Tweet
I am posting some results and the discussion from the paper below the abstract for more details.
Eur Urol. 2007 Mar;51(3):739-46; discussion 747-8. Epub 2006 Oct 27.
Failure of high-dose cyclophosphamide and etoposide combined with double-dose cisplatin and bone marrow support in patients with high-volume metastatic nonseminomatous germ-cell tumours: mature results of a randomised trial.Droz JP, et. al., Genito-Urinary Group of the French Federation of Cancer Centers (GETUG).
Centre Leon Berard, Lyon, France.
OBJECTIVES: To assess the impact on survival of high-dose chemotherapy with haematopoietic support in patients with high-volume, metastatic nonseminomatous germ cell tumours. METHODS: One hundred fifteen patients were randomised to receive either four cycles every 21 d of vinblastine (0.2 mg/kg on day 1), etoposide (100 mg/m2/d on days 1 through 5), cisplatin (40 mg/m2/d on days 1 through 5), and bleomycin (30 mg on days 1, 8, and 15) (arm A), or a slightly modified regimen followed by a high-dose chemotherapy including etoposide (350 mg/m2/d on days 1 through 5), cisplatin (40 mg/m2/d on days 1 through 5), and cyclophosphamide (1600 mg/m2/d on days 2 through 5) (arm B). RESULTS: In an intent-to-treat analysis, there were 32 (56%) and 24 (42%) complete responses in arms A and B, respectively (p=0.099). After a median follow-up of 9.7 yr, 31 and 27 patients have continuously shown no evidence of disease in arms A and B, respectively. There was no significant difference between the overall survival curves (p=0.167). According to the International Germ Cell Cancer Collaborative Group prognostic classification, the 5-yr survival rates were 88% and 82% in the intermediate group and 69% and 44% in the poor group (p=0.045) in arms A and B, respectively. CONCLUSIONS: This trial failed to demonstrate an impact on response and survival of high-dose chemotherapy with haematopoietic support in first-line treatment of patients with high-volume, metastatic nonseminomatous germ cell tumours.
Relapses and survival
After a median follow-up of 9.7 yr (4.5–12.1), 31 and 27 patients have continuously shown no evidence of disease in arms A and B, respectively. Patients failed from early death (0 and 6), toxic death (4 and 2), early incomplete response (6 and 9), relapse (10 and 10), intercurrent disease (1 and 0), death from progressive disease (4 and 3), and death from unknown cause (1 and 0), respectively. After failure of first-line treatment, seven and six patients, respectively, were successfully salvaged by further treatment. Four patients had salvage surgery. High-dose chemotherapy with haematopoietic stem cell support was given in three and two patients in arms A and B, respectively.
Sixteen and 23 patients died in arms A and B, respectively. There was no significant difference between the two arms as stratified on primary site (p = 0.167); however, there was a trend in favour of arm A with 5-yr survival rates of 75% and 61% in arms A and B, respectively. Fourteen and 18 patients are still alive after 10 yr of follow-up in arms A and B, respectively. Overall survival rates at 5 yr according to the IGCCCG prognostic classification were estimated at 100%, 85%, and 58% in the G, I, and P groups, respectively. None of the six patients in the good prognostic group died as compared with 6 of 33 (18%) in the I group and 31 of 71 (44%) in the P group. The 5-yr survival rates were 88% and 82% in the intermediate IGCCCG group (Fig. 1, p = ns), and 69% and 44% in the poor IGCCCG group (Fig. 2, p = 0.045) in the PVeBV and PVBV+PEC arms, respectively.
Discussion
The present trial was designed to assess the impact on survival of high-dose chemotherapy with bone marrow support in patients with high-volume metastatic NSGCT. It was based on the comparison of two treatment strategies that were developed in the mid-80s. The long-term results clearly confirm the preliminary observations on failure of the initial hypothesis [9] and [14]. The background of high-dose treatment (arm B) was (1) arguments for a dose-effect relationship with cyclophosphamide in first-line treatment [15] and with etoposide with or without cyclophosphamide in the salvage treatment [16] and [17]; (2) feasibility of the regimen in first-line treatment [5]; and (3) theoretical background for drug resistance acquisition on the basis of Norton and Simon hypothesis [18], which favours secondary introduction of drugs acting by different mechanisms. However, the retrospective analysis of trials during the last 15 yr has shown that the cornerstone study was published in 1987 by Williams et al. [19]. It demonstrated that the combination of bleomycin, etoposide, and cisplatin (BEP), four cycles, was superior and less toxic than the combination of PVB. From then on no prospective trial has shown any advantage to four cycles of BEP. In 1988, the PVeBV regimen had shown superiority over the PVB regimen in a randomised trial [4]. However this regimen introduced two differences as compared with PVB: the introduction of etoposide and the use of double-dose cisplatin. The role of double-dose cisplatin, which was specifically tested in a phase 3 randomised trial, was eventually more toxic, and induced similar response rates and survival [20]. Therefore the addition of etoposide was clearly the major new drug in germ cell tumour chemotherapy.
The failure of the high-dose chemotherapy regimen could be related to suboptimal cisplatin and etoposide total doses and/or dose intensities. Indeed these variables have been shown to impact the results of chemotherapy in the whole population of germ cell cancer patients [21] and good-risk patient populations [22] and [23]. A threshold of 33 mg/m2/wk for cisplatin [21] and [24] and approximately 167 mg/m2/wk for etoposide [22] and [23] seems mandatory to maintain an optimal efficacy. Actual administered dose intensities of cisplatin and etoposide were 57.4 mg/m2/wk and 145 mg/m2/wk in arm A, and 46.0 mg/m2/wk and 183 mg/m2/wk in arm B, respectively. Therefore, there was only a 13% mean reduction of etoposide dose intensities in arm A as compared with recommended values, which seems unlikely to influence the results of the trial. Regarding total doses of drugs, which may be less important than dose intensities as suggested by experimental models [25], the only background is the reference to four cycles of BEP (i.e., 400 mg/m2 for cisplatin and 2000 mg/m2 for etoposide). Actual administered total doses of cisplatin and etoposide were 692 mg/m2 and 1747 mg/m2 in arm A, and 511 mg/m2 and 2204 mg/m2 in arm B, respectively. Once again the 13% reduction of total dose of etoposide is unlikely to impact the results of the trial. Finally, the differences concerning the delivery of bleomycin in the two arms are minor and roughly similar to the total doses of bleomycin delivered in the BEP regimen.
The impact of high total doses delivered during one cycle of treatment and supported by haematopoietic support is based on the experience in refractory patients. Indeed previous results suggested that doses of etoposide up to 3 g/m2 [17] as well as high-dose etoposide plus cyclophosphamide [16] might induce response in cisplatin-refractory patients. In addition high-dose cyclophosphamide (6 g/m2) had induced promising responses in non-pretreated patients [15]. In this setting, the demonstration of the efficacy of carboplatin [26] and the ability to increase the doses without major extrahaematopoietic-limiting toxicities led us to study this drug in combination with etoposide [27], etoposide and cyclophosphamide [28] and [29], and etoposide and ifosfamide [30] and [31]. Tandem high-dose regimens with etoposide and carboplatin after two cycles of VAB-6 (vinblastine, cyclophophamide, dactinomycin, bleomycin, and cisplatin) or etoposide, cyclophoshamide, and carboplatin (CEC regimen) after two cycles of VIP (etoposide, ifosfamide, and cisplatin) have been studied in the first-line management of poor-risk germ cell tumours so far [32] and [33]. These studies demonstrated the feasibility and efficacy (50% response rates) of this approach. However the results of a US randomised trial comparing standard four cycles of BEP to two cycles of BEP + two cycles of high-dose CEC with haematopoietic support in intermediate- and poor-risk groups of the IGCCCG classification failed to demonstrate any survival advantage in the high-dose chemotherapy group [34].
During the 90s the German group developed an effective sequential cisplatin-based high-dose chemotherapy trial [35]. Bokemeyer [36] performed a match-paired analysis of patients included in this trial and others included in Indiana University trials of conventional chemotherapy. A benefit of high-dose first-line treatment was suggested with a 20% improvement in the 3-yr survival rate. A randomised trial comparing four cycles of standard BEP and one cycle of BEP followed by three, sequential, intensified cycles of VIP with haematopoietic support is ongoing. As the dose intensities delivered are clearly higher than those used in the current study, a different outcome might be observed.
Conclusions
On the basis of results reported in randomised trials so far, there is currently no role for high-dose chemotherapy with haematopoietic support in the daily front-line management of patients with metastatic NSGCT [37]. The standard treatment for intermediate- and poor-risk patients remains four cycles of BEP as recommended by the European Germ Cell Cancer Consensus Group [38]. The role of newer cytotoxic drugs, such as gemcitabine, paclitaxel, and oxaliplatin, is currently under evaluation in an international randomised trial taking into account the prognostic value of serum tumour marker decline after the first cycle of chemotherapy [39]. Other agents, such as temozolamide, or targeted therapies warrant further investigations.
Eur Urol. 2007 Mar;51(3):739-46; discussion 747-8. Epub 2006 Oct 27.
Failure of high-dose cyclophosphamide and etoposide combined with double-dose cisplatin and bone marrow support in patients with high-volume metastatic nonseminomatous germ-cell tumours: mature results of a randomised trial.Droz JP, et. al., Genito-Urinary Group of the French Federation of Cancer Centers (GETUG).
Centre Leon Berard, Lyon, France.
OBJECTIVES: To assess the impact on survival of high-dose chemotherapy with haematopoietic support in patients with high-volume, metastatic nonseminomatous germ cell tumours. METHODS: One hundred fifteen patients were randomised to receive either four cycles every 21 d of vinblastine (0.2 mg/kg on day 1), etoposide (100 mg/m2/d on days 1 through 5), cisplatin (40 mg/m2/d on days 1 through 5), and bleomycin (30 mg on days 1, 8, and 15) (arm A), or a slightly modified regimen followed by a high-dose chemotherapy including etoposide (350 mg/m2/d on days 1 through 5), cisplatin (40 mg/m2/d on days 1 through 5), and cyclophosphamide (1600 mg/m2/d on days 2 through 5) (arm B). RESULTS: In an intent-to-treat analysis, there were 32 (56%) and 24 (42%) complete responses in arms A and B, respectively (p=0.099). After a median follow-up of 9.7 yr, 31 and 27 patients have continuously shown no evidence of disease in arms A and B, respectively. There was no significant difference between the overall survival curves (p=0.167). According to the International Germ Cell Cancer Collaborative Group prognostic classification, the 5-yr survival rates were 88% and 82% in the intermediate group and 69% and 44% in the poor group (p=0.045) in arms A and B, respectively. CONCLUSIONS: This trial failed to demonstrate an impact on response and survival of high-dose chemotherapy with haematopoietic support in first-line treatment of patients with high-volume, metastatic nonseminomatous germ cell tumours.
Relapses and survival
After a median follow-up of 9.7 yr (4.5–12.1), 31 and 27 patients have continuously shown no evidence of disease in arms A and B, respectively. Patients failed from early death (0 and 6), toxic death (4 and 2), early incomplete response (6 and 9), relapse (10 and 10), intercurrent disease (1 and 0), death from progressive disease (4 and 3), and death from unknown cause (1 and 0), respectively. After failure of first-line treatment, seven and six patients, respectively, were successfully salvaged by further treatment. Four patients had salvage surgery. High-dose chemotherapy with haematopoietic stem cell support was given in three and two patients in arms A and B, respectively.
Sixteen and 23 patients died in arms A and B, respectively. There was no significant difference between the two arms as stratified on primary site (p = 0.167); however, there was a trend in favour of arm A with 5-yr survival rates of 75% and 61% in arms A and B, respectively. Fourteen and 18 patients are still alive after 10 yr of follow-up in arms A and B, respectively. Overall survival rates at 5 yr according to the IGCCCG prognostic classification were estimated at 100%, 85%, and 58% in the G, I, and P groups, respectively. None of the six patients in the good prognostic group died as compared with 6 of 33 (18%) in the I group and 31 of 71 (44%) in the P group. The 5-yr survival rates were 88% and 82% in the intermediate IGCCCG group (Fig. 1, p = ns), and 69% and 44% in the poor IGCCCG group (Fig. 2, p = 0.045) in the PVeBV and PVBV+PEC arms, respectively.
Discussion
The present trial was designed to assess the impact on survival of high-dose chemotherapy with bone marrow support in patients with high-volume metastatic NSGCT. It was based on the comparison of two treatment strategies that were developed in the mid-80s. The long-term results clearly confirm the preliminary observations on failure of the initial hypothesis [9] and [14]. The background of high-dose treatment (arm B) was (1) arguments for a dose-effect relationship with cyclophosphamide in first-line treatment [15] and with etoposide with or without cyclophosphamide in the salvage treatment [16] and [17]; (2) feasibility of the regimen in first-line treatment [5]; and (3) theoretical background for drug resistance acquisition on the basis of Norton and Simon hypothesis [18], which favours secondary introduction of drugs acting by different mechanisms. However, the retrospective analysis of trials during the last 15 yr has shown that the cornerstone study was published in 1987 by Williams et al. [19]. It demonstrated that the combination of bleomycin, etoposide, and cisplatin (BEP), four cycles, was superior and less toxic than the combination of PVB. From then on no prospective trial has shown any advantage to four cycles of BEP. In 1988, the PVeBV regimen had shown superiority over the PVB regimen in a randomised trial [4]. However this regimen introduced two differences as compared with PVB: the introduction of etoposide and the use of double-dose cisplatin. The role of double-dose cisplatin, which was specifically tested in a phase 3 randomised trial, was eventually more toxic, and induced similar response rates and survival [20]. Therefore the addition of etoposide was clearly the major new drug in germ cell tumour chemotherapy.
The failure of the high-dose chemotherapy regimen could be related to suboptimal cisplatin and etoposide total doses and/or dose intensities. Indeed these variables have been shown to impact the results of chemotherapy in the whole population of germ cell cancer patients [21] and good-risk patient populations [22] and [23]. A threshold of 33 mg/m2/wk for cisplatin [21] and [24] and approximately 167 mg/m2/wk for etoposide [22] and [23] seems mandatory to maintain an optimal efficacy. Actual administered dose intensities of cisplatin and etoposide were 57.4 mg/m2/wk and 145 mg/m2/wk in arm A, and 46.0 mg/m2/wk and 183 mg/m2/wk in arm B, respectively. Therefore, there was only a 13% mean reduction of etoposide dose intensities in arm A as compared with recommended values, which seems unlikely to influence the results of the trial. Regarding total doses of drugs, which may be less important than dose intensities as suggested by experimental models [25], the only background is the reference to four cycles of BEP (i.e., 400 mg/m2 for cisplatin and 2000 mg/m2 for etoposide). Actual administered total doses of cisplatin and etoposide were 692 mg/m2 and 1747 mg/m2 in arm A, and 511 mg/m2 and 2204 mg/m2 in arm B, respectively. Once again the 13% reduction of total dose of etoposide is unlikely to impact the results of the trial. Finally, the differences concerning the delivery of bleomycin in the two arms are minor and roughly similar to the total doses of bleomycin delivered in the BEP regimen.
The impact of high total doses delivered during one cycle of treatment and supported by haematopoietic support is based on the experience in refractory patients. Indeed previous results suggested that doses of etoposide up to 3 g/m2 [17] as well as high-dose etoposide plus cyclophosphamide [16] might induce response in cisplatin-refractory patients. In addition high-dose cyclophosphamide (6 g/m2) had induced promising responses in non-pretreated patients [15]. In this setting, the demonstration of the efficacy of carboplatin [26] and the ability to increase the doses without major extrahaematopoietic-limiting toxicities led us to study this drug in combination with etoposide [27], etoposide and cyclophosphamide [28] and [29], and etoposide and ifosfamide [30] and [31]. Tandem high-dose regimens with etoposide and carboplatin after two cycles of VAB-6 (vinblastine, cyclophophamide, dactinomycin, bleomycin, and cisplatin) or etoposide, cyclophoshamide, and carboplatin (CEC regimen) after two cycles of VIP (etoposide, ifosfamide, and cisplatin) have been studied in the first-line management of poor-risk germ cell tumours so far [32] and [33]. These studies demonstrated the feasibility and efficacy (50% response rates) of this approach. However the results of a US randomised trial comparing standard four cycles of BEP to two cycles of BEP + two cycles of high-dose CEC with haematopoietic support in intermediate- and poor-risk groups of the IGCCCG classification failed to demonstrate any survival advantage in the high-dose chemotherapy group [34].
During the 90s the German group developed an effective sequential cisplatin-based high-dose chemotherapy trial [35]. Bokemeyer [36] performed a match-paired analysis of patients included in this trial and others included in Indiana University trials of conventional chemotherapy. A benefit of high-dose first-line treatment was suggested with a 20% improvement in the 3-yr survival rate. A randomised trial comparing four cycles of standard BEP and one cycle of BEP followed by three, sequential, intensified cycles of VIP with haematopoietic support is ongoing. As the dose intensities delivered are clearly higher than those used in the current study, a different outcome might be observed.
Conclusions
On the basis of results reported in randomised trials so far, there is currently no role for high-dose chemotherapy with haematopoietic support in the daily front-line management of patients with metastatic NSGCT [37]. The standard treatment for intermediate- and poor-risk patients remains four cycles of BEP as recommended by the European Germ Cell Cancer Consensus Group [38]. The role of newer cytotoxic drugs, such as gemcitabine, paclitaxel, and oxaliplatin, is currently under evaluation in an international randomised trial taking into account the prognostic value of serum tumour marker decline after the first cycle of chemotherapy [39]. Other agents, such as temozolamide, or targeted therapies warrant further investigations.
