Randomised phase II trial of gemcitabine and paclitaxel second-line chemotherapy in patients with transitional cell carcinoma (AUO Trial AB 20/99)

The objectives are to evaluate and compare the response and toxicity of a 3-weekly and a 2-weekly regimen of gemcitabine (Gem) and paclitaxel (Pac) second-line treatment in patients with transitional cell carcinoma (TCC). Between June 2000 and July 2001, 30 patients with progressive disease (PD) during first-line chemotherapy (n = 11) or relapse after adjuvant cisplatin-based chemotherapy of a metastatic or locally advanced TCC (n = 18) have been randomised to receive either six cycles (schedule A) of 3-weekly Gem (1000 mg/qm, days 1 and 8) and Pac (175 mg/qm, day 1) or 2-weekly treatment until disease progression (schedule B) with Gem (1250 mg/qm, day 1) and Pac (120 mg/qm, day 2). Restaging was performed after every 6 weeks by clinical imaging. Of 30 patients, one patient in schedule A and two patients in schedule B were not evaluable for response due to serious adverse events (SAEs) during the first cycle. The overall objective response (OR) was 44% (12 of 27) with eight complete remissions (CRs) and four partial remissions. Median time to progression (TTP) was 11 (3-41) months in schedule A and 6 (1-15+) months in schedule B. Median survival was 13 (5-46) months in schedule A and 9 (0-16) months in schedule B. Schedule A showed a significantly higher rate of CRs (7 vs. 1, p < 0.05). With a median number of six (1-6) cycles (A) and nine (1-23) cycles (B), TTP and survival were not significantly different. In schedule B, one patient had WHO grade IV anaemia and leucopenia. WHO grade III toxicities were seen in schedule A/B as follows: anaemia 3 (23%)/2 (16%) patients, leucopenia 5 (38%)/2 (16%), thrombocytopenia 0/2 (16%) and alopecia 10 (76%)/4 (32%). The combination of Gem and Pac is an effective second-line regimen in patients with mainly poor prognosis due to PD after cisplatin-based chemotherapy. Except for three SAEs (uncertainly therapy related), both regimens were tolerated well. The 3-weekly schedule with a nonsplit Pac dose showed a significantly higher complete response rate in our small study population and, thus, might be superior to the 2-weekly schedule.