Risk factors for positive findings in patients with high-grade T1 bladder cancer treated with transurethral resection of bladder tumour (TUR) and bacille Calmette-Guérin therapy and the decision for a repeat TUR.

OBJECTIVE: To determine factors predictive of positive findings at the 3-month follow-up evaluation (after transurethral resection of bladder tumour [TUR] and bacille Calmette-Guérin [BCG] therapy) in patients with initial high-grade (HG)T1 bladder cancer, and to assess the depth of lamina propria (LP) invasion and effectiveness of BCG therapy. PATIENTS AND METHODS: In all, 138 patients with initial HGT1-transitional cell carcinoma (TCC) were prospectively assigned, after TUR + BCG and according to depth of LP invasion, to a postBCG-TUR (T1b) or cystoscopy/cytology (T1a) at 3 months. Any finding at 3 months was considered positive. The predictive value of 11 clinical and pathological variables was assessed by chi-squared, Mann-Whitney U and multivariate logistic regression. RESULTS: Of the 138 patients (14 women, mean age 69 years), 42% had T1a and 58% T1b TCC. Tumour size and carcinoma in situ (CIS) were significantly associated with positive findings and present in 26% (36/138) of the patients. The postBCG-TUR (T1b cases), was positive in 31% (25/80), including seven infiltrating tumours. On multivariate analysis, again a tumour size of >3 cm (odds ratio, OR, 7.02) and associated CIS (OR 5.4) were significantly related to a positive postBCG-TUR. A secondary finding was that at 20.3 months; patients with T1a TCC, who did not undergo a repeat TUR, did not have increased progression; only 3% (two of 58) had progressed compared with 21% (17/80) of those with T1b/c TCC (P < 0.002). CONCLUSIONS: In initial HGT1-TCC, tumour size and CIS were predictive factors of positive findings at 3 months after the initial TUR + BCG therapy. Patients with HGT1-TCC invading the LP (T1b TCC) had a seven times higher risk of a positive repeat TUR if the initial tumour was >3 cm and a five-fold increased risk if associated with CIS. The repeat TUR after BCG therapy allowed confirmation of complete resection and pathological evaluation of the BCG response. Although data are still preliminary, the strategy of performing a repeat TUR only in cases with LP involvement, i.e. T1b TCC, did not increase the risk of progression in cases with T1a TCC.