Phase I and II trials in the 1980s demonstrated that metastatic urothelial transitional cell cancers are chemosensitive. Currently, systemic combination chemotherapy is the only treatment that may result in long-term survival for some patients with advanced or metastatic disease. Although antitumour activity has been demonstrated with several agents, the median survival associated with single-agent therapy is 4–6 months. The median survival time for combination regimens (e.g. methotrexate+vinblastine, or doxorubicin+cisplatin) is approximately 8 months (Sternberg et al, 1989). It is known that pretreatment prognostic features have an impact on individual patient outcome, thus the variation in reported survival in patients treated with chemotherapy may be biased by these features (Bajorin et al, 1999; Calabro and Sternberg, 2002).
In 1985, the Memorial Sloan-Kettering cancer centre (MSKCC) investigated the four-drug regimen using MVAC (methotrexate, vincristine, doxorubicin, cisplatin) (Sternberg et al, 1985) in patients with TCC. They reported an overall response rate of 71%. These results were later confirmed in a series of 133 patients (Sternberg et al, 1989). Other studies using the MVAC regimen reported encouraging response rates including complete responses of 13–19% (Tannock et al, 1989; Igawa et al, 1990; Boutan-Laroze et al, 1991). Another study reported the use of MVAC in 21 patients with high-stage TCC. This study, with long-term follow-up, suggested that the durability of response was disappointing (Connor et al, 1989). Long-term follow-up evaluation of the intergroup trial confirmed that MVAC was superior to single-agent cisplatin in patients with advanced urothelial carcinoma. However, durable progression-free survival was rare (Saxman et al, 1997).
The combination of CMV (cisplatin, methotrexate, vincristine) has also been investigated with similar response rates to those reported with MVAC. A study of 58 patients with metastatic TCC reported complete responses (CR) in 14 of 50 evaluable patients and partial responses (PR) in 14 patients for an overall response rate of 56%. The median survival for CRs was 11 vs 7 months for PRs (P<0.05) and 6 months for nonresponders. Renal and haematological toxicities with this regimen were moderate (Harker et al, 1985).
Although response rates achieved using MVAC have been encouraging, this regimen has significant toxicity, particularly in older patients. This is relevant to bladder cancer since the median age of patients is approximately 68 years. Many of these patients have impaired renal function, which may compromise the safety of the use of cisplatin and increase the risk of toxicity. Indeed, a significant proportion of otherwise fit patients (around 50% in our institution) have a calculated glomerular filtration rate <60mlsmin−1, the usual cutoff for combination studies with cisplatin. Treatment with MVAC can result in clinically significant myelosuppression (up to 25% incidence of neutropenic fever and a 3% drug-related mortality) and significant mucositis (up to 50% incidence of grade 2–3 mucositis) (Sternberg et al, 1985, 1989; Connor et al, 1989; Tannock et al, 1989; Igawa et al, 1990; Logothetis et al, 1990; Boutan-Laroze et al, 1991; Loehrer et al, 1992). Recent efforts to improve the outcome for patients with metastatic transitional cell carcinoma have focused on the identification of new drugs with single-agent activity and on their incorporation into platinum-based combination regimens. Paclitaxel, docetaxel, ifosfamide, and gemcitabine are among the most active new agents (Hussain and James, 2003).
A large phase III trial randomised 405 patients with stage IV TCC and with no prior chemotherapy exposure between GC (gemcitabine 1000mgm−2 days 1, 8, and 15; cisplatin 70mgm−2 day 2) or MVAC every 28 days for a maximum of six cycles. Overall survival, time to disease progression, time to treatment failure, and response rate (GC, 49%; MVAC, 46%) were similar in each arm. Owing to a higher incidence of neutropenic fever and mucositis, more hospital admissions were required for patients receiving MVAC (49 admissions for a total of 272 days) compared to the GC group (9 admissions for a total of 33 days), resulting in considerably greater hospital resource utilisation (Von Der Maase et al, 2000). Thus, although this trial was not designed to show equivalence of the two regimens, many have interpreted the results as showing therapeutic noninferiority and adopted the GC regimen as the new standard on the basis of better tolerability (De Wit et al, 2001).