Endoscopic Ultrasound-Lavage Technique for Pancreatic Cancer: An In Vivo Pilot Study

Background: Pancreatic cancer (PC) has a very poor 5-year survival and prognosis. Even when CT or MRI shows no metastasis, staging laparoscopy(SL) still detects tiny peritoneal deposits in 20-30% of patients, making them ineligible for surgery. SL is invasive, requiring general anesthesia and substantial resources. Endoscopic ultrasound (EUS) allows the observation of the bile ducts, pancreas, and abdominal cavity, and EUS-guided fine-needle aspiration (EUS-FNA) is essential for pathological diagnosis. Reports on using EUS to perform peritoneal lavage cytology are currently not available. We hypothesized that combining EUS-FNA with peritoneal lavage (EUS-lavage technique; EUS-LT) could enhance staging accuracy and avoid unnecessary surgical procedures. Methods: Ten in vivo porcine models underwent EUS-LT. Using a 19G FNA needle, 800 mL saline was instilled into the intraperitoneal cavity and then recovered. Two refinements were introduced sequentially: an ENBD catheter with additional side holes and, subsequently, a side-hole introducer (EndoSheather) that eliminated balloon dilation. The primary endpoint was procedural success. Secondary endpoints included safety, complications, recovered volume, duration of endoscopic procedure, and time required to instill 800 mL. Nonparametric tests compared outcomes across iterations. Results: Ten-model porcine in vivo model series were included, and all procedures were successful. No device malfunctions or unanticipated technical failures; one minor mucosal injury during saline injection resolved after re-puncture. The average procedure time was 31.1 min. Stepwise refinements shortened procedure and infusion times and increased recovered volume. Recovered volume approached the instilled amount in later cases, indicating efficient performance. Conclusions: In this ten-model in vivo series, EUS-LT demonstrated technical feasibility and short-term safety.