The peritoneum is a relatively frequent metastatic site in digestive tumors (colon, stomach, pancreas) and is characterized by a poorer prognosis compared with other metastatic sites such as the lung or liver. Its dissemination pathway is complex and most often involves crossing the hemato-peritoneal barrier. This type of metastasis is difficult to visualize on imaging at an early stage, and surgical exploration may be required. In gastric cancer in particular, exploratory laparoscopy is part of the initial staging work-up for locally advanced tumors to assess the presence or absence of peritoneal metastases. It is therefore important to develop new, less invasive detection or prediction methods.

Circulating tumor DNA (ctDNA) is a promising non-invasive blood biomarker that can assist clinicians as a prognostic/predictive biomarker and/or a tool for monitoring response to anti-tumor therapies. This marker is most often assessed in plasma, but recent data suggest that tumor DNA may also be detected in other biological fluids such as peritoneal fluid. A preliminary study conducted by our team showed the ability to detect tumor DNA in peritoneal fluid from patients with peritoneal carcinomatosis of various origins, with a sensitivity of 75%. In gastric cancer, a recent meta-analysis demonstrated an increased risk of peritoneal metastases when peritoneal tumor DNA was positive (RR 13.81 [95% CI, 8.11-23.53]), as well as a reduction in 3-year recurrence-free survival (RR 5.37 [95% CI, 1.39-20.74]) and overall survival (HR 4.13 [95% CI, 1.51-11.32]).

The objective of this cohort is to evaluate the prognostic impact of circulating tumor DNA (ctDNA) in plasma and/or peritoneal fluid on the risk of developing peritoneal metastases. The primary endpoint is: Peritoneal recurrence rate according to tumor DNA positivity status (positive vs negative).

After screening, verification of inclusion and exclusion criteria, information and consent, patients will be enrolled in the study.

Several samples will be collected at different time points during patient management, in addition to samples already performed as part of routine care. No intervention or sampling will occur at a time point specifically required for research purposes. Blood and peritoneal fluid samples will always be collected concomitantly. They will then be transported to the molecular biology laboratory for initial DNA extraction. After extraction, samples will be stored at -80 °C until analysis.

The treatments offered to the patient follow standard-of-care recommendations and are not modified by participation in the study.

Three independent cohorts will be evaluated: gastric cancer, colon cancer, and pancreatic cancer.Depending on the type of cancer and the therapeutic strategy, different samples will be collected.

Systematic sampling:

Plasma: at diagnosis, at the time of surgery, in the early postoperative period, and 1 month after surgery (during the postoperative visit).

Peritoneal fluid: at the time of surgery and in the early postoperative period.

Specific sampling:

Peritoneal fluid: during pre-therapeutic exploratory laparoscopy (gastric cancer) or during diverting colostomy (obstructive colon cancer).

In total:

Gastric cancer: 3 peritoneal fluid samples and 4 blood samples.

Colon cancer: 2 peritoneal fluid samples and 3 blood samples systematically, plus 1 optional blood and peritoneal fluid sample (only in case of diverting colostomy).

Pancreatic cancer: 2 peritoneal fluid samples and 3 systematic blood samples, plus 1 optional blood sample (in case of neoadjuvant treatment).

Blood sampling:

2 EDTA tubes (molecular biology)

Peritoneal fluid sampling:

2 dry tubes (molecular biology, pathology)

For intraoperative sampling, a peritoneal washing is performed with 200 mL of 0.9% NaCl instilled into the abdominal cavity; then 50 mL are collected from the pouch of Douglas.

For the postoperative period, the sample will be collected between postoperative day 2 and day 5 directly from the drainage fluid (collected from the drain reservoir left in place at the end of surgery).

ctDNA analysis: NGS (next-generation sequencing), or high-throughput sequencing, is a molecular biology method that enables the rapid sequencing of millions of DNA molecules simultaneously and is used to identify point mutations as well as insertions, duplications, and deletions. It will be used for both plasma analysis and peritoneal fluid analysis.

Inclusion Criteria:

• Age > 18 years
• Signed non-opposition form
• For gastric cancer: histologically confirmed gastric adenocarcinoma >T2 and/or N+; for colon cancer: histologically confirmed colonic adenocarcinoma >T2; for pancreatic cancer: histologically confirmed pancreatic adenocarcinoma >T2
• No metastases on the initial staging work-up
• No contraindication to curative-intent surgical treatment

Exclusion Criteria:

• Primary tumor metastatic at diagnosis
• Presence of ascites or distant metastases
• Synchronous cancer or prior history of cancer within the past 5 years
• Contraindication to curative surgical treatment
• Any patient unable to comply with the study's medical follow-up for geographic, social, or psychological reasons
• Patients under legal guardianship/protection, or unable to read, understand, and sign the information notice and consent form
• Patients not affiliated with the social security system

All patients treated for a histologically proven, localized adenocarcinoma of gastric or pancreatic or colorectal, will be enrolled in this prospective cohort study after receiving and signed a specific consent information form.