Effect of a proton pump inhibitor on postprandial gastric volume, emptying and symptoms in healthy human subjects: a pilot study

Authors

  • A. B. M. GRUDELL,

    1. Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), Mayo Clinic College of Medicine, Rochester, MN, USA
    Search for more papers by this author
  • M. CAMILLERI,

    1. Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), Mayo Clinic College of Medicine, Rochester, MN, USA
    Search for more papers by this author
  • D. D. BURTON,

    1. Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), Mayo Clinic College of Medicine, Rochester, MN, USA
    Search for more papers by this author
  • D. A. STEPHENS

    1. Clinical Enteric Neuroscience Translational and Epidemiological Research (CENTER), Mayo Clinic College of Medicine, Rochester, MN, USA
    Search for more papers by this author

Dr M. Camilleri, Mayo Clinic, Charlton 8-110, 200 First Street SW, Rochester, MN 55905, USA. E-mail: camilleri.michael@mayo.edu

Summary

Background  In consensus guidelines, proton pump inhibitors (PPIs) are recommended for the treatment of functional dyspepsia. It is unclear whether PPIs change gastric volume or emptying.

Aim  To assess the effect of a PPI, rabeprazole, on gastric volume and emptying and postprandial symptoms.

Methods  In a double-blind, parallel-group placebo-controlled trial, 13 healthy participants were randomized to rabeprazole, 20 mg b.d., or placebo. On day 3, fasting gastric volume was measured using intravenous 99mTc-pertechnate and single photon emission computed tomography (SPECT). After the last dose of study medication, an 111In-chloride egg meal (300 kcal) was ingested, and postprandial gastric volume and emptying were measured by SPECT. Symptom ratings using a visual analogue scale (fullness, nausea, bloating, abdominal pain, aggregate score) were obtained at baseline and 15, 30, 45, 60 and 75 min postprandially. Group comparisons were performed using Mann–Whitney rank sum test.

Results  There were no statistically significant differences in gastric volume or emptying in the two groups. However, there was a borderline increase in gastric volume with rabeprazole compared with placebo. Rabeprazole treatment reduced aggregate postprandial symptoms, particularly fullness, 30 min after the meal (P = 0.01).

Conclusions  In healthy participants, rabeprazole, 20 mg b.d., did not significantly change gastric emptying, but reduced symptoms and had a borderline effect on gastric volume postprandially. The mechanism of reduced postprandial symptoms with a PPI requires further study.

Introduction

Functional dyspepsia, or unexplained postprandial symptoms in the upper abdomen, is a common diagnosis in the practice of primary care physicians and gastroenterologists. Several pathophysiological mechanisms have been proposed to contribute to functional dyspepsia, including delayed gastric emptying, gastric hypersensitivity to acid or distension, and impaired gastric accommodation.1 Proton pump inhibitors (PPIs) have been proposed to be an effective therapy for functional dyspepsia.2 However, some studies have shown that PPIs delay gastric emptying.3, 4 Such a delay in gastric emptying might be expected to aggravate dyspepsia if the baseline gastric emptying is normal or slow. Delayed gastric emptying is observed in ∼40% of patients with functional dyspepsia.1 The management guidelines for dyspepsia suggest that PPI therapy is more effective than placebo or H2 receptor antagonists (H2RAs) in relieving symptoms in patients with uninvestigated dyspepsia.5 The mechanism of PPI benefit in dyspepsia is unclear. Yet, PPI therapy is recommended as a first-line treatment of uninvestigated dyspepsia. Therefore, it is relevant to understand the effects of PPIs on gastric functions and to investigate potential mechanisms of relief of postprandial symptom with PPIs.

One potential benefit of PPIs is the reduction of gastric acid; however, in the postprandial period, the meal typically buffers acid, and the intragastric pH is often highest in this period. An alternative is that PPIs reduce intragastric volume and may thus reduce the volume of content available for reflux into the distal oesophagus. The literature suggests that gastric acid suppression with PPIs or H2RAs reduces intragastric volume. Indeed, some studies suggest that H2RAs reduce intragastric volume more than PPIs and that PPIs are superior to placebo. We elected to study PPIs in this study, because this class is recommended over H2RAs in the management guidelines for dyspepsia. Given the observation that a small proportion of patients with functional dyspepsia have accelerated gastric emptying,6 PPI-induced delay in gastric emptying3, 4 may restore normal emptying in these patients and reduce postprandial symptoms. The recent consensus criteria on functional dyspepsia differentiate an epigastric pain from a postprandial distress syndrome.7 The latter is characterized by bothersome postprandial fullness occurring after ordinary-sized meals and early satiation that prevents patients from finishing a regular meal. Fullness and early satiation are related to gastric volume, accommodation and emptying.1, 6

Therefore, our aim was to understand further the effect of PPIs on gastric volume and emptying and postprandial symptoms.

Methods

Participants

Thirteen normal weight, healthy volunteers were recruited from the local community by public advertisement. Each volunteer was screened by the same investigator physician (ABMG) during the first study visit. Exclusion criteria included (i) history of gastrointestinal surgery; (ii) any clinically significant abnormality in the prestudy screening; (iii) structural, metabolic or functional disease (determined through the shortened screening version of the Bowel Disease Questionnaire8), which affects the gastrointestinal system; (iv) significant disease affecting other systems; (v) intake of medications other than low stable doses of thyroid replacement, oestrogen replacement, low-dose aspirin and birth-control pills and (vi) alcoholics not in remission or known substance abusers.

Females of child-bearing potential underwent a pregnancy test within 48 h of administration of i.v. 99mTc-pertechnate and oral 111In. All participants signed a written consent form that was approved by the Mayo Institutional Review Board and Radiation Safety Committee of the Mayo Clinic.

Study design

In a double-blind, placebo-controlled fashion, participants were randomized to receive the PPI, rabeprazole, 20 mg p.o. b.d. or placebo b.d. during 48 h prior to scintigraphic imaging and during the day of study, as described below. Allocation was concealed.

Rabeprazole pharmacokinetics

Rabeprazole is 52% bioavailable, has a Tmax of 1.5 h, a linear area under the curve in the dose range 10–40 mg and a plasma half-life of 1–2 h. The elimination half-life of the thioether and sulphone metabolites ranges from 1 to 3 h. Rabeprazole is 96% protein bound in plasma and remains detectable in plasma for 9–12 h after single or multiple dosing, respectively.9 Biological activity with this and other PPIs is demonstrable with the first doses of administration. For example, 24-h intragastric pH monitoring shows that after 3 days of administration of rabeprazole 20 mg/day, pH is >4 for 78% of the time.10 This degree of acid suppression is virtually identical to that achieved (79%) with 7 days of administration.11 Moreover, 25–50% of patients receiving PPIs report complete and sustained heartburn relief within the first day of administration, and the daytime heartburn relief and night-time heartburn relief after 2 days of PPIs reach 63% and 72%, respectively.12 These data provide the rationale for the acute 3-day studies we performed to assess the effects of a PPI, rabeprazole, on gastric functions and symptoms.

Measurement of gastric volume and emptying

Prior to 1999, measurement of gastric volume involved the invasive intragastric placement of a polyethylene balloon connected to a barostatic device. In 1999, single photon emission computed tomography (SPECT) was introduced as a non-invasive method to measure fasting and postprandial gastric volume.13 The method used includes a semi-automated segmentation algorithm to identify the intra-abdominal organ with the highest uptake of 99mTc. In order to understand further the pathophysiology of upper gastrointestinal disorders, a method to simultaneously measure gastric volume and emptying via SPECT was developed and validated.14, 15 However, the ability of SPECT to measure gastric volume without a distending force and the contribution of acid secretion to gastric volume measurements have been questioned.16 The present study provided an opportunity to assess the effect of inhibition of acid secretion on gastric volume and emptying using SPECT.

Study procedure

After having received rabeprazole 20 mg or placebo b.d. for 2 days, participants presented to the General Clinical Research Center in the fasted state for gastric emptying and volume measurements using previously validated methods.13, 17–19 After study personnel confirmed that the participant was fasting (8 h) and had taken four doses of the study medication in the prior 2 days, 10 mCi of 99mTc-pertechnate was injected intravenously. Ten minutes later, a gastric image was obtained with SPECT while the participant was still fasting. The final dose of study medication was administered and, 30 min later, a 0.1 mCi 111In-chloride charcoal-labelled egg meal (300 kcal) was ingested. To obtain postprandial gastric volume measurements, images were obtained at 0–15 and 16–30 min after the meal. The average postprandial gastric volume was then calculated.

To obtain gastric emptying measurements, anterior and posterior static gamma camera images were obtained at 0 min and at 2 and 4 h after the meal.

Postprandial symptoms

Symptom data (fullness, nausea, bloating, abdominal pain) were also obtained using a 10-cm visual analogue scale anchored with the words ‘none’ and ‘worst ever’. These data were obtained at baseline prior to the meal and every 15 min for 75 min after the meal. The primary end point of analysis of symptoms was the aggregate score at 30 min postprandially, as in the prior studies.20

Statistical analysis and power

The primary end points were the change in gastric volume after the meal, proportion of gastric emptying at 4 h and aggregate symptoms score 30 min postprandially. Secondary end points were the proportion of gastric emptying at 2 h, the fasting gastric volume and individual postprandial symptoms. Group comparisons were then performed using the Mann–Whitney rank sum test.

Given the significance of postprandial fullness and satiation in dyspepsia and their relation to gastric accommodation,1 the study sample size was powered to detect a meaningful effect of the change in the stomach's volume after a meal. Thus, prior to the study, and based on a two-tailed, unpaired t-test, there was 80% power to detect the following effect sizes in response to treatment: 23% difference in postprandial gastric volume, 54% difference in fasting gastric volume and 48% difference in gastric emptying based on the results in prior studies in healthy human subjects in our laboratory using the same methods.13, 17–19 After the study, we analysed the sample size required to demonstrate a significant effect based on the pooled variance and differences in mean gastric emptying at 4 h and change in gastric volume after the meal.

Results

Thirteen volunteers were screened and were enrolled in the study. One participant's data were excluded after subcutaneous infiltration (rather than i.v. injection) of 99mTc-pertechnate. Participant demographics are shown in Table 1.

Table 1.   Patient demographics, fasting and postprandial gastric volumes, symptom rating 30 min after meal, and gastric emptying percentages
 Placebo (n = 5)Rabeprazole (n = 7)
Age (years), mean (SD)36 (22–45)32 (28–43)
Gender (m/f)3/23/4
Body mass index (kg/m2), mean (SD)25 (21–29)25 (21–29)
Fasting gastric volume (mL), median (IQR)222 (171–307)290 (221–316)
0–15 min postprandial gastric volume (mL), median (IQR)816 (775–883)951 (915–976)
16–30 min postprandial gastric volume (mL), median (IQR)798 (737–881)864 (840–922)
Average postprandial gastric volume (mL), median (IQR)810 (761–875)909 (877–944)
Delta (average) postprandial minus fasting gastric volume (mL), median (IQR)588 (486–670)609 (582–675)
Fullness (mm, VA scale), median (IQR)31 (22–45)15 (5–24)
Nausea (mm, VA scale), median (IQR)0 (0–9)0 (0–0)
Bloating (mm, VA scale), median (IQR)8 (0–34)0 (0–0)
Pain (mm, VA scale), median (IQR)0 (0–0)0 (0–0)
GE 2 h (%) median (IQR)34 (30–36)35 (21–40)
GE 4 h (%) median (IQR)85 (77–85)65 (59–84)

Effect of rabeprazole on gastric volume and emptying

An example of fasting and postprandial gastric images obtained with SPECT can be seen in Figure 1. Fasting and postprandial gastric volumes, symptom ratings 30 min after the meal, and gastric emptying percentages are shown in Table 1. There were no statistically significant differences in the two groups (rabeprazole vs. placebo). The interquartile ranges of the 0–15 min and average postprandial gastric volumes do not overlap between the two groups, which suggests a trend towards difference between the groups. However, these results were not statistically significant (P = 0.15 and P = 0.27, respectively). The change in volume (average postprandial minus fasting) was also not significant (P = 0.76).

Figure 1.

 Fasting (left panel) and postprandial (right panel) images of stomach via SPECT using Analyze PC 2.5 software system.

Effect of rabeprazole on postprandial symptoms

The aggregate symptom scores for each participant at baseline prior to the meal and at 15, 30, 45, 60 and 75 min postprandially are plotted in Figure 2. Participants in the rabeprazole group had lower aggregate symptoms 30 min after the meal, as seen in Figure 3. In Table 1, note that the fullness scores accounted for the largest component of the postprandial aggregate scores, and that rabeprazole was associated with virtually complete resolution of postprandial symptoms.

Figure 2.

 Aggregate symptoms for each patient at baseline prior to meal and 15, 30, 45, 60 and 75 min after meal ingested.

Figure 3.

 Aggregate postprandial symptoms 30 min after meal ingested.

Discussion

Our non-invasive study shows that a PPI, rabeprazole, reduced postprandial symptoms 30 min after ingestion of a relatively small test meal. Prior studies had shown delayed gastric emptying with PPIs or H2RAs.3, 4, 21 However, this study has demonstrated that the PPI, rabeprazole, 20 mg b.d., does not significantly alter gastric emptying or gastric volume, which were measured simultaneously using a validated, non-invasive imaging device, SPECT.15, 17 It is important to clarify that the method used does not evaluate the volume of intragastric content, but the volume estimated from a measurement of the cross-sectional area of transaxial sections of the stomach. The 99mTc injection is taken up by both the gastric parietal cells and the mucus-secreting epithelial cells and, therefore, the entire gastric mucosa is visualized by this technique. Intestinal mucus uptake is not shown in the gastric images obtained with the Analyze PC 2.5 (Biomedical Imaging Resource, Mayo Foundation, Rochester, MN, USA) software system, as the stomach was separated from background noise using the software's semi-automated segmentation algorithm that requires the user to identify an appropriate seed point using a greyscale threshold.22 The seed point used is the area with the highest 99mTc counts. On imaging the abdomen, the isotope accumulating in abdominal organs (other than stomach and kidneys) is negligible, suggesting either that the density of mucus-secreting cells in the intestine is too low or that the process that facilitates the uptake of 99mTc into gastric mucus-secreting cells is not present in intestinal mucus-secreting cells. There is significant heterogeneity in the function of mucus-secreting cells in different parts of the gastrointestinal tract.

The observation that healthy volunteers taking a PPI had a decreased sensation of fullness 30 min after a meal is intriguing. Protons are known to activate the release of bioactive mediators, such as serotonin and cholecystokinin, from the upper small intestine, which activate sensory mechanisms including vagal afferents.23, 24 In fact, several groups have reported increased sensitivity of the duodenum to infused acid.25, 26 However, the precise mechanism for the acid-induced postprandial symptoms is unclear. Tack and colleagues demonstrated that high concentrations of acid were associated with significant changes in motor function such as increased fundic tone or induction of migrating motor complex (MMC)-like activity that could conceivably alter thresholds for sensation to distensions of the fundus.26 It is also unclear whether function of acid-sensitive sensory nerves would be altered by PPIs; this seems unlikely in our study in healthy subjects in the absence of tissue ischaemia.

The fasting and postprandial gastric volumes measured in the two treatment groups were similar. Postprandial gastric secretion has been reported to range from 40 to 200 mL/h in normal healthy volunteers.27 Therefore, it would be reasonable to assume that this volume could contribute to the gastric volume measured with SPECT, although this would also depend on the rate at which the additional volume is emptied from the stomach. However, the present study suggests that the presumed inhibition of acid secretion with rabeprazole 20 mg b.d. does not significantly reduce the volume of the stomach measured either during fasting or postprandially. These data support previous studies of simultaneous measurement of gastric volume and emptying showing that gastric volume measured with SPECT includes more than the volume of ingested meal14, 15 and gastric acid secretion. Our data are also consistent with the observation by another group28 that acid inhibition does not negatively impact on the ability of the 99mTc-SPECT method to image the gastric mucosa, as seen in Figure 1.

We cannot exclude the possibility that other acid suppressing medications, such as H2RAs, may influence the volume of the stomach. In fact, studies in the anaesthesia literature show that an H2RA may be more effective than a PPI in reducing gastric fluid volume and the risk of aspiration pneumonitis; other studies show similar efficacies.29–31

The limitations of the study include the evaluation of healthy controls rather than patients with dyspepsia and the use of a small meal to assess postprandial symptoms. Clearly, further studies are needed in dyspepsia with a larger challenge meal to assess the influence of PPI therapy on symptoms in patients. Another potential limitation is the sample size tested. However, our study was based on a parallel-group design in a total of 12 subjects and had sufficient power to detect a 23% difference in postprandial change gastric volume and 48% difference in gastric emptying. The 23% difference in gastric accommodation (postprandial change in gastric volume) is consistent with a clinical difference between health and dyspepsia in studies using the barostatic method1 or SPECT.20 A post hoc analysis using the observed variation in gastric volumes and emptying in this study showed that 31 participants per treatment group would be required to show a significant effect of rabeprazole on t1/2gastric emptying, 25 participants per treatment group to show a significant effect on gastric emptying at 4 h and 48 per group to show an effect on postprandial gastric volume. Vidon et al. have shown that another PPI, lansoprazole, does not alter gastric emptying.32 However, the post hoc analysis suggests that, if there is a 0.2 (20%) difference in the amount of meal emptied at 4 h with rabeprazole treatment, we would be able to demonstrate a significant effect of rabeprazole vs. placebo with n = 12 per treatment group. Thus, our study cannot completely exclude the possibility that PPIs do not retard gastric emptying. Similarly, we cannot exclude the potential for a type II error in our appraisal of the postprandial gastric volume that shows a borderline difference in the two treatment groups.

The data acquired from this pilot study provide a basis for planning future studies of the effects of PPIs on gastric function in patients with functional dyspepsia.

Further study in patients with postmeal dyspepsia is needed to determine the role of PPIs in reducing such symptoms. To date, the clinical efficacy has been demonstrable in patients with reflux-like dyspepsia.2 The effect of PPIs in more carefully selected patients with non-reflux-associated dyspepsia, such as those with early satiety and postprandial fullness, would be worthy of further study.

Acknowledgements

This study was supported in part by General Clinical Research Center grant RR00585 and grants R01 DK067071 and K24 DK02638 to Dr Camilleri from National Institutes of Health.

Ancillary