Trapezoidal tunnel intracorporeal esophagogastric anastomosis technique in laparoscopic-assisted proximal gastrectomy: technical notes and preliminary experience

Abstract

Background

Proximal gastrectomy (PG) is a curative surgical option that can also improve quality of life. Although esophagogastric anastomosis is a simple one-site procedure, reflux esophagitis remains a concern. This study presents an esophagogastric anastomosis technique that simplifies intra-abdominal manipulation by forming a trapezoidal tunnel in the residual stomach through extra-abdominal manipulation. We then evaluated the feasibility and functional outcomes of this technique.

Methods

We retrospectively analyzed clinical data from 12 consecutive patients who underwent laparoscopic PG using the trapezoidal tunnel technique between November 2017 and September 2020 at Nippon Medical School Chiba Hokuso Hospital and between October 2020 and May 2022 at Nippon Medical School Hospital. Demographic and clinical pathological characteristics, preoperative details, and postoperative outcomes were analyzed. Reflux and stenosis status was assessed via endoscopy during follow-up conducted at least 1 year after surgery.

Results

Laparoscopic PG using the trapezoidal tunnel technique was successfully performed in all 12 cases. The mean operative time was 262 (195–362) minutes (including 57 [40–89] minutes for reconstruction). No postoperative complications of Clavien–Dindo grade II or higher were observed. The mean hospital stay was 12 (10–21) days. Endoscopic findings during postoperative follow-up (61 [14–89 months]) were normal, except for one patient who developed Los Angeles classification grade A esophagitis. Notably, no anastomotic strictures were identified.

Conclusion

The trapezoidal tunnel technique is a feasible approach for anti-reflux reconstruction following laparoscopic PG that simplifies intra-abdominal surgical techniques while maintaining favorable postoperative clinical outcomes. Given the retrospective nature of the cases analyzed, further large-scale clinical trials are essential to verify the safety and efficacy of this technique.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12957-026-04396-2.

Background

The prevalence of proximal gastric cancer (PGC)—particularly early PGC—has steadily increased in Western and Asian countries over the last two decades [1]. At present, the Japanese gastric cancer treatment guidelines recommend proximal gastrectomy (PG) with D1 + lymphadenectomy for patients with early-stage gastric cancer [2]. PG is considered superior to total gastrectomy (TG) because it increases food intake by preserving postoperative gastric capacity while suppressing weight loss by avoiding the ghrelin-secreting region. Preserving the stomach also enables the secretion of intrinsic factors and absorption of vitamin B12 [3]. Therefore, guidelines have recommended PG.

The Japanese Gastric Cancer Guidelines recommend three reconstruction techniques for PG: double-tract reconstruction, jejunal interposition reconstruction, and esophago-gastric anastomosis [2]. Although double-tract and jejunal interposition are considered effective reconstruction methods for reflux esophagitis [4, 5], they are time-consuming and associated with jejunal stagnation and intestinal obstruction as late complications. There is also evidence that jejunal interposition is quite difficult to perform [6].

Esophagogastric anastomosis is a simple reconstruction technique that can be completed relatively quickly. However, esophagogastric reflux remains a common and problematic complication [7]. To address this issue, esophagogastric anastomosis with anti-reflux function has been developed. Double-flap reconstruction is considered an experienced technique for preventing postoperative reflux esophagitis; however, other suturing techniques, such as hand-sewn flap closure, have become available. This paper presents a novel technique that facilitates intraperitoneal manipulation by reducing suturing, followed by an evaluation of its feasibility and functional outcomes.

Methods

Patients and inclusion criteria

This study retrospectively analyzed clinical data from 12 consecutive patients who underwent laparoscopic PG using the trapezoidal tunnel method between November 2017 and September 2020 at Nippon Medical School Chiba Hokuso Hospital and between October 2020 and May 2022 at Nippon Medical School Hospital. Indications for the trapezoidal tunnel method include the following: (1) upper gastric cancer; (2) gastric esophageal junction adenocarcinoma (AEG) with esophageal invasion depth < 1 cm, classified as Siewert type II or III; and (3) early gastric cancer (clinical stage T1N0M0) as defined in the Japanese Gastric Cancer Pathological Classification, 3rd Edition [8]. All cases required negative resection margins and residual gastric volume of at least half of the total gastric volume. Preoperative evaluation included endoscopy and computed tomography. Patients were selected based on the above indications and intraoperative feasibility, including sufficient esophageal length to allow safe construction of the trapezoidal tunnel. Among patients undergoing PG during the same period, five with an available esophageal length of less than 5 cm for tunneling underwent double-tract reconstruction. For cases involving T2 or deeper lesions, TG was performed. All procedures were performed by a single surgeon (NS) with experience including approximately 800 open and 300 laparoscopic gastric cancer surgeries.

Operative procedures

Intraperitoneal manipulation

A working space was secured at the anastomosis site. The esophagus was pulled toward the foot using a 5 cm margin and dissected along that line using a stapler.

Extraperitoneal manipulation

A 5 cm midline abdominal wall incision was made from the umbilical region. The stomach was exposed and the specimen was cut using a linear stapler to form a 5 cm wide residual stomach. A vertical trapezoidal tunnel (3.5 cm on the lesser curvature side, 5 cm on the greater curvature side, and 2.5 cm in width) was then created (Fig. 1A). Following proximal gastrectomy, blood flow to the residual stomach originates from the right gastric artery (except for the antrum, which receives blood from the right gastric artery due to the ligation of the left gastric artery). Consequently, the most oral portion of the residual stomach, which has a lesser curvature, experiences the poorest blood supply. A greater curvature results in a more elevated esophagus. For this reason, we designed the residual stomach to be shaped like a gastric tube with a width of approximately 5 cm. This creates a vertically oriented trapezoidal tunnel shape with a longer greater curvature and stapling along this greater curvature side.

Fig. 1.

Schema of the trapezoidal tunnel technique. A Creation of a trapezoidal tunnel at the residual stomach. B The oral and anal sides of the tunnel were temporarily closed. C The posterior esophageal wall and oral end of the residual stomach were fixed at a 5 cm margin using barbed sutures. D The esophagus was passed through the residual stomach tunnel. The insertion hole of residual stomach mucosa on the anal side and the esophageal hole was stitched on both sides. E 3 cm stapling using a linear stapler. F An anastomotic foramen was created. G Continuous suturing of the insertion hole using barbed sutures. An additional stitch was reinforced to overlap the starting point of the barbed suture. H Closure of the seromuscular layer on the anal and oral sides of the tunnel

Next, saline containing 0.00001% indigo carmine (the exact concentration is irrelevant as long as the saline turns light blue) was injected into the submucosa from the serosal side using a 23-gauge needle. To facilitate anastomosis, the submucosal layer was further dissected 1 cm distally from the anal margin of the tunnel.

A small hole was made in the mucosa for the stapler insertion hole, after which 4 − 0 polydioxanone stitches were attached to both ends (Fig. 1A, red arrow). The oral and anal sides of the tunnel were then loosely and temporarily closed to prevent the protrusion of mucosa from the trapezoidal tunnel entrance and anal opening, which may obscure the view during anastomosis (Fig. 1B). Figure 2 shows the corresponding intraoperative image. Intraoperative details of the procedure are shown in the Supplementary Video.

Fig. 2.

Extraperitoneal manipulation. A Creation of a trapezoidal tunnel in the residual stomach. Intraperitoneal manipulation. B Using a linear stapler, a 3 cm anastomosis was performed between the esophagus and residual stomach. C Closure of the insertion hole with a barbed suture

Intraperitoneal manipulation

Two traction stitches were placed at the left end of the esophagus to open a hole of approximately 1 cm (Fig. 1C, red arrow). Subsequently, the seromuscular layer and posterior wall of the esophagus on the oral side of the tunnel of the residual stomach were fixed 5 cm from the edge using barbed sutures (approximately four stitches were sufficient). During this procedure, the temporary closure stitches in the oral end of the tunnel were cut (Fig. 1C). After cutting the temporary closure stitches on the anal side, the esophagus was passed through the residual stomach tunnel (Fig. 1D). The small insertion hole in the residual stomach mucosa and the insertion hole in the esophagus was then sutured using two stitches on the left and right sides to prevent separation of the esophagus and gastric mucosa during anastomosis (Fig. 1D, red arrows).

3-cm stapling using a linear stapler

A linear stapler was carefully inserted in a straight direction, ensuring it aligned with the axes of the residual stomach and esophagus at the anastomosis. (Fig. 1E). As it is sometimes difficult to align the axis from the practitioner’s right-hand port, the key is to insert the stapler through the umbilical port. This facilitates easy alignment of the residual stomach and esophageal axis, leading to a more successful anastomosis. At this time, the camera was switched from the umbilical port to the port on the surgeon’s right hand.

Creation of the anastomotic foramen (Fig. 1F)

Closure of the insertion hole (Fig. 1G)

The insertion hole was continuously sutured using barbed sutures (Fig. 1G). An additional stitch at the beginning of the barbed suture was used to prevent leakage after surgery (Fig. 1G, red arrow).

Closure of the seromuscular layer on the anal and oral sides of the tunnel with barbed sutures (Fig. 1H)

The anal and oral sides of the tunnel were closed with a barbed suture. Finally, a 2 cm wide anti-reflux valve was formed (Fig. 1H).

Statistical analysis

Statistical analysis was limited to descriptive statistics. Continuous variables are presented as medians (ranges), while categorical variables are reported as counts and percentages.

Results

Our group introduced the trapezoidal tunnel technique in November 2017. By May 2022, we had performed laparoscopic-assisted PG and esophageal stump anastomosis using the trapezoidal tunnel technique in 12 patients with gastric cancer. The characteristics of these 12 patients are presented in Table 1.

Table 1.

Clinical characteristics of patients undergoing the trapezoidal tunnel intracorporeal esophagogastric anastomosis technique (n = 12)

Clinical characteristics	Value
Age (years)	77.5 (60–87)
Sex
Male	8
Female	4
BMI (kg/m2)	23.25 (20.2–26.3)
Histological type
tub	8
por	3
sig	1
Pathological stage
IA	9
IB	2
IIB	1

Data are presented as medians (ranges) or counts

All surgeries were performed by the first author, NS, who is a qualified surgeon in gastric surgery under the Endoscopic Surgical Skill Qualification System of the Japanese Society of Endoscopic Surgery. These operations were performed at two facilities, but the same surgeon transferred to these hospitals at different times to perform this surgery. Consequently, the postoperative follow-up period also varied from 14 to 89 months. The postoperative course and complications are summarized in Table 2. The median operative time was 262 min (range 195–362 min), the median reconstruction time was 57 min (range 40–89 min), and the median tunnel formation time was 18 min (range 10–23 min). The median blood loss was 5 g (range 0–30 g). No intraoperative complications were observed and no patients required conversion to open surgery. We found no postoperative complications classified as Clavien–Dindo classification grade II or higher. The median follow-up duration was 61 months. The median postoperative weight loss rate was − 14.15% [2.3 to (-34.8)].

Table 2.

Surgical outcomes of patients undergoing the trapezoidal tunnel intracorporeal esophagogastric anastomosis technique (n = 12)

Surgical outcomes	Value
Operation time (min)	262 (195–362)
Tunnel creation time (min)	18 (10–23)
Anastomosis time (min)	57 (40–89)
Blood loss (g)	5 (0–30)
Hospital stay (days)	12 (10–21)
GERD (Los Angeles classification)
N	9
M	2
A	1
Follow-up (months)	61 (14–89)
Serum albumin (g/dL)	4.15 (3.5–4.5)
Weight loss rate (%)	−14.15 (2.3 to − 34.8)

Data are presented as medians (ranges) or counts

Postoperative esophagogastric fluoroscopy was performed in all cases. Although good inflow was observed in the upright position, reflux was adequately prevented in the 10-degree head-down position (Fig. 3).

Fig. 3.

Postoperative esophagogastric fluoroscopy. A Standing position. B 10-degree head-down position

Endoscopic findings confirmed smooth passage of an approximately 1 cm diameter oral endoscope, with no residual food was observed in the esophagus. Postoperative endoscopic dilation was not performed. One case of Los Angeles classification [9] grade A reflux was observed. However, no reflux was observed in the other patients (Fig. 4). No symptoms of stenosis or reflux were observed. Proton pump inhibitor (PPIs) were administered routinely as postoperative prophylaxis, according to institutional practice. As a result, PPIs were used in 9 of 12 cases postoperatively.

Fig. 4.

Postoperative endoscopy. A Downward view of the anastomosis. B Upward view of the anastomosis

Discussion

Over the past two decades, the prevalence of PGC—especially early PGC—has steadily increased in Western and Asian countries [1]. The Japanese Gastric Cancer Society classification defines PGC as an adenocarcinoma of the upper third of the stomach with or without esophagogastric junction involvement [8] and recommends PG with D1 + lymph node dissection, particularly in the early stages [2]. Surgical treatment of gastric cancer can provide curative outcomes and preserve postoperative function. Evidence shows good 5-year survival rates among patients who undergo PG (94%–97%). Moreover, studies comparing TG and PG report no difference in recurrence and long-term survival rates [10–12], as well as postoperative complications [10, 13–15].

At present, the Japanese Gastric Cancer Guidelines recommend three reconstruction methods for PG, namely double-tract reconstruction, jejunal interposition, and esophagogastric anastomosis [2], each of which has specific advantages and disadvantages.

Double-tract reconstruction reduces reflux esophagitis by placing the jejunum between the esophagus and the remaining stomach [4, 6]. However, it results in difficulty in performing subsequent endoscopic examination of the remaining stomach [16].

Jejunal interposition reconstruction, wherein a portion of the jejunum is inserted between the esophagus and the remaining stomach, is reported to be beneficial for reflux esophagitis [5, 6]. Similar to double-tract reconstruction, its main disadvantage is difficulty performing endoscopic examination of the remaining stomach after surgery [17]. Esophagogastric anastomosis, while quick and simple, is associated with acid reflux.

To address the challenges of these existing methods, various residual esophageal anastomosis techniques with anti-reflux valves have been developed [6, 13, 18, 19]. Among these techniques, the modified SOFY technique uses a straight stapler to anastomose the esophagus and the remaining stomach from side to side, with the preserved dorsal esophageal wall preventing reflux by applying pressure from the pseudo-fornix of the remaining stomach [20]. This approach is a more refined and efficient version of the original SOFY method [21].

The modified SOFY method creates an anastomosis in the abdominal cavity using a linear stapler and effectively prevents reflux [20, 21]. Although this technique is considered promising, Yamashita et al. reported a lower incidence of reflux esophagitis with the double-flap approach than the modified SOFY method [20]. The similarity between our presented method and mSOFY lies in the simplified intraabdominal manipulation compared to the double flap technique. Where they differ is that, in the mSOFY technique, the anti-reflux valve is created laterally relative to the esophageal longitudinal axis, and the valve structure consists of the entire layer of both the residual stomach and esophagus. In our presented method, a 2 cm anti-reflux valve is created along the longitudinal axis of the esophagus. The anti-reflux valve is composed of the mucosal layer of the residual stomach and the entire layer of the esophagus. The serosal muscle layer of the residual stomach provides additional support on the anti-reflux valve by the external compress. According to Yamashita et al. [20], 80.6% of patients require some form of postoperative antiulcer medication. In our study, no symptoms of stenosis or reflux were observed. PPIs were routinely administered as postoperative prophylaxis, according to institutional practice; as a result, PPIs were used postoperatively in 9 of 12 cases. However, PPI use may have influenced the absence of symptoms.

Notably, one Grade A case involved a 61-year-old male. Preoperative measurements indicated that he was tall (181 cm) and large (85 kg), with a large stomach. He was also Helicobacter pylori-negative with only mild atrophy, suggesting potential hyperacidity. Although PPIs were not prescribed for this case, he did not report any reflux symptoms like heartburn or regurgitation and is being observed.

Another useful technique for esophagogastric anastomosis is the double-flap technique. This approach involves creating a double-flap (2.5 cm × 3.5 cm) in the remaining stomach using a seromuscular layer, with the mucosa and esophagus sutured to create an anti-reflux valve between the posterior esophageal wall and the remaining gastric mucosa. The left and right flaps are then closed to complete the seromuscular layer covering the anterior esophageal wall [22, 23]. Although this approach has outstanding anti-reflux effects, surgeons report finding it difficult due to the need for multiple ligations with the anastomosis [22, 23]. In the double-flap technique, the separated flaps are pulled together from both sides and then sutured; advanced skills are required for delicate ligation near the esophageal fissure. As our trapezoidal tunnel technique does not require this procedure, it may simplify the surgical process. However, further studies are needed to directly compare outcomes between techniques.

The duration of surgery and anastomosis of this technique varied due to the surgical learning curve. However, the time required for trapezoidal tunnel formation was stable and consistent because it was performed from the first surgery. This result highlights the core concept of this technique: ensuring temporal stability by performing as many steps as possible externally to avoid unnecessary complexity within the abdominal cavity.

Regarding weight loss rate, a multicenter prospective study reported a significantly lower percentage of body weight loss (%BWL) 1 year after surgery for PG compared to TG (-12.8% versus − 16.9%, respectively) [19]. This suggests that weight loss with our technique (-14.15%) is greater than that observed with PG, but lower than that of TG. Notably, the aforementioned study had an average age of 68.5 years in the PG group and 67.5 years in the TG group. This is approximately 10 years younger than the average age of 77.5 years in our study. In a study that used laparoscopic total gastrectomy (LTG), Suematsu et al. reported that the rate of weight loss was significantly lower in the older adult group than in the younger group (81% vs. 84%, p = 0.004) [24]. A decline in functional recovery capacity due to the older patient age could be a contributing factor to weight loss. Blood tests conducted during the final postoperative endoscopy showed a median serum albumin level of 4.15 g/dL (range, 3.5–4.5 g/dL), which was relatively favorable.

Additionally, no postoperative strictures were observed with our technique and no endoscopic dilation procedures were performed. As mentioned earlier, endoscopic findings suggest a favorable course regarding reflux. In terms of delayed emptying, although no cases showed food residue in the esophagus during endoscopy, food residue was found in the residual stomach in two cases, suggesting this is a possibility. This weight loss may have resulted from a decrease in food intake or from impaired gastric retention.

Our method requires a 5 cm esophageal length for anastomosis to ensure anti-reflux function. During the study period, five cases (all males) could not achieve sufficient esophageal length and thus underwent cardia-side gastrectomy with double-tract reconstruction at our institutions. The median age for both methods was 73 years [range: 61–77] and the median postoperative follow-up period was 36 months [range: 21–48], with no significant age difference between groups (p = 0.40). For these five cases, the postoperative LA reflux classification was N in three cases, A in one case, and C in one case, which did not differ significantly from patients treated using the proposed approach (p = 0.19). The median weight loss was − 15.5% [range: -5.0 to -40.7], which was also similar between groups (p = 0.91). Although no significant differences were observed between these two methods, double-tract reconstruction tended to have a longer operative time and poorer outcomes regarding reflux esophagitis and weight loss. Given the study’s small sample size, further investigation is needed.

Although our method was developed to minimize intra-abdominal manipulation, its applicability requires procedures near the esophageal hiatus and a continuous suturing technique with a running suture for insertion site closure. As a result, it may require a certain level of technical expertise.

Limitations

This retrospective study involved a small number of patients treated at two facilities. All operations were performed by a single surgeon. While surgeon had previously performed approximately 800 open surgeries and 300 laparoscopic surgeries for gastric cancer, they had no experience with robot-assisted surgery. In addition, all patients were Japanese, most were elderly, and none were obese. Because the procedure was intended to be curative, the study population was limited to early-stage gastric cancer patients. Given these limitations, particularly the technique’s application by novice surgeons and the study’s selected patient population, further accumulation of cases is needed to more fully assess the feasibility of generalizing this procedure to patients with advanced gastric cancer, younger age, obesity, or non-Asian ethnicity.

Conclusion

The described trapezoidal tunnel technique is a useful option for simplifying laparoscopic-assisted PG. Long-term follow-up involving a larger number of cases is necessary to evaluate the clinical benefits, including functional and nutritional outcomes of this approach.

Abbreviations

PG

proximal gastrectomy

PGC

proximal gastric cancer

TG

total gastrectomy

Authors’ contributions

NS designed the study. NS, HK, MS, and HK performed the surgeries. NS and DK conducted patient follow-up. YN and HY supervised the study.

Funding

Not applicable.

Data availability

The datasets supporting the conclusions of this article are included within the article.

Declarations

Ethics approval and consent to participate

This study was conducted following the Declaration of Helsinki. This protocol was reviewed and approved by the Institutional Review Board of the Nippon Medical School Chiba Hokusoh Hospital (approval number 829). The requirement for obtaining individual written informed consent was waived by the institutional review board because the study involved low risk and used an opt-out process. A public notice describing the study was posted on our institutional website to allow patients to refuse participation. No patients requested exclusion. The manuscript does not contain any personal information that could identify individual patients.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.