G-POEM in 2026: Select Refractory Gastroparesis, Not First-Line Pyloric Cutting

---

What is gastroparesis and how do I prove it?

Gastroparesis is a clinical diagnosis with a test attached. The patient has symptoms that suggest food is sitting in the stomach. The test shows delayed solid emptying. The EGD has already ruled out a mechanical reason for that delay. If any of those three is missing, you do not have gastroparesis.

ACG 2022 makes 4-hour solid-phase scintigraphy the diagnostic test (strong recommendation). The Tougas low-fat egg-white meal is the usual protocol. Retention under 10% at 4 hours is normal. Mild delay is often 10% to 20%. AGA's 2023 G-POEM commentary and the 2025 guideline both use at least 20% retained at 4 hours as the emptying threshold most associated with a later G-POEM response. A 1- or 2-hour study will overcall delay. Wireless motility capsule and ¹³C breath testing are alternatives when scintigraphy is not available. Radiopaque-marker studies are not.

EGD is not optional. Bezoar, pyloric stenosis, malignancy, and peptic stricture are not G-POEM diseases. Photograph the pylorus. If you are considering myotomy later, note whether the ring looks spastic, patulous, or scarred from prior surgery.

Functional dyspepsia with normal emptying is the usual mimic. Treating that patient with a pyloromyotomy is how you get a disappointed referral back at month 3.

What should I try before G-POEM?

Diet first: small particles, low fat, low fiber, frequent meals, liquid calories if solids will not empty. In diabetes, glucose control matters for future aggravation even if it does not immediately normalize emptying (ACG conditional).

Review the medication list. Opioids delay emptying and blunt G-POEM response. Wean them and repeat the emptying study. GLP-1 receptor agonists delay emptying; decide whether the drug or the stomach is the problem before you cut a pylorus. Anticholinergics and some calcium-channel blockers belong on the same list.

Metoclopramide is still the only FDA-approved prokinetic. ACG suggests it over no treatment (conditional, low). The label limits duration to 12 weeks because of tardive dyskinesia; contemporary estimates put that risk well under the old 1% to 10% figures, but it is not zero, and it is higher in older patients. Erythromycin works as a motilin agonist for days to a few weeks and then tachyphylaxis arrives. Use it as a bridge, not a lifestyle. Where domperidone is legal, ACG suggests it. AGA 2025 did not put prucalopride, aprepitant, nortriptyline, buspirone, or cannabidiol in the first-line chair.

Antiemetics treat nausea. They do not fix emptying. That is still useful, because nausea is what sends people to the emergency department.

Intrapyloric botulinum toxin looked plausible and then failed. Friedenberg 2008 and Arts 2007 were randomized and negative. ACG 2022 recommends against Botox (strong, moderate). AGA 2025 suggests against routine Botox in refractory disease. A "Botox response" is not a validated ticket to G-POEM. Transpyloric stenting is a temporizing trick, not a predictor.

Gastric electrical stimulation has a humanitarian-device exemption for diabetic nausea and vomiting. AGA 2025 suggests against using it as routine initial therapy, same posture as G-POEM: reserve it. Emptying often does not change even when vomiting does.

If oral intake is failing, a feeding jejunostomy (or a well-behaved PEG-J) is nutrition, not a delay tactic. Parenteral nutrition is last.

Who should I offer G-POEM?

Hold two documents in your head. They do not actually disagree once you read the comments.

ACG 2022 suggestion 20: in medically refractory gastroparesis, pyloromyotomy (endoscopic or surgical) over no treatment for symptom control. Conditional, low-quality evidence. That sentence exists because open-label series and then a sham-controlled RCT showed a symptom signal.

AGA 2025 recommendation 10: in medically refractory gastroparesis, suggest against the routine use of G-POEM. Conditional, low. The implementation notes are the selection filter: a proper 4-hour emptying study, generally at least 20% retained at 4 hours; 6 to 12 months of moderate nausea, vomiting, or postprandial fullness; a prior trial of prokinetics and antiemetics. Patients and clinicians who value symptom relief over procedural risk may still reasonably choose G-POEM.

The 2023 AGA Clinical Practice Update (Khashab, Wang, Cai) is the procedure-level checklist still worth using in the room:

• EGD already showed no mechanical obstruction.
• Solid-phase emptying is delayed, preferably more than 20% at 4 hours.
• Symptoms are moderate to severe, preferably nausea and vomiting, often a GCSI total above 2.
• Opioids weaned when possible, emptying rechecked.
• Most post-infectious gastroparesis should not be offered G-POEM.
• Pain-predominant patients respond poorly.
• Failure of a stimulator, a stent, or Botox is not a prerequisite.
• Do this at a center that already does third-space endoscopy.

Martinek 2022 is the number you quote at the consult. Forty-one patients, sham-controlled, enrollment stopped at interim analysis. GCSI success at 6 months: 71% after G-POEM versus 22% after sham. Diabetic gastroparesis looked strongest (89% versus 17% in a small subgroup). Postsurgical and idiopathic were less clean. Vosoughi's international prospective series is the uncontrolled complement: clinical success about 56% at 12 months, which is the honest longer-horizon number, not the 90% you will hear in a vendor hallway.

EndoFLIP across the pylorus may help you characterize a stiff ring and document that the myotomy changed distensibility. ACG gives it a conditional, very-low-quality nod as a predictor. It is not a substitute for emptying data or a symptom history.

How do I perform G-POEM?

This is not esophageal POEM. The tunnel runs uphill through thick antral folds toward a ring you can miss, next to a duodenal wall you can perforate, over a gastroduodenal artery you do not want to meet. Cricopharyngeal myotomy is a third operation entirely; that lives on the C-POEM page.

1. Inspect. Wash the antrum. Identify the pylorus on the clock face you plan to use (often greater curve, 5 o'clock). Retroflex and look at a fundoplication if one is there. Confirm there is no ulcer you are about to tunnel under.
2. Entry. Submucosal lift 4 to 5 cm proximal to the pylorus. Mucosotomy large enough to accept the cap. Some operators pre-inject a navigational bleb from the pylorus backward so the blue path points at the ring.
3. Tunnel. Dissect to the pyloric ring. The ring looks like a half-moon or a pale cable once you arrive. If you cannot find it, you are off-axis. Come out and look from the lumen rather than inventing a second tunnel.
4. Myotomy. Divide the pyloric sphincter. Most series cut 2 to 3 cm, starting at the ring and working a short distance back into the antrum. An insulated-tip knife is the usual protection against the duodenal wall and the GDA. Whether circular-only or full-thickness is better is not settled. Do not treat serosa as a trophy.
5. Close. Clips or endoscopic suture of the mucosotomy. A leak here is a perforation. Photograph the open pylorus at the end so the next endoscopist can see what you did.

PPI after the procedure. Overnight observation is typical until the learning curve is behind you. Early bleed, delayed ulcer at the pylorus, and capnoperitoneum are the expected complications. Dumping is uncommon but real once the ring is gone. Buried muscle and incomplete myotomy are how you get a second procedure.

What should I tell the patient?

G-POEM is not a cure for a paralyzed stomach. It lowers pyloric resistance so a weak antrum has less to push against. Nausea and vomiting are the symptoms most likely to move. Pain, bloating, and opiate-driven emptying often do not. Quote Martinek's 6-month sham result and Vosoughi's 12-month 56%, not a 90% brochure. AGA 2025 is why you are not offering this on the first visit. Shared decision is the actual recommendation: some refractory patients will still choose the cut, and that can be reasonable when the emptying is clearly delayed and the dominant symptoms are the ones pyloric therapy can touch.

If they need nutrition now, place the jejunal feeding tube. Do not make them wait for a myotomy slot to eat.

Selected references

1. Camilleri M, Kuo B, Nguyen L, et al. ACG Clinical Guideline: Gastroparesis. Am J Gastroenterol. 2022;117:1197-1220.
2. Staller K, et al. AGA Clinical Practice Guideline on Management of Gastroparesis. Gastroenterology. 2025;169:828-861.
3. Khashab MA, Wang AY, Cai Q. AGA Clinical Practice Update on Gastric Peroral Endoscopic Myotomy for Gastroparesis: Commentary. Gastroenterology. 2023;164:1329-1335.e1.
4. Martinek J, Hustak R, Mares J, et al. Endoscopic pyloromyotomy for severe and refractory gastroparesis: a pilot, randomised, sham-controlled trial. Gut. 2022;71:2170-2178.
5. Vosoughi K, Ichkhanian Y, Benias P, et al. Gastric per-oral endoscopic myotomy (G-POEM) for refractory gastroparesis: results from an international prospective trial. Gut. 2022;71:25-33.
6. Friedenberg FK, Palit A, Parkman HP, Hanlon A, Nelson DB. Botulinum toxin A for the treatment of delayed gastric emptying. Am J Gastroenterol. 2008;103:416-423.
7. Arts J, Holvoet L, Caenepeel P, et al. Clinical trial: a randomized-controlled crossover study of intrapyloric injection of botulinum toxin in gastroparesis. Aliment Pharmacol Ther. 2007;26:1251-1258.
8. Tougas G, Eaker EY, Abell TL, et al. Assessment of gastric emptying using a low fat meal: establishment of international control values. Am J Gastroenterol. 2000;95:1456-1462.

---

Last reviewed September 20, 2026. Written for clinicians deciding whether a patient with nausea, vomiting, and delayed emptying should go to G-POEM, and how to do the myotomy if they should.