Effect of netazepide, a gastrin/CCK₂ receptor antagonist, on gastric acid secretion and rabeprazoleinduced hypergastrinaemia in healthy subjects

Abstract

Aims

To compare gastric acid suppression by netazepide, a gastrin/CCK₂ receptor antagonist, with that by a proton pump inhibitor (PPI), and to determine if netazepide can prevent the trophic effects of PPI-induced hypergastrinaemia.

Methods

Thirty healthy subjects completed a double-blind, randomized, parallel group trial of oral netazepide and rabeprazole, alone and combined, once daily for 6 weeks. Primary end points were: basal and pentagastrin-stimulated gastric acid and 24 h circulating gastrin and chromogranin A (CgA) at baseline, start and end of treatment, gastric biopsies at baseline and end of treatment and basal and pentagastrin-stimulated gastric acid and dyspepsia questionnaire after treatment withdrawal.

Results

All treatments similarly inhibited pentagastrin-stimulated gastric acid secretion. All treatments increased serum gastrin, but the combination and rabeprazole did so more than netazepide alone. The combination also reduced basal acid secretion.

Rabeprazole increased plasma CgA, whereas netazepide and the combination reduced it. None of the biopsies showed enterochromaffin-like (ECL) cell hyperplasia. Withdrawal of treatments led neither to rebound hyperacidity nor dyspepsia.

Conclusions

Netazepide suppressed pentagastrin-stimulated gastric acid secretion as effectively as did rabeprazole. The reduction in basal acid secretion and greater increase in serum gastrin by the combination is consistent with more effective acid suppression. Despite our failure to show rabeprazole-induced ECL cell hyperplasia and rebound hyperacidity, the increase in plasma CgA after rabeprazole is consistent with a trophic effect on ECL cells, which netazepide prevented. Thus, netazepide is a potential treatment for the trophic effects of hypergastrinaemia and, with or without a PPI, is a potential treatment for acid-related conditions.

What is Already Known about this Subject

• Gastrin controls gastric acid secretion and mucosal cell growth, especially of enterochromaffin-like (ECL) cells, which express gastrin/CCK₂ receptors and release chromogranin A (CgA) into the circulation when stimulated by gastrin.

• In non-clinical studies, acid suppression by a proton pump inhibitor (PPI) causes hypergastrinaemia which results in ECL cell growth and, after PPI withdrawal, rebound hyperacidity. Netazepide is an orally active, selective gastrin/CCK₂ receptor antagonist, which suppresses acid production and prevents the trophic effects of PPI-induced hypergastrinaemia.

• In healthy subjects, oral netazepide causes dose-dependent, persistent inhibition of gastric acid secretion, which leads to increased serum gastrin.

What this Study Adds

• In healthy subjects, netazepide and the PPI rabeprazole were similarly effective at suppressing pentagastrin-stimulated gastric acid secretion and increasing serum gastrin.

• Rabeprazole increased plasma CgA, a sign of ECL cell hyperactivity, whereas netazepide reduced plasma CgA, a sign of ECL cell hypoactivity. Netazepide also prevented the increase in CgA resulting from rabeprazole-induced hypergastrinaemia, probably because netazepide blocks gastrin/CCK₂ receptors on ECL cells.

Introduction

Gastrin, a hormone produced by G cells in the gastric antrum, stimulates gastric acid secretion and the proliferation, migration, differentiation and anti-apoptosis of gastric epithelial cells 1,2. Gastrin activates gastrin (CCK₂) receptors on enterochromaffin-like (ECL) cells in the oxyntic mucosa to secrete histamine, which in turn stimulates adjacent parietal cells to secrete acid. Acid secretion is mediated by the proton pump via histamine H₂- and muscarinic M₃-receptors on parietal cells 3.

Acid suppression by a proton pump inhibitor (PPI) induces hypergastrinaemia 4. Prolonged PPI-induced hypergastrinaemia may cause ECL cell 5 and parietal-cell hyperplasia 6. PPI withdrawal may result in rebound hyperacidity 7,8 and dyspepsia 9,10. Other causes of hypergastrinaemia include chronic autoimmune gastritis (CAG), Zollinger-Ellison syndrome (ZES) and H. pylori-induced gastritis 11,12.

There are several potential clinical indications for a gastrin/CCK₂ receptor antagonist. Many have been described 13, but none has been developed into a medicine, mainly because of problems with potency, selectivity for CCK₁ or CCK₂ receptors, agonist activity and oral bioavailability. In non-clinical studies, netazepide (YF476) is a potent gastrin/CCK₂ receptor antagonist and is highly selective for the CCK₂ receptor, has good oral bioavailability and suppresses gastric acid secretion 14–17. In animal models, netazepide prevents ECL cell hyperplasia and the rebound hyperacidity that follows PPI withdrawal 18, and also prevents development 19,20 and causes regression 20 of ECL cell tumours. Acid suppression by netazepide leads to hypergastrinaemia in animals 19,21–24, but causes neither ECL cell growth nor rebound hyperacidity 18, because netazepide blocks the gastrin/CCK₂ receptors on ECL cells.

We have characterized the clinical pharmacology of netazepide as a gastric acid suppressant, in healthy subjects 25–27. Single oral doses caused dose-dependent inhibition of pentagastrin-stimulated gastric acid secretion, which persisted after repeated doses. Repeated doses also caused hypergastrinaemia, which is consistent with persistent acid suppression, in spite of partial tolerance to the effect of netazepide on gastric pH 26.

Our aim was to study healthy subjects to (i) compare acid suppression by netazepide and the PPI rabeprazole, alone and in combination, (ii) determine whether netazepide can prevent the effects of PPI-induced hypergastrinaemia, namely ECL cell hyperplasia and rebound hyperacidity and dyspepsia and (iii) show that hypergastrinaemia induced by netazepide causes neither ECL cell hyperplasia nor rebound hyperacidity and dyspepsia.

Methods

We complied with the ICH Guideline for Good Clinical Practice. The Medicines and Healthcare products Regulatory Agency, UK and Ravenscourt Ethics Committee approved the study. Subjects gave written, informed consent. We did the study during 2007–2010 and registered it as ClinicalTrials.gov NCT01699113.

Materials

Trio Medicines Ltd, England, supplied netazepide 50 mg capsules, matching placebo and rabeprazole (Pariet®; Eisai) 20 mg delayed-release, enteric-coated tablets. Over-encapsulation ensured that treatments looked identical. Cambridge Laboratories (Wallsend, England) supplied pentagastrin 500 μg 2 ml⁻¹. Subjects were supplied with an antacid/alginate (Gaviscon®; Reckitt Benckiser) that they could take after treatment withdrawal, if needed.

Study design

The study was randomized, double-blind and parallel group, and required three cohorts of 10 healthy, H. pylori negative (¹³C-urea breath test), non-smoking, men or women not at risk of pregnancy, with normal gastroscopy and no history of dyspepsia and taking no medicines. We deemed subjects healthy by: medical history and examination; ECG; tests for drugs of abuse, hepatitis B and C and HIV 1 and 2; and blood and urine tests. We allocated subject numbers sequentially, using separate randomization schedules for men and women. Treatments consisted of netazepide 100 mg and rabeprazole 20 mg, alone and in combination, once daily by mouth for 6 weeks.

We studied subjects on 3 days (days −14, −13 and −12) before starting treatment (baseline); the first 2 days (days 1 and 2) and last 3 days (days 40, 41 and 42) of treatment and 2 days (days 57 and 58) starting 15 days after the end of treatment. Subjects were resident on study days, and had clinic visits after 2 and 4 weeks’ treatment and 1 week after withdrawal, for safety checks. Table 1 shows the procedures during residence.

Table 1.

Study plan for procedures during residence

Procedure				Treatment*
	Baseline			Start		End			Withdrawal
	Day			Day		Day			Day
	−14	−13	−12	1	2	40	41	42	57	58
24 h gastric pH	•			•		•			•
Serum gastrin profile	•			•		•
Plasma CgA profile	•			•		•
Pharmacokinetics				•		•
Pentagastrin infusion		•			•		•			•
Gastric biopsies			•					•

^*42 days of netazepide and rabeprazole, alone and combined.

We gave subjects a diary card to record adverse events and compliance during treatment, and a validated questionnaire 28 and a diary card to record dyspepsia symptoms and antacid/alginate usage, respectively, during 3 weeks after treatment withdrawal.

We assessed safety and tolerability by medical examination, ECG, blood and urine tests and adverse events.

Ambulatory gastric pH

At baseline and on days 1, 40 and 57, we measured gastric pH continuously for 0.5 h before and 24 h after dosing, as described previously 25. Subjects fasted overnight before treatment, which they took at the same time (09.00 h−10.00 h) in each session. They rested for 6 h after dosing and were ambulant thereafter. After dosing, they consumed standard meals and water at standard times.

Pentagastrin tests

At baseline and on days 2, 41 and 58, subjects fasted overnight. At 10.00 h–11.00 h, we passed a nasogastric tube, to collect gastric aspirate by continuous suction while the subject was semi-recumbent, as described previously 27. First, we collected basal aspirate every 15 min for 30 min. We allowed 55 min for absorption of treatment 27, after which we collected aspirate for 5 min, to empty the stomach. Then, we gave pentagastrin intravenously 0.6 μg kg⁻¹ h⁻¹ for 2 h by infusion pump, and collected aspirate continuously in 15 min epochs. We measured volume, and then titratable acidity, of each collection by pH meter and automatic titrator (Radiometer) calibrated regularly with buffers. We used NaOH (0.1 m) as base and calculated H⁺ secretion rate in μmol min⁻¹.

Gastroscopy and mucosal biopsies

At baseline and on day 40, we took four biopsies of the gastric oxyntic mucosa from the anterior and posterior wall, and prepared and coded specimens for blinded assessment of ECL cells by histology, immuno-histochemistry of chromogranin A (CgA), histidine decarboxylase (HDC), vesicular monoamine transfer (VMAT2) and Ki67 and electron microscopy (EM) of ECL cell profile area, nuclear area and volume density, cytoplasmic area, numbers and volume densities of granules, secretory vesicles, microvesicles, secondary lysosomes, vacuoles, lipofuscin bodies, and volume densities of rough endoplasmic reticulum and Golgi complex. Preparation of specimens for assessment and methods are described elsewhere 18.

Serum gastrin and plasma CgA

At baseline and on days 1 and 40, we collected blood before and frequently for 24 h after dosing, separated serum or plasma and stored samples at −20°C until assay by ELISA (Gastrin: Immulite 2000, DPC. CV = 6.9%; CgA: DAKO. CV = 7.2%).

Plasma netazepide and rabeprazole

On days 1 and 40, we collected blood before and frequently for 24 h after dosing, separated plasma and stored samples at −20°C until assay by validated HPLC/MS methods 29. The lower limit of quantification was 0.5 and 0.25 ng ml⁻¹ for netazepide and rabeprazole, respectively.

Dyspepsia

Subjects completed the dyspepsia questionnaire and recorded antacid/alginate usage during each week of the 3 week post-treatment period. The questionnaire uses 4 or 5 point Likert scales to measure frequency and severity of 15 upper gastrointestinal symptoms, and the bother they cause 28.

Assessment of compliance

We assessed compliance by diary card and capsule counts. Subjects wore a wristwatch with an alarm and we telephoned them weekly, to remind them to take their treatment.

Statistics

Sample size

The study was exploratory in nature. We chose H⁺ secretion rate in pentagastrin-stimulated gastric aspirate as the primary end point, and calculated sample size and power (n = 10; α = 0.05; power at least 80%), using data from a previous study 27. We opted not to adjust P values for multiple comparisons because of the exploratory nature of the study.

Pharmacodynamics

For AUC(0,24 h) of 24 h gastric pH, pentagastrin-stimulated volume and H⁺ secretion rate of gastric aspirate (60−180 min after dosing), we summarized changes from baseline (mean and 95% confidence intervals on days −14, −13 and −12) at the start (days 1 and 2), end (days 40, 41 and 42) and after withdrawal (days 57 and 58) of each treatment, and used anova to test for significant differences (P < 0.05). For basal H⁺ secretion rate of gastric aspirate (0–30 min before dosing) and AUC(0,24 h) of serum gastrin and plasma CgA, before carrying out the analyses, we log-transformed the data because it was not normally distributed.

For immunoreactive cells and ECL cell ultrastructure we used the estimates from anova to make pairwise comparisons among subject groups at the start and end of treatment and a two-tailed t-test to compare each treatment.

Pharmacokinetics

We used WinNonlin to derive pharmacokinetic parameters by standard non-compartmental analysis and SAS for Windows for statistical analysis, as described previously 25.

Dyspepsia

We summarized the number of subjects per treatment group with symptoms during each week of the withdrawal period, added the scores for the three groups of symptoms to obtain a dyspepsia score, and compared treatment groups informally. We also summarized antacid/alginate usage and compared groups in terms of number of subjects and doses per treatment group.

Results

Subjects

We entered 32 subjects (17 men, 15 women; 28 White, four Black), of whom 30 completed the study, as required. One woman and one man (both on combination treatment) withdrew on day 2 and 42, respectively, for reasons unrelated to treatment. We used data from both for safety assessments and pharmacokinetic data from the subject who withdrew on day 42. Mean (range) age and body mass index of subjects receiving netazepide, rabeprazole and combination treatments were 38.7 (21–66) years and 25 kg m⁻², 30.4 (21–67) years and 24.4 kg m⁻² and 35.9 (21–57) years and 23.7 kg m⁻², respectively. There were similar numbers of men and women on each regimen.

Safety, tolerability and compliance

All treatments were well tolerated. Adverse events were minor, transient and occurred across treatments. There were no clinically relevant changes in safety assessments. Treatment compliance based on capsule counts and diary cards was 99%.

Pharmacokinetics (Tables S1 and S2)

Plasma netazepide concentrations were low, about 10–15% of those in our previous studies 25,26, and more variable. Mean AUC(0,24 h) of rabeprazole was higher when combined with netazepide than after rabeprazole alone, but not significantly so.

Pharmacodynamics

24 h ambulatory gastric pH (Table S3; Figure 1)

Figure 1.

24 h ambulatory pH: mean (SD) AUC(0,24 h) of pH at baseline, on days 1 and 40 (treatment), and on day 57 (withdrawal). *P < 0.05 *** P ≤ 0.0001 netazepide () rabeprazole () combination ()

Baseline: AUC(0,24 h) of pH was similar across groups.

Day 1: Compared with baseline, netazepide (mean change 18.7; 95% CI = 5.9, 31.5; P = 0.009) and rabeprazole (mean change 22.4; 95% CI = 8.3, 36.5; P = 0.006) were equally effective at increasing AUC(0,24 h) of pH. The combination was more effective than netazepide (mean difference −17.8; 95% CI = −33.2, −2.5; P = 0.03), but not rabeprazole (mean difference −14.1; 95% CI = −30.4, 2.2; P = 0.09).

Day 40: Netazepide was about as effective as it was on day 1 (mean change from baseline 12.0; 95% CI = 1.6, 22.4; P = 0.03), whereas rabeprazole (mean change 57.7; 95% CI = 46.4, 69.0; P ≤ 0.0001) and the combination (mean change 59.8; 95% CI = 51.5, 68.1; P ≤ 0.0001) were much more effective. Rabeprazole (mean difference 45.7; 95% CI = 31.5, 60.0; P ≤ 0.0001) and the combination (mean difference −47.9; 95% CI = −60.2, −35.5; P ≤ 0.0001) were more effective than netazepide, but there was no significant difference between the combination and rabeprazole (mean difference −2.1; 95% CI = −15.2, 10.9; P = 0.73).

Day 57: There were no significant changes from baseline, and no significant differences among treatments.

Pentagastrin stimulation test (Tables 2 and 3; Figure 2)

Table 2.

Basal H⁺ secretion rate (μmol min⁻¹) 0–30 min before dosing: (A) mean (SD) and geometric mean (SD) basal H⁺ secretion rate (μmol min⁻¹) at baseline, on days 2 and 41 of treatment and day 58 (withdrawal); (B) ratio of treatment means adjusted for baseline for basal H⁺ secretion rate on days 2, 41 and 58

A
Day	Treatment	Mean H+ secretion rate (SD) (μmol min−1)	Geometric mean H+ secretion rate (μmol min−1)	Geometric mean of ratio to baseline	95% confidence intervals
Baseline	Netazepide	81.4 (77.0)	49.9	–	–	–
	Rabeprazole	63.9 (68.4)	29.3	–	–	–
	Combination	36.4 (39.2)	24.4	–	–	–
2	Netazepide	44.7 (35.0)	34.3	0.69	0.23	2.01
	Rabeprazole	55.1 (83.9)	73.7	1.59	0.16	15.61
	Combination	8.4 (8.6)	8.8	0.29	0.06	1.30
41	Netazepide	59.8 (36.6)	45.9	0.92	0.33	2.57
	Rabeprazole	28.3 (31.8)	15.9	0.54	0.11	2.53
	Combination	6.1 (10.4)	5.9	0.16	0.05	0.58
58	Netazepide	72.4 (48.7)	57.8	1.16	0.37	3.65
	Rabeprazole	135.9 (132.3)	54.7	1.87	0.58	5.98
	Combination	60.4 (83.1)	19.7	0.77	0.16	3.75

B
Day	Treatment comparison	Ratio of geometric treatment means adjusted for baseline	95% confidence intervals		P values
2	Rabeprazole vs. netazepide	2.31	0.34	15.65	0.3601
	Rabeprazole vs. combination	5.51	0.59	51.57	0.1198
	Netazepide vs. combination	2.38	0.47	12.18	0.2750
41	Rabeprazole vs. netazepide	0.58	0.11	3.14	0.5083
	Rabeprazole vs. combination	3.31	0.45	24.33	0.2176
	Netazepide vs. combination	5.67	1.25	25.71	0.0273
58	Rabeprazole vs. netazepide	1.61	0.35	7.35	0.5173
	Rabeprazole vs. combination	2.43	0.42	14.11	0.3007
	Netazepide vs. combination	1.51	0.26	8.65	0.6250

Table 3.

Pentagastrin-induced gastric aspirate 60–180 min after dosing at baseline, on days 2 and 41 of treatment and day 58 (withdrawal): (A) mean (SD) volume (ml min⁻¹); (B) mean (SD) H⁺ secretion rate (μmol min⁻¹) at baseline, on days 2 and 41 of treatment and day 58 (withdrawal), and mean change from baseline on days 2, 41 and 58; (C) estimated differences between mean change from baseline for volume and H⁺ secretion rate on days 2, 41 and 58

A
Day	Treatment	Mean volume (ml min−1)	Mean change from baseline	95% confidence intervals
Baseline	Netazepide	4.0 (0.7)	–	–	–
	Rabeprazole	4.2 (1.5)	–	–	–
	Combination	3.9 (1.2)	–	–	–
2	Netazepide	2.1 (1.0)	−1.9	−2.88	−0.97
	Rabeprazole	1.8 (0.7)	−2.6	−3.52	−1.74
	Combination	2.6 (1.4)	−1.3	−2.66	0.04
41	Netazepide	2.3 (0.7)	−1.8	−2.64	−0.98
	Rabeprazole	2.2 (1.1)	−1.9	−2.91	−0.93
	Combination	1.9 (0.8)	−2.2	−3.49	−0.99
58	Netazepide	4.4 (1.0)	0.45	−0.57	1.48
	Rabeprazole	4.6 (1.7)	0.43	−0.49	1.34
	Combination	3.9 (1.4)	−0.26	−1.19	0.66

B
Day	Treatment	Mean H+ secretion rate (μmol min−1)	Mean change from baseline	95% confidence intervals
Baseline	Netazepide	317.0 (123.7)	–	–	–
	Rabeprazole	348.3 (166.2)	–	–	–
	Combination	286.5 (151.8)	–	–	–
2	Netazepide	83.6 (61.6)	−233	−330	−137
	Rabeprazole	96.8 (85.3)	−252	−349	−154
	Combination	33.6 (39.6)	−253	−355	−151
41	Netazepide	111.6 (55.1)	−205	−273	−138
	Rabeprazole	110.4 (52.6)	−238	−333	−143
	Combination	21.1 (24.3)	−263	−364	−162
58	Netazepide	389.8 (72.0)	72.8	−3.9	149
	Rabeprazole	466.4 (167.1)	118	−10.1	246
	Combination	304.3 (134.4)	12.5	−59.7	84.6

C
Day	Treatments compared	Gastric aspirate	Difference between treatment means adjusted for baseline	95% confidence intervals
2	Rabeprazole vs. netazepide	Volume	−0.71	−1.94	0.52
		H+ secretion rate	−18.1	−146	110
	Rabeprazole vs. combination	Volume	−1.32	−3.03	0.39
		H+ secretion rate	1.5	−133	136
	Netazepide vs. combination	Volume	−0.61	−2.23	1.00
		H+ secretion rate	19.5	−115	154
41	Rabeprazole vs. netazepide	Volume	−0.11	−1.32	1.09
		H+ secretion rate	−32.5	−140	75
	Rabeprazole vs. combination	Volume	0.32	−1.15	1.79
		H+ secretion rate	25.3	−105	155
	Netazepide vs. combination	Volume	0.43	−0.99	1.85
		H+ secretion rate	57.8	−58	174
58	Rabeprazole vs. netazepide	Volume	−0.03	−1.29	1.24
		H+ secretion rate	45.3	−93.4	184
	Rabeprazole vs. combination	Volume	0.69	−0.52	1.90
		H+ secretion rate	106	−30.9	242
	Netazepide vs. combination	Volume	0.72	−0.55	1.99
		H+ secretion rate	60.3	−37.4	158

Figure 2.

Gastric aspirate 60–180 min after dosing at baseline, on days 2 and 41 (treatment) and on day 58 (withdrawal). (A) mean (SD) pentagastrin-induced volume (ml min⁻¹), (B) mean (SD) pentagastrin-induced H⁺ secretion rate (μmol min⁻¹) and (C) geometric mean of basal H⁺ secretion rate (μmol min⁻¹). *significant ratio to baseline. netazepide () rabeprazole () combination ()

Baseline: The groups were not well matched for basal H⁺ secretion rate (μmol min⁻¹) before treatment, but mean secretion rate in all groups was low compared with secretion after pentagastrin.

Day 2: There were no significant differences in basal H⁺ secretion rate among treatments. Netazepide (mean change −1.9; 95% CI = −2.88, −0.97), rabeprazole (mean change −2.6; 95% CI = −3.52, −1.74), and the combination (mean change −1.3; 95% CI = −2.66, 0.04) all reduced pentagastrin-stimulated volume (ml min⁻¹). Also, netazepide (mean change −233; 95% CI = −330, −137), rabeprazole (mean change −252; 95% CI = −349, −154), and the combination (mean change −253; 95% CI = −355, −151) all reduced pentagastrin-stimulated H⁺ secretion rate. There were no significant differences among treatments for pentagastrin-stimulated volume or for H⁺ secretion rate.

Day 41: The combination significantly reduced basal H⁺ secretion rate compared with baseline. There was no significant difference between netazepide and rabeprazole (P = 0.5083) or rabeprazole and the combination (P = 0.2176). However, the combination (P = 0.0273) reduced basal H⁺ secretion rate more than netazepide alone. Netazepide (mean change −1.8; 95% CI = −2.64, −0.98), rabeprazole (mean change −1.9; 95% CI = −2.91, −0.93), and the combination (mean change −2.2; 95% CI = −3.49, −0.99) all reduced pentagastrin-stimulated volume. And, netazepide (mean change −205; 95% CI = −273, −138), rabeprazole (mean change −238; 95% CI = −333, −143), and the combination (mean change −263; 95% CI = −364, −162) all reduced H⁺ secretion rate. There were no significant differences among treatments for pentagastrin-stimulated volume or for H⁺ secretion rate.

Day 58: Compared with baseline, basal H⁺ secretion rate was higher after rabeprazole withdrawal, albeit not significantly so. There were no significant differences among treatments with respect to basal H⁺ secretion rate or to pentagastrin-stimulated responses, although the H⁺ secretion rate response to pentagastrin was highest after rabeprazole withdrawal.

Serum gastrin (Table 4; Figure 3) 

Table 4.

Serum gastrin: (A) mean (SD) and geometric mean AUC(0,24 h) at baseline and on days 1 and 40 of treatment, and geometric mean of ratio to baseline on days 1 and 40; (B) ratio of geometric means adjusted for baseline for AUC(0,24 h) of gastrin on days 1 and 40

A
Day	Treatment	Mean AUC(0,24 h) (pmol l−1 h)	Geometric mean (pmol l−1 h)	Geometric mean of ratio to baseline	95% confidence intervals
Baseline	Netazepide	581 (419)	505	–	–	–
	Rabeprazole	449 (168)	414	–	–	–
	Combination	629 (365)	559	–	–	–
1	Netazepide	1236 (1347)	913	1.81	1.47	2.23
	Rabeprazole	920 (528)	785	1.90	1.43	2.51
	Combination	1584 (1263)	1252	2.24	1.74	2.88
40	Netazepide	1552 (1504)	1177	2.33	1.75	3.10
	Rabeprazole	1841 (1030)	1594	3.86	2.67	5.56
	Combination	3874 (2612)	3081	5.52	3.81	7.98

B
Day	Treatments compared	Ratio of geometric treatment means adjusted for baseline	95% confidence intervals		P values
1	Rabeprazole vs. netazepide	1.05	0.76	1.45	0.7561
	Rabeprazole vs. combination	0.85	0.60	1.20	0.3340
	Netazepide vs. combination	0.81	0.60	1.09	0.1553
40	Rabeprazole vs. netazepide	1.66	1.08	2.55	0.0243
	Rabeprazole vs. combination	0.70	0.43	1.13	0.1368
	Netazepide vs. combination	0.42	0.27	0.65	0.0006

Figure 3.

Serum gastrin: (A) mean (SD) concentration vs time on day 40 of treatment. (B) geometric mean AUC(0,24 h) at baseline and on days 1 and 40 of treatment. *significant ratio to baseline. (A) , netazepide; , rabeprazole; , combination; (B) , netazepide; , rabeprazole; , combination

Baseline: AUC(0,24 h) of serum gastrin (pmol l⁻¹h) was similar across groups.

Day 1: Compared with baseline, netazepide, rabeprazole and the combination all increased AUC(0,24 h) of gastrin. There were no significant differences among treatments.

Day 40: Netazepide, rabeprazole and the combination all increased AUC(0,24 h) of gastrin significantly compared with baseline. There was no significant difference between rabeprazole and the combination (P = 0.1368). However, compared with netazepide, rabeprazole (P = 0.0243) and the combination (P = 0.0006) increased AUC(0,24 h) of gastrin.

Plasma chromogranin A (Table 5; Figure 4)

Table 5.

Plasma CgA: (A) mean (SD) and geometric mean AUC(0,24 h) at baseline and on days 1 and 40 of treatment, and geometric mean of ratio to baseline on days 1 and 40; (B) ratio of geometric means adjusted for baseline for AUC(0,24 h) of CgA on days 1 and 40

A
Day	Treatment	Mean AUC(0,24 h) (U l−1 h)	Geometric mean (U l−1 h)	Geometric mean of ratio to baseline	95% confidence intervals
Baseline	Netazepide	241 (55)	235	–	–	–
	Rabeprazole	218 (69)	209	–	–	–
	Combination	261 (49)	257	–	–	–
1	Netazepide	203 (51)	198	0.84	0.76	0.93
	Rabeprazole	241 (96)	226	1.09	0.96	1.23
	Combination	230 (68)	221	0.86	0.77	0.97
40	Netazepide	211 (55)	205	0.87	0.77	0.99
	Rabeprazole	718 (789)	494	2.37	1.54	3.66
	Combination	249 (55)	242	0.94	0.85	1.05

B
Day	Treatments compared	Ratio of geometric treatment means adjusted for baseline	95% confidence intervals		P values
1	Rabeprazole vs. netazepide	1.29	1.11	1.50	0.0022
	Rabeprazole vs. combination	1.26	1.08	1.47	0.0060
	Netazepide vs. combination	0.98	0.84	1.13	0.7183
40	Rabeprazole vs. netazepide	2.72	1.79	4.14	0.0001
	Rabeprazole vs. combination	2.52	1.66	3.81	0.0002
	Netazepide vs. combination	0.92	0.79	1.08	0.2961

Figure 4.

Plasma CgA: (A) mean (SD) concentration vs. time on day 40 of treatment. (B) geometric mean AUC(0,24 h) at baseline and on days 1 and 40 of treatment. *significant ratio to baseline. , netazepide; , rabeprazole; , combination

Baseline: AUC(0,24 h) of plasma CgA (U l⁻¹h) was similar across groups.

Day 1: Compared with baseline, netazepide and the combination reduced AUC(0,24 h) of CgA, whereas rabeprazole tended to increase it, albeit not significantly. Also on day 1, compared with rabeprazole, netazepide (P = 0.0022) and the combination (P = 0.0060) reduced AUC(0,24 h) of CgA. There was no difference between netazepide and the combination (P = 0.7183).

Day 40: Netazepide significantly reduced AUC(0,24 h) of CgA, whereas rabeprazole significantly increased it. Also, compared with either netazepide (P = 0.0001) or the combination (P = 0.0002), rabeprazole increased AUC(0,24 h) of CgA significantly. Again, there was no difference between netazepide and the combination (P = 0.2961).

Gastric biopsies (Tables S4 and S5)

Despite the increase in AUC(0,24 h) of CgA on day 40 by rabeprazole, overall there were no significant differences between baseline and day 42 biopsies for any treatment. There was a significant reduction in VMAT2 in the netazepide group. Biopsy results varied a lot within and among subjects.

Dyspepsia

The numbers of subjects per group with symptoms during withdrawal weeks 1, 2 and 3, respectively, were netazepide 2, 3 and 2, rabeprazole 3, 1 and 2 and combination 2, 2 and 2. Out of a possible total score of 1950 per group of 10 subjects, scores per group for weeks 1, 2 and 3, respectively, were netazepide 58, 30 and 8, rabeprazole 45, 19 and 44 and combination 52, 9 and 11. Symptoms occurred on 1–3 days, were mainly mild to moderate in severity and bothered subjects only a little or moderately.

In the netazepide group, one subject took one dose of antacid/alginate on day 43, and another subject took one dose on day 50. In the rabeprazole group, one subject took one dose on days 45 and 49. No subject in the combination group used antacid/alginate.

Discussion

The groups were matched at baseline in all respects, apart from basal H⁺ secretion rate in the pentagastrin test. All treatments were well tolerated and safe. Treatment compliance was remarkably high, probably because subjects were prompted by the daily wristwatch alarm and weekly calls, and because of good tolerability of the treatments.

Pharmacodynamics

Rabeprazole alone

As expected, rabeprazole increased AUC(0,24 h) of ambulatory intragastric pH, inhibited the effects of pentagastrin on gastric acid secretion and increased circulating concentrations of gastrin and CgA. The effects on pH, gastrin and CgA were greater after repeated doses. Overall, those results are consistent with suppression of gastric acid production, which leads to hypergastrinaemia and in turn stimulation of ECL cells and release into the circulation of CgA, a biomarker of ECL cell activity and mass 30–32.

Gastric biopsies showed no evidence of rabeprazole-induced ECL cell hyperplasia, whereas the same methods in rats given omeprazole and netazepide, alone and in combination, showed unequivocal ECL cell hyperplasia after omeprazole, but not after netazepide or the combination 18. Possible reasons why we found no evidence of ECL cell hyperplasia are (1) the short duration of and lower exposure to hypergastrinaemia, (2) marked variability of biopsy findings within and among subjects, (3) small numbers of subjects and (4) rigour of blinded assessments. In a study of patients on PPI treatment for up to 15 years, only a few had biopsy evidence of ECL cell hyperplasia 33. Transformation of ECL cells by hypergastrinaemia through a sequence of hyperplasia, dysplasia and metaplasia to gastric carcinoids in rats given high dose PPI seems to need prolonged and high exposure to gastrin 34. Likewise, the ECL cells of CAG and ZES patients take years of exposure to hypergastrinaemia to develop into gastric carcinoids 11,35,36.

We failed to show that rabeprazole withdrawal leads to significant rebound hyperacidity. However, even though there were no statistically significant differences among treatments, the fact that basal and pentagastrin-stimulated H⁺ secretion rates were higher than baseline after withdrawal of rabeprazole alone is at least consistent with the concept of PPI-induced rebound hyperacidity. Studies in animal models have shown that withdrawal of large doses of PPI, but not netazepide or a combination, results in rebound hyperacidity 18. However, a review of studies in healthy subjects and patients concluded that the evidence for rebound hyperacidity after PPI withdrawal is weak 37.

Despite our selecting subjects with no history of dyspepsia, several subjects in each group reported symptoms during the withdrawal period. However, we failed to show any significant differences among treatments with respect to dyspepsia after withdrawal, whereas others have reported an incidence of 40% after PPI withdrawal in healthy subjects 9,10.

Netazepide alone

The results in the netazepide group were similar to the rabeprazole group, with two exceptions. First, netazepide reduced rather than increased the AUC(0,24 h) of CgA. Non-clinical studies have shown that large, repeated doses of gastrin/CCK₂ receptor antagonists such as netazepide cause hypotrophy of ECL cells and a reduction in the weight and thickness of the oxyntic mucosa reminiscent of changes seen after fasting 18, so our finding of a reduction in AUC(0,24 h) of CgA is most likely real. Second, as reported previously 26,27, compared with a single dose, the effect of netazepide on AUC(0,24 h) of pH in healthy subjects tended to wane with repeated doses, whereas repeated doses of rabeprazole caused a greater increase in AUC(0,24 h) of pH than did a single dose. We have discussed elsewhere the possible reasons for the apparent attenuation of the effect of netazepide on ambulatory intragastric pH 26. However, whatever the reason, the results of the pentagastrin stimulation tests clearly show that the efficacy of netazepide as a gastrin/CCK₂ receptor antagonist is fully maintained during repeated dosing.

That netazepide did not increase CgA shows that netazepide-induced hypergastrinaemia, which is secondary to acid suppression, does not stimulate ECL cells, because netazepide blocks the gastrin receptors on ECL cells. The reduction in VMAT2 in gastric biopsy after netazepide is probably a false positive. Netazepide-induced hypergastrinaemia, and its lack of effect on ECL cells, is in accord with studies in animal models 18–24.

Combination of netazepide and rabeprazole

All three treatments inhibited pentagastrin-stimulated gastric acid secretion, but there were no differences among treatments. The combination reduced basal H⁺ secretion rate after repeated doses. This reduction was greater than with netazepide alone. Also, the combination increased AUC(0,24 h) of gastrin at steady-state more than did netazepide alone. Some studies have shown that serum gastrin concentrations tend to be inversely related to the degree of acid suppression by a PPI 38. So, the results of the AUC(0,24 h) of gastrin, together with the reduced basal H⁺ secretion rate, lends support to the concept that the combination suppressed gastric acid secretion more than either treatment alone, but that would require confirmation in further trials. A combination of netazepide and loxtidine, an insurmountable histamine H₂-receptor antagonist, was synergistic and caused nearly complete suppression of acid in an animal model 39.

The addition of netazepide to rabeprazole abolished the increase in AUC(0,24 h) of CgA induced by rabeprazole alone, consistent with netazepide blocking the trophic effect of rabeprazole-induced hypergastrinaemia on ECL cells. Also, after repeated doses, the combination was much more effective at reducing the basal H⁺ secretion increasing AUC(0,24 h) of pH than netazepide and as effective as rabeprazole. In previous studies 26,27, and to a lesser extent in this one, repeated doses of netazepide alone led to partial tolerance to its effect on gastric pH. However, there was no such tolerance when netazepide and rabeprazole were combined in this study.

Pharmacokinetics

Plasma concentrations of netazepide in both the netazepide and combination groups were much lower and more variable among subjects than in our previous studies 25,26. Mean C_max and AUC(0,24 h) after one dose of netazepide 100 mg were 39 ng ml⁻¹ and 86 ng ml⁻¹h, respectively, whereas mean C_max and AUC(0,24 h) after one dose of netazepide 100 mg in our previous study were 397 ng ml⁻¹ and 658 ng ml⁻¹h, respectively 25. Thus, in comparison, mean C_max and AUC(0,24 h) in the current study were 10-fold and 8-fold lower, respectively. That was because the formulation in this study contained crystalline netazepide, which proved to have low bioavailability, whereas the previous formulation contained amorphous netazepide, which was much more bioavailable. The present study shows that netazepide is very effective even at low plasma concentrations, but the results cannot be compared directly with those obtained with 100 mg doses of a superior formulation 25.

Rabeprazole pharmacokinetics were consistent with published data 40, apart from t_1/2, which was about 3 h in our study compared with 1.5 h, possibly an effect of over-encapsulation. Unlike other PPIs, rabeprazole is metabolized mainly via non-enzymatic pathways, with minor CYP2C19 and CYP3A4 involvement and its activity is little affected by CYP2C19 genotype 41, whereas netazepide is metabolized mainly via CYP3A4 and is a weak CYP3A4 inhibitor 42. Neither netazepide nor rabeprazole significantly affected the pharmacokinetics of the other. However, a formal study 43 would be required to confirm that finding.

Study design limitations

The omission of a placebo, short duration of the treatments and small groups of subjects are limitations of the study design. We decided against a placebo because the study was exploratory, demanding of subjects, the methods were mostly objective and baseline measurements could serve as controls where necessary, albeit less satisfactory than placebo values. Likewise, we decided against measuring circulating gastrin and CgA after treatment withdrawal, because the study was already demanding enough. Completion of the dyspepsia questionnaire during a run-in period before the start of treatment might have helped to assess the relevance of symptoms reported by some subjects after treatment withdrawal. A placebo treatment would have been even better.

In conclusion, netazepide suppressed gastric acid secretion as effectively as did rabeprazole. The reduction in basal acid secretion and the greater increase in serum gastrin after the combination are consistent with more effective acid suppression. The fact that basal and pentagastrin-stimulated H⁺ secretion rates were higher than baseline after rabeprazole withdrawal, albeit not significantly so, is also consistent with the concept of PPI-induced rebound hyperacidity. Further studies are required to clarify whether these assumptions are correct.

Hypergastrinaemia resulting from acid suppression by rabeprazole caused ECL cell hyperactivity, whereas hypergastrinaemia resulting from acid suppression by netazepide was associated with ECL cell hypoactivity, probably because netazepide blocked the gastrin/CCK₂ receptors on ECL cells. For the same reason, netazepide prevented ECL cell hyperactivity resulting from rabeprazole-induced hypergastrinaemia.

Thus, netazepide is a potential treatment for the trophic effects of gastrin and, with or without a PPI, netazepide is a potential treatment for acid-related conditions.

Competing Interests

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare netazepide (YF476) came from research by Ferring, Chilworth, England. In 2006, Ferring licensed netazepide to Trio Medicines Ltd, a subsidiary of Hammersmith Medicines Research (HMR), a contract research organization. MB, SD, GT and FvdB are employees of HMR and MB owns both companies. C-MZ and DC have no conflicts of interest. The study was funded by Trio Medicines Ltd. Sadly, Sir James Black, who was Scientific Adviser to Trio Medicines and helped design the study and interpret the results, died in March 2010.

We thank John Rennie, Paul Swain and Simon Leedham for the gastroscopies and biopsies, and Steven Whaley and Toni Mitchell for the statistical analyses.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's web-site: