Introduction

Surgery is the mainstay of gastric cancer treatment. A critical objective for gastric surgeons in both the curative and palliative settings is the development of less invasive procedures that result in an improved quality of life. Hence, laparoscopy-assisted gastrectomy is becoming increasingly prevalent. While it is considerably less invasive than the traditional laparotomy approach, the laparoscopy-assisted procedure still requires an additional incision, which is larger than a trocar site, through which to perform the anastomosis. Theoretically, however, this incision can be omitted if an intracorporeal anastomosis is performed, which, presumably, would produce less postoperative morbidity. Historically, intracorporeal anastomosis has been considered a technically difficult and unsecure procedure. However, a purely laparoscopic procedure can provide a safe anastomosis independent of the patient’s habitus. The establishment of a novel, secure, intracorporeal anastomosis, therefore, is one priority in gastric surgery. We have developed a new, simple, and secure method of intracorporeal gastroduodenostomy, using only endoscopic linear staplers, which we have named the delta-shaped anastomosis [1]. Because of its advantages, this procedure is gaining widespread acceptance among surgeons in Japan and Korea, where it is becoming a standard reconstruction procedure after laparoscopic distal gastrectomy [2–5]. In this report, we present the short- and long-term outcomes of our initial series of 100 consecutive procedures.

Patients and methods

Patients

One hundred consecutive gastric cancer patients who underwent the delta-shaped anastomosis procedure from May 2001 to November 2006 at Himeji Medical Center, Japan, were enrolled in this study.

All classifications, such as staging and lymph node numbering and grouping, were defined according to the Japanese classification of gastric carcinoma [6].

Delta-shaped anastomosis

Under general anesthesia, the patient was placed in the reverse Trendelenburg position with the legs apart. Five trocars were placed, as shown in Fig. 1. A laparoscopic distal gastrectomy was then performed with a CO2 pneumoperitoneum. Sufficient lymphadenectomy is performed and the stomach is transected. The resected specimen is removed through the extended umbilical incision, using a large plastic bag. An approximately 4-cm longitudinal incision is enough to remove the specimen, even for advanced gastric cancer. Hence, no additional incision was necessary. The extended umbilical incision normally shrinks well within a few months. Following the gastric resection, the delta-shaped anastomosis was completed as described previously [1]. Our first choice of reconstructive procedure is the Billroth-I with the delta-shaped anastomosis. In the case of a small stomach remnant or short first portion of the duodenum, in which the anastomosis would be under tension, or when gastro-esophageal reflux from a hiatal hernia is anticipated, a Roux-en-Y or Billroth-II reconstruction is selected.

Fig. 1
figure 1

Trocar placement. The endoscopic linear stapler is inserted through the left lower trocar

Full size image

The procedure for the delta-shaped anastomosis consists of the following steps. The duodenal bulb is first transected, using an endoscopic linear stapler from the posterior to the anterior wall, at a 90° angle from the usual line. This preserves the blood supply to the anastomotic line. The stomach is then divided in the usual fashion, from the greater to the lesser curvature. Small incisions are then created along the edge of the stomach and the duodenum (Fig. 2a). The posterior walls of both the stomach and the duodenum are approximated and joined with a 45-mm linear stapler (Fig. 2b). The staple line is then inspected for any defects and hemostasis is verified (Fig. 2c); the enterotomy is then closed with two applications of the linear stapler (Fig. 2d). Kocher’s maneuver is never performed as it can lead to rapid gastric transit and dumping syndrome.

Fig. 2
figure 2

Schematic outline of the delta-shaped anastomosis procedure. a Small incisions are placed in the remnant stomach and the duodenum. b The posterior walls are approximated and the stapler is fired. c Completed V-shaped anastomosis. d Closure of the common stab incision

Full size image

Postoperative follow-up

The nasogastric tube is removed immediately following the procedure and all stable patients, without significant medical co-morbidities, are transferred to the general ward for recovery. When this procedure was first implemented, oral intake was not initiated until the patient had a bowel movement. Since the introduction of a perioperative surgical care pathway for this surgery, from July 2006, patients have been started on clear liquids on postoperative day 1, and advanced to a soft diet on postoperative day 3. Once discharged, patients are followed on an outpatient basis every 6 months.

Quality of life was assessed 6 months or more postoperatively, with a questionnaire focusing on dumping syndrome symptoms and microgastria. Analysis was based on Sigstad’s score (Table 1) [7]. The clinical diagnostic indices of dumping syndrome were then calculated. In addition, the gastric remnant was evaluated endoscopically 1 year postoperatively using the RGB (Residue, Gastritis, Bile) classification (Table 2) [8].

Table 1 Relative weight of each sign or symptom of dumping syndrome [7]
Full size table
Table 2 RGB (Residue, Gastritis, Bile) classification [8]
Full size table

Results

All 100 patients in this study underwent a laparoscopic distal gastrectomy and a regional lymph node dissection. There were 66 men and 34 women. The mean age was 65.5 years. The final stage was IA in 80 patients, IB in 10, II in 4, and IIIA in 6. The lymph node dissection performed was D1 + β in 72 and D2 in 28 patients. Ninety-four patients underwent a curative potential A resection and six a B resection. The laparoscopic lymphadenectomy was nerve-sparing, as previously described [9]. Both the hepatic and celiac branches of the vagus nerve were preserved in most of the patients with early gastric cancer. Patient characteristics are shown in Table 3.

Table 3 Patient characteristics and operative records
Full size table

All operations were completed laparoscopically. Eight surgeons performed the delta-shaped anastomosis procedure. The mean time required for the anastomosis was 13 min (Table 4). Learning curves are shown in Fig. 3. All surgeons adapted quickly to the procedure. There was no intraoperative mortality. Only one case of minor leakage occurred as an anastomosis-related complication, and the patient recovered well with conservative management. Patients were able to tolerate clear liquids at a mean of 2.2 days after the procedure and a soft diet was initiated at a mean of 3.9 days after the procedure (Table 5). The mean postoperative length of hospital stay was 16.7 days; this duration was reduced to around 10 days after the perioperative surgical care pathway was incorporated in the postoperative management. Figure 4 shows the volume of oral intake tolerated at the end of the hospital stay. Most patients tolerated a solid diet, with two-thirds of the patients consuming 100% of the 1500 kcal/day soft diet at discharge.

Table 4 Operative data
Full size table
Fig. 3
figure 3

Times for anastomosis. Eight surgeons performed the delta-shaped anastomosis procedure. The time plateaued at around 10 min within a few experiences

Full size image
Table 5 Postoperative outcome
Full size table
Fig. 4
figure 4

Volume of oral intake at the end of hospital stay. Two-thirds of the patients ate 100% of the hospital diet, a 1500 kcal/day soft diet

Full size image

With a mean follow-up of 54.9 months (range 28–94 months), there were three cancer recurrences, two paraaortic lymph node recurrences, and one liver metastasis. Meanwhile, one patient died of a non-cancer related cause, and two died from non-gastric cancers. In addition to these, 13 patients were lost to follow-up, yielding a final response rate for the questionnaire of 78%. Subsequently, only one patient was diagnosed with dumping syndrome, based on a diagnostic index equal to or greater than +7. Seventy-four patients (94.9%) were diagnosed as not having dumping syndrome, based on an index equal to or less than +4 (Fig. 5). Table 6 shows the frequency distribution of postprandial symptoms. While regurgitation and eructation were rather frequent, five patients reported heartburn and only one patient required medical management of reflux.

Fig. 5
figure 5

Clinical diagnosis index of the dumping syndrome from analysis of the questionnaires that were sent to the patients more than 6 months after the operation. Border borderline dumping syndrome

Full size image
Table 6 Postprandial symptoms
Full size table

Figure 6 shows a graphical representation of the results of the endoscopic examination, which was performed in 83 patients. Two patients were diagnosed with reflux esophagitis. While 73.5% of the patients demonstrated bilious reflux into the remnant stomach, only 15 patients (23.8%) had second degree or worse remnant gastritis.

Fig. 6
figure 6

Results of the RGB (Residue, Gastritis, Bile) classification and presence of reflux esophagitis on endoscopy 1 year postoperatively

Full size image

Discussion

In this study, we have demonstrated that the delta-shaped anastomosis is a simple, easy and safe method of intracorporeal gastroduodenostomy. Eight surgeons performed all procedures as planned, without any conversion to other procedures or laparotomy. The mean time for the anastomosis was only 13 min and the learning curves were remarkably steep.

The Billroth-I procedure is generally believed to be the easiest reconstruction, as it only requires one anastomosis; however, it is also believed that complications after Billroth-I reconstruction occur more frequently than after other reconstructions and are likely to become serious [10]. In the present study, however, the rate of anastomosis-related complications was lower than what has previously been reported [2, 10], with only one patient experiencing a minor leakage, which responded to conservative management. Our experience with these 100 cases strongly suggests that the delta-shaped anastomosis is a simple and safe form of intracorporeal gastroduodenostomy.

To confirm that the oral intake in the early postoperative period was adequate, food and liquid intake was measured at the end of the hospitalization. Two-thirds of the patients ate 100% of the 1500 kcal/day soft diet. It is likely that the wide lumen of the delta-shaped anastomosis promoted oral intake in the early postoperative period. While a larger lumen facilitates the passage of solid food, it does increase the risk of dumping syndrome, particularly when the lumen is overly large. Dumping syndrome symptoms were assessed in the postoperative period with a standard symptom questionnaire. As both early and late dumping syndrome result from rapid gastric emptying, we did not distinguish between the two subtypes when analyzing Sigstad’s scores [7]. Based on the analysis of the questionnaire responses, only one patient was diagnosed with dumping syndrome. This may be explained by the following postoperative fluoroscopic and endoscopic findings. In the early postoperative period, contrast passes readily, which is consistent with the rapid resumption of adequate oral intake following surgery. Fluoroscopy a few months later, however, shows some retention of contrast material around the anastomotic site, with delayed passage. Endoscopy performed around the same time did not demonstrate any retained food and the lumen appeared to be of adequate diameter. The brief stasis evident on fluoroscopy may have been the result of duodenal twisting at the site of the delta-shaped anastomosis. This stasis may be sufficient to prevent dumping syndrome in most patients.

While only one patient was diagnosed with dumping syndrome, fifteen (19.2%) were diagnosed with gastroesophageal reflux, based on a Sigstad’s score of less than 0, and most patients reported “regurgitation” and “eructation”. Patients infrequently experienced heartburn (6.7%) and only two patients had endoscopic evidence of reflux esophagitis. On the other hand, bile reflux into the remnant stomach was observed endoscopically in 73.5% of our patients, a rate higher than what has previously been reported [8, 10, 11]. However, the incidence of remnant gastritis was low and severe symptoms were rarely observed. Thus, we concluded that while bile reflux was prevalent, it did not appear to be a clinically critical problem in our series.

A limitation of a completely laparoscopic technique is the difficulty of locating the tumor intraoperatively. Hyung et al. [12] reported that the resection line could be defined based on a combination of preoperative endoscopic clipping and intraoperative laparoscopic ultrasonography. Tanimura et al. [13] employed preoperative endoscopic inking. We used intraoperative endoscopy to identify the tumor and generally performed a total gastrectomy in any patients with tumor invasion in the upper third of the stomach. Another criticism of the laparoscopic distal gastrectomy with the delta-shaped anastomosis is its higher cost, primarily because it requires up to seven linear stapler cartridges to complete. Although this is an economically perplexing issue, this procedure has various advantages compared with other reconstructions. The smaller wound means a better cosmetic result and may lead to less invasiveness. In addition, purely laparoscopic surgery can provide a safe anasomosis independently of the patient’s habitus and brings about good postoperative respiratory function [14]. The delta-shaped anastomosis is safe in terms of minimal postoperative morbidity, and it is easy to perform, at least with the support of a well-experienced surgeon.

In conclusion, while an additional prospective clinical study should be planned to confirm the safety and efficacy of this procedure, the present review does support the idea that the delta-shaped anastomosis is simple and safe, and produces satisfactory short- and long-term results.