (BCa) is a common type of cancer that affects the urinary bladder. The early detection and management of BCa is critical for successful treatment and patient outcomes. In recent years, researchers have been exploring the use of biomarkers as a non-invasive and effective tool for the detection and monitoring of BCa. One such biomarker is programmed death-ligand 1 (PD-L1), which is expressed on the surface of cancer cells and plays a crucial role in the evasion of the immune system. Studies have shown that the PD-L1 expression is higher in BCa tumors than in healthy bladder tissue. Additionally, PD-L1 expression might even be detected in urine samples in BCa patients, in addition to the examination of a histological sample. The technique is being standardized and optimized. We reported how BCa patients had higher urinary PD-L1 levels than controls by considering BCa tumors expressing PD-L1 in the tissue specimen. The expression of PD-L1 in urinary BCa cells might represent both a diagnostic and a prognostic tool, with the perspective that the PD-L1 expression of exfoliate urinary cells might reveal and anticipate eventual BCa recurrence or progression. Further prospective and longitudinal studies are needed to assess the expression of PD-L1 as a biomarker for the monitoring of BCa patients. The use of PD-L1 as a biomarker for the detection and monitoring of BCa has the potential to significantly improve patient outcomes by allowing for earlier detection and more effective management of the disease.

Large-scale validation study of an improved semiautonomous urine cytology assessment tool: AutoParis-X

BACKGROUND

Adopting a computational approach for the assessment of urine cytology specimens has the potential to improve the efficiency, accuracy, and reliability of bladder cancer screening, which has heretofore relied on semisubjective manual assessment methods. As rigorous, quantitative criteria and guidelines have been introduced for improving screening practices (e.g., The Paris System for Reporting Urinary Cytology), algorithms to emulate semiautonomous diagnostic decision-making have lagged behind, in part because of the complex and nuanced nature of urine cytology reporting.

METHODS

In this study, the authors report on the development and large-scale validation of a deep-learning tool, AutoParis-X, which can facilitate rapid, semiautonomous examination of urine cytology specimens.

RESULTS

The results of this large-scale, retrospective validation study indicate that AutoParis-X can accurately determine urothelial cell atypia and aggregate a wide variety of cell-related and cluster-related information across a slide to yield an atypia burden score, which correlates closely with overall specimen atypia and is predictive of Paris system diagnostic categories. Importantly, this approach accounts for challenges associated with the assessment of overlapping cell cluster borders, which improve the ability to predict specimen atypia and accurately estimate the nuclear-to-cytoplasm ratio for cells in these clusters.

CONCLUSIONS

The authors developed a publicly available, open-source, interactive web application that features a simple, easy-to-use display for examining urine cytology whole-slide images and determining the level of atypia in specific cells, flagging the most abnormal cells for pathologist review. The accuracy of AutoParis-X (and other semiautomated digital pathology systems) indicates that these technologies are approaching clinical readiness and necessitates full evaluation of these algorithms in head-to-head clinical trials.

Iodinated Contrast Medium Affects Urine Cytology Assessment: A Prospective, Single-Blind Study and Its Impact on Urological Practice

During endoscopic procedures for suspected urothelial tumors of the upper urinary tract, radiographic imaging using an iodinated contrast medium is often required. However, following ureteropyelography, we detected changes in cytology characteristics not correlating with real cytology findings in naive urine. The aim of our study was to assess cytology changes between naive and postcontrast urine according to The Paris System of cytology classification. Methods: We prospectively assessed urine samples from 89 patients (23 patients with histologically proven urothelial cancer and 66 healthy volunteers). The absence of malignancy was demonstrated by CT urography and/or ureteroscopy. The study was single blind (expert cytopathologist) and naïve Paris system for urine cytology assessment was used. Furthermore, additional cytological parameters were analyzed (e.g., specimen cellularity, degree of cytolysis, cytoplasm and nucleus color, chromatin and nucleo-cytoplasmic ratio). Results: Our study showed statistically significant differences when comparing naïve and postcontrast urine in healthy volunteers (only 51 % concordance, p = 0.001) versus malignant urine specimens (82 % concordance). The most important differences were in the shift from The Paris System category 2 (negative) to 1 (non-diagnostic) and from category 2 (negative) to 3 (atypia). Other significant changes were found in the assessment of specimen cellularity (p = 0.0003), degree of cytolysis (p = 0.001), cytoplasm color (p = 0.003), hyperchromasia (p = 0.001), course chromatin (p = 0.002), nucleo-cytoplasmatic ratio (p = 0.001) and nuclear borders’ irregularity (p = 0.01). Conclusion: Our unique study found crucial changes in the cytological assessment of naive and postcontrast urine and we confirm that postcontrast urine is more often assessed as abnormal, suspect or non-diagnostic. Therefore, before urine collection for cytology, the clinician should avoid administration of iodinated contrast into the urinary tract.

Cytospin performance when using Paris system for reporting urinary cytology

Objectives:

The Paris System (TPS) for Reporting Urine Cytology has significantly improved the approach to evaluating urine cytology. TPS criteria were defined mainly according to ThinPrep and SurePath preparations, as they are widely utilized. The objective of this study is to validate urine cytology interpretation according to the TPS classification using cytospin technique in relation to the gold slandered histology.

Material and Methods:

This retrospective study examined and analyzed 316 urine specimens from King Abdulaziz University Hospital between 2015 and 2020. Cytospin technique is performed for all cases. Slides were recategorized using TPS criteria, then compared with the original histology diagnosis.

Results:

According to the TPS, 108 cases were classified as 101 AUC (32%), 95 NEG (30%), 59 HGUC (18.7%), 31 SHGUC (9.8%), and 30 (9.5%) others. The computed sensitivity of cytospin in urine cytology was 94.7%, with 73.9% specificity, a positive predictive value of 85.6%, a negative predictive value of 89.5%, and overall accuracy of 86.8%.

Conclusion:

Urine cytology testing is considered to be a non-invasive and sensitive method to screen for urothelial carcinoma. TPS defined standards are reliable on cytospin prepared slides for reporting urine cytology.

Evaluation of an artificial intelligence algorithm for assisting the Paris System in reporting urinary cytology: A pilot study

BACKGROUND

The Paris System for Reporting Urinary Cytology (TPS) has been shown to improve bladder cancer diagnosis. Advances in artificial intelligence (AI) may assist and improve the clinical workflow by applying TPS in routine diagnostic services.

METHODS

A deep-learning-based algorithm was developed to identify urothelial cancer candidate cells using whole-slide images (WSIs). In the testing cohort, 131 urine cytology slides were retrospectively retrieved and analyzed using this AI algorithm. The authors compared the performance of one cytopathologist and two cytotechnologists using AI-assisted digital urine cytology. Then, the AI-assisted WSIs were evaluated in the clinical workflow. The cytopathologist first made a diagnosis by reviewing the AI-inferred WSIs and quantitative data (nuclear-to-cytoplasmic ratio and nuclear size) for each sample. After a washout period, the same cytopathologist made a diagnosis for the same samples using direct microscopy. All diagnosis results were compared with the expert panel consensus.

RESULTS

The AI-assisted diagnosis by the two cytotechnologists and the one cytopathologist demonstrated performance results that were comparable to the expert panel consensus (sensitivity, 79.5% and 82.1% vs. 92.3%, respectively; specificity, 100% and 98.9% vs. 100%, respectively). Furthermore, the performance of the AI-assisted WSIs compared with the microscopic diagnosis by the cytopathologist demonstrated superior sensitivity (92.3% vs. 87.2%) and negative predictive value (96.8% vs. 94.8%). In addition, the AI-assisted reporting demonstrated near perfect agreement with the expert panel consensus (κ = 0.944) and the microscopic diagnosis (κ = 0.862).

CONCLUSIONS

The AI algorithm developed by the authors effectively assisted TPS-based reporting by providing AI-inferred WSIs and quantitative data.

[How has the Paris System contributed to urine cytology? Evaluating the contribution of the Paris System to urine cytology. A comparative study of the Paris System and the Papanicolaou method in a tertiary centre]

INTRODUCTION AND OBJECTIVES

The Paris System (PS) has replaced the classical Papanicolaou System (PapS) in reporting urine cytology, due to its improved sensitivity and negative predictive value (NPV) without loss of specificity. Furthermore, it has enabled the risk of malignancy to be established in each cytological category. The aim of this study is to compare the Paris System with previous results and determine the changes in sensitivity, specificity, positive predictive value, NPV and risk of malignancy in our centre, MATERIALS AND METHODS: Evaluation of the diagnostic power of urine cytology by means of a retrospective cohort study, comparing two series of 400 cytological studies, one using the Papanicolaou System and the other the Paris System.

RESULTS

In the detection of high-grade urothelial carcinoma, Paris System has better specificity (93.82% PapS vs 98.64% PS; P=.001) and PPV (39.5% PapS vs 70.6% PS; P=.044) than Papanicolaou System, without changes in sensitivity (53.5% PapS vs 37.5% PS; P=.299) or NPV (96.4% PapS vs 94.8% PS; P=.183). The risk of malignancy for the atypical category increases from low to high levels (1.6% PapS vs 40.0% PS; P=.001); the other categories showed no significant statistical changes.

CONCLUSION

The Paris System improves specificity and positive predictive value and establishes a better indication of risk of malignancy for each category, enabling specific clinical management in each case.

Comparison of Diagnostic Performances of Urine Cytology Before and After the Use of The Paris System Criteria: An Institutional Experience from Turkey

Background:

Urine cytology remains to be the test of choice in the detection of high-grade urothelial carcinomas (HGUC) due to its favorable sensitivity. However, a significant rate of cases is reported under atypical/indeterminate categories, which result in a decrease in its specificity. Providing standardized cytologic criteria, one of the aims of The Paris System (TPS) is to reduce the use of indeterminate diagnoses and provide a higher predictive value in these categories.

Aims:

We compared the diagnostic performances of TPS and our original reporting system, and also investigated the interobserver reproducibility of the cytologic criteria used.

Materials and Methods:

A total of 386 urine samples were reviewed retrospectively. Original cytologic diagnoses have been made using similar cytologic features proposed by TPS. All slides were recategorized after the use of the cytologic criteria as described by TPS guideline.

Results:

After TPS, specificity of the test increased from 39.6% to 63.5, sensitivity decreased from 92.5% to 88.8%, and diagnostic accuracy increased from 63.6% to 75%. The use of negative category increased threefold. Frequencies of indeterminate categories of atypical urothelial cells (AUC) and suspicious for HGUC (SHGUC) decreased by 36% and 56.5%, respectively. A subsequent detection of HGUC after AUC and SHGUC categories increased by 38% and 64%, respectively. Interobserver agreement for TPS categorization was 39%.

Conclusions:

TPS improved diagnostic accuracy of urine cytology by reducing the use of indeterminate categories, and resulted in increase in their predictive value for subsequent diagnosis of HGUC. However, reproducibility of diagnostic categories seemed to be imperfect.

Keratin 17 Is a Novel Cytologic Biomarker for Urothelial Carcinoma Diagnosis

OBJECTIVES

The microscopic features of urine cytology specimens are subjective and may not reliably distinguish between benign urothelial cells and low-grade urothelial carcinoma (UC). Prior studies demonstrated that keratin 17 (K17) detection in biopsies is highly sensitive for UC. The current study aimed to define K17 diagnostic test performance for initial screening and detect recurrent UC in urine specimens.

METHODS

K17 was detected by immunocytochemistry (ICC) in consecutively collected urine specimens (2018-2019). A qualitative score for the K17 test was determined in 81 samples (discovery cohort) and validated in 98 samples (validation cohort). K17 sensitivity and specificity were analyzed in both cohorts across all grades of UC.

RESULTS

Based on the discovery cohort, the presence of 5 or more K17 immunoreactive urothelial cells (area under the curve = 0.90; P < .001) was the optimal threshold to define a K17-positive test. The sensitivity of the K17 ICC test for biopsy-confirmed UC was 35 of 36 (97%) and 18 of 21 (86%) in the discovery and validation cohorts, respectively. K17 was positive in 16 of 19 (84%) specimens with biopsy-confirmed low-grade UC and in 34 of 34 (100%) of specimens with high-grade UC.

CONCLUSIONS

K17 ICC is a highly sensitive diagnostic test for initial screening and detection of recurrence across all grades of UC.

Deep learning based digital cell profiles for risk stratification of urine cytology images

Urine cytology is a test for the detection of high-grade bladder cancer. In clinical practice, the pathologist would manually scan the sample under the microscope to locate atypical and malignant cells. They would assess the morphology of these cells to make a diagnosis. Accurate identification of atypical and malignant cells in urine cytology is a challenging task and is an essential part of identifying different diagnosis with low-risk and high-risk malignancy. Computer-assisted identification of malignancy in urine cytology can be complementary to the clinicians for treatment management and in providing advice for carrying out further tests. In this study, we presented a method for identifying atypical and malignant cells followed by their profiling to predict the risk of diagnosis automatically. For cell detection and classification, we employed two different deep learning-based approaches. Based on the best performing network predictions at the cell level, we identified low-risk and high-risk cases using the count of atypical cells and the total count of atypical and malignant cells. The area under the receiver operating characteristic (ROC) curve shows that a total count of atypical and malignant cells is comparably better at diagnosis as compared to the count of malignant cells only. We obtained area under the ROC curve with the count of malignant cells and the total count of atypical and malignant cells as 0.81 and 0.83, respectively. Our experiments also demonstrate that the digital risk could be a better predictor of the final histopathology-based diagnosis. We also analyzed the variability in annotations at both cell and whole slide image level and also explored the possible inherent rationales behind this variability.

Effect of the Paris system for reporting urinary cytology with histologic follow-up

BACKGROUND

The Paris system (TPS) for Reporting Urinary Cytology provides a standardized reporting system whose main focus is the diagnosis of high-grade urothelial carcinoma (HGUC). We conducted a study to see the impact of The Paris System on our cytologic diagnoses with associated histology.

MATERIALS AND METHODS

We reviewed our pathology database regarding urinary specimens in the year before implementation of The Paris System and the year after. We gathered the data regarding cytologic diagnosis and concurrent/subsequent histology.

RESULTS

Over a 1-year period from 2016-2017, 486 urine cytology specimens were identified before implementation of The Paris System and diagnosed as follows: 83% benign/negative, 10% atypical, 2% suspicious, 5% HGUC, 0.2% low grade urothelial neoplasm (LGUN), and 0.2% unsatisfactory. Over a next 1-year period from 2017 to 2018, 602 specimens used TPS and diagnosed as follows: 85% negative for HGUC, 6% atypical, 3% suspicious, 4% HGUC, 0.17% LGUN, and 2% unsatisfactory. Although, not listed as a standardized category in The Paris System, our institution used "Negative for high-grade, cannot rule out low-grade urothelial neoplasm (NHL)" as a subcategory of Negative for HGUC. 4% of the cases fell into this category. Focusing on the Atypical category before TPS, histology was available in 15/49 (31%) cases. Of these, 40% had HGUC. Regarding the Atypical category after TPS, histology was available in 21/36 (58%) cases. Of these, 52% were HGUC. For the NHL category, concurrent histology was available in 13/26 (50%) cases. Of these, 67% were low grade urothelial neoplasms.

CONCLUSION

Our study showed that TPS lowered the rate of Atypical from 10% to 6%. After the implementation of TPS, Atypical corresponded to a higher rate of high-grade urothelial carcinoma. Also, the NHL subcategory had a high positive predictive value for diagnosing low grade urothelial neoplasms.

Urinary tract cytology: a cytologic-histopathologic correlation with The Paris System, an institutional study

INTRODUCTION

Urothelial carcinoma (UC) requires lifelong monitoring, commonly through urinary cytology and cystoscopy. Urine cytology has a relatively high sensitivity for detecting high-grade urothelial carcinoma (HGUC); however, its sensitivity for low-grade urothelial neoplasm (LGUN) is significantly lower with wide interobserver variability. The Paris System (TPS) was proposed to create standardized diagnostic categories with defined cytomorphologic criteria. We attempt to evaluate diagnostic efficacy of identifying UC using TPS through cytologic-histologic correlation.

MATERIALS AND METHODS

A retrospective search identified 170 cases of urine cytology cases with concurrent biopsies collected during a 2-year time period at University of Rochester Medical Center. Patient age, sex, smoking history, prior malignancy diagnoses, cystoscopy findings, specimen collection method, UroVysion results, and 1-year follow-up of surgical pathology cases were included.

RESULTS

Cytologic-histologic correlation was identified in 59% of cases, with 18% true positives and 41% true negatives. Discordant results were identified in 41% of cases; of these, 4% were false positives, 11% false negatives, 12% potential sampling bias, and 14% were low-grade urothelial carcinoma (LGUC). The analysis of this 2-year study finds a positive predictive value of urine cytology for HGUC to be 81%, a negative predictive value of 79%, a sensitivity of 61%, a specificity of 91%, and an accuracy of 79%.

CONCLUSIONS

Our results support TPS's ability to improve the reliability and accuracy of interpretations in urine cytology for HGUC. Nevertheless, additional studies are essential to improve the diagnostic accuracy of LGUN, and urine adequacy, in order to improve patient care and early detection, while identifying potential sampling bias.

Experience on the use of The Paris System for Reporting Urinary Cytopathology: review of the published literature

INTRODUCTION

The Paris System for Reporting Urinary Cytology (TPS) was first published in 2016 with clear objectives to standardize cytologic diagnostic criteria and provide uniform reporting, in order to improve patient stratification and associated clinical management. The aim of this paper is to evaluate the performance of TPS and review the literature published since TPS was introduced.

MATERIALS AND METHODS

Original articles focusing on the utilization and performance of TPS in urinary cytology specimens were identified using PubMed for publications from January 2016 to July 2020, using the keywords "Paris System", "urine cytology", and "urinary cytology".

RESULTS

Twenty-three relevant articles in the literature regarding the use of TPS were included in the review from a total of 30,802 urine cytology specimens, of which 21,485 (69.8%) had available diagnoses. Distribution of cases among categories ranged from 50.5% to 95.3% for negative for high-grade urothelial carcinoma (NHGUC), 1.2% to 23% for atypical urothelial cells (AUC), 0.2% to 6.6% for suspicious for high-grade urothelial carcinomas (SHGUC), and 2.2% to 14.1% for high-grade urothelial carcinomas (HGUC). The calculated risk of high-grade malignancy (ROHM) ranged from 8.7% to 36.8% for NHGUC, 12.3% to 60.9%% for AUC, 33.3% to 100% for SHGUC, and 58.8% to 100% for HGUC. Mean ROHM weighted by sample size was calculated at 15.7% (±7.8%), 38.5% (±14.3%), 76.2% (±17.2%), and 88.8% (±12.7%) for NHGUC, AUC, SHGUC, and HGUC, respectively. Reported sensitivity of TPS ranged from 40% to 84.7%, specificity from 73% to 100%, PPV from 62.3% to 100%, and NPV from 46% to 90%.

CONCLUSIONS

The application of TPS in the selected series has improved the screening and surveillance potential of urine cytology, while reducing high rates of indeterminate diagnoses, improving sensitivity and providing proper risk stratification for patients.

Should "suspicious for high-grade urothelial carcinoma" and "positive for high-grade urothelial carcinoma" remain separate categories?

BACKGROUND

The Paris System (TPS) for Reporting Urinary Cytology aims to standardize urine cytology reporting. Per TPS, the diagnosis of "suspicious for high-grade urothelial carcinoma (SHGUC)" is applied in cases that have few urothelial cells with severe atypia but are quantitatively insufficient for a diagnosis of "high-grade urothelial carcinoma (HGUC)." In our study, we compared the diagnostic accuracy and risk of malignancy (ROM) of these 2 categories to assess whether they could be combined in clinical practice to perhaps improve overall interobserver variability.

METHODS

All urine specimens with a diagnosis of either SHGUC or HGUC from January 2016 to July 2019 were retrieved from the pathology database of 2 large academic institutions. Only cases with follow-up biopsies within 6 months were included.

RESULTS

One hundred eighty-nine cases met the study criteria. Of these, 122 had a cytologic diagnosis of SHGUC, and 67 had a diagnosis of HGUC. Ninety-five (78%) cases from the SHGUC group and 64 (96%) cases from the HGUC group had biopsy-proven HGUC. The majority of cases with discordance had a history of treatment with either intravesical bacillus Calmette-Guérin or mitomycin. The difference in the rate of biopsy-proven HGUC between the SHGUC category and the HGUC category (95/122 vs 64/67, respectively) was statistically significant (P < .001).

CONCLUSIONS

The difference in ROM between SHGUC and HGUC was statistically significant in our study cohort. Intravesical chemotherapy was frequently observed in negative biopsy cases in both groups. Our preliminary findings suggest that the 2 TPS categories should remain separate.

Interobserver reproducibility of The Paris System of Reporting Urine Cytology on cytocentrifuged samples

BACKGROUND

The Paris System of Reporting Urine Cytology aims to screen samples for HGUC and to provide a universally acceptable reporting format for urine cytology specimens. However, studies detailing the reproducibility of this system, especially on cyto-centrifuge preparations, are lacking.

METHODS

149 voided urine samples received in Department of Pathology were reviewed independently by five cytopathologists. To estimate the overall agreement, Gwet's AC and AC1statistics between each pair of raters were used. To measure the random error component, polychoric correlations were used. To assess the systematic error, Friedman test was used.

RESULTS

There was moderately good inter-rater agreement between the raters. Gwets AC2 ranged between 0.67 and 0.89 for the classification of the cases once the sample was found adequate for assessment, while the Gwet's AC1 ranged between 0.61 and 0.94 in assessing for adequacy. There were significant systematic differences between raters in their thresholds for the different categories as well as in differentiating between an adequate and inadequate sample (P value by Friedman test <.001). The association between pathologists was moderately high (polychoric correlations ranging from 0.67 to 0.93). In the majority (108 of 149, 72.5%) of the cases, the range of differences between raters were of one category or less, suggesting satisfactory agreement, but having large disagreements in minority.

CONCLUSION

The interobserver reproducibility for the Paris System is moderately good, and is suitable for adoption. It is limited by the lack of agreement as to what constitutes an adequate specimen and differing threshold for categorizing the lesions in differing groups.