Cisplatin is the one of the most active and studied single agents in urothelial TCC. During the late 1970s, several trials of single-agent cisplatin in patients with advanced TCC reported overall response (OR) rates ranging from 26% to 65%. Complete responses (CR) from 5% to 16% were also observed. However, randomized trials that included single-agent cisplatin as a treatment arm reported lower OR rates and CR rates of 9% to 31% and 3% to 9%, respectively. The median overall survival (OS) of patients treated with single-agent cisplatin in these trials was 8 to 9 months (1,2,3). Poor renal function limits the ability to use cisplatin-based regimens due to the high median age for patients, the decline in glomerular filtration rate due to aging, and concomitant medical illnesses, such as vascular disease and diabetes mellitus that further compromise renal function. Carboplatin has far less renal toxicity than cisplatin, but efficacy of the two agents have not been directly compared. However, in a review of 327 patients with advanced/metastatic bladder cancer treated on 13 trials with single-agent carboplatin, 14% of patients achieved objective responses (3% CR and 11% partial responses [PR]) (4).

Agents also identified as active in the 1970s and 1980s included methotrexate (OR 30%), doxorubicin (OR 17%), and vinblastine (OR 22%) (5,6,7,8,9,10,11,12,13). However, unlike cisplatin and carboplatin, complete responses were extremely rare and response durations were typically 3 to 6 months.

Gemcitabine has been studied extensively at different doses and schedules in both previously treated and untreated patients. One trial performed in previously untreated elderly patients reported an OR of 45.5% with a dose of 1,200 mg/m² on a 3-week schedule (14). In two other studies, the same dose on days 1, 8, and 15 of a 4-week schedule resulted in a 28% OR and median OS of 13.5 months (15) and a 24% OR and median OS of 8 months, respectively (16). A reduced dose of gemcitabine at 1,000 mg/m² on the same schedule led to a 29% OR and a median OS that exceeded 13 months (17). Three studies have investigated gemcitabine in the second-line setting. A phase 1 study of 15 patients, 14 of whom were previously treated, reported an encouraging 27% OR with doses ranging from 875 to 1,370 mg/m² (18). Two phase 2 studies evaluating gemcitabine at 1,250 mg/m² on a 4- and 3-week schedule revealed a 23% OR and an 11% OR, respectively (19,20). The pooled OR from the 192 patients treated on these seven studies is 25%; however, the variable responses and different doses and schedules utilized make it difficult to identify the optimal dose and schedule of gemcitabine (21).

The taxanes have demonstrated modest activity, with both docetaxel and paclitaxel showing substantial activity when used as single agents in previously untreated patients. An Eastern Cooperative Oncology Group (ECOG) trial in 26 patients evaluated first-line therapy with paclitaxel by 24-hour continuous infusion administered every 21 days (22). Eleven of twenty-six patients (42%) demonstrated a response with seven achieving a CR. Similarly, in the first-line setting, docetaxel administered on a 3-week schedule achieved an OR of 31% (23).

As expected, second-line studies of taxanes report lower response rates than seen in chemotherapy-naive patients. Four of thirty patients (13.3%) treated with docetaxel every 21 days had a partial response, all in nodes or soft tissue; the estimated median duration of OS for all patients was 9 months (24). Myelosuppression was frequent and dose reductions for adverse events were needed in 18 patients (60%). Weekly paclitaxel was evaluated in patients with progressive urothelial carcinoma who had received one prior systemic chemotherapy regimen (25). Three of thirty-one patients (10%) achieved a PR with a median time to progression of 2.2 months and median survival of 7.2 months.

Ifosfamide, the oxazophosphorine analog of cyclophosphamide, is active in patients with both TCC and squamous cell carcinoma histologies. The ECOG evaluated ifosfamide as second-line therapy in 56 evaluable patients; there were 5 CR and 6 PR (20%, 95% CI, 10%-32%) (26).

In 1985, investigators at Memorial Sloan-Kettering Cancer Center (MSKCC) reported a landmark trial using the combination of methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC). In the initial report, 24 patients with advanced or unresectable urothelial TCC were treated with methotrexate (30 mg/m²) on days 1, 15, and 22; vinblastine (3 mg/m²) on days 2, 15, and 22, doxorubicin (30 mg/m²) on day 2 and cisplatin (70 mg/m²) on day 2 (27). Remarkably, responses were observed in 71% (95% CI, 53%-89%) of those treated with clinical CR in 50% (95% CI, 30%-70%). At the time of the report, the median duration of response had not been reached
at a median follow-up of 9.5 months. A second study from the same investigators confirmed the preliminary results with MVAC in a larger patient population (28).

A contemporary randomized trial proved that cisplatin was a necessary component of combination chemotherapy, comparing CMV (cisplatin, methotrexate, and vinblastine) with MV (methotrexate and vinblastine) (29). Of the 214 patients enrolled, 108 were randomized to CMV and 106 to MV. The hazard ration (HR) was 0.68 (95% CI, 0.51-0.90, p = 0.0065) in favor of the cisplatin-containing arm. MVAC remained the regimen of choice at this time with other groups publishing their experiences. In these small phase II studies, response rates were somewhat lower with OR rates ranging from 40% to 57% and CRs in 13% to 19% (30,31,32).

By the early 1990s, an intergroup trial comparing MVAC to single-agent cisplatin established this four-drug regimen as the original standard of care for patients with metastatic disease (3). Of the 246 assessable patients, 126 were randomized to cisplatin alone and 120 were randomized to MVAC. At a median follow-up duration of 19.7 months, the MVAC regimen resulted in a significantly greater response rate (39% vs. 12%; p < 0.0001) and OS (12.5 months vs. 8.2 months; p = 0.0002). As expected, treatment with MVAC was more toxic causing more drug related deaths (4% vs. 0%), febrile neutropenia (10% vs. 0%), grade 3 or 4 nausea/vomiting (12% vs. 1%), and mucositis (17% vs. 0%).

This randomized trial validated that cisplatin-based combination chemotherapy should be considered for patients with metastatic/unresectable urothelial TCC and that MVAC should be considered the treatment of choice. Despite its superiority, however, the limitations of MVAC were readily apparent. Although many patients responded to treatment, median survivals still approximated 1 year. Furthermore, the longterm durability of CRs was poor with only 3.7% of patients remaining continuously free of disease in an intergroup trial reported comparing MVAC with cisplatin (33).

Perhaps the most limiting factor associated with MVAC was the associated toxicity. In the reported trials, febrile neutropenia occurred in up to 25% of patients and grade 2-3 mucositis developed in up to 50%. Other prominent toxicities included decreased renal function, hearing loss, and peripheral neuropathy. Treatment-related deaths occurred in 2% to 4% of patients. Given that patients with metastatic TCC are typically older with comorbid conditions, this side effect profile is particularly relevant. However, the regimen also proved toxic in patients with a good renal function and preserved performance status.

In addition to MVAC, several other cisplatin-based regimens were extensively studied and warranted comparison in randomized studies. Given the encouraging activity of these regimens, comparison was necessary to define whether or not they improved upon MVAC as the standard of care (Table 27.1).

The first prospective randomized trial comparing combination chemotherapy was MVAC versus CISCA (cyclophosphamide, cisplatin, and doxorubicin) (34). One hundred and ten patients were enrolled with fifty-five randomized to CISCA and fifty-five randomized to MVAC. Treatment with MVAC was associated with a superior OR (65% vs. 46%; p <0.05) and median OS (48.3 weeks vs. 36.1 weeks; p = 0.0003). The toxicity associated with these two regimens was comparable.

In an effort to decrease the toxicity and improve the efficacy of MVAC, several investigators evaluated the use of altered doses/schedules of the regimen with colony stimulating factor support. Promising phase II data emerged for “high-dose” MVAC (HD-MVAC), which was actually standard dose MVAC but given every 2 weeks (in contrast to 4 weeks) with granulocyte colony stimulating factor (GCSF) (35). The European Organization for Research and Treatment of Cancer (EORTC) conducted a prospective randomized trial comparing this regimen to standard MVAC administered every 4 weeks (36). A total of 263 chemotherapy-naïve patients with metastatic/advanced TCC were enrolled. With the every 2-week schedule, it was possible to deliver twice the doses of cisplatin and doxorubicin in
half the time and with less toxicity. While there was a significant difference in the CR rates (21% vs. 9%, p = 0.009) and progression-free survival (p = 0.037; HR 0.75; 95% CI, 0.58-0.98) favoring the every 2 week schedule, there was no significant difference in median OS. Of note, this trial was powered to detect a 50% difference in median survival and a lesser benefit with this dose-dense regimen may have been missed. This trial was updated with long-term (>7 year) follow-up and OS was 22% at 5 years in the HD-MVAC versus 13.5% for MVAC. The 2-year survival rate for HD-MVAC was 37% versus 26% for MVAC. Despite no differences in median survivals, the mortality HR was 0.76 (p = 0.017) favoring the high-dose arm and reflecting the tails on the survival curves (37).

A phase III randomized trial comparing docetaxel plus cisplatin (DC) with MVAC was reported by the Hellenic Cooperative Oncology Group (38). Patients randomized to DC received docetaxel 75 mg/m² and cisplatin 75 mg/m² repeated every 3 weeks. Both treatment arms received GCSF support. Of the 224 patients enrolled, 109 were randomized to MVAC and 111 were randomized to DC. Although DC was associated with less hematologic toxicity and febrile neutropenia, OR (54.2 vs. 37.4; p = 0.017), median time to progression (9.4 months vs. 6.1 months; p = 0.003), and median OS (14.2 months vs. 9.3 months; p = 0.026) favored the MVAC arm.

Based on encouraging initial results with the gemcitabine plus cisplatin (GC) combination (39,40,41), a multicenter randomized phase III trial was performed to compare GC with MVAC in patients with advanced/metastatic TCC (42). Four hundred and five chemotherapy-naïve patients were enrolled with 203 randomized to GC (gemcitabine 1,000 mg/m² on days 1, 8, and 15; cisplatin 70 mg/m² on day 2) and 202 randomized to standard MVAC. Treatment on both arms was administered every 28 days for a maximum of 6 cycles. The response rates on both arms were similar with 12% CRs and 37% partial responses on the GC arm and 12% CRs and 34% PR on the MVAC arm (p = 0.51). Similarly, median OS was similar, 13.8 months with GC and 14.8 months with MVAC (HR 1.04; 95% CI, 82-1.32; p = 0.75). Importantly, GC was associated with a better safety profile and tolerability. While GC was associated with more grade ≥3 anemia and thrombocytopenia, MVAC was associated with more neutropenic fever (14% vs. 2%), neutropenic sepsis (12% vs. 1%), grade ≥3 mucositis (22% vs. 1%) and treatment related deaths (3% vs. 1%).

This randomized trial was not adequately powered to determine the therapeutic equivalence of these two regimens. However, these data can be interpreted as showing that, in terms of survival, GC is comparable to MVAC which has held up with long-term follow-up (43). In addition, GC appears to be associated with a more favorable risk-benefit ratio. Given the results of this trial, and the ease of administration of this regimen, GC has become a second standard of care regimen for patients with metastatic urothelial TCC.

The combination of paclitaxel, gemcitabine, and cisplatin was identified as quite active in a phase I/II study (Table 27.2) (44). Of the 58 evaluable patients, there were 16 CRs (28%) and 29 PRs (50%) for an OR rate of 77.6% (95% CI, 60%-98%). Toxicities consisted mainly of asthenia, thrombocytopenia, and neutropenia with 11 cases (22%) of febrile neutropenia and 1 toxic death. The median survival time for the phase I portion was 24 months and the median survival time of the whole group had not been reached at the time of publication. Based on this level of activity, this three-drug regimen was compared with GC in an international randomized phase III trial conducted by the EORTC (45). Six hundred twenty-seven patients were randomized. The OR was 57% for the three-drug arm and 46% for GC (p = 0.02). However, the increased response rate did not sufficiently impact on survival. The median survival was 15.7 months for paclitaxel plus GC and 12.8 months for GC, with no significant difference in OS (p = 0.12 adjusted for risk factors). Greater thrombopenia and bleeding was seen on the GC arm (12% vs. 7%) and more febrile neutropenia was seen with the addition of paclitaxel (13% vs. 4%). However, both treatments were well tolerated. The authors concluded that GC remained the standard of care and that paclitaxel should not be routinely added to the regimen.

The three-drug regimen of ifosfamide, paclitaxel, and cisplatin (ITP) has been explored in a phase II trial (46,47). Thirty of forty-four assessable patients treated with this regimen (68%; 95% CI, 52%-81%) demonstrated a major response with 10 (23%) CRs. Overall, myelosuppression was the predominant toxicity (45% grade 3-4 neutropenia) although the risk of febrile neutropenia was low (3.3% of all cycles). Grade 3 neuropathy and renal insufficiency occurred in 9% and 11%, respectively. The median survival of patients treated with ITP was 20 months. This survival is among the best reported result for patients with metastatic/advanced TCC and greater than the previously observed results with MVAC (12-13 months). However, the impact of baseline prognostic factors and aggressive posttreatment surgery cannot be discounted and may have contributed to the favorable results (48). The majority (>90%) of patients had either no poor risk factors (i.e., Karnofsky performance status (KPS) ≥80 and no visceral metastases) or one poor prognosis feature (either KPS <80 or visceral metastases).

In a pilot trial at MD Anderson Cancer Center, the combination of cisplatin, gemcitabine, and ifosfamide was explored (49). Fifty-one previously treated patients received cisplatin (30 mg/m²), gemcitabine (800 mg/m²), and ifosfamide (1 g/m²) given on day 1 and then repeated on days 8 and 15 unless there was dose-limiting toxicity. Treatment cycles were repeated every 28 days. Of the 51 patients enrolled, 48 (94%) had dose-limiting hematologic toxicity on day 8 or 15. Despite the difficulty in delivering this regimen, there were two CRs (4%) and 18 PRs (37%). Neither this regimen nor the ITP combination has been pursued in a randomized study.

Due to the high frequency of comorbidities and renal dysfunction in patients with advanced/metastatic TCC, recent studies have specifically addressed the efficacy and tolerability of chemotherapy in this patient population. The ECOG performed a phase II trial of paclitaxel and carboplatin in patients with advanced bladder cancer and renal dysfunction as defined by a serum creatinine of 1.6 to 4.0 mg/dL (50). Forty-two patients were accrued and thirty-seven were treated. Paclitaxel (225 mg/m²) and carboplatin at an area under the curve (AUC) dose of 6 were administered every 3 weeks for up to 6 cycles. Granulocytopenia (60%) and neurotoxicity (35%) were the most common ≥grade 3 toxicities. The regimen resulted in an objective response rate of 24.3% (95% CI, 11.9%-41.7%). Multiple other investigators examined the same combination with paclitaxel doses ranging up to from 150 to 225 mg/m² and carboplatin (AUC) doses from 5 to 6 without differences in CR, OR or survival.

Bellmunt et al. reported a trial of the combination of gemcitabine and carboplatin in patients “unfit” for cisplatin due to a WHO performance status of 2 and/or a glomerular filtration rate (GFR) below 60 mL/min (51). For the 16 evaluable patients, the OR rate was 44% with 1 CR and 6 PRs. The activity of this regimen as well as its tolerability justified a comparison to the standard of care for patients with poor renal function in Europe, the triplet of methotrexate, carboplatin, and vinblastine (M-CAVI). A randomized trial of carboplatin plus gemcitabine versus M-CAVI in patients with impaired renal function has been reported by the EORTC (Table 27.1) (52,53) One-hundred seventy-eight patients were randomized. Stratification parameters were well balanced including World Health Organization (WHO) performance status and renal function. The GFR was <60 mL/min in 58% of gemcitabinecarboplatin treated patients and 57% in M-CAVI patients. Severe mucositis was greater with M-CAVI (5.6% vs. 1.1%) as were neutropenic fever (14.6% vs. 5.6%) and deaths due to treatment (4.5% vs. 2.2%). Thrombocytopenia with bleeding was less with M-CAVI (1.1% vs. 18%). OR was greater for gemcitabine plus carboplatin (41.6% vs. 30.3%) but no differences in survival were evident; median survival was 9.3 months for gemcitabine plus carboplatin and 8.1 months for M-CAVI (HR 0.94; p = 0.64). This trial provides level 1 evidence that gemcitabine plus carboplatin should be recommended as the standard of care for cisplatin-ineligible patients based on impaired renal function, poor performance status, or both.