A Novel Method of Piggyback Jejunal Interposition Reconstruction Single-Tract Reconstruction After Proximal Gastrectomy for Siewert II and III Adenocarcinoma of the Esophagogastric Junction with a Diameter <4 cm

Background Digestive tract reconstruction after proximal gastrectomy (PG) has been extensively discussed. Herein, we introduce a novel method of PG with piggyback jejunal interposition reconstruction single-tract reconstruction (PJIRSTR) for Siewert II and III adenocarcinoma of the esophagogastric junction (AEG) with a diameter <4 cm, and investigate its safety, practicability, and short-term and long-term clinical outcomes of this procedure. Method The clinical data of 33 patients with Siewert II or Siewert III AEG who underwent PJIRSTR in Shanxi Cancer Hospital from July 2013 to November 2016 were retrospectively reviewed. Data of clinicopathologic characteristics, postoperative and surgical outcomes, and follow-up ndings, especially postoperative reux esophagitis and postoperative reux symptoms, were analyzed. Results The mean operation time was 136.7±22.4 (range: 110-180) min, including 32.3 ± 5.0(range: 26-45) min of the digestive tract reconstruction; the mean estimated blood loss was 87.6±18.1 mL; the mean number of dissected lymph nodes was 14.7±5.1; and the mean duration of postoperative hospitalization was 7.5±1.2 days. The early complication rate was 9% (n=3), including one case each of anastomotic bleeding, incision infection, and ileus. The late complication rate was 6% (n=2): both patients had gastroesophageal reux symptoms (Visick grade II), but only one patient had Los Angeles grade B reux esophagitis by endoscopy. Conclusion PJIRSTR is a safe, feasible, and innovative reconstruction method after PG for patients with Siewert II and III AEG with a diameter <4 cm. Furthermore, it has excellent eciency in terms of preventing reux symptoms and reux esophagitis after surgery.


Introduction
In the past few decades, the incidence of proximal gastric cancer has signi cantly increased worldwide [1][2][3]; accordingly, extensive research on adenocarcinoma of the esophagogastric junction (AEG) is also being carried out. For AEG treatment, total gastrectomy (TG) is usually preferred considering the oncological safety and the high incidence of complications such as re ux esophagitis and anastomotic strictures after proximal gastrectomy (PG). However, some studies have shown that PG and TG have similar 3-year survival rates in early AEG [4][5][6], and PG has been reported to be oncologically safe for patients with T2/T3 proximal gastric cancer [7]. In addition, PG has signi cant advantages over TG with respect to postoperative nutritional status [5][6]. However, re ux esophagitis and re ux symptoms after PG have plagued surgeons for many years. Although there are various digestive tract reconstruction methods after PG to solve this problem, no ideal method has found wide application so far owing to the ine cient re ux prevention or the technique not gaining adequate popularity for varied reasons.
Herein, we introduce a novel method called piggyback jejunal interposition reconstruction single-tract reconstruction (PJIRSTR) that can effectively prevent postoperative re ux symptoms and re ux esophagitis. To our knowledge, PJIRSTR in Siewert II and III AEG with a diameter <4 cm has not yet been reported in detail.

Patient And Methods
Patients From July 2013 to November 2016, 33 patients with Siewert II or III AEG who underwent PJIRSTR at Shanxi Cancer Hospital were selected. The inclusion criteria were as follows: (1) age range between 45 and 70 years; (2) Siewert II or III AEG con rmed by postoperative pathological ndings; (3) tumor diameter <4 cm without distant metastasis, and invasion of the lower esophagus <1 cm; and (4) no vital organ dysfunction and no surgical contraindications. Before the surgery, a routine gastroscopy was performed to assess whether the patients had re ux esophagitis and its severity. Endoscopic ndings were classi ed according to the Los Angeles Classi cation. Furthermore, the Visick score also was routinely used for all patients to evaluate gastroesophageal re ux symptoms.

Surgical procedure
After the patients were admitted to the hospital, relevant examinations and tests were completed, then the operation was performed after excluding contraindications. The same surgeon and their surgical team carried out all operations.

PG+ esophagojejunal Roux-en-Y anastomosis
For PJIRSTR, the esophagojejunal Roux-en-Y anastomosis was rst performed, and the jejunal loop was about 30 cm. On this basis, an anastomosis of the remnant stomach and jejunum was performed at a suitable position of the jejunal loop, which resembled the jejunum carrying the remnant stomach on its back.
After induction of general anesthesia, the patients were placed in the supine position. The entire procedure was carried out via a transabdominal approach, with a midline incision made to the upper abdomen. After entering the abdominal cavity, the greater and lesser omentum were released. According to the Third and Fourth edition Japanese gastric cancer treatment guidelines, D1+ or D2 lymph node dissection was performed based on the size, location, and degree of esophageal invasion of the tumor.
The digestive tract was reconstructed after PG, and at least half of the stomach was preserved. The jejunum and its mesangial blood vessels were cut at about 20-25 cm from the Treitz ligament, and the distal jejunum was raised to the end of the esophagus in front or behind the colon. The circular stapler was used to perform an end-to-side anastomosis between the esophagus and distal jejunum. The stump of the jejunum was closed with a linear closure device, and the length of the blind end <3 cm. Lateral anastomosis between the proximal and distal jejunum was performed about 30 cm away from the distal end of the esophageal jejunum anastomosis. The residual end was closed with a linear cutting closure device, and the length of the blind end was ≤3 cm.

PJIRSTR was performed after PG
The key of PJIRSTR: (i) The side-to-side anastomosis was performed between the anterior wall of the remnant stomach and the jejunum while keeping the anastomosis 3-5 cm away from the stump of the remnant stomach, which formed a structure of "arti cial gastric fundus" and anastomosis size was about 4-5cm. (ii) After anastomosis was completed, the angle formed by the long axis of the remnant stomach and the long axis of the jejunum should be <75°. (iii) The length of the interposition jejunum remains 12-15 cm. (iv) The jejunum was closed with a linear closure device at <3 cm below the anastomosis of the residual stomach jejunum to completely block the jejunum content channel and complete the PJIRSTR (Figure 1).

Postoperative care
All patients had indwelling nutrition tubes during surgery to provide nutritional support after surgery. Enteral nutrition support was provided for 1 week from day 1 after surgery; after 1 week, oral nutrition supplementation was started. The drainage tube was pulled out depending on each patient's condition. If the laboratory tests (blood test, liver, and kidney function) and upper gastrointestinal contrast 7-10 days after the operation were normal, and patients exhibited no abdominal pain or other discomfort, they were usually discharged from the hospital about 1 weeks after surgery.

Evaluation of the clinical indicators
Surgical outcomes, postoperative complications, and patient's surgical parameters (e.g., age, sex, weight, body mass index [BMI], histological type, pathologic ndings, number of retrieved and metastatic [LNs]) were recorded. Early complications (<30 days) including anastomotic bleeding, anastomosis leakage, anastomotic stenosis, incision infection, ileus, and organ dysfunction were analyzed. As for the late complications (>30 days), it mainly referred to re ux esophagitis, re ux symptoms, recannulation of intestinal tract, and changes in nutritional indicators, which were also analyzed. Visick scores were calculated for all patients via telephonic follow-up at postoperative 3, 6, and 12 months; all patients returned to the hospital for a gastroscopic review at postoperative 6 and 12 months. Gastroesophageal re ux symptoms were classi ed by Visick score, and re ux esophagitis was classi ed by the Los Angeles classi cation according to the endoscopy ndings. In addition, patients were required to return to the hospital for a review of body weight and levels of hemoglobin, total serum protein, and serum albumin levels to assess changes in nutritional indicators at postoperative 3, 6, and 12 months.

Statistical methods
Statistical analysis was performed using SPSS 26.0 software. All values were expressed as the mean ± standerd deviation (SD). The difference in complication rates and operative data between the Phase I study and the current study were analyzed using the Chi-square test and T-test, respectively. Paired-samples t-test was used to analyze the nutritional indicators. P < 0.05 was considered statistically signi cant.

Surgical outcomes and pathological ndings
All surgeries were open surgery via the abdominal transhiatal (TH) approach. Four patients underwent D1+ lymphadenectomy, and all others underwent D2 lymphadenectomy. The surgical outcomes of 33 patients are presented in Table 2. The mean operative time was 136.7±22.4 (range: 110-180) min, including 32.3 ± 5.0 (range: 26-45) min of digestive reconstruction. The intraoperative blood loss was 87.6±18.1 (range: 60-120) mL. The tumor size was 3.06±0.84 (range: 1-4) cm, and R0 resection was performed in all patients. There were no deaths or serious complications during the operation. Pathological ndings are shown in Table 3. According to the TNM staging (8th edition AJCC), three, three, ve, two, one, eight, and 12 cases were staged IA, IB, IC, IIB, IIIA, IIIB, and IVA, respectively. Histologic types

Early complications and late complications
The overall early complication rate was 9% (n=3), including one each of anastomotic bleeding, incision infection, and ileus that improved after conservative management without the need for reoperation. Anastomotic bleeding and incision infection were managed and improved by conservative treatment. The late complication rate was 6% (n=2): both patients had Visick grade II re ux symptoms (at postoperative 3 and 6 months, respectively); only one patient showed Los Angeles grade B re ux esophagitis upon endoscopy. Both patients regularly took proton-pump inhibitors (PPIs) and followed the doctor's advice and were relieved of the re ux symptoms upon review. The results of gastroscopy are shown in Figure 2. All patients showed promising imaging results upon upper gastrointestinal angiography 7-10 days after the operation with no anastomotic stenosis and no recanalization at the jejunal block. Furthemore, no recannulation of intestinal tract was observed during follow-up period. Figure 3 shows the post PG upper gastrointestinal radiography ndings, using PJIRSTR.

Recurrence and survival data
The median follow-up period was 60.7 (range: 27-90.8) months. One patient with pathological stage IVA was diagnosed with gastric stump cancer at postoperative 25 months and underwent re-surgery. The The average weight loss was 13.5%, 11.8%, and 10% at postoperative 3, 6, and 12 months, respectively.

Comparison with previous study
We compared the major surgical results and follow-up results of this study with the Phase I clinical study, and the results showed that postoperative hospital stays was signi cantly shorter(p<0.05) and there was no signi cant difference in early complications and stenosis(p> 0.05). However, the incidence of re ux symptoms and re ux esophagitis was signi cantly decreased in the current study. ( According to the Japanese gastric cancer treatment guidelines (Fifth edition), for early AEG, if more than half of the stomach after R0 resection can be preserved, PG can be selected as the surgical method [9]. However, it is still unclear whether PG can be performed for advanced AEG. Sugoor et al. [10] reported that as long as su cient surgical resection margins are ensured and enough remnant stomach is preserved, PG can be selected even for advanced proximal-third gastric cancer. Haruta et al. [11] believed that patients with advanced AEG whose tumor length is <4 cm could undergo PG.
As a functional preservation surgery, PG has some irreplaceable advantages compared with TG. First, PG ensures that food stays in the residual stomach for preliminary digestion, and then further digestion through the duodenum, which is conducive to full absorption in the jejunum and the nutritional status of patients after surgery. Second, PG provides a better pathway for further endoscopic treatment such as endoscopic retrograde cholangiopancreatography (ERCP) than TG. A recently conducted multi-center, prospective study reported that in terms of long-term e cacy, the 3-year survival rates of PG and TG were similar (96% and 92%, respectively), but the incidence of re ux esophagitis was signi cantly higher in the TG than PG group (14.5% vs. 5.4%; P=0.02) [12]. Thus, after ensuring oncological safety, if we improve the PG approach to reduce the incidence of postoperative complications such as re ux esophagitis and anastomotic stenosis to a level similar to or even lower than that of TG, PG has the potential to become the standard surgical method for AEG.
Yamashita [13] found that double-tract reconstruction (DTR), which allows food to pass through the residual stomach into the duodenum, can effectively maintain postoperative nutritional status. However, some studies indicated that dietary bolus after DTR cannot enter the duodenum through the residual stomach as scheduled, rather escape through the jejunum route, which ultimately results in nutritional results similar to TG with Roux-en-Y reconstruction (TGRY) [14,15]. In addition, a retrospective study have recently demonstrated DTR and esophagogastrostomy has similar nutritional outcomes and QOL [16].
PJIRSTR completely blocked the passage of the proximal jejunum and avoided food escaping through the jejunum. Futhermore, When food passes through the duodenum, it is fully mixed with bile and pancreatic juice, and stimulates the secretion and release of hormones such as pancreatic secretin and cholecystokinin in the gastrointestinal tract, which is conducive to the digestion and absorption of food. Compared to jejunal interposition (JI), PJIRSTR, a technique similar to RY, is relatively easy to perform. In addition, when cancer of the remnant stomach is found, PJIRSTR can deal better with the remnant stomach than JI. In terms of the length of the interposed jejunum, Tokunaga suggested that it should be <10 cm [17]. Some scholars have suggested that the ideal length is 10-15 cm[18-20], otherwise, it may cause intestinal food stasis and inconvenience during postoperative endoscopy. However, according to our clinical experience, when the length of the interposed jejunum is 12-15 cm, the above problems can be solved.
In this context, we speculated that if PJIRSTR could effectively solve long-term complications of postoperative re ux and anastomotic stenosis, it may become an attractive treatment method for Siewert II and III AEG with a diameter <4 cm. In our study, only one of the 33 patients (3%) had Los Angeles grade B re ux esophagitis as con rmed by endoscopy, and 2 (6%) patients had Visick grade II re ux symptoms. The postoperative anti-re ux effect of this method is completely comparable with some anti-re ux surgery methods that have been reported so far. For example, the incidence of re ux esophagitis after side-overlap esophagogastrostomy (SOFY) was 10%. However, the procedure requires an experienced surgeon to perform it under a laparoscope [21]. The incidence of re ux esophagitis after the double-ap technique (DFT) was 0%, and the incidence of re ux symptoms was 10% [22]. This technology has a good effect in preventing re ux esophagitis, but technical di culties and the risk of anastomotic stenosis is relatively high. Our analysis showed that PJIRSTR has better anti-re ux effect than other techniques and may be related to the following two mechanisms that have a dual anti-re ux effect. First, side-to-side anastomosis was performed between the jejunum and the anterior wall of the remnant stomach, while keeping the anastomosis 3-5 cm away from the stump of the remnant stomach. This method of anastomosis formed an "arti cial stomach fundus" structure, effectively blocking the food or digestive uid from owing into the esophagus. Second, is the anti-re ux effect of jejunum interposition; PJIRSTR completely blocked the passage of the proximal jejunum, making alkaline re ux esophagitis almost impossible.
In this study, we performed adjuvant chemotherapy for all patients whose postoperative pathology was not staged I at that time. Therefore, 23 of 33 patients underwent chemotherapy. The 3-year survival rate of patients was 90.9%; we observed that all patients who died within 3 years after surgery had stage IVA disease in pathologic staging. However, 18 of 21 patients with pathological stages of stage III and IV survived for more than 3 years. Thirteen patients survived more than 5 years at the last follow-up and are still alive. There were two patients still alive <5 years from the last follow-up. These results suggested that the pathological stage may not be the most important factor affecting the survival period of patients and PG may be feasible for Siewert II and III AEG with a diameter <4 cm, but further strong clinical data are required for con rmation.
In our study, we performed PJIRSTR, which achieved satisfactory surgical and postoperative results. No death during operation or severe perioperative complications were recorded in any of the 33 patients. Both early postoperative complications(9%) and late complications(6%) were low compared to DTR [23].For the postoperative nutritional status of the patients, all indicators decreased signi cantly within 3 months after surgery. Compared to the Phase I clinical stud, the nutritional indicators of the patients did not return to the preoperative level at postoperative 12 months, which may be related to the status of advanced tumors in some patients. In our study, the average weight loss was 13.5%, 11.8%, and 10% at postoperative 3, 6, and 12 months, respectively. Di culty maintaining weight is a typical feature after gastrectomy, which is connected with reduced stomach volume [23]. However, all the nutritional indicators have been rising after surgery.
There are several limitations to this study. First, this was a retrospective study prone to inherent bias. second, we performed a simple mean description of the patient's nutritional status indicators and did not use reliable clinical assessments for objective evaluation. Third, the sample size of the study is relatively small, and more multi-center, prospective clinical data support is needed to validate the ndings. Last, we did not consider the in uence of lymph node dissection on patients' survival period. However, we believe that although there were many methods of digestive tract reconstruction to prevent postoperative re ux, it is not necessary to establish a uni ed standard as long as the oncological safety and surgical results are guaranteed. Based on their own experience, surgeons can choose a surgical method with excellent postoperative outcomes that suits them. In addition, we believe that the current clinical research of AEG seems to have reached a bottleneck. Perhaps we should pay more attention to basic research and combine clinical research with basic research to bring the greatest bene ts to patients.
In conclusion, we report here our novel reconstruction method after PG-PJIRSTR-that shows satisfactory outcomes after the operation. Most importantly, PJIRSTR effectively addressed the problems of postoperative re ux and re ux esophagitis and improved the postoperative quality of life of patients. However, multi-center and prospective randomized trials are needed to verify its clinical application value.

Availability of data and materials
The data used to support the ndings of this study are available from the corresponding author upon request.

Ethics approval and consent to participate
This study was approved by the ethics committee of Shanxi Tumor Hospital. All patients signed informed consent.

Consent for publication
The patient was given his consent for information about himself to be published in World Journal of Surgical Oncology. Re ux symptom(n, %) 2(6%) Recannulation of intestinal tract(n, %) 0(0%)    Gastroscopic examination after the operation.

Figure 3
The upper gastrointestinal radiography ndings after PG with PJIRSTR. The contrast medium owing from the esophagus into the remnant stomach was extremely good.