Endoscopic retrograde cholangiopancreatography (ERCP) is the current standard technique for the treatment of pancreatobiliary disease. However, ERCP in patients with a surgically altered anatomy (SAA) remains a challenge. The short-type balloon enteroscope dedicated to pancreatobiliary intervention was recently introduced and has gained popularity as a primary modality for ERCP in patients with SAA. The currently available short-type single-balloon enteroscope (SBE) has a 3.2-mm enlarged diameter for the working channel and a 152-cm shortened working length, which can accommodate most conventional ERCP accessories and stent assemblies, and it is equipped with high-force transmission and passive bending, which facilitate passing the sharply angulated bowel segment. However, Roux-en-Y (R-Y) reconstruction anatomy is still challenging for the pancreatobiliary physician with regard to ERCP owing to the long length of the Roux and pancreatobiliary limb, and bowel angulations around the jejunojejunal anastomosis. The pooled procedural success of short SBE-assisted ERCP (SBE-ERCP) for R-Y reconstruction was reported to be 76.4% in a recent metaanalysis.

The most important factor for procedural success and safety of short SBE-ERCP for R-Y patients is to resolve and prevent various bowel types looping through the collaborative manipulation of an enteroscope and overtube. In the clinical field, there is an unmet need for a formulaic loop-handing technique that can be applied to most cases of R-Y reconstruction. Therefore, in the current study, we aimed to evaluate the efficacy and safety of a mechanistic loop resolution strategy for short SBE-ERCP in patients undergoing R-Y reconstruction.

SBE-ERCP and mechanistic loop resolution strategy The mechanistic loop resolution strategy was first introduced by an experienced senior professor with 30 years of experience with conventional balloon enteroscopy and ERCP. Based on this strategy, SBE-ERCP in this study was performed by a junior professor with 8 years of experience in colonoscopy and ERCP and had performed 5 cases of conventional BE-ERCP as an assistant before 2019. The short SBE was introduced in Korea in 2019, and from this time, the junior professor underwent training in SBE-ERCP with the mechanistic loop resolution strategy via 10 cases with Billroth-II or R-Y anatomy. From January 2020, he began performing SBE-ERCP as the main operator.

All ERCP procedures were performed with the patient in the prone position using an SBE (SIF-H290S; Olympus Corp., Japan) under CO2 insufflation and conscious sedation. A soft transparent hood (D-201-11804; Olympus Corp.) was used in all cases. The SBE was introduced alternately with an overtube apparatus (ST-SB1S; Olympus Corp.) following the mechanistic loop resolution strategy under endoscopic and fluoroscopic guidance. The overtube was advanced along the enteroscope by gently pulling the enteroscope, similar to the ERCP accessory advancement over the guidewire.

Mechanistic loop resolution strategy for total gastrectomy with R-Y reconstruction First (Step 1), an enteroscope was inserted beyond the esophagojejunal anastomosis into the jejunum, and the overtube was advanced across the esophagojejunal anastomosis to prevent recurrent reverse C-loop formation at the esophagojejunal junction. Second (Step 2), whenever an enteroscope was passed through the U-shaped or inverse U-shaped jejunal segment, an overtube was advanced along the enteroscope to sufficiently cover the U-shaped or inverse U-shaped jejunal segment. Thereafter, the enteroscope-overtube apparatus was retracted simultaneously after overtube ballooning to pleat the jejunum and to prevent U-loop or inverse U-loop reformation following subsequent enteroscope insertion. Subsequently (Step 3), when the U-shaped, inverse U loop became part of a three-dimensionally rotated N-loop, the enteroscope occasionally could not pass the U-shaped or inverse U-shaped jejunal segment, making a cane shape. At that time, the control section of the enteroscope was rotated 360° from its place, clockwise or counterclockwise toward the direction in which the loop formation was prevented. This preemptive extreme rotation maneuver aimed to minimize rotational vector forces from the loop and stiffen the enteroscope. Further (Step 4), when a three-dimensionally rotated loop containing a ring structure such as alpha, reverse alpha, and gamma loop was formed during enteroscope advance, before the trial of loop resolution, the tip of the overtube was positioned so that it did not reach the ring structure of the loop, with the ring structure not being covered with the overtube. Thereafter, the loop was corrected by rotating the enteroscope-overtube apparatus. This maneuver allowed the enteroscope to rotate in a three-dimensional spiral direction, whereas the overtube rotated in place, leading to the effective transmission of the rotational force generated by the operator's hand to the enteroscope-overtube apparatus. Biliopancreatic cannulation and therapeutic maneuvers were attempted by intentionally retroflexing the enteroscope tip near the inferior duodenal flexure. In Step 5, if the enteroscope tip was repetitively withdrawn before complete loop resolution due to a weak anchoring effect of the enteroscope tip, the overtube was further advanced into the ring structure of the loop to support the additional advancement of the enteroscope until a more suitable point for hooking and anchoring the enteroscope tip. When the enteroscope tip reached this anchoring point, the overtube was retracted back to the starting point of the ring structure of the loop, while maintaining the enteroscope tip in place. Thereafter, loop resolution was reattempted as described in Step 4. In Step 6, after passing the SBE into the pancreatobiliary jejunal limb and duodenum, a large reverse alpha loop that was formed through the Roux limb, jejunojejunal anastomosis, and pancreatobiliary limb was usually allowed without a trial of reduction because it facilitated the retroflex positioning of the enteroscope tip around the inferior duodenal flexure and guaranteed an enface view of the major papilla.

Mechanistic loop resolution strategy for R-Y hepaticojejunostomy with preserved stomach First (Step 1), once the enteroscope reached the second or third portion of the duodenum in a long-scope position, the enteroscope was straightened with rightward rotation and retraction, making a short-scope position, similar to the ERCP position. An overtube was passed into the duodenum over the straightened enteroscope, maintaining the short-scope position. The overtube balloon was positioned in the superior duodenal angle or duodenal bulb across the pyloric ring and inflated to prevent the overtube from being withdrawn back into the stomach. Thereafter (Step 2), the enteroscope was further advanced into the distal duodenum and jejunum, maintaining the short-scope position and being cautious of the recurrence of the long-scope position of the enteroscope in the stomach. In this step, continuous covering of the superior duodenal angle and pyloric ring with the overtube balloon was crucial for maintaining the short-scope position. In Step 3, after the tip of the enteroscope was inserted deep into the jejunum, the unstable loops over the superior and inferior duodenal flexure were corrected, making the esophagus, stomach, duodenum, and proximal jejunum lie in a straight line. The other basic loop resolution strategies were the same as those described for total gastrectomy with R-Y reconstruction.

Inclusion Criteria:

• Patients with a previous R-Y reconstruction
• Clinical and radiological diagnosis of biliary obstruction

Exclusion Criteria:

• Coagulopathy (International normalized ratio >1.5, platelet count <50,000)
• Peritoneal carcinomatosis.