During pediatric liver transplant, biliary reconstruction is often performed using Roux-en-Y choledoc-hojejunostomy or hepaticojejunostomy due to size mismatch, excessive tension caused by distance between donor and recipient ducts, or for transplant to treat primary biliary pathology. This method can be associated with additional small bowel-related complications compared with end-to-end ductal anastomosis. We report a case of late small bowel obstruction secondary to an impacted bezoar that formed at a patulous jejunojejunostomy portion of the biliary-enteric anastomosis. The patient was a 26-year-old male patient, who underwent deceased donor whole liver transplant for pediatric acute liver failure. Prior to his presentation to our institution with 2 days of abdominal pain, nausea, persistent burping, and intermittent vomiting, the patient reported an uneventful posttransplant course and was followed up at the institution where his transplant was performed. There were no reported changes in diet or lifestyle and no similar episodes during his follow-up. The patient was managed surgically after a brief trial of nonoperative management. At laparotomy, the anastomosis was resected and reconstructed to improve enteric drainage and prevent recurrent bezoar formation. The patient was discharged with no postoperative complications and remained asymptomatic at 11 months follow-up. We describe the clinical course and our technical approach at initial choledochojejunostomy creation and at jejunojejunostomy revision.
Key words : Liver transplantation, Roux-en-Y choledochojejunostomy, Small bowel obstruction
Introduction
Bezoars are foreign body aggregates that occur anywhere along the gastrointestinal (GI) tract but are predominantly found in the stomach. Intestinal bezoars have been identified as the cause of small bowel obstruction in 0.8% to 4.6% of cases in various published series.1-4 Gastrointestinal dysmotility, poor dentition, dietary habits, and trichophagia are common risk factors. Bezoars can occasionally form as sequelae of intestinal bypass procedures, but the incidence of bezoars after surgery is unclear. Here, we present a rare case of a late small bowel obstruction secondary to a bezoar that formed at the jeju-nojejunostomy anastomosis after Roux-en-Y choledoc-hojejunostomy at orthotopic liver transplant.
Case Report
A 26-year-old male patient with a history of liver transplant at 6 years of age for fulminant hepatic failure of suspected viral etiology presented to our institution with 2 days of abdominal pain, nausea, persistent burping, and intermittent vomiting. At a different transplant center, he received a whole liver deceased donor small adult graft with a Roux-en-Y choledochojejunostomy biliary anastomosis due to duct size mismatch. Prior to this presentation, the patient reported no change in diet or lifestyle and no similar episodes during his posttransplant course. On presentation, a computed tomography (CT) scan with oral contrast showed air and fluid-filled, dilated small bowel loops proximal to the jejunojejunostomy anastomosis (Figure 1). We compared this CT scan with previous results, and the patient did seem to have chronic dilatation of small bowel proximal to this anastomosis. A nasogastric tube was placed with consistent output of about 4 L/d. The patient passed flatus and had loose bowel movements but continued to have high nasogastric tube output despite several days of gastric decompression. Promotility agents were trialed without improvement in his symptoms.
The decision was made to operatively intervene. With a collaboration between bariatric surgery and transplant surgery teams, an exploratory laparotomy was performed. The jejunojejunostomy was noted to be patulous, approximately 10 × 15 cm, and filled with organic material. After careful delineation of the anatomy, the anastomosis was resected and reconstructed with 2 stapled anastomoses (Figure 2 and Figure 3). His postoperative course was uncomplicated. A nasogastric tube was kept in place for 48 hours, and an upper GI series was performed that revealed no leak. Concern for motility issues prompted a Gastrografin upper GI series due to chronic afferent bowel loop dilation dating back to 6 years before presentation. The nasogastric tube was removed, and his diet advanced as tolerated. He was discharged home on postoperative day 5 with a regular diet and no further postoperative comp-lications. He remained without recurrent bowel obstruction or oral intolerance at 11 months after exploration and revision.
Discussion
Transplant patients can be at risk for early and late complications related to the small bowel when biliary-enteric anastomoses are created during orthotopic liver transplant. Causes of small bowel obstruction commonly include adhesions, internal hernias, or a narrowing at the jejunojejunostomy anastomosis.5,6 Although there are a few case reports of GI bezoar formation following liver transplant, the literature is limited.7,8 There is a published case report of small bowel obstruction presenting 13 years after liver transplant, due to impaction of a bezoar embedded with gallstones in the common limb of Roux-en-Y hepaticojejunostomy.9 Others have reported enterolith formation after liver transplant, leading to intestinal obstruction.10-12 We speculate that in this case the original jejunojejunostomy was potentially too narrow and could have caused chronic upstream dilation, leading to stasis of enteric contents and bezoar formation. Although nono-perative management was trialed for 2 days, this was not successful. As in this case, patients in other reports of small bowel obstruction from bezoar formation after liver transplant required operative management.8,9 Several published series have shown that up to 80% of GI bezoars causing small bowel obstruction in the general population require operative intervention.4,13
Technical considerations for Roux-en-Y choledoc-hojejunostomy creation during orthotopic liver transplant
As described in the original operative report, a stapled jejunojejunostomy was performed. The length of Roux limb was not documented, and at his recent surgery (20 years after the transplant) we found a retrocolic Roux limb. As a teaching point, at the time of the initial creation of the Roux-en-Y at the original transplant operation, it is important not to narrow one of the limbs when closing the common enterotomy. Other steps to minimize complications include closing the mesenteric defect to reduce risk of internal hernia, placement of the choledoc-hojejunostomy anastomosis close to the end of the Roux limb to avoid bacterial overgrowth and reflux, and proper anastomosis of the distal jejunal limb (and not proximal divided jejunum) to the bile duct (to avoid Roux-en-O misconfiguration). Additionally, it is important to maintain adequate length of the limbs in the original anastomosis to prevent bile reflux and also to have adequate length of bowel in case subsequent revision is required. Utilization of absorbable suture is recommended, given that occurrence of small bowel obstructions at the jejunojejunostomy anastomosis have been linked to the use of nonabsorbable suture material during laparoscopic Roux-en-Y anastomosis.14
Limitations to endoscopic management
Endoscopic removal of bezoars, including gastric bezoars, is difficult due to the large size of the obstructive material and may be complicated by esophageal perforation. In some patients, push enteroscopy is required in order to reach the jejunojejunostomy anastomosis, which further complicates this option. In our case, at laparotomy, the jejunojejunostomy anastomosis was noted to be patulous and was approximately 10 × 15 cm. Endoscopic removal of the contents alone would have predisposed the patient to recurrent bezoar formation. Similarly, simple extraction of the bezoar intraoperatively with horizontal closure of the enterotomy would not reduce the likelihood of recurrence, given the patulous nature of the anastomosis. We therefore opted to resect the original anastomosis and repeat the jejunojejunostomy procedure.
Jejunojejunostomy revision technique
There is not a published report of the incidence of reoperation for bezoar following liver transplant; however, in a report of bariatric surgery operations, 3.3% of reoperations required a revision of the jejunojejunostomy.15 To perform the jejunostomy revision, we freed the anastomosis and then performed 2 stapled side-to-side functional end-to-end anastomoses using Endo GIA universal staples (60 mm, vascular/medium, tan). The first anastomosis was from the Roux limb to the common channel on the shared mesentery, and another anastomosis was performed 20 to 25 cm downstream, from the biliary limb to the common channel. The common enterotomies in both cases were closed with Endo GIA staples (60 mm, vascular/medium, tan). A crotch stitch and anti-kink/anti-obstruction stitch were placed at the anastomoses. As a general rule for mitigation of Roux-en-Y choledochojejunostomy anastomosis problems, it is usually best to resect and repeat the anastomosis. Resection of the whole anastomosis must be performed with care to avoid damage to the Roux mesentery. The Roux limb and the common (distal) channel have the same mesentery, so the first step should be reanastomosis of these structures. Then, the surgeon should choose the limb coming from the ligament of Treitz (eg, proximal limb, gastroduodenal limb) and perform a second anastomosis close to the first anastomosis (usually downstream on the common channel). These anastomoses can be stapled if the limbs are not narrowed. If there are adhesions and the anatomy is difficult to identify, then the surgeon should start distally at the terminal ileum, trace the common channel, work back to the jejunojejunostomy, and then follow the continuous mesentery onto the Roux and tag each limb for identification. Laparoscopic anastomotic resection and revision is also possible in the absence of extensive adhesions when an experienced surgeon is available.
Conclusions
We presented a case of a delayed bezoar at the jejunojejunostomy anastomosis following orthotopic liver transplant. We discussed management options and techniques for the initial Roux-en-Y creation and jejunojejunostomy revision
References:

Volume : 21
Issue : 7
Pages : 615 - 618
DOI : 10.6002/ect.2023.0102
From the 1Department of Surgery, Division of Transplantation, Massachusetts General Hospital, Boston, Massachusetts; the 2Boston College, Chestnut Hill, Massachusetts; the 3Department of Minimally Invasive Surgery, Bariatric Surgery and Foregut Surgery, Houston Methodist Academic Institute, Houston, Texas; and the 4Department of Surgery, Division of General Surgery, Massachusetts General Hospital, Boston, Massachusetts, USA
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Leigh Anne Dageforde, Massachusetts General Hospital, 55 Fruit Street, White 5, Boston, MA 02114, USA
Phone: +1 617 726 5277
E-mail: ldageforde@mgh.harvard.edu
Figure 1. Images From Abdominal Computed Tomography With Oral Contrast
Figure 2. Resected Jejunojejunostomy Opened to Show the Bezoar Contents
Figure 3. Intraoperative Images of Anastomosis of Roux Limb and Common (Distal) Channel