|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Multicenter Randomized Phase III Trial Comparing Protracted Venous Infusion (PVI) Fluorouracil (5-FU) With PVI 5-FU Plus Mitomycin in Inoperable Pancreatic CancerByFrom the Royal Marsden Hospital, Surrey; Cookridge Hospital, Leeds; Royal Bournemouth Hospital, Bournemouth; Salisbury General Hospital, Salisbury; and Maidstone General Hospital, Maidstone, United Kingdom. Address reprint requests to D. Cunningham, MD, Department of Medicine, Royal Marsden Hospital, Downs Rd, Sutton, Surrey, SM2 5PT United Kingdom; email: dcunn{at}icr.ac.uk
PURPOSE: To compare protracted venous infusion (PVI) fluorouracil (5-FU) with PVI 5-FU plus mitomycin (MMC) in patients with advanced pancreatic cancer in a multicenter, prospectively randomized study. PATIENTS AND METHODS: Two hundred eight patients were randomized to PVI 5-FU (300 mg/m2/d for a maximum of 24 weeks) or PVI 5-FU plus MMC (7 mg/m2 every 6 weeks for four courses). The major end points were tumor response, survival, toxicity, and quality of life (QOL). RESULTS: The two treatment groups were balanced for baseline demographic factors, and 62% had metastatic disease. The overall response rate was 8.4% (95% confidence interval [CI]) 3.2% to 13.7% for patients treated with PVI 5-FU alone compared with 17.6%; 95% CI 10.3% to 25.1% for PVI 5-FU plus MMC (P = .04). Median failure-free survival was 2.8 months for PVI 5-FU and 3.8 months for PVI 5-FU plus MMC (P = .14). Median survival was 5.1 months for PVI 5-FU and 6.5 months for PVI 5-FU plus MMC (P = .34). Toxicities in both arms were mild. There was an increased incidence of neutropenia in the 5-FU plus MMC arm (P < .01), although no differences in infection were seen. No patients developed hemolytic uremic syndrome. Global QOL improved significantly after 24 weeks of treatment compared with baseline for patients receiving 5-FU plus MMC, although there was no statistically significant difference in QOL between arms. CONCLUSION: PVI 5-FU plus MMC resulted in a superior response rate in comparison with PVI 5-FU alone in advanced pancreatic cancer, but this did not translate into a survival advantage. These results emphasize the importance of chemotherapy in this setting and the continuing value of the fluoropyrimidines in pancreatic cancer.
APPROXIMATELY 5% to 10% of patients with pancreatic cancer present with potentially operable disease. However, a large proportion of these will have significant comorbidity that precludes an operation, and even in those that reach a laparotomy, many will have unsuspected tumor load that makes tumor resection technically unfeasible. The vast majority of patients therefore have inoperable disease in which the prognosis is extremely poor. Despite chemotherapy, the median survival can be counted in months if not weeks. However, randomized data do support the use of chemotherapy in this setting, a survival benefit having been demonstrated in comparison with best supportive care.1-3 Although improvement in survival is possible, it remains a palliative treatment and quality of life (QOL) is an important consideration before making the decision to treat. Several recent studies have shown that chemotherapy has a beneficial effect on QOL. Glimelius et al2 reported an improvement in QOL in 36% of patients treated with chemotherapy, compared with only 10% in the best supportive care arm. Gemcitabine has been shown to have a clinically beneficial response in 24% of patients, in comparison with only 5% in fluorouracil (5-FU)-treated patients.4 At our own institution, a symptomatic improvement was noted after treatment with cisplatin and protracted venous infusion (PVI) 5-FU.5 An improved performance status was seen in 34%, with improvement of weight loss (71%), nausea and vomiting (70%), and pain (60%) also reported. To date, gemcitabine is the only drug licensed for use in advanced pancreatic cancer in North America, largely as a result of the study by Burris et al4 published in 1997. In this study, patients with locally advanced or metastatic pancreatic cancer were randomized to receive either weekly gemcitabine (given for 7 of the first 8 weeks, and 3 of 4 weeks thereafter) or to a schedule of 5-FU given over 30 minutes, weekly at a dose of 600 mg/m2. Not only did the authors report a clinical benefit response advantage for the gemcitabine arm as described above but also a survival benefit. The 1-year overall survival (OS) for the gemcitabine and 5-FU arm was 18% and 2%, respectively. Despite these results, several centers, particularly in North Europe, continue to use 5-FU-based chemotherapy regimens. Several of these combinations have shown superiority over best supportive care1-3 although to date trials have failed to demonstrate an advantage over single-agent 5-FU.6,7 Many investigators also believe that the regimen used in the Burris et al article was suboptimal. It is now recognized that the activity of 5-FU is schedule dependent. A meta-analysis of six randomized trials has demonstrated that continuous infusion 5-FU is superior to bolus 5-FU with respect to tumor response and survival in metastatic colorectal cancer.8 In addition, hematologic toxicity seems to be significantly less frequent in the infused regimens.9 The 30-minute 5-FU infusion used in the Burris et al article may in fact be inferior to a true bolus dose because of rapid degradation of the drug.10 A dose of 300 mg/m2/d was chosen for this study based on previous data demonstrating this to be both well tolerated and associated with a high response rate.11 Previous data have suggested that mitomycin (MMC) used as a single agent may have a similar response rate to 5-FU,12 and there is evidence for in vitro synergy between these drugs.13 Early phase II data in gastrointestinal cancers suggested that clinical results could be enhanced with response rates of 33% to 39% using this combination of drugs. However, there are no published randomized data investigating the combination of 5-FU and MMC in advanced pancreatic cancer. This study was designed to test the hypothesis that a combination of PVI 5-FU and MMC is a superior chemotherapy regimen than PVI 5-FU alone in patients with inoperable pancreatic cancer by using the end points of response rate, survival, and toxicity.
Patient Eligibility Patients were required to have histologically confirmed locally advanced or metastatic carcinoma of the pancreas that was not amenable to surgical resection or radical radiotherapy. They had to have adequate bone marrow reserve (platelets > 100 x 109/L, WBC count > 3 x 109/L, and neutrophils > 1.5 x 109/L), renal function (serum creatinine < 132 µmol/L, urea < 10.7 mmol/L), and hepatic function (bilirubin < 30 µmol/L). Patients had to have a good performance status (PS) (Eastern Cooperative Oncology Group PS 0 to 2), life expectancy of more than 3 months, and no intercurrent uncontrolled medical illnesses. They were excluded if there were intracerebral metastases, current alcohol or other drug abuse, history of other malignancy (apart from adequately treated nonmelanotic skin cancer, or carcinoma-in-situ of the uterine cervix), uncontrolled angina pectoris or clinically significant cardiac dysrhythmias, or any psychological condition precluding informed consent. Before randomization, written informed consent was obtained from all patients. The study was approved by the local research and ethics committee at each of the five participating centers.
Patient Randomization
Pretreatment Evaluation, Assessment During Treatment, and Follow-Up During the study patients were monitored every 3 weeks with medical history, physical evaluation, full blood count and serum chemistry. In addition, CT scans were performed every 12 weeks.
Intravenous Access
Chemotherapy
Evaluation of Response
Toxicity Evaluation and Dose Modification MMC was delayed for 1 week if WBC count was less than 3 x 109/L or platelet count less than 100 x 109/L. Hematologic toxicities were managed by reducing MMC doses as follows: grade 3 infection with neutropenia, 25% dose reduction; and grade 4 infection with neutropenia, 50% dose reduction. MMC was stopped if there was evidence of HUS or RBC fragmentation on peripheral blood film.
QOL
Statistical Methods
A total of 209 patients were randomized between July 1994 and October 2000. Four patients were ineligible due to nonpancreatic cancer pathology (one esophageal adenocarcinoma, one chronic pancreatitis, and one no histologic diagnosis) and due to inadequate PS (one patient). Two patients withdrew from treatment after randomization. The two groups were well matched for baseline characteristics (Table 1). Metastatic disease was present in 65% of patients treated with 5-FU alone and 56% in those treated with 5-FU and MMC.
Delivery of Chemotherapy The mean duration of chemotherapy was 12 and 13 weeks for the PVI and combination arms, respectively (P = .27). Dose-intensity of 5-FU was 40.9% of the intended total dose over 24 weeks in the 5-FU alone arm compared with 44.7% in the 5-FU and MMC arm (P = .27). Treatment interruptions occurred in 61.5% of patients receiving 5-FU alone compared with 58.8% of patients receiving 5-FU and MMC (P = .9).
Tumor Response
Multivariate logistic regression analysis showed that treatment arm (hazard ratio [HR], 3.2; 95% CI, 1.13 to 8.87) and PS (HR, 0.36; 95% CI, 0.16 to 0.78) were found to be statistically significant predictors of response. Disease extent (ie, locally advanced v metastatic) did not significantly predict for response.
Survival
Treatment with 5-FU plus MMC resulted in no improvement in OS. The median survival was 5.1 months and 6.5 months with 1-year survival of 23.5% (95% CI, 15.7% to 32.2%) and 26.2% (95% CI, 18.0% to 35.2%) for PVI 5-FU and PVI 5-FU plus MMC, respectively (P = .338) (Fig 2). One-year survival for all patients with locally advanced disease was 37.8% (95% CI, 27.1% to 48.4%), and was 32% (95% CI, 17.2% to 47.8%) and 43% (95% CI, 28.2% to 56.9%) for patients in the PVI 5-FU alone arm and PVI 5-FU plus MMC arm, respectively. One-year survival for all patients with metastatic disease was 16.9% (95% CI, 10.7% to 24.2%) and was 19.4% (95% CI, 10.9% to 29.8%) and 13.8% (95% CI, 6.1% to 24.5%) in the 5-FU alone arm and 5-FU plus MMC arm, respectively.
Multivariate Cox regression analysis showed that poor PS (HR, 1.4; 95% CI, 1.074 to 1.825), the presence of metastatic disease (HR, 1.883; 95% CI, 1.309 to 2.708), and alkaline phosphatase (HR, 1.001; 95% CI, 1.000 to 1.002) predicted survival. Treatment arm did not predict survival.
Toxicity
Only 11 patients developed RBC fragmentation in the peripheral blood film. This included five patients in the 5-FU arm and six in the 5-FU plus MMC arm, but no patients developed HUS in this series.
Central Venous Catheter Complications
Symptom Response Of the 90% of patients who were symptomatic at baseline, a response was observed in the majority, with no significant difference observed between study arms (Table 5).
QOL At baseline, 12 weeks, and 24 weeks, no significant differences in QOL between treatment arms were observed. However, at 24 weeks, global QOL was significantly superior to the pretreatment value in the 5-FU/MMC arm (P = .035). The pain score was also significantly improved at 24 weeks in comparison with baseline levels in the combination arm (P = .048). There was significantly less dyspnea at 12 weeks in comparison with pretreatment levels in the combination arm (P = .033), but no significant difference was noted between arms.
Second-Line Therapy and Response
This study has demonstrated that the combination of PVI 5-FU and MMC results in a superior response rate than 5-FU alone in patients with inoperable carcinoma of the pancreas. However, the results of this study failed to demonstrate a difference in OS between the two arms. Global QOL at 24 weeks was significantly superior to pretreatment values in the combination arm, although there was no difference between arms. Both treatments offered considerable symptomatic benefit with little in the way of grade 3 or 4 toxicity. The rate of neutropenia was significantly higher in the 5-FU/MMC arm, although still low at 3.2%. After the publication of the trial by Burris et al4 in 1997, gemcitabine has been widely adopted as the standard of care in the first-line treatment of inoperable pancreatic cancer. Before this, 5-FU was the only drug that had demonstrated significant single-agent activity. Three studies suggest that 5-FU-based combination therapy may result in longer survival than best supportive care. Palmer et al1 reported that 5-FU, doxorubicin, and MMC prolonged median OS (33 weeks v 15 weeks, P = .002). Similarly, Glimelius et al2 reported longer survival with 5-FU and folinic acid with or without etoposide in comparison with best supportive care (median 6 months v 2.5 months, P < .01). Finally, Mallinson et al3 reported superior survival with a five-drug regimen (5-FU, MMC, methotrexate, vincristine, and cyclophosphamide). The median OS in the chemotherapy arm was 44 weeks compared with 9 weeks in the control arm. A subsequent study failed to demonstrate an advantage for this regimen over single-agent bolus 5-FU,6 and similarly, in a three-arm study, the same schedule of 5-FU was not found to be inferior to 5-FU and doxorubicin or 5-FU, doxorubicin, and MMC.7 The study by Burris et al4 was reported as showing a statistical advantage for gemcitabine over single-agent 5-FU with respect to response, clinical benefit, and 1-year OS. While discussing the relative merits of chemotherapy regimens in pancreatic cancer, it is important to be wary when comparing response rates. Pancreatic adenocarcinoma is commonly associated with a marked desmoplastic reaction that can greatly overestimate the malignant cell mass. The borders of the tumor may be indistinct because of architectural changes such as those associated with pancreatitis and cyst formation, and anatomic location and proximity to the small bowel can all make an accurate response assessment difficult. So although the response rate for single-agent 5-FU in the present study seems to be superior to that of gemcitabine in the Burris et al trial (8.4% v 5.4%), this result must be interpreted with caution. A much more robust measure of efficacy is survival. Although there was no statistical difference in OS in the two arms of the present study, the 1-year survival for the 5-FU alone arm was 23.5%, which compares favorably with the gemcitabine arm in the Burris et al trial (18%). The presence of metastatic disease was found to be a significant independent poor prognostic factor for survival in this study. The proportion of patients with metastatic disease seems to be lower in the present study compared with the Burris et al article. In the gemcitabine arm, 72% of patients had stage IV disease compared with 64% with stage IVb in the 5-FU alone arm of the present study. It could be argued that the lower incidence of metastatic disease in the present study contributed to the relatively favorable 1-year OS of 23.3%. However, it is not clear in the Burris et al article how many patients had truly metastatic disease (ie, stage IVb disease). Second, in a subgroup analysis of the present study, the 1-year OS in patients with metastatic disease (n = 124) remained comparable to the results of Burris et al at 16.9%. Of the patients with metastatic disease treated with PVI 5-FU alone (n = 68), the 1-year OS was 19.4%. The 1-year OS in patients with locally advanced disease was 37.8%, and 32% in the PVI 5-FU alone arm. The low rate of toxicities and serious central venous catheter complications in this study underlines how well tolerated PVI 5-FU with or without MMC is in this palliative setting. The rate of grade 3 or 4 hematologic toxicities was extremely low, and other than lethargy, there was little in the way of grade 3 or 4 symptomatic toxicity. It must be remembered that the majority of the patients in this study were of good PS (72% PS 0 to 1). Not only is this likely to influence the tolerability of chemotherapy, but as was demonstrated in the multivariate Cox analysis, PS is a significant independent prognostic factor. A subgroup of patients of PS 2 will therefore have a particularly poor prognosis, and chemotherapy may not be indicated in this cohort. Such decisions must be taken on clinical grounds and after a full and honest discussion with the patient. Although we have reported a symptomatic benefit to treatment, it must be remembered that there was no control of other interventions that may influence symptom response such as analgesic use. However, this is unlikely to affect the relative efficacy of the two arms in the study, and indeed we failed to demonstrate a statistically significant difference between arms. It is widely recognized that there are limitations in QOL assessment in studies such as this. One of the greatest problems is the rate of attrition throughout the study period. For example, in this study, data were available for global QOL in 138 patients at baseline, 51 patients at 12 weeks, and only 29 patients at 24 weeks. The small patient number at later time points and the fact that those with superior QOL may be more likely to survive longer makes interpretation difficult. To date, the only direct comparison of single-agent gemcitabine and 5-FU in a phase III randomized controlled study has reported a significant superiority for gemcitabine.4 However, it is now well recognized that the efficacy of 5-FU is schedule dependent, with infused regimens being associated with a significantly higher response rate and a lower rate of toxicity in comparison with bolus regimens. This large randomized controlled study demonstrates the tolerability and clinical efficacy of infused 5-FU in pancreatic cancer and reinforces the continuing value of fluoropyrimidines in pancreatic cancer. The addition of MMC does not seem to offer a survival advantage, although response rates may be superior. Further research is required in an attempt to improve these results. The recent introduction of the oral fluoropyrimidines offers the exciting prospect of a useful, well-tolerated treatment, and the combination of these preparations with gemcitabine may prove to be an effective treatment strategy.
1. Palmer KR, Kerr M, Knowles G, et al: Chemotherapy prolongs survival in inoperable pancreatic carcinoma. Br J Surg 81: 882-885, 1994[Medline]
2. Glimelius B, Hoffman K, Sjoden PO, et al: Chemotherapy improves survival and quality of life in advanced pancreatic and biliary cancer. Ann Oncol 7: 593-600, 1996
3. Mallinson CN, Rake MO, Cocking JB, et al: Chemotherapy in pancreatic cancer: Results of a controlled, prospective, randomised, multicentre trial. Br Med J 281: 1589-1591, 1980
4. Burris HA III, Moore MJ, Andersen J, et al: Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: A randomized trial. J Clin Oncol 15: 2403-2413, 1997
5. Nicolson M, Webb A, Cunningham D, et al: Cisplatin and protracted venous infusion 5-fluorouracil (CF): Good symptom relief with low toxicity in advanced pancreatic carcinoma. Ann Oncol 6: 801-804, 1995 6. Cullinan S, Moertel CG, Wieand HS, et al: A phase III trial on the therapy of advanced pancreatic carcinoma: Evaluations of the Mallinson regimen and combined 5-fluorouracil, doxorubicin, and cisplatin. Cancer 65: 2207-2212, 1990[CrossRef][Medline]
7. Cullinan SA, Moertel CG, Fleming TR, et al: A comparison of three chemotherapeutic regimens in the treatment of advanced pancreatic and gastric carcinoma: Fluorouracil vs fluorouracil and doxorubicin vs fluorouracil, doxorubicin, and mitomycin. JAMA 253: 2061-2067, 1985
8. Meta-Analysis Group in Cancer: Efficacy of intravenous continuous infusion of fluorouracil compared with bolus administration in advanced colorectal cancer. J Clin Oncol 16: 301-308, 1998 9. Meta-Analysis Group in Cancer: Toxicity of fluorouracil in patients with advanced colorectal cancer: Effect of administration schedule and prognostic factors. J Clin Oncol 16: 3537-3541, 1998[Abstract] 10. Permert J, Hafstrom L, Nygren P, et al: A systematic overview of chemotherapy effects in pancreatic cancer. Acta Oncol 40: 361-370, 2001[Medline] 11. 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] 12. Crooke ST, Bradner WT: Mitoycin C: A review. Cancer Treat Rev 3: 121-139, 1976[CrossRef][Medline]
13. Sartorelli AC, Boothe BA: The synergistic anti-neoplastic activity of combinations of mitomycins with either 6-thioguanine or 5-fluorouracil. Cancer Res 25: 1393-1400, 1965 14. Miller AB, Hoogstraten B, Staquet M, et al: Reporting results of cancer treatment. Cancer 47: 207-214, 1981[CrossRef][Medline] 15. Macdonald JS, Haller DG, Mayer RJ: Grading of Toxicity. Philadelphia, PA, Lippincott, 1995, pp 519-523 16. Ringdal GI, Ringdal K: Testing the EORTC Quality of Life Questionnaire on cancer patients with heterogeneous diagnoses. Qual Life Res 2: 129-140, 1993[CrossRef][Medline]
17. Ross P, Norman A, Cunningham D, et al: A prospective randomised trial of protracted venous infusion 5-fluorouracil with or without mitomycin C in advanced colorectal cancer. Ann Oncol 8: 995-1001, 1997 18. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958[CrossRef] 19. Peto R, Peto J: Asymptotically efficient invariant procedures. J R Stat Soc A135: 185-206, 1972 Submitted September 4, 2001; accepted March 21, 2002.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test with Fisher’s exact test used where appropriate. Patient OS (defined as time from randomization to death from any cause) and FFS (time to progression or death from the time of randomization) were examined with the Kaplan-Meier product limit method,
