In Europe, pancreatic cancer (PC) is the 7th most common cancer and the 5th leading cause of cancer death in Europe. Each year, the number of deaths due to prostate cancer is almost as high as the number of new cases diagnosed reflecting the poor prognosis associated with this disease. PC is insidious and is often diagnosed late. Despite advances in the management of other more common gastrointestinal cancers, the treatment of PC has had few benefits inherent in recent advances in digestive oncology. Gemcitabine has thus remained the reference treatment for more than 10 years.

Recent studies have shown that gemcitabine/Nab-paclitaxel combination therapy is more effective in PC than gemcitabine-based therapy alone. In addition, multidrug therapy approaches (Irinotecan-5FU/LV) have also emerged to avoid the emergence of resistance to treatments while limiting toxicities. The recently developed Nal-IRI has also shown interesting efficacy in patients with metastatic PC previously treated with gemcitabine, with improved overall survival median and limited toxicity. Based on this information, the NAPOLI trial was conducted in patients with second line PC comparing the efficacy of Nal-IRI/5FU/LV or Nal-IRI and 5FU/LV alone; in this key study, the combination Nal-IRI/5FU/LV treatment was more effective than monotherapies (Nal-IRI or 5FU/LV alone).

Based on all these data, a Phase II trial testing the standard of care gemcitabine/nab-paclitaxel vs Nal-IRI/5FU/LV vs Nal-IRI/5FU/LV 2-months sequential regimen followed by gemcitabine/nab-paclitaxel will be performed. This will allow us to i) know the tolerance and efficacy of Nal-IRI/5FU/LV in the first line of treatment, ii) test a new sequential strategy with Nal-IRI but also iii) compare our results in the experimental arms with one of the two world standard therapeutic regimens: gemcitabine + nab-Paclitaxel. All this in order to improve the management of patients with PC from the first line of treatment.

Pancreatic cancer (PC) is the seventh cause from cancer and the fifth cause from cancer-related death in Europe. Nearly as many deaths occur from PC than the number of new cases diagnosed each year, reflecting the poor prognosis typically associated with this disease. PC is insidious in onset and is often diagnosed late at the stage of metastatic spread. In spite of advances made in the management of other more common gastrointestinal cancers, the treatment of PC did take only a small advantage of recent progresses in gastrointestinal oncology and targeted therapies did not significantly modify its prognosis to date.

Thus, for more than 10 years, gemcitabine has been the standard of care to treat metastatic PC. Recently, two positive phase III trials in metastatic PC were reported.

First, the gemcitabine with nab-paclitaxel combination therapy was compared to gemcitabine alone in 861 randomized patients with metastatic PC. Results showed a significant improvement in response rate (RR), progression free survival (PFS) and overall survival (OS). Second, the results of the PRODIGE 4/ACCORD 11 trial testing the FOLFIRINOX regimen in metastatic PC patients, finally gave a hope showing a major improvement in PFS and OS as compared to gemcitabine. However, though manageable, the safety profiles of FOLFIRINOX and of gemcitabine plus nab-paclitaxel were less favourable than that of gemcitabine. These regimens were associated with a higher incidence of grade 3-4 neutropenia, febrile neutropenia, thrombocytopenia, diarrhoea, and grade 2-3 sensory neuropathy.

Irinotecan has a stronger growth-inhibiting effect than cisplatin, mitomycin C and fluorouracil on cultured pancreatic adenocarcinoma cells. In most trials testing this molecule in PC patients, however, the response rates were low (<10%) and survivals were poor. Intensive regimens with irinotecan and 5-FU have been developed in colorectal cancer patients to increase the anti-tumor effects of this combination therapy. From those, the FOLFIRI.3, in which the irinotecan is administered before and after a 5-FU 46h continuous infusion gave promising results. The investigator has tested this regimen in a phase II trial in advanced PC patients, with an objective RR of 37.5%, a median OS time of 12 months and acceptable tolerability. Considering these encouraging results the investigator has proposed a new approach to improve outcomes for patients with metastatic PC, using FOLFIRI.3 and gemcitabine alone, sequentially, to increase patient survival with a preserved quality of life (QOL). Indeed, some authors have reported that the administration of different patterns of sequential polychemotherapy was independently associated with OS in patients with PC and other gastrointestinal tumors such as metastatic colorectal cancer. Such strategies using drugs without cross-resistance sequentially may increase anti-tumor effects and limit cumulative and non-cumulative toxicities.

The investigator thus performed a randomized multicentre phase II trial to assess this sequential treatment strategy using FOLFIRI.3 and gemcitabine alternately in one arm and gemcitabine alone in the other arm, in patients with metastatic non pre-treated pancreatic adenocarcinoma. In this study the FIRGEM strategy seems to be an effective first line treatment option in good condition patients with metastatic PC. The primary endpoint was reached with a rate of PFS at 6 months of 44.9% in the FIRGEM arm while gemcitabine alone failed (25.7%). This good PFS rate at 6 months was maintained at 12 and 18 months (26.2% and 18% respectively) though median PFS was 5.0 months. Moreover, an impressive objective response rate was observed in the FIRGEM arm (40%) as compared to the gemcitabine arm (11.4%). These results confirm those of the initial phase II trial evaluating the FOLFIRI.3 regimen (37.5% objective response rate)(15) in PC patients and compare favourably with the 31.6% and 23% reported in the two trials evaluating FOLFIRINOX and gemcitabine + nab-paclitaxel, respectively.

Median overall survival was 11 months with FIRGEM versus 8.2 months with gemcitabine (HR:0.710 95% CI: 0.457-1.103). Here again the experimental arm gave good results, with median OS in the range of those reported with FOLFIRINOX (11.1 months). The safety profile of the FIRGEM strategy showed that hematological and GI toxicities were more important than with gemcitabine alone. Interestingly no limiting sensory neuropathy was observed with our treatment schedule and a significant increase in the time to definitive deterioration of the QoL was observed in the FIRGEM group as compared with the gemcitabine group. This effect was observed for all domains.

Considering that nab-paclitaxel improves significantly patients outcome in metastatic PC when combined to gemcitabine, the addition of this drug to the FIRGEM strategy may be of particular interest. And the PRODIGE 37 trial tested this combination in a randomized phase II trial recently closed for inclusions. This trial allowed to fight the cancer with 4 different drugs, without any cross resistance described between them, given sequentially in the first 4 months of treatment. Moreover, the "resting period" without nab-paclitaxel may delay significantly the occurrence of the cumulative neuropathy induced by this molecule and optimised its used with such a "stop and go" like strategy. Results are awaited for early 2018.

More recently a liposomal irinotecan has been developed and tested in pancreatic cancer patients, it comprises irinotecan free base encapsulated in liposome nanoparticles, which keep irinotecan into the circulation sheltered from conversion to its active metabolite (SN-38) longer and use local macrophage-mediated activation, which would increase and prolong intratumoral levels of both irinotecan and SN-38. In a phase II study of 40 patients with MPA previously treated with gemcitabine-based therapy, monotherapy with nal-IRI resulted in a median overall survival of 5.2 months, and a manageable toxicity profile. The NAPOLI-1 phase III trial was then conducted, comparing three arms of chemotherapy in patients previously treated with gemcitabine-based therapy: liposomal irinotecan (MM-398 or nal-IRI) alone or combined with 5FU and folinic acid, and 5FU and folinic acid alone. Combination of nal-IRI and 5FU/LV was more effective than 5FU alone or Nal-IRI alone (median OS of 6.1 vs. 4.2 and 4.9 months respectively (HR: 0.67; p=0.012). Increased hematologic and GI toxicities were also seen in the Nal-IRI arms but were manageable. Nal-IRI plus 5FU and folinic acid extends survival in patients with metastatic pancreatic ductal adenocarcinoma who previously received gemcitabine-based therapy. In a setting where there is a paucity of second line treatment option, this combination is an important emerging treatment option for metastatic adenocarcinoma of the pancreas.

Considering all these data, the investigator proposes to run a randomized phase II trial testing the standard continuous Gemcitabine + Nab-paclitaxel schedule vs Nal-IRI+5FU/LV vs Nal-IRI+5FU/LV for two months followed by Gemcitabine + Nab-paclitaxel for two months before starting again Nal-IRI. This will allow i) to generate efficacy and tolerability data on the Nal-IRI/5FU combination in the first line setting, ii) to test a new sequential strategy with Nal-IRI , in regards of the interesting results obtained in second and third line pancreatic cancer treatment, iii) to control our results in the experimental arms with one of the two first line worldwide standard regimen: Gemcitabine + Nab-paclitaxel.

Inclusion Criteria:

• Histopathologically or cytologically proven pancreatic adenocarcinoma (on primitive or metastatic lesion)
• Metastatic disease at a distance
• At least one measurable lesion according RECIST v1.1 criteria
• 18 ≤ age ≤ 75 years
• Life expectancy >12 weeks
• Performance status WHO < 2
• No prior chemotherapy : adjuvant chemotherapy by gemcitabine +/- capecitabine is allowed if ended at least 12 months before the inclusion and adjuvant or neo-adjuvant FOLRIFINOX chemotherapy is allowed if ended at least 12 months prior the inclusion
• Pain well controlled before the inclusion of the patient
• ANC ≥ 1,500 cells/μL (without the use of hematopoietic growth factors); platelet count ≥ 100,000 cells/μL, hemoglobin ≥ 9 g/dL (blood transfusions is permitted for patients with hemoglobin levels below 9 g/dL)
• Adequate hepatic function as evidenced by: Serum total bilirubin within normal range for the institution (Serum bilirubin ≤ 1,5 UNL) Biliary drainage allowed for biliary obstruction.
• Albumin levels ≥ 3.0 g/dL
• Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) ≤ 2.5 x ULN (≤ 5 x ULN acceptable if liver metastases were present)
• Normal renal function test (creatinine clearance ≥ 50 ml/min)
• Normal ECG or ECG without any clinically significant findings
• Patient able to understand and sign an informed consent
• Females of child-bearing potential are required to test negative for pregnancy at the time of enrollment based on a urine or serum pregnancy test.
• Both male and female patients of reproductive potential were required to agree to use a reliable method of birth control, during the study and for 3 months following the last dose of study drug.
• Patient affiliated to social security
• Regular follow-up possible

Exclusion Criteria:

• Uncontrolled brain or meningeal metastasis, or bone metastasis (no need of systematic CT scan)
• Prior radiation therapy (except if there is at least one measurable target outside irradiation area)
• Clinically significant gastrointestinal disorder including hepatic disorders, bleeding, inflammation, occlusion, or diarrhea > Grade 1
• History of chronic inflammatory bowel disease
• Other types of pancreatic tumours, in particular endocrine or acinar cell tumours
• Ampulloma
• Gilbert's syndrome
• Presence of neuropathy > grade 1 according to NCI-CTC
• History of any second malignancy in the last 5 years; subjects with prior history of in-situ cancer or basal or squamous cell skin cancer are eligible. Subjects with other malignancies are eligible if they had been continuously disease free for at least 5 years.
• Severe arterial thromboembolic events (myocardial infarction, unstable angina pectoris, stroke) less than 6 months before inclusion.
• NYHA Class III or IV congestive heart failure, ventricular arrhythmias or uncontrolled blood pressure.
• Known hypersensitivity to any of the drugs /constituents or non-liposomal irinotecan
• Any other medical or social condition deemed by the investigator to be likely to interfere with a patient's ability to sign informed consent, cooperate and participate in the study, or interferes with the interpretation of the results.
• Use of CYP3A4/UGT1A inducers/inhibitors
• Use of strong CYP2C8 inhibitors or inducers, or presence of any other contraindications for nab-paclitaxel or gemcitabine
• ILD presence
• Pregnant or breast feeding