Application of Hao’s Esophagogastrostomy by Fissure Technique (HEFT) in proximal gastrectomy: protocol for a prospective, multicentre, randomised controlled study

Abstract

Background

Proximal gastrectomy (PG) has emerged as the preferred surgical approach for adenocarcinoma of the upper 1/3 stomach and selected cases of oesophagogastric junction adenocarcinoma. We developed a novel oesophagogastric anastomosis technique with an antireflux mechanism (Hao’s Esophagogastrostomy by Fissure Technique). It may have a superior effect on patient weight maintenance compared with the double-tract reconstruction. We intend to conduct a prospective, multicentre, randomised controlled clinical trial to validate this hypothesis.

Methods and analysis

The primary objective evaluates body weight loss at 12 months postoperatively. Secondary objectives assess surgical safety through comprehensive analysis of complication rates and nutritional parameters, including serial haematological evaluations during follow-up. The study will enrol 52 participants across multiple centres with planned 3-year longitudinal monitoring to evaluate both immediate postoperative outcomes and intermediate-term clinical impacts.

Ethics and dissemination

This study was approved by the hospital institutional review board of Huashan Hospital, Fudan University (2024-1173) and is being conducted in accordance with the Declaration of Helsinki and the Good Clinical Practice guidelines. On completion of the study, the results will be published in a peer-reviewed journal.

Trial registration number

NCT06679244.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a prospective randomised controlled clinical trial.

• This expanded trial is designed based on preliminary data obtained from the lead centre.

• The subject of the study is a novel, promising anastomosis technique.

• Body weight loss serves as an objective and highly primary endpoint.

• As this novel technique risks bias from centre variations in surgical duration and the study was unmasked, the objective quality-of-life indicators require large-scale confirmation.

Introduction

The incidence of gastric cancer has been declining globally in the past few decades. However, the incidence of proximal gastric cancer (upper 1/3 of the stomach) and adenocarcinoma of the oesophagogastric junction has increased significantly.^{1 2} Total gastrectomy (TG) remains a conventional surgical approach, yet its associated nutritional and metabolic sequelae—particularly malnutrition and anaemia—are inevitable, significantly impairing QOL (Quality of life) in long-term survivors such as those with early-stage tumour.^{3 4} With improved understanding of lymphatic metastasis patterns and oncological principles, proximal gastrectomy (PG) has gained clinical traction.⁵,⁹ The development of antireflux reconstruction techniques now enables PG to preserve partial gastric function while mitigating severe postoperative reflux oesophagitis. For upper gastric cancer or adenocarcinoma of the oesophagogastric junction with favourable prognoses, the ideal approach should retain the distal stomach to optimise QOL and employ physiologically sound reconstruction to prevent reflux.¹⁰

Currently, no consensus exists on optimal post-PG reconstruction, with debates centring on reflux oesophagitis incidence and residual gastric utilisation. Among existing methods, double-flap technique (DFT) and double-tract reconstruction (DTR) are the most recognised.^{11 12} DFT, introduced by Kamikawa, demonstrates validated antireflux efficacy and is widely adopted in Japan and Korea. However, its technical complexity, prolonged operative time and high anastomotic stenosis rates limit laparoscopic implementation.¹³ DTR, proposed by Aikou in 1988, requires less residual gastric volume and bypasses delayed gastric emptying via jejunal pathways. Yet, most ingested food bypasses the remnant stomach, negating its functional role, while multiple anastomoses increase procedural complexity and costs.¹⁴,¹⁶ These limitations underscore the need for simpler, more effective PG reconstruction methods.

Since 2013, our team has pioneered totally laparoscopic PG, recognising that effective antireflux mechanisms require both ‘pseudofundus’ reconstruction and anastomotic valve formation. The former establishes reflux resistance, while the latter prevents oesophagitis. Through iterative refinement, we developed Hao’s Esophagogastrostomy by Fissure Technique (HEFT). Unlike DFT, HEFT eliminates complex muscular flap construction, reduces technical difficulty and enables full laparoscopic completion. By minimising muscular wrapping around the anastomosis, HEFT theoretically lowers stenosis risk. The ‘pseudofundus’ design and dynamic valve mechanism create a unique physiological state, maintaining effective antireflux functionality during both feeding and fasting without compromising food intake. This approach preserves natural anatomical pathways, maximises meal capacity and enhances nutritional support.^{17 18}

Preliminary single-centre studies confirm HEFT’s safety and feasibility as a functional oesophagogastric anastomosis.¹⁹ Building on these findings, we designed this prospective, multicentre, randomised controlled trial comparing HEFT with DTR in laparoscopic PG. The study evaluates postoperative nutritional status, short-term/medium-term safety and survival outcomes to identify optimal reconstruction strategies and advance minimally invasive gastric cancer treatment. The surgical steps for HEFT are illustrated in figure 1.

Figure 1. Surgical steps for Hao’s Esophagogastrostomy by Fissure Technique. (a) The position of the fissure in the anterior wall of the stomach, the fixed position of the oesophagus and stomach and the position for full-thickness incision of the gastric wall at the distal end of the fissure. (b) Single-layer continuous suture of the oesophageal posterior wall and the gastric mucosa. (c) Single-layer continuous suture between the anterior wall of the oesophagus and the entire layer of stomach to reconstruct ‘pseudofundus’. (d) Fold the ‘pseudofundus’ 270° and rotate it around the lower end of the oesophagus.

Methods and analysis

Trial design

This study employs a prospective, multicentre, randomised, open-label, superiority trial design across three participating institutions. A total of 52 eligible patients will be enrolled and randomised into a 1:1 ratio via centralised minimisation-based dynamic randomisation to either the experimental group (receiving HEFT) or the control group (DTR). Baseline patient characteristics will be recorded preoperatively, with perioperative safety profiles collected within 7 days postoperatively. In addition to standard postoperative surveillance for patients with gastric cancer, protocol-specific evaluations at 12 months will include: body weight measurements, haemoglobin and serum albumin levels and endoscopic findings (assessing anastomotic integrity and reflux complications). Longitudinal follow-up will continue until 36 months postoperatively to document overall survival (OS) and disease-free survival (DFS). The trial design is depicted as a ‌flow chart‌ in figure 2.

Figure 2. Study flowchart. BWL, body weight loss; DFS, disease-free survival; DTR, double-tract reconstruction; HEFT, Hao’s Esophagogastrostomy by Fissure Technique; OS, overall survival; PG, proximal gastrectomy.

Study setting

The clinical trial will be conducted at three hospitals in Shanghai, China: ‌Huashan Hospital‌, Fudan University (lead institution), responsible for ‌patient enrolment (26 cases)‌, ‌data collection‌, ‌final data aggregation‌ and ‌result analysis. Fudan University Shanghai Cancer Center‌ and Ruijin Hospital, Shanghai Jiaotong University School of Medicine are responsible for ‌surgical procedures‌, ‌postoperative follow-up and submission of follow-up data‌ for enrolled patients at their respective centres.

Objectives and endpoints

PG demonstrates nutritional advantages over TG in postoperative recovery, with body weight loss (BWL) serving as a validated prognostic indicator following radical gastrectomy.²⁰ Our preliminary single-centre study demonstrated superior BWL outcomes with HEFT (11.12%) compared with literature-reported BWL rates for DTR (13.66%).²¹,²⁷ Higher BWL is associated with elevated nutritional risk and increased incidence of postoperative complications such as anastomotic leakage and infection.²⁸ A 2–3% reduction in BWL significantly decreases malnutrition-related readmission rates by approximately 15–20%.²⁹ Consequently, HEFT achieving a 2.5% lower BWL than DTR would reduce nutritional risk and potentially yield clinical benefits including fewer complications and reduced need for nutritional interventions, demonstrating clear clinical decision-making value. The 2.54% discrepancy between literature-derived DTR outcomes and our retrospective HEFT data closely approximates the predefined 2.5% superiority margin. This alignment suggests that despite retrospective study biases, the observed BWL difference likely reflects real-world clinical variation, while ensuring the margin remains within a feasible ‘potential benefit magnitude’, thereby mitigating the risk of setting impractically high or clinically insignificant thresholds. Postoperative BWL stabilises at the 12-month plateau phase, reflecting long-term surgical adaptation independent of short-term stress responses, and correlates significantly with patients’ dietary satisfaction and functional capacity.^{30 31} Consequently, 12-month postoperative BWL is designated as the primary objective. BWL is calculated as (preoperative weight−postoperative weight)/preoperative weight×100% . For weight measurement across all centres, electronic scales with a minimum division value of 0.1 kg are used. Patients are uniformly dressed in a unified, lightweight and thin patient gown, with personal items such as jewellery removed and stand upright in the centre of the scale to complete the measurement. For the same patient at the same time point, measurements are taken two times, and the average is used as the final result (if the difference between the two measurements is >0.5 kg, a third measurement is conducted, and the average of the two closest values is taken). The scales used in the experiment are calibrated with standard weights every 3 months. Secondary objectives are categorised as follows: (1) Efficacy metrics: haemoglobin and serum albumin levels at 12 months postoperatively; incidence of anastomotic stenosis and reflux oesophagitis at 12 months. (2) Short-term safety outcomes (≤7 days): operative duration, intraoperative blood loss, rates of anastomotic leakage, pancreatic fistula and intra-abdominal infection. (3) Medium-to-long-term safety outcomes (36 months): 3-year OS and 3-year DFS.

ECOG performance status

A standardised tool for assessing the physical functional status of patients with cancer, where a score of 0 indicates fully normal activity ability and a score of 1 indicates the ability to engage in light physical activity.

Eligibility criteria

A consecutive cohort of 52 patients with ‌Siewert II or III gastric adenocarcinoma and early upper gastric cancer‌ will be enrolled. ‌PG with D2 lymphadenectomy‌ is anticipated to achieve ‌R0 resection‌ outcomes‌. Detailed eligibility criteria‌ are provided in table 1‌.

Table 1. Eligibility criteria.

Inclusion criteria	Exclusion criteria
1. 18 years ≤age ≤80 years	1. Preoperative imaging showing fused regional lymph nodes (maximum diameter≥3 cm)
2. Primary tumour located in upper stomach or oesophagogastric junction (Siewert type II/III) with planned proximal gastrectomy and D2 lymphadenectomy achieving R0 resection	2. Pregnancy or lactation
3. Histopathologically confirmed adenocarcinoma by endoscopic biopsy	3. History of other malignancies within 5 years
4. cT1N0M0 for upper gastric adenocarcinoma; cT1-3N0-1M0 for oesophagogastric junction adenocarcinoma without distal perigastric lymph node metastasis	4. Preoperative fever≥38°C or active infection requiring systemic therapy
5. Expected survival>6 months	5. Severe psychiatric disorders
6. No prior upper abdominal surgery (except laparoscopic cholecystectomy)	6. Severe respiratory disease (FEV1<50% of predicted value)
7. No preoperative chemotherapy/radiotherapy/targeted therapy/immunotherapy	7. Severe hepatic/renal dysfunction
8. Preoperative Eastern Cooperative Oncology Group performance status 0/1	8. Unstable angina or myocardial infarction within 6 months
9. Preoperative ASA classification I-III	9. Cerebral infarction/haemorrhage within 6 months (except old lacunar infarcts)
10. Adequate organ function	10. Systemic corticosteroid use within 1 month
11. Signed informed consent	11. Emergency surgery required for complications (bleeding/perforation/obstruction)
	12. Participation in other clinical trials within 6 months

ASA, American Society of Anesthesiologists; FEV₁, forced expiratory volume in one second.

Interventions

All enrolled patients will undergo totally laparoscopic PG with preservation of

approximately 1/2–2/3 of the distal stomach and D1+ to D2 lymphadenectomy. The experimental group will receive HEFT, involving a longitudinal seromuscular incision (3 cm in length) on the anterior gastric wall, positioned 2–5 cm distal to the resection margin. Mucosal integrity is preserved while incising the serosal and muscular layers. A single-layer continuous anastomosis is performed between the distal oesophagus and the inferior end of the gastric fissure, creating a functional valve through mucosal apposition of the fissure’s gastric wall and the oesophageal posterior wall. The ‘pseudogastric fundus’ (residual gastric body) is then folded 270° around the distal oesophagus to reinforce anterior oesophageal support. The control group will undergo DTR, initiated with Roux-en-Y oesophagojejunostomy following proximal gastric transection. A side-to-side gastrojejunostomy is then performed 10–15 cm distal to the esophagojejunal anastomosis, followed by a subsequent side-to-side jejunojejunostomy 25 cm distal to the gastrojejunal anastomosis.^{11 32}

Outcomes

Short-term surgical safety will be evaluated by collecting perioperative data within 7 days postoperatively. Patients with stage II or higher gastric cancer who undergo R0 resection will receive guideline-directed adjuvant chemotherapy, while subsequent management of non-R0 resections or recurrences will be determined independently by participating centres based on clinical judgement. At 12 months postoperatively, weight measurements and endoscopic evaluations will assess BWL and document anastomotic stenosis/reflux oesophagitis incidence and severity. All participants will undergo survival and safety monitoring for up to 36 months.

Recruitment and participant timeline

All participants will sign an informed consent form before surgery, and they will receive a copy of the document. Patient enrolment is planned to be completed within 20 months, targeting 52 surgically treated cases. Follow-up initiates on first patient enrolment, with maximum individual monitoring duration of 36 months. The study will conclude concurrently when the final participant completes their 36-month postoperative follow-up. The informed consent document is attached to the main text as patient consent form.

During the study period, participants may discontinue involvement prematurely due to medical indications, incomplete follow-up, personal reasons or other factors. Discontinued cases will subsequently transition to routine clinical follow-up based on the clinical discretion of participating research centres.

Sample size

This study employs a superiority design with 1-year BWL as the primary superiority evaluation metric, using 1:1 randomisation. Based on the existing literature, the reported 1-year BWL for the DTR group is 13.66%, while preliminary retrospective analyses indicate an 11.12% BWL (SD 2.46) for HEFT.²¹,²⁷Assuming a superiority margin of 2.5% (1−β= 0.80, α=0.05, attrition rate=10%), the calculated sample size is 52 cases, with 26 patients allocated to each the experimental (HEFT) and control (DTR) groups.

Assignment of interventions

This study employs a customised randomisation system from the lead institution, integrating the Pocock-Simon minimisation algorithm, a dedicated database, and an authority management module to implement minimisation-based allocation. The assignment is executed by a remote server independent of participating research centres, where local investigators can only input patient data through a restricted interface without direct access to allocation algorithms or historical sequence records. On patient enrolment, dedicated staff at each participating centre will input case-specific parameters (age, preoperative stage, histopathological type) into the centralised randomisation system, which will immediately generate and transmit the randomisation assignment to the originating centre. Research personnel entering patient data remain completely blinded to allocation outcomes. Their accounts exclusively permit ‘data entry’ functionality, with no access to ‘allocation result query’ buttons or interfaces. Assignment results are solely transmitted to authorised principal investigators, requiring input of patient-specific identification codes for retrieval. Furthermore, allocation results are generated instantaneously by the system on data entry completion, with no allocation-related feedback provided to data entry personnel during the input process. Given that both interventions involve surgical procedures, masking will not be implemented.

Data collection, management and analysis

Continuous variables will be expressed as mean±SD (x̄±s) and analysed using Student’s t-test. Categorical data will be presented as frequencies (%) and evaluated via χ² or Fisher’s exact tests. Correlation analyses will use Pearson or Spearman methods. Survival outcomes will be assessed through Kaplan-Meier curves with Log-rank testing for categorical variables. Univariate Cox regression analysis will screen continuous variables, followed by multivariate Cox regression employing a forward likelihood ratio method. For superiority margin evaluation, a one-sided test (α=0.05) will be implemented using an independent samples t-test for normally distributed data or the Mann-Whitney U test for skewed distributions. This study will calculate the 95% two-sided CI for intergroup differences between experimental (HEFT) and control (DTR) groups, with simultaneous confirmation of statistically significant BWL reduction and HEFT’s superiority established when the one-sided p value is <0.05 and the CI upper limit falls below −2.5%. If there are cases of loss to follow-up, patient withdrawal, etc, after clarifying the missing data mechanism, the multiple imputation method (using the R language mice package or SPSS (V.23.0) multiple imputation module) shall be prioritised, and sensitivity analysis shall be conducted to verify the conclusions. For the analysis of primary results, the intention-to-treat analysis shall be adopted. If applicable, the perprotocol analysis will be used as an auxiliary method to supplementally explain the impact of compliance on the results. All statistical analyses will be performed using SPSS V.23.0 (IBM Corp), with statistical significance defined as p<0.05.

Monitoring

Two uniformly trained clinical researchers will oversee trial implementation and manage data collection/processing throughout the study duration.

The evaluation of adverse reactions in this study refers to Common Terminology Criteria for Adverse Events V.5.0 and Accordion Severity Grading System. In the event of study-related harms (excluding surgical complications or pharmacological adverse events), participating medical staff will provide immediate medical intervention, with the host clinical centre delivering necessary healthcare measures per institutional protocols. Pursuant to regional legal regulations, study sponsors will bear associated medical expenses and provide appropriate financial compensation for affected participants.

Patient and public involvement

Patients and/or the public will not be involved in the design, conduct, reporting or dissemination plans of this research.

Discussion

Surgical intervention remains the primary curative approach for gastric cancer. For tumours localised in the upper third of the stomach or the adenocarcinoma of the oesophagogastric junction, TG—historically the standard procedure—often results in significant postoperative complications and diminished QOL.³³ Consequently, PG has emerged as a viable alternative to mitigate severe reflux oesophagitis, particularly with advancements in surgical techniques. However, PG remains investigational, with varying indications across international guidelines.^{34 35} The Chinese guidelines recommend PG for early-stage upper gastric cancers (≤4 cm at the adenocarcinoma of the oesophagogastric junction) or cT2-3 upper gastric cancers after rigorous evaluation of resection margins and lymph node metastasis, offering substantial nutritional and QOL benefits.³⁶ A recent study demonstrates comparable 5-year OS and recurrence-free survival between PG and TG in patients with locally advanced proximal gastric cancer receiving neoadjuvant chemotherapy, further supporting PG’s therapeutic potential and oncologic efficacy.³⁷

Despite these benefits, no universally accepted reconstruction method exists for PG. HEFT, an innovative reconstruction approach, has demonstrated preliminary safety and superior postoperative nutritional outcomes. Compared with existing methods like DTR, HEFT’s unique design—featuring a ‘pseudogastric fundus’ and a valvular mechanism at the anastomotic site—provides effective antireflux properties while reducing technical complexity. The procedure is fully laparoscopic, enhancing its clinical applicability.^{17 18} Notably, while PG may increase the risk of metachronous gastric cancer compared with DTR (which risks gastric conduit disuse and remnant cancer), HEFT optimises residual gastric functionality. Refined patient selection criteria and comprehensive postoperative surveillance protocols may further mitigate remnant cancer risks.^{38 39}

In summary, HEFT exhibits distinct advantages in PG reconstruction. This prospective multicentre randomised trial aims to validate its efficacy and safety, potentially establishing HEFT as a standardised digestive tract reconstruction method and providing high-level evidence for optimising postoperative management in PG.

Ethics and dissemination

This study was approved by the hospital institutional review board of Huashan Hospital, Fudan University (2024-1173) and is being conducted in accordance with the Declaration of Helsinki and the Good Clinical Practice guidelines. The lead research centre (affiliated with the authors) has established a well-defined collaborative ethical review mechanism with two participating centres. Primary ethical review procedures are conducted at the lead institution, with the ethics committees of collaborating centres endorsing the finalised ethical determinations. All institutions recognise the ethical review outcomes issued by the lead centre. On completion of the study, the results will be published in a peer-reviewed journal.