Introduction

Gastric carcinoma is associated with 5-year survival rates of between 8 and 85% after R0 resection, depending on tumor stage [1, 2]. The outcome after surgical treatment is significantly influenced by postoperative mortality and morbidity (e.g., mortality has been reported from various surgical centers to be between 0 and 15%) [3–7]. Perioperative morbidity is determined by the occurrence of infectious complications (such as wound infection, pneumonia, or urinary tract infections), the number of red blood units or required surgical revisions, and the insufficiency rate at the esophagojejunal anastomosis as considerable factors. Various authors from surgical centers with adequate caseloads have reported rates ranging from 1.3 to 15.9% [8–16]. Unfortunately, data obtained in prospective or multicenter studies are rare or missing and are dominated by results from retrospective studies [9–11, 13, 15, 16].

Based on prospectively documented cases with gastrectomy due to gastric cancer in a database with a representative case number for descriptive statistics, the aims of the study were: (1) to determine the frequency and influencing factors of the insufficiency of esophagojejunal anastomosis in a multicenter clinical observational study and (2) to derive a preventive concept.

Patients and methods

The study was designed as a prospective multicenter observational study led by the Institute for Quality Control in Operative Medicine, University Hospital, Otto-von-Guericke University (Magdeburg, Germany). Data were obtained from 80 East German hospitals of all levels of clinical care, such as regional and university hospitals, through a time period of 12 months (1 January 2002 to 31 December 2002). All patients with a primary and malignant gastric tumor growth [such as gastric carcinoma, carcinoma of the esophagogastral junction, or gastrointestinal stroma tumor (GIST)] who were admitted to one of the participating surgical departments were enrolled in the study. Exclusion criteria were tumor recurrence and other histological entities.

Study participation was voluntary; no hospitals were excluded. The participating surgical departments committed a complete prospective documentation of all consecutive patients meeting the inclusion criteria independent of therapeutic modalities (conservative or operative). To control the completeness of documentation, spot checks comparing the data with lists of clinic documentation systems were performed. The rate of registration was >95%.

Precise descriptions of the study design and performance have been summarized by Meyer et al. [17].

For specific aims, all subjects who underwent total abdominal gastrectomy due to gastric carcinoma, including esophagojejunal anastomosis, were investigated.

The patients were evaluated with regard to the following:

  • Total number, age, and sex

  • Operative vs. conservative (nonoperative) treatment

  • Curative vs. palliative intention of surgical intervention

  • Various surgical procedures

  • Reconstruction

  • Type of anastomosis (Table 1).

Table 1 Characteristics of the study “Quality Control in Gastric Carcinoma 2002 (Primary Tumor Growth)”—general aspects and patients
Full size table

In particular, the patients who underwent total abdominal gastrectomy were investigated, since this group of patients formed the case series with the greatest case number and relatively homogenous characters, including the criterion of an esophagojejunal anastomosis. Anastomotic insufficiency was defined as a radiologically and/or endoscopically detectable leakage of the stapler or suture area at the esophagojejunal anastomosis and was related to the type of reconstruction of the upper gastrointestinal tract after gastrectomy, anastomosis technique, and therapeutic intention (curative vs. palliative).

Statistics

Descriptive statistics of the data were achieved using SPSS for Windows (Version 12.0). All available preoperative and intraoperative variables with possible correlation to anastomotic healing were investigated (see Table 1). Logistic regression was performed as a setup procedure. A univariate analysis of all investigated parameters was followed by a multivariate analysis to elucidate independent variables as appropriate, which influence the development of anastomotic insufficiency.

Results

Through the 12 months of 2002, 1,199 patients from 80 hospitals with gastric carcinoma, carcinoma of the esophagogastric junction, or GIST who were diagnosed by histological investigation were prospectively documented using a standardized protocol. Out of all these patients, gastric carcinoma was diagnosed in 1,139. One-thousand thirty-one individuals (90.5%) underwent surgical intervention [nonsurgical treatment, n=108 (9.5%)]. Altogether, 649 patients were approached with total gastrectomy, resulting in a gastrectomy rate of 62.9% (649/1,031). Considering the group of resected patients (n=891), the rate was 72.8%, whereas the gastrectomy rate related to the subjects with radical resection, such as subtotal resection or gastrectomy, was 79.8%. Out of the 726 resected patients with curative intention, 527 (72.6%) underwent gastrectomy (Table 2).

Table 2 Characteristics of the study “Quality Control in Gastric Carcinoma 2002 (Primary Tumor Growth)”—number of surgical interventions
Full size table

The case series of total, exclusively abdominal gastrectomy comprised 586 patients, out of whom 475 individuals (81.1%) were surgically approached with curative intention and 111 subjects were surgically approached with (18.9%) with palliative intention. The spectrum of various types of anastomoses was dominated by a Roux-Y loop (n=526; 89.8%), subdivided into 380 cases (72.2%) without a pouch and 146 cases (27.8%) with a pouch. The following most frequently used types of anastomosis were: omega loop in 27 patients (4.6%; without a pouch, n=12; with a pouch, n=15) and interposition of a small intestine segment by Longmire in 25 cases (4.3%). Stapler anastomosis dominated the profile of surgical techniques (n=550, 93.9%). Hand sewing of the anastomosis (n=36) was infrequently used: one layer, n=31 (86.1%); two layers, n=5 (13.9%) (Table 3).

Table 3 Characteristics of the study “Quality Control in Gastric Carcinoma 2002 (Primary Tumor Growth)”—total abdominal gastrectomy
Full size table

The overall insufficiency rate of esophagojejunal anastomosis in abdominal gastrectomy was 5.5% (32/586), with no significant difference between the various types of surgical reconstruction of the upper gastrointestinal tract and various techniques for the esophagojejunal anastomosis. Related to the type of surgical reconstruction of the upper gastrointestinal tract, the insufficiency rates were as follows: Roux-Y loop without a pouch, 5.3% (20/380) vs. 6.2% with pouch (9/146); interposition of a segment of the small intestine by Longmire, 0% (0/25); and omega loop with and without a pouch, n=0 (0/15) and 8.3% (1/12), respectively.

Although stapler anastomosis was insufficient in 29 of 550 cases (5.3%), hand sewing (one layer) caused an insufficiency rate of 9.7% (3 of 31 patients). In the five subjects who received two-layered hand-sutured anastomoses connecting the aboral end of the esophagus with the oral end of the jejunal loop, insufficiency or leakage was not observed. Comparing curative intention with palliative intention in surgical treatment, there was also no significant difference in the postoperative development of anastomotic insufficiency: 5.1% (24/475) and 7.2% (8/111), respectively (Table 4).

Table 4 Characteristics of the study “Quality Control in Gastric Carcinoma 2002 (Primary Tumor Growth)”—insufficiency of anastomosis
Full size table

The results of the univariate and multivariate analyses of all investigated variables (see Table 1) are summarized in Table 5. In the univariate analysis, we establish a significant connection of anastomotic insufficiency with preoperative dysphagia, gastric stenosis, positive (metastatic) lymph nodes, nicotine abuse, and multivisceral resection. In the following multivariate analysis, dysphagia and gastric stenosis proved to be independent, significantly influencing factors for the occurrence of postoperative insufficiency of the esophageal anastomosis with odds ratios of 3.408 and 3.762, respectively.

Table 5 Characteristics of the study “Quality Documentation in Gastric Cancer 2002 (Primary Tumor Growth)”—results of univariate and multivariate analyses for anastomotic insufficiency (only significant results)
Full size table

In the group of total abdominal gastrectomies, 6.7% (39/586) of the patients showed preoperative dysphagia and 5.1% (30/586) showed gastric stenosis. Patients with dysphagia suffered from an anastomotic leakage in 12.8% (5/39), whereas patients without dysphagia suffered from an anastomotic leakage in 4.9% (27/547). In addition, insufficiency rate in patients with gastric stenosis was 16.7% (5/30) and was therefore pronouncedly higher than the 4.9% (27/556) in patients without gastric stenosis.

Discussion

The insufficiency of the esophagojejunal anastomosis after gastrectomy alters considerably the overall outcome in the surgical treatment of gastric cancer. Various authors have reported insufficiency rates to be between 1.3 and 15.9% [9–12, 14–16]. In the present study, this rate is 5.7%. The much comparable German Gastric Cancer Trial (GGCS) found an insufficiency rate of 7.2% [8]. Although 19 hospitals were involved in the GGCS at the end of the 1980s, in particular and almost exclusively university hospitals (gastrectomies, n=787), 80 hospitals of various levels of clinical care participated in this study (gastrectomies, n=586).

Comparing the same parameter (the number of radical resections), gastrectomy rate is higher compared with GGCS: 72.1 vs. 47.7%; including the extended gastrectomies, the percentages are 79.8% in this study vs. 71.1% in the GGCS. Other authors report gastrectomy rates to be between 40 and 55% [18–21].

The majority of participating hospitals favored the Roux-Y loop without a pouch for the reconstruction of the upper gastrointestinal tract after gastrectomy, which was used in 380 patients (64.8%). Similarly, stapler anastomosis (93.9%) dominated hand-sewn anastomoses. Neither the comparison between the various procedures for the reconstruction of the upper gastrointestinal tract and the various anastomosing techniques after gastrectomy nor that between gastrectomies with curative and palliative intention revealed any significant difference in the occurrence of postoperative insufficiency in the esophagojejunal anastomosis. Therefore, identification of further influencing factors in order to achieve a preventive effect is of great importance in univariate and multivariate logistic regression analyses (a list of selected reports is presented in Table 6). Important factors established in these papers are advanced tumor growth, extended surgical intervention, patient’s nutritional status, and duration of the intervention. In addition, the caseload of a surgical center and the factor “surgeon” are considered as further criteria for the final outcome. However, there are only very few data on controlled studies that are available in the literature [12, 14].

Table 6 Comparison of data from the literature on insufficiency rates of esophagojejunal anastomosis
Full size table

This study elucidates that dysphagia and gastric stenosis are independent variables with a significant impact on the development of esophagojejunal anastomosis insufficiency. Both are characteristic aspects of an advanced tumor growth and underline the required early detection and diagnosis of gastric cancer.

In 52.5% (n=598) of the documented gastric cancer cases of this study, stages III and IV, according to the Union International Contre le Cancer classification, were diagnosed. In addition, dysphagia is the leading symptom of an obstruction of the upper gastrointestinal tract, which is frequently associated with a severe nutritional deficit. This might offer possible options for a preventive approach prior to gastric resection. First, the preoperative nutritional status should be objectively characterized beginning with the documentation of current nutritional aspects of a patient’s medical history and a clinical examination of nutritional status. According to the literature, this can be completed by some special laboratory parameters, such as serum albumin level and lymphocyte count as well as physiological parameters (e.g., body impedance analysis). The influence of nutritional deficit on the complication rate and postoperative outcome has been investigated several times and has resulted in a significant impact of this relevant factor [22–24]. Thus, an adapted, short-term, hyperkaloric preoperative nutritional supplementation [25] can improve the nutritional status prior to the surgical intervention and can be supportive in the prevention of esophagojejunal anastomosis insufficiency [26].