Bile duct tumours are rare. They are the 6th most common type of digestive cancer. Their therapeutic management is complex and must be multidisciplinary in nature. Most of the time, an endoscopic or radiological biliary drainage is necessary before any tumour treatment.

Their prognosis is poor due to the fact that they are normally diagnosed late, which makes curative surgery impossible. A population study in the Côte d'Or region of France reported a survival rate at 5 years of approximately 10%.

For the locally advanced or metastatic forms, treatment has not been properly codified. With respect to chemotherapy, prospective studies, most often phase II, are difficult to interpret due to a limited number of patients and due to the heterogeneity of this type of tumour (bile duct and pancreas tumours). Treatment with 5FU alone provides an objective response in approximately 10% of cases. In combination with mitomycin or carboplatin, the objective response rate is 20%, with a median survival period of 5 months. Interferon combined with 5FU has a better response rate (30%), but occurrences of different types of toxicity are more frequent.

More recently, gemcitabine and the 5FU-cisplatin combinations demonstrated objective tumour control in 50% of patients with a median survival period of 10 months. Gemcitabine combined with oxiplatin or with cisplatin has shown the same response rate but a median survival period of approximately 12 months.

The benefit of this combination has been confirmed in a phase III trial that compared the gemcitabine-cisplatin combination to gemcitabine alone, in 410 patients with locally advanced unresectable and/or metastatic bile duct cancer. The results were in favour of the combined treatment with a median survival period of 11.7 months (versus 8.1 months - HR 0.64 [0.52 - 0.80]). This combination is currently the reference first-line treatment.

At the same time as these results, triple therapies involving 5FU + oxiplatin + irinotecan have objectively shown a significant increase in overall survival of patients with metastatic pancreatic adenocarcinoma compared to gemcitabine alone (median of 11.1 months versus 6.8 months, HR = 0.57 [0.45 - 0.73] p < 0.0001). The response rate and progression-free survival (PFS) have also been improved with these triple therapies; the response rates were 31.6% versus 9.4% p < 0.001 and the median PFS 6.4 months versus 3.3 months p < 0.001, respectively.

The adverse events observed with the triple therapy occurred more frequently, for febrile neutropaenia (5.4%), with the need to treat with growth factors (G-CSF for 42.5% of patients). Haematological and digestive toxicity was also higher: grade 3-4 neutropaenia was observed for 45.7% of patients in the FOLFIRINOX arm and 18.7% of patients in the gemcitabine arm (p = 0.0001); vomiting was noted for 14.5% of patients in the FOLFIRINOX arm and 4.7% in the gemcitabine arm (p = 0.002). Quality of life was improved in the FOLFIRINOX arm.

Due to the histological, therapeutic and prognostic similarities between pancreatic and bile duct cancer, it is interesting to assess this triple therapy compared to the current reference treatment in bile duct cancers: gemcitabine combined with cisplatin (GEMCIS). Due to the known higher levels of toxicity for this triple therapy (digestive and haematological), the investigators modified the conventional FOLFIRINOX regimen (mFOLFIRINOX) by removing the 5-FU bolus at D1 of each cycle. This modification to the regimen would appear not to decrease the efficacy of the treatment.

Inclusion Criteria:

• - WHO 0 or 1
• Age ≥ 18 years
• Tumour of the intrahepatic or extrahepatic (and/or hilar) bile ducts, or of the gallbladder
• Measurable abdominal metastases (at least a lesion >10 mm) and/or measurable, unresectable primary tumour
• Disease proven by histopathology or cytology (on metastasis or primary tumour)
• If there are no abdominal metastases, the unresectability must be confirmed by a hepatobiliary surgeon in a multidisciplinary team (MDT) meeting
• Bilirubin <1.5 N (after endoscopic or trance hepatic optimum biliary drainage, if necessary), AST and ALT <10N
• Serum creatinine <130 µmol/L, creatinine clearance >60 mL/min
• Neutrophils ≥ 1500/mm3 and platelets ≥ 75,000/mm3
• Prothrombin index > 70%
• Serum albumin > 25 g/L
• Patient registered with a social security scheme (including CMU)
• Signed informed consent form

Exclusion Criteria:

• - Non-measurable metastases and primary tumour
• Ampullary carcinoma or cancer of the pancreas with infiltration of the bile ducts or mixed tumours (hepatocholangiocarcinoma)
• Chemotherapy and/or radiotherapy within the last 4 months
• Other malignant tumour except in situ basal cell carcinoma or curatively treated carcinoma of the uterine cervix or other malignant tumour that has been treated and has been considered cured for at least 5 years
• Major comorbidity factors (unstable angina, myocardial infarction that has occurred within the last 6 months, heart failure ≥2 according to the NYHA classification, uncontrolled high blood pressure)
• Woman who is pregnant or breastfeeding, or patient of either sex who is of childbearing age and not using an adequate contraceptive method
• Not able to undergo the trial medical follow-up for geographical, social or psychological reasons.