First-intention EUS-guided transluminal drainage with LAMS: an effective and safe method for management of fluid collections after any kind of surgery.

Background: Symptomatic postoperative collections (PCs) frequently complicate surgery with significant morbidity and mortality. In contrast with pancreatic inflammatory collections, little is known about endoscopic ultrasound-guided drainage of PCs (EUS-PCD). The aim of this study is to evaluate the safety and efficacy of EUS-PCD using lumen-apposing metal stent (LAMS) as the first-line drainage approach for PCs of any kind.

Methods: This is a monocentric retrospective study. We retrieved all consecutive symptomatic PCs treated at our center between February 2019 and September 2024. All cases were considered suitable for EUS-PCD after multidisciplinary discussion. Rates of technical success, clinical success, and AEs were calculated.

Results: We retrieved 66 PCs, mainly resulting from pancreatic and lower gastrointestinal tract surgery. The median size of collections was 7.6 cm and infection occurred in 54 of the cases. The median time from surgery to drainage was 19 days (IQR 13-29); in 10 cases, this occurred ≤ 7 days after surgery. 51 drainages were performed from the gastric/duodenal window, 15 transrectally. LAMS were removed after a median time of 18.5 days (IQR 12-27). After removal, double-pigtail stents were placed in 25 PCs and at least one necrosectomy session was performed in 13. Technical success was achieved in 97.0% of cases. Clinical success was achieved in 95.2%; in 3 cases, collection recurrence occurred and retreatment with LAMS was successful. Overall AEs rate was 9.1%, but only one was severe, requiring surgery. Rates of technical and clinical failure and AEs were not affected by surgery type (pancreatic, non-pancreatic), timing of drainage (≤ 7, 7-10, > 10 days), size of collections (≤ 4, 4-10, > 10 cm), and access window (transgastric/duodenal/rectal). Necrosectomy performance was the only predictor of AEs occurrence (OR 6.9, C.I.: 1.1-46.9, p = 0.048) at univariable analysis.

Conclusion: First-intention EUS-PCD seems to be a safe and effective treatment, regardless of the origin and size of the collection and drainage timing.