S0924 Is Afferent Limb Length in Post-Pancreaticoduodenectomy Patients a Limiting Factor for Endoscopic Retrograde Cholangiopancreatography?

Abstract

INTRODUCTION: Despite the advances in modern endoscopy, endoscopic retrograde cholangiopancreatography (ERCP) remains a challenge in patients who are post-pancreaticoduodenectomy (PD). For post-PD reconstruction, the choledochojejunostomy and pancreaticojejunostomy are frequently placed on a single limb of the jejunum to create the afferent limb. This study evaluated the technical and clinical success rates of ERCP in post-PD patients relative to their afferent limb length (ALL). METHODS: We retrospectively evaluated patients with prior classic or pylorus-preserving PD who received ERCP from 01/13/06 to 01/16/19. Operative reports were analyzed for anatomical post-PD reconstruction details. Outcomes included: endoscopic access to the anastomosis of interest, technical success rate—the ability to treat stricture with dilation or stent placement, and clinical success rate—a documented decrease in bilirubin to normal or 50% of peak value within 2 weeks. Statistical analysis was done using Stata/SE version 16.0. RESULTS: A total of 28 patients (75% male; median age 66 years, range 22–86 years) with prior PD (20 classic PD; 8 pylorus-preserving PD) received 66 ERCPs [Table 1]. The median ALL was 42.5 cm (range 40–65 cm) in both surgical groups, 40 cm (range 40–60 cm) in those with pylorus-preserving PD, and 45.0 cm (range 40–65 cm) in those with classic PD. The most commonly used endoscope to reach the anastomosis of interest was a therapeutic upper endoscope (n = 31, 47.0%) [Table 2]. For biliary decompression, the choledochojejunostomy was reached on 16/25 initial ERCPs (64%), and on 54/63 (85.7%) total ERCPs when repeat attempts were included. All 3 ERCPs performed for pancreatic intervention were unsuccessful. The technical success rate was 81.0% (51/63 ERCPs) and the clinical success rate was 76.2% (48/63 ERCPs) [Table 3]. In the 9 patients where the enterobiliary anastomosis was not identified, 2 had pylorus-preserving PD, 7 had classic PD, and the median ALL was 40.0 cm (range 40-65). Unsuccessful ERCPs were attributed to sharp bowel angulations (n = 5), poor visibility (n = 1), and ALL (n = 2; length >65 cm). There were no perforations or bleeding. CONCLUSION: Overall, ERCP was limited by excessively long afferent limbs (lengths >65 cm), but more commonly by sharp bowel angulations that prevented endoscope advancement. Improvements are necessary for the endoscopic management of the pancreaticojejunostomy in patients who are post-PD.Table 1.: All enteric anastomosis were placed on a single limb of the jejunum to create the afferent limbTable 2Table 3