Radical cystectomy remains the gold standard treatment for invasive non metastatic transitional cell cancer (TCC) of the bladder. In contemporary series, specific survival rates are about 60 to 65% at 5 years, decreasing for locally advanced disease to 45-50% in patients with nonorgan-confined lymph-node negative tumours and to 30-35% in patients with lymph node positive tumours. Perioperative chemotherapy (adjuvant ou neoadjuvant) has been developed in order to improve these results. Thanks to randomized trials and meta-analysis, it can be concluded that perioperative chemotherapy increases overall survival with an absolute benefit of 5%, equating to a survival rate of 50% at 5 years for nonorgan-confined tumours. However, the chemotherapy administration time and the optimal chemotherapy regimen to be delivered are not yet determined. Meta-analyses have shown that the benefit is only observed for chemotherapy regimens including cisplatin. In daily management 4 to 6 cycles of gemcitabine and cisplatin are delivered since this combination has been shown to yield a similar efficacy with a better tolerance as compared to the MVAC regimen (methotrexate, vinblastine, doxorubicin and cisplatin) in the metastatic setting. As HD-MVAC has been shown to be associated with higher response rates than MVAC in bladder metastatic disease, also a better efficacy of HD-MVAC can be suspected in the perioperative setting. Investigators therefore designed a randomized phase III study to compare the efficacy of GC and HD-MVAC in term of progression-free survival in patients for whom chemotherapy has been decided, before or after radical cystectomy. Secondary endpoints include overall survival, side effects, response rate in the neoadjuvant setting and ancillary studies focusing on gemcitabine and cisplatin sensitivity. The total number of patients projected is 500. The number of patients is based on the median progression-free survival rate of 50% at 3 years observed in patients treated with GC (standard arm A) in the perioperative setting. An absolute improvement of 10% (HR=0.74) is expected with HD-MVAC (experimental arm B) with a=0.05 and b=0.20. An interim analysis is planned after the occurrence of 174 events. With an estimated uniform accrual rate of 140 patients per year for 3.5 years and exponential survival, the final analysis is expected to occur 8 years after the start of the trial.

Inclusion Criteria:

• Primary tumour of the bladder
• Histologically confirmed infiltrating urothelial carcinoma (epidermoid and/or glandular variants are accepted if combined with TCC)
• Disease defined by a T2, T3 or T4a N0 (lymph node £ 10 mm on CT scan) M0 stadification for patients receiving neoadjuvant chemotherapy OR pT3 or pT4 OR pN+ whatever pT and M0 for patients receiving adjuvant chemotherapy
• 18 ≤ age ≤ 80 years
• General condition 0 or 1 as per the WHO scale
• Absence of previous chemotherapy for muscle-invasive disease
• Haematological function: Haemoglobin > 11 g/dl, neutrophils ≥ 1500/mm3, platelets ≥ 100,000/mm3
• Liver function: Grade* 0 ASAT and ALAT, grade* 0 alkaline phosphatases, normal bilirubin
• Renal function: calculated (or measured) creatinine clearance ³ 40 ml/min - --- Patients covered by a social security scheme
• Patient having read the information sheet and signed the informed consent form.

Exclusion Criteria:

• Pure adenocarcinoma or pure epidermoid carcinoma or mixed or pure small-cell neuroendocrine carcinoma
• Ventricular ejection fraction < 50%
• History of cancer in the 5 years prior to entry in the trial other than basal cell skin cancer or in situ epithelioma of the cervix
• Male or female patients not agreeing to use an effective method of contraception throughout the duration of treatment and for 6 months after treatment discontinuation
• Pregnant women, or female subjects liable to become pregnant or currently breast-feeding,
• Patient already included in another therapeutic trial on an investigational medicinal product,
• Persons deprived of their freedom or under judicial protection (including guardianship)
• Unable to receive medical follow-up during the trial owing to geographical, social or psychological reasons.