Gastroparesis: A Review.

Scope: individuals with suspected or confirmed idiopathic or diabetic gastroparesis, including medically refractory disease. All 12 recommendations are CONDITIONAL ('suggests'), at low or very low certainty of evidence; none are strong. DIAGNOSIS (Rec 1): AGA suggests against a 2-hour (or shorter) gastric emptying study and in favour of a 4-hour study [conditional, low certainty], because patients with normal emptying at 2 hours may be reclassified as delayed at 4 hours. The 4-hour value must be measured directly at 4 hours, not mathematically derived from earlier time points. Testing should be done without confounders: hyperglycaemia and drugs that alter emptying (opioids, GLP-1 receptor agonists, prokinetics). Two meals have well-established normal values: the low-fat Tougas/EggBeaters meal (liquid egg white, bread, jam, water; 250 kcal, 2% fat) and the Mayo Clinic meal (two whole eggs, bread, skim milk; 320 kcal, 30% fat), the latter more thoroughly validated. PHARMACOTHERAPY: AGA suggests USING metoclopramide (Rec 2) and USING erythromycin (Rec 3), each conditional/very low certainty. Metoclopramide, oral tablet/liquid or intranasal, start 5 mg before meals, may increase up to 10 mg before meals; discuss potential side effects before starting (tardive dyskinesia risk noted, absolute risk low; higher risk of adverse events leading to discontinuation in older patients and those on psychotropics); assess efficacy and side effects at 4-8 weeks to decide on continuation, then monitor for the duration of therapy. Patients placing higher value on adverse-event risk and lower value on symptom improvement may reasonably decline it. Erythromycin, low dose (e.g. 100-150 mg by mouth, 30 min before meals) to reduce side effects seen at 250-500 mg; ethylsuccinate oral suspension allows smaller doses since tablets come only in high strengths; prokinetic via motilin agonism with no direct antiemetic effect; tachyphylaxis with prolonged continuous use is managed by drug holidays (e.g. 3 weeks on, 1 week off); cautions include QT prolongation, CYP3A interactions and antibiotic resistance. AGA suggests AGAINST, as first-line therapy, with shared-decision carve-outs, domperidone (Rec 4), prucalopride (Rec 5), aprepitant (Rec 6), nortriptyline (Rec 7.1) and buspirone (Rec 7.2) [conditional; low to very low certainty], and against cannabidiol EXCEPT in the context of a clinical trial (Rec 8, conditional/low certainty). Each carries an explicit shared-decision-making carve-out naming who may still reasonably use it: domperidone for patients who had neurological side effects on metoclopramide or who have movement disorders such as Parkinson's disease (US use requires an FDA expanded-access IND, and the sole US supplier is exiting; obtain baseline potassium, magnesium and ECG QTc, monitor on therapy, reassess at 4-8 weeks); prucalopride particularly in idiopathic gastroparesis (patients with diabetic or connective-tissue-disease-related gastroparesis are less likely to respond) and in those with coexisting chronic idiopathic constipation, noting the depression/suicidality monitoring warning; aprepitant for nausea and vomiting not helped by 5-HT3 antagonists such as ondansetron, recognising it has no effect on gastric motor function; nortriptyline for coexisting IBS or significant abdominal pain, started at low dose and titrated slowly, framed to patients as a peripheral neuromodulator; buspirone for predominant early satiety and bloating. For cannabidiol the panel notes the single positive trial used pharmaceutical-grade Epidiolex, which is not available for clinical use in gastroparesis, that commercial formulations are unregulated and variable in potency, and that THC-containing products raise cannabinoid hyperemesis concern. PROCEDURES (medically refractory disease): AGA suggests against G-POEM EXCEPT for selected patients with medically refractory disease (Rec 9, conditional/low certainty), candidates should have gastroparesis confirmed on an appropriately performed 4-hour gastric emptying study, generally with at least moderate delay (20% retention at 4 hours with the EggBeaters meal), at least six months of moderate cardinal symptoms (nausea, vomiting and/or postprandial fullness), and a prior trial of other treatments including a prokinetic such as metoclopramide plus an antiemetic. AGA suggests against pyloric botulinum toxin injection (Rec 11, conditional/very low certainty), noting that retreatment as often as every 3 months limits cost-effectiveness and that repeat treatment may cause pyloric scarring complicating later G-POEM; patients placing higher value on endoscopic intervention and lower value on chronic medical therapy may reasonably attempt it. AGA suggests against the ROUTINE INITIAL use of gastric electrical stimulation (Rec 12, conditional/very low certainty), reserving it, as with G-POEM, for select patients whose symptoms are refractory to medical therapy; patients placing higher value on potential nausea/vomiting improvement and lower value on the increase in serious adverse events may reasonably select GES, and should be told it aims to improve nausea and vomiting, not abdominal pain. For surgical pyloric interventions (pyloromyotomy or pyloroplasty) AGA makes NO recommendation and designates a knowledge gap, advising use only in the context of clinical trials (Rec 10). NON-GRADED adjuncts appear only as implementation considerations, not recommendations: panel content experts use small-particle, low-fat, low-residue diets before or alongside pharmacotherapy, and as-needed antiemetics (5-HT3 antagonists e.g. ondansetron, H1 antagonists e.g. promethazine, D2 antagonists e.g. prochlorperazine); efficacy of pharmacological interventions is assessed at 4-8 weeks. The guideline does NOT address exclusion of mechanical obstruction, enteral/J-tube nutrition, glycaemic targets, or perioperative GLP-1 RA management, those remain with the supporting documents.