The number of laparoscopic total gastrectomies (LTGs) for gastric cancer has recently been increasing in Japan [1–4]. It has been suggested that the reduction in adhesions after laparoscopic surgery could increase the risk of internal hernia [5, 6]. Jejunojejunostomy mesenteric defect, Petersen’s space, and mesenterium of the transverse colon are known internal hernia orifices after gastrectomy with Roux-en-Y reconstruction [1, 2, 7–9]. In the case of total gastrectomy, postoperative esophageal hiatal hernia (EHH) is an important complication [1, 3, 4]. The purpose of this non-randomized, retrospective cohort study was to analyze the incidence, clinical features, and prevention of EHH after LTG for gastric cancer.

Patients and methods

Patients

Between October 2008 and July 2014, 78 patients underwent LTG for gastric cancer that was diagnosed by endoscopy and biopsy in our hospital. All operations were performed by an experienced surgeon who was also a licensed attending doctor for laparoscopic surgery. All patients in this study underwent the following standard operations: (1) LTG, which was indicated because of the location and macroscopic appearance of the primary tumor, and (2) D0, D1, D1+, or D2 lymphadenectomy as per the guideline of the Japanese Gastric Cancer Association [10].

Definition of EHH

Preoperative EHH was defined as dilation of cardia, which was diagnosed on the basis of endoscopic observation. Postoperative EHH was divided into two types: (1) the conventional type involving herniated contents other than a jejunum limb (Fig. 1) and (2) the migration type involving migration of the anastomotic staple from the esophageal hiatus (≧3 cm) (Fig. 2) [11–20]. Both diagnoses were based on computed tomography (CT) scan findings. A CT scan was performed every 6 months after surgery at minimum.

Fig. 1
figure 1

Postoperative esophageal hiatal hernia (conventional type: herniated contents other than the jejunum limb)

Full size image
Fig. 2
figure 2

Postoperative esophageal hiatal hernia [migration type: migration of anastomotic staple from the esophageal hiatus (≥3 cm)]

Full size image

Surgical technique for esophagojejunostomy

Functional end-to-end anastomosis [21]

The jejunal limb and the esophagus were excised sufficiently to allow passage of the jaws of the endoscopic linear stapler (ETS45, Ethicon Endo Surgery, blue cartridge) or (Endo-GIA 45, Covidien, purple cartridge) (Fig. 3). The stapler device was closed and fired, creating an anastomosis. These firing procedures converted the two holes into one common hole, which was then closed with one or two linear staplers (Fig. 4). Functional end-to-end anastomosis was completely established (Fig. 5).

Fig. 3
figure 3

The jejunal limb and the esophagus were excised sufficiently to allow passage of the jaws of the endoscopic linear stapler

Full size image
Fig. 4
figure 4

These firing procedures converted the two holes into one common hole, which was then closed with one or two linear staplers

Full size image
Fig. 5
figure 5

Functional end-to-end anastomosis was completely established

Full size image

Overlap method anastomosis [22]

An enterotomy was made 5 cm distal to the stapler line on the antimesenteric side of the jejunum. Another enterotomy was made on the left edge of the esophageal stump. After each fork was completely inserted into each lumen, the firing of the linear stapler converted the two openings into an entry hole to create an end-to-side anastomosis (Fig. 6). An AV-shaped anastomotic staple line between the esophagus and the jejunum was created, and intraluminal hemostasis was confirmed. The entry hole of the stapler was closed by an intracorporeal interrupted hand-sewn technique with 3-0 monofilament absorbed fiber (Fig. 7). Overlap method anastomosis was completely established (Fig. 8).

Fig. 6
figure 6

After each fork was completely inserted into each lumen, the firing of the linear stapler converted the two openings into an entry hole to create an end-to-side anastomosis

Full size image
Fig. 7
figure 7

The entry hole of the stapler was closed by an intracorporeal interrupted hand-sewn technique with 3-0 monofilament absorbed fiber

Full size image
Fig. 8
figure 8

Overlap method anastomosis was completely established

Full size image

Crus incision and repair

We divided the patients who received LTG into three groups: (1) crus preserving, (2) crus incision and repair, and (3) crus incision without repair. During LTG, the left crus of the diaphragm was widely incised for overlap anastomosis or intrathoracic functional end-to-end anastomosis (Fig. 9). For crus repair, we sutured the hiatus and anchored the jejunal limb to the hiatus using 3-0 non-absorbable monofilament fiber (Fig. 10).

Fig. 9
figure 9

During laparoscopic total gastrectomy, the crus was widely incised for overlap anastomosis

Full size image
Fig. 10
figure 10

We sutured the hiatus and anchored the jejunal limb to the hiatus using 3-0 absorbable monofilament fiber

Full size image

Statistical analysis

Patient characteristics and operative details were analyzed using the Chi-square test, the unpaired t test, and the Mann–Whitney test. The association between the intervention of the crus of the diaphragm, and the EHH was analyzed using the Chi-square test. All statistical analyses were performed using the Stata/IC (STATA Statistical Software, version 14.0; Stata Corp., College Station, TX, USA). Two-sided probability (p) values of <0.05 were considered significant.

Results

Baseline characteristics (Table 1)

Of the 78 identified patients who underwent LTG for gastric cancer, there were 51 males (65.4 %) and 27 females (34.6 %), aged 34–93 years (median: 66 years). The follow-up period after surgery was 1–73 months (median: 25 months). There was no significance in the baseline characteristics of patients who underwent LTG with respect to preoperative factors, intraoperative factors, or postoperative factors. There were 15 cases of preserving the crus, 14 cases of the incision and repair of the crus, and 49 cases of crus incision.

Table 1 Patient characteristics and operative details
Full size table

Patients with EHH after LTG (Table 2)

Of the 78 patients, seven (9.0 %) were diagnosed with postoperative EHH. All patients were diagnosed by CT. The mean interval between LTG and diagnosis of EHH was 17.6 months (1–36 months). Of the seven patients with postoperative EHH, three (42.9 %) were symptomatic and required an emergency operation for intestinal obstruction. There were no complications in the three patients who required an operation. Four patients (57.1 %) were asymptomatic and did not require an operation. In the symptomatic patients, two were of the conventional type and one had comorbid hernias (migration type and Petersen’s hernia) and underwent crus repair and jejunum fixation to crura. The patient with comorbid hernias also underwent closure of the Petersen’s hernia orifice.

Table 2 Patients of esophageal hiatal hernia after laparoscopic total gastrectomy
Full size table

EHH risk factors (Chi-square test) (Table 3)

Table 3 shows the relationship between the intervention of the crus of the diaphragm and EHH. Only crus incision without repair during LTG was associated with an increased risk in the development of postoperative EHH (0 of 29 for preserving the crus or the incision and repair of the crus vs. 7 of 49 for crus incision without repair; p = 0.033).

Table 3 Analysis of risk factor for postoperative esophageal hiatal hernia
Full size table

Discussion

Laparoscopic surgery reduces adhesions, and patients undergoing LTG are thus at risk of developing internal hernias such as EHH [5, 6] Internal hernias after laparoscopic gastrectomy, such as jejunojejunostomy mesenteric defect, Petersen’s space, and mesenterium of the transverse colon, are common and major post-operative complications of Roux-en-Y reconstruction [1, 2]. Although reports of EHH after LTG are rare [3, 4], with a reported incidence rate of 0.5 % [1], there are many papers on EHH following esophagectomy [11–20]. The incidence of EHH after esophagectomy is reported to be 0.4–13 %, though this may vary if asymptomatic EHH is included [19, 20]. There is some pathogenesis involved in postoperative EHH, including the suction effect of the negative intrathoracic pressure and positive abdominal pressure, reduced adhesions after laparoscopic surgery, and enlargement of the hiatus [19, 20, 23, 24].

EHH is difficult to diagnose because many cases are asymptomatic and the symptoms have a nonspecific and broad spectrum, including chest or abdominal pain, respiratory distress, nausea, vomiting, constipation, and gastrointestinal bleeding, all of which are easily interpreted as complications of chemotherapy [11–18]. Because a delay in diagnosis and treatment can result in intestinal necrosis with the need for resection of the small intestine, mortality rates increase from 10 to 20 to 80 % depending on the length of diagnostic delay [19, 20]. Although the accuracy has not been firmly established, CT scans are useful for evaluating hernia contents and for providing indications for operative intervention [11–20].

Some authors argue that all postoperative EHH should be repaired because of the high mortality related to delayed diagnosis and the high rate of hernia recurrence [12, 13]. In contrast, some authors hold that, in the case of small hernia, asymptomatic hernia, or a short life expectancy due to progressive cancer, a wait-and-see approach is another treatment option [15, 19, 25]. It is well accepted that patients presenting with severe complications such as obstruction or strangulation need immediate surgical repair. For surgical repair, there are several options: transabdominal or transthoracic, simple hernia closure with or without a relaxing incision in the left hemi diaphragm, mesh repair, or another repair (prolene suture web-shoelace-like pattern) [11–18, 26]. We believe that a wait-and-see approach is preferred in the case of small hernia, asymptomatic hernia, or a short life expectancy. The patient’s condition determines which repair should be performed.

The results revealed that crus incision is the only risk factor for EHH after LTG. Van Sandick concluded that only iatrogenic enlargement of the hiatus during the operation was significantly associated with an increased risk of EHH after esophagectomy [19]. In this report, our data suggest that crus incision without repair is associated with EHH after LTG. For reconstruction after LTG, especially in overlap anastomosis, some surgeons believe that the wider the crus incision, the easier it is to perform esophagojejunostomic anastomosis [22]. Previously, we cut the left crus and dissected the abdominal and lower thoracic esophagus widely in order to make anastomosis easy. As of August 2014, however, if crus incision and division of the esophagus are needed, to repair the crus, we now always attempt to suture the hiatus and anchor the jejunal limb to the hiatus using 3-0 non-absorbable monofilament fiber (Fig. 2) [27–29].

This study has several limitations. According to a previous report, the frequency of EHH after LTG was 0.5 % (1 in 218 cases); this case was symptomatic and required an operation [1]. In the present series, 9.0 % (7 of 78 patients) had radiological evidence of herniated bowel into the chest and included both symptomatic and asymptomatic cases. The actual incidence of EHH could thus be even greater than reported. This may result from a technical issue (iatrogenic enlargement of the hiatus). A second limitation is that there were too few cases to conduct multivariate analysis. Moreover, in the crus preservation and repair group, because there was no EHH, we were not able to use the logistic regression model and perform the Chi-square test to analyze the risk factors for EHH.

Conclusions

EHH after LTG is an important complication that can require an emergency operation. Our data suggested that crus incision without repair is associated with EHH after LTG. Reduced division of the crus and dissection of the esophagus is preferable, but if necessary, suture repair may help avoid postoperative EHH.