Jan Tack, Florencia Carbone; Alessandra Rotondo


Curr Opin Gastroenterol. 2015;31(6):499-505. 

In This Article

Treatment Options

Dietary Interventions

Although dietary measures (smaller more frequent meals low in fat and nondigestible residue) are a first management step in gastroparesis and related severe upper gastrointestinal motility disorders, their efficacy is not proven. Wytiaz et al.[25] studied self-reported aggravation or alleviation by nutrients in 45 patients with gastroparesis (39 idiopathic), using food toleration and aversion rating questionnaires. Patients reported food substances, which they considered to aggravate (orange juice, fried chicken, cabbage, oranges, sausage, pizza, peppers, onions, tomato juice, lettuce, coffee, salsa, broccoli, bacon, and roast beef) or improve (saltine crackers, jello, and graham crackers) symptoms, and indicated good tolerance for a number of food items (ginger ale, gluten-free foods, tea, sweet potatoes, pretzels, white fish, clear soup, salmon, potatoes, white rice, popsicles, and applesauce).[25] These data can be used as a basis for designing dietary interventions in gastroparesis.

The same group studied 12 gastroparesis patients who received one of four meals (high-fat solid, high-fat liquid, low-fat liquid, and low-fat solid meal) in a randomized crossover fashion on four separate days. The severity of eight symptoms was rated on a 0–4 scale every 15 min, before and up to 4 h after meal ingestion.[26] Both high-fat meals as well as low-fat solid meal were associated with higher symptom burden. The total symptom score was significantly elevated after the high-fat solid meal compared with all other meals. Nausea was highest with the high-fat solid meal and lowest with the low-fat liquid meal. This provides support for recommending low-fat and more liquid meals in gastroparesis to reduce symptoms, but longer term diet studies are needed.[26]

In a longer term dietary intervention study, 56 diabetic gastroparesis patients were randomized to a 20-week treatment with an intervention diet (small particle size) or a control diet (diabetes diet).[27] All symptoms, measured by the PAGI-SYM questionnaire[12] except for upper abdominal pain, improved significantly with the intervention diet but not the control diet. The improvement in the three PAGI-SYM subscales that constitute the GCSI[15] was significantly higher after the intervention diet, and the same was true for the heartburn/regurgitation scale.[27] Based on this study, a diet with small particle size should be incorporated in the dietary management of patients with diabetic gastroparesis.


Prokinetics are considered the first pharmacotherapeutic option for gastroparesis, although evidence of efficacy is limited.[2,28] In a meta-regression analysis of the association between improvement in symptoms and in emptying rate across multiple gastroparesis studies, no significant correlation was found between both aspects.[5] Ghrelin agonists are a novel class of prokinetics.[29,30] TZP-102, an orally administered ghrelin agonist, was evaluated in two phase 2b controlled trials in patients with type 1 or type 2 diabetes with gastroparesis symptoms and delayed gastric emptying.[31] In one study, a total of 201 patients were randomized to once-daily treatment with placebo or TZP-102 (10 or 20 mg) for 12 weeks. The second study randomized patients to placebo or TZP-102 (10 mg t.i.d.) for 12 weeks. In both studies, symptoms showed similar improvement in all treatment arms.[31] Gastric emptying at baseline and week 12 in the first study also showed no difference between groups. An accompanying editorial outlined some limitations of these negative studies, including allowed concomitant use of antiemetics and opioid analgesics and discordance between symptom responses and gastric emptying results in previous studies with TZP-102.[32] Relamorelin, a subcutaneously administered ghrelin receptor agonist, shown to enhance gastric emptying and to reduce symptoms in a pilot trial in 10 type 1 diabetes patients, is currently undergoing phase 2 testing.[33]

It has been argued that efficacy of traditional prokinetics administered orally may be limited due to poor emptying from the stomach resulting in poor absorption from the duodenum. Parkman et al.[34] evaluated a metoclopramide nasal spray (10 or 20 mg) and tablets (20 mg) q.i.d. in 89 diabetic patients with symptoms suggestive of gastroparesis. Total symptom score improved with all three treatment modalities with the nasal spray showing better tolerance and similar or higher efficacy compared with the tablet. In a placebo-controlled follow-up trial, the metoclopramide nasal spray was studied in diabetic patients with a clinical diagnosis of 'gastroparesis', based on scores on a gastroparesis daily diary. In this US multicenter study, 285 diabetic gastroparesis patients (82.5% type 2 diabetes) were randomized to 4 weeks of placebo, metoclopramide 10 mg or 14 mg sprays t.i.d., 30 min before meals.[35] Symptom assessment used a modified Gastroparesis Cardinal Symptom Index Daily Diary, which assessed the severity of four symptoms (nausea, bloating, early satiety, and upper abdominal pain) on a 0–5 scale. Both metoclopramide doses and placebo generated similar symptom improvement. In a prespecified analysis by sex, female patients showed significantly greater symptom relief to metoclopramide than placebo nasal spray. Metoclopramide was generally well tolerated, although dysgeusia, headache, and dizziness occurred more frequently than with placebo.[35] As this trial did not require documented delayed gastric emptying, it remains unclear how many of these patients truly had gastroparesis and how many had dyspeptic symptoms with normal emptying. Furthermore, as metoclopramide crosses the blood–brain barrier, the nasal spray formulation does not eliminate the risk of extrapyramidal and other central nervous system adverse events.

Psychotropic Agents

Based on the assumption that visceral hypersensitivity contributes to symptom generation, psychotropics, especially tricyclic antidepressants, are often used to treat gastroparesis, although convincing evidence of their efficacy is lacking.[1,2] The issue was addressed in a placebo-controlled trial with nortriptyline in idiopathic gastroparesis, in which a total of 130 patients from nine tertiary care centers were randomized to 15 weeks treatment at bedtime with placebo or nortriptyline in increasing dose every 3 weeks (10, 25, 50, and 75 mg in case of good tolerance). Symptom outcomes were not significantly different between both treatment arms, and interruption rate for adverse events was greater with nortriptyline.[36]

The efficacy of antidepressants in patients with dyspeptic symptoms, including those with delayed emptying, was also addressed in a placebo-controlled multicenter trial using amitriptyline and escitalopram.[37] In this study, 292 functional dyspepsia patients were recruited from eight US centers and randomized to 12 weeks treatment in the evening with placebo, amitriptyline 50 mg, or escitalopram 10 mg. Gastric emptying rate was obtained at baseline and the study evaluated adequate relief of functional dyspepsia symptoms at weekly intervals. Responder rates, defined as adequate relief during at least 5 of the last 10 weeks in the trial, were 40% for placebo, 53% for amitriptyline, and 38% for escitalopram (P = 0.05). In the subgroup with ulcer-like dyspepsia symptoms (pain as the most bothersome symptom), amitriptyline resulted in more adequate relief compared with placebo and escitalopram (67% versus, respectively, 39 and 27%, P = 0.06), whereas no significant benefit was seen in motility-like dyspepsia (pain is not the most bothersome symptom) (46, 41, and 43%). Delayed gastric emptying at baseline was found in 21% of the patients. In the group with normal emptying, responses were similar to the total group, with higher response rates for amitriptyline. In contrast, no difference in response between the three treatment arms was seen in those with delayed emptying at baseline.[37]

Surgery and Endoscopic Therapeutic Approaches

Surgery is often considered the last resort in severe, medically refractory, gastroparesis, and few outcome studies are available.[38] A consecutive series of 35 patients (86% women) who underwent laparoscopic total or near-total gastrectomy for gastroparesis symptoms failing to respond to prokinetic and antiemetic therapies was reported.[39] The primary symptoms were reflux, followed by nausea and abdominal pain. Remarkably, 46% had previously undergone pyloromyotomy, 54% fundoplication, and 23% gastric electrical stimulation treatment. Total gastrectomy with esophagojejunostomy was performed in six, and subtotal gastrectomy with gastrojejunostomy in the others. Surgical anastomotic leak, requiring reintervention, occurred in 17%. No questionnaires were used, but quantification based on clinical reporting with a median follow-up of 6 months, showed that surgery resulted in major improvement of nausea, bloating, and belching while no significant effect was obtained for pain.[39] Although significant weight loss was the rule after surgery (median BMI decreased from 32 to 26), all patients were able to resume or maintain oral feeding. Although this study shows that some patients may improve after near-total gastrectomy, the patient group is unusual with a high rate of previous surgical pyloromyotomies and fundoplications. Moreover, the manuscript lacks details on gastric emptying rate of solids and liquids (may be rapid after vagus lesions, e.g., with fundoplication or pyloromyotomy). Mancini et al.[40] reported a retrospective analysis on the outcome of pyloroplasty in 46 patients with refractory gastroparesis. During follow-up of up to one year, both the GCSI and gastric emptying rate were significantly improved compared with preoperative results.

These surgical studies report favorable outcomes, but in an uncontrolled setting, with relatively short follow-up. Taking into account older literature, the surgical option in gastroparesis should be considered with caution, and temporary nasointestinal tube feeding can be used to evaluate tolerance of nutrients rapidly entering the small bowel.[38] A number of case series of gastric electrical stimulator insertion for refractory gastroparesis were also published, with response rates of up to 60%.[41–43] However, these uncontrolled case series should be considered cautiously, as the two most recent controlled trials with gastric electrical stimulation failed to show significant differences in the blinded periods with stimulator on or off.[44,45]

There are also recent case reports of endoscopic pyloromyotomy, performed through a gastric submucosal tunnel similar to the peroral endoscopic myotomy for achalasia, to provide symptom relief in gastroparesis.[46–48] Although these establish feasibility, longer follow-up in larger series and, if possible, comparative or controlled trial data are needed before implementation in clinical practice at larger scale.