Split-sample comparison of urothelial cells in ThinPrep and cytospin preparations in urinary cytology: Do we need to adjust The Paris System for Reporting Urinary Cytology criteria?

Background The Paris System for Reporting Urinary Cytology (TPS) defines clear morphologic criteria to classify urinary specimens into 7 diagnostic categories. According to TPS, a nuclear-to-cytoplasmic ratio (N:C ratio) >0.7 and hyperchromasia must be observed to render a diagnosis of high-grade urothelial carcinoma (HGUC). TPS was established using only liquid-based preparation techniques, and to the authors' knowledge it is unknown whether TPS can be applied using other preparation methods. Methods In the current prospective study, voided urine samples from patients with HGUC and negative for HGUC (NHGUC) were prepared using both ThinPrep and cytospin methods. ImageJ image processing software was used to measure the N:C ratio and hyperchromasia. For each patient, the N:C ratio and degree of hyperchromasia of urothelial cells present in both cytopreparations were compared. Results A total of 10 HGUC cases and 9 NHGUC cases, represented by a total of 688 cells (mean, 36.7 cells in HGUC cases; and mean, 35.8 cells in NHGUC cases), were evaluated in the current study. An overall comparison of HGUC cells with NHGUC cells demonstrated that HGUC cells had a higher average N:C ratio (0.5465 vs 0.2846) and greater hyperchromasia as measured by the average nuclear pixel gray value (100.8 vs 120.7). The N:C ratio was statistically different in 4 NHGUC cases, demonstrating higher N:C ratios in the ThinPrep preparations. Hyperchromasia was found to be statistically different in 6 cases, 5 of which demonstrated increased hyperchromasia in the ThinPrep specimens. Conclusions The morphologic features of HGUC cells appear to be similar in samples prepared using the ThinPrep and cytospin methods, and therefore TPS criteria may be applied successfully in laboratories that use these methods.