With an 81.8% observed 1-year progression-free survival and a lower limit of the 95% CI above 70%, this phase II study reached its target demonstrating results of a true 1-year progression-free survival rate of >80%. The 2-year overall survival rates in, respectively, intermediate and poor prognosis patients were 89 and 78%. The C-BOP/BEP schedule was feasible in a multicentre setting with predictable and manageable short-term toxicity without any toxic death.
In patients with metastatic germ cell tumours of unfavourable prognosis, attempts to improve the outcome of the standard regimen BEP as induction chemotherapy have focused on three approaches: firstly, intensification of cisplatin application; secondly, introduction of new drugs together with the sequential use of alternating drug combinations; and thirdly, increasing the cumulative doses of the cytotoxic, drugs including the use of high-dose chemotherapy followed by autologous stem cell support. Not rarely, clinical investigators have combined these strategies.
Intensification did not improve the 1-year failure-free survival rate in the EORTC GU group's phase III study that in poor prognosis patients (Royal Marsden Hospital categorisation system) compared three cycles of BOP (bleomycin, vincristine, cisplatin) given every 10 days followed by three cycles of VIP-B (etoposide, ifosfamide, cisplatin, bleomycin) with four cycles BEP plus two cycles of EP (Kaye et al, 1998
). Concerning the introduction of new agents, four courses of VIP were compared with four cycles of BEP (de Wit et al, 1998b
), but both regimens led to similar 2-year failure-free survival (VIP: 64%; BEP: 60%) and 2-year overall survival (VIP: 74%; BEP: 71%). The greatest experience with the use of alternative drugs has been with the POMB/ACE regimen (cisplatin, vincristine, methotrexate,bleomycin, actinomycin-D, cyclophosphamide, etoposide) with two institutions treating 339 patients with MGCT over a 20-year period with a median follow-up of 8 years (Bower et al, 1997
). In total, 92 patients were retrospectively identified as IGCCC poor prognosis. The POMB-EPI regimen, essentially consisting of the components of POMB alternating with a modified VIP regimen, resulted in a 2-year overall survival of 64% in 22 patients within the poor prognosis group (Germa-Lluch et al, 1999
). Taxanes in combination with cisplatin-based chemotherapy have also been tested. In a phase I/II study with BEP combined with Taxol, all 13 evaluable patients with intermediate or poor prognosis MGCT achieved a complete response and none of these patients relapsed with a median follow-up of 18 months (de Wit et al, 1999a
). Currently the EORTC GU group conducts a phase III study comparing four cycles of T-BEP with four cycles of BEP in patients with intermediate prognosis features. Fizazi et al (2002a)
reported a 5-year overall survival of 88% in the intermediate prognosis group and 83% (95% CI: 58–100%) in the poor prognosis group using the CISCA/VB regimen (cisplatin, cyclophosphamide, adriamycin, vinblastine, bleomycin). The same group described a phase II study using the BOP-CISCA-POMB-ACE regimen comprising eight drugs plus granulocyte colony-stimulating factor (Fizazi et al, 2002b
). Investigators of the Genitourinary Group of the French Federation of Cancer Centers have embarked on a prospective trial of BEP vs
CISCA (II)/VB (IV) in poor-risk patients (Culine et al, 1997
). The third approach implies the attempts to use high-dose combination chemotherapy with autologous haematopoetic stem cell support. In several phase II trials, this strategy has resulted in promising survival rates with acceptable toxicity without toxic death, for example, given as CEC (cisplatin, etoposide, cyclophosphamide) after induction with BEP (Decatris et al, 2000
), or as sequential high-dose VIP plus paclitaxel (Hartmann et al, 2002
). The German Testicular Cancer Group reported a 79% 2-year survival in 182 poor prognosis patients (IGCCCG criteria) who were treated with escalating doses of VIP followed by autologous stem cell support (Schmoll et al, 2003
). In poor prognosis patients, high-dose schedules using CEC or VIP together with autologous blood stem cell transplantation are currently investigated in prospective randomised trials by the US Intergroup (BEP vs
high-dose CEC) and the EORTC GU Group (BEP vs
The C-BOP/BEP schedule used in this study principally comprises two manoeuvers to improve the efficacy of induction chemotherapy during the first 6 weeks: an increase of the cumulative cisplatin dose and frequent cycling of the cisplatin, vincristine and bleomycin in an effort to overcome rapid proliferation, together with alternating the C-BOP schedule with BEP. For example, in the C-BOP schedule, the dose intensity of platinum during the first 6 weeks is 63
(based on a carboplatin AUC3 being equivalent to cisplatin 50
) compared to 50
in the BOP/VIP schedule and 33.3
in the standard BEP. The results of the present study are promising. When C-BOP/BEP is given in a multicentre setting, the complete response rates and the 2-year overall survival rates for, respectively, patients with intermediate and poor prognosis MGCT is at least comparable with the results obtained with other intensive cytotoxic regimens (). Our 2-year overall survival in the poor prognosis group may seem slightly inferior to Christian et al
's figure of 88% achieved at three large cancer centres. This may be due to differences between poor prognosis patients with MGCT as to the type and number of poor risk factors (Decatris et al, 2000
). Interstudy differences in outcome may reflect differences in the prognostic features or patient selection rather than real differences in treatment efficacy. Christian et al
's slightly superior survival figure can also be viewed as a consequence of broader experience with MGCT patients in general at each of the three involved cancer centres as compared to the experience of institutions with few cases of MGCT (Aass et al, 1991
; Collette et al, 1999
The C-BOP/BEP schedule has been shown to be safe in a multicentre design, with no toxic deaths in the current study. Based on nonrandomised historical studies (International Germ Cell Cancer Collaborative Group, 1997
), its superior efficacy compared to BEP cannot be excluded. During the first weeks of the induction chemotherapy, the proportion of myelotoxic drugs in C-BOP is less than in the BEP schedule. This may be advantageous in patients who are very sick due to a large tumour burden at the time of diagnosis and who are at particularly high risk of neutropenic fever and sepsis. Nevertheless, the occurrence of ‘any grade 3/4 toxicity' in 90% of the patients and of grade 3/4 haematological toxicity in 77% warrants that these patients should be managed at institutions with experience in the treatment of MGCT.
Until the ongoing three phase III studies comparing four cycles of BEP with experimental chemotherapy regimens are finalised, BEP remains the standard treatment of MGCT, even in patients with intermediate and poor prognosis features. The current and published results with c-BOP/BEP support ongoing plans to conduct a multicentre phase III study comparing C-BOP with four cycles of BEP in poor prognosis patients.
The following investigators and centres contributed patients to the study:
SD Fossa, Norwegian Radium Hospital, Oslo, Norway (21); B Paluchowska, Maria Sklodowska – Curie Memorial Cancer Center, Warsaw, Poland (16); A Horwich, Royal Marsden Hospital, Sutton, UK (13); G Kaiser, Klinikum Nürnberg, Nürnberg, Germany (8); P de Mulder, Sint Radboud University Hospital, Nijmegen, The Netherlands (4); O Koriakine, Medical Radiological Research Center, Obninsk, Russia (formerly: Cancer Research Center, Moskow, Russia) (3); ATM Van Oosterom, Universitair Ziekenhuis Gasthuisberg, Leuven, Belgium (3); C Sternberg, San Camillo and Forlanini Hospitals (formerly: San Raffaele Hospital), Roma, Italy (2); G Mead, Royal South Hants Hospitaln Southampton, UK (2); JB Vermorken, Universitair Ziekenhuis Antwerpen, Edegem, Belgium (1); JJ Croles, Bosh Medicentrum, s'Hertogenbosch, The Netherlands (1); R de Wit, Erasmus Medical Center, Rotterdam, The Netherlands (1); HJ Keizer, University Hospital, Leiden, The Netherlands (1); A Bono, Ospedale di Circolo e Fundacione Macchi, Varese, Italy (1); M Williams, Addenbrooke National Health Service, Cambridge, UK (1); RE Coleman, Weston Park Hospital, Sheffield, UK (1).
L de Prijck was the data manager at the EORTC Data Centre, responsible for this trial. L Collette was the statistician at the Data Centre.