Performance Characteristics of Urinary Tract Cytology: Observations From the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic Cytopathology.

ABSTRACT

CONTEXT Urine cytology is used in the evaluation of hematuria or to follow patients with known urothelial neoplasia. OBJECTIVE: To investigate the performance characteristics of urinary cytology challenges in the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic Cytopathology. DESIGN: Participant responses from 2000 to 2010 were evaluated for the reference diagnoses of high-grade urothelial carcinoma (HGUC), squamous cell carcinoma, adenocarcinoma, and benign diagnoses (including polyomavirus infection and ileal loop urine). Responses were compared for differences in sample preparation (conventional, liquid based, and cytospin) and participant type (laboratory, pathologist, cytotechnologist). RESULTS: There were 96 093 responses (46 637 pathologists [48.5%], 29 976 cytotechnologists [31.2%], and 19 480 laboratories [20.3%]); 69 814 of 74 821 responses (93.3%) for the general category positive for malignancy and 18 698 of 21 272 responses (87.9%) for positive for malignancy were concordant. Of the malignant reference diagnoses, 83.3% (59 985 of 71 581), 43.9% (732 of 1667), and 49.1% (370 of 756) were correctly identified as HGUC, adenocarcinoma, and squamous cell carcinoma, respectively. However, 802 of 1669 adenocarcinoma challenges (48.1%) and 275 of 755 squamous cell carcinoma challenges (36.4%) were misdiagnosed as HGUC. For the benign cases, the most common diagnostic pitfall was overinterpretation of ileal loop urines (330 of 5291; 6.2%) and Polyomavirus challenges (220 of 3535; 6.2%) as HGUC. For the general diagnosis of positive for malignancy, cytotechnologists performed slightly better that pathologists (94.5% [23 553 of 24 924] versus 92.3% [36 210 of 39 230]), whereas the reverse occurred for the negative category (85.6% [6423 of 7503] versus 88.8% [10 427 of 11 742]). CONCLUSIONS: Although the participants performed well in accurately classifying cases as benign and malignant, there were difficulties with the correct identification of adenocarcinoma and squamous cell carcinoma cases and with overinterpretation of ileal loop and Polyomavirus challenges as HGUC.