Predicting the pathology of post-BCG lesions

Abstract

14573 Background: This study evaluated the accuracy of office cystoscopy and urinary cytology in predicting the histopathology of post-BCG bladder lesions. Methods: A retrospective chart review was performed of all patients who had received intravesical BCG at two Veteran Affairs Hospitals from July 1995 to July 2005. The initial pathology, the post-treatment cystoscopy findings, urinary cytology and post-BCG biopsy were correlated. Results: A total of 136 biopsies from 64 patients who had received 79 courses of BCG were reviewed. The mean time to the first repeat bladder biopsy and to the first recurrence was 8.8 months and 11.7 months, respectively. During the study, 83% of patients developed recurrence. Malignancy was confirmed in 61% of all biopsies. For all lesions, cytology alone had a sensitivity of 53.7%, specificity of 90.5%, PPV 85%, NPV 47%. The biopsy was positive in 84% of patients with visible papillary lesions and a negative cytology. Improved accuracy was seen in papillary lesions with a positive cytology; 94% positive biopsy. The bladder biopsy was positive in 91% of patients with erythema and a positive cytology in comparison to 9.5% of patients with erythema and a negative cytology. Conclusions: There is controversy regarding the necessity of obtaining transurethral biopsies of erythematous lesions in patients who have undergone intravesical BCG therapy. Combining the cystoscopic findings and the urine cytology increases the ability to delineate BCG effect from recurrence. This study suggests that patients with post-BCG erythematous patches and a negative cytology can avoid an unnecessary transurethral bladder biopsy. This study also demonstrates that most BCG failures occur after the traditionally believed 3-month evaluation. No significant financial relationships to disclose.