Mo1255 Management of Duodenal Perforations After Upper Endoscopic Procedures: Is a Non-Surgical Approach Tenable?

Abstract

Management of Duodenal Perforations After Upper Endoscopic Procedures: Is a Non-Surgical Approach Tenable? Alejandro L. Suarez*, Dennis Collins, Anand Gupte, Shailendra S. Chauhan, Peter V. Draganov, Christopher E. Forsmark, Mihir S. Wagh Division of Gastroenterology, University of Florida, Gainesville, FL Background: Duodenal perforation is a rare but serious endoscopic complication. Management of duodenal perforations is not standardized and surgical intervention has been the traditional approach. Aim: To review the incidence, management and outcomes of duodenal perforations after upper endoscopic procedures. Methods: A retrospective review of our electronic endoscopic complications database and medical records was performed to identify patients with duodenal perforations after upper endoscopic procedures (EGD/ enteroscopy, EUS, ERCP) from 1/2000 8/2011. The study was approved by the IRB at the University of Florida. Medical records were reviewed to collect data on incidence, management and outcomes after duodenal perforation. Data collected included procedure indication, type of endoscopic procedure, time of diagnosis, radiologic test results, management via surgery or conservative measures, length of hospital stay and mortality. Results: 43,572 upper endoscopic procedures were performed during this period [30,038 EGDs (69%), 7,761 ERCPs (18%), 5,773 EUS (13%) ] leading to 18 duodenal perforations (0.04%). 6/18 (33%) perforations resulted from diagnostic procedures while 12/18 (67%) were from therapeutic endoscopies including duodenal EMR/polypectomy, dilation, EUS-FNA, sphincterotomy and stent placement. 8/18 (44%) were identified during endoscopy, 5/18 (28%) within 24 hours and 5/18 (28%) after 24 hours after endoscopy. 16/18 (88%) had radiographic imaging (15 CT scans and 1 abdominal X-ray) while 2 went straight to surgery. Overall, 12/18 (67%) underwent surgery with a mean length of stay of 28 days while 6/18 (33%) were treated conservatively with a mean length of stay of 10 days. Intraperitoneal free air was seen in 73% and retroperitoneal air was seen in 27%. Abdominal CT scan with oral contrast was used to detect duodenal extravasation and leak. Contrast extravasation was seen in 4/15 (27%) with all 4 patients (100%) undergoing surgery (2/4, 50% survived with a mean length of stay 13.5 days). Extravasation of contrast was not seen on CT in 11/15 (73%). Only 5/11 (45%) had surgery (all 5 survived with a mean length of stay 14 days) while 6/11 (55%) were treated conservatively (NPO, NG tube, IV fluids, antibiotics, percutaneous drain if needed), of which all 6 (100%) survived with a mean length of stay 10.3 days. Complications were seen in 4/18 (22%) patients (1 with fever that resolved with antibiotics, 1 sepsis, 1 post-ERCP pancreatitis and 1 abdominal abscess drained percutaneously). Overall, 2/18 (11%) patients died after duodenal perforation (both patients had contrast extravasation requiring surgery). Conclusion: Though limited by small numbers, our data suggests that patients with duodenal perforation without duodenal contrast extravasation on CT scan may be effectively managed conservatively without surgery.