THE IMPORTANCE OF MORPHOMETRIC PARAMETERS IN DIFFERENTIATING BENIGN/REACTIVE UROTHELIAL CELLS FROM LOW-GRADE UROTHELIAL CARCINOMA: COMPUTER- ASSISTED STUDY ON URINE SPECIMENS

Musayev J. , Metilli N. , Sholan R. , Hasanov A. , Damirli A. , Bakhshaliyeva K. The importance of morphometric parameters in differentiating benign/reactive urothelial cells from low-grade urothelial carcinoma: computer-assisted study on urine specimens. Azerbaijan Medical University, Republican Treatment and Diagnostic Center, Bureau of Pathological Anatomy, Unification of Forensic Medical Expertise and Pathological Anatomy, Baku, Azerbaijan, Canakkale Onsekiz Mart University, Canakkale, Turkey. ABSTRACT. Background. Urine cytology is deemed a sensitive method in detection of high-grade urothelial carcinoma. In contrast, detection of low-grade urothelial carcinoma (LGUC) and its differentiation from reactive lesions is difficult with urinary cytology. Objective. Our study aims to determine the effectiveness of morphometric parameters in differentiating reactive urothelial cells from LGUC by cytological examination of urine specimens. Methods. Voided urine samples were used for the study, while the cases were randomized into two groups: those diagnosed with LGUC (first group; N=10) and those which were not diagnosed with LGUC (second group; N=10). The morphometric parameters of major nuclear diameter (MaND), minor nuclear diameter (MiND), mean nuclear area (MNA), cell diameter (CD), mean cell area (MCA), as well as MaND/CD, MiND/CD, MiND/MaND and MNA/MCA ratios were measured on 100 urothelial cells for each case through ScopeImage® 9.0 software. Results. A statistically significant difference was found between the mean values of MiND/CD (p=0.017) and MNA/MCA (p=0.002) ratios of groups. The mean value of both parameters in the first group constituted 0.2 and higher, and below 0.2 in the second group. Conclusion. The ratios of MiND/CD and MNA/MCA in urothelial cells proved significantly higher in patients with LGUC than benign/reactive cases. The reliability of these findings in differentiating LGUC from benign/reactive lesions needs to be verified through studies examining a large number of cases. These parameters can be assessed much faster through a special software enabling an automatic measurement and thus can be used in routine cytological examination.


Introduction
Urinary cytology constitutes a significant percentage of non-gynecological cases in routine cytopathology practice, and it continues to be one of the difficult specimens for (cyto)pathologists. The inadequate cellularities of samples, cellular degeneration prior to fixation, as well as unrealistic expectations for diagnosing low-grade urothelial carcinoma (LGUC) are the main challenges encountered in urinary cytology practice (1). The ability to consistently diagnose LGUC in urine specimens remains controversial, as the reported sensitivity of urinary cytology in detection of LGUC is as low as 10% (2). Some cytological diagnostic criteria were attempted to be determined for LGUC, while McCroskey et al. reported that most of the features described previously as diagnosticwere observed almost equally in cases negative for LGUC (2).
These features of urine cytology were also reflected in the reporting system elaborated in 2015. Accordingly, The Paris System for Reporting Urinary Cytology (TPS) focused primarily on detection of high grade urothelial carcinoma (HGUC) while minimizing the detection of LGUC, since cytology has a high sensitivity of detecting the former with a poor sensitivity for the latter (1).
While the presence of three-dimensional papillary structures with fibrovascular cores is a unique diagnostic clue of LGUC in cytology, however, it is only seen in about 2% of cases. Nevertheless, this type of papillary structures can be observed in any low-grade papillary lesion, including papillomas, papillary urothelial neoplasia of low malignant potential, and LGUC. To this end, "low-grade urothelial neoplasm (LGUN)" was recommended as a diagnostic category for similar cases in TPS (1,2).

Aim
The main purpose is to determine the effectiveness of morphometric parameters in differentiating benign/reactive urothelial cells from LGUC by cytological examination of voided urine specimens.

Material and methods
Case selection: Ten urine cytology specimens of cases with histologically confirmed diagnosis of LGUC (1st group) and the urine cytology specimens of 10 benign/reactive cases (2nd group) were selected for the current study. All specimens were Papanicolaou stained conventional smears prepared from voided urine specimens through centrifugation, and each case had two slides. Cases with LGUC diagnosis were confirmed histologically within one month after collecting the urine sample. Other 10 benign/reactive cases were patients who underwent poliomavirus screening following renal transplantation, or who were diagnosed with cystitis without any history of urothelial carcinoma before and within six months after the urine sample has been collected.
Morphometric examination: A morphometry was performed on randomly selected 100 urothelial cells for each case through ScopeImage® 9.  Only non-overlapping isolated cells with wellpreserved cytoplasm and individual cells with clear cytoplasmic and nuclear borders in some clustered groups were selected for measurement. Threedimensional cell clusters, degenerative and swollen cells with vacuolar cytoplasm, as well as cells thoroughly or partially covered with inflammatory and red blood cells were excluded from the study.
Statistics: The mean value of morphometric parameters for each case, as well as the mean value and standard deviation of these parameters in each group were calculated. Independent sample t test was used for determination of the p value and p value ≤ 0.05 was considered significant.

Results
The mean value of morphometric parameters for each case, the mean value and standard deviation of these parameters for each group were shown in tables 1 and 2. The mean values of parameters such as MaND, MiND and MNA, as well as MaND/CD, MiND/CD and MNA/MCA ratios were higher in urothelial cells of LGUC cases. However, the mean values of MCA and CD were higher in benign/reactive cases. The MiND/MaND ratio presented almost a similar value in both groups.
A statistically significant difference was revealed only between the mean values of MiND/CD (p=0.017) and MNA/MCA (p=0.002) ratios of two groups. The mean value of both parameters comprised 0.2 and higher in all cases of the first group, and below 0.2 in all cases of the second group.
Prior studies on urine samples had concurred with the fact, that cytomorphometric analysis is superior to conventional urine cytology in detecting malignant cells (20,24,25). These studies stated that especially the MNA and N/C ratio are generally successful in differentiating malignant cells from benign cells (20,25

Conclusion
In our study, MiND/CD and MNA/MCA ratios of urothelial cells derived from voided urine was significantly higher in cases with LGUC than benign/reactive cases. The reliability of these findings in differentiating LGUC from benign/reactive lesions needs to be verified through studies examining a large number of cases. The most remarkable limitation of this study lies in the high number of working hours allotted for the measurement of morphometric parameters, as the measurement of all parameters on 100 cells for each case by one person encompasses a full working day. This is our humble opinion that these parameters can be assessed much faster with a special software enabling automatic measurement, which thus can be used in a routine diagnosis process.
Prospects for further development is to determine the effectiveness of morphometric parameters in differentiating cells from LGUC by cytological examination.
Information about conflicts of interest Potential or obvious conflicts of interest, related to this manuscript, at the time of publication does not exist and is not expected.