Comparison of Conventional Sediment Smears and Liquid-Based Cytology and Utility of Cell Block Preparation in Urinary Cytology.

INTRODUCTION

The Paris System for Reporting Urinary Cytology (TPS) provides a non-invasive screening tool for high-grade urinary bladder carcinomas. In resource-poor settings, urinary cytology continues to be performed on conventional sediment smears (CSS) only. Cell blocks (CBs), though an essential part of cytology, are not part of routine protocol for urinary cytology. The present study was undertaken to compare CSS and liquid-based cytology (LBC) in urinary cytology, and to assess the utility of CB preparation in decreasing the indeterminate/grey zone diagnosis (Non-diagnostic/Atypical Urothelial Cells/Suspicious for High-Grade Urothelial Carcinoma).

MATERIALS AND

METHODS

This prospective study evaluated 1051 urine samples from 417 patients. All these samples were processed using CSS and LBC. Diagnostic parameters (sensitivity/specificity/positive predictive value [PPV]/negative predictive value [NPV]/diagnostic accuracy) were analyzed for each preparation with their individual and combined comparisons. CBs were also prepared for each sample.

RESULTS AND DISCUSSION: For all urine samples in terms of percentage reduction of indeterminate diagnosis, results of CSS alone (29.88% of indeterminate diagnosis) versus LBC alone (30.83% of indeterminate diagnosis) showed no statistically significant difference (p-value 0.468). When the findings were combined for CSS with LBC, the percentage of indeterminate diagnosis was reduced to 27.12%. When the combined results were compared with the findings of CSS alone, the results were statistically significant (p-value < 0.001), and when the combined results were compared with the findings of LBC alone, the results were statistically insignificant (p-value 0.68). Follow-up histopathological diagnosis and/or relevant radiology and/or positive CBs were considered the gold standard, which was available for 620 cases. Diagnostic parameters, along with the Risk of High-Grade Malignancy (ROHM) for these 620 cases, were calculated. Evaluation of the CB reduced the grey zone diagnosis in four cases and helped to reach a definitive diagnosis. However, it was statistically insignificant (p-value 0.125).

CONCLUSION

CSSs give as good and comparable cytological results as LBC. Though LBC provides a more determinate diagnosis (Negative for High-Grade Urothelial Carcinoma or High-Grade Urothelial Carcinoma) when compared to CSS, this difference is not statistically significant. A combination of LBC with CSS provides superior results than CSS or LBC alone. This study highlights the promising results CSSs provide, which is relevant in resource-poor settings where LBC facilities are not readily available. CBs are useful in reducing indeterminate diagnoses on cytology; however, the results are not statistically significant.