Background Velusetrag is an orally active 5-HT 4 receptor agonist of potential benefit in treating chronic idiopathic constipation.

Transcription

1 Alimentary Pharmacology and Therapeutics Clinical trial: the efficacy and tolerability of velusetrag, a selective 5-HT 4 agonist with high intrinsic activity, in chronic idiopathic constipation a 4-week, randomized, double-blind, placebo-controlled, dose response study M. Goldberg, Y.-P. Li, J. F. Johanson, A. W. Mangel à, M. Kitt, D. T. Beattie, K. Kersey & O. Daniels Theravance, Inc., South San Francisco, CA, USA. University of Illinois, College of Medicine, Rockford, IL, USA. à RTI Health Solutions, Research Triangle Park, NC, USA. Correspondence to: Dr O. Daniels, Theravance, Inc., 901 Gateway Blvd, South San Francisco, CA 94080, USA. odaniels@theravance.com Publication data Submitted 26 April 2010 First decision 14 May 2010 Resubmitted 12 August 2010 Accepted 23 August 2010 EV Pub Online 22 September 2010 SUMMARY Background is an orally active 5-HT 4 receptor agonist of potential benefit in treating chronic idiopathic constipation. Aim To evaluate the efficacy, safety and tolerability of velusetrag in chronic idiopathic constipation. Methods After a 2-week baseline period, patients [<3 spontaneous bowel movements (SBM) week] received placebo or velusetrag (15, 30 or 50 mg) daily for 4 weeks in a randomized, double-blind design, followed by a 1-week follow-up period. The primary endpoint was the change from baseline in weekly SBM frequency averaged over the 4-week treatment period. Results Patients receiving velusetrag (15, 30 and 50 mg) achieved statistically and clinically significant increases in weekly SBM frequency relative to those receiving placebo. Mean increases were 3.6, 3.3 and 3.5 SBM week respectively, compared with 1.4 SBM week for placebo (P < ). Statistically significant increases in the weekly frequency of complete SBM (CSBM) were also reported (mean increases of 2.3, 1.8 and 2.3 for 15, 30 and 50 mg velusetrag respectively, compared with 0.6 for placebo). Common adverse events associated with velusetrag were diarrhoea, headache, nausea and vomiting, generally occurring during the initial days of dosing. Conclusion was efficacious and well tolerated in patients with chronic idiopathic constipation (ClinicalTrials.gov identifier NCT ). Aliment Pharmacol Ther 2010; 32: doi: /j x

2 Clinical trial: efficacy of velusetrag in chronic idiopathic constipation INTRODUCTION Chronic constipation affects up to 27% of the population of the United States and significantly impacts quality of life. 1 Current therapies often include dietary fibre and osmotic or stimulant laxatives. However, with these remedies, relief of symptoms is often inadequate, particularly upon chronic use in severely constipated patients; in addition, bloating and abdominal discomfort are sometimes observed. 2 7 Lubiprostone (Amitiza; Sucampo Pharma Americas, Inc., Bethesda, MD, USA, and Takeda Pharmaceuticals America, Inc., Deerfield, IL, USA), a chloride channel activator, has been approved in the United States for the treatment of chronic idiopathic constipation and irritable bowel syndrome with constipation (IBS-C). The incidence of nausea in patients receiving the approved dose of lubiprostone for chronic idiopathic constipation was approximately 29% in clinical trials, and resulted in 9% of patients discontinuing in these studies. 8, 9 Thus, an unmet need still exists for a highly effective and well-tolerated treatment that addresses the reduced motility underlying this troublesome chronic disorder. The serotonin 5-HT 4 receptor plays an important role, both physiologically and pathophysiologically, in the regulation of GI motility. 10, 11 Based on the robust prokinetic activity of 5-HT 4 receptor agonists in the upper and lower GI tract of healthy human subjects, this class of agents has been investigated for the treatment of patients with disorders of reduced GI motility. Tegaserod (Zelnorm; Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA) was approved in the USA and other countries for treatment of chronic idiopathic constipation and IBS-C. However, clinical 15, 16 trials demonstrated only modest efficacy; mean weekly increases of approximately 1 spontaneous bowel movement (SBM) and 0.6 complete SBM (CSBM) were reported. 17 The limited clinical efficacy of tegaserod has been attributed to its lack of 5-HT 4 receptor selectivity Tegaserod has significant affinity for 5-HT 1B, 5-HT 1D, 5-HT 2A, 5-HT 2B and 5-HT 7 receptor subtypes. Antagonism of 5-HT 2B receptors in the colon by tegaserod may counteract 5-HT 4 receptor-mediated prokinetic activity. An interaction of tegaserod with non-5-ht 4 receptors may also be responsible for its possible association with increased cardiovascular risk, which resulted in suspension of its marketing in March An analysis of pooled clinical data revealed that 0.1% of patients treated with tegaserod had serious cardiovascular ischaemic events (myocardial infarction, unstable angina or stroke), compared with 0.01% in placebotreated patients. The occurrence of adverse events with tegaserod and older generation prokinetics such as cisapride, domperidone and metoclopramide, which also interact with multiple non-5-ht 4 receptors (i.e. dopamine D 2, 5-HT 2A, 5-HT 2B, or 5-HT 3 ) or herg potassium channels, has stimulated efforts to identify and develop highly selective 5-HT 4 receptor agonists One such compound, prucalopride (Resolor; Movetis NV, Turnhout, Belgium), has demonstrated robust efficacy in patients with chronic idiopathic constipation and has recently received approval in some countries in Europe To date, prucalopride has not been linked to cardiovascular risk. (N-[(1R,3R,5S)-8-{(2R)-2-hydroxy-3-[methyl (methylsulfonyl)amino]propyl}-8-azabicyclo[3.2.1]oct-3-yl]- 1-(1-methylethyl)-2-oxo-1,2-dihydroquinoline-3-carboxamide; previously known as TD-5108) is a highly selective 5-HT 4 receptor agonist. demonstrates good potency (pec 50 = 8.3) and high intrinsic activity at human recombinant 5-HT 4 receptors and at native 5-HT 4 receptors in human and rodent GI tissue. 28 has no significant affinity for the herg potassium channel in contrast to cisapride (IC 50 values of >3 lm and <20 nm respectively, in Chinese Hamster ovary cells). 28 Furthermore, unlike tegaserod, velusetrag has no appreciable affinity for the 5-HT 1B, 5-HT 1D, 5-HT 2A, 5-HT 2B or 5- HT 7 receptor subtypes. 28 produces robust increases in GI motility in rats, guinea pigs and dogs, 29 and has good oral bioavailability in rats and dogs. 30 Single doses of up to 70 mg and repeated daily doses of up to 50 mg for 14 days in healthy human subjects were associated with dose-related prokinetic activity, including increases in stool production. 31 has also been shown to increase gastric emptying as well as intestinal and colonic transit in healthy volunteers, suggesting potential utility in both upper and lower GI disorders. 32 was well tolerated in these healthy volunteer 31, 32 studies. Consistent with its mechanism of action, more healthy subjects treated with velusetrag than with 31, 32 placebo experienced diarrhoea in Phase 1 studies. The systemic exposure of velusetrag in healthy subjects increased in a dose-dependent manner, exhibiting an elimination half-life (12 14 h), supportive of once daily dosing. 31 Pharmacokinetics were similar in patients with chronic constipation. 32 This Phase 2 clinical trial was conducted to investigate the efficacy and tolerability of velusetrag in patients with chronic idiopathic constipation. Aliment Pharmacol Ther 2010; 32:

3 M. Goldberg et al. MATERIALS AND METHODS Study design This randomized, double-blind, placebo-controlled, parallel-group, multicentre trial (known as ACCORD: A randomized, placebo-controlled, double-blind study of TD-5108 for the treatment of chronic constipation) in patients with chronic idiopathic constipation was conducted at 49 sites in the United States. Following a 2-week baseline period, eligible patients were randomized in equal proportions to receive one of three doses of velusetrag (15, 30 or 50 mg) or placebo, administered orally once daily for 4 weeks, followed by a 1-week follow-up period (Figure 1). Randomization was stratified by centre using a permuted block algorithm and was made by a telephonic interactive voice response system (IVRS). If the study blind was broken by site staff, the date, time, and reason were to be recorded in the case report forms and the sponsor was to be notified immediately. The study protocol and consent form were reviewed by the Institutional Review Board for each site, and each patient provided written informed consent prior to initiation of study procedures. The study was conducted in accordance with the US code of Federal Regulations and the principles of the Declaration of Helsinki. The study was registered in the publicly accessible registry Clinical Trials.gov (identifier NCT ). Patients Eligible patients were males and females at least 18 years of age diagnosed with functional constipation as defined by the Rome 3 criteria. 33 Patients were excluded if they met criteria for irritable bowel syndrome with constipation at the time of enrolment or if they had structural bowel disease. Other major exclusion criteria included a history of secondary gastrointestinal, neurological, metabolic, or pharmacological causes of constipation, a history of stable angina, Raynaud s syndrome, peripheral vascular disease or ischaemic colitis, any untreated, active infections or medical illnesses, elevated serum transaminases (alanine aminotransferase or aspartate aminotransferase >1.3 times the upper limit of normal for the reference laboratory) or hypercalcaemia (based on the normal range for the reference laboratory) on the screening evaluation, and a history of major GI surgery. For randomization to the 4-week treatment period, the patient had to report <3 SBM week and laxative use <2 days week during the 2-week baseline period. Screening and baseline period (14 days) n = screen failures 401 subjects randomized n = mg n = mg n = mg n = 97 Treatment period (28 days) 99 completed 8 withdrawn 1 adverse event 5 withdrew consent 2 other 94 completed 7 withdrawn 4 adverse event 1 withdrew consent 2 other 86 completed 11 withdrawn 3 adverse event 4 withdrew consent 3 other 78 completed 19 withdrawn 11 adverse event 2 withdrew consent 1 major deviation 5 other Follow-up period (7 days) Figure 1 Overview of the study design Aliment Pharmacol Ther 2010; 32:

4 Clinical trial: efficacy of velusetrag in chronic idiopathic constipation Assessments Patients used the IVRS to record stool diary information daily throughout the baseline, double-blinded treatment and follow-up periods. The diary captured information on the time of the first bowel movement after the start of study medication, ratings of consistency, completeness of evacuation and straining for each bowel movement, the presence and severity of bloating and laxative use in the previous 24 h, and, on a weekly basis, whether patients had adequate relief of constipation over the previous 7 days. As rescue medication, up to 15 mg per day of bisacodyl could be used if no bowel movement had occurred for at least 96 h and the patient experienced discomfort. Any other laxative use was prohibited. Adverse events were collected by treatment-blinded study personnel throughout the treatment and follow-up periods. Measurements of vital signs, clinical laboratory assessments (haematology, biochemistry and urinalysis), and 12-lead electrocardiograms (ECGs) were obtained and physical examinations were performed at clinic visits during the double-blind treatment period. A final assessment was conducted via telephone at the end of the follow-up period. Patients were instructed to take blinded study medication once daily, in the morning at approximately the same time (1 h) each day, on an empty stomach. All study medication was supplied in blister packages with the same number of capsules (active plus placebo, as required) per dose; all capsule strengths were matching in appearance. All study personnel and patients were blinded to treatment. At the week 2 and week 4 visits, compliance was assessed by a count of the number of capsules returned. Study endpoints The primary efficacy endpoint was the change from baseline in the average number of SBM per week averaged over the 4-week treatment period, i.e. the total number of SBM during the treatment period divided by 4. An SBM was defined as a bowel movement unrelated to administration of a laxative, i.e. bowel movements reported within 24 h after taking bisacodyl were not considered spontaneous. An SBM was further defined as a bowel movement not requiring manual extraction of stool and was characterized by the presence of visible quantities of stool in the toilet bowl, as judged by the patient. The baseline average number of SBM per week was the sum of the SBM over the 2-week baseline period divided by 2. The change from baseline in the average number of SBM per week averaged over the 4-week period was selected as the primary endpoint as no bowel movement data in any patient population were available to provide estimates of SBM and or CSBM increases each week for this first study of velusetrag in patients with chronic constipation. Phase 2 studies of lubiprostone 34 and linaclotide 35 also used change in overall weekly SBM frequency as primary endpoints. Secondary endpoints included changes from baseline in the average number of CSBM per week over the 4- week treatment period, as well as changes from baseline in the number of SBM and CSBM in each of the four treatment weeks. A CSBM was defined as an SBM with a positive response to the IVRS question, Did you completely empty your bowels? The percentages of patients with an increase of at least 1 CSBM week, an increase of at least 1 SBM week, and at least three total SBM week (i.e. responder analyses in terms of BM frequency) in each week, as well as over the treatment period, were also determined. Additional secondary endpoints included average stool form consistency each week [assessed using the modified 7-point Bristol Stool Form scale (1 = pellets, 2 = formed and hard, 3 = formed and soft, 4 = semi-formed, 5 = mushy, 6 = loose, and 7 = watery)] 36 and average straining score each week (0 = no straining, 1 = acceptable straining, and 2 = too much straining). The time to first SBM within 24 h of the first dose of study medication (evaluated by Kaplan Meier curves) and the median time to use of rescue medication each week were determined. Responder analyses related to symptom improvement included the percentage of patients each week that had a mean Bristol stool score of 3 5, who did not report any straining, who experienced any bloating, and who reported no rescue laxative use. In addition, patient satisfaction was assessed by the proportion who reported adequate relief of their constipation each week (answered 0 for no and 1 for yes in response to the question, Have you had adequate relief from constipation over the last 7 days? ). Statistical analysis Power sample size considerations were based on the primary efficacy variable. The table below shows the sample size per group for 80% power and a two-sided type I error rate of 5%. Assuming that average improvement for an active dose group compared with placebo in weekly SBM frequency during weeks 1 and 4 adjusted for baseline differences was 1.5, and the standard deviation on the change from baseline scale was 3.4, approximately 82 patients per group would afford 80% power. Estimates of 1.5 and 3.4 for the treatment effect and Aliment Pharmacol Ther 2010; 32:

5 M. Goldberg et al. standard deviation respectively were interpolated from data available from tegaserod 17 and lubiprostone studies. 9 The sample size of up to 400 patients in total was selected to allow for the possibility that some of the randomized patients might not be included in the efficacy analysis population (i.e. patients were included in the analysis if they received treatment for at least 7 days, made at least five IVRS calls for reporting the diary data and were under observation for at least 7 days). 4, r 2, , , 3.2 N is the improvement in average SBM over 4 weeks relative to placebo, adjusted for baseline differences. r is the standard deviation on the change from baseline scale. Sample size per group for 80% power. Comparisons were made between each velusetrag dose and placebo based on an analysis of covariance (ANCOVA) model with a main effect for treatment group and baseline average number of SBM per week as a covariate. The 95% confidence intervals (CI) of the least square (LS) means from the ANCOVA model were summarized for the treatment group differences. In the primary analysis, missing values were imputed by the last observation carried forward (LOCF) method; if there were three or more days for which no IVRS call was made, that entire week s score was imputed by LOCF method. The primary efficacy endpoint was summarized for each treatment group (placebo and 15, 30 and 50 mg velusetrag). Hochberg s method, 37 a step-up test, was used to adjust alpha levels for the comparison of each dose level of velusetrag against placebo (if the least significant P-value was 0.05, then all velusetrag groups were declared significantly different from placebo; if the largest P-value was >0.05, then the other two P-values had to be to be considered significantly different from placebo; and, if the second largest P-value was >0.025, the third P-value had to be to be considered significantly different from placebo). The primary efficacy analysis was based on the efficacy analysis population (with LOCF imputation). All patients in the efficacy evaluable population (n = 95, 91, 84 and 98 for 15 mg, 30 mg, 50 mg and placebo treatment groups respectively) had week 1 data and up to 8%, 13% and 18% of patients in each treatment group had missing weekly data on weeks 2, 3 and 4 respectively. Sensitivity analyses (not shown) were performed with respect to the intent-to-treat population (consisting of all randomized patients; missing weekly bowel data were assigned a value of zero) and the per-protocol population (all patients in the efficacy population who completed 4 weeks of treatment) as confirmatory support for the primary analysis described herein. In the sensitivity analyses, treatment weeks considered missing were also imputed by the baseline observation carried forward method (BOCF). All analyses were performed using SAS version 8.2 or later on a PC platform. The results of the principal analyses of SBM and CSBM were similar in the intention-to-treat and per protocol populations as in the efficacy population analysis. RESULTS In total, 401 patients were randomized to receive once daily treatment with 15, 30, or 50 mg velusetrag or placebo (Figure 1). More patients in the 15 or 30 mg velusetrag and placebo groups (90 to 93%) completed the study than in the 50 mg velusetrag group (80%), owing principally to a greater number of discontinuations for adverse events in the 50 mg velusetrag group. In total, 5% of velusetrag-treated patients discontinued due to adverse event; adverse events associated with withdrawal were primarily GI-related and or headache. The first patient randomization occurred on 6 October 2006, and the last patient visit occurred on 11 May Patient demographics and baseline characteristics There were no significant differences among the treatment groups in demographic or other baseline characteristics (Table 1). Patients had chronic constipation for a median duration of 18 years prior to the study. Laxative use was reported as daily or weekly by 38% of the study population. Twenty-two percent of patients had previously received tegaserod. Based on the average weekly SBM and CSBM frequency during the 2- week baseline period [i.e SBM week (Table 2) and CSBM week], patients were significantly constipated. Efficacy assessments SBM and CSBM frequency. In the primary efficacy analysis, for each dose of velusetrag, the increase from baseline in the number of SBM per week averaged over the 4-week treatment period was significantly greater than for placebo (P < , Table 2). The mean increase was 3.6 SBM week for the 15 mg dose, 3.3 SBM week for the 30 mg dose and 3.5 SBM week for the 50 mg dose of velusetrag compared with 1.4 SBM week for placebo. These increases resulted in a mean frequency of more than 4 SBM week 1106 Aliment Pharmacol Ther 2010; 32:

6 Clinical trial: efficacy of velusetrag in chronic idiopathic constipation Table 1 Demographic and baseline characteristics of patients who received at least one dose of study medication (n =107) 15 mg (n = 101) 30 mg (n = 96) 50 mg (n =97) Overall (n = 401) P value Age (years) Mean (s.d.) 44.4 (11.7) 44.3 (10.7) 46.1 (10.0) 45.8 (9.3) 45.1 (10.5) Gender Male 9 (8%) 7 (7%) 9 (9%) 7 (7%) 32 (8%) Female 98 (92%) 94 (93%) 87 (91%) 90 (93%) 369 (92%) Race American Indian or Alaska native 1 (1%) 0 (0%) 0 (0%) 0 (0%) 1 (0%) Asian 3 (3%) 1 (1%) 2 (2%) 3 (3%) 9 (2%) Black, of African 33 (31%) 32 (32%) 26 (27%) 23 (24%) 114 (28%) heritage White 70 (65%) 67 (66%) 68 (71%) 71 (73%) 276 (69%) Mixed 0 (0%) 1 (1%) 0 (0%) 0 (0%) 1 (0%) BMI (kg m 2 ) Mean (s.d.) 27.6 (5.4) 28.6 (6.4) 27.7 (6.5) 28.0 (5.6) 27.9 (6.0) Constipation history (months) Median (Min, Max) 240 (8, 720) 168 (6, 672) 234 (9, 720) 240 (7, 684) 216 (6, 720) Prior tegaserod use 25 (23%) 21 (21%) 22 (23%) 20 (21%) 88 (22%) Treatment groups were compared on continuous measures using the Kruskal Wallis test. Treatment groups were compared on categorical parameters using the Fisher s Exact test. in each of the velusetrag dose groups compared with 2.8 SBM week for placebo (Figure 2). In exploratory analyses (not shown), none of the patient baseline characteristics appeared to influence the response with respect to the primary efficacy endpoint. Changes from baseline in SBM frequency in each of the 4 weeks of treatment were significantly greater for each dose of velusetrag than for placebo (P < for 15 and 50 mg velusetrag and P < for 30 mg velusetrag in each week of treatment), with the largest responses occurring in the first week. Similar results were observed for CSBM. The mean increases from baseline in CSBM frequency averaged over the 4-week treatment period were 2.3, 1.8, and 2.3 CSBM week for the 15, 30 and 50 mg velusetrag doses respectively, compared with a mean increase of 0.6 CSBM week for placebo (P < ). The mean CSBM frequencies for the 15, 30 and 50 mg doses of velusetrag were 2.6, 2.0 and 2.6 CSBM week respectively, compared with 0.8 CSBM week for placebo. Increases in CSBM frequency were also statistically significant in each week of treatment (P for each dose and week). Both SBM (Figure 2) and CSBM frequency decreased in the follow-up week after discontinuation of treatment. Time to first bowel movement. The first SBM in the 24 h following the initial dose of study medication occurred more rapidly after velusetrag than after placebo. The median times to the first SBM were approximately 21, 14 and 18 h in the velusetrag 15, 30 and 50 mg groups respectively, compared with 47 h in the placebo group (P < for all comparisons with placebo). Consequently, 57%, 70% and 64% of patients receiving 15, 30 and 50 mg velusetrag respectively had an SBM within 24 h compared with 30% for placebo. Responder analyses. Statistically significant differences between each dose of velusetrag and placebo were demonstrated for multiple responder endpoints related to SBM and CSBM frequency (Table 3). These responder endpoints include the percentages of patients reporting at least 3 SBM week with a weekly average increase of 1 SBM week over the 4-week treatment period (P 0.004), and the percentages reporting at least 3 SBM week in each of the 4 weeks of therapy (P 0.007). Similarly, the percentage of patients reporting an average of at least 3 Aliment Pharmacol Ther 2010; 32:

7 M. Goldberg et al. Table 2 Primary efficacy analysis: change from baseline in the average number of spontaneous bowel movements (SBM) per week over the 4-week treatment period (n = 98) 15 mg (n = 95) 30 mg (n = 91) 50 mg (n = 84) Baseline average number of SBM week Mean (s.d.) 1.3 (0.86) 1.2 (0.82) 1.1 (0.80) 1.2 (0.81) Median Min, Max 0.0, , , , 3.0 Average number of SBM week over the 4-week treatment period Mean (s.d.) 2.8 (2.03) 4.8 (2.88) 4.4 (2.85) 4.6 (3.13) Median Min, Max 0.0, , , , 15.5 Change from baseline Mean (s.d.) 1.4 (1.88) 3.6 (2.66) 3.3 (2.69) 3.5 (3.04) Median Min, Max )2.3, 8.0 )3.0, 12.0 )1.3, 14.8 )2.5, 15.0 P value < < < % CI [1.43, 2.90] [1.15, 2.64] [1.28, 2.79] P values are based on an ANCOVA model with baseline average number of SBM per week as covariate. The 95% CI for LS means difference comparing each velusetrag treated group and placebo is based on an ANCOVA model with baseline average number of SBM per week as covariate mg 4 30 mg 3 50 mg (post-treatment) (Baseline) Time point (weeks) Mean number of SBM/week Figure 2 The effect of velusetrag (15, 30 and 50 mg) and placebo on mean weekly spontaneous bowel movement (SBM) frequency (efficacy population) during each week of treatment (weeks 1 to 4) and during week 5 after treatment was discontinued. CSBM week with an average increase of 1 CSBM week over the 4-week treatment period, and the percentages reporting at least 3 CSBM week for each of the 4 weeks were significantly greater for each dose of velusetrag than placebo (P < and P respectively). There was a 5.4-fold difference in the proportion of patients reporting at least 3 CSBM week for the 15 mg group vs. placebo. Diary data. produced statistically significant activity, relative to placebo, with respect to stool consistency, bloating and straining, the use of rescue laxative and patient assessment of adequate relief. Representative data for week 4 are presented in Figure 3. The need for rescue laxative was reduced over the entire 4-week treatment period in patients receiving velusetrag (Figure 3). The median time to first use of rescue laxative was approximately 12 days (292 h; 95% CI 200, 394) for patients in the placebo group compared with approximately 22 days (631 h; 269, upper limit not calculable) in the 50 mg group. No median time could be calculated for patients in the 15 and 30 mg velusetrag groups because fewer than 50% of those patients required rescue laxative. In each of the 4 weeks of treatment, the proportion of patients reporting adequate relief of their constipation was statistically significantly greater (P ) for the velusetrag 15 and 50 mg groups compared with the placebo group. For the velusetrag dose of 30 mg, the difference from placebo was statistically significant (P ) for each of the first 3 weeks of treatment. Week 4 adequate relief data are presented in Figure 3. Safety and tolerability Adverse events were reported by 47 patients (47%) taking velusetrag 15 mg, 51 (53%) taking velusetrag 30 mg, 47 (48%) taking velusetrag 50 mg, and 39 (36%) taking 1108 Aliment Pharmacol Ther 2010; 32:

8 Clinical trial: efficacy of velusetrag in chronic idiopathic constipation Table 3 Comparison of the activities of velusetrag and placebo on a variety of secondary efficacy endpoints, expressed as the percentage of responders; all comparisons statistically significantly different (P < 0.05) compared with placebo (n = 98) 15 mg (n = 95) 30 mg (n = 91) 50 mg (n = 84) % with 1 SBM week increase in week % with an average of 3 SBM week and an average increase of 1 SBM week % with 3 SBM week in week % with 1 CSBM week increase in week % with 1 SBM week increase for all 4 weeks % with 3 SBM week for all 4 weeks % with 1 CSBM week increase for all 4 weeks % with 3 CSBM week in week % with an average of 3 CSBM week and an average increase of 1 CSBM week % with 3 CSBM week for all 4 weeks % with 3 CSBM week and an increase of 1 CSBM week for all 4 weeks SBM, spontaneous bowel movement; CSBM, complete spontaneous bowel movement. Responders (%) mg Bristol score Absence of straining Bloating No rescue laxative use 30 mg 50 mg Adequate relief Figure 3 Percentage of responders with respect to stool consistency (average Bristol score of 3 5 in week 4), straining (average straining score of <1 in week 4), bloating (absence in week 4), no rescue laxative use over the 4-week treatment, and adequate relief in week 4(P < 0.01, P < vs. placebo). placebo. The majority of adverse events were mild or moderate in severity. The most frequently reported adverse events (occurring in 5% of patients in any treatment group) were GI-related and headache (Table 4). Diarrhoea, an expected event for a prokinetic agent, was reported by 15% at 50 mg, 11% at 30 mg and 12% at 15 mg compared with 1% for the placebo group. Electrolyte imbalances were not reported in any patient experiencing diarrhoea. The incidence of headache was identical for the velusetrag 15 mg and placebo groups, but higher at the 30 and 50 mg doses. The onset of diarrhoea and nausea as well as headache was frequently on the first or second day of dosing (Figure 4) with a mean duration of 1 to 7 days. No events in the cardiovascular system organ class were reported in patients receiving velusetrag, while there were single reports of palpitations and ventricular extrasystoles in the placebo group. A total of 19 patients discontinued the study due to adverse events (Figure 1). Events leading to discontinuation were diarrhoea, nausea, vomiting and headache, almost all occurring early in the treatment period. Nine patients in the 50 mg group discontinued for early onset (within the first few days after initiation of treatment) diarrhoea loose or watery stools compared with Aliment Pharmacol Ther 2010; 32:

9 M. Goldberg et al. Number (%) of patients, by treatment group (n =107) 15 mg (n = 101) 30 mg (n= 96) 50 mg (n= 97) Headache 6 (6) 6 (6) 10 (10) 20 (21) Diarrhoea 1 (1) 12 (12) 11 (11) 15 (15) Nausea 3 (3) 5 (5) 4 (4) 15 (15) Vomiting 1 (1) 4 (4) 2 (2) 7 (7) Abdominal pain upper 1 (1) 3 (3) 5 (5) 3 (3) Flatulence 3 (3) 3 (3) 5 (5) 2 (2) Table 4 Incidence of adverse events reported by at least 5% of patients in any treatment group (safety population) one patient each in the 15 and 30 mg groups. Two serious adverse events (abdominal pain reported by one patient on placebo and another on 15 mg velusetrag) (a) 22 Patients (%) (b) Patients (%) Including day 1 (n = 107) 15 mg (n = 101) 30 mg (n = 96) 50 mg (n = 97) Headache Diarrhoea Nausea Excluding day 1 (n = 107) 15 mg (n = 100) 30 mg (n = 96) 50 mg (n = 87) Headache Diarrhoea Nausea Figure 4 Incidence rates for headache, diarrhoea and nausea over the entire treatment period (all patients) (a) and excluding those events with onset on Day 1 (excluding those dosed only on Day 1) (b). were reported in the study. Neither event was considered by the reporting investigator to be treatmentrelated. There were no clinically relevant changes in haematology, biochemistry, urinalysis, vital signs or ECG parameters in any of the treatment groups. In particular, no velusetrag-treated subject had a QTcF value of 500 ms or an increase of 60 ms from baseline at their week 2 or 4 visits. The incidence of QTcF and QTcB values in the 450 to <500 ms range and of increases from baseline of 30 to <60 ms was similar across all treatment groups and visits. After 4 weeks of treatment, mean (standard deviation) QTcF values were (19.1), (15.3), (18.0) and (18.4) ms for the 15, 30 and 50 mg velusetrag and placebo groups respectively. The percentage of patients with maximum QTcF values of 450 to <500 ms was 3%, 2% and 0% for each of the 15, 30 and 50 mg velusetrag treatment groups respectively, and 1% for the placebo group and the percentage with QTcF increases from baseline of 30 to <60 ms was 6%, 7% and 6%, for each of the 15, 30 and 50 mg velusetrag treatment groups respectively, and 2% for the placebo group. DISCUSSION The results of this randomized, placebo-controlled, double-blind study demonstrate that velusetrag is efficacious and well tolerated over 4 weeks of treatment in patients with severe chronic constipation. Improvements in spontaneous and complete spontaneous bowel movement frequency, as well as in stool characteristics and symptoms and in patient satisfaction, were evident in each week of treatment. Relative to placebo, the time to first SBM after initiation of treatment was reduced and the time to first rescue medication use was increased with velusetrag Aliment Pharmacol Ther 2010; 32:

10 Clinical trial: efficacy of velusetrag in chronic idiopathic constipation For these severely constipated patients ( CSBM week at baseline), the magnitude of effect of velusetrag (an increase to approximately 2.5 CSBM week) represents a clinically meaningful improvement. 38 On the basis of increases in the numbers of SBM and CSBM per week, on average, treatment with velusetrag normalized bowel movement frequency. Stool consistency was more often in the normal range during treatment with velusetrag, and straining and bloating were reduced. These findings are consistent with data from studies with velusetrag in healthy human subjects and observations with other drugs demonstrating activity at the 5-HT 4 receptor , 27, 31, 32 Responder rates for tegaserod have been defined based on an average increase of 1 CSBM week from baseline during weeks 1 to 4 of treatment. By this definition, response rates with 2 and 6 mg tegaserod daily were 36 and 40% respectively, compared with 27% for placebo 15 and the corresponding numbers needed to treat (NNT; the reciprocal of the absolute difference in responder rates between each active dose and placebo) were 11.1 and 7.3 respectively. 15 By the same criterion, the percentages of responders to velusetrag 15 mg and placebo were 42 and 13 respectively (Table 3) and the NNT was 3.4. However, in view of differences in study design and lack of direct head-to-head comparison between tegaserod and velusetrag, comparison of clinical response rates should not be made. 39 The activity of velusetrag with respect to SBM frequency, CSBM frequency and other response criteria appeared more pronounced during the first week of treatment. In addition, for those patients who experienced diarrhoea, the onset was typically in the first few days after the initiation of treatment. These results may reflect an initial exaggeration of prokinetic activity, possibly in the absence of endogenous 5-HT 4 tone, leading to a flushing effect, or alternatively, some degree of receptor desensitization. Nevertheless, the increase in SBM and CSBM frequency relative to placebo was statistically significant in each week of the 4-week treatment period. In this Phase 2 study, velusetrag appeared to be generally well tolerated over the 4 weeks of treatment. Nausea, diarrhoea and headache generally occurred early after initiation of therapy, and are also commonly observed with other 5-HT 4 receptor agonists. 15, 27 The incidence of diarrhoea was greater at 15 mg than in the placebo group, but the incidence of headache and nausea was similar to placebo, and the incidence of all these adverse events was less than at the highest dose of velusetrag (50 mg). A higher discontinuation rate for such events was reported for the 50-mg dose group than for the other treatment groups, particularly after the first dose of velusetrag, although 80% of patients at this high dose still completed the study. No cardiac adverse events were reported by velusetrag-treated patients in this study, and there were no clinically relevant trends on 12-lead ECGs after 2 and 4 weeks of treatment. In conclusion, velusetrag showed significantly greater efficacy than placebo in patients with chronic idiopathic constipation on each endpoint evaluated. Potentially beneficial effects were maintained for the 4-week duration of this study. Overall evaluation of efficacy and safety data suggests that the 15 mg dose level is at least as efficacious as higher doses and has a superior adverse event profile. Additional studies of longer duration are justified to obtain a more complete understanding of the potential clinical utility of velusetrag in chronic idiopathic constipation. ACKNOWLEDGEMENTS Declaration of personal interests: M. Goldberg is an employee of Amgen, Inc. and a former employee of Theravance, Inc. He owns stock in Amgen, Inc. J. F. Johanson has served as a speaker, a consultant, and an advisory board member for Citius, Prometheus, Salix, Schering Plough, Takeda and Theravance. A. Mangel has served as a consultant for Theravance, Tioga, Pharmos, Ardelex, Novartis, Glaxo and Synergy. He is an employee of RTI Health Solutions and, as such, has received no personal compensation from any of these companies. M. Kitt has served as a consultant for Theravance. He is an employee of Nuon Therapeutics and a former employee of Theravance. Y-P. Li, D. T. Beattie, K. Kersey and O. Daniels are employees of Theravance, and own stock or shares in Theravance. K. Kersey and O. Daniels also own stock or shares in Amgen. The contributions of the site investigators and their staff and patients who participated in the study are gratefully acknowledged. We thank Ms B. Garofalo, who coordinated the ACCORD study as an employee of Theravance, Inc. Declaration of funding interests: This study and writing of this paper were funded by Theravance, Inc. Writing support during the final stages of manuscript preparation was provided by Kim Hoag, an employee of Theravance, Inc. Aliment Pharmacol Ther 2010; 32:

11 M. Goldberg et al. REFERENCES 1. Higgins PD, Johanson JF. Epidemiology of constipation in North America: a systematic review. Am J Gastroenterol 2004; 99: Johnson DA. Treating chronic constipation: How should we interpret the recommendations? Clin Drug Investig 2006; 26: Johanson JF, Kralstein J. Chronic constipation: a survey of the patient perspective. Aliment Pharmacol Ther 2007; 25: Edwards CA, Tomlin J, Read NW. Fibre and constipation. Br J Clin Pract 1988; 42: Ramkumar D, Rao SS. Efficacy and safety of traditional medical therapies for chronic constipation: systematic review. Am J Gastroenterol 2005; 100: Voderholzer WA, Schatke W, Muhldorfer BE, et al. Clinical response to dietary fiber treatment of chronic constipation. Am J Gastroenterol 1997; 92: Gattuso JM, Kamm MA. Review article: the management of constipation in adults. Aliment Pharmacol Ther 1993; 7: Lacy BE, Chey WD. Lubiprostone: chronic constipation and irritable bowel syndrome with constipation. Exp Opinion Pharmacother 2009; 10: Sucampo Pharmaceuticals, Inc. AMITIZA Ò (Lubiprostone) Prescribing Information, Revised May Baker DE. Rationale for using serotonergic agents to treat irritable bowel syndrome. Am J Health Syst Pharm 2005; 62: Pan H, Galligan JJ. 5-HT 1A and 5-HT 4 receptors mediate inhibition and facilitation of fast synaptic transmission in enteric neurons. Am J Physiol 1994; 266: G Bouras EP, Camilleri M, Burton DD, et al. Selective stimulation of colonic transit by the benzofuran 5HT 4 agonist, prucalopride, in healthy humans. Gut 1999; 44: Degen L, Matzinger D, Merz M, et al. Tegaserod, a 5-HT 4 receptor partial agonist, accelerates gastric emptying and gastrointestinal transit in healthy male subjects. Aliment Pharmacol Ther 2001; 15: Emmanuel AV, Kamm MA, Roy AJ, et al. Effect of a novel prokinetic drug, R093877, on gastrointestinal transit in healthy volunteers. Gut 1998; 42: Kamm MA, Müller-Lissner S, Talley NJ, et al. Tegaserod for the treatment of chronic constipation: a randomized, double-blind, placebo-controlled multinational study. Am J Gastroenterol 2005; 100: Tack J, Müller-Lissner S, Bytzer P, et al. A randomized controlled trial assessing the efficacy and safety of repeated tegaserod therapy in women with irritable bowel syndrome with constipation. Gut 2005; 54: Johanson JF, Wald A, Tougas G, et al. Effect of tegaserod in chronic constipation: a randomized, double-blind, controlled trial. Clin Gastroenterol Hepatol 2004; 2: Beattie DT, Smith JAM. Serotonin pharmacology in the gastrointestinal tract: a review. Naunyn Schmiedeberg s Arch Pharmacol 2008; 377: Sanger GJ. 5-hydroxytryptamine and the gastrointestinal tract: where next? Trends Pharmacol Sci 2008; 29: De Maeyer JH, Lefebvre RA, Schuurkes JA. 5-HT 4 receptor agonists: similar but not the same. Neurogastroenterol Motil 2008; 20: Nilsson T, Longmore J, Shaw D, et al. Characterisation of 5-HT receptors in human coronary arteries by molecular and pharmacological techniques. Eur J Pharmacol 1999; 372: Centurion D, Sanchez-Lopez A, Ortiz MI, et al. Mediation of 5-HT-induced internal carotid vasodilation in GR and ritanserin-pretreated dogs by 5-HT 7 receptors. Naunyn-Schmiedeberg s Arch Pharmacol 2000; 362: Ellis ES, Byrne C, Murphy OE, et al. Mediation by 5-hydroxytryptamine 2B receptors of endothelium-dependent relaxation in rat jugular vein. Br J Pharmacol 1995; 114: Ishida T, Kawashima S, Hirata K, et al. Nitric oxide is produced via 5-HT 1B and 5-HT 2B receptor activation in human coronary artery endothelial cells. Kobe J Med Sci 1998; 44: Coremans G, Kerstens R, De Pauw M, et al. Prucalopride is effective in patients with severe chronic constipation in whom laxatives fail to provide adequate relief Results of a double-blind, placebocontrolled clinical trial. Digestion 2003; 67: Emmanuel AV, Roy AJ, Nicholls TJ, et al. Prucalopride, a systemic enterokinetic, for the treatment of constipation. Aliment Pharmacol Ther 2002; 16: Quigley EMM, Vandeplassche L, Kerstens R, et al. Clinical trial: the efficacy, impact on quality of life, and safety and tolerability of prucalopride in severe chronic constipation a 12-week, randomized, double-blind, placebocontrolled study. Aliment Pharmacol Ther 2009; 29: Smith JAM, Beattie DT, Marquess D, et al. The in vitro pharmacological profile of TD-5108, a selective 5-HT 4 receptor agonist with high intrinsic activity. Naunyn-Schmiedeberg s Arch Pharmacol 2008; 378: Beattie DT, Armstrong SR, Shaw JP, et al. The in vivo gastrointestinal activity of TD-5108, a selective 5-HT 4 receptor agonist with high intrinsic activity. Naunyn-Schmiedeberg s Arch Pharmacol 2008; 378: Shaw J-P, Beattie D, Cheong S-K, et al. Preclinical Pharmacokinetics of TD-5108, a selective, high intrinsic activity and orally bioavailable 5-HT 4 receptor agonist. AAPS Ann Meet Expos 2007; 9(S2): Goldberg MR, Wong SL, Ganju J, et al. TD-5108, a selective 5-HT 4 agonist with high intrinsic activity, shows immediate and sustained prokinetic activity in healthy subjects. Gastroenterology 2007; 132(4, Suppl 2): Manini ML, Camilleri M, Goldberg M, et al. Effects of velusetrag (TD-5108) on gastrointestinal transit and bowel function in health and pharmacokinetics in health and constipation. Neurogastroenterol Motil 2010; 22: Drossman DA. The functional gastrointestinal disorders and the Rome III process. Gastroenterol 2006; 130: Johanson JF, Ueno R. Lubiprostone, a locally acting chloride channel activator, in adult patients with chronic constipation: a double-blind, placebo-controlled, dose-ranging study to evaluate efficacy and safety. Aliment Pharmacol Ther 2008; 25: Lembo AJ, Kurtz CB, MacDougall JE, et al. Efficacy of linaclotide for patients with chronic constipation. Gastroenterology 2010; 138: Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol 1997; 32: Westfall PH, Tobias RD, Rom D, Wolfinger RD, Hochberg Y. Multiple Comparisons and Multiple Tests Using SAS Ò. Cary, NC: SAS Institute, Inc, 1999; Paré P, Ferrazzi S, Thompson WG, et al. An epidemiological survey of constipation in Canada: definition, rates, demographics, and predictors of health care seeking. Am J Gastroenterol 2001; 96: Kirsch I, Moncrieff S. Clinical trials and the response rate illusion. Contemporary Clinical Trials 2007; 28: Aliment Pharmacol Ther 2010; 32: