Gastric cancer, one of the most common cancers in the world, has a high mortality rate [1]. Despite improvements in surgical techniques and postoperative management, gastrectomy can cause many postoperative complications [2,3]. The incidence of esophagojejunostomy leakage is reportedly between 4% to 15% after total gastrectomy for gastric cancer [4,5]. Furthermore, esophagojejunostomy leakage is one of the most fatal complications following total gastrectomy. Its postoperative mortality rate can reach up to 60% [6].

Recently, a multicenter randomized clinical trial for laparoscopic versus open gastrectomy for gastric cancer (LOGICA) has been conducted successfully in the West [7], reporting a short-term mortality of 3.5% to 6.9% and anastomotic leakage rate of 8.7% to 10.0%. According to a study of risk factors for esophagojejunostomy leakage after laparoscopy-assisted total gastrectomy for gastric cancer in Japan, esophagojejunostomy leakage occurred in 13 of 131 patients (9.9%) [8]. Esophagojejunostomy leakage is one of the big issues in both the East and the West. While laparoscopic surgery is increasing, esophagojejunostomy leakage after laparoscopic surgery can be a bigger disappointment than that after open surgery in terms of patient’s surgical satisfaction.

There have been many studies on risk factors for esophagojejunostomy leakage following total gastrectomy for gastric cancer patients [8–10]. However, reports on treatment for esophagojejunostomy leakage are limited. A covered self-expandable metal stent (SEMS) has been proposed as a promising treatment for esophagojejunostomy leakage following total gastrectomy [11,12]. However, SEMS has some disadvantages and limitations.

Endoscopic monomeric n-butyl-2-cyanoacrylate (Histoacryl, B. Braun) injection has long been used to prevent bleeding from esophageal varices [13]. Histoacryl can be used surgically for skin closure [14] and mesh fixation [15]. Pramateftakis et al. [16] reported successful treatment for esophagojejunostomy leakage by endoscopic Histoacryl injection. The purpose of this study was to evaluate complications after total gastrectomy in patients with gastric cancer during the period when Histoacryl injection was performed. Therapeutic outcomes of endoscopic Histoacryl injection for esophagojejunostomy leakage were also determined.

Study design and patient enrollment

This was a single-center retrospective study using a prospectively collected database. Endoscopic Histoacryl injection at an esophagojejunostomy leakage site after total gastrectomy in patients with gastric cancer was first implemented in 2016. Between January 2016 and December 2021, a total of 205 patients who underwent total gastrectomy with Roux-en-Y esophagojejunostomy for gastric cancer (performed by MC Kim) at Dong-A University Hospital were enrolled. The indications of total gastrectomy in our institute were middle or upper one-third gastric cancer except an absolute indication of endoscopic submucosal dissection (ESD). After ESD, additional total gastrectomy was performed in patients with incomplete resection or positive lymphovascular invasion [17].

All patients who underwent laparoscopic or open surgery were managed routinely using the same clinical pathway as follows: (1) no nasogastric insertion or preoperative mechanical bowel preparation; (2) administer antibiotics once before and after surgery; (3) the use of two closed suction drains; (4) sips of water and clear liquid diet at postoperative 2 and 3 days, respectively; and (5) discharge of patients with tolerable diet and no complication on computed tomography (CT) at postoperative 8 days. The diagnosis of esophagojejunostomy leakage was made using clinical symptoms, CT, fluoroscopy, and endoscopy (Fig. 1A, B, and C).

Endoscopic injection technique

A single endoscopist at our centers performed all endoscopic procedures. Endoscopic treatment was performed using a standard upper gastrointestinal endoscope (EPK-i7010 with EG29-i10, Pentax Medical). All patients were given intravenous midazolam (0.05 mg/kg) for sedation before the endoscopic procedure. The endoscopist observed the anastomotic site and measured the size of anastomotic defect on the basis of the diameter of the endoscope. n-butyl-2-cyanoacrylate as tissue sealant (Histoacryl) and iodized oil (Lipiodol, Guerbet) were used to close the anastomotic leakage site. The endoscopist prepared a total of 1 mL with a mixed solution of 0.5 mL of Histoacryl and 0.5 mL of Lipiodol. It was then injected 1 mL at a time (Fig. 1D). The number of injected syringes depended on the size of the leakage. After injection of the mixture, additional saline was injected to detach the adhesive from the needle tip. Fluoroscopy with Gastrografin (Bayer) was performed 2 days after injection (Fig. 2A). Water, clear liquid, and soft diet were served sequentially if there was no leakage.

Data collection

Clinicopathologic characteristics and surgical outcomes such as type of surgery, operation time, postoperative hospital day, 30 days postoperative complication, Clavien-Dindo classification of complication, and mortality were investigated for 205 patients. Baseline characteristics and clinical outcomes of 10 patients with esophagojejunostomy leakage such as the duration from the first surgery to leakage, symptom for leakage, intervention for leakage, additional percutaneous catheter drainage, time to water feeding from intervention, postintervention hospital days, number of Histoacryl injections, leakage size, injection time, and survival were investigated.

Two hundred and five patients underwent total gastrectomy between January 2016 and December 2021 in Dong-A University Hospital by a single surgeon (MC Kim). Postoperative complication and mortality rates were 25.4% and 0.9%, respectively. Laparoscopic surgery was performed in almost half of the patients. One-third of patients had early gastric cancer. Mean operation time and postoperative hospital day were 186 minutes and 11 days, respectively (Table 1).

Sixty-four postoperative complications occurred in 52 patients within 30 days (Table 2). Ten (4.9%) esophagojejunostomy leakages occurred in 205 patients. Serious complications of more than Clavien-Dindo IIIb accounted for 6.3%. Eight patients underwent reoperation due to esophagojejunostomy leakages (n = 3), intraabdominal bleeding (n = 1), feeding jejunostomy site leakage (n = 1), left diaphragmatic hernia (n = 1), efferent loop adhesion (n = 1), and afferent loop perforation by an endoscope (n = 1). Two patients died of sepsis from esophagojejunostomy leakage (Table 2).

Among the 10 patients with esophagojejunostomy leakage, nine were male and one was female. Mean age and mean body mass index (BMI) were 64.6 years and 23.4 kg/m², respectively. Eight patients had comorbidities. Seven patients underwent laparoscopic surgery. Six patients had advanced gastric cancer. A palliative surgery was performed on one patient. All patients had cancer-free proximal resection margins. Mean operation time was 210 minutes. Mean postoperative hospital stay of six patients was 47 days except for two mortalities and two patients with leakage after discharge. One patient (No. 2) died of postoperative sepsis from a late diagnosis. One patient (No. 9) who had an American Society of Anesthesiologists physical status score of 4 before surgery was discharged after injection due to leakage. However, the leakage occurred again and the patient died of sepsis (Table 3).

Ten esophagojejunostomy leakages occurred at mean postoperative 11.2 days. Two patients (No. 6 and 7) had leakage after discharge after the first surgery. Clinical symptoms of patients with leakages included fever (n = 4), abdominal pain (n = 2), chest and epigastric pain (n = 1), low blood pressure (n = 1), and asymptomatics (n = 2). Two patients without symptoms were diagnosed with leakage on a routine discharge CT. Among 10 patients with esophagojejunostomy leakage, endoscopic Histoacryl injection was performed on eight patients and leakage was successfully managed with endoscopic Histoacryl injection in seven patients (87.5%). Of three patients who underwent reoperation, one was successfully treated with three times of endoscopic Histoacryl injections due to persistent postoperative leakage, one was treated after video-assisted thoracic surgery due to leakage to the thorax, and one died of sepsis after reoperation. Additional percutaneous catheter drainage was not performed for the remaining five (excluding the three who underwent reoperation and the two who had prophylactic closed drain of the first surgery). Mean postinjection hospital stay of seven successfully managed patients was 13.8 days. They were able to drink water at 1 to 6 days after injection. Among eight patients with endoscopic Histoacryl injection, six patients were injected once and two patients were injected three times. The sizes of leakages were 4 to 6 mm in eight injected patients. The injection took 3 to 10 minutes at a time (Table 4).

Although surgical technique and perioperative management are gradually improving, anastomosis site leakage after gastrectomy is a fatal and life-threatening complication. It is associated with prolonged hospital stay and poor prognosis [18–20]. In a previously reported study, anastomosis site leakage was the cause of death within 30 days for 30% of patients who underwent gastrectomy [21]. When anastomotic leakage occurs, patients may complain of symptoms such as epigastric pain, nausea, and vomiting. Patients may also present with fever, tachycardia, and/or hypotension. When a patient presents these clinical features after gastrectomy, complications of gastric surgery should be suspected. If anastomosis site leakage is suspected, it should be confirmed by CT, fluoroscopy with Gastrografin, and endoscopy.

The optimal management for esophagojejunostomy leakage is controversial. According to a review of the literature [6], three options (conservative, endoscopic, and surgical treatment) have been proposed. The size of the leakage also affects the choice of appropriate treatment and the patient’s prognosis. Conservative treatment is helpful in clinically stable patients with small leakages. Even with mild clinical symptoms, patients with stable vital signs and small anastomotic leakage (less than 10% of the circumference) may improve with conservative treatment such as fasting, total parenteral nutrition, and appropriate intravenous antibiotics for optimization of cardiopulmonary function [22,23]. Percutaneous catheter drainage for intraabdominal abscess from leakage can shorten the duration of treatment for leakage [24]. Surgical treatment can be considered as the last resort because it is related to a higher mortality rate. This treatment should be performed in the case of severe sepsis or a large defect of leakage.

Several endoscopic treatments have been suggested: self-expanding metal stent [11], fibrin glue injection and endoclips [25], Over-the-Scope-Clip [26], and endoluminal vacuum therapy [27]. Among them, self-expanding metal stents are widely accepted for treating esophagojejunostomy leakage after total gastrectomy [28]. The use of stents has been reported with variable success. However, stents for treating esophagojejunostomy leakage can have several limitations. It can cause stent migration and perforation of leakage site. It can also cause chest or epigastric pain, leading to its removal. Traditionally, endoscopic monomeric n-butyl-2-cyanoacrylate (Histoacryl) injection has long been used effectively to prevent bleeding from esophageal varices [13]. Recently, Histoacryl application has been used safely in various fields such as skin closure [14], gastrointestinal varices or ulcer [29], meniscal repair [30], endovascular leakage [31], placental neoplasm [32], and keratitis [33]. Endoscopic treatment of recurrent tracheoesophageal fistula with Histoacryl has also been reported [34]. Histoacryl seals leakage with a mechanism that can improve vascularity and healing by promoting an inflammatory response immediately after contact with body fluids [16]. In a previous report, sealants were used to treat postoperative leakage. Histoacryl was successfully used to repair esophagojejunostomy leakage after 3 weeks of successful conservative treatment [16].

Anastomosis insufficiency after gastrectomy usually occurs at 2 to 7 days after surgery [35]. In the present study, the mean duration from the first surgery to leakage was 11.2 days. The recent use of various surgical materials might have led to a late occurrence of leakage. The size of delayed leakage after surgery seemed to be smaller than the size of immediate leakage (2–3 days) after surgery.

In our institute, endoscopic Histoacryl injection was first attempted in 2016 in stable patients with small esophagojejunostomy leakage who had not been treated spontaneously for postoperative 39 days. Endoscopic Histoacryl injection was performed in eight patients. Of them, seven (87.5%) were successfully managed without injection-related complications. Unfortunately, one patient was discharged after confirming that there was no leakage in fluoroscopy with Gastrografin after a single injection, but the leak recurred and the patient died of sepsis. Therefore, additional CT scans and blood tests are required before the patient’s discharge because fluoroscopy with Gastrografin cannot accurately detect leakage after injection.

According to a systemic review [28], an overall success rate of 88% was reported among 371 SEMS-treated patients with a hospital mortality rate of 7.5%. SEMS was proposed as the main treatment option in patients with a leaking esophagus. This result of endoscopic Histoacryl injection is comparable to that of SEMS for the treatment of leakage. Furthermore, this method can be simple and effective in stable patients with small-sized esophagojejunostomy leakage after total gastrectomy. The advantage of this treatment is that the patient’s compliance is very high. In addition, it can be treated repeatedly. Moreover, there is no pain before or after the procedure. Furthermore, no additional drainage is required and water can be consumed in a short period of time at a low cost. In this study, the mean hospital stay of seven successfully injected patients was 13.8 days. Patients who were treated by endoscopic Histoacryl injection recovered faster than those receiving SEMS, although there was a difference in the degree of patient’s leakage between the two groups. Endoscopic Histoacryl injection might also be applicable to various leakages such as stapler line leakage of sleeve gastrectomy or artificial lesser curvature of subtotal gastrectomy and gastrojejunostomy leakage.

However, this study has some limitations. This is a retrospective and single-center study with a small number of cases. This treatment (endoscopic Histoacryl injection) cannot be applied to all patients with esophagojejunostomy leakage. In addition, the treatment protocol for endoscopic Histoacryl injection has not been standardized yet. Thus, this treatment needs to be further studied by many researchers. A prospective randomized and multicenter study is also needed in the future.

In conclusion, the overall morbidity and mortality of 205 total gastrectomies in patients with gastric cancer during 6 years were comparable to those of previously published studies. While a variety of treatment options are available, early detection and multidisciplinary approaches are keys to achieving successful outcomes regardless of the strategy adopted. The current experience in our hospital suggests that endoscopic Histoacryl injection for esophagojejunostomy leakage after total gastrectomy can be considered as a useful treatment for some selected cases.

Ethical statements

Patients with endoscopic injections received sufficient explanation and consent regarding the endoscopic procedure. This study was approved by the Institutional Review Board of Dong-A University Hospital with a waiver of informed consent (No. DAUHIRB-22-102). All authors had access to the study data.

Author’s contributions

Conceptualization, Formal analysis, Methodology, Visualization: MCK, MK

Data curation, Investigation: MCK, SS

Writing–original draft: MCK

Writing–review & editing: all authors

All authors read and approved the final manuscript.

Conflict of interest

All authors have no conflicts of interest to declare.

Funding/support

This study was supported by the Dong-A University Research Fund.

Clinicopathologic characteristics and surgical outcomes of 205 total gastrectomy patients with gastric cancer

Thirty days postoperative complications of 205 total gastrectomy patients with gastric cancer

Baseline characteristics of 10 patients with esophagojejunostomy leakage

Clinical outcomes of 10 patients with esophagojejunostomy leakage