A non-invasive protocol for the treatment of gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy

doi:10.21203/rs.3.rs-2633579/v1

Background: We aimed to study the safety and efficacy of three tubes with or without covered esophageal stent placement for the management of gastro-mediastinal or gastro-pleural fistula.

Methods: We retrospectively assessed the clinical data of 31 consecutive patients with gastro-mediastinal or gastro-pleural fistula treated by using a non-invasive protocol. Patients received three tubes (jejunal feeding tube, gastrointestinal drainage tube and abscess drainage tube) with or without esophageal covered stent placement. All patients received continue abscess drainage and nutritional support after procedure. The tubes and/or esophageal covered stents were removed after fistula healing.

Results: All patients received three tubes placement and 11 patients with lumenal narrowing received esophageal covered stent placement. Technically success was found in all patients, with no perioperative death, esophageal rupture or massive hemorrhage. Abscess cavity disappeared in 22 patients, with a clinical success rate of 71.0%. All patients received esophageal stent placement were cured and stents were removed, for the median duration of 1.6 months (IQR 1.4, 3.7). Three patients showed clinical improved, with markedly decreased abscess cavity and markedly shrunk fistula. The median survival was 30.8 months. The 1-, 3-, 5-year survival rates were 71.1%, 46.1% and 39.5%, respectively.

Conclusions: A non-invasive protocol of three tubes with or without covered esophageal stent placement is safe and effective for gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy.

Esophagogastric anastomotic leakage

Esophageal stent

Drainage

Interventional radiology

Gastric fistula

Patients with localized and/or resectable esophagogastric carcinoma are usually treated with esophagogastrectomy (1–4). Gastro-mediastinal or gastro-pleural fistula, a pathological communication between intrathoracic stomach and mediastinum or pleural cavity, is a relatively rare complication, mediastinitis or pleuritis can cause severe infections or malnutrition, and may corrode vessels and result in massive bleeding or even death.

Gastro-pleural fistula (5) or gastric airway fistula (6, 7) have been reported in literature. However, there are very few reports regarding gastro-mediastinal or gastro-pleural fistula. This complication proved to be fatal due to its corrosive actions of gastric juice, massive bleeding and nutritional debility. Cesario et al. (1) reported a case of gastric stump–pleural fistula after an Ivor–Lewis esophagogastrectomy for esophageal cancer and was successfully treated by endoscopic approach. Three-tube method and esophageal covered metallic stent have been used for esophageal fistula (3, 4, 8). In this study, we designed a non-invasive protocol for gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy, consisting of three tubes (jejunal feeding tube, gastrointestinal drainage tube and abscess drainage tube) with or without covered esophageal stent.

Patient selection

This study was approved by the Ethics Committee and Medical Records Management Department of our university. Informed consent was obtained from all the patients. All patients’ clinical data were retrospectively analyzed. This study enrolled all patients with gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy who received three tubes (jejunal feeding tube, gastrointestinal drainage tube and abscess drainage tube) with or without esophageal covered stent placement in Department of Interventional Radiology between February 2013 and July 2022. The diagnosis of a gastro-mediastinal or gastro-pleural fistula was made based on chest computed tomography and esophagogram using a water-soluble contrast agent (Figure 1; Figure 2A). Patients with gastric airway fistula, esophago-tracheal fistula, iatrogenic or spontaneous esophageal perforation were excluded from this study.

Three-tube method

All non-invasive procedures were performed under fluoroscopic guidance similar to previous reports (2, 4, 8, 9). The lidocaine gel was orally taken to anesthetize the esophagus. A 5-F catheter was transnasally inserted into the abscess cavity through the fistula orifice. The stiff guidewire tip was inserted into the bottom of the abscess cavity, along which, a 5-F pigtail or straight catheter (Cook Medical, Inc., Bloomington, IN) was placed. Continuous negative pressure drainage was performed for the abscess cavity through the abscess drainage tube. A jejunal feeding tube was introduced into the proximal segment of jejunum for nutritional support, and gastrointestinal drainage tube was inserted into the gastral cavity for drainage of gastric juice (Figure 2BC).

Esophageal covered stent placement

Patients with narrowing stomach tube received covered stent placement at the level of the anastomosis. Retrievable esophageal covered stents (Nanjing Micro-Tech Medical Company, Nanjing, China) were used in 11 patients (3, 8, 9). The median diameter of the stents was 20 mm, ranging from 18 to 22 mm. The median length of the stents was 120 mm, ranging from 80 to 160 mm. A stent delivery system was introduced along a stiff guidewire, and then released to cover the fistula orifice. The retrieval wire of the stent was left in place and the stent delivery system was withdrawn slowly. Fistula closure was confirmed by using esophagography.

Postoperative care

Enteral nutrition solution (Milupa Gmbh & Co. KG, Friedrichsdorf, Germany) was infused through the jejunal feeding tube after the procedure. The patient was permitted to feed by mouth after the sealing of fistula confirmed by esophagography. All patients received broad-spectrum antibiotic treatment before and after procedure. Physiological saline was injected and irrigated through the drainage tube twice a day. Continuous negative pressure suction was used for the drainage of abscess cavity. Esophagography and radiographic examination of the abscess cavity via drainage catheter was performed once every 5 to 7 days to show the position of abscess drainage tube and the change of abscess cavity (Figure 2D). The tube was adjusted to make sure its tip was within the abscess cavity if necessary. The covered stent and drainage tube was removed when chest CT showed complete disappearance of abscess cavity, full expansion of the lungs, with no pleural effusion (Figure 3).

General information

This study involved a total of 31 patients with gastric mediastinal or pleural fistula, including 25 men and 6 women (Table 1). The ages of the patients ranged from 52 years to 78 years, with a mean age of 64.9±7.4 years. Types of esophageal surgery included McKeown operation (n=11), Sweet procedure (n=17), and lvor lewis procedure (n=3). Akiyama stomach tube was performed in 20 patients. Eleven patients had cervical anastomosis and 20 patients had intrathoracic anastomosis. Five patients were hand sewn and 26 patients were machine sewn by mechanical staplers. Chest drainage tubes were inserted for transthoracic drainage before transfer to our department in 6 patients, and all transthoracic drainage tubes were replaced with transnasal abscess drainage tubes. All remained 25 patients received transnasal abscess drainage tubes. The median duration from surgery to fistula diagnosis was 25 days (range: 11.0 to 53.5 days). There were 23 cases of gastro-pleural fistula (74.2%), and 8 patients (25.8%) showed gastro-mediastinal fistula.

Intervention and complications

Three-tube method was used for all patients, of which, 11 patients received esophageal covered stent placement. A total of 14 covered stents were used, with a median 20 mm (IQR 18, 20mm) in diameter and 120 mm (100, 140mm) in length. All operations were technically successful, with appropriate positions and satisfactory expansion of stents or drainage of tubes. For patients received stent placement, immediate postprocedural esophagography showed all fistulae were completely blocked after stenting. No massive hemorrhage, esophageal rupture, or other complications occurred during procedures. There were 5 complications, including 2 stent migrations, 2 clogging of tube and one slipping of jejunal feeding tube. Esophageal stents were adjusted for a median time of 2.0 (IQR, 0.5 to 2.0 times). The abscess drainage tubes were adjusted or exchanged for a median time of 3.0 (IQR, 1.8 to 5.0 times). No intensive care or longer hospitalization duration was needed due to a worsening condition after interventional procedure.

Postoperative management and outcomes

All patients underwent continuous negative pressure suction via the drainage tube after procedure, and about 30 mL/day to 230 mL/day of pus was drained daily. Chest CT examination showed that the abscess cavity disappeared in 22 patients, with a clinical success rate of 71.0%. All patients received esophageal stent placement showed disappear of abscess and healing of fistula. All stents were removed once the patients are cured; the median duration from placement to removal of stent was 1.6 months (IQR, 1.4, 3.7). Three patients showed clinical improved, with markedly decreased abscess cavity and markedly shrunk fistula confirmed by esophagogram (Table 2).

Follow-up

Except 2 patients (6.7%) lost to follow-up, the remained 29 patients were followed up for a median time of 9.7 months (IQR 3.1 to 30.8 months). By the last follow-up date of September 8, 2022, 16 patients were still alive. Eleven patients were able to return to their normal living conditions without any symptom. One patient showed a shrunk fistula and abscess cavity but refused to place esophageal stent, he had a nasal feeding diet through the transnanal nutrition tube. Eleven patients died of cancer recurrence, of which, 3 patients received chemotherapy after interventional procedure. The median survival was 30.8 months. The 1-, 3-, 5-year survival rates were 71.1%, 46.1% and 39.5%, respectively.

Gastro-pleural fistula is described rarely in adults following penetrating trauma(10, 11), bariatric surgery(12, 13), malignancy (14, 15), peptic ulcer disease (15–17), and chemoradiation therapy (18). Gastro-mediastinal or gastro-pleural fistula is a relatively rare condition that occurs as a consequence of esophagogastrectomy, which could be fatal due to its corrosive actions of gastric juice, massive bleeding and nutritional debility. Stomach fistula drawn into the chest without anastomotic insufficiency is extremely rare. Thus, anastomotic insufficiency is the key factor of stomach fistula. Besides, chemotherapy and tumor recurrence may induce the formation of stomach fistula.

Cesario et al. (1) reported a case of esophageal cancer developed gastric stump–pleural fistula after an Ivor–Lewis esophagogastrectomy. This patient was successfully treated by endoscopic approach. There are no previous reports of non-invasive protocol for the treatment of this kind of complication, although three-tube method and esophageal covered metallic stent have been used for esophageal fistula (3, 4, 8).

In contrast with an esophagogastric anastomotic leakage, management of gastro-mediastinal or gastro-pleural fistula is more challenging and optimal treatment has not yet been determined. Anastomotic leakage has been treated by surgical repair, esophageal stent and three tubes placement (2, 4).

Successful management of a gastro-mediastinal or gastro-pleural fistula requires adequate and effective therapy of thoracic infection and mediastinitis by abscess drainage. We present 15 consecutive patients managed by non-invasive protocol. Abscess drainage tube placement via fistula orifice into abscess cavity is relatively difficult in a relatively large gastric cavity. In this study, a patient failed to place a abscess drainage tube into the abscess cavity by endoscopic approach, and was then placed under fluoroscopic guidance. Transnasal abscess drainage tubes were placed in 11 patients, and 4 patients with chest drainage tubes did not exchanged into transnasal drainage to reduce the fluoroscopic procedure time.

Moreover, esophageal covered stent placement was also difficult in a relatively large gastric cavity rather than esophagus for anastomotic leakage. Thus, only 4 patients with lumenal narrowing received esophageal covered stent placement in this study, considering that stent migration may not be avoided for those without lumenal narrowing.

After esophageal stent placement, it was still allowed to perform drainage from the pleural space or mediastinum by means of transnasal or transthoracic drainage tube. All patients received esophageal stent placement showed disappear of abscess and healing of fistula. All stents were removed once the patients are cured. However, three patients showed decreased fistula and abscess cavity in the remained 20 patients without esophageal stent placement. This result seemed to indicate that esophageal covered stents helps improve clinical efficacy. Besides, there were 5 complications and only one stent migration was observed.

There are some limitations in this study. This is a retrospective analysis with a relatively small patient sample. Esophageal stent may not be suitable for all patients, especially those without lumenal narrowing.

A non-invasive protocol of three tubes with or without covered esophageal stent placement is safe and effective for gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy. Esophageal covered stent may be suitable for patients with lumenal narrowing.

CT: Computerized tomography

Ethics approval and consent to participate：This study was approved by the Ethics Committee of The First Affiliated Hospital of Zhengzhou University.
Consent for publication：Not applicable.

Funding：Not applicable.
Competing interests：The authors declare that they have no competing interests.

Authors' contributions：HXW and RJZ were responsible for the conceptionand design of the study. BYH wrote the paper. BYH collected and analyzed the data. All authors discussed the results and revised the manuscript.
Availability of data and materials: The datasets used are available from the corresponding author on reasonable request.

Acknowledgements：Not applicable.

Cesario A, Galetta D, Margaritora S, Tringali A, Petruzziello L, Porziella V, et al. Persistent gastric stump-pleural fistula with circumscribed pleural empyema: a single approach for a twofold problem. Ann Thorac Surg. 2004 Sep;78(3):e56-7.
Bi Y, Zhu X, Yu Z, Wu G, Han X, Ren J. Interventional radiology protocol for treatment of esophagogastric anastomotic leakage. Radiol Med. 2019 Dec;124(12):1253-61.
Bi Y, Li J, Yu Z, Han X, Wu G. Modified Type of Double-Covered Self-Expandable Segmental Metallic Stents for Palliation of Esophageal Fistula. J Laparoendosc Adv Surg Tech A. 2019 Feb 20.
Bi Y, Wu Z, Yi M, Han X, Ren J. Three-tube method and covered metallic stent for the treatment of anastomotic leakage after esophagectomy. BMC Gastroenterol. 2020 Oct 7;20(1):330.
Darbari A, Tandon S, Singh GP. Gastropleural fistula: rare entity with unusual etiology. Ann Thorac Med. 2007 Apr;2(2):64-5.
Han X, Li L, Zhao Y, Liu C, Jiao D, Ren K, et al. Individualized airway-covered stent implantation therapy for thoracogastric airway fistula after esophagectomy. Surg Endosc. 2017 Apr;31(4):1713-8.
Li TF, Duan XH, Han XW, Wu G, Ren JZ, Ren KW, et al. Application of combined-type Y-shaped covered metallic stents for the treatment of gastrotracheal fistulas and gastrobronchial fistulas. J Thorac Cardiovasc Surg. 2016 Aug;152(2):557-63.
Bi Y, Yi M, Yu Z, Han X, Ren J. Covered metallic stent for the treatment of malignant esophageal fistula combined with stricture. BMC Gastroenterol. 2020 Jul 30;20(1):248.
Bi Y, Li J, Yi M, Yu Z, Han X, Ren J. Interventional Protocol for Treatment of Complications after Esophagojejunostomy for Esophagogastric Carcinoma. Gastroenterology research and practice. 2019;2019:1465301.
Chowdary PB, Sadashivaiah SB, Gangappa RB, Shivashankar SC. Gastro pleural fistula: a rare entity presenting as a complication of empyema thoracis following stab injury to the chest. J Clin Diagn Res. 2015 Apr;9(4):PD05-6.
Muzaffar MS, Umair B, Asghar A, Ali MZ, Hanif MS, Kamal D. Gastropleural fistula: an unusual sequel of blunt chest trauma. J Coll Physicians Surg Pak. 2009 Jul;19(7):447-9.
Nguyen D, Dip F, Hendricks L, Lo Menzo E, Szomstein S, Rosenthal R. The Surgical Management of Complex Fistulas After Sleeve Gastrectomy. Obes Surg. 2016 Feb;26(2):245-50.
Mendoza Ladd A, Al-Bayati I, Shah P, Haber G. Endoscopic closure of a gastropleural fistula. Endoscopy. 2015;47 Suppl 1 UCTN:E131-2.
Dohrmann T, Kutup A, Mahner S, Muller V, Jaenicke F, Izbicki JR, et al. Gastropleural fistula in a patient with recurrent ovarian cancer receiving combination therapy with Carboplatin, gemcitabine, and bevacizumab. J Clin Oncol. 2013 Apr 20;31(12):e208-10.
Neri A, Lambert Y, Marrelli D, Di Mare G, Mastrogiacomo D, Corso G, et al. Gastro-pleuro-pericardial fistula following combined radiation and chemotherapy for lung metastases from renal cell carcinoma: report of a case. Surg Today. 2013 Dec;43(12):1457-60.
Tzeng JJ, Lai KH, Lo GH, Hsu JH, Mok KT. Gastropleural fistula caused by incarcerated diaphragmatic herniation of the stomach. Gastrointest Endosc. 2001 Mar;53(3):382-4.
Virlos I, Asimakopoulos G, Forrester-Wood C. Gastropleural fistula originating from the lesser curve: a recognised complication, an uncommon pathway of communication. Thorac Cardiovasc Surg. 2001 Oct;49(5):308-9.
Dias OM, Barbosa CC, Teixeira LR, Vargas FS. Gastropleural fistula from gastric perforation due to renal cell carcinoma after bevacizumab chemotherapy: a case report. Clinics (Sao Paulo). 2011;66(8):1495-8.

Table 1. Patient and cancer characteristics

Characteristics	Values
Sex, male, n (%)	25 (80.6%)
Mean age, Range, years	64.9±7.4 (52-78)
Duration from surgery to fistula, Days	25 (11.0, 53.5)
Median course of disease, months	1.5 (1.0, 3.0)
Fistula type
Gastro-pleural fistula	23 (74.2%)
Gastro-mediastinal fistula	8 (25.8%)
Comorbidities, n (%)
Hypertension	5 (16.1%)
Diabetes mellitus	3 (9.7%)
Location of fistula, n (%)
Upper segment	8 (25.8%)
Middle and lower segment	15 (48.4%)
Lower segment	8 (25.8%)
Previous treatment before fistula
Preoperative chemotherapy	3 (9.7%)
Postoperative chemotherapy	4 (12.9%)
Postoperative radiotherapy	2 (6.5%)
Pre-operative laboratory tests
Leukocytes (×10⁹/L)	6.7 (5.1, 10.1)
Neutrophil (%)	68.6 (62.6, 78.4)
Maximal body temperature after fistula occurred (°C)	36.6 (36.5, 36.8)

Table 2. Clinical outcomes of interventional treatment

Variables	value
Duration from diagnosis to healing of fistula, Months	3.1 (2.0, 7.3)
Duration from placement to removal of stent, Months	1.6 (1.4, 3.7)
Interventional technique, n (%)
Three tubes placement without covered stent	20 (64.5%)
Three tubes and covered stents placement	11 (35.5%)
Transthoracic drainage tube	6 (19.4%)
Complications, n (%)
Stent migration	2 (6.5%)
Clogging of tube	2 (6.5%)
Slipping of tube	1 (3.2%)
Outcomes of fistula and abscess cavity, n (%)
Cured	22 (71.0%)
Improved	3 (9.7%)
Loss of follow up	2 (6.5%)
inefficacy	4 (12.9%)
Adjustment time of abscess drainage tube	3.0 (1.8, 5.0)
Adjustment time of covered stent	2.0 (0.5, 2.0)

No competing interests reported.

A non-invasive protocol for the treatment of gastro-mediastinal or gastro-pleural fistula after esophagogastrectomy

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Tables

Additional Declarations

Status:

Version 1