P1.042. Utility of a Fenestrated Surgical Drain for the Prevention and Treatment of Esophageal Leak

Abstract

Abstract Topic Benign Disease: Iatrogenic Esophageal Disease, Perforation and Postsurgical Complications Background Esophageal transmural defects are associated with high morbidity and mortality. Sponge-based endoscopic vacuum therapy (EVT) is effective but costly and requires frequent exchanges under general anesthesia. We describe a modified, cost-effective EVT approach using a fenestrated, spongeless radio-opaque silicone drain for both prevention and treatment of esophageal leaks. Methods A retrospective case series was conducted across three tertiary upper gastrointestinal centers in Singapore (2024–2026). Eighteen consecutive patients received an 18Fr fenestrated drain treatment, connected to continuous wall suction (-150 to -200 mmHg), with a daily 30-minute rest period. Therapeutic group (n=8) included post-esophagectomy anastomotic leaks (n=4), Boerhaave’s syndrome (n=2), squamous cell carcinoma perforation (n=1) and post-leiomyoma excision leak (n=1). The fenestrated drain was placed endoscopically, either intraluminally or within the leak cavity. Prophylactic group (n=10) were deemed high-risk pre-operatively (age >65, ASA>2, ECOG>1, BMI>29, HbA1c>8.0%, smoking, or end-stage renal failure) and received intraoperative intraluminal drain placement spanning the anastomosis following esophagectomy. A water-soluble contrast study was performed on postoperative days 5–7 prior to drain removal. Primary outcomes were complete leak healing (therapeutic) and leak prevention (prophylactic). Secondary outcomes included drainage duration, exchange interval(s) of the drain, anesthesia requirements during drain exchanges and time to oral feeding. Results Technical success of drain placement was 100%. In the therapeutic group (mean age 66 ±13.3 years), all eight patients achieved leak closure. Median treatment duration was 27.5 days (range 7–52). Six patients (75%) healed with the fenestrated drain alone (median: 20.5 days [range: 7–50]); two (25%) transitioned to sponge-based EVT due to surgeon preference despite adequate initial sepsis control. Median time to oral feeding was 23 days (range: 8–52). In the prophylactic group (mean age 57.8 ±8.8 years), nine of ten patients (90%) achieved uneventful anastomotic healing. Median time to oral feeding was 8 days (range: 6–18). One patient with an aorto-esophageal fistula developed conduit necrosis requiring re-operation. Eleven patients required no drain exchanges. Among those needing adjustments, the median exchange interval was 10 days (range: 4–30); 82.4% were performed under local anesthesia or sedation. One superficial gastric conduit ulceration occurred. No drain-related mortality occurred. Conclusion EVT using a fenestrated surgical drain appears safe and cost-effective when compared to conventional sponge-based EVT. Its reduced anesthesia requirements and infrequent exchange intervals address key limitations of existing techniques. While these findings support its utility for both therapeutic and prophylactic purposes, larger prospective studies are needed to confirm its effectiveness.