Phase III trial of methotrexate, vinblastine, doxorubicin, and cisplatin versus carboplatin and paclitaxel in patients with advanced carcinoma of the urothelium

A trial of the Eastern Cooperative Oncology Group

Authors

  • Robert Dreicer M.D.,

    Corresponding author
    1. Department of Hematology/Oncology and Urologic Institute, Cleveland Clinic Foundation, Cleveland, Ohio
    • Department of Hematology/Oncology R35, The Cleveland Clinic Foundation 9500 Euclid Avenue, Cleveland OH 44195
    Search for more papers by this author
    • Fax: (216) 444-9464

    • Dr. Dreicer has received research funding from Millennium, Aventis, Celgene, AstraZeneca, and Abbott Laboratories.

  • Judith Manola M.S.,

    1. Department of Biostatistical Sciences, Dana Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author
  • Bruce J. Roth M.D.,

    1. Division of Hematology/Oncology, Vanderbilt University Medical Center, Nashville, Tennessee
    Search for more papers by this author
    • Dr. Roth has received research funding from Millennium, Aventis, AstraZeneca, and Abbott Laboratories.

  • William A. See M.D.,

    1. Department of Urology, Medical College of Wisconsin, Milwaukee, Wisconsin
    Search for more papers by this author
    • Dr. See has received research funding from AstraZeneca and Abbott Laboratories.

  • Steven Kuross M.D.,

    1. Duluth Clinic, Duluth, Minnesota
    Search for more papers by this author
  • Martin J. Edelman M.D.,

    1. Division of Hematology/Oncology, University of Maryland Cancer Center, Baltimore, Maryland
    Search for more papers by this author
    • Dr. Edelman has received research funding and honoraria from Bristol-Myers Squibb.

  • Gary R. Hudes M.D.,

    1. Fox Chase Cancer Center, Philadelphia, Pennsylvania
    Search for more papers by this author
  • George Wilding M.D.

    1. University of Wisconsin Clinical Cancer Center, Madison, Wisconsin
    Search for more papers by this author

  • This study was coordinated by the Eastern Cooperative Oncology Group (Robert L. Comis, M.D.)

Abstract

BACKGROUND

The regimens of carboplatin plus paclitaxel (CP) and methotrexate, vinblastine, doxorubicin, and cisplatin (M-VAC) were compared in patients with advanced urothelial carcinoma.

METHODS

Patients with metastatic urothelial carcinoma were randomized to receive either CP (paclitaxel at a dose of 225 mg/m2 and carboplatin [targeted area under the concentration-time curve (AUC) of 6] given every 21 days) or the standard M-VAC dosage.

RESULTS

Eighty-five patients were randomized to the respective treatment regimens (41 to CP and 44 to M-VAC). Response rates and overall survival were similar for both treatment arms. Patients treated with CP had an overall response rate of 28.2% (95% binomial confidence interval, 15.0–44.9%) compared with an overall response rate of 35.9% for the M-VAC arm (95% binomial confidence interval, 21.2–52.8%) (P = 0.63, Fisher exact test). The median progression-free survival among patients who were treated with M-VAC was 8.7 months and was 5.2 months for patients receiving CP (P = 0.24, log-rank test). At a median follow-up of 32.5 months, the median survival for patients treated with M-VAC was 15.4 months versus 13.8 months for patients treated with CP (P = 0.65, log-rank test). Patients treated with M-VAC were found to have more severe worst-degree toxicities compared with patients treated with CP (P = 0.0001). There were no significant differences with regard to quality of life as assessed by the Functional Assessment of Cancer Therapy–Bladder (FACT-BL) instrument (P = 0.33).

CONCLUSIONS

Interpretation of the results of this study must be made with caution because the study failed to reach its accrual goal. Patients treated with CP had a median survival of 13.8 months compared with 15.4 months for patients treated with M-VAC. Patients treated with CP appeared in general to better tolerate their treatment; however, there were no significant differences noted with regard to measured quality of life parameters. Cancer 2004. © 2004 American Cancer Society.

Advanced transitional cell carcinoma is a moderately chemosensitive neoplasm. In the 1980s cisplatin-based chemotherapy regimens demonstrated relatively high objective response rates with some suggestion that therapy was impacting on the natural history of the disease. The methotrexate, vinblastine, doxorubicin, and cisplatin (M-VAC) regimen subsequently became the de facto standard of care.1–3 However, this regimen has a number of limitations that were highlighted by the update of the Intergroup trial,3 which compared cisplatin and M-VAC in patients with advanced urothelial carcinoma. With a minimum follow-up of 6 years, only 3.7% of patients treated on the M-VAC arm were alive and free of disease.3 Toxicity remains a serious issue, with a substantial percentage of patients experiencing nausea and emesis, stomatitis, and myelosuppression.2

Over the last decade, a number of new chemotherapeutic agents have been introduced into clinical practice, and several of these have demonstrated modest-to-significant activity against transitional cell carcinoma of the urothelium. In an early multicenter Phase II trial in previously untreated patients with advanced urothelial carcinoma, paclitaxel demonstrated an overall response rate of 42%.4 Vaughn et al. subsequently performed a Phase I/II trial of carboplatin plus paclitaxel and demonstrated a 50% response rate in the Phase II portion of their study, with sensorimotor neuropathy and granulocytopenia being the most common toxicities reported.5

Given the known toxicity of the M-VAC regimen and the observation of the activity of the carboplatin plus paclitaxel (CP) combination, we designed a randomized trial comparing the M-VAC regimen with the CP regimen in previously untreated patients with advanced urothelial carcinoma.

MATERIALS AND METHODS

Eligible patients had histologically confirmed transitional cell carcinoma (or mixed histologies containing a component of transitional cell carcinoma) of the urothelium with evidence of progressive, bidimensionally measurable, regional or metastatic disease. Patients must have been free of disease from prior malignancies for at least 5 years, with an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 at the time of entry into the study. Patients must not have received prior systemic chemotherapy or biologic response modifier therapy and 4 weeks must have elapsed since any major surgery. Adequate renal and hepatic function was required with a serum creatinine level ≤ 1.7 mg/dL, aspartate aminotransferase (AST) ≤ 2.0, and bilirubin ≤ 1.5 times the upper limit of normal. Adequate bone marrow reserve was mandated with a requirement for a granulocyte count ≥ 1500 mm3 and a platelet count ≥ 100,000 mm3 at study entry. Patients with uncontrolled cardiac dysrhythmias or American Heart Association Class III or Class IV disease were excluded. Each study site was required to have received approval of this clinical trial by a local Human Investigations Committee in accord with an assurance filed with and approved by the Department of Health and Human Services. All patients provided written informed consent prior to registration onto the current study.

Patients randomized to the M-VAC arm received methotrexate intravenously (i.v.) at a dose of 30 mg/m2 on Days 1, 15, and 22; vinblastine at a dose of 3 mg/m2 i.v. on Days 2, 15, and 22; doxorubicin at a dose of 30 mg/m2 i.v. on Day 2; and cisplatin at a dose of 70 mg/m2 i.v. over 2 hours on Day 2 with adequate hydration. Therapy was repeated every 28 days to a maximum of 6 cycles of therapy. Patients randomized to the experimental arm received paclitaxel over 3 hours at a dose of 225 mg/m2 i.v. on Day 1 followed by a fixed dose of carboplatin (targeted area under the concentration-time curve [AUC] of 6) i.v. over 30 minutes every 3 weeks for a maximum of 6 treatment cycles. Standard dose modifications for hematologic and nonhematologic toxicities for both regimens were stipulated. The use of growth factor support was in accord with the American Society of Clinical Oncology guidelines.6

Prior to study entry, all patients underwent physical examination and standard laboratory evaluations, as well as computed tomography (CT) scans of the chest, abdomen, and pelvis. Tumor measurements were performed every other cycle.

Tumor responses were analyzed using ECOG criteria.7 National Cancer Institute (NCI) common toxicity criteria (Version 2) were used to analyze toxicity. Patients who achieved objective responses or those with stable disease continued to receive therapy for a maximum of six cycles.

The study was designed to compare the M-VAC regimen with the CP regimen in patients with advanced carcinoma of the urothelium. The primary endpoint on which the sample size estimates were based was a comparison of survival times between the two treatment arms. Secondary analyses were performed to compare response, duration of response, toxicity, and quality of life between the two arms. Power calculations assumed a two-sided significance level of 0.05 using the log-rank test. Assuming a median survival of 12 months for the M-VAC regimen, a total of 300 patients followed for a minimum of 1 year would be required to ensure 80% power to detect a 50% improvement in the CP arm. Based on an intent-to-treat analysis and allowing for a 10% unevaluability rate, 330 patients was the target accrual.

RESULTS

The study was open for enrollment between September 1998 and June 2001, during which time 85 patients were entered onto the study. There were five ineligible cases (three on the M-VAC arm and two on the CP arm). Reasons for ineligibility included 1 patient with previous radiotherapy to a metastatic site, 2 patients who did not have on-study measurements made within 4 weeks of registration, and 2 patients who were found to have incidental prostate carcinoma on cystoprostatectomy (which took place prior to an amendment to the study to permit such an occurrence). One patient randomized to the M-VAC arm withdrew consent prior to therapy, but to minimize potential bias this patient was included in the analysis. After monitoring the study for 2.5 years, the ECOG Data Monitoring Committee recommended that the study be terminated because of slow accrual (the accrual rate was 31 cases per year).

Patient Characteristics

Patient characteristics at the time of study entry were well balanced across treatment arms, although there were no patients representing racial minorities who were randomized to the CP arm. There were no statistically significant differences with regard to performance status or visceral metastatic sites, factors shown by Bajorin et al. to have prognostic significance (Table 1).8

Table 1. Baseline Patient Characteristics
ParameterM-VAC armCP arm
No.%No.%
  1. M-VAC: methotrexate, Vinblastine, doxorubicin, and Cisplatin; CP: Carboplatin and Paclitaxel; ECOG: Eastern Cooperative Oncology Group.

No. of eligible patients41 39 
Male34832769
Median age (yrs)6465
Race, White368839100
ECOG performance status    
 018451334
 118451847
 2410718
Unknown1 1 
Bone and or liver metastases12291333
Risk factors (based on Bajorin et al.6)    
 014341128
 123562359
 2410513
Creatinine (median mg/dL)1.2 (0.5–1.6)1.0 (0.5–1.5)

Response Assessment

Five patients, all of whom were randomized to the M-VAC arm, were unevaluable for response. One patient had a stroke during the second therapy cycle and was removed from the study without follow-up assessment, three patients withdrew after the first cycle because of for toxicity and did not undergo repeat tumor assessment, and one patient achieved an unconfirmed complete response as assessed by magnetic resonance imaging at baseline and by CT at follow-up. In the M-VAC arm, 5 patients (12.8%) achieved a complete response (CR) and 9 patients (23.1%) achieved a partial response (PR), for an overall response rate of 35.9% (95% exact binomial confidence interval, 21.2–52.8%). Of the patients treated on the CP arm, 1 (2.6%) achieved a CR and 11 patients (25.6%) achieved a PR, for an overall response rate of 28.2% (95% exact binomial confidence interval, 15.0–44.9%). There was no statistically significant difference noted with regard to these response rates (P = 0.63, Fisher exact test). Two patients were excluded from the response analysis (the one patient who withdrew before treatment and one patient for whom no follow-up data were submitted). Conclusions regarding response differences between the treatment arms were similar when these patients were included as unevaluable (P = 0.63, Fisher exact test).

Progression-Free Survival

Progression-free survival was defined as the time from randomization to documentation of disease progression or death from any cause. Patients who were alive and free of disease at the time of last follow-up were censored at the date of last disease assessment. Because cycles were 3 weeks long for the CP arm and 4 weeks long for the M-VAC arm, the potential for differential follow-up exists and thus the analyses of progression-free survival should be interpreted with caution. The median progression-free survival among those patients treated with M-VAC was 8.7 months and was 5.2 months for patients receiving CP (P = 0.24, log-rank test) (Fig. 1).

Figure 1.

Progression-free survival in the two treatment arms. M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin; CP: carboplatin and paclitaxel.

Survival

Survival was defined as the time from randomization until death. Patients who were alive at the time of last follow-up were censored at the date they were last known to be alive. There were no statistically significant differences in overall survival that were believed to be attributable to treatment (P = 0.65, log-rank test) (Fig. 2). At a median follow-up of 32.5 months, the median survival for patients treated with the M-VAC regimen was 15.4 months, whereas patients treated with the CP regimen experienced a median survival of 13.8 months. An analysis of overall survival by treatment arm was conducted considering all patients as randomized, including those determined to be ineligible; the results were consistent with the analysis of eligible patients with no statistically significant differences noted with regard to overall survival (P = 0.75, log-rank test).

Figure 2.

Overall survival in the two treatment arms. M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin; CP: carboplatin and paclitaxel.

Toxicity

Patients on the M-VAC arm received a total of 150 cycles (median, 5.5 cycles) compared with 137 cycles (median, 6 cycles) for patients on the CP arm. Eighteen patients on each treatment arm completed 6 planned cycles of therapy. Six patients on the M-VAC arm (17%) discontinued therapy because of toxicity compared with 3 patients on the CP arm (9%).

There were two treatment-related deaths: one on each treatment arm. One patient who was randomized to the CP arm and who had a history of coronary artery disease and arrhythmia died of a myocardial infarction at home the evening after the administration of the third cycle. The other patient, who was treated on the M-VAC arm, was hospitalized after the fourth cycle of treatment for the management of dehydration and dyspnea; this patient became neutropenic and died secondary to complications from sepsis.

Table 2 provides a comparison by treatment arm of selected hematologic and nonhematologic toxicities. Table 3 provides a comparison between treatment arms of the worst-degree toxicities, with P values from the Mehta test9 shown in the last column comparing the categories of Grade 1–5 toxicity. Patients treated with M-VAC were found to have more severe worst-degree toxicities compared with patients treated with CP. Approximately 33% of patients on the M-VAC arm had a worst-degree toxicity of Grade 4 or higher compared with 15% of patients treated with CP.

Table 2. Selected Toxicities of ≥ Grade 3
Toxicity typeM-VAC (n = 43)CP arm (n = 41)
Grade 3 (%)Grade 4 (%)Grade 5 (%)Grade 3 (%)Grade 4 (%)Grade 5 (%)
  1. M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin; CP: carboplatin and paclitaxel.

Anemia3355
Neutropenia442329
Febrile neutropenia725
Thrombocytopenia19210
Dyspnea142
Mucosititis9
Emesis122
Nausea167
Thrombosis/embolism52
Fatigue19510
Neuropathy-sensory215
Table 3. Differences in Selected Toxicities by Treatment Arm
Toxicity gradeM-VAC (n = 43)CP arm (n = 41)P value
1234512345
  1. M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin; CP: carboplatin and paclitaxel.

Neutropenia048310539120< 0.0001
Neutropenic fever00601005000.57
Thrombocytopenia1688101134000.20
Sensory neuropathy341001486000.055
Stomatitis313400420000.12
Creatinine194000820001.00

Quality of Life

Quality of life was measured using the Functional Assessment of Cancer Therapy-Bladder (FACT-BL) instrument and was assessed after randomization but prior to therapy, at the end of Cycles 2 and 4, immediately after therapy, and approximately 10 months after study entry. As a consequence of the differences in cycle lengths between the 2 treatment arms, the second assessment occurred approximately 1 week later for patients randomized to the M-VAC arm and the third assessment occurred 3–4 weeks later. The off-treatment assessment occurred approximately 9 weeks later for patients treated on the M-VAC arm.

Two different strategies were used to determine whether there were differences in quality of life associated with the treatment arm. In the first approach, we assumed that the increase in the proportion of missing assessments over time was independent of survival. Given this assumption and using a log-linear model, there were no statistically significant differences noted with regard to quality of life over time by treatment arm (P = 0.33) (Fig. 3). A second analysis that attempted to correlate the missing data with survival was similarly without evidence of statistical significance

Figure 3.

Quality of life as determined using the Functional Assessment of Cancer Therapy-Bladder (FACT-BL) instrument. M-VAC: methotrexate, vinblastine, doxorubicin, and cisplatin; CP: carboplatin and paclitaxel. Midpoints show mean values by group; vertical bars show standard deviations.

DISCUSSION

During the period of time in which the current trial was designed, the M-VAC regimen had been widely utilized for > 10 years and was the de facto standard of care. The M-VAC regimen had been compared with both single agent cisplatin and the cyclophosphamide, cisplatin, and doxorubicin (CISCA) regimen and found to be superior.2, 10 Two Phase III trials have reported the median survival rates achieved with M-VAC to be between 12.5–14.8 months.2, 11 Even in expert hands, M-VAC is a moderately toxic regimen. The regimen is particularly problematic in the small but important subset of patients with advanced urothelial carcinoma with concomitant renal insufficiency.12

The CP regimen has had a somewhat mixed Phase II experience in patients with advanced urothelial carcinoma. In their Phase I/II study, Vaughn et al.5 demonstrated an objective response rate of 50% with a median survival of approximately 9 months, similar to the response rate of 52% and the 10-month median survival demonstrated by Redman et al.13 Both Pycha et al. and Zielinski et al. conducted Phase II trials with similar dosages and schedules and observed response rates > 65% with median survivals of > 1 year.14, 15 The Southwest Oncology Group conducted a Phase II trial in 29 patients and reported a 21% overall response rate with a median survival of 9 months.16 Differences in response rates and survival among these Phase II studies may in part be a consequence of imbalances in identified prognostic factors such as the presence or absence of visceral metastatic disease and poor performance status.8

Although not directly addressed by the current study, the relative utility of carboplatin versus cisplatin in patients with advanced urothelial carcinoma remains an open question. Critics of the utility of carboplatin in urothelial carcinoma point to the randomized Phase II study by Petrioli et al. as evidence of its relatively inferiority.17 Petrioli et al. randomized 57 patients to receive either the methotrexate, vinblastine, epirubicin, and cisplatin (M-VEC) regimen or the methotrexate, vinblastine, epirubicin, and carboplatin (M-VECa) regimen. Patients receiving M-VEC had an overall response rate of 71% with a median survival of ≥ 13 months versus an overall response rate of 41% and a median survival of ≥ 9.5 months in the M-VECa arm.16

The current trial designed to compare response rates and survival between patients treated with CP and those treated with M-VAC failed to meet its target accrual goal of 330 patients. Regrettably, this experience reflects a growing challenge in this country for conducting and completing clinical trials in patients with advanced urothelial carcinoma. This contrasts with the recent successful completion of two major trials in Europe.11, 18 In addition to our trial, which was closed early because of slow accrual, the Intergroup neoadjuvant study led by the Southwest Oncology Group took well over a decade to accrue 317 patients. Poor trial design, lack of consensus with regard to the important questions to ask, lack of adequate resources to conduct trials, and the availability of practitioners in the community to deliver a range of alternative chemotherapeutic agents off protocol all may contribute to our inability to complete trials in a timely fashion.

Despite the important limitations in interpreting a Phase III trial that is significantly underpowered, relevant information can be gleaned from this prospective experience. There were approximately 40 patients in each arm of a well balanced, prospectively randomized study. The median survival in the M-VAC arm was 15.4 months, which was similar to that noted in other Phase III trials. In the CP arm, the median survival was 13.8 months, a finding that is similar to that reported in several Phase II trials. Although there were no statistically significant differences with regard to quality of life as assessed with the FACT-BL, treatment-related toxicity in the M-VAC arm was generally reported to be more severe.

Even if the current trial had reached its planned accrual goal, it was not designed as an equivalency study. Although the trial comparing gemcitabine plus cisplatin with the M-VAC regimen was similarly not designed as an equivalency study, it was a well powered study and, understandably, has been interpreted as demonstrating with reasonable certainty that the combination of gemcitabine plus cisplatin is a clinically useful alternative to M-VAC.10 The current study, with its limitations, should not be used to support similar conclusions.

Not to be lost in the discussion regarding the optimal treatment regimen for advanced urothelial carcinoma is one unmistakable conclusion. Although advanced transitional cell carcinoma represents a chemosensitive neoplasm, to our knowledge essentially no progress has been made in improving the survival of patients since the introduction of the M-VAC regimen nearly 20 years ago. Although novel chemotherapy agents continue to be introduced into our armamentarium and warrant evaluation, it has become increasingly clear that more rationale, directed therapies will need to be integrated into the management of advanced urothelial carcinoma to move the bar forward.

Acknowledgements

The contents of this study are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.

Ancillary