Atypical urothelial cells classified according to the Paris System for Reporting Urinary Cytology: A 2‐year experience with histological correlation from a Finnish tertiary care center—low rate and high risk of malignancy

The Paris System for Reporting Urinary Cytology (TPS) was issued to shift the focus of urine cytology to high‐grade lesions to increase the diagnostic accuracy of urine cytology. The aim of this study was to evaluate the power of TPS in the atypical urothelial cells (AUC) category with histological correlation and follow‐up.


INTRODUCTION
The Paris System for Reporting Urinary Cytology (TPS) was released in 2016. TPS aims to standardize the cytologic diagnostic criteria in urine cytology and focuses on high-grade malignancies to improve diagnostic output in cytology. TPS contains six diagnostic categories: negative for high-grade urothelial carcinoma (NHGUC), atypical urothelial cells (AUC), suspicious for high-grade urothelial carcinoma (SHGUC), high-grade urothelial carcinoma (HGUC), low-grade urothelial neoplasm (LGUN), and other malignances. Each category has strictly defined cytomorphological criteria, to ensure that the cytological evaluation is homogeneous and reproducible. 1 The AUC category is frustrating for clinicians and especially for patients. On histological follow-up, it includes both benign and malignant entities. 2 Reviews by Pastorello et al. 3 and Cowan and Van-denBussche 4 found that the number of atypical cytological samples decreased in most studies after the introduction of TPS. However, in one study, the number conversely increased from 3% to 24.2%. 5 Our study is based on cases from a Finnish tertiary care university hospital and focuses on AUC category of TPS with cytohistological correlation and a follow-up of 2 years.

MATERIALS AND METHODS
All urine cytology samples were fixed in 50% ethanol, cytospun, and stained by Papanicolaou stain. The samples were first screened by a cytotechnologist, after which a cytopathologist evaluated the samples and signed them out. Three cytopathologists, each with at least 20 years of experience, signed out all the non-NHGUC cases. Only the NHGUC cases were signed-out by general pathologists. All cases were diagnosed at the Department of Pathology, Fimlab Laboratories, Tampere, Finland, during a 2-year period (January 2017-December 2018). The cytological diagnoses were made prospectively according to TPS, and the timeframe for histological follow-up lasted until 2019. The time spans between cytological and histological diagnoses were also calculated. Statistical analysis was performed using IBM SPSS Statistics 27 (IBM, Armonk, New York). This study was conducted according to the Declaration of Helsinki and approved by the Regional Ethics Committee of Pirkanmaa Hospital District (R17174). Informed consent from patients was not requested.
Of the 205 samples diagnosed as AUC, 34 cases (16.6%) had no cytological or histological follow-up in the 0.5-to 2.5-years period  Table 2.
At least one histological follow-up sample was available in 97 (47.3%) of the cases. In total, there were 152 histological follow-up samples. The first biopsy was taken within 1 month of the AUC diagnosis in 18 (8.8%) cases, between 1 and 2 months after diagnosis in 33 (16.1%) cases, and after 2 months in the rest (22.4%). The longest time between AUC diagnosis and follow-up histological sample was 25 months (one case).
The distribution of histological diagnoses is presented in   (Figure 2). Cytohistological correlation of an AUC case with final histopathological diagnosis of high-grade noninvasive papillary urothelial carcinoma is illustrated in Figure 3.
There were 36 (17.6%) benign histological diagnoses. The "no tumor" category contained both normal histological findings and various reactive and inflammatory changes (see Table 3). Cytohistological correlation of an AUC case with final histopathological diagnosis of inflammation is illustrated in Figure 4. The inflammation was granulomatous in four (2.0%) cases, acute in one (0.5%) case, and chronic in three (1.5%) cases. There were seven (3.4%) cases with metaplastic changes ( Table 3). -577

DISCUSSION
George Papanicolaou introduced the term atypia/atypical into cytology without a strict definition. As attempts at terminological standardization (e.g., TPS) progress, the term "atypia" has come to succeed due to well-defined criteria and ROM determination as well as close follow-up management. 2,6 The main goal of TPS is to reduce the atypia rate in urine cytology at the individual, laboratory, and community levels and consequently to increase clinicians' and the patients' confidence in cytological diagnoses. 2,6 During the 2-year prospective study period, our rate of AUC diagnoses in urine cytology was 5.5%, which is considered good performance. In the second edition of TPS, the suggested frequency of AUC diagnosis is 5%-15%, 7 so our results are in the lower limit of the range. All our specimens, except the negative ones, were signed out by three cytopathologists. This may be one reason for the low rate of AUC diagnosis, because it is easier to keep the criteria strict and the threshold for consultations low among only a few cytopathologists.
In the pre-TPS era, there was high variability in the use of the atypia category with some studies reporting rates as high as 54.3% 8 and 59.8%. 9 A massive literature analysis of 30,802 cases produced a mean value of 23.2%, with a range of 4.7%-59.8%. 3 Importantly, the highest reported rates of atypia diagnoses each decreased in the post-TPS era to 8.5% and 41.5%, respectively. 8 were also found in benign and low-grade neoplasms and this consequently led to a 3.1-fold increase in false-positive cases. 5 The study failed to recognize that according to TPS AUC diagnosis requires more than just an elevated N/C ratio. A recent study importantly pointed out that enlarged nuclear size and the presence of nucleolus are also important diagnostic features in AUC cases. 12 Several studies have investigated N/C ratio reproducibility. In one study, interobserver variance decreased with increased N/C ratio. 13 Similarly, Zhang et al. 14  12.5% as decreased from the pre-TPS rate of 46%, 20 14.7%, 11 and 20%. 23 The percentage of benign findings among AUC cases is also variable, having been reported as 5%, 20 11.6%, 11 18.6% (present study), and 30%. 23 Importantly, there was a steep decrease in benign follow-up diagnoses in the AUC cases, from 53% in the pre-TPS era to 5% in a recent study. 20 The main limitations of the present study are the sample-based (rather than patient-based) analysis and the limited follow-up time of 0.5-2.5 years. Nevertheless, the data are robust in volume and represent unselected material from community health care centers and from regional and university hospitals.
TPS has been endorsed by clinicians 24