Pancreatic ductal adenocarcinoma (PDAC) exhibits significant heterogeneity, making the optimal choice of chemotherapy challenging. While targeted therapies benefit from companion biomarkers, few tools exist to guide the selection of cytotoxic chemotherapy. Transcriptomic signatures now allow for the prediction of sensitivity to cytotoxic agents. Several molecular classifications (such as basal-like, classical, etc.) have been established and correlated with prognosis, but they are rarely used in clinical practice. The PaCaOmics program has developed robust predictive signatures, grouped under the name Pancreas-View Signature, capable of analyzing FFPE samples using minimal material.

Locally advanced or borderline resectable pancreatic cancer (BR-PDAC) accounts for approximately 20% of cases. Neoadjuvant chemotherapy (NAC) has become the standard of care, improving R0 resection rates and overall survival. The two main chemotherapy regimens used are mFOLFIRINOX and GEM/NAB-paclitaxel, which show comparable efficacy and toxicity profiles. However, no clear consensus exists on the superiority of one over the other. Therefore, predictive biomarkers are crucial to help select the most appropriate neoadjuvant regimen, avoid unnecessary toxicities, and maximize the chances of curative surgery. The NEOPREDICT trial aims to evaluate the efficacy of treating patients with borderline resectable PDAC identified with a GEM+ sensitivity transcriptomic signature with GEMCITABINE + NAB-PACLITAXEL regimen compared to standard of care mFOLFIRINOX as NAC.

Inclusion Criteria:

• Borderline resectable pancreatic ductal adenocarcinoma (BR-PDAC) defined by National Comprehensive Cancer Network (NCCN) criteria v2.2025 on contrast-enhanced CT-scan (including non-contrast acquisition, pancreatic phase and portal venous phase), with stage confirmed by a local pancreatic expert review board including at least medical oncologist/onco-gastroenterologist, pancreatic surgeon and pancreatic expert radiologist. No central review is required,
• Measurable pancreatic lesion according to RECIST 1.1 (CT scan or MRI < 28 days),
• WHO PS 1 or 0,
• Histologically proven pancreas ductal adenocarcinoma,
• Available FFPE from pancreatic tumor sample with > 10% of tumor cells (assessed by a local expert pancreatic pathologist),
• No prior chemotherapy or radiation for pancreatic cancer (except one cycle of mFOLFIRINOX during waiting time for GEM + signature) or resection of pancreatic cancer,
• Age ≥18 years old and ≤ 80 years old if geriatric standardized evaluation validates the study chemotherapy regimen administration for patients between 75-80,
• Written informed consent obtained from patient before any protocol related intervention,
• Acceptation and ability to conform to the protocol requirement during all the duration of the investigation including treatment, scheduled visits, clinical and biological examinations and follow up,
• Adequate organ function, as defined by the following:
• AST and ALT < 3.5 x upper limit of normal (ULN),
• Total serum bilirubin < 3 x ULN, (for patients with total serum bilirubin between 1.5 and 3 x ULN, the dose of irinotecan will be adjusted in accordance with the SmPC).
• Serum albumin >30 g/L,
• Hemoglobin >9.0 g/dl,
• Absolute neutrophil count (ANC) >1.5 G/L,
• Platelets >100 G/L,
• Creatinine clearance > 50 mL/min (according to CKD-EPI),
• Normal Kalemia, calcemia, magnesemia
• Women of childbearing potential must agree to use contraception during treatment and for at least 15 months after discontinuation of the experimental treatments. Men who have sexual relationship with women of childbearing potential must agree to use contraception during treatment and for at least 12 months after discontinuation of the experimental treatments.
• Patient affiliated to French social security

Exclusion Criteria:

• Strictly resectable or locally advanced PDAC according to NCCN criteria,
• Distant metastases (including inter aortic lymph nodes),
• Any condition that contraindicates the use of IRINOTECAN, OXALIPLATIN, 5FU, GEMCITABINE or NAB-PACLITAXEL,
• Partial or complete DPD deficiency (uracilemia ≥ 16 ng/mL),
• Any progressive pathology not stabilized over the past 6 months: liver impairment, renal impairment, respiratory or cardiac failure.
• Other concomitant cancer or a history of cancer during the previous 3 years, except for localized cancer in situ, basal or squamous cell skin cancer adequately treated,
• Other interventional clinical trial except for non-interventional trial (ie not modifying 1-year EFS),
• QT/QTc (Fredericia Correction) interval > 450 msec for men and > 460 msec for women,
• Pregnant or breastfeeding woman,
• Known Gilbert's syndrome or known homozygosity for UGAT1A1*28 polymorphism,
• Treatment with millepertuis,
• Uncompensated asthma,
• Potentially severe infection < 7 days,
• Inflammatory bowel disease and/or intestinal obstruction,
• Known severe allergy to contrast dye (for CT or MRI) without possible substitution,
• Hypersensitivity to the active substance or to one of the excipients of one of the study treatments,
• Treatment with brivudine within 4 weeks prior to the administration of protocol treatment,
• Concomitant treatment with a strong inhibitor (i.e. ketoconazole) or inducer (i.e. rifampicin, carbamazepine, phenobarbital, phenytoin, apalutamide) of cytochrome P450 3A4 or 2C8 (CYP3A4 or CYP2C8),
• Patient who has received a live attenuated vaccine (against yellow fever, chickenpox, shingles, measles, mumps, rubella, tuberculosis, rotavirus) in the 6 weeks prior to randomization,
• Patient with sensitive peripheral neuropathy with functional discomfort,
• Impossibility of undergoing medical monitoring during the trial for geographical, social, or psychological reasons,
• Patient who is under judicial protection (patient who is legally institutionalized or under guardianship or curatorship) or not able to give consent. (as refered at art. art. L. 1121-6, art. L. 1121-8, art. L. 1121-8-1 du Code de la Santé Publique)