TESTICULAR CANCER has been one of the major success

Transcription

1 Intensive Induction Chemotherapy With CBOP/BEP in Patients With Poor Prognosis Germ Cell Tumors By J.A. Christian, R.A. Huddart, A. Norman, M. Mason, S. Fossa, N. Aass, E.J. Nicholl, D.P. Dearnaley, and A. Horwich Purpose: Despite a high cure rate in patients with testicular cancer, there remain patients in the poor prognosis group who have a less favorable outcome. Intensive induction chemotherapy using a regimen consisting of carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin (CBOP/BEP), developed at the Royal Marsden Hospital, is designed to overcome the rapid proliferation seen in germ cell tumors. This study assesses the outcome of patients with poor-prognosis nonseminomatous germ cell tumors (NSGCT) treated with CBOP/BEP. Patients and Methods: Patients with NSGCT from three centers, classified as poor prognosis according to International Germ Cell Classification Consensus Group criteria, were treated with CBOP/BEP regimen during the period from 1989 to Data on treatment toxicity, relapse-free survival (RFS), and overall survival (OS) were collected prospectively on a hospital database. TESTICULAR CANCER has been one of the major success stories for modern multiagent cytotoxic chemotherapy. The majority of patients with metastatic germ cell tumors can expect to be cured. However, patients with nonseminomatous germ cell tumors (NSGCT) and high tumor markers, nonpulmonary visceral metastases, or a mediastinal primary site at presentation have a less certain outcome. When treated with standard chemotherapy regimens such as bleomycin, etoposide, and cisplatin (BEP), cure rates of less than 50% have been achieved in an international pooled analysis. 1 Patients with a mediastinal primary site are known to have a particularly poor outcome. Clearly, treatment results in this poor-prognosis group of patients needs to be improved. A number of strategies have been tried, including the use of multiagent regimens such as cisplatin, vincristine, methotrexate, bleomycin, dactinomycin, cyclophosphamide, and etoposide, 2 new chemotherapy drugs such as ifosfamide, 3 and high-dose chemotherapy. 4 To date none have been proven to be better than standard BEP in randomized trials. Several studies have demonstrated that germ cell tumors have a rapid proliferation rate. Malaise et al 5 found the mean doubling time of embryonal lung metastases to be 19.5 days, whereas a study at the Royal Marsden Hospital suggested a median doubling time of 8 days for tumors producing beta human chorionic gonadotropin ( -HCG). 6 It was further demonstrated using flow cytometry that these tumors could have an S-phase fraction of up to 51% (between two and four times that of other cancers), 7 and two studies have shown labeling indices of 44% and 80%, respectively. 8,9 In addition to the rapid proliferation rate, there is likely to be a phase of accelerated tumor proliferation at the start of treatment, as is well described in radiation therapy, 10 when potential doubling times could decrease by more than half. Dearnaley 11 calculated that a tumor with a doubling time of 2 days could produce cells as a result of cell Results: Fifty-four male patients with poor prognosis NSGCT were treated with CBOP/BEP. The RFS at 3 and 5 years for all patients was 83.2% (95% confidence interval [CI], 68.8% to 91.3%). After a median follow-up of 4 years, the OS of the 54 patients was 91.5% (95% CI, 78.6% to 96.8%) at 3 years and 87.6% (95% CI, 71.3% to 94.9%) at 5 years. Three-year OS in patients with a primary mediastinal germ cell tumor was 77.1% (95% CI, 34.5% to 93.9%) compared with 95.4% (95% CI, 82.8% to 98.8%) in patients with a testicular primary tumor (P.24). Conclusion: The results reported here compare favorably with the historical results of alternative regimens used in the management of poor-prognosis NSGCT. We suggest a phase III trial to confirm our findings. J Clin Oncol 21: by American Society of Clinical Oncology. repopulation during four cycles of standard chemotherapy given every 3 weeks. This proliferation could possibly contribute to treatment failure in some patients. In an effort to overcome rapid proliferation, we have previously described an intensive induction regimen (bleomycin, vincristine, and cisplatin [BOP]/BEP) on the basis of the schedule of Wettlaufer. 11,12 Features of this regime include weekly cisplatin for 4 weeks with weekly bleomycin and vincristine (substituting for etoposide or vinblastine because of vincristine s lesser myelosuppression) for 6 weeks. Included in this schedule is a 5-day infusion of bleomycin in the second and fourth weeks according to the work of Samuels et al, 13 who found that this was the most efficacious way of using bleomycin in large-volume tumors, and Gerson et al, 14 who found that continuous infusions may be less toxic than bolus injections. This is followed by three courses of BEP chemotherapy. Development of the current CBOP/BEP schedule adding carboplatin weeks 2 and 4, modifying the bleomycin schedule using From the Academic Department of Radiotherapy and Oncology, Royal Marsden Hospital, Sutton, Surrey, and Velindre Hospital, Cardiff, United Kingdom; and Det Norske Radium Hospital, Montebello, Oslo, Norway. Submitted May 24, 2002; accepted November 21, Supported by the Institute of Cancer Research and Cancer Research UK (grant no. C46/A2131), London; and Bob Champion Cancer Trust, Surrey, UK. This work was undertaken in The Royal Marsden National Health Service (NHS) Trust, which received a proportion of its funding from the NHS Executive. The views expressed in this publication are those of the authors and not necessarily those of the NHS Executive. Address reprint requests to J.A. Christian, MD, Academic Department of Radiotherapy and Oncology, Royal Marsden Hospital, Downs Rd, Sutton, Surrey SM2 5PT, United Kingdom; j.christian@icr.ac.uk by American Society of Clinical Oncology X/03/ /$20.00 Journal of Clinical Oncology, Vol 21, No 5 (March 1), 2003: pp DOI: /JCO

2 872 CHRISTIAN ET AL Fig 1. Diagram of the regimen consisting of carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin (BOP/BEP) to show bolus (1), daily-infusion o, and continuous-infusion f drug scheduling, with dose details of the BEP phase. *Bleomycin 75,000 U continuous infusion over 5 days weeks 2 and 4. #Cisplatin is given at 50 mg/m 2 /d via protracted infusion days 1 and 2 in weeks 1 and 3. The schedule changes to 20 mg/m 2 /d via protracted infusion days 1 through 5 in weeks 7, 10, and 100 mg/m 2 /d days 1 through 5 via protracted infusion in weeks 7, 10, and ,000-U aliquots, and increasing the etoposide dosage in weeks 7, 10, and 13 has been previously described. 11 Initial data on 21 patients treated during the period of August 1989 to January 1992 were published in 1994, 15 when a 2-year overall survival rate (OS) of 90% and 2-year failure-free survival rate (FFS) of 86% were reported. We have reclassified our patients according to the International Germ Cell Cancer Collaborative Group (IGCCCG) classification, 1 and in this report we update these results on a larger group of patients with poor-prognosis NSGCT. Eligibility PATIENTS AND METHODS Patients were included from three centers: the Royal Marsden Hospital, Surrey, United Kingdom; the Radium Hospital, Oslo, Norway; and the Velindre Hospital, Cardiff, United Kingdom. Patients with a histologic diagnosis of NSGCT and a primary site in the testis or at an extragonadal primary site were eligible. The criteria for inclusion in the poor-prognosis group were as follows: (1) alpha-fetoprotein (AFP) level greater than 10,000 ng/ml, (2) -HCG greater than 50,000 U/L, (3) lactate dehydrogenase (LDH) greater than 10 times the upper limit of normal, (4) nonpulmonary visceral metastases, or (5) mediastinal primary site. Patients consented to receive the CBOP/BEP regimen as the standard regimen used in these centers for poor-prognosis NSGCT. The diagnosis was made using histologic data combined with staging information and serum tumor markers. In the absence of a histologic diagnosis (because of need for urgent chemotherapy treatment or difficulty in obtaining an adequate biopsy), an elevated serum -HCG and/or AFP level with a clinical picture and staging compatible with poor-prognosis NSGCT was accepted. Patients could not have received previous chemotherapy or radiotherapy. Before 1997, patients were categorized as poor prognosis using the Medical Research Council Prognostic Factor Analysis 16 and staged according to the Royal Marsden Hospital staging classification. 17 With the development of the IGCCCG classification in 1997, this has been used to reclassify patients treated before this date and has been prospectively used since. Patients who were reclassified according to IGCCCG criteria to intermediate prognosis have been excluded from analysis. Initial Assessment All patients were assessed before commencing treatment with standard hematologic and biochemical parameters, serum tumor markers (AFP and -HCG), chest x-ray, and baseline computed tomography (CT) scans of the chest, abdomen, and pelvis. From 1995, LDH level was measured for all patients but was not universally available before then. These investigations, CT scans excluded, were repeated weekly during treatment. Patients underwent a glomerular filtration rate (GFR) test to assess renal function. The GFR was determined by clearance of chromium-51 EDTA. All patients were required to have a creatinine clearance greater than 60 ml/min before commencing chemotherapy unless an obstructive uropathy was present, in which case the ureter was stented before commencing treatment. CT (or magnetic resonance imaging) of brain was performed in symptomatic patients and in men who had high -HCG levels ( 20,000 U/L) and/or more than 20 lung metastases. CBOP/BEP Regimen The CBOP/BEP regimen is shown in Fig 1 and is as previously described. 15 It is structured in two parts; the accelerated schedule CBOP lasts 6 weeks. Cisplatin is administered at the dose of 100 mg/m 2 on weeks 1 and 3 and at a dose of 40 mg/m 2 in weeks 2 and 4. Carboplatin is given in weeks 2 and 4 as an infusion after the 6-hour cisplatin infusion is completed. When the regimen was first designed, weeks 1 and 3 cisplatin was given as a daily infusion of 20 mg/m 2 /d for 5 days. Since 1993, cisplatin has been administered as a daily infusion of 50 mg/m 2 /d for 2 days. All cisplatin is given with at least 3 L of 0.9% saline hydration and 20% mannitol to encourage diuresis. The carboplatin is given as a 60-minute infusion in 500 ml of 5% dextrose. The dose is calculated according to the Calvert formula: 18 carboplatin dose (mg) AUC (25 GFR), where AUC (area under the plasma concentration-time curve) the desired plasma concentration of carboplatin, which is set at 3 mg/ml. Vincristine is given as a weekly bolus dose of 1.4 mg/m 2 (maximum 2 mg) by intravenous bolus injection. Bleomycin 75,000 U is given as a continuous infusion over 5 days during weeks 2 and 4 and by bolus intravenous injection (15,000 U) with 100 mg of hydrocortisone in other weeks. All of the above chemotherapy is given with appropriate antiemetics. Once the 6-week CBOP schedule is completed, the patients are then treated with three cycles of BEP chemotherapy, as described in Fig 1. Bleomycin is given weekly during BEP at 15,000 U to limit the total maximum dose for the CBOP/BEP regimen to 345,000 U. During chemotherapy, weekly tumor marker levels are assessed and chest x-rays (weekly during CBOP phase and once every three weeks during BEP phase) are carried out to monitor toxicity and response. Outcome Measures and Follow-Up After completion of the CBOP/BEP regimen, response was assessed by CT scans of chest, abdomen, and pelvis in addition to measurement of tumor markers. Patients with residual masses after completing chemotherapy were considered for surgical resection, although often if the tumor markers were normal and multiple sites were present, an initial policy of monitoring was undertaken. Radiotherapy was considered for sites of residual disease where surgical excision was not possible and when tumor markers had failed to normalize at the end of chemotherapy. Patients with brain metastases were considered for whole-brain radiotherapy at the end of treatment or intercalated between treatment cycles. After treatment was complete, patients were followed-up every 2 months in the first year, every 3 months in the second year, every 4 months in the third year, every 6 months up to 5 years, and annually thereafter. Relapse was defined as either a consistent increase in the serum tumor markers, the serial increase in size of a residual mass, or the development of new tumor lesions in patients. Patients were additionally classified as having a complete response (CR) or a partial response (PR) to chemotherapy. A CR was defined as the presence of normal tumor marker levels after chemotherapy was completed, with no clinical or radiologic evidence of residual

3 CBOP/BEP FOR POOR-PROGNOSIS GERM CELL TUMORS 873 masses and no further treatment being required. Patients who had normal tumor marker levels after completing chemotherapy but who had clinical or radiologic evidence of residual masses were still defined as having CR if a complete surgical resection of all masses showed no evidence of viable tumor (ie, histology showed necrosis/fibrosis or mature differentiated teratoma). Occasionally, the tumor markers had not normalized at the end of chemotherapy but decreased consistently from extremely high levels to normal after the patient completed chemotherapy and had no additional treatment. Such a pattern was considered normal because of the physiologic tumor marker half-life, and these patients were also included in the CR category. Patients who had normal tumor marker levels after completing chemotherapy in addition to residual masses, with or without surgery, were classified as tumor marker negative PR. Both CR and tumor marker negative PR are considered favorable responses to treatment. Patients were described as incomplete responders when markers failed to normalize at the end of primary treatment. When surgery for residual masses showed malignant viable tumor cells, this was considered an incomplete response, but if the disease was completely removed and no further treatment was required, these patients were considered for analysis of relapse-free survival. Disease progression during chemotherapy was defined as a serial increase in tumor markers, serial increase in size of a residual mass, or the development of new tumor lesions; these patients were also called incomplete responders. Toxicity Treatment-related toxicity was assessed during and after chemotherapy was complete. Lung toxicity related to the use of bleomycin was assessed during the initial accelerated phase by documenting patient symptomatology, clinical examination, and plain x-rays on a weekly basis. During the BEP phase, the same investigations were repeated every 3 weeks. Pulmonary function tests and CT scans were carried out only if the patient developed respiratory symptoms. Neurotoxicity was documented during chemotherapy and at clinic follow-up visits. Common toxicity criteria (CTC) scoring was used to document toxicity. Treatment Modifications Patients with dyspnea related to bulky lung disease or with hydronephrosis at the time of starting chemotherapy had their week 1 CBOP cisplatin substituted by carboplatin to reduce the fluid load required with chemotherapy. A small number of patients had their week 1 CBOP bleomycin omitted because of bulky lung disease. Patients were switched to a non bleomycincontaining regimen if there was evidence of bleomycin lung toxicity. During the accelerated phase of treatment (ie, weeks 1 to 6), treatment was not given if total leukocytes decreased to less than /L or platelet count decreased to less than /L. Such cycles were simply omitted from the schedule and the time course of the regimen did not alter so that the first cycle of BEP always commenced week 7. Cycles of BEP were delayed if leukocytes decreased to less than /L or platelets decreased to less than /L until the first day that leukocyte or platelet counts were demonstrated to be increasing above the nadir. Data Collection and Statistical Methods Data at the Royal Marsden Hospital have been collected prospectively onto a tumor database since Some of the data from Cardiff and Norway have been collected retrospectively. A proportion of patients were followed-up for long-term side effects of treatment as part of another study. Toxicity data were collected from hematologic and biochemical records and from regular assessments during and after treatment. Kaplan-Meier statistical analysis has been used to represent relapse-free survival and OS curves. A subanalysis has independently assessed the differences in relapse-free survival and OS between primary testicular tumors and primary mediastinal tumors. RESULTS Between August 1989 and September 2000, 54 male patients with primary presentation poor-prognosis NSGCT were treated with CBOP/BEP chemotherapy. All such patients treated in the three centers were included in this analysis except for patients Table 1. Distribution of Poor Prognostic Factors Among the 54 Patients Who Received CBOP/BEP No. of Patients % Serum tumor markers only, ie. AFP 10,000 ng/ml or -HCG 50,000 U/L of LDH 10 upper limit of normal Nonpulmonary visceral metastases only; eg, brain, liver metastases Serum tumor markers and nonpulmonary visceral metastases Mediastinal primary site Abbreviations: CBOP/BEP, carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin; AFP, alpha-fetoprotein; HCG, human chorionic gonadotropin; LDH, lactate dehydrogenase. treated in European Organization for Research and Treatment of Cancer (EORTC) Trial The median age was 27 years (range, 16 to 53 years). The median follow-up duration was 4 years. Forty patients were treated at Royal Marsden Hospital, eight at the Radium Hospital, and six at the Velindre Hospital. The series included 44 patients with primary testicular germ cell tumor and 10 patients with primary mediastinal germ cell tumors (MGCT). The reasons these patients were classified as poor-prognosis presentations are listed in Table 1. Biopsy-confirmed disease was present in 42 of 54 patients. In 12 patients, the diagnosis was made on the basis of the clinical picture of NSGCT, with elevated serum marker levels to confirm the diagnosis. This included five of 10 MGCT patients in whom there was difficulty obtaining a good-quality tissue biopsy. All nonbiopsied patients had to have an elevated AFP level to confirm NSGCT rather than seminoma alone. However, four patients have been included in the series, each of whom had a very high -HCG level alone, without histologic confirmation. These patients were considered to have choriocarcinoma. Nineteen patients did not receive a complete course of CBOP/BEP. In 13 patients, weeks 5 and/or 6 bleomycin was omitted because of myelosuppression, and one patient also omitted week 4 chemotherapy, in addition to weeks 5 and 6, because of myelosuppression. Two patients developed bleomycin lung toxicity diagnosed on plain x-ray during the BEP phase; BEP chemotherapy was replaced with ifosfamide, etoposide, and cisplatin for one patient, and the bleomycin was simply omitted henceforth for the other patient. One patient experienced a prolonged period of neutropenia during the BEP phase, and during his last cycle of BEP, the etoposide was omitted. Two patients experienced disease progression during CBOP/BEP, which required a change of therapy. Response Rates Response data were available for all 54 patients. At the end of primary treatment, 16 (29.6%) of 54 patients had a CR. In this group, one patient (1.9%) had no residual masses, whereas the remaining 15 patients (27.8%) had surgery to remove all residual masses (11 had necrosis/fibrosis and four had mature teratoma). Thirty-two patients (59.3%) with a tumor marker negative PR at the end of chemotherapy had residual masses that were partially resected, were monitored during follow-up, or were not able to be surgically excised. Two patients with tumor marker negative

4 874 CHRISTIAN ET AL Fig 2. Relapse-free survival of 54 patients treated with carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin (CBOP/BEP). PR demonstrated viable undifferentiated teratoma cells on histology after resection of a tumor mass and were classified as incomplete responders. Both of these patients remained relapsefree on long-term follow-up, having received no further treatment. Of the three patients who did not achieve negative tumor markers at the end of chemotherapy, two of these patients developed disease progression during CBOP/BEP chemotherapy, and one patient developed disease progression soon after CBOP/BEP was complete, requiring further chemotherapy. Thus, five patients (9.3%), including the patients with viable tumor in the postchemotherapy mass, had an incomplete response to CBOP/BEP. One patient (1.9%) died during chemotherapy as a result of treatment-related toxicity and was categorized as a nonresponder. Relapse-Free Survival The relapse-free survival rate at 3 and 5 years for all patients was 83.2% (95% confidence interval [CI], 68.8% to 91.3%). The relapse-free survival curve is shown in Fig 2. Eight patients experienced relapse during follow-up. The median relapse-free survival has not yet been reached. Five patients were successfully treated with salvage therapy with a combination of additional chemotherapy, radiotherapy, and surgical resection of residual masses. However, in three patients attempts at salvage therapy failed; these patients died of progressive disease. Overall Survival After a median follow-up of 48.5 months, the OS rate of all 54 patients was 91.5% (95% CI, 78.6% to 96.8%) at 3 years and 87.6% (95% CI, 71.3% to 94.9%) at 5 years. The OS curve is shown in Fig 3. Seven patients had died at the end of the follow-up period; three of these died from progressive disease. Three patients died from treatment-related toxicity: one from neutropenic sepsis complicating bleomycin pneumonitis, one from a second malignancy (acute myeloid leukemia) after completing treatment, and one from the effects of somnolence syndrome after whole-brain radiotherapy. The remaining patient died from an unrelated accidental cause. In the patients with a primary mediastinal tumor, the results were slightly less favorable. Three-year OS in patients with MGCT was 77.1% (95% CI, 34.5% to 93.9%) compared with Fig 3. Overall survival for 54 patients treated with carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin (CBOP/BEP).

5 CBOP/BEP FOR POOR-PROGNOSIS GERM CELL TUMORS 875 Fig 4. Comparison of the overall survival of patients treated with carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin (CBOP/BEP) when primary testicular tumors and primary mediastinal tumors were analyzed separately. 95.4% (95% CI, 82.8% to 98.8%) in patients with a testicular primary (P.240) (Fig 4). Toxicity Hematologic, neurologic, and pulmonary toxicity data were available for all 54 patients. Hematologic toxicity. The major toxicity was hematologic, with CTC grade 3/4 leukopenia (WBC count 1.0) in 44.4% of patients. CTC grade 3/4 thrombocytopenia (platelet count 25) was seen in 42.6% of patients (Table 2). Twenty-two (40.7%) of 54 patients required antibiotics for neutropenic sepsis during treatment. Neurologic toxicity. Twenty-two patients (40.7%) experienced no sensory neuropathy, 26 (48.1%) had CTC grade 1 toxicity, five (9.3%) had CTC grade 2 toxicity, and one (1.9%) had CTC grade 3 toxicity. Two patients had CTC grade 3 high-tone hearing loss. Pulmonary toxicity. There were three cases of bleomycin lung toxicity. Two patients, as previously described, developed chest x-ray shadowing and mild respiratory symptoms during the BEP phase of treatment and had a complete resolution of respiratory symptoms after cessation of bleomycin. The other patient died as a result of pneumonitis complicated by Klebsiella pneumoniae during a period of neutropenia, having received only 3 weeks of CBOP chemotherapy. One patient who received radiotherapy to the mediastinum as part of the primary treatment developed CTC grade 2 pulmonary toxicity after completing the radiotherapy. DISCUSSION In this series we have updated previously reported results of this regimen in patients treated at the Royal Marsden Hospital 11,12,15 with a larger series of patients who have been defined as having poor prognosis according to the IGCCCG classification; we have also included patients from two additional centers that have used this approach over recent years. We have achieved a 3-year relapse-free survival rate of 83.2% and a 3-year OS rate of 91.5%, which compare favorably with the reported international figures and our historical data. 19 This is also in line with historical series of patients treated with weekly chemotherapy, which have reported survival rates of 85% to 93% in patients with advanced disease defined in a variety of ways. 12,20,21 The outcome of patients with a mediastinal primary tumor is particularly encouraging; we have achieved a 77.1% 3-year OS rate. We believe the most meaningful way to describe outcome in NSGCT is using measures of relapse-free survival and OS. Interpretation of response rates in advanced NSGCT can be difficult because of residual masses that contain mature differentiated teratoma or necrosis/fibrosis. In our data, we have had an overall favorable response rate of 88.9%, which is not dissimilar to a 3- and 5-year relapse-free survival rate of 83.2%. The CR rate in this series is an underestimate of true disease eradication. This is in part due to patients with multiple masses who, having had one mass resected showing necrosis/fibrosis, had other masses simply observed, and also because a significant number of the patients continue to have resolution of residual masses on monitoring over time, effectively rendering them as CRs. Although the majority of men with metastatic germ cell tumors are curable with platinum-based chemotherapy, an international collaboration has defined prognostic groups on the basis of data from 5,202 patients with NSGCT. The group with the poorest prognosis, comprising 14% of the total number of patients, was defined by the presence of high markers ( -HCG 50,000 U/L, AFP 10,000 ng/ml, LDH 10 upper limit of normal), and/or nonpulmonary visceral metastases, and/or a mediastinal primary site. This group had a predicted 5-year survival rate of 48% (95% CI, 42% to 54%), with a progression- Table 2. Hematologic Toxicity of CBOP/BEP Using Common Toxicity Criteria No. of Patients % No. of Patients % No. of Patients % No. of Patients % No. of Patients % Total (n) Leucopenia Thrombocytopenia Abbreviation: CBOP/BEP, carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin.

6 876 CHRISTIAN ET AL Table 3. Drug Comparison of Drug Dose-Intensity of CBOP/BEP With the MRC Regimen BOP/VIP-B Dose-Intensity (mg/m 2 /wk) During Weeks 1-6 CBOP/BEP MRC BOP/VIP-B 22 Relative Dose-Intensity of CBOP Platinum* Bleomycin Vincristine Abbreviations: CBOP/BEP, carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin; MRC, Medical Research Council; BOP/VIP-B, bleomycin, vincristine, and cisplatin, plus etoposide, ifosfamide, cisplatin, and bleomycin. *Based on carboplatin area under the plasma concentration-time curve 3 being equivalent to cisplatin 50 mg/m 2. free survival rate of 41%. When tested in validation sets of more current patients, 5-year survival rates between 50% and 57% were found. The type of chemotherapy used for these patients is not defined, but most patients were treated in the 1980s, and a significant proportion was likely to have received BEP chemotherapy. A recent randomized trial of a different intensive induction-sequential chemotherapy schedule consisting of bleomycin, vincristine, and cisplatin, plus etoposide, ifosfamide, cisplatin, and bleomycin (BOP/VIP-B) was compared with a regimen based on bleomycin, etoposide, and cisplatin (BEP/ EP). 22 There was no significant difference between the two arms in the proportion of patients who achieved a CR with chemotherapy alone, time to first disease progression, FFS, or OS between the two arms. The 1-year FFS rates for BEP/EP and BOP/VIP-B were 60% and 53%, respectively. Although this group of patients was defined as poor prognosis by Medical Research Council (MRC) criteria, 39% would be defined as good or intermediate prognosis on the current IGCCCG classification. Specific results for those currently defined as poor prognosis treated with BEP/EP are not reported, but the FFS at 1 year was 49% (95% CI, 42% to 55%), with no differences between the arms. It is on the background of these results that we report the results of CBOP/BEP. Caution is required in interpreting the results of single-center phase II trials because a number of biases are possible and regimens that show promise in the phase II setting may not be shown to be better than standard treatment in randomized trials. It is possible that the results are due to treatment in a specialist center, where the staff are familiar with the extra complexity required when using intensive regimens compared with standard once every three weeks chemotherapy. There are data developed in the BEP/EP versus BOP/VIP-B trial 22 showing that patients treated in larger centers (defined as entering five or more patients onto the study) had better survival as a result of enhanced response rates, fewer relapses, fewer toxic deaths, and greater use of postchemotherapy surgery. We cannot exclude this as a possibility, but it should be noted that the data used for the international prognostic factor were mainly from the larger international centers involved in research in the treatment of testicular cancer. There was also no difference in outcome among the three centers contributing to this study. To respond to this issue, the CBOP/BEP regimen has been tested in a multiinstitutional phase II trial under the auspices of the EORTC (Trial 30948), the results of which are pending. A limited degree of dose intensification has been tested in the MRC/EORTC phase III trial discussed above, with no significant advantage for the more dose-intense arm. There are differences between the schedule used in this trial and the CBOP/BEP schedule, not least the use of weekly induction chemotherapy (compared with 10-day chemotherapy in the MRC/EORTC trial) and the use of prolonged bleomycin infusions. This means that during the first 6 weeks, CBOP/BEP is significantly more dose-intense (Table 3). The effect of these differences may be significant; a 2-year progression-free survival rate of 63% was reported with BOP/VIP in the initial phase II work, which was similar to the FFS rate reported in the phase III trial of 53% (95% CI, 47% to 61%). 22,23 In comparison, the results of CBOP/BEP (3-year RFS rate of 83.2%) are superior. Several other approaches have shown promise in phase II trials (Table 4). Alternating chemotherapy administered at 2-week intervals has been tested on a group of patients and has resulted in a 75% 3-year survival rate in the poor-prognosis group. 2 Dose intensification by the use of high-dose chemotherapy either as consolidation or cyclical chemotherapy has been attempted. An initial phase III trial of high-dose chemotherapy and autologous bone marrow transplantation after standard chemotherapy versus standard-dose chemotherapy showed worse results in the high-dose arm. 4 This trial could be criticized for the nonstandard nature of the conventional-dose chemotherapy and the high mortality rate attributed to the high-dose Table 4. Comparison of Reported Series of Chemotherapy Schedules Used for Advanced or Poor-Prognosis Germ Cell Tumors Regimen Type of Trial Definition of Poor Prognosis Assessment Criteria Outcome (%) Reference Various IGCCCG analysis IGCCCG 5-year PFS 41 IGCCCG 1 BEP/EP Phase III MRC 1-year FFS 60 Kaye et al, VIP Phase III Indiana advanced 2-year FFS 64 Nichols et al, POMB/ACE Phase II IGCCCG 3-year overall 75 Bower et al, survival High-dose VIP Phase II Indiana 2-year PFS 75 Bokemeyer et al, PEBV with high dose PBC Phase III Institut Gustave Roussy mathematical model 2-year overall survival 60 Chevreau et al, CBOP/BEP Phase II IGCCCG 3-year PFS 83 Current series Abbreviations: BEP/EP, bleomycin, etoposide, and cisplatin, plus etoposide and cisplatin; VIP, etoposide, ifosfamide, and cisplatin; POMB/ACE, cisplatin, vincristine, methotrexate, bleomycin, dactinomycin, cyclophosphamide, and etoposide; PEBV, cisplatin, etoposide, bleomycin, and vincristine; PEC, cisplatin, etoposide, and carboplatin; CBOP/BEP, carboplatin, bleomycin, vincristine, and cisplatin, followed by bleomycin, etoposide, and cisplatin; IGCCCG, International Germ Cell Cancer Collaborative Group; MRC, Medical Research Council; PFS, progression-free survival; FFS, failure-free survival.

7 CBOP/BEP FOR POOR-PROGNOSIS GERM CELL TUMORS chemotherapy, but the approach is currently being retested using BEP with or without high-dose chemotherapy. An additional high-dose trial reported by Bokemeyer et al 24 reported a 2-year progression-free survival rate of 75%, which was higher than for matched patients treated with standard chemotherapy. In an ongoing EORTC phase III high-dose trial (EORTC Trial 30974), cyclical high-dose VIP chemotherapy with granulocyte colonystimulating factor and stem-cell support is being tested. The cost of this approach is significant bone marrow suppression and extended hospital inpatient stay during treatment. Compared with these reported results, the results with CBOP/ BEP regimen are favorable. Like other more intensive regimens, CBOP/BEP is also more toxic, with a significant toxic death rate resulting from greater risk of bone marrow suppression than standard BEP chemotherapy. The increased dose of cisplatin and the use of vincristine also increase the risk of neuropathy and 877 infertility after treatment. Bleomycin lung toxicity is known to increase with total bleomycin doses of greater than 300,000 IU 25 ; the CBOP/BEP regimen has a cumulative dose of 345,000 IU, although in this series, the number of patients with lung toxicity has been acceptably low, perhaps owing to the infusional nature of the bleomycin. These negative aspects should be balanced against improvements in efficacy; however, CBOP/BEP is inappropriate for patients with better prognosis. We suggest that these results support the need to test CBOP/BEP in a phase III setting to provide the information on the precise position of this balance. The role of intensive induction chemotherapy for patients with poor-prognosis metastatic germ cell tumors has yet to be finally established. Our results with CBOP/BEP compare favorably with the reported results of alternative regimens used in the management of poor-prognosis NSGCT. We suggest a phase III trial to establish the role of the CBOP/BEP regimen. 1. International Germ Cell Cancer Collaborative Group: International Germ Cell Consensus Classification: A prognostic factor-based staging system for metastatic germ cell cancers. J Clin Oncol 15: , Bower M, Newlands ES, Holden L, et al: Treatment of men with metastatic nonseminomatous germ cell tumors with cyclical POMB/ACE chemotherapy. Ann Oncol 8: , Nichols CR, Catalano PJ, Crawford ED, et al: Randomized comparison of cisplatin and etoposide and either bleomycin or ifosfamide in treatment of advanced disseminated germ cell tumors: An Eastern Cooperative Oncology Group, Southwest Oncology Group, and Cancer and Leukemia Group B Study. J Clin Oncol 16: , Chevreau C, Droz JP, Pico JL, et al: Early intensified chemotherapy with autologous bone marrow transplantation in first line treatment of poor risk nonseminomatous germ cell tumors: Preliminary results of a French randomized trial. Eur Urol 23: , Malaise EP, Chavaudra N, Charbit A: Relationship between the growth rate of human metastases, survival and pathological type. Eur J Cancer 10: , Price P, Hogan SJ, Horwich A: The growth rate of metastatic nonseminomatous germ cell testicular tumors measured by marker production doubling time: I. Theoretical basis and practical application (erratum appears in Eur J Cancer 26:925, 1990). Eur J Cancer 26: , Fossa SD, Pettersen EO, Thorud E, et al: DNA flow cytometry in human testicular cancer. Cancer Lett 28:55-60, Silvestrini R, Costa A, Pilotti S, et al: Cell kinetics in human germ call tumors of the testis, in Pavone-Macaluso M, Smith PH, Bagshaw MA (eds): Testicular Cancer and Other Tumors of the Genitourinary Tract. New York, NY, Plenum, 1985, pp Rabes HM: Proliferation of human testicular tumors. Int J Androl 10: , Withers HR, Taylor JM, Maciejewski B: The hazard of accelerated tumor clonogen repopulation during radiotherapy. Acta Oncol 27: , Dearnaley DP: Intensive induction chemotherapy for poor risk tumors, in Horwich A (ed): Testicular Cancer, Investigation and Management (ed 2). London, United Kingdom, Chapman & Hall, 1996, pp Horwich A, Brada M, Nicholls J, et al: Intensive induction chemotherapy for poor risk nonseminomatous germ cell tumors. Eur J Cancer Clin Oncol 25: , Samuels ML, Holoye PY, Johnson DE: Bleomycin combination chemotherapy in the management of testicular neoplasia. Cancer 36: , 1975 REFERENCES 14. Gerson R, Tellez BE, Lazaro LM, et al: Low toxicity with continuous infusion of high-dose bleomycin in poor prognostic testicular cancer. Am J Clin Oncol 16: , Horwich A, Dearnaley DP, Norman A, et al: Accelerated chemotherapy for poor prognosis germ cell tumors. Eur J Cancer 30A: , Mead GM, Stenning SP, Parkinson MC, et al: The Second Medical Research Council study of prognostic factors in nonseminomatous germ cell tumors: Medical Research Council Testicular Tumor Working Party. J Clin Oncol 10:85-94, Peckham MJ: Investigations and staging: General aspects and staging classification, in Peckham MJ (ed): The Management of Testicular Tumours, Edward Arnold, London, 1981, p Calvert AH, Newell DR, Gumbrell LA, et al: Carboplatin dosage: Prospective evaluation of a simple formula based on renal function. J Clin Oncol 7: , Dearnaley DP, Nicholl J, Saunders M: BEP chemotherapy for metastatic teratoma: Royal Marsden experience Br J Cancer 58:526, 1988 (abstr) 20. Wettlaufer JN, Feiner AS, Robinson WA: Vincristine, cisplatin and bleomycin with surgery in the management of advanced metastatic nonseminomatous testis tumors. Cancer 53: , Murray N, Coppin C, Swenerton K: Weekly high intensity cisplatin etoposide (HIPE) for far advanced germ cell cancers (GCC). Proc Am Soc Clin Oncol 6:101, 1987 (abstr 394) 22. Kaye SB, Mead GM, Fossa S, et al: Intensive induction-sequential chemotherapy with BOP/VIP-B compared with treatment with BEP/EP for poor-prognosis metastatic nonseminomatous germ cell tumor: A Randomized Medical Research Council/European Organization for Research and Treatment of Cancer study. J Clin Oncol 16: , Lewis CR, Fossa SD, Mead G, et al: BOP/VIP A new platinumintensive chemotherapy regimen for poor prognosis germ cell tumors. Ann Oncol 2: , Bokemeyer C, Kollmannsberger C, Meisner C, et al: First-line high-dose chemotherapy compared with standard-dose PEB/VIP chemotherapy in patients with advanced germ cell tumors: A multivariate and matched-pair analysis. J Clin Oncol 17: , O Sullivan JM, Huddart RA, Norman AR, et al: Predicting the risk of bleomycin lung toxicity in patients with germ cell tumors. Ann Oncol 14:91-96, 2003