ThinPrep® imaging system‐assisted vs manual screening of urinary cytology slides in the detection of the “suspicious for high‐grade urothelial carcinoma” category

• Published in: Cytopathology (Oxford) 2022-11, Vol.33 (6), p.716-724
• Main Authors: Piaton, Eric, Prat, Jean‐Jacques, Nennig, Cindy, Hutin, Karine, Colombel, Marc, Ruffion, Alain
• Format: Article
• Language: English
• Subjects: atypical urothelial cells; bladder; Cancer; Carcinoma, Transitional Cell - diagnosis; Carcinoma, Transitional Cell - pathology; Cellular biology; Cervix; Cystoscopy; Cytodiagnosis - methods; Cytology; Cytopathology; high‐grade urothelial carcinoma; Humans; Hyperplasia - pathology; Mass Screening; Original; Paris system; Thinprep® imaging system; Urinary Bladder Neoplasms - diagnosis; Urinary Bladder Neoplasms - pathology; urinary cytology; Urologic Neoplasms - pathology; Urothelial carcinoma; Urothelium - pathology
• ISSN: 0956-5507; 1365-2303
• DOI: 10.1111/cyt.13173

Background The ThinPrep® Imaging System (TIS) is a Food and Drug Administration‐approved review system for cervical cytopathology, where it has been shown to increase performance over manually reviewed slides. Application of the TIS to urinary cytology has only been reported in a single study, in 2013. Methods We aimed to compare the agreement of two cytotechnologists' and a pathologist's manual screening (dots) with the fields of view (FOVs) selected by the TIS. We also aimed to track cases in which the TIS could identify missed abnormals and reduce the false‐negative fraction. Electronically marked TIS fields (EMTFs) suspicious for high‐grade urothelial carcinoma (SHGUC) were controlled by follow‐up cystoscopy and histology, where available. Results A total of 826 consecutive specimens were studied. Of those, 94 (11.4%) were unreadable by the TIS. There were 710 possible comparisons, of which 380 (53.5%) received no dot after manual screening. Of the 330 remaining slides, 149 (45.1%) had at least one dot matching with the TIS FOVs. After TIS reading, EMTFs were noted in 13 of 636 (2.0%) negative cytology cases. Surveillance showed that 3/13 (23.1%, 0.4% of the 710 possible comparisons) of those cases matched with high grade urothelial carcinoma (HGUC), whereas 6/13 (46.1%, 0.8% of the 710 possible comparisons) had negative follow‐up at 24 months, and 4/13 (30.8%) were lost for follow‐up. Conclusion The TIS increases the detection rate of SHGUC cells, potentially leading to a slight decrease in the false‐negative fraction, but at the expense of a slight but larger increase in the number of false‐positive cases. These findings stress the importance of a careful approach to the evaluation of the FOVs. This study compares manual screening with the ThinPrep® Imaging System (TIS) in urine cytopathology. Man and machine produced comparable results, but the study shows a slight risk of overinterpretation when using the TIS.