SEARCH

SEARCH BY CITATION

Keywords:

  • germ cell tumors;
  • BEP (bleomycin, etoposide, cisplatin);
  • VIP (etoposide, ifosfamide, cisplatin);
  • International Germ Cell Cancer Collaborative Group;
  • toxicity

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

Various staging systems have been proposed for disseminated germ cell neoplasms. The Indiana University staging system was based on clinical and radiographic findings only, whereas the newly created International Germ Cell Cancer Collaborative Group (IGCCCG) staging system also utilized serum markers as a prognostic factor. This study updated the intergroup trial that compared the standard therapy of bleomycin, etoposide, and cisplatin (BEP) with etoposide, ifosfamide, and cisplatin (VIP) in advanced germ cell tumors and reanalyzed the results using the IGCCCG staging system.

METHODS

From October 1987 to April 1992, 304 patients with advanced-stage germ cell tumors (using the Indiana University staging system) were randomized to receive four cycles of BEP or VIP. Two hundred and eighty-six patients were eligible and fully evaluable. With a median follow-up of 7.3 years, 283 of the 286 evaluable patients from the Eastern Cooperative Oncology Group protocol, E3887, were reclassified using the IGCCCG staging system. Progression-free survival (PFS), overall survival (OS), and toxicity were assessed for the treatment arms.

RESULTS

With a longer follow-up of 7.3 years and using the Indiana University staging system, the PFS rates were 64% versus 58% and the OS rates were 69% versus 67% in the VIP and BEP arms, respectively. For patients reclassified with the IGCCCG staging system, the PFS rates were 81%, 72%, and 54% and the OS rates were 89%, 81%, and 60% for good, intermediate, and poor-risk patients, respectively. Differences in OS (VIP, 62%; BEP, 57%) and PFS (VIP, 56%; BEP, 49%) for the subset of patients reclassified as poor risk by the IGCCCG staging system were not significantly different. More toxicity, primarily hematologic toxicity, occurred on the VIP arm.

CONCLUSIONS

With a median follow-up of 7.3 years and with a reclassification based on the IGCCCG, OS and PFS rates were comparable between BEP and VIP. Toxicity, primarily hematologic, was modestly greater with the ifosfamide-containing arm. The VIP regimen may be considered a treatment alternative for patients with underlying pulmonary disease. In most patients with poor and intermediate-risk germ cell tumors, four cycles of BEP remain the standard therapy. Cancer 2003;97:1869–75. © 2003 American Cancer Society.

DOI 10.1002/cncr.11271

The evolution of chemotherapy in germ cell tumors is well documented.1 Current therapy cures the majority of patients. Certain clinical factors (e.g., markedly elevated tumor markers, nonpulmonary visceral metastases, and primary mediastinal nonseminomatous germ cell tumors) identify patients less likely to be cured with current therapy. Utilizing these observations, researchers in the United States and Europe devised various schemes in an attempt to clarify the prognosis and to determine the most appropriate therapy for individual germ cell tumor patients.2–6 Two of the more commonly used staging systems were developed at Indiana University and Memorial Sloan Kettering Cancer Center (MSKCC).1–3 The former staging system utilized clinical and radiographic findings, whereas the MSKCC system used an equation that also incorporated serum markers. The additional prognostic information from serum markers was believed to be sufficiently important to incorporate into prognostic clinical trials.

In 1997, the International Germ Cell Cancer Collaborative Group (IGCCCG) published a new classification system based on data collected from more than 5000 germ cell patients treated from 1975 to 1990 (Table 1).7 This system classifies patients with disseminated germ cell tumors into good, intermediate, or poor-risk groups. Historically, the Indiana classification identified 27–37% of all patients with disseminated disease as poor risk, whereas the International Staging System identified only 16% of patients as poor risk. This smaller number suggests a more select group who may be candidates for more intense therapy. This retrospective analysis was more precise in allocating patients with variable risks to various induction therapies.

Table 1. International Germ Cell Consensus Classification
NonseminomaSeminoma
  1. AFP: α-fetoprotein; HCG: human chorionic gonadotropin; LDH: lactate dehydrogenase; ULN: upper limit of normal.

  2. Data from International Germ Cell Consensus Classification ÷ A prognostic factor-based staging system for metastatic germ cell cancers. International Germ Cell Cancer Collaborative Group. J Clin Oncol. 1997;-15:594–603.7

Good prognosis 
 Testis/retroperitoneal primary, no nonpulmonary visceral metastases, and all of the following markers: AFP (< 1000 ng/mL), HCG (< 5,000 IU/L), and LDH (< 1.5 × ULN)Any primary site, no nonpulmonary visceral metastases, and normal AFP, any HCG, any LDH
Intermediate prognosis 
 Testis/retroperitoneal primary, no nonpulmonary visceral metastases, and any of the following markers: AFP (≥ 1000 and ≤ 10,000 ng/mL), HCG (≥ 5000 and ≤ 50,000 IU/L), or LDH (1.5 and ≤ 10 × ULN)Any primary site, nonpulmonary visceral metastases, and normal AFP, any HCG, any LDH
Poor prognosis 
 Mediastinal primary, nonpulmonary visceral metastases, or any of the following markers: AFP (> 10,000 ng/mL), HCG (> 50,000 IU/L), or LDH (> 10 × ULN)No patients were classified as poor risk

The last completed intergroup trial (E3887) in advanced germ cell tumors evaluated cisplatin, etoposide, and either bleomycin (BEP) or ifosfamide(VIP) and is the largest prospective trial conducted to date in this patient population.8 The E3887 trial used the Indiana University classification system (Table 2) to select patients with advanced or poor-risk disease. The progression-free survival (PFS) rates (VIP, 64%; BEP, 60%) and overall survival (OS) rates (VIP, 74%; BEP, 71%) at 2 years were not statistically significantly different. This article updates the 1998 intergroup trial8 that compared BEP and VIP and reanalyzes patients using the IGCCCG classification system to verify its utility in patients with advanced disease.

Table 2. Indiana Staging System: Advanced Disease
  1. Data from Birch R, Williams S, Cone A, et al. Prognostic factors for favorable outcome in disseminated germ cell tumors. J Clin Oncol. 1986;-4:400–407.3

Primary mediastinal nonseminomatous germ cell tumor, advanced pulmonary disease > 10 pulmonary metastases per lung field, or multiple pulmonary metastases > 3 cm
Palpable abdominal mass plus pulmonary metastases
Hepatic, osseous, or central nervous system metastases

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

From October 1987 to April 1992, 304 patients with advanced-stage disseminated germ cell tumors by the Indiana staging system (Table 2) were enrolled on E3887. The trial was coordinated by the Eastern Cooperative Oncology Group (ECOG) with participation by the Southwest Oncology Group and the Cancer and Leukemia Group B. All patients, regardless of primary site, histology, age, or performance status, were eligible provided the aforementioned staging criteria were met. Patients previously treated with chemotherapy or who had elevated levels of serum creatinine (unless due to obstructive uropathy) were ineligible.

Staging evaluation for all patients included chest, abdominal, and pelvic computed tomographic scans and measurement of beta subunit human chorionic gonadotropin (HCG), α-fetoprotein (AFP), and lactate dehydrogenase. Patients were assigned randomly to the standard arm, BEP, or the investigational arm, VIP. The accrual goal of 300 patients was to ensure 140 assessable patients per treatment arm to provide a 90% probability of detecting a 20% difference in the complete remission (CR) rate.

The BEP regimen and dose modifications have been published.8 Postchemotherapy surgery was performed approximately 4–6 weeks after the last course of chemotherapy in patients with normal serologic markers and disease that was amenable to complete surgical excision. After December 1991, granulocyte–colony-stimulating factor (G-CSF) became available and was incorporated as primary prophylaxis in both treatment arms (34 patients).

As part of this update and reanalysis, the primary site, histology, presence of nonpulmonary visceral metastases, and serum tumor markers (obtained on review of treatment enrollment forms and study flowsheets) were used to reclassify patients as good, intermediate, or poor prognosis patients according to the IGCCCG classification. Cases in which the histology was reported as pure seminoma, but tumor markers indicated nonseminomatous components (i.e., any elevation of AFP or marked elevation of HCG postorchiectomy [>1000 IU/L]), were reclassified using the nonseminomatous criteria.

The PFS and OS rates were classified according to the original Indiana Staging System, and then by treatment arm for each prognostic group. They were updated based on follow-up reported to the ECOG coordinating center.

Statistical Consideration

The exact midrank Wilcoxon test for ordered categorical outcomes was used to compare the distributions of toxicity grades between treatment arms and the exact Jonckheere–Terpstra test was used to compare toxicity and international stage. Survival curves were estimated by the Kaplan–Meier method and time-to-event comparisons were performed using the log rank test.9, 10 All reported P values are two-sided. SAS software version 6.12 (SAS Institute, Cary NC) was used for all computations.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Of the 286 evaluable patients with advanced disease using the Indiana Staging System, 283 patients had sufficient data to be reclassified using the IGCCCG staging system. Of the three patients who had insufficient data, one was excluded for not meeting the Indiana advanced criteria. The remaining two patients did not have tumor markers available before treatment and otherwise did not meet the International Staging System poor prognosis group on the basis of histology, mediastinal primary site, or nonpulmonary visceral metastases.

Using the IGCCCG classification, 37 patients (13.1%) were reclassified as good prognosis, 65 patients (23.0%) as intermediate prognosis, and 181 patients (63.9%) as poor prognosis. With a median follow-up of 7.3 years at the time of reanalysis (range, 0.6–11 + years), the PFS of 64% versus 58% and the OS of 69% versus 67% were observed for the VIP and BEP-treated patients, respectively. This compares favorably to the previously reported 2-year PFS (64% vs. 60%) and 2-year OS rate (74% vs. 71%). Four patients had late disease recurrences (beyond 2 years) and there were three additional deaths (two patients, unknown cause; one patient, acute leukemia). For all patients, the 5-year PFS rates were 81% for good prognosis patients, 72% for intermediate prognosis patients, and 54% for poor prognosis patients (Fig. 1). The 5-year OS rates were 89% for good prognosis patients, 81% for intermediate prognosis patients, and 60% for poor prognosis patients (Fig. 2).

thumbnail image

Figure 1. Progression-free survival by International Germ Cell Cancer Collaborative Group stage regardless of treatment arm. The number of events and the number of patients at risk for each time period are shown below the graph.

Download figure to PowerPoint

thumbnail image

Figure 2. Overall survival by International Germ Cell Cancer Collaborative Group stage regardless of treatment arm. The number of events and the number of patients at risk in each time period are shown below the graph.

Download figure to PowerPoint

Of the 37 good prognosis patients using the IGCCCG classification, 24 were treated with BEP and 13 were treated with VIP. Twenty-six of the 65 intermediate prognosis patients were treated with BEP and the remaining 39 were treated with VIP. Of the 181 poor prognosis patients, 89 were treated with BEP and 92 were treated with VIP. Table 3 lists the 5-year PFS and OS rates by prognosis and treatment arm. None of the differences in the treatment arms for OS or PFS reached statistical significance. Figures 3 and 4 show the PFS and OS curves, respectively, by treatment arm for the 181 patients reclassified as poor prognosis by the IGCCCG classification.

Table 3. 5-Year Survivals by Treatment Arma
 Overall survival (%)Progression-free survival (%)
BEPVIPBEPVIP
  • BEP: bleomycin, etoposide, and cisplatin; VIP: etoposide, ifosfamide, and cisplatin.

  • a

    None of the P values between arms were statistically significant.

Good prognosis (n = 37)88927592
Intermediate prognosis (n = 65)84777372
Poor prognosis (n = 181)57624956
thumbnail image

Figure 3. Progression-free survival by treatment arm for patients reclassified as having poor prognosis by the International Germ Cell Cancer Collaborative Group staging system. The number of events and the number of patients at risk for each time period for both treatment arms are shown below the graph.

Download figure to PowerPoint

thumbnail image

Figure 4. Overall survival by treatment stage for patients reclassified as having poor prognosis by the International Germ Cell Cancer Collaborative Group staging system. The number of events and the number of patients at risk in each time period for both treatment arms are shown below the graph.

Download figure to PowerPoint

Seven of the 286 (2.8%) evaluable patients have had disease recurrence or have died 2 or more years posttherapy (range, 2.2–8.5 years). One patient had disease recurrence 3.9 years post-BEP therapy and is now currently alive without evidence of disease at 10.2 years after resection of mediastinal teratoma. Another patient experienced disease recurrence (BEP) at 3.9 years and is alive at 10.2 years, but the nature of the disease recurrence and subsequent therapy are not available. The remaining five patients with late events have died. One BEP patient died 3.8 years posttherapy and 1 VIP patient died 8.3 years posttherapy without details about the cause of death or disease status. One VIP patient had disease recurrence at 4.5 years with an elevated AFP level and received four courses of cisplatin-based chemotherapy and subsequent high-dose chemotherapy with stem cell rescue, but died in the peritransplant period. One patient (BEP) had disease recurrence at 8.5 years and died at 9.5 years without details of treatment. The last late disease recurrence was treated with BEP for a primary mediastinal tumor. In this case, acute myelogenous leukemia (AML) developed at 3.1 years. Cytogenetics revealed Klinefelter sydrome and trisomy 21 mosaicism. He died 14 months after the diagnosis of AML without evidence of germ cell tumor recurrence.

Toxicity results are presented in Table 4. On the BEP arm, 75%, 73%, and 79% of good, intermediate, and poor-risk patients, respectively, had Grade 3–5 toxicity. On the VIP arm, 85%, 93%, and 93% of good, intermediate, and poor-risk patients, respectively, had Grade 3–5 toxicity. Hematologic toxicity was the most common toxicity, regardless of treatment arm. Grade 3–5 toxicity occurred in 63%, 73%, and 76% of good, intermediate, and poor-risk patients, respectively, on the BEP arm and in 85%, 93%, and 90% of good, intermediate, and poor prognosis patients, respectively, on the VIP arm.

Table 4. Toxicity of Poor-Risk Patients (IGCCCG)
ToxicityTreatment arm
BEP (n = 89) Toxicity grade (%)aVIP (n = 92) Toxicity grade (%)a
345345
  • IGCCCG: International Germ Cell Cancer Collaborative Group; BEP: bleomycin, etoposide, and cisplatin; VIP: etoposide, ifosfamide, and cisplatin.

  • a

    Worse degree.

  • b

    Aggregate Grade 3–5 hematologic toxicity between BEP and VIP was significant (P = 0.0003). All other comparison P values are not significant.

Nausea/vomiting7.90012.000
Infection2.31.22.34.403.3
Bleeding00001.11.1
Neurologic4.51.104.45.50
Respiratory1.31.33.84.62.30
Genitourinary1.1004.400
Hepatic4.7003.600
Hematologicb39.337.1028.362.00

Comparison of toxicity for poor-risk patients by the IGCCCG classification between the BEP and VIP arms did show more differences in overall Grade 3–5 toxicity (79% vs. 93%; P = 0.0002) and more Grade 3–5 hematologic toxicity (76% vs. 90%; P = 0.003). However, this increased hematologic toxicity did not translate into a statistically greater incidence of infection (P = 0.08).

Three possible drug-related deaths occurred in good or intermediate-risk patients. One good-risk patient on the VIP arm died of sepsis after Cycle 1, and 2 intermediate-risk patients on the BEP arm died, 1 of sepsis after Cycle 1 and 1 of pulmonary hemorrhage after Cycle 1.

Of the poor-risk patients, three on the VIP arm and two on the BEP arm died of sepsis during therapy. One patient on the VIP arm died of intracranial bleeding during recovery of his complete blood counts at which time the platelet count was 76,000/mm3 with the suspicion that progressive brain metastases contributed to his death. Two patients on the BEP arm died of pulmonary complications. One of the deaths occurred during Cycle 1 after only 30 U of bleomycin. Respiratory difficulty began before the administration of chemotherapy. Therefore, death was most likely a cause of progressive disease. The other pulmonary death was more typical of bleomycin, occurring after a total bleomycin dosage of 300 U. This patient experienced decreasing pulmonary functions and worsening bilateral infiltrates without an infectious etiology.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

This update confirms the durable CR rate even in patients with poor-risk germ cell tumors. All patients in this trial were evaluated with a median follow-up of more than 7 years. This reanalysis compares favorably with the original report.8

The 2-year and 7+ year PFS rates for VIP (64% and 64%, respectively) and BEP (60% and 58%, respectively) remain very similar. Likewise, the 2-year and 7+ year OS rates for VIP (74% vs. 69%) and BEP (71% vs. 67%) are also similar. Given the known propensity for germ cell tumors to recur within 2-years of diagnosis, the similar PFS and OS rates at 2- and 7-years are an expected finding.

Late-recurring disease, defined as occurring 2 or more years after cisplatin-based chemotherapy, occurs in 2–4% of cases.11 In this trial, late disease recurrences were documented clearly in 4 of 286 patients (1.4%). In the three additional patients, two had unknown disease status at death and one died of AML of unknown FAB subtype. This patient had a primary mediastinal nonseminomatous germ cell tumor, which has a known association with Klinefelter syndrome and hematologic neoplasms including AML (acute megakaryocyctic leukemia [M7] is the most common type).12 Hematologic neoplasms were usually diagnosed within a year of the primary mediastinal tumor, but cases have been diagnosed later. The patient completed four cycles of BEP, including a total dose of 2000 mg/m2 etoposide. Etoposide exposure is a risk factor for AML, although it is seen most typically at doses exceeding 2000 mg/m2.13 However, the typical cytogenetic abnormality associated with treatment-related AML due to etoposide exposure (and other topoisomerase II inhibitors) involving 11q23 was not present.14 A trisomy 21 mosaicism could also account for this patient's leukemia.15

The advantages of using the IGCCCG classification were also shown in the current analysis. The reclassification identifies 102 patients (about one-third of the total number) who fell into a different, more favorable, prognostic subset. Many of the advances in the treatment of germ cell tumors have involved amelioration of toxicity for patients with good prognosis.16–19 The improved survival in the advanced disease subgroup provides the rationale to test more aggressive therapies in patients with poor prognosis disease. These changes resulted in the current therapy for disseminated germ cell patients with good and intermediate risk of three cycles of BEP or four cycles of etoposide and cisplatin.19–21 Therefore, the routine use of the International Staging System could spare some patients with disseminated germ cell tumors the added inconvenience and toxicity of more intensive therapy.

It is noteworthy that for the 181 patients (regardless of treatment arm) reclassified as poor risk by the IGCCCG classification, the OS and PFS rates of 60% and 54% are superior to the 48% and 41% previously reported for similarly staged patients in the initial publication by the IGCCCG.7 The data set used by the IGCCCG to define the various prognostic categories included patients treated from 1975 to 1990. Some of the earliest regimens did not contain etoposide, the second most active drug in the therapy for germ cell tumors. The IGCCCG showed improved PFS and OS rates for patients treated in the 1980s and 1990. The improved rates reflect the incorporation of more active regimens in frontline and, subsequently, salvage therapy. Results of modern treatment regimens are probably the most appropriate sources when counseling poor-risk patients.

Although only 16% of patients with disseminated germ cell tumors meet the poor-risk criteria for the IGCCCG classification, they still comprise a heterogeneous group. Kollmannsberger et al.22 analyzed 332 patients with IGCCCG poor-risk disease employing classification and regression-tree modeling. Within this poor-risk subset were patients with 2-year PFS rates ranging from 28% (mediastinal primary tumor with lung metastases) to 75% (gonadal or retroperitoneal primary tumor without visceral metastases) and with 2-year OS rates ranging from 49% (mediastinal primary tumor and visceral metastases) to 84% (gonadal or retroperitoneal primary tumor without visceral metastases). These very poor-risk patients are the most likely nonresponders to standard therapy and, therefore, have the most to gain from investigational therapy. However, these patients comprised only 6–8% of the IGCCCG poor-risk patients, so practical issues of sufficient enrollment of patients in trials do arise.

This reanalysis showed increased Grade 3–5 toxicity for the ifosfamide-containing arm (79% vs. 93%), which reached statistical significance (P = 0.0002). This difference was largely due to Grade 3–5 hematologic toxicity, which occurred in 76% poor-risk BEP-treated patients versus 90% poor-risk VIP-treated patients. Despite these factors, the increase in hematologic toxicity did not result in an increased risk of neutropenic infection or life-threatening bleeding, which are the most clinically relevant outcomes. This trial was conducted before the routine use of G-CSF. Only 34 patients received growth factors after they became available in December 1991. Based on the high incidence of neutropenia with cisplatin-based combination therapy, G-CSF is used routinely at Indiana University to treat poor-risk patients who receive cisplatin-based chemotherapy.

Most patients treated with bleomycin will not suffer from any long-term clinical sequelae. The survival results presented in the current study provide reassurance that outcome would not be compromised with the use of VIP rather than BEP in select individuals. These candidates include patients who develop bleomycin-induced changes during therapy (e.g., basilar rales, inspiratory lag, radiographic changes suggestive of interstitial disease, and changes in pulmonary function) or who have significant underlying pulmonary pathology (a rare finding in a young population).

This reanalysis has several limitations, including its retrospective nature. The reclassification resulted in even fewer poor-risk patients on both arms of the study, which decreases the statistical power to determine differences in therapeutic efficacy and toxicity between the VIP and BEP arms. It is possible that significant clinical differences may exist between these two regimens in poor-risk disease.

The prognosis of patients with poor-risk disease warrants investigation of novel and potentially more toxic regimens. The currently active intergroup study, which randomizes poor and intermediate-risk patients between standard therapy of BEP for four cycles and experimental therapy of BEP for two cycles followed by two cycles of high-dose chemotherapy with autologous peripheral blood stem cell transplant, is one such approach. This trial will determine whether the cost and morbidity of the more aggressive treatment are offset by improved survival rates. If no survival difference is seen, future advances will likely rely on a better understanding of the molecular basis of the disease. Until that time, larger trials utilizing ifosfamide or other new agents may be reasonable.

In summary, this trial confirms the activity of both BEP and VIP in poor-risk germ cell tumor patients with a follow-up of more than 7 years. Current therapy suggests that the prognosis in this patient population is somewhat better than previously reported by the IGCCCG. This reanalysis allows comparison by risk factors to other current or future trials using the IGCCCG classification. These two regimens are therapeutically similar and have comparable PFS and OS rates in the two arms. The VIP regimen is associated with more hematologic toxicity, making it the preferred regimen for most patients. However, in patients with underlying pulmonary disease, VIP is an alternative. Use of the IGCCCG staging system should be required for all future trials to allow proper comparison of results. Patients with poor-risk disease should be encouraged to participate in ongoing clinical trials that explore novel treatment approaches.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Einhorn LH. Treatment of testicular cancer: a new and improved model. J Clin Oncol. 1990; 8: 17771781.
  • 2
    Bosl GJ, Geller NL, Cirrincione C, et al. Multivariate analysis of prognostic variables in patients with metastatic testicular cancer. Cancer Res. 1983; 43: 34033407.
  • 3
    Birch R, Williams S, Cone A, et al. Prognostic factors for favorable outcome in disseminated germ cell tumors. J Clin Oncol. 1986; 4: 400407.
  • 4
    Stoter G, Sylvester R, Sleijfer DT, et al. Multivariate analysis of prognostic factors in patients with disseminated nonseminomatous testicular cancer: results from a European Organization for Research on Treatment of Cancer Multiinstitutional Phase III Study. Cancer Res. 1987; 47: 27142718.
  • 5
    Droz JP, Kramar A, Ghosn M, et al. Prognostic factors in advanced nonseminomatous testicular cancer. A multivariate logistic regression analysis. Cancer. 1988; 62: 564568.
  • 6
    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 Tumour Working Party. J Clin Oncol. 1992; 10: 8594.
  • 7
    International Germ Cell Consensus Classification. A prognostic factor-based staging system for metastatic germ cell cancers. International Germ Cell Cancer Collaborative Group. J Clin Oncol. 1997; 15: 594603.
  • 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. 1998; 16: 12871293.
  • 9
    Kaplan EL, Meier P. Nonparametric estimation of incomplete observations. J Am Stat Assoc. 1958; 53: 457481.
  • 10
    Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966; 50: 163170.
  • 11
    Baniel J, Foster RS, Gonin R, et al. Late relapse of testicular cancer. J Clin Oncol. 1995; 13: 11701176.
  • 12
    Nichols CR, Roth BJ, Heerema N, et al. Hematologic neoplasia associated with primary mediastinal germ-cell tumors. N Engl J Med. 1990; 322: 14251429.
  • 13
    Nichols CR, Breeden ES, Loehrer PJ, et al. Secondary leukemia associated with a conventional dose of etoposide: review of serial germ cell tumor protocols. J Natl Cancer Inst. 1993; 85: 3640.
  • 14
    Bredeson CN, Barnett MJ, Horsman DE, et al. Therapy-related acute myelogenous leukemia associated with 11q23 chromosomal abnormalities and topoisomerase II inhibitors: report of four additional cases and brief commentary. Leuk Lymphoma. 1993; 11: 141145.
  • 15
    Fong CT, Brodeur GM. Down's syndrome and leukemia: epidemiology, genetics, cytogenetics and mechanisms of leukemogenesis. Cancer Genet Cytogenet. 1987; 28: 5576.
  • 16
    Einhorn LH, Williams SD, Troner M, et al. The role of maintenance therapy in disseminated testicular cancer. N Engl J Med. 1981; 305: 727731.
  • 17
    Stoter G, Sleyfer DT, ten Bokkel Huinink WW, et al. High-dose versus low-dose vinblastine in cisplatin-vinblastine-bleomycin combination chemotherapy of non-seminomatous testicular cancer: a randomized study of the EORTC Genitourinary Tract Cancer Cooperative Group. J Clin Oncol. 1986; 4: 11991206.
  • 18
    Williams SD, Birch R, Einhorn LH, et al. Treatment of disseminated germ-cell tumors with cisplatin, bleomycin, and either vinblastine or etoposide. N Engl J Med. 1987; 316: 14351440.
  • 19
    Culine S, Kerbrat P, Bouzy J, et al. Are 3 cycles of bleomycin, etoposide and cisplatin (3bep) or 4 cycles of etoposide and cisplatin (4ep) equivalent regimens for patients (pts) with good-risk metastatic non seminomatous germ cell tumors (NSGCT)? Preliminary results of a randomized trial [abstract]. Proc Am Soc Clin Oncol. 1999; 18: 309a.
  • 20
    Einhorn LH, Williams SD, Loehrer PJ, et al. Evaluation of optimal duration of chemotherapy in favorable-prognosis disseminated germ cell tumors: a Southeastern Cancer Study Group protocol. J Clin Oncol. 1989; 7: 387391.
  • 21
    Loehrer PJ, Johnson D, Elson P, et al. Importance of bleomycin in favorable-prognosis disseminated germ cell tumors: an Eastern Cooperative Oncology Group trial. J Clin Oncol. 1995; 13: 470476.
  • 22
    Kollmannsberger C, Nichols C, Meisner C, et al. Identification of prognostic subgroups among patients with metastatic ‘IGCCCG poor-prognosis’ germ-cell cancer: an explorative analysis using cart modeling. Ann Oncol. 2000; 11: 11151120.