|
|||||
|
|
||||||
© 2000 American Society for Clinical Oncology Exploratory Phase III Study of Paclitaxel and Cisplatin Versus Paclitaxel and Carboplatin in Advanced Ovarian CancerFrom the University Medical Center Utrecht, Utrecht; Spaarneziekenhuis, Haarlem; Department of Medical Statistics, Leiden University Medical Centre, Leiden, the Netherlands; Finsen Center, Copenhagen University Hospital; Herlev Amtssygehus, University of Copenhagen, Copenhagen; Roskilde Hospital, Roskilde; Hillerød Hospital, Hillerød; Bispebjerg Hospital and Bristol Myers-Squibb, København, Denmark; and Ospedale San Giovanni, Bellinzona, Switzerland. Address reprint requests to J.P. Neijt, MD, Utrecht University Medical Center, locatie Academisch Ziekenhuis F02.126, PO Box 85500, 3508 GA Utrecht, the Netherlands; email j.p.neijt{at}digd.azu.nl
PURPOSE: To determine the side effects and feasibility of cisplatin and carboplatin each in combination with paclitaxel as front-line therapy in advanced epithelial ovarian cancer. PATIENTS AND METHODS: Patients were randomly allocated to receive paclitaxel 175 mg/m2 intravenously as a 3-hour infusion followed by either cisplatin 75 mg/m2 or carboplatin (area under the plasma concentration-time curve of 5), both on day 1. The schedule was repeated every 3 weeks for at least six cycles. Women allocated to paclitaxel-cisplatin were admitted to the hospital, whereas the carboplatin regimen was administered to outpatients. RESULTS: A total of 208 eligible patients were randomized. Both regimens could be delivered in an optimal dose and without significant delay. Paclitaxel-carboplatin produced significantly less nausea and vomiting (P < .01) and less peripheral neurotoxicity (P = .04) but more granulocytopenia and thrombocytopenia (P < .01). The overall response rate in 132 patients with measurable disease was 64% (84 of 132 patients), and in patients with elevated CA 125 levels at start, it was 74% (132 of 178 patients). With a median follow-up time of 37 months, the median progression-free survival time of all patients was 16 months and the median overall survival time was 31 months. The small number of patients entered onto the study caused wide confidence intervals (CIs) around the hazards ratio for progression-free survival of paclitaxel-carboplatin compared with paclitaxel-cisplatin (hazards ratio, 1.07; 95% CI, 0.78 to 1.48) and did not allow conclusions about efficacy. CONCLUSION: Paclitaxel-carboplatin is a feasible regimen for outpatients with ovarian cancer and has a better toxicity profile than paclitaxel-cisplatin.
FOR MANY YEARS, the combination of cyclophosphamide-cisplatin was used as the standard of care for the treatment of ovarian cancer; then, however, a large randomized study proved that a combination of paclitaxel with cisplatin resulted in a better response rate and an improvement in progression-free interval and survival.1 Because of the survival advantage, paclitaxel-cisplatin was recommended as the new standard of care.2 In this schedule, paclitaxel was infused over a time period of 24 hours and administration of the regimen required admission to the hospital for at least 48 hours. For this reason, other dose-time schedules were investigated.3 In this study, a higher, compared with that of the previous study, dose of paclitaxel (175 mg/m2) was administered in 3 hours with an equal dose of cisplatin (75 mg/m2). The shift in dose and administration time is likely to result in a different toxicity profile. Indeed the first reports of the use of this schedule reported a high incidence of neurotoxicity.4 The substitution of the analog carboplatin for cisplatin in this combination may improve the toxicity profile, and admission to the hospital could be avoided. Carboplatin is less nephro-, oto-, and neurotoxic than the parent compound cisplatin but is more myelotoxic when used as a single drug. Unfortunately, the analog was less effective compared with the parent drug cisplatin in a number of solid tumors.5 Also, for the initial treatment of ovarian cancer with the aim to cure and achieve long-term survival, carboplatin was not uniformly recommended.5,6 So, before the combination paclitaxel-carboplatin could be recommended in previously untreated patients, we needed answers to questions such as the following: for how many cycles is the schedule safe and feasible when administered to outpatients, is neurotoxicity indeed less compared with the cisplatin-based schedule, and, finally, if both regimens are safe and feasible, is carboplatin equally effective as cisplatin in combination with paclitaxel? In 1994, we decided to investigate cisplatin and carboplatin, each with paclitaxel, in an exploratory phase III trial. Our main objective was to explore the side effects and feasibility of both regimens to justify the costs for a larger randomized study.
Patients Eligible patients were required to have histologically verified epithelial ovarian carcinoma (based on the 1973 World Health Organization [WHO] histologic classification), International Federation of Gynecology and Obstetrics (FIGO) stages IIB to IV. Exclusion criteria included the following: WHO performance status of 4, age greater than 75 years or less than 18 years, complete bowel obstruction or the presence of symptomatic brain metastases, previous treatment with chemo- or radiotherapy, inadequate bone marrow function (WBC count < 3.0 x 109/L or platelet count < 100 x 109/L), inadequate renal function (serum creatinine level > 120 µmol/L or creatinine clearance < 60 mL/min/1.73 m2), and inadequate liver function (bilirubin level > 25 µmol/L). Patients were excluded if they had a history of ventricular arrhythmia, congestive heart failure, or documented myocardial infarction within the 6 months preceding randomization. Patients with borderline tumors or abdominal adenocarcinoma of unknown origin were also excluded, as were patients with a second malignant disease, expected inadequacy of follow-up, an active infection, or other serious medical conditions that would impair the ability of the patient to receive protocol treatment, including prior allergic reactions to drugs that contained Cremophor EL.
Ethics Review
Toxicity and Response Criteria
Treatment Dose reductions were performed according to nadir and nadir duration.11 Patients received at least six cycles of protocol treatment unless they developed progressive disease or unacceptable toxicity. In patients with assessable disease and no change in disease status after six cycles, treatment was continued for another two courses, but subsequent treatment in these patients was left to the discretion of the investigator. Patients who achieved a partial response continued treatment until progression or unacceptable toxicity occurred. If complete response was achieved, the patient received another three cycles of therapy after the date of documented response. In patients with nonassessable disease, the study treatment was also given for six cycles, unless progressive disease or unacceptable toxicity occurred. Subsequent treatment was at the discretion of the investigator.
Surgery
Investigations
Statistical Methods
Sample size and power calculations were based on the comparison of the progression-free survival in both arms. Under the paclitaxel-cisplatin regimen, the median progression-free survival time was approximately 20 months. Our study was designed to obtain a significant result ( All randomized patients were included in the analysis of survival and progression-free survival, which were performed strictly according to the intention-to-treat principle. Kaplan-Meier survival curves were drawn with SPSS for Windows, version 8.0 (SPSS, Inc, Chicago, IL), and P values were calculated with the log-rank and Breslow tests for significance.12 Progression-free survival was measured from the day of randomization to the time of first relapse or to last follow-up. We analyzed the neurotoxicity-free period by means of the Kaplan-Meier method with censoring if the treatment was stopped before reaching neurotoxicity. This method corrects for stopping for other reasons than neurotoxicity and is generally used in the setting of competing risk analysis.
The Cox proportional hazards regression model was used for univariate and multivariate analysis.13 Factors tested in univariate analyses for progression-free survival included treatment, age, WHO performance status, residual tumor size (> or
Randomization and Eligibility In the accrual period from March 1994 to March 1997, 213 patients were randomized. Five patients were considered ineligible, one because of borderline histology, one because of serious illness, and three because they were randomized by a center not allowed to participate in the study. At the time of analyses, the median follow-up time was 37 months (75th percentile, 11 months; 25th percentile, 42 months). One patient randomized to paclitaxel-carboplatin was lost to follow-up.
Patient Characteristics
Chemotherapy Delivery One patient did not receive any chemotherapy. More cycles were administered in the patients treated with the carboplatin-based schedule (701 v 666; P = .003) (Table 2). Slightly more delay between cycles was seen in the carboplatin-treated patients. The actual mean doses of drug delivered as percentage of the dose planned were, for the paclitaxel-cisplatin and paclitaxel-carboplatin regimens, respectively, 99% for paclitaxel, 96% for cisplatin, 98% for paclitaxel, and 103% for carboplatin (using the Cockcroft formula to calculate the planned dose in milligrams). The actual mean doses of the drugs delivered by treatment cycle are shown in Figs 1 and 2.
Toxicity The major toxicities encountered are summarized in Table 3. No differences in toxicity distribution were found for the following: hair loss, fever, mucositis, diarrhea, allergic reactions, pulmonary complications, cutaneous complications, cardiac events, arthralgia, myalgia, constipation, or renal side effects. Neurotoxicity (> WHO grade 1) occurred earlier in patients randomized to the paclitaxel-cisplatin regimen (Fig 3). Six patients on paclitaxel-cisplatin versus one on paclitaxel-carboplatin refused further treatment before the protocol treatment was completed. Eighteen patients (17%) discontinued paclitaxel-cisplatin because of toxicity, compared with five (5%) in the paclitaxel-carboplatin arm. In both patient groups, 7% of the patients were hospitalized for other reasons than the administration of chemotherapy.
Response and Survival Overall response data in patients with measurable disease are summarized in Table 4. In 178 patients, the CA 125 level was greater than 40 U/mL at the start of treatment. These patients were assessable for determination of CA 125 response. A remission measured by CA 125 level was found in 68 (76%) of 89 patients allocated to the cisplatin-based schedule and in 64 (72%) of 89 patients allocated to carboplatin-paclitaxel. The overall CA 125 response rate was 74% (132 of 178). In 86 (73%) of 118 patients, the CA 125 level increased before other signs of clinical progression were diagnosed. This increase in CA 125 level occurred a median of 84 days before progressive disease was measured with CT or ultrasound scan. The time between randomization and the first clinical sign of progressive disease (other than an increase of CA 125 level) was 16 months for both treatments (Fig 4). Median overall survival times were 30 and 32 months for patients allocated to paclitaxel-cisplatin and paclitaxel-carboplatin, respectively.
The hazards ratios for carboplatin compared with cisplatin were as follows: 1.07 for progression-free survival (95% confidence interval [CI], 0.78 to 1.48), 0.85 for overall survival (95% CI, 0.59 to 1.24), and 1.27 for CA 125 progression-free survival (95% CI, 0.90 to 1.80). Univariate analysis of factors that might affect progression-free survival showed that patients with a residual tumor of less than 1 cm before chemotherapy, FIGO stage II or III, platelet values below the median value (405 x 109/L), or a high hemoglobin level (> 7.5 mmol/L) had longer progression-free survival (Table 5). In the multivariate model, the tumor diameter before chemotherapy (P < .01) and stage (P < .05) were the only factors with a significant impact on progression-free survival. The treatment effect remained unchanged when a multivariate model with treatment and residual tumor as a prognostic factor was considered. After adjustment for tumor diameter and stage, the hazards ratio for carboplatin compared with cisplatin on progression-free survival was similar: 1.03 (95% CI, 0.75 to 1.42).
The results of our study show that the paclitaxel-carboplatin regimen used is safe and easy to administer in outpatients. The regimen is overall less toxic than the cisplatin-based combination. Paclitaxel, carboplatin, and cisplatin were able to be delivered in an adequate dose. When the dose of paclitaxel delivered as part of the paclitaxel-carboplatin regimen is compared with that delivered with the paclitaxel-cisplatin regimen (Figs 1 and 2), it is obvious that during the first six cycles the dose delivered of paclitaxel was similar for the two schedules. But after six cycles, the dose of paclitaxel in paclitaxel-carboplatin was decreased to less than 170 mg/m2, most likely as a result of myelotoxicity in combination with carboplatin. In combination with cisplatin, this dose adjustment was not seen in later cycles. The dose of carboplatin used in our study is accepted by many as the optimal dose for carboplatin in a combination regimen. The dose of AUC 5 allows prolonged treatment beyond six cycles, and proof is lacking that a higher dose is more efficacious. In comparative studies, no difference was found in efficacy between carboplatin AUC 4 and 8, both in combination with cyclophosphamide, and between carboplatin as a single drug dosed AUC 6 or 12.14,15 A large study by the International Collaborative Ovarian Neoplasm group reported no significant differences between carboplatin AUC 5 every 3 weeks and a combination of cisplatin, doxorubicin, and cyclophosphamide.16 In our study, treatment duration was defined differently from what is common in other studies (see Patients and Methods). This strategy resulted in a number of patients who received more than six cycles of treatment. More patients randomized to the carboplatin combination received prolonged treatment, and fewer patients stopped prematurely compared with the women allocated to the cisplatin treatment. This is most likely a result of the better tolerability of the paclitaxel-carboplatin treatment. It is unsure whether this prolonged treatment in the paclitaxel-carboplatin group did have an impact on progression-free survival. Proof is lacking in the literature that longer treatment is better. The worst hematologic toxicities encountered were grade 3 or 4 granulocytopenia, and this occurred more frequently in women treated with the paclitaxel-carboplatin regimen. Earlier phase II studies suggested that the paclitaxel-carboplatin combination causes less thrombocytopenia, as expected.17 Our data do not confirm this finding, but thrombocytopenia was not a problem, even for patients who received treatment beyond six cycles. Except for alopecia, which occurred in almost all patients, nonhematologic toxicities such as neurotoxicity were reported more frequently in the cisplatin arm. Peripheral neurotoxicity remained a frequent side effect in the carboplatin-based regimen, but it occurred later on and was less frequent and less severe compared with the neurotoxicity caused by cisplatin-paclitaxel. The finding that neurotoxicity occurs later is especially important for patients who experience early relapse and need neurotoxic second-line treatment. It was encouraged in our study to measure CA 125 level before each cycle of chemotherapy to gain experience with this marker as an indicator of early progressive disease during treatment and as a marker useful to determine response. Indeed, in the marker increase was a first sign of progressive disease in 70% of the patients with an elevated level at entry. The overall response rate determined with a CA 125 level of 74% was as expected. Univariate analysis of our data revealed the well-known prognostic factors of residual disease and stage to be predictors for progression-free survival but also showed low hemoglobin levels, high platelet counts, and a high number of granulocytes to be negative predictive factors before treatment. Interleukin-6 (IL-6), a multifunctional cytokine, may be responsible for this phenomenon. This cytokine, which has a diversity of functions, leads to the induction of C-reactive protein, increased platelet counts, a high granulocyte count, and low hemoglobin levels. Different epithelial ovarian cancer cell lines are found to produce varying amounts of IL-6, and a possible relationship between IL-6 levels in malignant ovarian tumors and these pretreatment laboratory values is suggested.18 The larger the tumor mass, the more likely growth factors are released from the tumor, and this may explain why the prognostic effect disappears in the multivariate analysis that includes tumor mass. Our study is the first randomized study to compare paclitaxel-carboplatin with paclitaxel-cisplatin. Unfortunately, we cannot draw any conclusions about the efficacy of the two regimens in comparison with each other. Taking into account the broad CIs of the hazards ratios, the results are as expected in this patient population. At least three larger studies have followed. These studies have more power to answer the main question: is the carboplatin combination as effective as the cisplatin combination? A German phase III trial was initiated by the Arbeitsgemeinschaft Gynakologische Onkologie Ovarian Cancer Study Group and completed accrual in 1997. The regimens used were similar to those of the present study, although the carboplatin dose was AUC 6. In total, 800 patients were enrolled within 2 years. The Gynecologic Oncology Group initiated another large study. Again, similar regimens were used, but this time the carboplatin was dosed AUC 7.5. The International Collaborative Ovarian Neoplasm group collaborators continued their efforts to determine the best standard treatment for ovarian cancer and randomized the carboplatin schedule, as used in our study, against a three-drug combination of cyclophosphamide, doxorubicin, and cisplatin or single-agent carboplatin. At present, the data of these studies are not fully matured, and so far, survival differences cannot not be precluded. In summary, the results of our study show a favorable toxicity profile for paclitaxel-carboplatin. The regimen is so well tolerated that its widespread use at this point in time is recommended by several authorities in the field.19 However, a word of caution is needed because the follow-up of most carboplatin-paclitaxel studies is short and differences in long-term survival could become apparent in the future.
Supported by grants from Bristol-Meyers Squibb, København, Denmark, and Bristol-Meyers Squibb, Woerden, the Netherlands. We acknowledge Benedicte Palle, Ole Roer, Lene Stigaard, Cinthy Dekker, and Annelies Epping for assistance with data management and the significant contribution of the following investigators to this work: Steen W. Hansen, Petronella O. Witteveen, Louis H. Siegenbeek van Heukelom, and Ed Maartense.
Presented in part at the Thirty-Third Annual Meeting of the American Society of Clinical Oncology, Denver, CO, May 17-20, 1997.
1. McGuire WP, Hoskins WJ, Brady MF, et al: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334: 1-6, 1996
2.
Neijt JP: New therapy for ovarian cancer. N Engl J Med 334: 50-51, 1996 (editorial)
3.
Eisenhauer EA, ten Bokkel Huinink WM, Swenerton KD, et al: European-Canadian randomized trial of paclitaxel in relapsed ovarian cancer: High-dose versus low-dose and long versus short infusion. J Clin Oncol 12: 2654-2666, 1994 4. Connelly E, Markman M, Kennedy A, et al: Paclitaxel delivered as a 3-hr infusion with cisplatin in patients with gynecologic cancers: Unexpected incidence of neurotoxicity. Gynecol Oncol 62: 166-168, 1996[Medline]
5.
Lokich J, Anderson N: Carboplatin versus cisplatin in solid tumors: An analysis of the literature. Ann Oncol 9: 13-21, 1998 6. Vermorken JB, ten Bokkel Huinink WW, Eisenhauer EA, et al: Carboplatin versus cisplatin. Ann Oncol 4: S41-S48, 1993 (suppl 4) 7. Miller AB, Hoogstraten B, Staquet M, et al: Reporting results of cancer treatment. Cancer 47: 207-214, 1981[Medline] 8. Rustin GJS, Nelstrop AE, Bentzen SM, et al: Use of tumour markers in monitoring the course of ovarian cancer. Ann Oncol 10: S25-S32, 1999 (suppl 1) 9. Calvert AH, Newell DR, Gumbrell LA: Carboplatin dosage: Prospective evaluation of a simple formula based on renal function. J Clin Oncol 17: 1748-1756, 1989 10. Cockcroft DW, Gault MH: Prediction of creatinine clearance from serum creatinine. Nephron 16: 31-41, 1976[Medline] 11. Neijt JP, Engelholm SA, Witteveen PO, et al: Paclitaxel (175 mg/m2 over 3 hours) with cisplatin or carboplatin in previously untreated ovarian cancer: An interim analysis. Semin Oncol 24: S36-S39, 1997 (suppl 15) 12. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958 13. Cox DR: Regression models in life-tables. J R Stat Soc 34: 187-220, 1972 14. Gore M, Mainwaring P, AHern R, et al: Randomized trial of dose-intensity with single-agent carboplatin in patients with epithelial ovarian cancer: London Gynaecological Oncology Group. J Clin Oncol 16: 2426-2434, 1998[Abstract]
15.
Jakobsen A, Bertelsen K, Andersen J, et al: Dose-effect study of carboplatin in ovarian cancer: A Danish Ovarian Cancer Group study. J Clin Oncol 15: 193-198, 1997 16. The ICON Collaborators: ICON2: Randomised trial of single-agent carboplatin against three-drug combination of CAP (cyclophosphamide, doxorubicin, and cisplatin) in women with ovarian cancer. Lancet 352: 1571-1576, 1998[Medline] 17. Calvert AH: A review of the pharmacokinetics and pharmacodynamics of combination carboplatin/paclitaxel. Semin Oncol 24: S85-S90, 1997 (suppl 2) 18. Van der Zee AG, de Cuyper EM, Limburg PC, et al: Higher levels of interleukin-6 in cystic fluids from patients with malignant versus benign ovarian tumors correlate with decreased hemoglobin levels and increased platelet counts. Cancer 75: 1004-1009, 1995[Medline] 19. Berek JS, Bertelsen K, du Bois A, et al: Advanced epithelial ovarian cancer: 1998 consensus statements. Ann Oncol 10: S91-S96, 1999 (suppl 1) Submitted October 13, 1999; accepted May 1, 2000.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||
|
Copyright © 2000 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|