Skip to main content
Internal Medicine logoLink to Internal Medicine
. 2023 Jul 5;63(4):487–491. doi: 10.2169/internalmedicine.1500-22

Endoscopic Nasopancreatic Drainage Contributes to Early Resolution of Postgastrectomy Gastropancreatic Fistula

Ryogo Minami 1, Jun Nakahodo 1, Masataka Kikuyama 2, Kazuro Chiba 1, Hiroki Tabata 1, Ayu Tachibana 1, Terumi Kamisawa 1
PMCID: PMC10937146  PMID: 37407450

Abstract

A 76-year-old man experienced abdominal pain 43 days after gastric cancer resection. Computed tomography revealed a gastric wall defect extending to the pancreas, and endoscopic retrograde pancreatography revealed a gastropancreatic fistula. Afterward, a nasopancreatic duct drainage tube was inserted. Seven days later, no leakage of the contrast medium from the duct was observed, and the patient was discharged 22 days after endoscopic nasopancreatic duct drainage. Endoscopic nasopancreatic duct drainage prevents pancreatic juice leakage and promotes gastric ulcer healing.

Keywords: endoscopic nasopancreatic drainage, pancreatic fistula

Introduction

Postoperative pancreatic fistula is a common complication of pancreatoduodenectomy; however, it is occasionally observed after gastric cancer surgery. While conservative treatment is recommended in most cases, reoperation may be needed. Endoscopic nasopancreatic duct drainage (ENPD) is effective for treating pancreatic duct disruption caused by acute pancreatitis, pseudopancreatic cysts, and postoperative pancreatic effusion (1).

We herein report a case of a postoperative anastomotic ulcer and pancreatic duct fistula following gastric cancer surgery in which ENPD helped resolve the complications.

Case Report

A 76-year-old man underwent extended lymphadenectomy (D2) and distal gastrectomy (Billroth I reconstruction) for advanced gastric cancer in the prepyloric region and was discharged on postoperative day 9. The resected gastric cancer was a Union for International Cancer Control 8th edition Stage IIIB T3(SS)N1M0 lesion extending to the subserosal layer without pancreatic invasion and located contralaterally to the pancreas on the ventral side. The patient was in good condition after discharge and had no abdominal pain or fever.

However, epigastric pain developed on postoperative day 43, and the patient visited the emergency department of our hospital. Blood test results demonstrated a white blood cell count of 11,800 /μL, hemoglobin of 11.2 g/dL, C-reactive protein of 2.03 mg/dL, and pancreatic amylase of 71 IU/L. Computed tomography (CT) revealed inflammatory fat with areas of high density around the swollen pancreatic body, ascites between the stomach and pancreas, and a thin gastric wall extending from the anastomosis to the pancreatic surface (Fig. 1a, b). The pancreatic body parenchyma showed poor enhancement and mild dilatation of the upstream pancreatic duct (Fig. 1c). Upper endoscopy revealed a large, deep ulcer at the gastroduodenal anastomosis of the greater curvature of the stomach (Fig. 1d).

Figure 1.

Figure 1.

(a) Contrast-enhanced CT (axial image): dense, fatty areas were observed around the pancreatic body. The white arrowhead indicates a large depression in the gastric wall extending to the pancreatic parenchyma. (b) Contrast-enhanced computed tomography (coronal image): the white arrowhead indicates a large gastric ulcer. The black arrowhead indicates a staple at the gastroduodenal anastomosis. (c) Pancreatic body parenchyma adjacent to the gastric ulcer demonstrates poor contrast, and the main pancreatic duct of the pancreatic tail is mildly dilated. (d) Endoscopic image. A large, deep ulcer can be seen at the gastroduodenal anastomosis on the greater curvature of the stomach.

Endoscopic retrograde pancreatography (ERP) revealed leakage of the contrast medium from the pancreatic duct into the gastric lumen, which severely disrupted the duct. The patient had a fistula associated with a gastroduodenal anastomotic ulcer that perforated the pancreas and communicated with the pancreatic duct (Fig. 2a, b). An ENPD tube (5-Fr Pigtail tube; QuickPlaceV™ Model No. PBD-V803 W-05; Olympus, Tokyo, Japan) was placed in the pancreatic duct. The distal end of the tube was positioned at the pancreatic tail, and proton pump inhibitor (PPI) administration and fasting were initiated.

Figure 2.

Figure 2.

(a) Endoscopic retrograde pancreatography. Leakage of contrast medium from the pancreatic duct into the gastric lumen can be seen along with the gastroduodenal anastomotic ulcer forming a fistula that perforates the pancreas. (b) Magnification of the extrapancreatic ductal leak. (c) Pancreatography does not show leakage of contrast medium from the pancreatic duct 7 days after placement of the nasopancreatic duct drainage tube. (d) Endoscopic image at 12 days after nasopancreatic drainage. The anastomotic ulcer had resolved and become covered by regenerated mucosa.

Pancreatography was performed seven days later. No leakage of the contrast medium from the pancreatic duct was observed (Fig. 2c). CT performed immediately after pancreatography demonstrated no intragastric accumulation of the contrast medium. The ENPD tube was removed on day 12 after the procedure, and endoscopic retrograde pancreatic drainage (ERPD) (7-Fr 7-cm straight; Flexima™; Boston Scientific, Marlborough, USA) was performed. By this point, the anastomotic ulcer had healed and been covered with regenerated mucosa (Fig. 2d). An analysis of a biopsy specimen showed no malignancy. The patient was able to resume oral food intake on the same day and was discharged 22 days after ENPD (Fig. 3). The ERPD tube was removed four months later.

Figure 3.

Figure 3.

Clinical course. The pain gradually resolved after nasopancreatic drainage.

Discussion

Postoperative pancreatic effusion is generally recognized as a complication of pancreaticoduodenectomy; however, it also frequently occurs after surgery for gastric cancer. The incidence of pancreatic fistula is 2-30% after open distal gastrectomy and 0-12% after laparoscopic distal gastrectomy (2). It is one of the most serious complications associated with gastric cancer surgery and can cause intra-abdominal bleeding, malnutrition, sepsis, suture failure, and pseudoaneurysms. Complications can be fatal in some cases (3,4).

D2 lymph node dissection, as in the present case, is reportedly associated with a high risk of postdistal gastrectomy pancreatic leakage, in which compression of the pancreas by the forceps is a contributory factor (2,4,5). However, routine surgery for gastric cancer involves resection of at least two-thirds of the stomach and D2 lymph node dissection, involving dissection of the pyloric region and peripancreatic margin. Postoperative pancreatic leakage is common immediately after surgery and is often treated with fasting (3). The median length of hospital stay for the treatment of a postdistal gastrectomy pancreatic fistula was 37 days, and 28.6% of patients were hospitalized for >50 days (3). Many patients usually require prolonged hospitalization.

Two potential mechanisms underlying the fistula formation that occurred at postoperative month 2 in the present case are hypothesized. According to the first hypothesis, an anastomotic ulcer or suture failure may have formed a fistula with the pancreas. The incidence of anastomotic ulcer formation after distal gastrectomy is 0.4-3.3% (6). However, there have been no reports of a fistula facilitating the communication of the anastomotic ulcer with the pancreas. The second hypothesis holds that postoperative leakage of pancreatic fluid led to the formation of the anastomotic ulcer.

The treatment of a postoperative pancreatic fistula requires drainage of the pancreatic juice. ENPD is reportedly effective for mediastinal pancreatic pseudocysts and pancreatic effusion in chronic pancreatitis caused by pancreatic juice leakage (1). In the present case, the pancreatic leakage resolved in one week, despite the large size of the fistula, owing to the efficacy of ENPD in reducing the exposure of the Billroth I anastomotic ulcer to pancreatic juice via intragastric reflux. ENPD facilitates pancreatic ductography after implantation; in addition, it allows for rapid decision-making in the treatment of pancreatic fistulas, such as removing an ENPD tube or switching to an ERPD tube. In the present case, ERP was performed again four months after ENPD was switched to ERPD. Because pancreatic duct imaging revealed no stenosis, the ERPD tube was removed. There are no established criteria concerning the duration of ERPD tube placement. Pancreatic duct stenosis is a late complication following drainage of the pancreatic duct, and investigating the presence of duct stenosis is necessary after ERPD tube removal. Furthermore, our patient underwent periodic contrast-enhanced CT during follow-up after gastric cancer surgery. No pancreatic duct stenosis was observed at 1 year after surgery.

In cases of pancreatic duct disruption, the location of the end of the ENPD tube is important. Varadarajulu et al. reported that a stent bridging the disruption correlated with a successful outcome in the endoscopic treatment of pancreatic disruption (7). Removing the pancreatic juice secreted by the pancreatic portion distal to the main pancreatic duct injury by ENPD was effective in the present case.

In addition, a previous study reported the efficacy of ENPD in the treatment of duodenal ulcer perforation (8); controlling the secretion of pancreatic juice into the duodenum suppressed the conversion of trypsinogen to trypsin, a proteolytic enzyme, thereby preventing the progression of tissue damage at the ulcer site. Furthermore, ENPD reduces pressure in the pancreatic duct, thereby facilitating the healing of fistulas. The ulcer of the patient in the present case was thought to have resolved via a similar mechanism.

To date, there have been no reports of a fistula facilitating the communication of an anastomotic ulcer with the pancreas. However, there have been seven reports of a fistula facilitating the communication of a gastrointestinal ulcer with the pancreas (Table), and surgery was performed in five of these cases. If symptoms of peritonitis and abdominal pain do not improve, surgery is recommended (9). Shibukawa et al. treated an ulcer with a PPI, and endoscopy on hospital day 31 demonstrated a decrease in the size of the ulcer and persistence of the mucosal defect, confirming the diagnosis of an open ulcer (10). Tana et al. reported that symptoms improved after four weeks of conservative treatment with antibiotics and PPIs (11). Conservative treatments require prolonged healing periods compared with surgical approaches.

Table.

Previous Cases of Fistula Formed by a Gastrointestinal Ulcer and Communicating with the Pancreas.

Reference Sex Age (years) Symptom Location of pancreatic fistula Treatment Clinical course
(12) M 41 Anemia Posterior stomach wall Surgery The patient experienced intraoperative coagulopathy and died on hospital day 11.
(13) M 57 Anemia and hematemesis Lesser curvature of stomach Surgery Symptoms resolved after surgery
(10) F 62 Melena Greater curvature of stomach Conservative therapy with PPI Endoscopy on hospital day 31 confirmed gradual healing of the ulcer.
(11) M 49 Epigastric pain Duodenum Conservative therapy with PPI Symptom resolution after 4 weeks
(14) M 67 Epigastric pain Lesser curvature of stomach Surgery Symptoms resolved after surgery
(15) M 50s Hematemesis Posterior stomach wall Surgery Discharged on postoperative day 7
(16) M 43 Abdominal pain Duodenum Surgery Multiple debridement and washouts were required in the subsequent 4-month admission period.
Our case M 76 Abdominal pain Anastomosis of stomach ENPD Discharged on postoperative day 22

PPI: proton pump inhibitor

Pancreatic perforations caused by peptic ulcers, particularly anastomotic ulcers, are rare. In such cases, ENPD is recommended to drain the pancreatic fluid and reduce exposure of the ulcer to the fluid.

The authors state that they have no Conflict of Interest (COI).

References

  • 1. Watanabe S, Toki M, Kambayashi K, et al. Successful treatment of mediastinal pancreatic pseudocyst and pancreatic pleural effusion with endoscopic pancreatic duct drainage: a case report. DEN Open 3: e133, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Washio M, Yamashita K, Niihara M, Hosoda K, Hiki N. Postoperative pancreatic fistula after gastrectomy for gastric cancer. Ann Gastroenterol Surg 4: 618-627, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Yajima K, Iwasaki Y, Takahashi K. Incidence of and risk factors for severe pancreatic fistula formation following radical gastrectomy. Nihon Fukubu Kyukyu Igakukai Zasshi (J Abdom Emerg Med) 35: 549-555, 2015. [Google Scholar]
  • 4. Tanaka T, Takayama T, Matsumoto S, et al. Evaluation of postoperative pancreatic fistula after gastrectomy. Nihon Shokaki Geka Gakkai Zassi (Jpn J Gastroenterol Surg) 44: 657-664, 2011. [Google Scholar]
  • 5. Tsujiura M, Hiki N, Ohashi M, et al. “Pancreas-compressionless gastrectomy”: a novel laparoscopic approach for suprapancreatic lymph node dissection. Ann Surg Oncol 24: 3331-3337, 2017. [DOI] [PubMed] [Google Scholar]
  • 6. Nomura Y, Haruma K. Anastomotic ulcer. I to Cho (Stomach Intest) 52: 913-920, 2017. (in Japanese). [Google Scholar]
  • 7. Varadarajulu S, Noone TC, Tutuian R, Hawes RH, Cotton PB. Predictors of outcome in pancreatic duct disruption managed by endoscopic transpapillary stent placement. Gastrointest Endosc 61: 568-575, 2005. [DOI] [PubMed] [Google Scholar]
  • 8. Enokida K, Kikuyama M, Kurokami T, et al. [Endoscopic nasobiliary and nasopancreatic drainage contributing to healing of duodenal ulcer perforation: a case report]. Nippon Shokakibyo Gakkai Zasshi (J Jpn Soc Gastroenterol) 112: 1836-1842, 2015. [DOI] [PubMed] [Google Scholar]
  • 9. Dhruv S, Gurala D, Philipose J, Gumaste V. Gastropancreatic fistula: a rare complication of peptic ulcer disease. ACG Case Rep J 8: e00622, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Shibukawa N, Ouchi S, Wakamatsu S, Wakahara Y, Tatsumi N, Kaneko A. A rare case of gastric ulcer penetrating the pancreas that was successfully managed by conservative therapy. Intern Med 56: 2277-2279, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Tana C, Silingardi M, Giamberardino M, Cipollone F, Meschi T, Schiavone C. Emphysematous pancreatitis associated with penetrating duodenal ulcer. World J Gastroenterol 23: 8666-8670, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Hughes JJ, Blunck CE. CT demonstration of gastropancreatic fistula due to penetrating gastric ulcer. J Comput Assist Tomogr 11: 709-711, 1987. [DOI] [PubMed] [Google Scholar]
  • 13. Fujihara S, Mori H, Nishiyama N, Kobara H, Masaki T. Giant gastric ulcer penetrating into the pancreas. Arab J Gastroenterol 13: 158-160, 2012. [DOI] [PubMed] [Google Scholar]
  • 14. Dhruv S, Gurala D, Philipose J, Gumaste V. Gastropancreatic Fistula: A Rare Complication of Peptic Ulcer Disease. ACG Case Rep J 8: e00622, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Bartsokas C, Voulgaris S, Economopoulos N, Vassiliu P. Giant gastric ulcer penetrating the pancreas in a haemodynamically unstable patient. BMJ Case Rep 13: e250527, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. van der Mark M, Hartslief M. Massive emphysematous pancreatitis associated with duodenal microperforation. J Surg Case Rep 9: rjac392, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Internal Medicine are provided here courtesy of Japanese Society of Internal Medicine

RESOURCES