|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Phase III Study of Cyclophosphamide, Doxorubicin, and Fluorouracil (CAF) Plus Leucovorin Versus CAF for Metastatic Breast Cancer: Cancer and Leukemia Group B 9140
From the National Cancer Institute, Bethesda, MD; University of Texas M.D. Anderson Cancer Center, Houston, TX; CALGB Statistical Center, Durham; Wake Forest University School of Medicine, Winston-Salem, NC; Cancer Institute of New Jersey, New Brunswick; Hoffman LaRoche Inc, Nutley, NJ; Weill Medical College of Cornell University, Memorial Sloan-Kettering Cancer Center, New York; SUNY Upstate Medical Center, Syracuse; North Shore University Hospital, Manhasset, NY; University of Missouri/Ellis Fischel Cancer Center, Columbia, MO; and University of California at San Francisco, San Francisco, CA. Address reprint requests to Howard L. Parnes, MD, Division of Cancer Prevention, National Cancer Institute, National Institutes of Health, 6130 Executive Plaza EPN Room 2100, Rockville MD 20852; email: hp24c{at}nih.gov.
Purpose: To determine whether biochemical modulation with LV (leucovorin) enhances the efficacy of CAF (cyclophosphamide, doxorubicin, and fluorouracil) against metastatic breast cancer. Patients and Methods: Women with histologically confirmed stage IV breast cancer, Cancer and Leukemia Group B (CALGB) performance status 0 to 2, and no prior chemotherapy for metastatic disease were randomly assigned to receive CAF (cyclophosphamide 500 mg/m2 day 1, doxorubicin 40 mg/m2 day 1, and fluorouracil [FU] 200 mg/m2 intravenous bolus days 1 to 5) with or without LV (LV 200 mg/m2 over 30 minutes days 1 to 5 given 1 hour before FU). Results: Two hundred forty-two patients were randomly assigned to treatment; 124 patients had visceral crisis and 40 patients had a CALGB performance status score of 2. The median follow-up was 6 years. The two study arms were similar with regard to serious adverse events; four patients died from treatment-related causes, two patients on each study arm. Predictive variables for time to treatment failure and survival were visceral disease and performance status. The overall response rate was 29% for CAF versus 28% for CAF plus LV. The median time to treatment failure (9 months) and median survival (1.7 years) did not differ by treatment arm. Conclusion: Modulation of CAF with LV improved neither response rates nor survival among women with metastatic breast cancer, compared with CAF alone. Multivariate analyses confirmed the prognostic importance of performance status and visceral crisis. However, the overall and complete response rates, response durations, time to treatment failure, and survival were the same in the two treatment arms.
CYCLOPHOSPHAMIDE, DOXORUBICIN, and fluorouracil (CAF) is an active regimen for the initial therapy of women with advanced breast cancer;1–4 however, given the short half-life of fluorouracil (FU), the intermittent bolus administration of this agent is not likely to be optimal. One potential method of enhancing the efficacy of FU is by continuous infusion administration. Although infusion studies in colon cancer5,6 have indicated an improved response for this mode of delivery, continuous infusion therapy is inconvenient, requires continued venous access or pump devices, and is associated with infectious complications and risk of thrombosis. Another approach for enhancing the activity of FU is through biochemical modulation with leucovorin (LV), a source of reduced folates.7 An important mechanism of FU cytotoxicity is through the inhibition of thymidylate synthetase, an enzyme necessary for DNA synthesis.8,9 Fluoro deoxyuridine monophosphate, a metabolite of FU, covalently binds to thymidylate synthetase and inhibits the formation of deoxythymidine triphosphate, one of the four nucleotides necessary for DNA synthesis. The interaction of FdUMP with thymidylate synthetase in the presence of N5–10-methylene tetrahydrofolic acid leads to the formation of a ternary complex resulting in the inactivation of the enzyme and the depletion of deoxythymidine triphosphate. In the absence of excess reduced folates, the ternary complex has a short half-life. LV increases the extracellular concentration of reduced folates, thus stabilizing the ternary complex and decreasing the concentration of free enzyme available for normal nucleotide synthesis.10 FU plus LV, given either weekly (FU 600 mg/m2 plus LV 500 mg/m2 6 weeks on, 2 weeks off)11 or on days 1 to 5 (FU 425 mg/m2 plus LV 20 mg/m2 every 4 weeks),12 became a standard regimen for the treatment of patients with colon cancer in the late 1980s and early 1990s.13 This combination had also been used in the treatment of metastatic breast cancer, with response rates ranging from approximately 20% in heavily pretreated patients to 48% in patients with no prior chemotherapy for metastatic disease.14–19 These trials used a variety of doses and schedules, and all showed some level of activity, indicating that the addition of LV could enhance the activity of CAF chemotherapy in women with advanced breast cancer. Although the initial trials of FU plus LV did not produce excessive hematologic toxicity, there was an increase in the incidence and severity of mucositis and diarrhea compared with the effects of FU alone. Therefore, we elected not to incorporate LV into the standard CAF regimen, in which a relatively high dose of FU (500 mg/m2) is given on days 1 and 8 of each 28-day cycle. Instead, a 21-day cycle was used with patients receiving cyclophosphamide (500 mg/m2) and doxorubicin (40 mg/m2) on day 1 and FU, with or without LV, on days 1 to 5 of each cycle. To determine the maximum-tolerated dose (MTD) of FU for use in this regimen, a phase I trial was conducted at the University of Maryland Cancer Center20 in which FU was escalated in cohorts of three patients beginning at 150 mg/m2 days 1 to 5 both in the presence and absence of LV 200 mg/m2 days 1 to 5. The MTD of FU in this study was 250 mg/m2 days 1 to 5 in the presence of LV and 200 mg/m2 in the absence of LV. Myelosuppression was the dose-limiting toxicity among patients treated with and without LV; however, FU myelotoxicity was reduced in the presence of LV, which is consistent with observations from prior studies.19,20 At the dosages used in this study, stomatitis and diarrhea were infrequent complications and were not dose limiting. Although the MTD for FU in the presence of LV was 250 mg/m2 days 1 to 5, the dose was limited to 200 mg/m2 days 1 to 5 in the phase III comparison study to provide a balanced comparison with the arm without leucovorin.
The study was open to accrual between September 1991 and August 1995. To be eligible, patients were required to have histologically documented female breast cancer, metastatic disease, and a Cancer and Leukemia Group B (CALGB) performance status (PS) score of 0, 1, or 2. Prior chemotherapy for metastatic disease was not allowed. To avoid competition with another CALGB breast cancer study, patients were also required to have any one of the following: visceral crisis (ie, lymphangitic lung metastases, bone marrow replacement, or significant liver metastases), nonmeasurable disease, or PS score of 2 (symptomatic, in bed less than 50% of the day). When the competing trial was closed to accrual in January 1994, these three eligibility qualifications were dropped. Institutional review boards at all participating centers approved the trial, and all patients gave written informed consent before being enrolled in the study.
Patients were randomly assigned with equal probability to receive CAF (cyclophosphamide 500 mg/m2 day 1, doxorubicin 40 mg/m2 day 1, FU 200 mg/m2 intravenous [IV] bolus, days 1 to 5), with or without LV (LV 200 mg/m2 in 100 mL IV fluid over 30 minutes days 1 to 5, completed 1 hour before FU was administered). This regimen was used to avoid the excessive mucosal toxicity that would have been expected had LV been incorporated into a standard 28-day CAF regimen. Cycles were repeated every 21 days provided the patient had recovered from drug toxicity; that is, granulocytes Patients continued to receive CAF or CAF plus LV until disease progression was documented or until a cumulative dose of 440 mg/m2 of doxorubicin (adjuvant plus metastatic) was reached. When a total cumulative dose of 440 mg/m2 of doxorubicin had been reached, stable and responding patients were switched to a standard cyclophosphamide, methotrexate, and fluorouracil regimen of the treating physician’s choice without LV. Patients received a total of 15 cycles of therapy for metastatic disease unless progressive disease was documented.
Statistical Methods
Overall response rate was the percentage of patients assessable for tumor response whose best response to therapy was complete or partial. Tumor measurements were required on the first day of each treatment cycle if accessible by physical exam. Chest radiographs were required every two cycles during the study (every cycle if results were abnormal because of metastases), and liver scans were required every two cycles if results were abnormal because of metastases. Confirmatory scans were required to document adequate duration of response, that is, more than 4 weeks, if not accessible by physical exam. Patients who did not have follow-up tumor assessments because of early death or early termination of treatment because of toxicity were considered assessable but nonresponders. Complete response (CR) was defined as the disappearance of all evidence of measurable disease for more than 4 weeks or disappearance of all nonmeasurable abnormalities. Partial response (PR) was defined as more than 50% reduction in the sum of the products of the perpendicular diameters of all measurable lesions. Progressive disease was defined as an increase of For those patients who responded to therapy (CR or PR), response duration was the length of time from either CR or PR, whichever occurred first, until disease progression or death from any cause. Surviving patients still in response were censored at the date last known to be free from disease. TTF was the interval from study entry to disease progression or death from any cause, whichever occurred first. Removal from study because of a deterioration of condition without documentation of progression was not considered an event for the TTF analysis. Progression-free survivors were censored at the date last known to be progression-free. Overall survival was the time from study entry until death from any cause. Surviving patients were censored at the date of last follow-up. Toxicity was graded according to the CALGB expanded common toxicity criteria.
Distributions of overall survival, TTF, and response duration were shown using the Kaplan-Meier product limit method.21 Multivariate proportional hazards regression models related prognostic variables with time-to-event end points.22 Models included treatment arm as well as recognized prognostic factors for metastatic breast cancer as candidate variables. All models included treatment arm; the remaining variables were selected using a forward or step-up procedure that required a significance level of
Patients A total of 242 patients were accrued to this study. One patient, randomly assigned to the CAF alone arm, never received treatment, and no data are available. She was therefore excluded from all analyses. Five patients were ineligible: three patients on the CAF-alone arm and two patients on the CAF plus LV arm. Reasons for ineligibility included that the patient began protocol treatment before registration or that baseline laboratory values were not obtained, were out of range, or were performed too long before treatment began. All 241 treated patients were included in analyses. The median follow-up was 6 years.
The pretreatment characteristics for patients by study arm are shown in Table 1
Adverse Events The two study arms were similar with regard to serious adverse events. Four patients died from treatment-related causes: two patients on the CAF arm (cardiac dysfunction and infection) and two patients on the CAF plus LV arm (also cardiac dysfunction and infection). The two deaths on the CAF arm occurred 5 and 9 months, respectively, after protocol treatment began. On the CAF plus LV arm, the patient who experienced lethal infection died 2 weeks after therapy began; the other patient died of congestive heart failure 43 months after therapy began. The most frequent toxicities for both treatment arms were hematologic. The two arms differed most regarding the frequency of grade 4 lymphocytopenia and grade 3 neutropenia. Grade 4 lymphocytopenia (< 500/µL) occurred in 54% of patients on the CAF arm versus 46% on the CAF plus LV arm. Grade 3 leukopenia was reduced among patients receiving CAF plus LV; however, grade 4 leukopenia (55% CAF v 54% CAF plus LV) was nearly identical in the two treatment groups. Nonhematologic toxicities were infrequent. The most common grade 3+ nonhematologic toxicities for patients receiving CAF versus CAF plus LV were infection (10% v 18%; P = .11), diarrhea (8% v 11%; P = .48), stomatitis (3% v 7%; P = .21), and dyspnea (3% v 10%; P = .062); however, none of these differences was of statistical significance. The only reported secondary malignancy was in a patient with contralateral breast cancer occurring 11 months after therapy with CAF alone was initiated.
Tumor Response
Among patients with measurable disease randomly assigned to CAF, there were seven CRs and 25 PRs for a response rate of 43% (95% CI, 32% to 55%); among those randomized to CAF plus LV, there were one CR and 30 PRs for a response rate of 36% (95% CI, 26% to 48%; P = .42). The response rate for patients with visceral crisis at study entry was 21% (95% CI, 12% to 34%) for CAF versus 27% for CAF plus LV (95% CI, 17% to 40%; P = .53). For patients without visceral crisis the response rates were 38% (95% CI, 24% to 53%) for CAF versus 29% (95% CI, 18% to 42%) for CAF plus LV (P = .33). Treatment arm did not influence length of response among responders (Fig 1  ). The median duration of response was 12 months for CAF and 11 months for CAF plus LV (P = .45).
Overall Survival Median survival was about 1.7 years (Fig 2 ) for both the CAF and CAF plus LV arms. In the univariate analysis, the addition of LV to CAF did not improve survival (P = .87). Table 3 summarizes the results of multivariate regression modeling. After the effects of the clinical variables indicated in Methods were adjusted for, survival was not different between the two treatment arms (P = .95). Important predictor variables for survival were visceral crisis (P < .0001), PS (P = .0002), and patient age (P = .015). Specifically, the hazard of death was 78% higher for patients with visceral involvement at study entry compared with patients without visceral disease (risk ratio = 1.78). A one-unit increase in performance level (from 0 to 1 or from 1 to 2) corresponds to a 43% increase in the hazard of death. For every 10 additional years in age, there is a 17% increase in hazard.
TTF Figure 3 shows that TTF did not differ by treatment arm (P = .91). The median TTF was 0.75 and 0.79 years for CAF and CAF plus LV, respectively. Table 4 gives the multivariate model for TTF. After other variables were controlled for, TTF was not different between the two treatment arms (P = .68). Variables significantly associated with TTF were visceral crisis (P = .0022) and performance score (P = .0014). The hazard of treatment failure was 57% higher for patients with visceral involvement compared with patients without visceral involvement (risk ratio = 1.57). With each unit increase in PS (from 0 to 1 or from 1 to 2) there was a 36% increase in the hazard of failure (risk ratio = 1.36).
CAF was one of the most active regimens available for the treatment of metastatic breast cancer in the early 1990s.1–4 However, despite the relatively high initial response rates observed with this and other regimens, most patients with advanced breast cancer eventually experience disease progression and die from their disease. Although the advantage of combination chemotherapy over standard single-agent therapy is generally accepted, methods to enhance the activity of FU could potentially improve the therapeutic index of this regimen. Biochemical modulation with LV increases the activity of FU in the treatment of colorectal cancer11,12,23 and had been studied with promising results in women with refractory breast cancer.14–19 Therefore, the CALGB conducted this randomized phase III study to determine whether biochemical modulation of FU with LV enhances the activity of CAF against metastatic breast cancer. The primary objectives of this randomized controlled trial were to compare the response rate, TTF, and survival of patients with metastatic breast cancer treated with CAF versus patients treated with CAF plus LV. Although the study population comprised poor-risk (PS 2, visceral crisis) as well as good-risk (PS 0 to 1, nonvisceral disease) patients, the treatment arms were well balanced for pretreatment variables such as disease burden, age, PS, and hormone receptors. The frequency and type of adverse events were similar in the two arms. The most frequent life-threatening (grade 4) toxicities were hematologic, and the frequency of granulocytopenia was equivalent in the two arms. The low overall incidences of grades 3 and 4 stomatitis and diarrhea and the observation that these toxicities were not significantly greater among patients receiving CAF plus LV than among those receiving CAF alone, raises two questions. First, was biochemical modulation of FU actually achieved with the dose and schedule of LV used? Second, was an effective dose of FU used? Regarding the first question, LV doses as low as 20 mg/m2 days 1 to 5 have been shown to significantly increase both response rates and survival of patients with metastatic colon cancer compared with the use of FU alone. In addition, in the phase I trial of CAF plus LV conducted in preparation for the current study, 200 mg/m2 of LV days 1 to 5 did reduce CAF myelotoxicity. In that study, none of six patients treated with CAF plus LV at the 250 mg/m2 dose level of FU had grade 4 myelotoxicity during the first cycle of chemotherapy.20 In contrast, three of six patients treated with the identical CAF regimen without LV experienced grade 4 myelotoxicity (unpublished data). This observation regarding the attenuation of FU myelotoxicity by LV is consistent with previously published data11,12 and indicates that the dose and schedule of LV used in this study were sufficient to produce biochemical modulation of FU. Of note, grades 3 and 4 mucosal toxicity were observed in the phase I study of CAF plus LV when the dose of FU was increased to 300 mg/m2 days 1 to 5. However, because dose-limiting myelotoxicity had been encountered at 250 mg/m2 of FU in the absence of LV, no comparative data are available regarding the mucosal toxicity of CAF at an FU dose of 300 mg/m2 without LV. Regarding the second question, it is possible that the outcome of the patients randomly assigned to the leucovorin arm would have been improved if they had received 250 mg/m2 rather than 200 mg/m2 of FU days 1 to 5. However, if the dose of FU were different in the two study arms, it would not have been possible to determine whether any observed benefit was because of the higher dose of FU or the addition of LV. Therefore, it was decided to use the same dose of FU in the two study arms. No differences in efficacy were observed between CAF and CAF plus LV. The overall response rate, CR rate, duration of response, TTF, and survival were the same in the two treatment arms. As expected, the response rates observed in this study are somewhat lower than those commonly reported for taxane-based regimens; however, the outcome measures of time to progression and overall survival are similar to those observed with more modern regimens.24,25 No differences were apparent among selected subgroups, such as those with visceral crisis versus those without visceral crisis or those with measurable versus assessable disease. Therefore, it is unlikely that the broad entry criteria used had any effect on the outcome of this trial. As expected, results of the multivariate regression analyses confirmed the prognostic importance of PS and visceral crisis. In conclusion, modulation of CAF with LV, in the dose and schedule used, improved neither response rates nor survival among patients with metastatic breast cancer compared with CAF alone.
Supported in part by grants from the National Cancer Institute (CA31946), National Institutes of Health, Department of Health and Human Services, Bethesda, MD, to the Cancer and Leukemia Group B. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
1. Aisner J, Weinberg V, Perloff M, et al: Chemotherapy versus chemoimmunotherapy (CAF v CAFVP v CMF each ± MER) for metastatic carcinoma of the breast: A CALGB study—Cancer and Leukemia Group B. J Clin Oncol 5:1523–1533, 1987 2. Aisner J, Cirrincione C, Perloff M, et al: Combination chemotherapy for metastatic or recurrent carcinoma of the breast: A randomized phase III trial comparing CAF versus VATH versus VATH alternating with CMFVP—Cancer and Leukemia Group B Study 8281. J Clin Oncol 13:1443–1452, 1995[Abstract] 3. Hayes DF, Henderson IC, Shapiro CL: Treatment of metastatic breast cancer: Present and future prospects. Semin Oncol 22:5–21, 1995[Medline]
4. Berry DA, Broadwater G, Klein JP, et al: High-dose versus standard chemotherapy in metastatic breast cancer: Comparison of Cancer and Leukemia Group B trials with data from the Autologous Blood and Marrow Transplant Registry. J Clin Oncol 20:743–750, 2002 5. Seifert P, Baker LH, Reed ML, et al: Comparison of continuously infused 5-fluorouracil with bolus injection in treatment of patients with colorectal adenocarcinoma. Cancer 36:123–128, 1975[CrossRef][Medline] 6. Lokich JJ, Ahlgren JD, Gullo JJ, et al: A prospective randomized comparison of continuous infusion fluorouracil with a conventional bolus schedule in metastatic colorectal carcinoma: A Mid-Atlantic Oncology Program Study. J Clin Oncol 7:425–432, 1989[Abstract] 7. Ardalan B, Luis R, Jaime M, et al: Biomodulation of fluorouracil in colorectal cancer. Cancer Invest 16:237–251, 1998[Medline] 8. Peters GJ, van der Wilt CL, van Triest B, et al: Thymidylate synthase and drug resistance. Eur J Cancer 31A:1299–1305, 1995[CrossRef][Medline] 9. Evans RM, Laskin JD, Hakala MT: Effect of excess folates and deoxyinosine on the activity and site of action of 5-fluorouracil. Cancer Res 41:3288–3295, 1981[Medline] 10. Lockshin A, Danenberg PV: Biochemical factors affecting the tightness of 5-fluorodeoxyuridylate binding to human thymidylate synthetase. Biochem Pharmacol 30:247–257, 1981[CrossRef][Medline] 11. Petrelli N, Douglass HO, Jr., Herrera L, et al: The modulation of fluorouracil with leucovorin in metastatic colorectal carcinoma: A prospective randomized phase III trial—Gastrointestinal Tumor Study Group. J Clin Oncol 7:1419–1426, 1989[Abstract] 12. Poon MA, O’Connell MJ, Moertel CG, et al: Biochemical modulation of fluorouracil: Evidence of significant improvement of survival and quality of life in patients with advanced colorectal carcinoma. J Clin Oncol 7:1407–1418, 1989[Abstract] 13. Einhorn LH: Improvements in fluorouracil chemotherapy? J Clin Oncol 7:1377–1379, 1989[Medline] 14. Swain SM, Lippman ME, Egan EF, et al: Fluorouracil and high-dose leucovorin in previously treated patients with metastatic breast cancer. J Clin Oncol 7:890–899, 1989[Abstract] 15. Marini G, Simoncini E, Zaniboni A, et al: 5-Fluorouracil and high-dose folinic acid as salvage treatment of advanced breast cancer: An update. Oncology 44:336–340, 1987[Medline] 16. Loprinzi CL, Ingle JN, Schaid DJ, et al: 5-Fluorouracil plus leucovorin in women with metastatic breast cancer: A phase II study. Am J Clin Oncol 14:30–32, 1991[Medline] 17. Doroshow JH, Leong L, Margolin K, et al: Refractory metastatic breast cancer: Salvage therapy with fluorouracil and high-dose continuous infusion leucovorin calcium. J Clin Oncol 7:439–444, 1989[Abstract] 18. Fine S, Erlichman C, Kaizer L, et al: Phase II trial of 5-fluorouracil and folinic acid in the treatment of advanced breast cancer. Breast Cancer Res Treat 30:205–209, 1994[CrossRef][Medline]
19. Zaniboni A, Meriggi F, Arcangeli G, et al: L-folinic acid and 5-fluorouracil in the treatment of advanced breast cancer: A phase II study. Ann Oncol 4:41–43, 1993 (suppl 2)
20. Parnes HL, Abrams JS, Tait N, et al: Phase I/II study of cyclophosphamide, doxorubicin, fluorouracil, and leucovorin for treatment of metastatic adenocarcinoma. J Natl Cancer Inst 83:1017–1020, 1991 21. Kaplan E, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481, 1958[CrossRef] 22. Cox D: Regression models and life tables. J R Stat Soc 34:187–220, 1972 23. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: Evidence in terms of response rate—Advanced Colorectal Cancer Meta-Analysis Project. J Clin Oncol 10:896–903, 1992[Abstract] 24. Perez EA, Hillman DW, Stella PJ, et al: A phase II study of paclitaxel plus carboplatin as first-line chemotherapy for women with metastatic breast carcinoma. Cancer 88:124–131, 2000[CrossRef][Medline] 25. Friedrichs K, Holzel F, Janicke F: Combination of taxanes and anthracyclines in first-line chemotherapy of metastatic breast cancer: An interim report. Eur J Cancer 38:1730–1738, 2002[CrossRef][Medline] Submitted May 16, 2002; accepted February 6, 2003. This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
1,800/µL, platelets 
.05 to be included in the final model. P values were based on Wald 
2 statistics (regression analyses) and log-rank statistics (Kaplan-Meier curves). Response rates with 95% confidence intervals (CIs) were calculated using exact binomial methods. We used Pearson correlations to assess the association of pairs of explanatory variables. We transformed variables, when appropriate, to increase their predictive ability. We analyzed continuous variables in regression analyses on a continuous scale but give two reference points for interpretation of risk ratios. 
