Dacarbazine, Cisplatin, and Interferon-Alfa-2b With or Without Interleukin-2 in Metastatic Melanoma: A Randomized Phase III Trial (18951) of
http://www.100md.com
《临床肿瘤学》
the Department of Medicine III, Charité, Campus Benjamin Franklin, Berlin
Department of Dermatology, University of Ulm, Ulm
Department of Dermatology, University of Heidelberg, Heidelberg
Haematologisch-Onkologische Praxis Altona, Hamburg, Germany
Department of Medical Oncology, University Medical Center, Nijmegen
Daniel den Hoed Cancer Center, University of Rotterdam, Rotterdam, Netherlands
Royal Marsden Hospital, London
Cancer Research UK Clinical Center, St James’s University Hospital, Leeds, United Kingdom
Centre Pluridisciplinaire d’Oncologie, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland
Department of Oncology, University Hospital Gasthuisberg, Leuven
European Organisation for Research and Treatment of Cancer Data Center
Department d’Oncologie, Hospital Universitaire Erasme, Brussels, Belgium
Department of Medical Oncology, Centre Leon Berard, Lyon, France
ABSTRACT
BACKGROUND: Based on phase II trial results, chemoimmunotherapy combinations have become the preferred treatment for patients with metastatic melanoma in many institutions. This study was performed to determine whether interleukin-2 (IL-2) as a component of chemoimmunotherapy influences survival of patients with metastatic melanoma.
PATIENTS AND METHODS: Patients with advanced metastatic melanoma were randomly assigned to receive dacarbazine 250 mg/m2 and cisplatin 30 mg/m2 on days 1 to 3 combined with interferon-alfa-2b 10 x 106 U/m2 subcutaneously on days 1 through 5 without (arm A) or with (arm B) a high-dose intravenous decrescendo regimen of IL-2 on days 5 through 10 (18 x 106 U/m2/6 hours, 18 x 106 U/m2/12 hours, 18 x 106 U/m2/24 hours, and 4.5 x 106 U/m2 for 3 x 24 hours). Treatment cycles were repeated in the absence of disease progression every 28 days to a maximum of four cycles.
RESULTS: Three hundred sixty-three patients with advanced metastatic melanoma were accrued. The median survival was 9 months in both arms, with a 2-year survival rate of 12.9% and 17.6% in arms A and B, respectively (P = .32; hazard ratio, 0.90; 95% CI, 0.72 to 1.11). There was also no statistically significant difference regarding progression-free survival (median, 3.0 v 3.9 months) and response rate (22.8% v 20.8%).
CONCLUSION: Despite its activity in melanoma as a single agent or in combination with interferon-alfa-2b, the chosen schedule of IL-2 added to the chemoimmunotherapy combination had no clinically relevant activity.
INTRODUCTION
The activity of interleukin-2 (IL-2) in the treatment of metastatic melanoma has been tested extensively during the past decade. IL-2 administered intravenously (IV) in high doses results in up to 20% of patients obtaining an objective tumor response,1-7 including some patients with apparently durable complete responses. Interferon-alfa (IFN-) is also an active agent in metastatic melanoma, and phase II results have suggested that when it is combined with IL-2, response rates (RRs) of 20% to 40% can be obtained.8-13 The combination of IFN- and IL-2, using a decrescendo regimen of IL-2, lead to an RR of 41% in a phase II trial,13 which was among the best results achieved with the combination of these two drugs in various schedules. In subsequent studies of combination therapy, the addition of a number of different cytotoxic agents with IFN- and various regimens and doses of IL-2 were investigated. RRs of up to 56% were achieved in the initial phase II studies with these combinations, which sometimes incorporated as many as six different agents.14-25 RRs exceeding 50% were seen often in phase II studies incorporating cisplatin, IFN-, and high-dose IL-2 in the treatment regimen21-25; on the basis of this observation, these three drugs became the preferred chemoimmunotherapy regimen in many institutions for patients with metastatic melanoma.
Combination regimens comprising intermediate doses of IFN- given with IL-2 have been investigated in randomized trials. In the first such randomized study completed by the European Organisation for Research and Treatment of Cancer (EORTC) Melanoma Group, the IL-2 was given in serial reducing doses ("decrescendo" regimen), a regimen that had shown promising activity with manageable toxicity.26 Patients were treated with IFN- and a decrescendo regimen of IL-2 with or without cisplatin; the addition of cisplatin resulted in a doubling of the RR and time to progression. There was no overall survival benefit, although a significant proportion of durable complete remissions (CRs) and partial remissions (PRs) were observed.27 Given the fact that IL-2 has the most promising activity when given as a single agent and is also the most toxic agent in melanoma when given at high dose, we asked the question in the EORTC Melanoma Group’s second multicenter randomized trial (EORTC trial number 18,951) of whether the addition of IL-2 to dacarbazine, cisplatin, and IFN- would have an impact on survival. This is one of the largest randomized trials to date that involves patients with metastatic melanoma.
PATIENTS AND METHODS
Patients were eligible for this trial if they had metastatic melanoma with measurable disease that could not be controlled by surgery and a Karnofsky performance status (PS) of at least 60%. Exclusion criteria were the presence of brain metastases on a brain computed tomography scan or magnetic resonance imaging; prior therapy with IL-2 or two other components of the regimen; symptomatic cardiac, pulmonary, renal, liver, or thyroid disease; autoimmune diseases; corticosteroid treatment; and significant bone marrow dysfunction. The protocol was approved by the EORTC Protocol Review Committee and the institutional review committees of all participating hospitals, and informed consent was obtained from all patients before being randomly assigned. Only centers with previous experience in the administration of high-dose IL-2 participated in this trial.
Treatment Plan
All patients received dacarbazine, 250 mg/m2/d IV on days 1 to 3, cisplatin 30 mg/m2/d IV on days 1 to 3, and IFN--2b (Schering-Plough, Kenilworth, NJ) 10 x 106 U/m2/d subcutaneously on days 1 to 5. Patients randomly assigned to arm B received IL-2 (Proleukin; Chiron, Amsterdam, Netherlands) in addition as a continuous IV infusion in a decrescendo schedule starting on day 5 with 18 million U/m2 over 6 hours followed by 18 million U/m2 over 12 hours, 18 million U/m2 over 24 hours, and a maintenance dose of 4.5 million U/m2/24 hours for an additional 72 hours. This schedule resulted in the administration of a total dose of 67.5 million U/m2 of IL-2 over 114 hours. Cycles were repeated every 4 weeks for a maximum of four cycles; treatment was stopped at any time if there was evidence of disease progression or in the event of unacceptable toxicity. Patients with stable disease after two cycles of treatment received two additional courses of treatment.
Recommended concomitant medications included prophylactic 5HT3 antagonists before dacarbazine and cisplatin, and paracetamol or indomethacin to ameliorate cytokine-related fever. Low-dose furosemide or renal-dose dopamine was recommended for IL-2-related oliguria, and IV saline and colloids were recommended for hypotension.
The development of any grade 4 toxicity resulted in protocol treatment being discontinued, except in the case of grade 4 hematologic toxicities, provided they had resolved by the time the next cycle was due. The IL-2 infusion was interrupted if the following grade 3 toxicities occurred during treatment: hypotension not responding to concomitant therapy, cardiac arrhythmia, suspicion of myocardial ischemia, agitation or persistent confusion, elevation of bilirubin (> 60 μmol/L), bacterial sepsis, or dyspnea at rest. In these circumstances, toxicity was reassessed every 2 hours until it resolved to grade 1 or less and restarted at 50% of the original dose. Dose reductions were not carried over to subsequent treatment cycles. Only if there had been a transient rise in serum creatinine exceeding 265 μmol/L or grade 4 neurotoxicity during the previous cycle was the dose of IL-2 reduced by 50% for the next cycle.
Response and Toxicity Assessments
Tumor response was assessed with appropriate imaging procedures after the second and fourth treatment cycles according to standard WHO criteria.26 Patients achieving a PR or CR were in some analyses combined as "responding" patients. Toxicity was evaluated by using standard WHO criteria and assessed before the next cycle of treatment. The toxicity of the last cycle was assessed after 4 weeks. After the completion of study treatment, patients were followed every 2 months for the first 6 months and then once every 3 months.
Statistical Design
Randomization was performed centrally (EORTC Data Center), with stratification for center and serum lactate dehydrogenase (LDH) level (three groups: < upper limit of normal [ULN], between ULN and 2x ULN, and 2x ULN) by using the minimization technique.
The primary end point was the duration of survival after being randomly assigned to a treatment arm. This was defined as time from assignment until death, whatever the cause; patients still alive were censored at their latest date of follow-up. Secondary end points were RR (best response observed during the study: CR or PR), progression-free survival (PFS; time from assignment until progression or death, whatever the cause; patients still alive without progression were censored at their latest date of follow-up), and relapse-free survival (RFS; for patients who reached CR or PR, time from PR or CR until relapse or death, whatever the cause; patients still alive without relapse were censored at their latest date of follow-up).
The aim of the study was to detect a difference between the two treatment arms in terms of overall survival rate at 2 years. The goal was set initially to enter 156 patients to detect a 15% difference (10% v 25%) in the 2-year survival rates between the two arms (two-tailed test: = 5%; ? = 20%). Because the accrual of patients was achieved quickly and almost 100 patients were accrued, it was decided to increase the study population up to a minimum of 338 eligible patients to allow detection of a smaller difference of 10%, which was the minimum felt to be medically relevant. A total of 286 patients followed until death would allow the detection with an 85% statistical power ( = 5%) of a 10% difference in the 2-year survival rates (10% v 20%), corresponding to a hazard ratio (HR) of 0.70 in case of an exponential distribution.27 To make up for potential noneligible patients and patients lost to follow-up, another 7.5% of patients were added to a target accrual goal of 363 patients.
The actuarial curves were computed by using the Kaplan-Meier technique and the SEs of the estimates were obtained by using the Greenwood formula.28 The difference between curves was tested for statistical significance by using the two-tailed log-rank test.28 The Cox proportional hazards model was used to obtain the estimate and the 95% CI of the HR of the instantaneous event rate in the experimental group versus the one in the control group, adjusting by possible confounding factors; the Wald test was used to determine the prognostic importance of each variable included in the model.29 Prognostic interaction between variables was tested by including products of variables into the model. All analyses were performed according to the intent-to-treat-principle. The 2 test was used to compare the RRs of treatment arms; adjustments by other factors were performed via the logistic model.
The database was frozen on September, 2002. SAS 8.1 software (SAS Institute, Cary, NC) was used for the statistical analyses.
RESULTS
Patient Characteristics and Study Flow
Between May 1995 and April 2000, 363 patients from 25 centers were accrued and randomly assigned. Their pretreatment characteristics are listed in Table 1 . This patient population was largely a group with advanced metastatic melanoma (large majority with M1c disease) who had good PS (> 80% of patients had a Karnofsky index of 90% and 100%). Patient characteristics were well balanced between treatment arms. The study-flow summary is listed in Table 2. In essence, 3% of the patients did not start the allocated treatment, usually because of rapid disease progression or withdrawn consent. Eleven patients (3%) were found to be ineligible. However, all patients were included in the statistical evaluations, because all analyses were based strictly on an intent-to-treat analysis.
Adherence to Treatment Plan
The treatment was administered as planned in 89% and 72% of patients in arms A and B, respectively. Details of treatment administration are summarized in Table 3. The most common reason for treatment termination was progressive disease, occurring as a reason for discontinuation among 50% of patients in arm A and 36% of patients in arm B. Toxicity of treatment or patient refusal were the next most frequent reasons for treatment discontinuation, which were reported for 6.7% of patients in arm A and 21.8% of patients in arm B. Chemotherapy or IFN was rarely dose-reduced in cycle one, whereas the IL-2 dose was reduced during cycle one in 24% of patients who received it (arm B). Overall, chemotherapy and IFN were usually administered at the full planned dose, but the IL-2 dose was reduced or interrupted in 15.5% and 33.9% of patients, respectively.
Toxicities
As expected, treatment in arm B was associated with a higher incidence of grade 3 to 4 hypotension, fever without infection, lethargy, anorexia, and diarrhea (Table 4). The nadirs of the total white blood cell and neutrophil counts were lower in arm A, whereas the platelet nadirs were lower in arm B.
Survival
The median follow-up was 3.38 years at the time of the final analysis, by which time a total of 328 deaths had occurred. The median survival was 9.0 months in each arm. The treatment difference was not significant (P = .31; Fig 1). Stratification by the initial LDH (three categories) indicated at randomization yielded P = .31, the estimated HR (arm B v A) was 0.89, and the 95% CI was 0.72 to 1.11. The 2-year survival-rate estimates were 12.9% (SE, 2.5%) and 17.6% (SE, 2.8%) in arms A and B, respectively. An apparent separation of the survival curves after 2 years was based on too few patients at risk to contribute to overall significance. Nonetheless, the number of patients surviving at the time of study lock was greater in arm B than in arm A (24 v 11 patients; Table 5).
Secondary analyses were performed to investigate the contribution of established major prognostic factors. Univariate analyses (not shown) considered LDH, American Joint Committee on Cancer (AJCC) M stage, and PS. AJCC M stage did not reveal any differences, but few patients were M1a, and the majority were M1c. LDH and PS revealed significant differences. As for LDH, only two of the initial three categories lead to distinct survival curves, resulting in a median survival of 5.1 months with the LDH above 2x ULN and 10.1 months with LDH below 2x ULN. Also, PS was of predictive value, with the distinction already occurring between a Karnofsky index of 100% (median survival, 11.4 months) versus 90% (median survival, 7.6 months) and only few changes below that level. The Cox proportional hazards model considering LDH and PS showed that the initial LDH (two categories: 2x ULN v < 2x ULN; HR, 2.43; P < .0001) and PS (two categories: < 100% v 100%: HR, 1.53; P = .0002) seemed to be important independent prognostic factors (Fig 2). The estimated HR (arm B v A) adjusted for these factors was 0.895 (95% CI, 0.72 to 1.11; P = .32). In the Cox model, the interaction between treatment and a score based on LDH and PS was not significant (P = .16).
PFS
The analysis for PFS included all 363 patients, of whom 359 were followed until progression or death. No significant (P = .28) difference was detected between the two arms (Fig 3). The median PFS was 3.0 months (arm A) versus 3.96 months (arm B), and the estimated 2-year PFS rate was 3.89% (SE, 1.44%) versus 4.50% (SE, 1.50%). The estimated HR adjusted for LDH (three categories) at randomization was 0.87 (95% CI, 0.70 to 1.07; P = .19). In a Cox model, LDH (< 2x ULN v > 2x ULN; P = .002) and PS (< 100% v 90%; P = .04) seemed to be of prognostic importance, whereas treatment remained not significant (P = .19; HR, 0.87; 95% CI, 0.70 to 1.07).
RR
The overall objective RR (CR + PR) was similar in the two arms: 22.8% (SE, 3.1%) versus 20.8% (SE, 3.0%) in arms A and B, respectively (Table 5). The observed difference (2%, with a 95% confidence limit of –6.5%, 10.5%) was not significant (P = .74). The logistic model indicated that the initial LDH (< 2x ULN v > 2x ULN; P = .66) and the PS (< 100% v 100%; P = .15) had no prognostic importance, and the treatment impact was not significant (P = .65; odds ratio, 0.89; 95% CI, 0.54 to 1.47).
RFS
The RFS analysis is based on 79 patients who achieved a PR or CR, of whom 70 relapsed or died during follow-up (Fig 4). There was no significant difference in RFS between the two treatment arms (P = .89; HR, 0.97; 95% CI, 0.60 to 1.52). A Cox model confirmed these results: the initial PS was of no prognostic importance (HR, 1.24; P = .2), whereas the initial LDH was borderline significant (HR, 2.16; P = .06). The type of response had prognostic importance (P = .002; Fig 5): those who reached a CR (n = 13) had a longer RFS than those who reached a PR (n = 66). However, there was no impact of treatment on RFS in CR patients nor in PR patients considered separately. The time to response was quite similar between the treatment arms, because most PRs were first noted after two treatment cycles, and most CRs were first noted after four cycles.
DISCUSSION
This study shows that IL-2 added to a complex regimen given for a maximum of four cycles and consisting of dacarbazine, cisplatin, and IFN- does not confer a clinically meaningful survival benefit for most patients with advanced melanoma, nor does it confer an increase in RR or time to progression. These results are in line with several smaller previous randomized phase II and III trials that evaluated the effect of adding intermediate-dose IL-2 to dacarbazine/IFN-,30 low-dose subcutaneous IL-2 and IFN- to dacarbazine/cisplatin/carmustine/tamoxifen,31 low-dose subcutaneous IL-2 and IFN- to dacarbazine/cisplatin/carmustine,32 intermediate-dose IL-2 and IFN- to dacarbazine/cisplatin/carmustine/tamoxifen,33 and high-dose IL-2 and IFN- to dacarbazine/cisplatin/tamoxifen.34
The only randomized trial with borderline significance for overall survival improvement (P = .06 on two-sided log-rank test) is a single-institution study35 that investigated the effect of adding IL-2 and IFN- to a polychemotherapy regimen consisting of cisplatin, vinblastine, and dacarbazine. However, recent data suggest that the trend for a survival benefit observed in this trial cannot be confirmed in a larger, multi-institutional setting using a modified regimen and dose-adjustment schema given for a maximum of four cycles (M.B. Atkins, personal communication, May 2003).
Most previous trials30-33 have been criticized because they have investigated IL-2 regimens that were never tested for their capacity to induce remission in phase II studies. However, the decrescendo regimen used in the trial reported here was active in prior phase II studies, and its activity was confirmed in a previous phase III trial.26 Despite these promising early data, in the trial reported here, we have not been able to show that IL-2 adds to the efficacy of the combination of dacarbazine, cisplatin, and IFN-.
There are several possible explanations for the failure of this trial to detect improved efficacy. First, it is conceivable that melanoma has a subset that does not respond to either chemotherapy or cytokine treatment. Second, combination chemotherapy could impair the mechanisms by which IL-2 induces long-term remissions. Our finding that RRs between this IL-2-based biochemotherapy regimen and non–IL-2-based therapy are the same would be entirely consistent with both of these hypotheses. Alternatively, the dose intensity of IL-2 achieved in this multicenter trial may have been below an as-yet-undefined threshold necessary for efficacy, but because the study was performed only in centers with experience in IV IL-2 regimens and dose reductions were mandated by toxicity and treatment termination was mandated by patient desire in 6%, an increase in dose intensity would be difficult to achieve in a multicenter setting.
Physicians treating patients with metastatic melanoma need to acknowledge the negative results of the study reported here and similar data from smaller trials. The combination of chemotherapy and immunotherapy in chemoimmunotherapy regimens, which have dominated melanoma treatment strategy for the past 10 years, has not proven successful; to date, no such treatment regimen has been shown in a multicenter setting to significantly prolong survival in patients with stage IV melanoma. The efficacy of IL-2 in melanoma is established, but not to an extent that, in the view of the EORTC Melanoma Group, justifies its use in the context of biochemotherapy outside of clinical trials. Future research should focus on enhancement of the immunologic activity of IL-2, as currently pursued by combinations with histamine or vaccines.36,37
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Acknowledgment
We greatly acknowledge the tremendous efforts of the staff in all participating hospitals and the superb work by the melanoma team of the European Organisation for Research and Treatment of Cancer Data Center.
NOTES
Supported by educational grants from Chiron BV (Amsterdam, Netherlands), Schering Plough (Kenilworth, NJ), and the National Cancer Institute (Bethesda, MD; grants 2U10 CA11488-25 through 5U10 CA11488-32).
The article’s contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Sparano J, Fisher R, Sunderland M, et al: Randomized phase III trial of treatment with high-dose interleukin-2 either alone or in combination with interferon-alfa-2a in patients with advanced melanoma. J Clin Oncol 11:1969-1977, 1993
Rosenberg S, Lotze M, Yang J, et al: Experience with the use of high-dose interleukin-2 in the treatment of 652 cancer patients. Ann Surg 210:474-485, 1989
Thatcher N, Dazzi H, Gosh A, et al: Recombinant IL-2 given intra-splenically and intravenously in advanced malignant melanoma: A phase I/II study. Cancer Treat Rev 16:49-52, 1989 (suppl A)
Parkinson D, Abrams J, Wiernik P, et al: Interleukin-2 therapy in patients with metastatic malignant melanoma: A phase II study. J Clin Oncol 8:1650-1656, 1990
Whitehead R, Kopecky K, Samson M, et al: Phase II study of intravenous bolus recombinant interleukin-2 in advanced malignant melanoma: Southwest Oncology Group study. J Natl Cancer Inst 83:1250-1251, 1991
Dorval T, Mathiot C, Brandely M, et al: Lack of effect of tumour infiltrating lymphocytes in patients with metastatic melanoma who failed to respond to interleukin 2. Eur J Cancer 27:99, 1991 (suppl 2)
Legha S, Gianan M, Plager C, et al: Evaluation of interleukin-2 administered by continuous infusion in patients with metastatic melanoma. Cancer 77:89-96, 1996
Rosenberg S, Lotze M, Yang J, et al: Combination therapy with interleukin-2 and alpha-interferon for the treatment of patients with advanced cancer. J Clin Oncol 7:1863-1874, 1989
Kruit W, Goey S, Monson J, et al: Clinical experience with the combined use of recombinant interleukin-2 (IL-2) and interferon-alpha-2a (IFN alpha) in metastatic melanoma. Br J Haematol 79:84-86, 1991 (suppl 1)
Dillman R, Church C, Oldham R, et al: Inpatient continuous-infusion interleukin-2 in 788 patients with cancer. The National Biotherapy Study Group experience. Cancer 71:2358-2370, 1993
Marincola F, White D, Wise A, et al: Combination therapy with interferon alfa-2a and interleukin-2 for the treatment of metastatic cancer. J Clin Oncol 5:1110-1122, 1995
Kruit W, Punt C, Goey S, et al: Dose efficacy study of two schedules of high-dose bolus administration of interleukin-2 and alpha-interferon in patients with metastatic melanoma. Br J Cancer 74:951-955, 1996
Keilholz U, Scheibenbogen C, Tilgen W, et al: Interferon- and interleukin-2 in the treatment of malignant melanoma: A comparison of two phase II trials. Cancer 72:607-614, 1993
Stoter G, Shiloni E, Aamdal S, et al: Sequential administration of recombinant human interleukin-2 and dacarbazine in metastatic melanoma. A multicentre phase II study. Eur J Cancer Clin Oncol 25:41-43, 1989 (suppl 3)
Dillman R, Oldham R, Barth N, et al: Recombinant interleukin-2 and adoptive immunotherapy alternated with dacarbazine therapy in melanoma: a National Biotherapy Study Group trial. J Natl Cancer Inst 82:1345-1349, 1990
Flaherty L, Redman B, Chabot G, et al: A phase I-II study of dacarbazine in combination with outpatient interleukin-2 in metastatic malignant melanoma. Cancer 65:2471-2477, 1990
Demchak P, Mier J, Robert N, et al: Interleukin-2 and high-dose cisplatin in patients with metastatic melanoma: A pilot study. J Clin Oncol 9:1821-1830, 1991
Flaherty L, Robinson W, Redman B, et al: A phase II study of dacarbazine and cisplatin in combination with outpatient administered interleukin-2 in metastatic malignant melanoma. Cancer 71:3520-3525, 1993
Atkins M, O’Boyle K, Sosman J, et al: Multiinstitutional phase II trial of intensive combination chemotherapy for metastatic melanoma. J Clin Oncol 12:1553-1560, 1994
Dummer R, Gore M, Hancock B, et al: A multi-center phase II clinical trial using dacarbazine and continuous infusion of interleukin-2 in metastatic melanoma: Clinical data and immunomonitoring. Cancer 75:945-948, 1995
Richards J, Mehta N, Ramming K, et al: Sequential chemoimmunotherapy in the treatment of metastatic melanoma. J Clin Oncol 10:1338-1343, 1992
Khayat D, Borel C, Tourani J, et al: Sequential chemoimmunotherapy with cisplatin, interleukin-2, and interferon alfa-2a for metastatic melanoma. J Clin Oncol 11:2173-2180, 1993
Legha SS, Ring S, Eton O, et al: Development of a biochemotherapy regimen with concurrent administration of cisplatin, vinblastine, dacarbazine, interferon alfa, and interleukin-2 for patients with metastatic melanoma. J Clin Oncol 16:1752-1759, 1998
Legha SS, Ring S, Bedikian A, et al: Treatment of metastatic melanoma with combined chemotherapy containing cisplatin, vinblastine and dacarbazine (CVD) and biotherapy using interleukin-2 and interferon-alpha. Ann Oncol 7:827-835, 1996
Proebstle TM, Scheibenbogen C, Sterry W, et al: A phase II study of dacarbazine, cisplatin, interferon-alpha, and high dose IL-2 in ‘poor risk’ metastatic melanoma. Eur J Cancer 32:1530-1533, 1996
Keilholz U, Goey S, Punt C, et al: Interferon alfa-2a and interleukin-2 with or without cisplatin in metastatic melanoma: A randomized trial of the European Organization for Research and Treatment of Cancer Melanoma Cooperative Group. J Clin Oncol 15:2579-2588, 1997
Keilholz U, Martus P, Punt CJA, et al: Prognostic factors for survival and factors associated with long-term remission in patients with advanced melanoma receiving cytokine-based treatments: second analysis of a randomised EORTC Melanoma Group trial comparing interferon-alpha2a (IFNalpha) and interleukin 2 (IL-2) with or without cisplatin. Eur J Cancer 38:1501-1511, 2002
Freedman LS: Tables of the numbers of patients required in clinical trials using the logrank test. Stat Med 1:121-129, 1982
Hosmer DW, Lemershow S. Applied Survival Analysis: Regression Models of Time to Event Data. New York, NY, John Wiley & Sons, 1999
Hauschild A, Garbe C, Stolz W, et al: Dacarbazine and interferon alpha with or without interleukin 2 in metastatic melanoma: A randomized phase III multicenter trial of the Dermatologic Oncology Group (DeCOG). Br J Cancer 84:1036-1042, 2001
Atzpodien J, Neuber K, Kamanabrou D, et al: Combination chemotherapy with or without s.c. IL-2 and IFN-alpha: Results of a prospectively randomized trial of the Cooperative Advanced Malignant Melanoma Chemoimmunotherapy Group (ACIMM). Br J Cancer 86:179-184, 2002
Ridolfi R, Chiarion-Seleni V, Guida M, et al: Cisplatin, dacarbazine with or without subcutaneous interleukin-2, and interferon alpha-2b in advanced melanoma outpatients: Results from an Italian multicenter phase III randomized clinical trial. J Clin Oncol 20:1600-1607, 2002
Johnston S, Constenia D, Moore J, et al: Randomized phase II trial of BCDT [carmustine (BCNU), cisplatin, dacarbazine (DTIC) and tamoxifen] with or without interferon alpha (IFN-alpha) and interleukin (IL-2) in patients with metastatic melanoma. Br J Cancer 77:1280-1286, 1998
Rosenberg SA, Yang JC, Schwartzentruber DJ, et al: Prospective randomized trial of the treatment of patients with metastatic melanoma using chemotherapy with cisplatin, dacarbazine, and tamoxifen alone or in combination with interleukin-2 and interferon alfa-2b. J Clin Oncol 17:968-975, 1999
Eton O, Legha S, Bedikian A, et al: Sequential biochemotherapy versus chemotherapy for metastatic melanoma: Results from a phase III randomized trial. J Clin Oncol 20:2045-2052, 2002
Agarwala SA, Glaspy J, O’Day SJ, et al: Results from a randomized phase III study comparing combined treatment with histamine dihidrochloride plus interleukin-2 versus interleukin-2 alone in patients with metastatic melanoma. J Clin Oncol 20:125-133, 2002
Rosenberg SA, Yang JC, Schwartzentruber DJ, et al: Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma. Nat Med 4:321-327, 1998(Ulrich Keilholz, Cornelis)
Department of Dermatology, University of Ulm, Ulm
Department of Dermatology, University of Heidelberg, Heidelberg
Haematologisch-Onkologische Praxis Altona, Hamburg, Germany
Department of Medical Oncology, University Medical Center, Nijmegen
Daniel den Hoed Cancer Center, University of Rotterdam, Rotterdam, Netherlands
Royal Marsden Hospital, London
Cancer Research UK Clinical Center, St James’s University Hospital, Leeds, United Kingdom
Centre Pluridisciplinaire d’Oncologie, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland
Department of Oncology, University Hospital Gasthuisberg, Leuven
European Organisation for Research and Treatment of Cancer Data Center
Department d’Oncologie, Hospital Universitaire Erasme, Brussels, Belgium
Department of Medical Oncology, Centre Leon Berard, Lyon, France
ABSTRACT
BACKGROUND: Based on phase II trial results, chemoimmunotherapy combinations have become the preferred treatment for patients with metastatic melanoma in many institutions. This study was performed to determine whether interleukin-2 (IL-2) as a component of chemoimmunotherapy influences survival of patients with metastatic melanoma.
PATIENTS AND METHODS: Patients with advanced metastatic melanoma were randomly assigned to receive dacarbazine 250 mg/m2 and cisplatin 30 mg/m2 on days 1 to 3 combined with interferon-alfa-2b 10 x 106 U/m2 subcutaneously on days 1 through 5 without (arm A) or with (arm B) a high-dose intravenous decrescendo regimen of IL-2 on days 5 through 10 (18 x 106 U/m2/6 hours, 18 x 106 U/m2/12 hours, 18 x 106 U/m2/24 hours, and 4.5 x 106 U/m2 for 3 x 24 hours). Treatment cycles were repeated in the absence of disease progression every 28 days to a maximum of four cycles.
RESULTS: Three hundred sixty-three patients with advanced metastatic melanoma were accrued. The median survival was 9 months in both arms, with a 2-year survival rate of 12.9% and 17.6% in arms A and B, respectively (P = .32; hazard ratio, 0.90; 95% CI, 0.72 to 1.11). There was also no statistically significant difference regarding progression-free survival (median, 3.0 v 3.9 months) and response rate (22.8% v 20.8%).
CONCLUSION: Despite its activity in melanoma as a single agent or in combination with interferon-alfa-2b, the chosen schedule of IL-2 added to the chemoimmunotherapy combination had no clinically relevant activity.
INTRODUCTION
The activity of interleukin-2 (IL-2) in the treatment of metastatic melanoma has been tested extensively during the past decade. IL-2 administered intravenously (IV) in high doses results in up to 20% of patients obtaining an objective tumor response,1-7 including some patients with apparently durable complete responses. Interferon-alfa (IFN-) is also an active agent in metastatic melanoma, and phase II results have suggested that when it is combined with IL-2, response rates (RRs) of 20% to 40% can be obtained.8-13 The combination of IFN- and IL-2, using a decrescendo regimen of IL-2, lead to an RR of 41% in a phase II trial,13 which was among the best results achieved with the combination of these two drugs in various schedules. In subsequent studies of combination therapy, the addition of a number of different cytotoxic agents with IFN- and various regimens and doses of IL-2 were investigated. RRs of up to 56% were achieved in the initial phase II studies with these combinations, which sometimes incorporated as many as six different agents.14-25 RRs exceeding 50% were seen often in phase II studies incorporating cisplatin, IFN-, and high-dose IL-2 in the treatment regimen21-25; on the basis of this observation, these three drugs became the preferred chemoimmunotherapy regimen in many institutions for patients with metastatic melanoma.
Combination regimens comprising intermediate doses of IFN- given with IL-2 have been investigated in randomized trials. In the first such randomized study completed by the European Organisation for Research and Treatment of Cancer (EORTC) Melanoma Group, the IL-2 was given in serial reducing doses ("decrescendo" regimen), a regimen that had shown promising activity with manageable toxicity.26 Patients were treated with IFN- and a decrescendo regimen of IL-2 with or without cisplatin; the addition of cisplatin resulted in a doubling of the RR and time to progression. There was no overall survival benefit, although a significant proportion of durable complete remissions (CRs) and partial remissions (PRs) were observed.27 Given the fact that IL-2 has the most promising activity when given as a single agent and is also the most toxic agent in melanoma when given at high dose, we asked the question in the EORTC Melanoma Group’s second multicenter randomized trial (EORTC trial number 18,951) of whether the addition of IL-2 to dacarbazine, cisplatin, and IFN- would have an impact on survival. This is one of the largest randomized trials to date that involves patients with metastatic melanoma.
PATIENTS AND METHODS
Patients were eligible for this trial if they had metastatic melanoma with measurable disease that could not be controlled by surgery and a Karnofsky performance status (PS) of at least 60%. Exclusion criteria were the presence of brain metastases on a brain computed tomography scan or magnetic resonance imaging; prior therapy with IL-2 or two other components of the regimen; symptomatic cardiac, pulmonary, renal, liver, or thyroid disease; autoimmune diseases; corticosteroid treatment; and significant bone marrow dysfunction. The protocol was approved by the EORTC Protocol Review Committee and the institutional review committees of all participating hospitals, and informed consent was obtained from all patients before being randomly assigned. Only centers with previous experience in the administration of high-dose IL-2 participated in this trial.
Treatment Plan
All patients received dacarbazine, 250 mg/m2/d IV on days 1 to 3, cisplatin 30 mg/m2/d IV on days 1 to 3, and IFN--2b (Schering-Plough, Kenilworth, NJ) 10 x 106 U/m2/d subcutaneously on days 1 to 5. Patients randomly assigned to arm B received IL-2 (Proleukin; Chiron, Amsterdam, Netherlands) in addition as a continuous IV infusion in a decrescendo schedule starting on day 5 with 18 million U/m2 over 6 hours followed by 18 million U/m2 over 12 hours, 18 million U/m2 over 24 hours, and a maintenance dose of 4.5 million U/m2/24 hours for an additional 72 hours. This schedule resulted in the administration of a total dose of 67.5 million U/m2 of IL-2 over 114 hours. Cycles were repeated every 4 weeks for a maximum of four cycles; treatment was stopped at any time if there was evidence of disease progression or in the event of unacceptable toxicity. Patients with stable disease after two cycles of treatment received two additional courses of treatment.
Recommended concomitant medications included prophylactic 5HT3 antagonists before dacarbazine and cisplatin, and paracetamol or indomethacin to ameliorate cytokine-related fever. Low-dose furosemide or renal-dose dopamine was recommended for IL-2-related oliguria, and IV saline and colloids were recommended for hypotension.
The development of any grade 4 toxicity resulted in protocol treatment being discontinued, except in the case of grade 4 hematologic toxicities, provided they had resolved by the time the next cycle was due. The IL-2 infusion was interrupted if the following grade 3 toxicities occurred during treatment: hypotension not responding to concomitant therapy, cardiac arrhythmia, suspicion of myocardial ischemia, agitation or persistent confusion, elevation of bilirubin (> 60 μmol/L), bacterial sepsis, or dyspnea at rest. In these circumstances, toxicity was reassessed every 2 hours until it resolved to grade 1 or less and restarted at 50% of the original dose. Dose reductions were not carried over to subsequent treatment cycles. Only if there had been a transient rise in serum creatinine exceeding 265 μmol/L or grade 4 neurotoxicity during the previous cycle was the dose of IL-2 reduced by 50% for the next cycle.
Response and Toxicity Assessments
Tumor response was assessed with appropriate imaging procedures after the second and fourth treatment cycles according to standard WHO criteria.26 Patients achieving a PR or CR were in some analyses combined as "responding" patients. Toxicity was evaluated by using standard WHO criteria and assessed before the next cycle of treatment. The toxicity of the last cycle was assessed after 4 weeks. After the completion of study treatment, patients were followed every 2 months for the first 6 months and then once every 3 months.
Statistical Design
Randomization was performed centrally (EORTC Data Center), with stratification for center and serum lactate dehydrogenase (LDH) level (three groups: < upper limit of normal [ULN], between ULN and 2x ULN, and 2x ULN) by using the minimization technique.
The primary end point was the duration of survival after being randomly assigned to a treatment arm. This was defined as time from assignment until death, whatever the cause; patients still alive were censored at their latest date of follow-up. Secondary end points were RR (best response observed during the study: CR or PR), progression-free survival (PFS; time from assignment until progression or death, whatever the cause; patients still alive without progression were censored at their latest date of follow-up), and relapse-free survival (RFS; for patients who reached CR or PR, time from PR or CR until relapse or death, whatever the cause; patients still alive without relapse were censored at their latest date of follow-up).
The aim of the study was to detect a difference between the two treatment arms in terms of overall survival rate at 2 years. The goal was set initially to enter 156 patients to detect a 15% difference (10% v 25%) in the 2-year survival rates between the two arms (two-tailed test: = 5%; ? = 20%). Because the accrual of patients was achieved quickly and almost 100 patients were accrued, it was decided to increase the study population up to a minimum of 338 eligible patients to allow detection of a smaller difference of 10%, which was the minimum felt to be medically relevant. A total of 286 patients followed until death would allow the detection with an 85% statistical power ( = 5%) of a 10% difference in the 2-year survival rates (10% v 20%), corresponding to a hazard ratio (HR) of 0.70 in case of an exponential distribution.27 To make up for potential noneligible patients and patients lost to follow-up, another 7.5% of patients were added to a target accrual goal of 363 patients.
The actuarial curves were computed by using the Kaplan-Meier technique and the SEs of the estimates were obtained by using the Greenwood formula.28 The difference between curves was tested for statistical significance by using the two-tailed log-rank test.28 The Cox proportional hazards model was used to obtain the estimate and the 95% CI of the HR of the instantaneous event rate in the experimental group versus the one in the control group, adjusting by possible confounding factors; the Wald test was used to determine the prognostic importance of each variable included in the model.29 Prognostic interaction between variables was tested by including products of variables into the model. All analyses were performed according to the intent-to-treat-principle. The 2 test was used to compare the RRs of treatment arms; adjustments by other factors were performed via the logistic model.
The database was frozen on September, 2002. SAS 8.1 software (SAS Institute, Cary, NC) was used for the statistical analyses.
RESULTS
Patient Characteristics and Study Flow
Between May 1995 and April 2000, 363 patients from 25 centers were accrued and randomly assigned. Their pretreatment characteristics are listed in Table 1 . This patient population was largely a group with advanced metastatic melanoma (large majority with M1c disease) who had good PS (> 80% of patients had a Karnofsky index of 90% and 100%). Patient characteristics were well balanced between treatment arms. The study-flow summary is listed in Table 2. In essence, 3% of the patients did not start the allocated treatment, usually because of rapid disease progression or withdrawn consent. Eleven patients (3%) were found to be ineligible. However, all patients were included in the statistical evaluations, because all analyses were based strictly on an intent-to-treat analysis.
Adherence to Treatment Plan
The treatment was administered as planned in 89% and 72% of patients in arms A and B, respectively. Details of treatment administration are summarized in Table 3. The most common reason for treatment termination was progressive disease, occurring as a reason for discontinuation among 50% of patients in arm A and 36% of patients in arm B. Toxicity of treatment or patient refusal were the next most frequent reasons for treatment discontinuation, which were reported for 6.7% of patients in arm A and 21.8% of patients in arm B. Chemotherapy or IFN was rarely dose-reduced in cycle one, whereas the IL-2 dose was reduced during cycle one in 24% of patients who received it (arm B). Overall, chemotherapy and IFN were usually administered at the full planned dose, but the IL-2 dose was reduced or interrupted in 15.5% and 33.9% of patients, respectively.
Toxicities
As expected, treatment in arm B was associated with a higher incidence of grade 3 to 4 hypotension, fever without infection, lethargy, anorexia, and diarrhea (Table 4). The nadirs of the total white blood cell and neutrophil counts were lower in arm A, whereas the platelet nadirs were lower in arm B.
Survival
The median follow-up was 3.38 years at the time of the final analysis, by which time a total of 328 deaths had occurred. The median survival was 9.0 months in each arm. The treatment difference was not significant (P = .31; Fig 1). Stratification by the initial LDH (three categories) indicated at randomization yielded P = .31, the estimated HR (arm B v A) was 0.89, and the 95% CI was 0.72 to 1.11. The 2-year survival-rate estimates were 12.9% (SE, 2.5%) and 17.6% (SE, 2.8%) in arms A and B, respectively. An apparent separation of the survival curves after 2 years was based on too few patients at risk to contribute to overall significance. Nonetheless, the number of patients surviving at the time of study lock was greater in arm B than in arm A (24 v 11 patients; Table 5).
Secondary analyses were performed to investigate the contribution of established major prognostic factors. Univariate analyses (not shown) considered LDH, American Joint Committee on Cancer (AJCC) M stage, and PS. AJCC M stage did not reveal any differences, but few patients were M1a, and the majority were M1c. LDH and PS revealed significant differences. As for LDH, only two of the initial three categories lead to distinct survival curves, resulting in a median survival of 5.1 months with the LDH above 2x ULN and 10.1 months with LDH below 2x ULN. Also, PS was of predictive value, with the distinction already occurring between a Karnofsky index of 100% (median survival, 11.4 months) versus 90% (median survival, 7.6 months) and only few changes below that level. The Cox proportional hazards model considering LDH and PS showed that the initial LDH (two categories: 2x ULN v < 2x ULN; HR, 2.43; P < .0001) and PS (two categories: < 100% v 100%: HR, 1.53; P = .0002) seemed to be important independent prognostic factors (Fig 2). The estimated HR (arm B v A) adjusted for these factors was 0.895 (95% CI, 0.72 to 1.11; P = .32). In the Cox model, the interaction between treatment and a score based on LDH and PS was not significant (P = .16).
PFS
The analysis for PFS included all 363 patients, of whom 359 were followed until progression or death. No significant (P = .28) difference was detected between the two arms (Fig 3). The median PFS was 3.0 months (arm A) versus 3.96 months (arm B), and the estimated 2-year PFS rate was 3.89% (SE, 1.44%) versus 4.50% (SE, 1.50%). The estimated HR adjusted for LDH (three categories) at randomization was 0.87 (95% CI, 0.70 to 1.07; P = .19). In a Cox model, LDH (< 2x ULN v > 2x ULN; P = .002) and PS (< 100% v 90%; P = .04) seemed to be of prognostic importance, whereas treatment remained not significant (P = .19; HR, 0.87; 95% CI, 0.70 to 1.07).
RR
The overall objective RR (CR + PR) was similar in the two arms: 22.8% (SE, 3.1%) versus 20.8% (SE, 3.0%) in arms A and B, respectively (Table 5). The observed difference (2%, with a 95% confidence limit of –6.5%, 10.5%) was not significant (P = .74). The logistic model indicated that the initial LDH (< 2x ULN v > 2x ULN; P = .66) and the PS (< 100% v 100%; P = .15) had no prognostic importance, and the treatment impact was not significant (P = .65; odds ratio, 0.89; 95% CI, 0.54 to 1.47).
RFS
The RFS analysis is based on 79 patients who achieved a PR or CR, of whom 70 relapsed or died during follow-up (Fig 4). There was no significant difference in RFS between the two treatment arms (P = .89; HR, 0.97; 95% CI, 0.60 to 1.52). A Cox model confirmed these results: the initial PS was of no prognostic importance (HR, 1.24; P = .2), whereas the initial LDH was borderline significant (HR, 2.16; P = .06). The type of response had prognostic importance (P = .002; Fig 5): those who reached a CR (n = 13) had a longer RFS than those who reached a PR (n = 66). However, there was no impact of treatment on RFS in CR patients nor in PR patients considered separately. The time to response was quite similar between the treatment arms, because most PRs were first noted after two treatment cycles, and most CRs were first noted after four cycles.
DISCUSSION
This study shows that IL-2 added to a complex regimen given for a maximum of four cycles and consisting of dacarbazine, cisplatin, and IFN- does not confer a clinically meaningful survival benefit for most patients with advanced melanoma, nor does it confer an increase in RR or time to progression. These results are in line with several smaller previous randomized phase II and III trials that evaluated the effect of adding intermediate-dose IL-2 to dacarbazine/IFN-,30 low-dose subcutaneous IL-2 and IFN- to dacarbazine/cisplatin/carmustine/tamoxifen,31 low-dose subcutaneous IL-2 and IFN- to dacarbazine/cisplatin/carmustine,32 intermediate-dose IL-2 and IFN- to dacarbazine/cisplatin/carmustine/tamoxifen,33 and high-dose IL-2 and IFN- to dacarbazine/cisplatin/tamoxifen.34
The only randomized trial with borderline significance for overall survival improvement (P = .06 on two-sided log-rank test) is a single-institution study35 that investigated the effect of adding IL-2 and IFN- to a polychemotherapy regimen consisting of cisplatin, vinblastine, and dacarbazine. However, recent data suggest that the trend for a survival benefit observed in this trial cannot be confirmed in a larger, multi-institutional setting using a modified regimen and dose-adjustment schema given for a maximum of four cycles (M.B. Atkins, personal communication, May 2003).
Most previous trials30-33 have been criticized because they have investigated IL-2 regimens that were never tested for their capacity to induce remission in phase II studies. However, the decrescendo regimen used in the trial reported here was active in prior phase II studies, and its activity was confirmed in a previous phase III trial.26 Despite these promising early data, in the trial reported here, we have not been able to show that IL-2 adds to the efficacy of the combination of dacarbazine, cisplatin, and IFN-.
There are several possible explanations for the failure of this trial to detect improved efficacy. First, it is conceivable that melanoma has a subset that does not respond to either chemotherapy or cytokine treatment. Second, combination chemotherapy could impair the mechanisms by which IL-2 induces long-term remissions. Our finding that RRs between this IL-2-based biochemotherapy regimen and non–IL-2-based therapy are the same would be entirely consistent with both of these hypotheses. Alternatively, the dose intensity of IL-2 achieved in this multicenter trial may have been below an as-yet-undefined threshold necessary for efficacy, but because the study was performed only in centers with experience in IV IL-2 regimens and dose reductions were mandated by toxicity and treatment termination was mandated by patient desire in 6%, an increase in dose intensity would be difficult to achieve in a multicenter setting.
Physicians treating patients with metastatic melanoma need to acknowledge the negative results of the study reported here and similar data from smaller trials. The combination of chemotherapy and immunotherapy in chemoimmunotherapy regimens, which have dominated melanoma treatment strategy for the past 10 years, has not proven successful; to date, no such treatment regimen has been shown in a multicenter setting to significantly prolong survival in patients with stage IV melanoma. The efficacy of IL-2 in melanoma is established, but not to an extent that, in the view of the EORTC Melanoma Group, justifies its use in the context of biochemotherapy outside of clinical trials. Future research should focus on enhancement of the immunologic activity of IL-2, as currently pursued by combinations with histamine or vaccines.36,37
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Acknowledgment
We greatly acknowledge the tremendous efforts of the staff in all participating hospitals and the superb work by the melanoma team of the European Organisation for Research and Treatment of Cancer Data Center.
NOTES
Supported by educational grants from Chiron BV (Amsterdam, Netherlands), Schering Plough (Kenilworth, NJ), and the National Cancer Institute (Bethesda, MD; grants 2U10 CA11488-25 through 5U10 CA11488-32).
The article’s contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Sparano J, Fisher R, Sunderland M, et al: Randomized phase III trial of treatment with high-dose interleukin-2 either alone or in combination with interferon-alfa-2a in patients with advanced melanoma. J Clin Oncol 11:1969-1977, 1993
Rosenberg S, Lotze M, Yang J, et al: Experience with the use of high-dose interleukin-2 in the treatment of 652 cancer patients. Ann Surg 210:474-485, 1989
Thatcher N, Dazzi H, Gosh A, et al: Recombinant IL-2 given intra-splenically and intravenously in advanced malignant melanoma: A phase I/II study. Cancer Treat Rev 16:49-52, 1989 (suppl A)
Parkinson D, Abrams J, Wiernik P, et al: Interleukin-2 therapy in patients with metastatic malignant melanoma: A phase II study. J Clin Oncol 8:1650-1656, 1990
Whitehead R, Kopecky K, Samson M, et al: Phase II study of intravenous bolus recombinant interleukin-2 in advanced malignant melanoma: Southwest Oncology Group study. J Natl Cancer Inst 83:1250-1251, 1991
Dorval T, Mathiot C, Brandely M, et al: Lack of effect of tumour infiltrating lymphocytes in patients with metastatic melanoma who failed to respond to interleukin 2. Eur J Cancer 27:99, 1991 (suppl 2)
Legha S, Gianan M, Plager C, et al: Evaluation of interleukin-2 administered by continuous infusion in patients with metastatic melanoma. Cancer 77:89-96, 1996
Rosenberg S, Lotze M, Yang J, et al: Combination therapy with interleukin-2 and alpha-interferon for the treatment of patients with advanced cancer. J Clin Oncol 7:1863-1874, 1989
Kruit W, Goey S, Monson J, et al: Clinical experience with the combined use of recombinant interleukin-2 (IL-2) and interferon-alpha-2a (IFN alpha) in metastatic melanoma. Br J Haematol 79:84-86, 1991 (suppl 1)
Dillman R, Church C, Oldham R, et al: Inpatient continuous-infusion interleukin-2 in 788 patients with cancer. The National Biotherapy Study Group experience. Cancer 71:2358-2370, 1993
Marincola F, White D, Wise A, et al: Combination therapy with interferon alfa-2a and interleukin-2 for the treatment of metastatic cancer. J Clin Oncol 5:1110-1122, 1995
Kruit W, Punt C, Goey S, et al: Dose efficacy study of two schedules of high-dose bolus administration of interleukin-2 and alpha-interferon in patients with metastatic melanoma. Br J Cancer 74:951-955, 1996
Keilholz U, Scheibenbogen C, Tilgen W, et al: Interferon- and interleukin-2 in the treatment of malignant melanoma: A comparison of two phase II trials. Cancer 72:607-614, 1993
Stoter G, Shiloni E, Aamdal S, et al: Sequential administration of recombinant human interleukin-2 and dacarbazine in metastatic melanoma. A multicentre phase II study. Eur J Cancer Clin Oncol 25:41-43, 1989 (suppl 3)
Dillman R, Oldham R, Barth N, et al: Recombinant interleukin-2 and adoptive immunotherapy alternated with dacarbazine therapy in melanoma: a National Biotherapy Study Group trial. J Natl Cancer Inst 82:1345-1349, 1990
Flaherty L, Redman B, Chabot G, et al: A phase I-II study of dacarbazine in combination with outpatient interleukin-2 in metastatic malignant melanoma. Cancer 65:2471-2477, 1990
Demchak P, Mier J, Robert N, et al: Interleukin-2 and high-dose cisplatin in patients with metastatic melanoma: A pilot study. J Clin Oncol 9:1821-1830, 1991
Flaherty L, Robinson W, Redman B, et al: A phase II study of dacarbazine and cisplatin in combination with outpatient administered interleukin-2 in metastatic malignant melanoma. Cancer 71:3520-3525, 1993
Atkins M, O’Boyle K, Sosman J, et al: Multiinstitutional phase II trial of intensive combination chemotherapy for metastatic melanoma. J Clin Oncol 12:1553-1560, 1994
Dummer R, Gore M, Hancock B, et al: A multi-center phase II clinical trial using dacarbazine and continuous infusion of interleukin-2 in metastatic melanoma: Clinical data and immunomonitoring. Cancer 75:945-948, 1995
Richards J, Mehta N, Ramming K, et al: Sequential chemoimmunotherapy in the treatment of metastatic melanoma. J Clin Oncol 10:1338-1343, 1992
Khayat D, Borel C, Tourani J, et al: Sequential chemoimmunotherapy with cisplatin, interleukin-2, and interferon alfa-2a for metastatic melanoma. J Clin Oncol 11:2173-2180, 1993
Legha SS, Ring S, Eton O, et al: Development of a biochemotherapy regimen with concurrent administration of cisplatin, vinblastine, dacarbazine, interferon alfa, and interleukin-2 for patients with metastatic melanoma. J Clin Oncol 16:1752-1759, 1998
Legha SS, Ring S, Bedikian A, et al: Treatment of metastatic melanoma with combined chemotherapy containing cisplatin, vinblastine and dacarbazine (CVD) and biotherapy using interleukin-2 and interferon-alpha. Ann Oncol 7:827-835, 1996
Proebstle TM, Scheibenbogen C, Sterry W, et al: A phase II study of dacarbazine, cisplatin, interferon-alpha, and high dose IL-2 in ‘poor risk’ metastatic melanoma. Eur J Cancer 32:1530-1533, 1996
Keilholz U, Goey S, Punt C, et al: Interferon alfa-2a and interleukin-2 with or without cisplatin in metastatic melanoma: A randomized trial of the European Organization for Research and Treatment of Cancer Melanoma Cooperative Group. J Clin Oncol 15:2579-2588, 1997
Keilholz U, Martus P, Punt CJA, et al: Prognostic factors for survival and factors associated with long-term remission in patients with advanced melanoma receiving cytokine-based treatments: second analysis of a randomised EORTC Melanoma Group trial comparing interferon-alpha2a (IFNalpha) and interleukin 2 (IL-2) with or without cisplatin. Eur J Cancer 38:1501-1511, 2002
Freedman LS: Tables of the numbers of patients required in clinical trials using the logrank test. Stat Med 1:121-129, 1982
Hosmer DW, Lemershow S. Applied Survival Analysis: Regression Models of Time to Event Data. New York, NY, John Wiley & Sons, 1999
Hauschild A, Garbe C, Stolz W, et al: Dacarbazine and interferon alpha with or without interleukin 2 in metastatic melanoma: A randomized phase III multicenter trial of the Dermatologic Oncology Group (DeCOG). Br J Cancer 84:1036-1042, 2001
Atzpodien J, Neuber K, Kamanabrou D, et al: Combination chemotherapy with or without s.c. IL-2 and IFN-alpha: Results of a prospectively randomized trial of the Cooperative Advanced Malignant Melanoma Chemoimmunotherapy Group (ACIMM). Br J Cancer 86:179-184, 2002
Ridolfi R, Chiarion-Seleni V, Guida M, et al: Cisplatin, dacarbazine with or without subcutaneous interleukin-2, and interferon alpha-2b in advanced melanoma outpatients: Results from an Italian multicenter phase III randomized clinical trial. J Clin Oncol 20:1600-1607, 2002
Johnston S, Constenia D, Moore J, et al: Randomized phase II trial of BCDT [carmustine (BCNU), cisplatin, dacarbazine (DTIC) and tamoxifen] with or without interferon alpha (IFN-alpha) and interleukin (IL-2) in patients with metastatic melanoma. Br J Cancer 77:1280-1286, 1998
Rosenberg SA, Yang JC, Schwartzentruber DJ, et al: Prospective randomized trial of the treatment of patients with metastatic melanoma using chemotherapy with cisplatin, dacarbazine, and tamoxifen alone or in combination with interleukin-2 and interferon alfa-2b. J Clin Oncol 17:968-975, 1999
Eton O, Legha S, Bedikian A, et al: Sequential biochemotherapy versus chemotherapy for metastatic melanoma: Results from a phase III randomized trial. J Clin Oncol 20:2045-2052, 2002
Agarwala SA, Glaspy J, O’Day SJ, et al: Results from a randomized phase III study comparing combined treatment with histamine dihidrochloride plus interleukin-2 versus interleukin-2 alone in patients with metastatic melanoma. J Clin Oncol 20:125-133, 2002
Rosenberg SA, Yang JC, Schwartzentruber DJ, et al: Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma. Nat Med 4:321-327, 1998(Ulrich Keilholz, Cornelis)