Clinical Enteric Neuroscience Translational and Epidemiological Research (C.E.N.T.E.R.) Group
Pharmacodynamic effects of a novel prokinetic 5-HT4 receptor agonist, ATI-7505, in humans
Article first published online: 13 DEC 2006
Neurogastroenterology & Motility
Volume 19, Issue 1, pages 30–38, January 2007
How to Cite
Camilleri, M., Vazquez-roque, M. I., Burton, D., Ford, T., Mckinzie, S., Zinsmeister, A. R. and Druzgala, P. (2007), Pharmacodynamic effects of a novel prokinetic 5-HT4 receptor agonist, ATI-7505, in humans. Neurogastroenterology & Motility, 19: 30–38. doi: 10.1111/j.1365-2982.2006.00865.x
- Issue published online: 13 DEC 2006
- Article first published online: 13 DEC 2006
- Received: 16 August 2006 Accepted for publication: 9 October 2006
- 5-HT4 receptor;
Abstract ATI-7505, an investigational 5-HT4 receptor agonist, was designed to have similar activity as cisapride without the cardiac adverse effects, i.e. without QT prolongation. In addition, ATI-7505 is not metabolized by CYP450. The aim of the study was to assess the effect of ATI-7505 on gastrointestinal (GI) and colonic transit in healthy humans. A randomized, parallel-group, double-blind, placebo-controlled study evaluated effects of 9-day treatment with ATI-7505 (3, 10 or 20 mg t.i.d.) on scintigraphic GI and colonic transit in healthy volunteers (12 per group). Primary endpoints were gastric-emptying (GE) T1/2, colonic geometric centre (GC) at 24 h and ascending colon (AC) emptying T1/2. Daily stool diaries were kept. Analysis of covariance assessed overall treatment group differences, followed by post hoc unadjusted pairwise comparisons. There were borderline overall treatment effects (decrease) on GE T1/2 (P = 0.154); the 20 mg t.i.d. of ATI-7505-accelerated GE vs placebo (P = 0.038). ATI-7505 increased colonic transit (GC24, P = 0.031) with fastest transit at 10 mg t.i.d. vs placebo (P = 0.065). ATI-7505 accelerated AC emptying T1/2 (overall P = 0.075) with 10 mg dose vs placebo (P = 0.042). There was looser stool (Bristol stool form scale, overall P = 0.056) with the 10 and 20 mg t.i.d. doses. No safety issues were identified. ATI-7505 accelerates overall colonic transit and tends to accelerate GE and AC emptying and loosen stool consistency.
ATI-7505 is a serotonin type-4 (5-HT4) receptor agonist belonging to the benzamide series of similar compounds, which is being developed as a gastro-prokinetic agent for upper gastrointestinal (GI) diseases, such as GORD and gastroparesis. The design of ATI-7505 was based on the prototypical agent, cisapride. Cisapride was withdrawn from the US market in July 2000 following reports of drug-related pro-arrhythmic events. These events occurred notably in patients taking other medications that are known to inhibit the CYP450 3A4 isozyme, e.g. erythromycin, fluconazole and amiodarone. The withdrawal of cisapride created a medical need in people suffering from severe gastroparesis, such as diabetic patients.
Other prokinetics available for use in clinical practice, such as erythromycin and domperidone, also have potential for similar adverse effects including time-dependent CYP3A4 inhibition with erythromycin that has been implicated in many drug–drug interactions (e.g. with cisapride and with statins)1–3 or the potent hERG channel antagonism4 by domperidone. Thus, when administered intravenously, domperidone has been associated with cardiac arrhythmias, cardiac arrest and sudden death.5–8
Differences in benzamide-induced 5HT4-receptor-mediated responses between different tissues or species mainly originate from either partial or full agonistic behaviour of these compounds in different test systems. This is a function of the receptor density in a particular tissue and of the nature of the receptor subtype or splice variant.9,10 In the GI tract, 5-HT4 receptors are abundant presynaptic receptors that are located on cholinergic neurons projecting to both muscle layers in the colon and to the circular muscle layer in the stomach.11–13 On the contrary, in the atrium, the 5-HT4 receptors represent a small population of saturable binding sites (Bmax approximately 4 fmol mg−1 protein) with a pK(D) of 9.7.14 A porcine model to quantify these differences has been described recently.15, 16
ATI-7505 was designed to retain potent, specific 5-HT4 full agonist activity in the GI tract, but a partial agonist activity in the heart. In order to address safety concerns, ATI-7505 was required to be devoid of other 5-HT receptor activities and to have negligible inhibitory activity at the hERG channel, with an affinity ratio between IKr and 5-HT4 receptors of at least 1000-fold. In addition, the compound was to be, as much as possible, immune to drug–drug interactions.
Early phase-I clinical results and further preclinical tests will be published elsewhere. Briefly, routine measurements in well-validated test systems in cellular functional assays, measuring cAMP levels in HEK293 cells transfected with the human 5-HT4b splice variant, showed that ATI-7505 is a potent agonist at the human 5-HT4b receptor. As expected from the molecular design, inhibition of potassium current in patch-clamped HEK cells transfected with the human IKr channel is very weak, suggesting that there would be an adequate safety window between activity on the GI tract and potential cardiac toxicity. In addition, the primary metabolite is 100-fold less active than the parent drug at the 5-HT4 receptor and has no detectable hERG channel inhibitory activity at concentrations up to 100 μmol L−1. Finally, preliminary results of intensive cardiac safety monitoring, both in preclinical phase and in phase-I, suggest that ATI-7505 has a very safe cardiac profile.17
The objectives of this study were to (i) compare the effects of three different doses of ATI-7505 and placebo on gastric, small bowel and colonic transit in healthy human volunteers and (ii) evaluate the safety and tolerability of three different oral doses of ATI-7505, compared with placebo when given to healthy human volunteers.
Materials and methods
Fig. 1 shows the chemical structures of ATI-7505 and of the prototype compound cisapride. The molecule, like cisapride, belongs to the benzamide series of 5-HT4 receptor agonists. Unlike cisapride, which is a mixture of (3R, 4S) and (3S, 4R) isomers of substituted piperidine-based scaffolds, ATI-7505 is the pure (3S, 4R) isomer. ATI-7505, with its (R)-quinuclidinyl moiety, is metabolized by ubiquitous carboxylesterases to a single metabolite ATI-7500. Hydrolytic esterase metabolism, unlike oxidative CYP450 metabolism, is a large-capacity metabolic system that can easily handle therapeutic amounts of xenobiotics. Drugs like ATI-7505 that are metabolized by carboxylesterases therefore are not expected to undergo drug–drug interactions with other drugs that are metabolized by esterases. Obviously, this is even more so if the competing drugs are metabolized by a different enzymatic system, like CYP450.
Pharmacodynamic study design and dosing
This trial was a double-blind, randomized, placebo-controlled study evaluating the GI transit effects of three different doses of ATI-7505 and placebo for up to 9 days. Three different doses of ATI-7505 and placebo were administered in parallel to four independent groups of 12 subjects each, stratified on gender. The maximum duration of drug exposure for any subject was 9 days. Allocation was concealed, and all study personnel including the study statistician were blinded to the randomization code and assignments until all the measurements were communicated in blinded manner and stored securely.
The study consisted of a screening visit and a 9-day treatment phase. Scintigraphic gastric emptying (GE) and small bowel and colonic transit studies were performed after receiving ATI-7505 or placebo for 7 days. Subjects then received an additional 2 days of double-blind study medication during the scintigraphic transit study. A daily bowel habit diary (using validated instruments including the Bristol stool form scale)18 was kept during the double-blind treatment period. All visits were held in the General Clinical Research Center (GCRC) at Mayo Clinic, Rochester, MN, USA. The study was approved by the Mayo Clinic Institutional Review Board.
After informed consent was obtained, subjects underwent a screening evaluation to document physical and baseline medical status. Eligible subjects were randomized to one of four dose groups as follows: 3, 10, 20 mg, or placebo p.o. t.i.d. for 9 days. The placebo tablets were of the same size and appearance as the ATI-7505 tablets. The placebo tablets were formulated with identical excipients.
Subjects were instructed to take one tablet of study drug with liquid in the morning (between 6 a.m. and 8 a.m.) 5 min before breakfast, one tablet with liquid at midday (between 12p.m and 2p.m.) 5 min before lunch and one tablet with liquid in the evening (between 6p.m. and 8p.m.) 5 min before dinner each day. Subjects who did not eat a full meal at any of these times took the tablets during the specified period with 8 oz of liquid.
Safety was assessed throughout the study by monitoring adverse events and obtaining laboratory tests and vital signs.
Participants and inclusion criteria
A sufficient number of subjects were enrolled to complete 48 studies in healthy male and female subjects between the ages of 18 and 65 years (inclusive). The racial/ethnic characteristics of the group reflected the communities in Olmsted County, Minnesota, which are 95% white, 3% Asian and 2% other minorities. There are no known ethnic or gender differences in bioavailability or effects of ATI-7505.
Male and female volunteers of any ethnic group were eligible to participate in this clinical investigation, provided they met the following criteria: 18–65 years (inclusive) of age; a BMI ≥18 and ≤32 and PR, QRS and QT intervals within normal limits on screening electrocardiography (ECG). Specifically, they had to have PR <220 ms, QRS <140 ms, QTc <450 ms; be able to understand study requirements; follow instructions; attend all required study visits and sign IRB-approved, written informed consent. They also had to have negative urine screen and clinical history (prior 2 years) for drugs of abuse at the screening visit.
Study procedures and measurement of gastrointestinal and colonic transit
All female subjects of childbearing potential had a serum pregnancy test at screening and within the 48-h period prior to the scintigraphic study. This method has been validated and performance characterized and used extensively in the literature in drug development studies of novel medications (Fig. 2).19–23
On the first scintigraphy day (study day 8), all participants ingested a methacrylate-coated, delayed-release capsule containing 111In adsorbed on activated charcoal particles at approximately 6a.m. Subjects received assigned study medication after the capsule containing 111In was shown by gamma camera to be level or below the anterior superior iliac crest, or after 2 h, whichever occurred first. 99mTcsulphur colloid-labelled scrambled eggs were ingested with one slice of whole wheat bread and one glass of skim milk (300 kcal) to facilitate measurement of gastric and small bowel transit. The meal was ingested 5 min after the study medication.
All participants received standardized meals (chicken meal and roast beef sandwich snack) 4 and 8 h after the ingestion of the radiolabelled breakfast meal. Scintigraphic camera scans with abdominal images were performed prior to radiolabelled breakfast meal and at 0, 15, 30, 45, 60, 75, 90 and 105 min, every half hour between 2 and 6 h inclusive, and 8 h following the meal.
Subjects were discharged from the GCRC following completion of the 8-h scan and consumption of the standardized roast beef sandwich snack. Upon discharge from the GCRC, subjects were encouraged to return to their usual diet (with instructions regarding overnight fast for the morning scans on the next 2 days).
Subjects returned to the GCRC on study day 9 after an overnight fast for a 5-min camera scan the next morning [24 h (±1 h) after ingestion of the radiolabelled breakfast] and for an additional 5-min camera scan in the afternoon [approximately 32 h (±1 h) after ingestion of the radiolabelled breakfast meal]. Subjects were discharged from the GCRC, following completion of the 32-h (±1 h) scan and were again reminded to continue their usual diet. Subjects were instructed that their last dose of study medication was to be taken as the evening dose on day 9.
A final camera scan was taken (approximately 48 ± 1 h after ingestion of the radiolabelled breakfast). Upon completion of the camera scan, an end-of-study physical examination and all end-of-study laboratory evaluations were obtained. The bowel habit diary for the double-blind period was collected. Subjects were then discharged from the GCRC.
Safety and tolerability were evaluated by serial assessment of vital signs, physical examination, ECG, adverse events, concomitant medications and blood sample analysis.
A variable region of interest programme was used to measure transit, as in previous studies. The primary efficacy outcome measures were GE T1/2 (as measured with scintigraphic method following 99mTc-egg meal) and ascending colon (AC) T1/2 and colonic geometric centre (GC) at 24 h (based on the transit of 111In-labelled activated charcoal). The following key parameters were also evaluated: percentage GE at 1, 2 and 4 h, colonic filling at 6 h (a surrogate for small bowel transit profile) and colonic GC at 4, 8 and 48 h.
Sample size determination
The study was powered to detect a clinically meaningful difference in colonic transit based on the colonic GC at 24 h. The sample size of 12 subjects per treatment arm (i.e. 12 subjects in the placebo group, and 12 subjects in each of the three ATI-7505 groups) was based on a simple, two-sample t-test (e.g. placebo vs specific dose level). Assuming a coefficient of variation of 40% (based on historical data using the scintigraphic method) and assuming a type I error of 0.05, 12 subjects per treatment arm would provide 80% power to detect an effect size of 46% (estimated difference as a percentage of the overall mean value of the response). A drug effect corresponding to this size difference is clinically significant and has been previously demonstrated with other agents, such as tegaserod and renzapride.
Based on the parallel-group design, the primary and secondary response endpoints were analysed using an analysis of covariance (ancova) with gender as the covariate. Post hoc pairwise comparisons, unadjusted for multiple pairwise comparisons of each dose vs placebo, were also examined. Consistent with an intention-to-treat analysis paradigm, missing data were imputed for continuous endpoints using the corresponding overall (subjects with non-missing values) mean value for a given specific endpoint. The error degrees of freedom in the respective ancova model was reduced by 1 for each missing value imputed.
After the conclusion of the study, descriptive statistics were summarized for the safety parameters for all subjects enrolled in the study. A safety analysis was performed on treatment-emergent adverse events, changes in physical examination, changes in vital sign measurements and changes in clinical laboratory evaluations.
The study was successfully completed in 47 participants. One participant was withdrawn, after taking three doses of study drug when the study ECG showed a junctional rhythm. The participant was withdrawn and at the request of his treating cardiologist, his study medication assignment was revealed. The participant had been randomized to placebo. The data for this subject were imputed as described above and the values included in the analyses with adjustment of the degrees of freedom (reduction by 1 for each data imputation) as prespecified in the study protocol.
Upper gastrointestinal transit
The overall drug effect on GE T1/2 (Fig. 3) and GE at 4 h (Table 1) was borderline (ancova, P = 0.154, P = 0.176, respectively). Specific pairwise comparisons with placebo suggested that 20 mg t.i.d. ATI-7505 accelerates GE (comparison with placebo, T1/2, P = 0.038, unadjusted; GE at 4 h, P = 0.028, unadjusted). There were no significant treatment differences for colonic filling at 6 h, a surrogate for small bowel transit time.
|Parameter||Placebo||3 mg t.i.d.||10 mg t.i.d.||20 mg t.i.d.|
|Gender (% female)||67||75||67||69|
|Age, year*||32 (±3)||33 (±4)||33 (±3)||35 (±2)|
|BMI, kg m−2*||25.2 (±1.0)||23.9 (±0.8)||24.4 (±1.0)||23.8 (±0.7)|
|GE 60 min (%)†||16 (±2)||16 (±2)||17 (±2)||20 (±2)|
|GE 120 min (%)†||48 (±3)||49 (±3)||50 (±3)||58 (±3)|
|GE 240 min (%)†||90 (±3)||93 (±3)||93 (±3)||98 (±3)|
|GE T1/2(min)†||126 (±6)||124 (±6)||122 (±6)||109 (±6)|
|CF (6 h), (%)†||50 (±8)||53 (±8)||57 (±8)||45 (±8)|
|Colon GC 8 h†||1.5 (±0.2)||1.5 (±0.2)||2.1 (±0.2)||1.9 (±0.2)|
|Colon GC 24 h†||2.7 (±0.2)||2.3 (±0.2)||3.2 (±0.2)||2.9 (±0.2)|
|Colon GC 48 h†||4.1 (±0.2)||3.5 (±0.3)||4.2 (±0.2)||4.2 (±0.2)|
|AC T1/2 (hours)†||11.6 (±1.7)||12.6 (±1.7)||6.7 (±1.7)||11.1 (±1.7)|
|Mean, bm day−1†||1.26 (±0.17)||1.09 (±0.17)||1.50 (±0.17)||1.27 (±0.17)|
|Mean stool consistency†||3.4 (±0.2)||3.2 (± 0.2)||3.9 (±0.2)||3.9 (±0.2)|
|Mean ease of passage score†||4.0 (±0.1)||3.9 (±0.1)||4.0 (±0.1)||3.9 (±0.1)|
There was a significant overall drug effect of ATI-7505 on the primary endpoint of colonic transit, colonic GC at 24 h (P = 0.031, ancova, Fig. 4). The comparison of 10 mg t.i.d. vs placebo was borderline significant, P = 0.065, unadjusted.No significant differences for the colonic GCs at other time points were detected (GC 8, P = 0.196; GC 48, P = 0.178).
The overall drug effect on emptying of the AC (T1/2) suggested borderline treatment differences, P = 0.075 (Fig. 5). The pairwise comparison between 10 mg t.i.d. and placebo was significant, unadjusted P = 0.042.
While there were no significant effects on stool frequency or ease of passage (Table 1), there was a trend for looser bowel movement consistency with ATI-7505 (overall treatment differences, P = 0.056), with the 10 and 20 mg t.i.d. doses being associated with numerical scores suggesting looser stool consistency.
Drug safety was assessed by adverse event monitoring, routine laboratory studies (chemistry, haematology and urinalysis), vital sign measurements and ECG. ATI-7505 had a favourable safety profile and was well tolerated. No serious adverse events occurred and almost all of the adverse events reported were of mild or moderate intensity. As expected from the pharmacology of the drug, symptoms related to the GI system were the most commonly reported adverse events. Diarrhoea, flatulence and abdominal pain were reported most frequently, with the incidence generally increasing with dose. Table 2 summarizes those adverse events reported by two or more treated participants. Mild headache was also reported, but less commonly than the GI events. Clinical laboratory studies and vital signs did not show any consistent or dose-related changes from baseline values. Electrocardiograms were measured twice before treatment and then on four occasions during dosing. There were no changes in the heart rate, the PR, the QRS or in the QT intervals, which were corrected by both the Bazett and the Fridericia methods. There was no evidence of QT prolongation as measured by central tendency or by outliers with QTc >500 ms or change from baseline >60 ms.
|Adverse event preferred term*||Placebo n = 12||3 mg t.i.d. n = 12||10 mg t.i.d. n = 13||20 mg t.i.d. n = 12|
|Any adverse event (%)||67||75||83||83|
|Sinus headache (%)||0||0||17||0|
|Abdominal pain (%)||0||0||17||8|
|Abdominal pain lower (%)||0||0||8||8|
|Abdominal pain upper (%)||0||0||17||0|
|Loose stools (%)||8||0||17||0|
|Stomach discomfort (%)||0||0||0||17|
This study has demonstrated that ATI-7505 has GI prokinetic properties in humans. A novel design approach was used to develop an analog of the well-known prokinetic, cisapride. This approach eliminated CYP3A4 metabolism and potential for drug–drug interactions. Very weak affinity for the hERG K+ channel also minimized greatly the risk of cardiac arrhythmias. This article summarizes a first pharmacodynamic study that demonstrates the potential therapeutic targets and efficacy in humans. The demonstrated effects on transit suggest that this compound may be effective in upper and lower gut motility disorders and, potentially, in conditions that are associated with regional dysmotility at both sites.
The pharmacodynamic results suggest that the 20-mg t.i.d. dose is effective in accelerating GE in healthy humans. This is an important observation, given the prior reports on the equivocal effects of 5-HT4 receptor agonists, such as cisapride and metoclopramide, on GE in healthy humans24 and the clear improvement of gastric or small bowel transit observed with cisapride in gastroparesis and pseudo-obstruction.25–28 Thus, the reduction of mean GE T1/2 in healthy volunteers observed with 20-mg t.i.d. ATI-7505 relative to placebo suggests that further studies are indicated in diseases such as gastroparesis and functional dyspepsia, both of which are associated with delayed GE, in order to determine the potential clinical benefit of ATI-7505.
The effects of ATI-7505 on colonic transit in healthy participants also suggest that there is a potential for the drug to be beneficial in the treatment of disorders associated with slow colonic transit. It has been previously shown that a change in GC at 24 h of 1 unit is associated with a dramatic difference in bowel function from normal to diarrhoea29 or from constipation to normal.30 The magnitude of change in overall colonic transit and AC emptying in healthy volunteers is consistent with the trend to change in stool consistency demonstrated in parallel by means of the stool diaries. It is intriguing that the change in AC-emptying T1/2 with 10 mg t.i.d. of ATI-7505 is consistent with the magnitude of change observed with other 5-HT4 receptor agonists such as tegaserod,31 prucalopride32 or renzapride33 using the same method in the same laboratory over the last decade.
Moreover, as ATI-7505 is not metabolized by CYP450, the risk of drug levels escalating as a result of competitive inhibition of the enzyme by other agents should not be an issue. This property would make it easier for the concomitant use of ATI-7505 as a prokinetic with compounds such as the motilin agonist erythromycin, or with antiemetics. Such combination treatments of erythromycin or antiemetics with cisapride were contraindicated because of the competition for CYP3A4 metabolism.
There has been a significant clinical need for a safe and effective prokinetic for the management of patients with gastroparesis and chronic intestinal pseudo-obstruction since the withdrawal of cisapride. Given the concerns associated with stimulation of 5-HT4 receptors in the heart, much attention has been directed towards motilin agonists. The commonly used macrolide antibiotic, erythromycin, is effective when used intravenously, although efficacy with long-term oral administration is unclear at best. A recent meta-analysis suggests that it is ineffective in the longer term.34 Other non-antibiotic motilides have been ineffective in clinical trials.35,36 While atilmotin was effective in pharmacodynamic studies, it has a short duration of action, necessitating parenteral administration.37 After initial enthusiasm over the efficacy of the motilin agonist mitemcinal, subsequent phase-II trials in diabetic gastroparesis raised questions regarding optimal dosing.38,39
Ghrelin agonists are being developed as gastric prokinetics, based on observations with synthetic human ghrelin.40–42 However, our studies, using doses of exogenous ghrelin that stimulate growth hormone release in the physiological range and are 10–100 times lower than the doses used in the other articles in the literature, suggest that ghrelin does not have a significant effect on GE.43
Hence, there is the need to develop novel medications with a similar or improved spectrum of efficacy relative to cisapride and improved benefit-to-risk ratio. In summary, these first studies prove the concept that the novel chemical synthetic approach is able to produce a medication with potentially greater safety while maintaining efficacy potential. ATI-7505 appears to have prokinetic properties in both stomach and colon in healthy volunteers, and further studies in disease states appear warranted.
The study was supported in part by General Clinical Research grant RR00585 from National Institutes of Health and was funded by ARYx Therapeutics, Fremont, California. The excellent secretarial support of Mrs Cindy Stanislav is gratefully acknowledged.
- 17ATI-7505 is a novel, selective 5HT(4) receptor agonist that causes gastrointestinal prokinetic activity in dogs. Gastroenterology 2004; 126: A641., , , , .
- 36Failure of a motilin receptor agonist (ABT-229) to relieve the symptoms of functional dyspepsia in patients with and without delayed gastric emptying: a randomized double-blind placebo-controlled trial. Aliment Pharmacol Ther 2000; 14: 1653–61., , et al.
- 38Effect of mitemcinal fumarate (GM-611) on gastric emptying in patients with idiopathic or diabetic gastroparesis. Gastroenterology 2004; 126: A483., , , , .
- 39Mitemcinal fumarate (GM-611) provided symptomatic relief of diabetic gastroparesis, especially in type I diabetes: results of a 12-week, multi-center, double-blind, placebo-controlled, randomized phase 2b study (gm-611-05). Gastroenterology 2005; 128: 0A467., , , , , .
- 42Effect of ghrelin on gastric emptying in patients with neurogenic gastroparesis. Gastroenterology 2005; 128: A59., , et al.
- 43Effect of synthetic human ghrelin on gastric motor function and postprandial symptoms in normal weight and obesity: randomized, double-blind, placebo-controlled study. Neurogastroenterol Motil 2005; 17: 631–2., , et al.