|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 1999 American Society for Clinical Oncology Phase III Clinical Trial of the Combination of Cisplatin, Dacarbazine, and Carmustine With or Without Tamoxifen in Patients With Advanced Malignant MelanomaFrom the Mayo Clinic and Mayo Foundation, Rochester, and Duluth Community Clinical Oncology Program, Duluth, MN; Illinois Oncology Research Association CCOP, Peoria, and Carle Cancer Center Community Clinical Oncology Program, Urbana, IL; and Siouxland Hematology-Oncology Associates, Sioux City, IA. Address reprint requests to Edward T. Creagan, MD, Mayo Clinic, 200 First St, SW, Rochester, MN 55905; email creagan.edward{at}mayo.edu
PURPOSE: A prospective randomized phase III clinical trial was conducted to assess whether the addition of tamoxifen (TAM) to the three-agent regimen of cisplatin (CDDP), dacarbazine (DTIC), and carmustine (BCNU) significantly increased the progression-free survival and overall survival of patients with advanced malignant melanoma.
PATIENTS AND METHODS: Patients with advanced malignant melanoma were treated with CDDP + DTIC + BCNU (CDB) with or without TAM. The dose schedule was CDDP 25 mg/m2 given intravenously (IV) for 30 to 45 minutes in 500 mL of dextrose and RESULTS: There were 184 eligible patients enrolled. These patients were observed until death or for a minimum of 1.3 years. At last contact, 12 were still alive. The median time to progression was 3.4 months on the CDB arm and 3.1 months on the CDB + TAM arm. The median survival time was 6.8 months with CDB and 6.9 months with CDB + TAM. Progression-free survival (P = .429) and overall survival (P = .545) were not found to differ by treatment. CONCLUSION: The addition of TAM to this three-agent regimen of CDB was not found to provide a meaningful clinical advantage in the treatment of patients with advanced malignant melanoma.
SYSTEMIC STRATEGIES for patients with malignant melanoma are a vexing conundrum. Among single-agent cytotoxic regimens, most notably imidazole carboxamide and the nitrosoureas, the response rates are approximately 15% to 20%.1-3 Median response durations are measured in months, with few long-term survivors. Single-agent and combination regimens using biologic response modifiers such as interferon and interleukin-2 have held great promise. Unfortunately, responses are likewise limited to a relatively small proportion of patients, although there may be some long-term responders.4-7 Initial uncontrolled clinical trials using the three-agent regimen of cisplatin (CDDP), carmustine (BCNU), and dacarbazine (DTIC) produced response rates approximately two- to three-fold in excess of those achieved by single-agent regimens but little evidence of enhanced survival.1-3 Toxicities, primarily gastrointestinal and hematologic, were manageable in most reported series. Early studies of the addition of tamoxifen (TAM) to the three-agent regimen of CDDP, DTIC, and BCNU (CDB) seemed promising. Del Prete et al8 treated 20 patients with the four-agent regimen and obtained a response rate of 55%. McClay et al9 reported a response rate of 50% in the 20 assessable patients they treated with the regimen, but most of the patients had to be re-treated at 4-week intervals rather than 3-week intervals because of leukopenia, thrombocytopenia, and anemia. Six (30%) patients developed a deep vein thrombosis (DVT), and four of these six patients had a pulmonary embolism. McClay et al10 subsequently opened a study assessing the three-agent regimen alone, having noted that single-agent trials of TAM found it had little activity in the treatment of malignant melanoma and that of the four agents in their previous trial, only TAM was associated with DVT. Of the 20 patients treated, none developed a DVT, but the response rate was 10%. Based on these findings, McClay et al recommended the addition of TAM to CDB and that patients be monitored for vascular complications. After these reports, the three-agent regimen plus TAM became popular among many oncologists in the management of patients with disseminated malignant melanoma. However, there had not been a randomized trial to assess the addition of TAM to this cytotoxic regimen. In early 1991, we opened a phase III clinical trial among patients with advanced malignant melanoma to assess the addition of TAM to the regimen of CDB.
A randomized phase III study was conducted among patients with advanced malignant melanoma to assess whether the addition of TAM to the drug combination of CDB could significantly increase progression-free survival and overall survival. The dose levels and routine of administration were as follows: CDDP 25 mg/m2 was given intravenously (IV) for 30 to 45 minutes in 500 mL of dextrose and  normal saline on days 1 to 3 of a 3-week cycle; DTIC 220 mg/m2 was given IV for 1 hour in 500 mL of dextrose and NaCl on days 1 to 3 of a 3-week cycle; BCNU 150 mg/m2 was given IV for 2 to 3 hours in 750 to 1,000 mL of dextrose and 5% water on day 1 of every odd 3-week cycle; and 20 mg of TAM was taken orally every morning. Patients who responded to treatment continued on therapy until disease progression was documented. For those who remained stable after 18 weeks of therapy, treatment was discontinued and monthly evaluations continued until progression. Patients whose disease progressed in the CNS continued therapy for an additional 18 weeks if they had stable disease or continued therapy until disease progression was noted if they were responding to therapy. Colony-stimulating growth factors were used at the physician's discretion. This trial accrued patients with histologically confirmed measurable or assessable malignant melanoma with an Eastern Cooperative Oncology Group (ECOG) performance score of at most 2 and an estimated caloric intake of at least 1,200 kcal/day. To be considered measurable, a lesion must have had clearly measurable perpendicular diameters on physical examination or chest x-ray with ruler or palpable hepatomegaly. The minimum size of the largest tumor diameter was required to be 1.0 cm if measured by physical examination or chest x-ray; 3.0 cm if measured by computed tomography scan, magnetic resonance imaging scan, or ultrasound; or 5.0 cm if measured by radioisotope liver scan. For palpable hepatomegaly, the liver must have been palpable at least 5 cm below the costal margin in the midclavicular line or below the xiphoid process on quiet respiration. A lesion was considered to be assessable if it was clinically apparent but not bidimensionally measurable. Contraindications to study entry included a leukocyte count less than 4,000 cells/µL; platelet count less than 130,000 platelets/µL; serum creatinine greater than 25% above the upper limit of institutional normal limit; total bilirubin greater than 1.5 times upper limit of institutional normal limit or AST greater than two times the upper limit of the institutional normal limit; major surgery less than 3 weeks before registration; any prior chemotherapy; prior radiation therapy of more than 15% of bone marrow; weight loss of more than 10% within 3 months before registration; CNS metastases, ascites, or pleural effusion as the sole indicator lesion; nonmelanomatous malignant disease within 5 years before registration except for carcinoma-in-situ or basal-cell skin cancer; and active infection. Pregnant or lactating women were not eligible for this trial. This trial was approved by the institutional review boards of all participating institutions, and written informed consent was obtained from each patient before study entry. Patients were randomly assigned to a treatment group using a dynamic allocation procedure that was designed to balance the marginal distribution of dominant metastatic disease (visceral v nonvisceral), type of indicator lesion (measurable v assessable), performance status (< 2 v 2), and institution.11 Before randomization and subsequent re-treatments every 3 weeks, patients underwent a complete medical examination, including measurement of the indicator lesion, hematologic and chemistry group readings (WBC, platelets, hemoglobin, creatinine, total bilirubin, AST, and alkaline phosphatase), toxicity monitoring, and chest x-rays. Creatinine clearance was also obtained if creatinine was found to be more than 25% above pretreatment levels. Pelvic examinations were to be performed annually for all female patients. At each evaluation, a patient with measurable disease was classified as having either a complete response (CR), partial response (PR), stable disease (STAB), or progression of disease (PROG). CR was defined as the disappearance of all evidence of tumor. PR was defined as at least a 50% reduction in the product of the perpendicular diameters of the indicator lesions or at least a 30% reduction in the sum of the linear measurements of the liver below both costal margins in the midclavicular lines and xiphoid without progression in any lesion or appearance of new lesions. PROG was defined as 25% or more increase in the product of the perpendicular diameters of the indicator lesion if the criteria for CR or PR had not been met, an increase in the tumor at the point of its maximum reduction that was 50% or more of the size of the maximum tumor reduction if the criteria for CR or PR had been met, the appearance of a new lesion, or a significant clinical deterioration such as weight loss of more than 5% of body weight, worsening of tumor-related symptoms, a decline in performance status of two or more levels, or need for palliative radiation therapy. STAB was defined as failure to meet the criteria for CR, PR, or PROG. A patient with assessable disease was classified as either CR, regression (REGR), STAB, or PROG. CR was defined as the complete disappearance of all evidence of disease. REGR was defined as a definite decrease in tumor size without the appearance of new lesions. PROG was defined as the definite increase in tumor size, the appearance of new lesions, or significant clinical deterioration. STAB was defined as no definitive decrease or increase in tumor size or appearance of new lesions. Time to progression was defined as the time from registration to disease progression. Patients who died without documentation of progression were censored at the date their disease status was last evaluated (eight patients). Three patients died before having an evaluation. One of these three patients was considered to have progressed on their death date, having died 3 days after study entry because of hypoxia and cardiac arrest secondary to lung metastases. The other two patients were censored at their death date: one died of a pulmonary embolism 11 days after study entry and the other died 32 days after entry onto the study after a cardiovascular collapse. Survival time was defined as the time from registration to death. A patient was considered to have achieved an objective response if the patient maintained a CR, PR, or REGR on two consecutive evaluations at least 3 weeks apart. The duration of treatment response was defined as time from the date at which the response was first noted until the date at which disease progression was noted. Confidence intervals (CI) for the true proportion of responses were constructed using the properties of the binomial distribution. Time-to-event distributions were estimated using the Kaplan-Meier method.12
Fisher's exact test was used to assess whether response rates differed with respect to treatment group.13 The log-rank test was used to assess whether the distribution of response time, progression-free survival (PFS) time, or overall survival (OS) time differed with respect to treatment.14 For PFS and OS, a univariate Cox's proportional hazards model was fit to the data to obtain an estimate of the risk of the event for the CDB + TAM arm relative to that for the CDB arm.15 The following factors were assessed for their prognostic value in terms of response duration, PFS, and OS: age (
There were 188 patients accrued onto this study between September 1991 and April 1996. Two patients (1%) were declared ineligible: one patient had received prior radiation therapy to more than 15% of his bone marrow, and the other patient had a breast primary tumor. Two patients (1%) were removed from study after they refused their treatment assignments. The pretreatment characteristics of the remaining 184 eligible patients are listed in Table 1.
Response
The risk of not responding to treatment was found univariately to be significantly increased for those patients with an assessable indicator lesion (P = .024). The risk of not responding to treatment was not found to differ univariately with respect to type of dominant disease, sex, age 60 years or older at randomization, or ECOG performance status. Multivariate logistic regression analysis indicated that the risk of not responding was significantly increased for those patients who had visceral disease and an assessable indicator lesion. After these two factors were adjusted for, the risk of not responding was not found to differ significantly between the two treatment regimens (P = .451).
Disease Progression
The progression-free survival was found univariately to be significantly increased for patients with an ECOG performance score of at most 1 and for patients with nonvisceral dominant disease. Progression-free survival was not found to differ univariately with respect to sex, age 60 or more years at registration, or type of indicator lesion. Multivariate Cox regression analysis revealed three risk groups. They are listed in descending order of risk: patients with a performance score of 2, patients with a performance score less than 2 and visceral disease, and patients with a performance score less than 2 and nonvisceral disease. After these risk factors were adjusted for, there was no evidence to suggest that progression-free survival differed with respect to treatment regimen (P = .451). The adjusted hazards ratio of CDB + TAM to CDB was estimated to be 1.13 (95% CI, 0.83 to 1.53).
Survival
Survival was found univariately to be significantly increased for patients with an ECOG performance score of at most 1 and for patients with nonvisceral dominant disease. Survival was not found to differ univariately with respect to sex, age 60 or more years at registration, or type of indicator lesion. Multivariate Cox regression analysis revealed three risk groups. They are listed in descending order of risk: patients with a performance score of 2, patients with a performance score less than 2 and visceral disease, and patients with a performance score less than 2 and nonvisceral disease. After these risk factors were adjusted for, there was no evidence to suggest that survival differed with respect to treatment regimen (P = .592). The adjusted hazards ratio of CDB + TAM to CDB was estimated to be 1.09 (95% CI, 0.80 to 1.47).
Toxicity Severe (grade 3 or higher) nonhematologic toxicities reported on each treatment arm were similar. They were lethargy (12% CDB + TAM v 9% CDB), nausea (10% v 9%), vomiting (9% v 3%), and neuromotor (5% v 4%). A grade 3 gastrointestinal bleed and a grade 2 phlebitis was also reported on the CDB + TAM arm.
Administration of Chemotherapy
Studies conducted by Del Prete et al8 and McClay et al9 of the addition of TAM to the three-agent regimen of CDB for the treatment of malignant melanoma were encouraging. Response rates of 50% to 55% were achieved, but these trials treated a small number of patients and a limited number of institutions participated. Also, the incidence of thromboembolic events, associated with the addition of TAM, was worrisome. A similar response rate was achieved by Lattanzi et al.17 They reported a response rate of 54% in 26 patients, but these patients were part of a larger trial in which the study design was changed twice. First the dose of CDDP was increased and then TAM was eliminated. This study suffered from not only small numbers of patients and few participating institutions but changes in the study design as well. Margolin et al18 reported the results of a Southwest Oncology Group phase II trial of CDB and TAM. The response rate achieved among the 79 assessable melanoma patients was 15%, considerably less than that previously reported. They attributed the lower response rate to differences seen in patients who enroll in trials with fewparticipating institutions and patients who enroll in cooperative group trials. In 1991, Rusthoven et al19 opened a randomized double-blind placebo-controlled trial to assess the impact of the addition of TAM to CDB on response. Their response rates were similar to those in our study, with a 30% response rate among the 104 patients receiving CDB + TAM and a 21% response rate among the 100 patients receiving CDB + placebo. They concluded that the addition of TAM to CDB did not significantly increase the response rate relative to that of the three-agent regimen alone. In 1991, we, the North Central Cancer Treatment Group, opened a randomized phase III clinical trial to assess the relative efficacy of the addition of TAM to CDB, not only in terms of response but also progression-free survival and overall survival. Using the schedule of the agents in this study, we could not demonstrate any meaningful advantage from the addition of TAM. The response rate, progression-free survival, and overall survival of each study arm were comparable. Unfortunately for patients with malignant melanoma, we cannot offer the prolongation of life with the addition of TAM to CDB compared with CDB alone. Emerging developments from phase I trials and the development of genetically oriented technologies may provide promise for patients with disseminated malignant melanoma.
Additional participating institutions include the following: Allan Blair Cancer Centre, Regina, Saskatchewan, Canada (Andrew Maksymiuk, MD); Iowa Oncology Research Association Community Clincal Oncology Program (CCOP), Des Moines, IA (Roscoe F. Morton, MD); Meritcare Hospital CCOP, Fargo, ND (Ralph Levitt, MD); Ochsner Community Clinical Oncology Program, New Orleans, LA (Carl G. Kardinal, MD); Toledo Community Clinical Oncology Program, Toledo, OH (Paul L. Schaefer, MD); Sioux Community Cancer Consortium, Sioux Falls, SD (Loren K. Tschetter, MD); Quain and Ramstad Clinic, Bismarck, ND (Delano M. Pfeifle, MD); Cedar Rapids Oncology Project CCOP, Cedar Rapids, IA (Martin Wiesenfeld, MD); Nebraska Oncology Group-Creighton University, University of Nebraska Medical Center, and Associates, Omaha, NE (James A. Mailliard, MD); CentraCare Clinic, St Cloud, MN (Harold E. Windschitl, MD); Mayo Clinic Scottsdale CCOP, Scottsdale, AZ (Richard Wheeler, MD); Rapid City Regional Oncology Group, Rapid City, SD (Larry P. Ebbert, MD); Ann Arbor Regional CCOP, Ann Arbor, MI (Philip J. Stella, MD); and Grand Forks Clinic, Ltd, Grand Forks, ND (Daniel J. Walsh, MD).
This study was conducted as a collaborative trial of the North Central Cancer Treatment Group and Mayo Clinic, and was supported in part by United States Public Health Service grants no. CA-25224, CA-37404, CA-15083, CA-35113, CA-35195, CA-35269, CA-35103, CA-35101, CA-37417, CA-35272, CA-35415, CA-52352, CA-60276, CA-63848, and CA-63849 from the National Cancer Institute, Department of Health and Human Services.
1. McClay EF, McClay ME: Systemic chemotherapy for the treatment of metastatic melanoma. Semin Oncol 23:744-753, 1996[Medline] 2. Hansson J: Systemic therapy of malignant melanoma. Med Oncol 14:73-81, 1997[Medline] 3. Anderson CM, Buzaid AC, Legha SS: Systemic treatments for advanced cutaneous melanoma. Oncology 9:1149-1168, 1995[Medline] 4. Philip PA, Flaherty L: Treatment of malignant melanoma with interleukin-2. Semin Oncol 24:S32-S38, 1997 (1 suppl 4) 5. Kirkwood JM, Agarwala SS: Systemic cytotoxic and biologic therapy of melanoma. PPO Updates 7:1-16, 1993 6. Eton O, Talpaz M, Lee KH, et al: Phase II trial of recombinant human interleukin-2 and interferon-alpha-2a: Implications for the treatment of patients with metastatic melanoma. Cancer 77:893-899, 1996[Medline] 7. Punt CJ, van Herpen CM, Jansen RL, et al: Chemoimmunotherapy with bleomycin, vincristine, lomustine, dacarbazine (BOLD) plus interferon alpha for metastatic melanoma: A multicentre phase II study. Br J Cancer 76:266-269, 1997[Medline] 8. Del Prete SA, Maurer LH, O'Donnell J, et al: Combination chemotherapy with cisplatin, carmustine, dacarbazine, and tamoxifen in metastatic melanoma. Cancer Treat Rep 68:1403-1405, 1984[Medline] 9. McClay EF, Mastrangelo MJ, Bellet RE, et al: Combination chemotherapy and hormonal therapy in the treatment of malignant melanoma. Cancer Treat Rep 71:465-469, 1987[Medline] 10. McClay EF, Mastrangelo MJ, Sprandio JD, et al: The importance of tamoxifen to a cisplatin-containing regimen in the treatment of metastatic melanoma. Cancer 63:1292-1295, 1989[Medline] 11. Pocock SL, Simon R: Sequential treatment assignment with balancing for prognostic factors in controlled clinical trials. Biometrics 31:103-115, 1975[Medline] 12. Kaplan E, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:547-581, 1958 13. Fisher RA: Statistical methods for research workers. Edinburgh, Scotland, United Kingdom, Oliver and Boyd, 1941 14. Peto R, Peto J: Asymptotically efficient rank invariant procedures (with discussion). J R Stat Soc 135:185-207, 1972 15. Cox DR: Regression models and life-tables. J R Stat Soc 34:187-220, 1972
16.
Walker SH, Duncan DB: Estimation of the probability of an event as a function of several independent variables. Biometrika 54:167-179, 1967 17. Lattanzi SC, Tosteson T, Chertoff J, et al: Dacarbazine, cisplatin and carmustine, with or without tamoxifen, for metastatic melanoma: 5-year follow-up. Melanoma Res 5:365-369, 1995[Medline] 18. Margolin KA, Liu PY, Flaherty LE, et al: Phase II study of carmustine, dacarbazine, cisplatin, and tamoxifen in advanced melanoma: A Southwest Oncology Group study. J Clin Oncol 16:664-669, 1998[Abstract]
19.
Rusthoven JJ, Quirt IC, Iscoe NA, et al: Randomized, double-blind, placebo-controlled trial comparing the response rates of carmustine, dacarbazine, and cisplatin with and without tamoxifen in patients with metastatic melanoma: National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 14:2083-2090, 1996 Submitted December 2, 1998; accepted February 16, 1999. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 1999 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
60 years, < 60 years), sex, dominant disease (visceral v nonvisceral), ECOG performance status (0 v 1 v 2), and type of indicator lesion (measurable v assessable). For each of these factors, a univariate logistic regression model was fit to the response data to assess whether the response rates differed with respect to that factor.
