O2146: Risk Factors and Treatment Strategies for Anastomotic Ulcers in Pediatric Short Bowel Syndrome

Background: Anastomotic ulcers (AUs) are a rare cause of morbidity in short bowel syndrome (SBS). Prior studies of AUs have been small, descriptive reports. This study evaluates a large cohort of SBS patients to comparatively identify risk factors and describe treatment strategies for AUs.

Methods: This was a retrospective, case-control study of SBS patients within an interdisciplinary pediatric intestinal rehabilitation program from 2013-2023. Cases were patients with localized, peri-anastomotic ulceration on GI endoscopy and were matched 1:2 with SBS controls without an AU diagnosis, based on sex and time since intestinal failure diagnosis. Cases and controls were compared using conditional logistic regression on clinically relevant risk factors.

Results: Of 588 patients followed in our program and screened, 32 (5.4%) cases were identified. Median duration from initial surgery to AU diagnosis was 5.9 years, the anastomosis was enterocolonic in 23 (71.9%). Median follow up after AU diagnosis was 2.17 years. Compared to controls, AU patients were more likely to have bacterial overgrowth (74.2% vs 25.8%), and less likely to have volvulus (12.5% vs 37.5%) or an intact ileocecal valve (15.6% vs 42.2%). On multivariable analysis (Table 1), AU was associated with increased bowel length (OR 2.3 for each 10% increase in percent expected bowel length, CI 1.31,4.05), longer duration of parenteral nutrition in years (OR 1.69, CI 1.08, 2.63) and decreased weight-for-age Z-score (OR 0.38, CI 0.16, 0.92). Of cases, 10 (31%) required at least one red blood cell transfusion. Initial medical management with antibiotics, anti-inflammatories, and/or enteral steroids was utilized in 29 (91%) cases. Of these, 8 ultimately failed medical management, as defined by the clinical team, and underwent surgery (7) or endoscopic coagulation (1). Initial procedural intervention was performed with endoscopic coagulation in 1 (3%) case and surgery in 2 (6%) cases, of which 1 required repeat surgery (Figure 1).

Conclusion: In this large cohort of SBS patients, AU occurred in 5.4% of patients. Most patients can be managed medically but surgery may be ultimately required in select patients. The odds of AU increased with greater percent expected bowel length and longer PN duration and decreased with higher weight-for-age Z-scores. The association with longer PN duration and lower weight-for-age Z-scores may be due to ongoing bowel inflammation impacting growth and absorption. This, along with greater percent expected bowel length, may also potentially be due to salvage of marginally viable bowel at surgery in attempt to preserve bowel length, thus resulting in poorer quality bowel at these anastomoses. Further exploration, such as more detailed pathologic evaluation of resected anastomoses, is needed to provide additional clarity regarding these hypotheses.