Correlation of cystoscopic impression with histologic diagnosis of biopsy specimens of the bladder.

There is a paucity of information in the contemporary literature that would permit assessment of the urologist's ability to endoscopically discriminate between benign and malignant lesions of the bladder or to predict the grade and stage of papillary neoplasms. This prospective study evaluates the correlation between cystoscopic impression of urothelial lesions and final histologic diagnoses. Sixty-four patients with 68 urothelial abnormalities requiring formal biopsy or endoscopic resection were evaluated prospectively. At the time of endoscopy, treating urologists completed questionnaires documenting the surgeon's endoscopic impression of disease type and extent and performed standard biopsy or resection of all suspicious lesions. Specimens were submitted for routine histopathologic analysis, and the results were correlated with the questionnaire data. Endoscopic evaluation correctly discriminated between dysplastic/malignant and benign/reactive lesions in this study with a sensitivity of 100%, specificity of 100%, and positive and negative predictive values of 100%. Urologists could not readily distinguish between low- and high-grade papillary urothelial lesions and were frequently unable to determine if a tumor was invasive, particularly if the degree of invasion was microscopic. Endoscopic impression at the time of bladder biopsy or resection is accurate and discriminates between the presence and absence of cancer. Endoscopic impression alone is a relatively poor staging tool with respect to extent of invasive disease and must be coupled with careful histopathologic analysis of biopsy material, bimanual examination when appropriate, and axial imaging for complete assessment of a given tumor.