High-dose chemotherapy (HD-CT) is able to circumvent platinum-resistance of resistant/refractory germ-cell tumors (GCTs), but expectancy of cure remains low. New strategies are needed with new drugs and a sequential approach.

Patients with relapsed (but not absolutely refractory to Cisplatinum-based chemotherapy) poor-prognosis GCTs are scheduled to receive 2 cycles combining epirubicin and paclitaxel followed by 3 consecutive HD-CT supported by stem cell transplantation. One course will combine Taxol, 360 mg/m² + thiotepa, 720 mg/m², followed by two ICE regimens (Ifosfamide, 12 g/m², carboplatin, AUC 20, etoposide, 1500 mg/m²).

This phase II study is designed as a Gehan method. The main objective of the study is the complete response rate. With this aim in view, it is planned to enroll in its first step 14 patients to insure that if no complete response (CR) is noticed, study would be stopped for inefficacy (i.e., a CR rate lower than 20%). If one or more CR are noticed, protocol specified that up to 45 patients will be included in order to reduce the confidence interval (CI) of the CR rate. Secondary objectives are the overall response rate (RR), the overall survival (OS) and the progression-free survival (PFS) rates, toxicity and toxic death rate. The statistical analysis is done in terms of intent-to-treat.

The treatment is designed for relapsed poor prognosis patients with testicular or extra-gonadal GCTs previously treated with cisplatin-containing regimens.

Poor prognosis patients are defined as either relapsed/refractory patients more than one month after cisplatin administration, whether in the course of first-line CT or in that of salvage treatment, or patients who showed evidence of progression after at least 2 lines of cisplatin-containing CT (i.e., BEP and VeIP).

Eligibility requirements includes the following criteria: age >18 years and < 65, performance status < 3, histologically or biologically documented GCTs, testicular, abdominal or mediastinal tumors, measurable or evaluable disease, life expectancy > 3 months, normal cardiac, liver, and renal function tests, absence of infection, HIV negative test, and signed informed consent. All patients had to have been previously treated with at least one line of a cisplatin-containing regimen and were included if they were refractory after one or two line(s) of cisplatin-based CT, or had relapsed after two lines of a cisplatin-based CT.

Non-inclusion criteria are: patients with 'BEYER' score > 3, growing teratoma syndrome, possibility of being treated with a conventional cisplatin-based CT, and previous HD-CT regimen.

During the initial evaluation: each patient must have a clinical evaluation that includes measurements of tumor marker levels, and an imaging work up. The tumor marker levels are determined every week during all the sequence of treatment (normal level < 10 ng/ml for alpha-foetoprotein and < 2 mU/mL for HCG). The tumor mass is measured after the first two cycles of induction/mobilization therapy and at the end of the treatment.

Chemotherapy, patients are scheduled to receive 2 courses of front-line mobilization CT followed by 3 HD-CT supported by PBSCT. The front-line treatment consists in a combination of epirubicin, (100 mg/m² in a 30-minute infusion), and paclitaxel (Taxol, 250 mg/m² given in a 3-hour infusion), administered both on days 1 and 14. These 2 cycles are supported by filgrastim (Neupogen), 5 µg/kg, twice a day from days 2 and 15, respectively, until apheresis performed on days 10-13 and 24-27. Apheresis is stopped when at least 9 x 106 CD34+ cells/Kg of BW are obtained for the 3 grafts. A third cycle is permitted if the number of CD34+ cells is not achieved after the first 2 cycles, provided the patient was responsive to CT, or for any reason decided upon by the investigator.

The first HD-CT regimen consists in an association of thiotepa, 720 mg/m² and taxol, 360 mg/m², both administered in a continuous infusion over 3 days (D34 to D36). The first pack of CD34+ cells is infused on day 39. The second and the third courses (ICE) scheduled on days 62-66 and 90-94 respectively, consist in a combination of etoposide (150 mg/m² twice a day, in a 2-hour infusion, for 5 days), ifosfamide (2 400 mg/m²/d in a 3-hour infusion, for 5 days) supported by sodium mercaptoethanesulfonate (mesna, in a 30-minute infusion every 3 hours, during a 12-hour period, initiated at the same time as the ifosfamide infusion), and carboplatin (AUC 4/d, in a 6-hour infusion, for 5 days). Infusion of PBSCs is planned on days 71 and 99. During the 3 high-dose therapies, G-CSF is administered at a daily dose of 5µg/kg, from the day of PBSC infusion until PMN recovery (i.e., PMN > 1.5 x 109/L). Each of these ICE regimens is delayed if the PMN level is less than 1.5 x 109/L and/or the platelet level less than 100 x 109/L.

Toxicity and response to therapy are evaluated according to the ECOG and WHO criteria. The duration of response is calculated from the date of documented response to the date of progression. The duration of PFS and OS are calculated from the date of inclusion to the date of progression, or death if no progression (PFS), and the date of death (OS), according to Kaplan-Meier's method. Survival curves are established according to the classification proposed by BEYER et al. Progression, death from treatment and withdrawal from protocol for whatever reason are considered as treatment failures. Whenever possible, patients in clinical partial response (PR) with normal tumor markers (PRm-) will be proposed for surgery of residual masses at the end of the whole procedure. Sequential procedures are proposed in the case of multiple metastatic sites. If surgery is complete and the pathological examination does not show any viable tumor, patients will be considered as complete responders. If surgery is complete and the pathological examination showed persistent viable tumor, they will be considered as surgical complete response.

Inclusion Criteria:

Eligibility requirements includes the following criteria:

• Age >18 years and < 65
• Performance status < 3
• Histologically or biologically documented GCTs
• Testicular, abdominal, or mediastinal tumors
• Measurable or evaluable disease
• Life expectancy > 3 months
• Normal cardiac, liver, and renal function tests
• Absence of infection
• HIV negative test
• Signed informed consent
• All patients had to have been previously treated with at least one line of a cisplatin-containing regimen and were included if they were refractory after one or two line(s) of cisplatin-based CT, or had relapsed after two lines of a cisplatin-based CT

Exclusion Criteria:

• Fireproof diseases (progress unless month with regard to the last cycle of chemotherapy or in the course of chemotherapy)
• Relapses after complete answer obtained by surgery ( sCR )
• Neuropathy of superior rank or = II - renal Function (Office) superior creatinine or = 125 mmol/l and/or clearance of the creatinine subordinate or = II 60ml / mn
• Antecedents of congestive even compensated cardiac insufficiency
• Hurts of growing teratoma that is measurable hurts increasing by size (cutting) in the absence of rise of marker pens
• Extensive chemotherapy with support of haematopoietic stem cells. NB: A previous preventive irradiation under diaphragmatitis for a seminoma stage I (dose from 24 to 30 Gy in classic spreading) does not establish one against formal indication. However, an estimation clarifies capacities of the haematopoietic marrow is recommended with observation of the evolution of the NFP in the course of chemotherapy and quantification of cells CD 34 + in the peripheral blood. It's the same of the case where a chemotherapy by carboplatine was realized