Transmural migration of an internal pancreaticojejunostomy stent causing delayed intra-abdominal sepsis: a case report

Abstract

Background

Internal pancreatic duct stents are commonly used during pancreaticojejunostomy (PJ) reconstruction following pancreaticoduodenectomy to facilitate drainage and potentially reduce postoperative pancreatic fistula (POPF). These stents are typically expected to migrate distally and pass without consequence. However, rare complications related to stent migration may occur and can present with atypical clinical and biochemical findings, creating diagnostic challenges.

Case Description

A 63-year-old woman underwent pancreaticoduodenectomy for pancreatic adenocarcinoma with placement of an internal PJ stent. Her initial postoperative course was uneventful, with normal drain amylase levels. On postoperative day (POD) eight, she presented with fever, leukocytosis, and abdominal pain. Computed tomography (CT) demonstrated peripancreatic fluid and gas collections suspicious for POPF; however, drain amylase levels remained low, creating diagnostic uncertainty. Despite initial nonoperative management, the patient’s clinical status deteriorated, prompting exploratory laparotomy. Intraoperatively, the PJ anastomosis was intact, but diffuse enteric contamination was present. Intraoperative ultrasound revealed transmural migration of the pancreatic duct stent through the jejunal wall into the mesentery, resulting in a concealed perforation with extraluminal leakage of enteric contents. The stent was retrieved via controlled enterotomy, the defect was repaired, and the abdomen was irrigated. The patient subsequently improved with resolution of sepsis and was discharged in stable condition.

Conclusions

Transmural migration of an internal PJ stent is a rare but serious complication that may present with intra-abdominal sepsis in the absence of elevated drain amylase levels. This case highlights an important diagnostic pitfall, as enteric contamination rather than pancreatic leakage may predominate. In patients with unexplained postoperative deterioration following pancreaticoduodenectomy, particularly when imaging is equivocal and biochemical markers are discordant, stent-related complications should be considered. Early recognition and prompt surgical intervention are critical for optimal outcomes.

Highlight box.

Key findings

• Transmural migration of an internal pancreaticojejunostomy (PJ) stent caused concealed mesenteric perforation and intra-abdominal sepsis despite an intact anastomosis.

• Drain amylase levels remained low, masking the diagnosis and creating a diagnostic pitfall.

• Intraoperative ultrasound was critical in identifying the etiology when cross-sectional imaging was inconclusive.

What is known and what is new?

• Internal pancreatic duct stents are commonly used during PJ reconstruction and are generally expected to migrate distally without clinical consequence. Rare complications such as biliary migration, infection, and obstruction have been described.

• This manuscript describes a rare mechanism of stent-related morbidity in which transmural migration into the mesentery caused enteric leakage and sepsis in the absence of elevated drain amylase.

What is the implication, and what should change now?

• Low drain amylase does not exclude serious postoperative complications following pancreaticoduodenectomy.

• In patients with unexplained postoperative sepsis and equivocal imaging, surgeons should maintain a high index of suspicion for stent-related complications, particularly when internal stents are used.

• There is a role for consideration of operative exploration with the use of adjunctive tools such as intraoperative ultrasound to facilitate timely diagnosis and improve outcomes.

Introduction

Pancreaticoduodenectomy remains the cornerstone surgical treatment for malignancies of the pancreatic head and periampullary region. Despite substantial advances in operative technique and perioperative care, the procedure continues to be associated with significant morbidity, with overall complication rates reported between 30% and 50%, even at high-volume centers (1,2). Among these complications, postoperative pancreatic fistula (POPF) represents one of the most clinically significant and potentially life-threatening events following pancreatic resection (3).

The development of POPF is largely related to failure of the pancreaticoenteric anastomosis and is influenced by multiple factors including pancreatic duct diameter, gland texture, operative technique, and surgeon experience (4). Consequently, numerous technical modifications of pancreatic reconstruction have been proposed in an effort to reduce fistula rates. These strategies include variations in anastomotic technique, reinforcement methods, and the use of pancreatic duct stents placed across the pancreaticojejunostomy (PJ) to facilitate drainage of pancreatic secretions and reduce intraductal pressure at the anastomosis (5).

Internal pancreatic duct stents are commonly used during duct-to-mucosa PJ and are designed to bridge the anastomosis, directing pancreatic secretions into the jejunal lumen during the early postoperative period. However, the routine use of these stents remains controversial. Several randomized trials and systematic reviews have demonstrated inconsistent benefit, with some studies suggesting no significant reduction in clinically relevant POPF compared with non-stented anastomoses (6,7). As a result, the decision to place a pancreatic duct stent is often guided by surgeon preference, institutional practice patterns, and intraoperative POPF risk stratification via examination of the pancreatic duct size and gland texture.

Internal stents, when placed, are typically intended to migrate distally through the gastrointestinal tract and pass spontaneously without clinical consequence. Although this process is usually uneventful, rare complications related to stent migration have been described, including migration into the biliary tree, cholangitis, hepatic abscess, and intestinal obstruction (8-10). Here, we report a case of transmural PJ stent migration causing concealed mesenteric perforation and delayed intra-abdominal sepsis, highlighting an important diagnostic pitfall and emphasizing the need to consider stent-related complications in patients with discordant clinical, biochemical, and imaging findings following pancreaticoduodenectomy. We present this article in accordance with the CARE reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0237/rc).

Case presentation

Initial presentation

A 63-year-old woman with a history of tobacco use, obstructive sleep apnea, hypertension, and migraine headaches and no past surgical history presented with progressive epigastric and right upper quadrant abdominal pain associated with abnormal liver function tests and clinical jaundice. Initial laboratory evaluation demonstrated markedly elevated transaminases and alkaline phosphatase with biliary obstruction suspected on abdominal ultrasonography, which revealed dilation of the common bile duct and pancreatic duct. No family history of pancreatic cancer.

Contrast-enhanced computed tomography (CT) demonstrated intrahepatic and extrahepatic biliary dilation with abrupt truncation of the distal common bile duct at the ampullary region and a subtle lesion within the pancreatic head. Endoscopic ultrasound (EUS) subsequently identified a 20×15 mm mass in the pancreatic head and uncinate process suspicious for pancreatic adenocarcinoma. Fine needle biopsy confirmed pancreatic adenocarcinoma, and staging imaging demonstrated no evidence of metastatic disease. Following multidisciplinary evaluation, the patient was deemed an appropriate candidate for pancreaticoduodenectomy.

Operative technique and PJ stent placement

The patient underwent pancreaticoduodenectomy after diagnostic laparoscopy confirmed the absence of metastatic disease. A standard pancreaticoduodenectomy with antrectomy, cholecystectomy, and regional lymphadenectomy was performed without the need for vascular reconstruction. Frozen section analysis confirmed negative pancreatic and bile duct margins.

Reconstruction was performed using a duct-to-mucosa PJ with a modified Blumgart technique. Posterior outer-layer pledgeted 4-0 Prolene U-stitches were placed through the pancreatic parenchyma and jejunal wall. The pancreatic duct was then anastomosed to the jejunal mucosa using interrupted 5-0 PDS sutures.

An internal pancreatic duct stent was fashioned from a 5 Fr × 41 cm Cardinal Health Argyle™ pediatric feeding tube (reference number 460802E), composed of soft, malleable polyvinyl chloride (PVC), and trimmed to a total length of 5 cm with a rounded, blunt tip and no side holes incorporated. Prior to stent insertion, a fine right-angle instrument was passed through both the pancreatic duct and the jejunal mucosal opening to confirm proper duct-to-mucosa alignment and verify an unobstructed intraluminal pathway.

Under our standard technique, the stent is intended to reside approximately equally within the pancreatic duct and jejunal lumen. However, in this patient, focal narrowing of the pancreatic duct limited advancement of the stent to approximately 1.5 cm into the duct despite appropriate intraluminal positioning, leaving a relatively longer intrajejunal segment. The stent was confirmed to move freely across the anastomosis and was secured to the jejunal mucosa with a 5-0 PDS suture. The anterior outer layer of the PJ was then completed using pledgeted 4-0 Prolene sutures, invaginating the jejunal wall over the pancreatic remnant.

Reconstruction proceeded with an end-to-side hepaticojejunostomy (HJ) and antecolic isoperistaltic gastrojejunostomy. Two closed-suction drains were placed adjacent to the PJ and HJ prior to abdominal closure, one anterior and one posterior.

Initial postoperative course

The patient’s early postoperative course was uncomplicated. Drain amylase measured on postoperative day (POD) 3 remained within normal limits. Both drains demonstrated persistently low-volume (<40–50 mL/day) serosanguinous output, with the posterior drain producing approximately 10 mL/day and the anterior drain approximately 30 mL/day. Given the minimal output, the posterior drain was removed prior to discharge, while the anterior drain was left in place. The patient was discharged home on POD 6 with the anterior drain in place.

Clinical deterioration

On POD 8, the patient presented to an outside emergency department with diffuse abdominal pain, fever, nausea, and leukocytosis. On examination, the patient was febrile with diffuse abdominal tenderness and guarding without peritonitis. Laboratory evaluation demonstrated a white blood cell count of 23.5×10⁹/L. CT demonstrated postoperative fluid and gas collections near the pancreatic resection bed measuring approximately 8.5×2.9 cm (Figure 1). Careful review of the imaging also suggested that the pancreatic duct stent appeared to extend beyond the jejunal limb. However, this finding was subtle and difficult to interpret in the immediate postoperative setting.

Figure 1.

Contrast-enhanced coronal CT images demonstrating findings suspicious for PJ stent migration. (A) Post–pancreaticoduodenectomy anatomy. (B) Internal PJ stent coursing from the pancreatic remnant into the jejunal limb. (C) Distal portion of the PJ stent appearing to traverse or abut the jejunal wall, raising suspicion for transmural migration. (D) Associated postoperative fluid and gas collection adjacent to the pancreatic resection bed. CT, computed tomography; PJ, pancreaticojejunostomy.

Given the patient’s clinical deterioration and imaging findings, differential diagnoses included POPF, intra-abdominal abscess, anastomotic leak, and stent-related complication. However, analysis of fluid from the remaining surgical drain demonstrated low amylase levels (28 U/L), which remained only mildly elevated at 100 U/L after drain manipulation, creating uncertainty regarding the source of the patient’s sepsis. On POD 9, the patient was transferred back to our institution and initially managed with nasogastric tube decompression, intravenous antibiotics and octreotide therapy. Interventional radiology (IR) attempted CT-guided drainage of the suspected collection but was unable to obtain significant fluid.

Reoperation and identification of stent migration

Given persistent leukocytosis and concern for uncontrolled intra-abdominal sepsis, the decision was made to pursue an exploratory laparotomy on POD 10. Upon entry into the abdomen, approximately one liter of enteric succus was encountered throughout the peritoneal cavity. Inspection of the PJ, HJ, and gastrojejunostomy demonstrated intact anastomoses without evidence of leak.

Intraoperative ultrasound was then performed to evaluate the PJ and revealed that the internal pancreatic duct stent had migrated through the jejunal wall into the mesentery (Figure 2). This transmural migration resulted in a concealed mesenteric perforation with extraluminal leakage of enteric contents. Direct inspection of the PJ confirmed the distal portion of the stent exiting the jejunal wall and extending into the surrounding mesentery, with formation of an extraluminal mesenteric tract at the site of perforation (Figure 3).

Figure 2.

Intraoperative ultrasound demonstrating PJ stent position. (A) Ultrasound image showing the pancreatic duct stent appropriately positioned within the lumen of the PJ. (B) Ultrasound image demonstrating migration of the stent with the distal portion traversing the jejunal wall and extending into the adjacent mesentery, confirming transmural stent migration. PJ, pancreaticojejunostomy.

Figure 3.

Intraoperative photograph demonstrating transmural migration of the PJ stent. The distal portion of the pancreatic duct stent is visualized exiting the jejunal wall and extending into the surrounding mesentery. An associated extraluminal mesenteric tract is present at the site of perforation. PJ, pancreaticojejunostomy.

A controlled enterotomy was created directly at the PJ to access the migrated stent. Inspection confirmed that the perforation was located distal to the anastomosis, where the distal tip of the stent had traversed the jejunal wall and extended into the adjacent mesentery, creating an extraluminal mesenteric tract. The migrated stent was not completely removed or replaced; instead, it was withdrawn, shortened flush with the pancreatic duct, and resecured to the jejunal mucosa with a 5-0 PDS suture. The enterotomy was then closed primarily in two layers, the mesenteric defect at the site of perforation was repaired, and the abdominal cavity was thoroughly irrigated before placement of multiple drains.

Outcome and follow-up

Clinician-assessed outcomes showed resolution of sepsis and leukocytosis, while the patient reported symptomatic improvement and recovery to baseline status. Subsequent imaging demonstrated decreasing intra-abdominal collections, which were managed with adjunctive percutaneous drainage. She was discharged home in stable condition. At two-week postoperative clinic follow-up, she continued to recover well, with resolution of symptoms and removal of the remaining surgical and IR drains. At most recent follow-up three months postoperatively, the patient remained clinically well with no evidence of recurrent intra-abdominal infection. All procedures performed in this case were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Patient perspective

The patient reported significant distress during the period of clinical deterioration and uncertainty regarding the diagnosis. Following reoperation and recovery, she expressed relief at the resolution of symptoms and satisfaction with the care received. She emphasized the importance of clear communication during postoperative complications and follow-up.

Timeline

• Preoperative: diagnosis of pancreatic adenocarcinoma following EUS-guided biopsy;

• POD 0: pancreaticoduodenectomy with internal PJ stent placement;

• POD 1–3: normal drain amylase levels; low drain output;

• POD 6: discharged home with one drain in place;

• POD 8: readmission with fever, leukocytosis, abdominal pain; CT showing peripancreatic fluid and gas collection;

• POD 9: transfer and initiation of antibiotics and octreotide;

• POD 10: exploratory laparotomy with stent migration and contained perforation identified;

• Post-operative course: clinical improvement and resolution of sepsis;

• two-week follow-up: recovery progressing, drains removed;

• three-month follow-up: no recurrent complications.

Discussion

This case illustrates a rare complication of internal pancreatic duct stenting following pancreaticoduodenectomy. In this patient, transmural migration of the PJ stent through the jejunal wall resulted in a concealed mesenteric perforation and intra-abdominal sepsis despite an intact PJ. Notably, drain amylase levels remained low, initially obscuring the diagnosis and reducing suspicion for pancreatic leak. This unusual presentation highlights an important diagnostic pitfall and underscores the need to consider stent-related complications when evaluating unexplained postoperative deterioration after pancreaticoduodenectomy.

Several studies have evaluated the role of pancreatic duct stents in preventing clinically relevant POPF. A position statement from the International Study Group of Pancreatic Surgery (ISGPS) notes that while stenting may theoretically reduce anastomotic tension and improve drainage, the available evidence supporting routine use remains inconclusive (5). Meta-analyses and systematic reviews comparing stented and non-stented PJ have demonstrated inconsistent results, with some studies showing no significant difference in the incidence of clinically relevant POPF between groups (6,7). Consequently, many surgeons individualize stent use based on intraoperative factors such as pancreatic texture, duct size, and surgeon preference.

When internal pancreatic duct stents are used, they are typically intended to migrate distally into the gastrointestinal tract and pass spontaneously without clinical consequence. Studies evaluating the natural history of these stents have shown that most internal pancreatic stents eventually migrate distally, with a reported median time to spontaneous passage of approximately 454 days, although this can vary depending on stent size and configuration (8). Alternative reconstructive strategies include performing the PJ without a transanastomotic stent or using different stent calibers based on surgeon preference and patient-specific factors. In our practice, a 5-Fr internal stent was selected; absorbable pancreatic duct stents are not routinely available at our institution, and evidence supporting their routine use remains limited.

Despite this expected benign course, stent-related complications have occasionally been reported. Migration into the biliary tree is one of the more frequently described patterns and may result in cholangitis or hepatic abscess formation (9,11,12). In such cases, endoscopic retrieval is often successful, with reported success rates approaching 78%, particularly with the use of advanced endoscopic techniques such as double-balloon enteroscopy and EUS in patients with altered postoperative anatomy (13,14). Surgical intervention may still be required when endoscopic retrieval fails or when complications such as obstruction or perforation occur.

More unusual complications have also been described in isolated case reports. A case report described migration of the stent forming a stent–stone complex within the jejunum, leading to small bowel obstruction requiring operative intervention (10). A particularly rare delayed presentation described stent migration causing an entero-spinal fistula with subsequent vertebral osteomyelitis several years after pancreaticoduodenectomy (15).

Our case illustrates a distinct and rarely described mechanism of stent-related morbidity. Rather than migrating distally within the intestinal lumen, the pancreatic duct stent migrated transmurally through the jejunal wall into its mesentery. This resulted in a concealed mesenteric perforation with leakage of enteric contents and diffuse intra-abdominal contamination, despite a grossly intact PJ.

Although the precise mechanism of transmural perforation cannot be established with certainty, we hypothesize that patient-specific anatomic factors played a key role. The pancreatic duct in this patient demonstrated focal narrowing that limited advancement of the stent to approximately 1.5 cm within the duct, resulting in a relatively longer intrajejunal segment than intended. Following completion of the anterior layer of the PJ, which invaginates the jejunal wall over the pancreatic remnant, this longer intraluminal segment may have exerted increased focal pressure against the jejunal wall (Figure 4). Over time, this localized mechanical stress may have predisposed to gradual transmural erosion and subsequent perforation despite appropriate initial intraluminal placement and the use of a soft, malleable stent material.

Figure 4.

Proposed mechanism of pancreaticojejunostomy stent migration and transmural jejunal perforation. (A) Focal narrowing of the pancreatic duct limited stent advancement, resulting in a relatively longer intrajejunal segment. (B) Following completion of the anterior layer of the modified Blumgart PJ , invagination of the jejunal wall may have increased focal pressure exerted by the stent tip against the bowel wall, potentially leading to gradual transmural erosion and perforation. PJ, pancreaticojejunostomy.

Drain amylase levels remained low despite the patient’s septic presentation, initially reducing suspicion for a pancreatic leak. Because the stent had migrated extraluminally through the jejunal wall, enteric contents rather than pancreatic secretions were responsible for the intra-abdominal contamination. As a result, the typical biochemical hallmark of POPF, a markedly elevated drain amylase, was absent. In this case, when preoperative CT scan was inconclusive, intraoperative ultrasound proved critical in identifying the true etiology of the patient’s deterioration and guiding definitive management.

This case report has several strengths. It describes a rare and clinically significant complication of PJ stent migration with a clearly defined mechanism, supported by intraoperative findings and imaging correlation. The case highlights an important diagnostic pitfall, demonstrating how discordant biochemical, clinical, and radiographic findings can obscure the underlying etiology. Additionally, it emphasizes the utility of intraoperative ultrasound in identifying otherwise occult pathology. However, this report is limited by its single-patient design, which precludes assessment of incidence or generalizability. Furthermore, the rarity of this complication limits the ability to draw definitive conclusions regarding risk factors or optimal prevention strategies.

Conclusions

Although internal pancreatic duct stents are widely used and generally considered safe, this case highlights the importance of recognizing rare stent-related complications. In patients presenting with unexplained postoperative sepsis following pancreaticoduodenectomy and suspicious CT scan findings, particularly when imaging demonstrates collections near the pancreatic bed but drain amylase levels remain low, migration of an internal pancreatic duct stent should be considered in the differential diagnosis.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this case were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.