Bladder washing cytology. Comparison of two analytic methods and two proposed quantitative criteria for carcinoma in situ.

ABSTRACT

OBJECTIVE: To compare bladder washing cytology preparations created by the Nucleopore filter and slide centrifuge techniques and to evaluate a marker for carcinoma in situ (CIS). STUDY DESIGN: Nucleopore filter and slide centrifuge preparations from 27 patients with urothelial carcinoma were compared and used to create two criteria for CIS. To study reproducibility, three observers evaluated 25 filter preparations for these CIS criteria. RESULTS: The filter technique displayed more better-preserved single cancer cells (P = .02) and a higher percent group count (the number of cancer cell groups divided by the sum of the number of single cancer cells plus the number of cancer cell groups) (P = .005) than did the cytocentrifuge technique. The initial study showed that patients with many single tumor cells and lower percent group counts were more likely to have CIS than patients without this combined condition (P = .001). This CIS marker had moderate reproducibility (kappa = 0.47 +/- 0.12). CONCLUSION: The filter technique had better cellular recovery and preservation of tumor cells than did the centrifuge technique. Quantitative cytologic criteria proposed in this study may be an indication that CIS may be present; improved sensitivity and specificity may be obtained if they are combined with other criteria.