1183 Surgery Type Predicts Post-Operative Jejunal Polyp Detection in Patients With Familial Adenomatous Polyposis After Surgical Duodenectomy

Abstract

INTRODUCTION: In patients with familial adenomatous polyposis (FAP), advanced duodenal polyposis by Spigelman stage requires consideration of duodenal resection to prevent duodenal cancer. Pancreas-sparing duodenectomy (PSD) is preferred over pancreaticoduodenectmy (PD). Jejunal polyps occur 41 to 78% in post-duodenectomy patients, however, the differences by surgery type have not been evaluated. METHODS: We identified consecutive FAP patients following duodenal resection, including PD, PSD, or segmental duodenectomy (SD), at Cleveland Clinic, 08/1978 - 01/2018. Jejunal polyp-free survival (JPFS) time was calculated as time from surgery to detection of first lesion, or if no polyps, time to most recent endoscopy. Severity of jejunal polyposis was classified using the Spigelman score (henceforth, Neo-Spigelman). Kaplan-Meier curves were used to estimate JPFS by surgery type. Cox proportional hazards models adjusted for age, sex, race, and presence of pre-operative jejunal polyps were used. RESULTS: 64 patients were identified (Table 1). Jejunal polyposis was detected after surgery in 38/64 patients (59.4%, Table 2). Surgery type (PSD vs PD vs SD) was a significant predictor of JPFS overall (Figure 1, P = 0.008). 5-year JPFS was 49.5% (CI 32.3–64.7%) for patients after PSD, 80.7% (CI 51.1–93.4%) after PD, and 66.7% (CI 19.5–90.4%) after SD. Comparison of PSD vs PD yielded a significant difference in JPFS (P = 0.006), as did comparison of PSD vs SD (P = 0.048). PD vs SD was not significant (P = 0.12). In comparison PD, adjusted analyses showed hazard ratio (HR) of polyp detection at 4.0 (CI 1.6 – 10.0) for PSD. HR estimate was not significant for SD (HR 2.6, CI 0.6 – 11.1). Surgery type had a significant association with Neo-Spigelman score (P < 0.037), with mean score at 4.5 ± 0.5 for PD, 5.5 ± 1.4 for PSD, and 6.3 ± 0.6 for SD. No association was found between surgery type and management of jejunal polyps (surveillance vs endoscopic resection vs surgery, P = 0.09). CONCLUSION: Jejunal polyposis occurs in the majority of FAP patients after surgical duodenectomy. We demonstrate differences in polyp-free survival by surgery type, with PSD demonstrating higher risk of jejunal polyposis in comparison to PD. Surgery type was also associated with severity of polyposis, with a tendency for higher Neo-Spigelman scores after PSD. The long-term outcomes of jejunal polyps in FAP patients requires further study, and the surgical determinants of their occurrence may merit consideration.