At the time of diagnosis, 25% of patients with colorectal cancer present with liver metastasis (CRLM). Among patients with localized colorectal cancer (Stages I-III), 50% to 70% will develop liver metastases during the course of their disease. Surgery in combination with intravenous (IV) chemotherapy represents the only chance of cure for selected patients by removing all liver metastases and treat residual microscopic disease by postoperative chemotherapy for 3 months. However, up to two-thirds of patients will experience a relapse, with about two-thirds of recurrences occurring in the liver.

Hepatic arterial infusion (HAI) chemotherapy has been proposed to improve the efficacy of chemotherapy by increasing the concentration of the drug in the liver. This treatment is currently administered by infusion through a specific catheter placed in the artery feeding the liver parenchyma, connected to a subcutaneous port-a-cath system. Several trials have shown that the administration of floxuridine or oxaliplatin via HAI combined with IV chemotherapy achieves a higher response rate compared to IV chemotherapy alone in patients with unresectable colorectal liver metastases. HAI chemotherapy has thus become an attractive therapeutic option for patients who underwent curative-intent surgery to reduce the risk of hepatic recurrence.

The investigators recently demonstrated in the PACHA-01 phase II randomized study a 47% decrease of hepatic recurrence risk by HAI of oxaliplatin compared to IV chemotherapy alone, despite a higher but manageable toxicity among 99 patients who underwent curative surgery considered at high risk of recurrence. Moreover, this study showed promising results in terms of time to recurrence and survival. Moreover, feasibility has improved in recent years with the development of non-invasive techniques for HAI.

The investigators propose to conduct the PACHA-02 trial to evaluate the efficacy in terms of disease-free survival of oxaliplatin administered via HAI in combination with IV chemotherapy after curative resection in patients with colorectal cancer at high risk of recurrence. A total of 272 patients who will undergo curative surgery for at least 4 CRLM with no residual disease on imaging performed within 4 weeks after surgery will be included. Patients will then be randomized to receive oxaliplatin-based chemotherapy either via HAI or IV combined with standard IV chemotherapy, every 2 weeks for at least 3 months.

The primary objective of this study will be to determine if the administration of oxaliplatin via HAI increases the time between treatment and disease recurrence compared to IV administration. The secondary objectives include overall survival, hepatic recurrence-free survival, safety, pattern of recurrence, and quality of life.

Inclusion Criteria:

• Patient must have signed a written informed consent prior to any trial specific procedures. When the patient is physically unable to give their written consent, a trusted person of their choice, independent from the investigator or the sponsor, can confirm in writing the patient's consent
• Age > 18 years
• ECOG performance status 0-1
• Histologically confirmed stage IV pMMR CRC
• Resected CRLM by one- or two-stage procedures including reverse strategy
• Partial Response or Stability Disease (RECIST 1.1) to preoperative cytotoxic doublet or triplet IV chemotherapy +/- targeted agent before surgery
• Curative-intent ( R0/R1 resection ± local ablation) surgery of 4 or higher CRLM
• No macroscopic residual (hepatic or extra-hepatic) disease on postoperative CT scan within 4 weeks after surgery confirmed during local multidisciplinary tumor board (except up to 3 lung nodules < 10 mm deemed amenable to curative-intent resection/local ablation and non-resected primary tumor with no or mild symptoms)
• Eligible to HAI of oxaliplatin by (permanent or selective) catheterization defined as the absence of medical (any contraindication to oxaliplatin administration, mainly residual peripheral sensory neuropathy grade < 2) and technical (vascular anatomy to perform HAI chemotherapy) contraindications to administer oxaliplatin-based doublet or triplet chemotherapy within 8 weeks from surgery during at least 4 cycles evaluated by interventional radiologist and medical oncologist
• Normal liver function (bilirubin < 1.5 x upper limit of normal values (ULN), aminotransferases < 5 ULN, alkaline phosphatase < 5 ULN, International normalized ratio (INR) < 1.5 ULN, platelets > 100,000/mm3)
• Women of childbearing potential must have a negative pregnancy test done within 30 days before randomisation
• Potentially reproductive patients must agree to use an effective contraceptive method or practice adequate methods of birth control or practice complete abstinence while on treatment, and for at least 6 months after the last dose of study drug
• Patients must be willing and able to comply with the protocol for the duration of the study including scheduled visits, treatment plan, laboratory tests and other study procedures
• Patients must be affiliated to a Social Security System (or equivalent)

Exclusion Criteria:

• Stage IV dMMR CRC
• Progressive Disease (RECIST 1.1) to preoperative cytotoxic doublet or triplet IV chemotherapy +/- targeted
• Incomplete (R2) surgery or residual (hepatic or extrahepatic) disease on postoperative CT scan within 4 weeks after surgery or symptomatic primary tumours in case of reverse strategy
• Extra hepatic metastasis disease, except ≤ 3 lung nodules < 10 mm deemed amenable to curative-intent resection/local ablation and non-resected primary tumor with no or mild symptoms
• Impossibility to receive at least 4 postoperative cycles with oxaliplatin
• Patients with contraindications for HAI or IV doublet or triplet administration as limiting anatomical variations of hepatic artery, peripheral sensory neuropathy ≥ grade 2 (NCI-CTAE v.5.0), gastric/duodenal ulcer or significant chronic liver disease (resulting in portal hypertension and/or liver failure)
• Peripheral neuropathy grade ≥ 2
• Patient with a dihydropyrimidine dehydrogenase deficiency (DPD)
• Medical history of other concomitant or previous malignant disease, except adequately treated in situ carcinoma of the uterine cervix, basal or squamous cell carcinoma of the skin, or cancer in complete remission for ≥5 years
• Pregnant women or women who are breast-feeding
• Participation in another therapeutic trial within the 30 days prior to randomisation
• Patients unwilling or unable to comply with the medical follow-up required by the trial because of geographic, familial, social, or psychological reasons
• Persons deprived of their liberty or under protective custody or guardianship