Correlation of office-based cystoscopy and cytology with histologic diagnosis: how good is the reference standard?

To evaluate the accuracy of cystoscopy and cytology in predicting the histopathologic features of suspicious cystoscopic lesions.

We reviewed the bladder biopsy records at two institutions from July 2001 to July 2004. Intraoperative biopsies were performed for positive (papillary or sessile) (n = 155) and equivocal (n = 101) lesions found during office cystoscopy. The specimens were submitted for histopathologic analysis, and the results were correlated with the preceding office-based cystoscopy and cytology findings.

Malignancy was found in 27 (26.7%) of 101 equivocal lesions and 128 (82.5%) of 155 sessile or papillary lesions. Of the 72 biopsies performed for equivocal lesions in patients with a history of transitional cell carcinoma (TCC), 24 (33%) had malignancy compared with 3 (10.3%) of 29 biopsies performed in patients without a history of TCC. For patients with no history of TCC and equivocal lesions, cytology was able to identify all neoplasms (100% sensitivity) and negative cytology always predicted a normal biopsy result (100% negative predictive value). However, for patients with a history of TCC and equivocal lesions, cytology had a sensitivity and negative predictive value of 81.8% and 88.2%, respectively.

Our findings suggest that patients without a history of TCC who have an equivocal lesion at cystoscopy may avoid an unnecessary operation if they have normal cytology findings. A large number of patients with equivocal cystoscopic findings undergo unnecessary biopsies in the operating room, and future studies should be performed to assess the value of other bladder markers in predicting oncologic status in this setting.