EUS-guided Versus Laparoscopic Gastrojejunostomy for Malignant Gastric Outlet Obstruction

A Multicentre Cohort Study and a Randomized Controlled Trial of EUS-guided Versus LaparOscopic GAsTrojejunostomY for malIgNant Gastric Outlet Obstruction: EATING Trial

Gastric outlet obstruction (GOO) is a common complication of various types of malignancy arising from the pancreas, bile ducts, retroperitoneum and distal stomach. As these tumours grow they may invade or compress the distal stomach or gastric outlet causing patients to develop the classical manifestations of GOO including vomiting, weight loss, electrolyte disturbances loss of quality of life as well as a reduced life expectancy. Until recently these patients were managed using either endoscopic stenting or surgery. Laparoscopic gastrojejunostomy (L-GJ) has been regarded as the golden standard in the management of GOO. However surgery is invasive and may lead to peri-operative complications and longer hospitalization which further greatly impacts on the quality of life of these patients in which survival is already greatly impaired by their oncological problem. Recently it has been shown that a special stent can be utilized to circumvent the tumour obstruction without the need for surgery. This technique utilizes endoscopic ultrasound to connect the stomach to the jejunum at a site distal to the tumour obstruction. This techniques is minimally invasive compared to surgery. In a large retrospective propensity matched study we recently compared endoscopic ultrasound placement of an AXIOS stent to laparoscopic surgery and could show that the EUS-guided approach led to less complications (2.6 vs 26%) and shorter hospital stay compared to surgery.1 Our study and that of other groups, has led the European Society of Gastrointestinal Endoscopy (ESGE) to recommend that EUS-guided gastrojejunostomy (EUS-GJ) can be considered for patients with malignant GOO as an alternative for surgery and endoscopic stenting, acknowledging that further randomized controlled trials comparing the EUS-GJ approach to surgery, are necessary.2 Based on the above clinical need we set out to design a multicentre international trial with the main aim of comparing EUS-guided gastrojejunostomy to laparoscopic gastroenterostomy. The planned study has two components: 1. Randomized study (EATING RCT)- patients that meet the strict inclusion criteria for the (EATING-RCT) will be randomized to either EUS-GJ or L-GJ. 2. Patients in whom inclusion criteria for the RCT are not met, will be invited to participate in the observational arm of the study (EATING-C) comparing EUS-GJ, enteral stenting and surgical gastrojenostomy in a "real world scenario" setting. With this combined design we aim to include all patients with malignant gastric outlet obstruction undergoing treatment. This will be an international multicentre effort, with expert centers in Belgium and other countries in Europe, including UZ Brussel (BE), UCL (BE), Hôpital Erasme (BE), UCL Namur (BE), UZ Gent (BE), Alicante (Spain), Gemelli Rome (Italy), San-Rafaelle Milan (Italy) and G.B. Rossi University Hospital Verona (Italy) as participating centres (total n=11).

This envisaged investigator driven study will be performed using an AXIOS stent produced by the company BOSTON SCIENTIFIC. This stent is registered in Europe and CE approved for the drainage of pancreatic fluid collections, the gall bladder and bile ducts. As described above we and other groups have used this product off-label to succesfully perform EUS-guided GJ. High-quality randomized controlled data are however required that will be generated in the proposed EATING RCT study. It is expected that the data, generated by the EATING RCT trial will contribute significantly to resolve the current question regarding the best management of patients with GOO. Ultimately the objective should be to provide high quality affordable care that is minimally invasive to cancer patients with GOO. Such high quality data will also be crucial when applying for registration and CE approval of this product for use in the management of GOO .

Gastric outlet obstruction (GOO) is a relatively common disease manifestation amongst patients with various types of malignancy. Until now, laparoscopic gastrojejunostomy (L-GJ) has been regarded as the golden standard for these specific patients, although recent (low-quality) evidence suggests that off-label EUS-guided gastrojejunostomy (EUS-GJ) with the Hot-AXIOS stent might provide a safer alternative with similar effectiveness and shorter hospital stay. The available comparative studies are unfortunately mainly of retrospective design, include low numbers of EUS-GJ procedures and therefore prone to bias.

Currently, EUS-GJ is performed in the participating centres with the off-label use of the Hot-AXIOS stent outside study context continuously. Moreover, whilst awaiting randomized confirmation, the European Society for Gastrointestinal Endoscopy has suggested EUS-GJ as an alternative to L-GJ and enteral stenting in centers with expertise.

Study hypothesis: In patients with malignant Gastric Outlet Obstruction we hypothesize that endoscopic ultrasound guided placement of a lumen apposing stent, to create a gastrojejunostomy, will result in reduced procedure related morbidity, shorter time-to-tolerate oral intake, shorter hospital stay and a higher quality of life compared to surgery, while reducing procedure related and overall costs. Primary endpoint is the overall rate of procedure-related adverse events happening within 90 days of the endoscopic or surgical procedure. Key secondary endpoints are time to oral intake and post-procedural hospital stay.

The EATING study will utilize a combined design. In the EATING-C prospective, multicentre, real-life, cohort study all patients with mGOO will be enrolled irrespective of their performance status, their disease stage and treatment (i.e. EUS-GJ, OR surgical [laparoscopic or not] GJ OR enteral stenting).

The subgroup of patients with expected survival over 2 months and without radiographic signs of significantly obstructive peritoneal disease will be included in the EATING-RCT, an international, open-label, parallel group, superiority randomized controlled trial in which patients who fulfill the inclusion/exclusion criteria (see 4.2 inclusion criteria) and give written informed consent will be randomized to either EUS-GJ or L-GJ. The same outcomes will be prospectively evaluated in the EATING-C and EATING-RCT up to 12 months or death of the patient, whichever comes first.

During the assessment of eligibility, patients will receive explanation of the treatment possibilities for mGOO. If they fall within criteria for EATING-RCT they will be asked informed consent for a randomization to L-GJ versus EUS-GJ. If they meet EATING-C, but not EATING-RCT criteria or if they are willing to receive one specific procedure only, these patients will not be randomized and thus be included in the EATING-C.

Expected full trial duration: 12 months after inclusion of last patient (target n=120).

Individual trial duration for patient: 12 months after treatment or time until death (whichever comes first).

In previous published experiences, surgical Gastro-Jejunostomy has been associated with a 20 to 40% rate of adverse events (10.1055/s-0043-101695, 10.1097/MCG.0000000000000887). Previous experiences of EUS-guided Gastro-Jejunostomy have found a 10 to 15% rate of adverse events (10.1097/MCG.0000000000000887, 10.1016/j.dld.2020.04.021). We performed an international, multicenter, retrospective, propensity score-matched analysis of consecutive EUS-GJ and L-GJ procedures in 3 European centres In this study we observed a 2.7% vs. 27.0% rate of overall adverse events in EUS-GJ vs. L-GJ respectively.

Inclusion Criteria:

Participants eligible for inclusion in the EATING-C must meet all of the following criteria:

1. Voluntary written informed consent of the participant or their legally authorized representative has been obtained prior to any screening procedures
2. ≥ 18 years old
3. Endoscopic or radiological confirmation of a gastric outlet obstruction due to an obstructive malignant neoplastic lesion extending from the distal one third of the stomach or the duodenum
4. GOOSS score(8) < 2 (no oral intake or liquids only)
5. Performance status justifying palliative treatment for mGOO
6. No contra-indications to undergo deep conscious sedation or general anesthesia All participants that are considered for Trial participation, per the above criteria will be documented on the Screening Log, including Screen Failures.

Participants eligible for inclusion in the EATING-RCT must meet all of the following criteria:

1. EATING-C inclusion criteria
2. Endoscopic and surgical creation of a gastrojejunostomy are deemed both technically feasible by the treating physician.
3. Expected survival exceeds 2 months.
4. ECOG Performance Status ≤2

Exclusion Criteria:

Participants eligible for the EATING-C Trial must not meet any of the following criteria:

1. Patients with benign GOO
2. Patients who are candidates for curative surgical resection of the primary disease
3. Patients whose mGOO is likely to resolve very quickly (within days or weeks) under anti-tumoral treatment (such as lymphoproliferative diseases).

Participants eligible for the EATING-RCT Trial must not meet any of the following criteria:

1. ECOG Performance status > 2

2. Expected survival of less than 2 months according to disease extent and comorbidities, as assessed by an oncologist or a treating physician. If unclear, the Coordinating Investigator is to be contacted, after which in- or exclusion will be decided upon by means of consensus.

3. Peritoneal carcinomatosis with signs of obstruction on cross-sectional imaging (on either small or large bowel level).

   In case these parameters leads to ambiguity in individual patients at participating sites, the Coordinating Investigator will be contacted and a final conjoint decision for in/exclusion will be made after revision of radiological findings.

4. Grade II and III ascites OR ascites interfering with the EUS-GJ trajectory, complicating successful creation of an EUS-GJ.

5. Previous surgical procedures or reconstructions impeding EUS-GJ.

6. Diffuse tumor involvement of the gastric wall impeding EUS-GJ

7. Participation in other interventional procedures which may be of influence on primary or secondary outcome parameters

8. Presence of other strictures in gastrointestinal tract leading to radiological or clinical signs of obstruction or patency issues.

Participants who meet one or more of the above exclusion criteria must not proceed to be enrolled/randomized in the Trial and will be identified on the Screening Log as Screen Failure.