Covered Self-Expanding Metal Stents for the Treatment of Refractory Esophageal Nonvariceal Bleeding: A Case Series

Abstract

Introduction:

This study aimed to assess the safety and effectiveness of covered self-expanding metal stents for the treatment of nonvariceal esophageal bleeding in patients for whom routine therapies have failed.

Patients and Methods:

A retrospective analysis was conducted on patients with esophageal bleeding in our hospital. Data on hemostatic effects and complications were collected from patients who underwent esophageal stenting.

Results:

In total, 4 patients were treated with five stents. In all 4 patients, the placement of esophageal stents immediately stopped the ongoing bleeding. One patient experienced recurrent bleeding 4 days after the removal of the first stent. Hemostasis was achieved after the insertion of a second stent. No stent-related complications occurred during or after stent implantation in the other 3 patients.

Conclusions:

The implantation of a covered self-expandable metal stent is a safe and effective alternative to treat acute, nonvariceal esophageal bleeding after routine therapies have failed.

Introduction

Acute esophageal bleeding remains a major medical problem of the digestive tract. Although various pharmacologic and interventional methods to treat this disease are available, urgent endoscopic intervention has become the first-line therapy for the management of variceal and nonvariceal esophageal bleeding.¹ However, in 10%–20% of patients, routine endoscopic methods, such as injection or endoclips, fail to stop the bleeding or are followed by recurrent bleeding.² Therefore, a new method is required to overcome this problem.

Self-expandable stent implantation in the esophagus is a safe, minimally invasive, and effective treatment for esophageal strictures and fistulas.³ However, this treatment has not been investigated in the management of refractory nonvariceal esophageal bleeding. Here, we summarize our preliminary experience in managing patients with self-expanding, removable, polyurethane membrane-covered metal stents (SEMSs).

Patients and Methods

Records of patients with esophageal bleeding who were admitted to our department between January 2007 and June 2011 were reviewed. During this period, 512 patients with esophageal bleeding were treated at our hospital, including 21 patients with nonvariceal esophageal bleeding. In 17 of the 21 patients, hemostasis was achieved with routine methods (e.g., medicine, injection of epinephrine or sclerosants, endoclip closure, etc.). In the remaining 4 patients, routine therapies failed, and stent placement was performed.

In these four patients, an SEMS (Sigma Medical Corp., Nanjing, China) was used to arrest the acute hemorrhage when routine therapies had failed. Stents were placed according to the manufacturer's instructions. Under upper endoscopic guidance, the location and size of the lesion were determined. A stent was inserted using a unique delivery system, which was passed over a guidewire with a soft tip. The stent was required to cover the entire length of the lesion.

After definitive hemostasis was achieved, the stent was confirmed to be in the correct position by means of an immediate repeat endoscopic examination and a chest X-ray examination performed 3 days later. The aim of the second procedure was mainly to confirm that the position and dilatation of the stent were appropriate and to check whether active bleeding had stopped. Generally, patients underwent a third endoscopic examination 7 days after stent placement, to assess the lesion condition and stent position. When the patients' general condition and laboratory tests, including the hemoglobin level, had improved, endoscopic stent removal was performed under upper endoscopic monitoring.

Informed consent was obtained from all the patients prior to treatment. The institutional review board of our hospital approved this study.

Results

Clinical features

This study involved 4 patients (age range, 27–75 years), none of whom had a history of taking anticoagulants. Their clotting times were all normal. All of these patients presented with massive hematemesis, melena, and hypovolemia. In all patients, the first upper endoscopy examination was performed within 12–24 hours after stabilization of their hemodynamic status with medications. All patients were treated with a proton-pump inhibitor and octrotide in conjunction with routine endoscopic therapies. However, these therapies failed in all 4 patients, and the bleeding recurred 48–72 hours later. Therefore, stents were placed to control the bleeding. The characteristics of the patients are summarized in Table 1.

Table 1.

Characteristics of the Patients

Gender	Age (years)	Diagnosis	Distance of lesion from the incisors (cm)	Rockall score	Forrest classification	Diameter×length of stent (mm)	Number of stents	Stent retention (weeks)	Re-bleeding after stent removal	Adverse events
M	27	Diffuse erosion and massive ulcer	27–39	5	Ib	16×160	2	1 (first) 8 (second)	Yes	Distal migration of stent; death
M	75	Anastomotic ulcer	30–32	8	IIa	17×65	1	4	No	None
M	67	Ulcer	27–29	8	IIb	18×85	1	1	No	None
F	62	Ulcer	26–28	6	Ib	18×85	1	1	No	None

F, female; M, male.

Endoscopic findings

One patient presented with a massive ulcer (approximately 3×12 cm) and diffuse erosion of the middle-to-lower esophagus, with esophageal stricture (Fig. 1). One patient presented with an anastomosis ulcer in the lower third of the esophagus. This patient underwent surgery for esophageal cancer. Each of the 2 remaining patients presented with a deep ulcer in the middle third of the esophagus. In 1 patient, the ulcer was in the base of a deep diverticulum. In both patients, swollen and calcified lymph nodes were found adjacent to the esophagus on computed tomography of the chest. These 2 patients were suspected of having tuberculosis of the lymph nodes.

FIG. 1.

(a) Continual oozing of blood from the esophageal lesion. (b) Epinephrine solution (diluted 1:10,000) and endoclips could not be used to arrest the bleeding because of the diffuse erosion. (c) A 16-×160-mm (diameter×length) stent was inserted to stop the bleeding after the endoclip treatment had failed. (d) A large and extensive ulcerative erosion (approximately 3×12 cm) was found in the lower third of the esophagus after the stent was extracted 1 week later. (e) The patient had three episodes of esophageal bleeding. A second stent migrated distally 4 days after it was placed. The bleeding was stopped by repositioning the stent.

Stent placement

In total, five stents were placed. The diameters of the stents were 16–18 mm, and their lengths were 65–160 mm. All five stents were placed successfully under endoscopic monitoring. No adverse events, such as bleeding, perforation, or tears, occurred during stent placement.

Therapeutic outcomes

In all 4 patients, the massive bleeding was stopped immediately, and no recurrent bleeding was observed in the following 72 hours. Stent migration was absent on the chest X-rays taken 3 days after stent placement. All 4 patients had hemoglobin levels below 65 g/L before stent insertion. During the 72 hours after stent placement and blood transfusion, the patients showed stable hemoglobin levels exceeding 75 g/L. Changes in hemoglobin levels are shown in Table 2.

Table 2.

Hemoglobin Level at Different Time Points Before and After Stent Placement

Age (years)/		Time before stent placement		CRBC transfusion	Hemoglobin level (g/L) at time after stent placement
gender	Diagnosis	1st visit	12 hours	(mL)	12 hours	24 hours	72 hours	7 days	14 days	30 days	35 days	60 days
27/M	Diffuse erosion and massive ulcer	61	46	800	80	82	84	104	97	93 (stent removal)	72 (re-bleeding; 2nd stent placement)	45 (death)
75/M	Anastomotic ulcer	73	64	400	78	79	81	82	85	91 (stent removal)	—	98
67/M	Ulcer	85	61	600	87	87	90	93 (stent removal)	96	102	—	104
62, F	Ulcer	76	59	400	68 (CRBC transfusion 200 mL)	75	74	78	86 (stent removal)	95	—	102

CRBC, concentrated red blood cell; F, female; M, male.

Adverse events and follow-up

All of the stents were extracted under endoscopic monitoring, without any technical difficulties or adverse events, 7–30 days after their placement. Generally, patients remained in the hospital for 48–72 hours after stent removal, under observation for potential re-bleeding. We monitored their blood pressure, heart rate, defecation, and hemoglobin levels to predict if re-bleeding would occur. After the patients were discharged from the hospital, they were followed with upper gastrointestinal endoscopy every month in our outpatient clinic. During this period, they were given oral proton-pump inhibitor to cure the lesions. Two patients also received diagnostic antituberculosis treatment.

One patient had another massive hemorrhage 4 days after stent removal and required repeat stent insertion for hemostasis. This patient presented with massive hematemesis for the third time 3 days after the second stent placement. An upper gastrointestinal endoscopy performed at this time showed that the stent had migrated to the distal esophagus (Fig. 1e). The bleeding was stopped instantly by repositioning the stent under endoscopic monitoring. Unfortunately, this patient died after a fourth episode of massive hematemesis, which occurred 60 days later, 2 days after the second stent was extracted in another hospital. His diagnosis was uncertain. In the other 3 patients, stents were extracted 7–30 days after their placement. No adverse events, such as recurrent bleeding, perforation, or necrosis related to stent placement, occurred.

Discussion

Therapeutic endoscopy is the method predominantly used for the treatment of both variceal and nonvariceal esophageal bleeding.^1,4 All 4 patients in our study had acute massive bleeding and unstable hemodynamic status. Although all of them were treated with repeated injections of epinephrine or sclerosants during endoscopy, the effectiveness of this treatment was extremely limited. In 2 patients, this method failed to achieve hemostasis, and we could not identify the bleeding sites owing to the massive amount of blood covering the entire lumen. In the other 2 patients, the hemostatic effect of the injections lasted for only 1 day, and then the bleeding recurred. We considered that further injections may increase the risk of bleeding recurrence because of ulcers or erosions at the injection site. Moreover, further injections may have increased the risk of esophageal stricture or delayed perforation.⁵

Endoscopic clipping has the theoretical advantages of minimal tissue injury and high hemostatic ability.⁶ However, this method requires skill to maintain a frontal view of the exposed blood vessel. We attempted to perform clipping in the 2 patients whose lesions were located in the middle section of the esophagus but were unable to do so owing to the difficulty in assessment or insufficient visibility of the bleeding point. In some patients, the size and fibrosis of the lesions may limit the application of clips. Currently marketed endoclip devices allow only single-clip deployment and may not be suitable for ongoing massive bleeding in the esophagus.⁷ The lesion size in the other 2 patients was larger than the clip size, and the lesions were diffuse and fragile, rendering endoclips unsuitable. Therefore, we used stents to treat these patients.

Esophageal stents have been used to treat variceal esophageal bleeding.^1,8,9 The management of nonvariceal esophageal bleeding with stents has not yet been reported. We think that the hemostatic effect of stents in refractory, nonvariceal esophageal bleeding is mainly based on the following three aspects: (1) mechanical compression, (2) protection of the wound surface, and (3) antireflux protection. Furthermore, nutrition can be maintained via the oral route during the recovery period. Because the patient is able to swallow, the blood can descend into the gastrointestinal tract, thereby reducing the risk of asphyxia because of pulmonary aspiration. Moreover, these stents may be implanted immediately after the occurrence of acute bleeding without the need for X-ray assistance. The stents can be placed easily and extracted conveniently. They can be left in situ for a relatively long time before extraction, facilitating the patient's recovery and limiting early re-bleeding. Thus, this method may be appropriate for patients in whom surgery is contraindicated or impossible.

On the basis of our initial experience, we consider that stent selection must be performed using the following rules: (1) As far as possible, select a covered stent that can be extracted easily. (2) The diameter of the stent must be individualized. Generally, we choose an 18-mm-diameter stent for hemostasis. A smaller stent should be considered if stricture exists. (3) The length of the stent must be longer than that of the lesion. (4) Antireflux stents must be considered if the lesion is located in the lower part of the esophagus or cardia.

The duration of stent retention remains to be clarified. A study has shown that prolonged esophageal stenting (124–1980 days; mean, 297 days) is safe and effective for various indications.¹⁰ This duration mainly depends on the nature of the lesion. For variceal bleeding, stents are left in situ for 2–14 days.^1,4 Stent retention for too short a time may lead to early recurrent bleeding. However, longer retention times may lead to adverse events, such as migration, obstruction, and difficult removal.^10,11

This study was limited because of its small sample size; nevertheless, our initial experience showed that endoscopic SEMS placement was a well-tolerated and feasible intervention, which caused no method-related mortality or adverse events. This technique can be used as a salvage or alternative therapy after routine methods have failed. We consider this to be a promising technique, but more data are required to confirm our results.

Disclosure Statement

No competing financial interests exist.

Author Contributions

Y.Z. contributed significantly to analysis and wrote the manuscript. J.H. performed the data analyses and wrote the manuscript. X.W. helped perform the analysis with constructive discussions. D.L. contributed to the conception of the study.

References

Maufa

, Al-Kawas

. Role of self-expandable metal stents in acute variceal bleeding. Int J Hepatol, 2012; 2012:418369.

Bosch

, Berzigotti

, Garcia-Pagan

, et al. The management of portal hypertension: Rational basis, available treatments and future options. J Hepatol, 2008; 48(Suppl 1):S68–S92.

Swinnen

, Eisendrath

, Rigaux

, et al. Self-expandable metal stents for the treatment of benign upper GI leaks and perforations. Gastrointest Endosc, 2011; 73:890–899.

Escorsell

, Bosch

. Self-expandable metal stents in the treatment of acute esophageal variceal bleeding. Gastroenterol Res Pract, 2011; 2011:910986.

Park

, Yeh

, Triadafilopoulos

. Injection therapies for nonvariceal bleeding disorders of the GI tract. Gastrointest Endosc, 2007; 66:343–354.

Kapetanos

, Beltsis

, Chatzimavroudis

, et al. The use of endoclips in the treatment of nonvariceal gastrointestinal bleeding. Surg Laparosc Endosc Percutan Tech, 2009; 19:2–10.

Yeh

, Kaltenbach

, Soetikno

. Endoclips. Tech Gastrointest Endosc, 2006; 8:2–11.

Zehetner

, Shamiyeh

, Wayand

, et al. Results of a new method to stop acute bleeding from esophageal varices: Implantation of a self-expanding stent. Surg Endosc, 2008; 22:2149–2152.

Hubmann

, Bodlaj

, Czompo

, et al. The use of self-expanding metal stents to treat acute esophageal variceal bleeding. Endoscopy, 2006; 38:896–901.

10.

Schoppmann

, Langer

, Prager

, et al. Outcome and complications of long-term self-expanding esophageal stenting. Dis Esophagus, 2012; 26:154–158.

11.

Turkyilmaz

, Eroglu

, Aydin

, et al. Complications of metallic stent placement in malignant esophageal stricture and their management. Surg Laparosc Endosc Percutan Tech, 2010; 20:10–15.