Secondary Logo

Journal Logo


Bimodal Release Ondansetron Improves Stool Consistency and Symptomatology in Diarrhea-Predominant Irritable Bowel Syndrome: A Randomized, Double-Blind, Trial

Plasse, Terry F. MD1; Barton, Gary MD2; Davidson, Evelyne MD3; Abramson, Danielle PhD1; Kalfus, Ira MD1; Fathi, Reza PhD1; Raday, Gilead MS1; Harris, M. Scott MD4

Author Information
The American Journal of Gastroenterology: September 2020 - Volume 115 - Issue 9 - p 1466-1473
doi: 10.14309/ajg.0000000000000727



Irritable bowel syndrome (IBS) is a functional bowel disorder associated with abdominal pain and disordered defecation (1). Unlike inflammatory bowel diseases, no obvious pathology is noted macroscopically or microscopically. However, biochemical and enteric neural abnormalities have been found when biopsies and fecal material from normal patients and patients with IBS are compared (2–4). In addition, IBS has been linked with low-grade inflammation (5,6).

Four IBS subtypes have been described; diarrhea-predominant (IBS-D) accounts for 30%–45% of all US adult patients with IBS (7).

Several 5-HT3 antagonists have been studied for the treatment of IBS-D. 5-hydroxytryptamine (5-HT) acts on 5-HT3 receptors in enteric neurons, inducing intestinal secretion and propulsion; 5-HT3 inhibition decreases intestinal motility and increases bowel transit time (8). Alosetron is approved for the treatment of IBS-D in women with warnings regarding severe constipation and ischemic colitis. Other 5-HT3 antagonists, including ramosetron, have demonstrated activity in IBS-D as well (9). Immediate release ondansetron has been evaluated in several small studies at doses from 4 mg daily to 16 mg tid, with favorable results, and no evidence of ischemic colitis or other significant side effects have been observed (10–12).

RHB-102 is an investigational bimodal release formulation of ondansetron. The 12 mg dosage strength used in this study contains 9 mg ondansetron in a controlled-release matrix and 3 mg in an immediate release outer coating. It is designed for once daily administration. Overall bioavailability of ondansetron administered once daily as RHB-102 is similar to that of 3 times daily immediate release ondansetron, but once daily administration of the bimodal release formulation results in immediate bioavailability and prolonged plasma levels of ondansetron, allowing once daily dosing (13). The 12 mg dose represents an intermediate dose in the range previously studied in IBS-D and below the labeled dose for the treatment of nausea and vomiting (14).

We report below the outcome of a phase 2 study evaluating the efficacy and safety of RHB-102 in patients with IBS-D.


This was a randomized, double-blind, placebo-controlled study conducted at 16 sites in the United States (see Table 1, Supplemental Digital Content 1, between May 2016 and July 2017. The primary objective of this study was assessment of the effect of RHB-102 on stool consistency based on evaluation by the Bristol Stool Scale. Secondary objectives were assessment of other efficacy parameters and safety of RHB-102 in patients with IBS-D.

The Institutional Review Boards approved the study protocol at all participating sites. Informed consent was obtained from all patients, and translations were provided for non-English-speaking patients. All study procedures were consistent with the International Committee on Harmonisation Good Clinical Practice Guidelines and followed the Declaration of Helsinki for protection of human subjects. The trial was registered in (NCT02757105).


Male and female patients aged 18 and older who met the Rome III criteria for IBS-D (15) were eligible to participate; patients with IBS-M were excluded. Patients with evidence of other causes for bowel disease were excluded, as were patients with a history of abdominal surgery (other than appendectomy or cholecystectomy) at any time; use of alosetron, eluxadoline, or other treatments for IBS-D within 2 weeks (for alosetron or eluxadoline) to 4 months (for rifaximin) before the formal screening program; and the use of apomorphine within 24 hours of screening, based on the reports of hypertension and loss of consciousness when coadministered with ondansetron (14). Additional exclusions were corrected QT interval greater than 450 ms on baseline ECG (because of reports of QT prolongations with doses of 32 mg or greater), use of medications known to prolong the QT interval, other significant cardiac abnormalities, standard hematologic and biochemical parameters worse than grade 1 per NCI CTCAE v 4.03, and baseline C-reactive protein (CRP) greater than twice the upper limit of normal (5.0 mg/L) for the reference laboratory.

Study design

After initial assessment, patients underwent a two-week screening phase. Patients were required to document loose or watery stools (Bristol Stool Scale 6 or 7) at least 2 days per week and an average worst daily pain intensity at least 3.0 on a Likert scale of 0–10 each week during the screening period. Baseline symptoms present and their severity were collected over the screening period. Throughout the study, including the screening phase, patients were not allowed to take antispasmodics. Patients were not allowed to take any loperamide during the screening phase, although they were allowed a limited amount during the treatment phase of the study, maximum of 4 mg per day up to 3 days per week, as rescue medication. Qualifying patients who had not undergone colonoscopy within the 5 years before consent or had one or more concerning signs or symptoms were required to undergo colonoscopy before randomization. Biopsies were performed on patients undergoing colonoscopy for this study, and findings indicating presence of other conditions, such as microscopic colitis, were exclusionary.

Study medication was provided based on the allocation sequence. Patients were stratified by gender and randomly assigned to receive either RHB-102 12 mg or matching placebo in a 3:2 ratio. RHB-102 and placebo were identical in appearance and were centrally packed and labeled by an independent contract manufacturing organization before being distributed to sites. The randomization schedule used a permuted block structure with blocks of 5. The schedule and allocation sequence were generated using the PROC PLAN procedure from SAS Software Version 9.3 (SAS Institute, Cary, NC).

Throughout the study, patients kept daily diaries of bowel movement occurrences and consistencies, symptoms, adverse events (AEs), and concomitant medications.

Patients who developed constipation, defined as no bowel movement for at least 48 hours, were instructed to reduce dosing to once every other day. If IBS symptoms recurred or worsened, the patient was instructed to increase the dose back to once daily. If constipation persisted, study medication was discontinued.

Efficacy end points

The primary end point was the proportion of patients who were overall stool consistency responders, defined as patients who were weekly stool consistency responders for at least 4 of the 8 study weeks. A weekly stool consistency responder, per the US FDA guidance on IBS-D (16), was a patient who experienced a 50% or greater reduction in the number of days in the week with at least one stool with a consistency of 6 or 7 on the Bristol Stool Scale as compared to the two-week baseline. In addition, the patient could not have an increase in average worst daily pain >10% over baseline for the week. Patients taking rescue medication on a given day were considered treatment failures for that day. Stool consistency, rather than the composite end point of stool consistency and abdominal pain, was chosen as the primary end point because it was felt that the former was more sensitive to the effect of ondansetron and, hence, would show a statistically significant effect in a moderate-sized study.

Secondary end points included the proportion of patients who were overall abdominal pain responders and composite weekly responders for at least 4 weeks of 8 weeks. A weekly abdominal pain responder was defined as a patient who experienced a decrease in the weekly average of worst abdominal pain in the past 24 hours score ≥30% compared with the baseline and the number of days per week with at least one stool with consistency of type 6 or 7 that is the same as the baseline or decreased and the number of stools of type 6 or 7 on those days remained unchanged or decreased. A composite weekly responder was a patient who was both a stool consistency and pain responder in the same week.

Additional efficacy evaluations included stool frequency, urgency, and abdominal discomfort, which were assessed on numeric rating scales of 0–10 and interference with daily activities, which was assessed on a scale of 0–4.

Safety assessments

All patients were assessed while on treatment and for 28 days thereafter for the occurrence of AEs, which were coded using MedDRA version 13.1 and graded according to the NCI CTCAE v4.03 criteria. Biochemical and hematologic laboratory abnormalities developing on the study and those present at baseline that worsened by at least one NCI CTCAE v4.03 grade were considered AEs, regardless of investigator assessment of clinical relevance.

Statistical analysis

Based on responses rates of 70% and 40% in the ondansetron and placebo groups, as suggested in an earlier study (12) and 3:2 randomization,104 patients were required to demonstrate significance with 80% power and a two-sided significance level of α = 0.05 and 104 patients would be needed. The sample size was increased to 120 (72 RHB-102 and 48 placebo) to allow for unevaluable patients and early dropouts.

All analyses were performed on the modified intent to treat (mITT) population, defined as all randomized patients who received at least one dose of study medication. Patients lacking more than 3 days' data during any given week were considered nonresponders for the parameter lacking data for that week.

The safety population was all patients who received any medication by treatment received.

Continuous variables were summarized using standard descriptive statistics which included the number of patients (n), arithmetic mean, SD, minimum, median, and maximum. Categorical variables were described using absolute and relative frequencies. Percentages by categories were based on the number of patients with no missing data. Confidence intervals were two-sided 95% confidence intervals if not otherwise specified. Statistical tests were two-sided with a significance level of α = 0.05. Prospective analyses were performed for all efficacy end points for patients above and below the median baseline CRP.

A χ2 test with continuity correction at the significance level α = 0.05 was used for comparison of the treatment groups regarding overall and weekly pain response. No corrections for multiple comparisons were made.

Efficacy data were analyzed prospectively by predesignated subgroups by gender and baseline CRP (above or below the median for CRP for study patients).

All data processing, summarization, and analyses were performed using SAS Version 9.3 or higher.



A total of 326 patients were screened, of whom 127 were randomized and 126 treated (Figure 1). The most common reasons for exclusion were CRP above allowable per protocol (n = 52), average worst daily pain <3.0 (n = 36), proportion of days during baseline evaluation with stool consistency 6 or 7 lower than required by protocol (n = 19), comorbidities which could interfere with patient evaluation or safety (n = 17), and failure to meet the Rome III criteria (n = 10).

Figure 1.
Figure 1.:
CONSORT diagram.

All 75 patients randomized to RHB-102 received RHB-102. Fifty-one of 52 patients randomized to placebo received placebo; one patient randomized to placebo withdrew before treatment. These 126 patients constitute both the modified intent to treat and safety populations.

Demographics and patient characteristics are shown in Table 1. Patients were well-matched between the treatment groups. Overall, the median age was 40, with a range of 18–74. Thirty percent of the patients were men. IBS symptomatology was similar in intensity and frequency of symptoms between the treatment groups. Median (range) duration of symptoms was 5.1 (1–42) years in the RHB-102 group and 3.5 (1–43) years in the placebo group. Forty-nine percent of RHB-102 and 45% of placebo patients had at least one IBS symptom self-described as severe; overall, 93% had at least one moderate symptom, with similar frequency in both treatment groups. One patient was entered into the study despite a baseline CRP of 26.02 mg/L in violation of the protocol requirement for baseline CRP ≤10 mg/L.

Table 1.
Table 1.:
Demographics and disease characteristics: mITT population

Previous treatment with IBS-D-related medication was more common in the RHB-102 group (32%) as compared to the placebo group (22%). This difference was accounted for by the use of loperamide, taken by 28% of patients in the RHB-102 group as compared to 16% of patients in the placebo group.


Fifty-six percent of patients in the RHB-102 group vs 35.3% in the placebo group achieved the primary end point of the study, overall stool consistency response, P = 0.036. Overall pain response rates were 50.7% in the RHB-102 group and 39.2% in the placebo group (P = 0.278); composite response rates were 40.0% and 25.5%, respectively (P = 0.135). These results are shown in Figure 2. Differences in stool consistency response rates were noted within the first week of treatment and persisted throughout the study (Figure 3); differences were statistically significant (P ≤ 0.05) from weeks 3 to 6, inclusive. Differences between RHB-102 and placebo dissipated within one week of stopping study medication. Approximately 25% of patients in each treatment group maintained their responses after discontinuation of treatment.

Figure 2.
Figure 2.:
Response rates in the mITT population. *P = 0.036.
Figure 3.
Figure 3.:
Stool consistency response rates by week on study. *P ≤ 0.05.

Thirty patients (40.0%) in the RHB-102 and 16 patients (31.4%) in the placebo group used rescue medication at least once during the study. Stool consistency response rates when the use of rescue medication was not considered in the definition of response were similar to those calculated when excluded (57.3% vs 37.3%, RHB-102 vs placebo, P = 0.042). Paralleling the results for stool response rate, the mean number of days with the Bristol Stool Scale stool grade 6 or 7 each week was lower for active than for placebo through week 8, the last treatment week, as shown in Figure 1, Supplemental Digital Content 2, The curves merged by week 10, 2 weeks after the end of treatment.

Trends to greater improvements in interference with daily activities, mean abdominal discomfort, and mean urgency were observed over time in patients treated with RHB-102 as compared to placebo (see Figures, Supplemental Digital Contents 3–5,,, No differences between treatment groups in weekly mean daily stool frequency were observed (see Figure 5, Supplemental Digital Content 6,

The effect of gender on each response parameter is shown in Table 2. The effect of treatment on stool consistency was similar in both genders, but while the effect of treatment on pain was substantial in men, it was minimal in women.

Table 2.
Table 2.:
Response rate by gender, mITT population

Data were analyzed prospectively by baseline CRP above vs below median (2.09 mg/L), as shown in Figure 4. For the primary end point, stool consistency response, there was virtually no difference in response rates between active and placebo for patients with baseline CRP below the median. However, in patients with baseline CRP above the median, there was a greater difference in response rates compared with the general population (59.5% vs 23.1%, RHB-102 vs placebo, P = 0.009). Similarly, for patients with CRP above the median, worst pain response rates were 56.8% and 30.8% for RHB-102 and placebo, respectively (P = 0.075), and composite response rates were 45.9% and 19.2%, respectively (P = 0.055). There was no treatment effect for either pain or composite response in patients with baseline CRP below median.

Figure 4.
Figure 4.:
Response rates by baseline CRP. ***P = 0.009; **P = 0.075; *P = 0.055. Differences between treatment groups for patients with baseline CRP ≤ median: P > 0.9 for each endpoint. CRP, C-reactive protein.


Treatment was well tolerated. AEs experienced by 5% or more of patients in either treatment group, regardless of relation to study medication, are shown in Table 3. Most AEs in both treatment groups were mild, and most were balanced between treatment groups. No patient in either group experienced a life-threatening or serious AE.

Table 3.
Table 3.:
Adverse events experienced by ≥5% of patients in either treatment group

Gastrointestinal disorders were more common in the RHB-102 than in the placebo group. This difference was because of the increased incidences of constipation and flatulence in the RHB-102 compared with the placebo groups: for constipation, the incidences were 13% and 4% and for flatulence, the incidences were 8% and 0% in the active and placebo groups, respectively. In the RHB-102 group, constipation was mild in 8%, moderate in 4%, and severe in 1.3% of patients; in the placebo group, constipation was mild in all patients. Eleven patients (14.7%) in the active group and 2 in the placebo group (3.9%) decreased dosing frequency because of constipation. Two patients in the RHB-102 group discontinued therapy because of constipation: one in conjunction with painful defecation and the other patient had previously reduced dosing frequency before discontinuing completely. In each case, the constipation resolved within several days without sequelae. Two additional patients reduced dosing to alternate days in accordance with the protocol-specified procedures. No patient in the placebo group discontinued because of an AE. No clinical cardiac AEs were noted.

Laboratory abnormalities first noted on the study or worsening by at least one grade were routinely considered AEs, regardless of investigator assessment or attribution, accounting for the high incidence of laboratory (investigation) AEs. The only laboratory abnormality which occurred more than 5% more frequently in the RHB-102 than in the placebo group was blood glucose increased, which was noted in 22.7% of RHB-102 and 17.6% of placebo patients at least once during the study.


This randomized, double-blind, placebo-controlled study demonstrated that once daily dosing with 12 mg bimodal release formulation of ondansetron significantly improved stool consistency response rates in patients with IBS-D, the primary end point for this study. This trial was not powered to detect significant differences in other end points. Fifty-six percent of patients treated with RHB-102 vs 35.3% of those in the placebo group met the FDA guidance definition of overall stool consistency response over the duration of the study (P = 0.036). Differences in stool consistency response rates were noted within the first week of treatment and persisted throughout the study, although less pronounced at week 8; this difference subsided after treatment was stopped. A greater proportion of RHB-102 than placebo patients met the pain and composite end points, although these differences were not statistically significant. The magnitude of effects was comparable with those with other approved products for IBS-D, alosetron (17,18), rifaximin (19), and eluxadoline (20) across all study end points.

The treatment effect for stool consistency response was similar in both men and women, as shown in Table 2. Although numerically there was a large worst pain treatment effect in men, RHB-102 had essentially no effect on pain in women.

Screening procedures used in this study were pragmatic, i.e., those generally used by clinicians when deciding on whether symptoms are because of IBS-D or to other causes. We routinely assessed CRP level at baseline, as a nonspecific indicator of inflammation. Although patients with rigorously diagnosed IBS-D may have modestly elevated CRP levels (21), a marked elevation suggests presence of another condition. Hence, we excluded patients with baseline CRP greater than twice the upper limit of normal for the study reference laboratory.

Mildly elevated CRP might represent low level inflammation linked to increased intraluminal proteases (5,6). We prospectively analyzed the data by baseline CRP above and below the median. Treatment effects for the primary and secondary end points were limited to the subgroup of patients with CRP above median for our study population, as shown in Figure 4. The median value, 2.09 mg/L, was well within the normal range for our reference laboratory, 5 mg/L. In addition, 84% of patients in the RHB-102 group and 78% in the placebo group had normal baseline CRP values. Although response rates to active drug were somewhat higher in patients with baseline CRP above as compared to below the median, the greater difference was in response rates in the placebo group in which the response rates were far lower in patients with baseline CRP above than below the median. To the best of our knowledge, this difference in response rates has not previously been reported. Baseline CRP may represent a marker for individuals responsive to drugs that slow intestinal transit.

Treatment with RHB-102 was well tolerated. Some patients experienced constipation, a known side effect of 5-HT3 antagonists, but this was controlled in most patients with a reduction of dosing to every other day. No patient experienced a serious AE, such as ischemic colitis, which has been noted with alosetron, although the number of patients in this study may preclude detection of a low frequency event. Ischemic colitis has not been reported in the general use of ondansetron or with ramosetron, a 5-HT3 antagonist used in Japan. In larger studies and eventual clinical use, awareness of the possibility of serious AEs as sequelae of constipation, with appropriate dosing modification, may reduce the occurrence of serious AEs.

In summary, RHB-102, a once daily bimodal release formulation of ondansetron, significantly improved stool consistency in patients with IBS-D. Abdominal pain and composite response were also improved, although the differences were not statistically significant in this relatively small study. Treatment appeared to work at least as well in men as in women. An intriguing relationship between baseline CRP level, even when normal, and response to treatment was noted, suggesting that low level inflammation may affect outcome of therapy. This remains to be confirmed in future studies.


Guarantor of the article: Terry F. Plasse, MD.

Specific author contributions: T.F.P. was the primary author of the protocol and of the manuscript and managed the study on behalf of RedHill Biopharma. G.B. and E.D. participated as investigators in the study and reviewed and revised the manuscript. R.F. was responsible for management of the clinical trial material and participated in discussion of all pharmacokinetic issues during the study. D.A. and G.R. were involved management of the study on behalf of the sponsor. D.A., I.K. and M.S.H. reviewed the study protocol, assisted in the interpretation of the data, participated in discussion of issues during the study, and reviewed and revised the manuscript. T.F.P. accepts full responsibility for the conduct of the study and the veracity and accuracy of the data and analysis.

Financial support: This study was supported by the developer of RHB-102, RedHill Biopharma, Ltd, Tel Aviv, Israel.

Potential competing interests: Drs Plasse, Abramson, Kalfus, Fathi, and Harris are consultants to RedHill Biopharma, the sponsor of the study. Mr Raday is an employee of RedHill Biopharma. Drs Barton and Davidson were investigators compensated for their participation by RedHill Biopharma.

Study Highlights


  • ✓ IBS-D is a common ailment causing substantial discomfort and interference with normal activities.
  • ✓ Ondansetron, a 5-HT3 antagonist, has shown activity in IBS-D over a range of doses in several small studies.


  • ✓ Once daily bimodal release ondansetron demonstrated a statistically significant improvement in the FDA end point of stool consistency response over placebo.
  • ✓ Baseline CRP may be a strong predictor of response to ondansetron in IBS-D.

Clinical trials gov identifier: NCT02757105.


1. Lacy BE, Mearin F, Chang L, et al. Bowel disorders. Gastroenterol 2016;150:1393–407.
2. Buhner S, Hannes H, Hartwig K, et al. Protease signaling through protease activated receptor 1 mediate nerve activation by mucosal supernatants from irritable bowel syndrome but not from ulcerative colitis patients. PLOS One 2018;13:e0193943.
3. Rolland-Fourcade C, Denadai-Souza A, Cirillio C, et al. Epithelial expression and function of trypsin-3 in irritable bowel syndrome. Gut 2017;66:1767–78.
4. Tooth D, Garsed K, Singh G, et al. Characterisation of faecal protease activity in irritable bowel syndrome with diarrhea: Origin and effect of gut transit. Gut 2014;63:753–60.
5. Collins SM, Denou E, Verdu EF, et al. The putative role of the intestinal microbiota in the irritable bowel syndrome. Dig Liver Dis 2009;41:850–3.
6. Spiller R, Major G. IBS and IBD—separate entities or on a spectrum? Nat Rev Gastroenterol Hepatol 2016;13:613–21.
7. Hungin APS, Chang L, Locke GR, et al. Irritable bowel syndrome in the United States: Prevalence, symptom patterns and impact. Aliment Pharmacol Ther 2005;21:1365–75.
8. Mawe GM, Hoffman JM. Serotonin signaling in the gastrointestinal tract; functions, dysfunctions, and therapeutic targets. Nat Rev Gastroenterol Hepatol 2013;10:473–86.
9. Zheng Y, Yu T, Tang Y, et al. Efficacy and safety of 5-hydroxytryptamine 3 receptor antagonists in irritable bowel syndrome: A systemic review and metaanalysis of randomized controlled trials. PLOS One 2017;12:e0172846.
10. Steadman CJ, Talley NJ, Phillips S, et al. Selective 5-hydroxytryptamine type 3 receptor antagonism with ondansetron as treatment for diarrhea-predominant irritable bowel syndrome: A pilot study. Mayo Clin Proc 1992;67:732–8.
11. Maxton DG, Morris J, Whorwell PJ. Selective 5-hydroxytryptamine antagonism: A role in irritable bowel syndrome and functional dyspepsia? Aliment Pharmacol Ther 1996;10:595–9.
12. Garsed K, Cernova J, Hastings M, et al. A randomised trial of ondansetron for the treatment of irritable bowel syndrome with diarrhea. Gut 2013;63:753–60.
13. Sayegh L, Paquette JM, Essalihi R, et al. Comparative bioavailability study of RHB-102 (24 mg ER tablets QD) versus ondansetron 8 mg trablets BID and a single dose of ondansetron 24 mg in healthy volunteers. Clin Pharm Ther 2015;97:S47. PI-092.
14. US Food and Drug Administration. Zofran (ondansetron) Oral Formulation Approved Labeling (,020605s020,020781s020lbl.pdf) (2017).
15. Ford AC, Bercik P, Morgan DG, et al. Characteristics of functional bowel disorder patients: A cross-sectional survey using the Rome III criteria. Alim Pharmacol Ther 2014;39:312–21.
16. US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research. Guidance for Industry: Irritable Bowel Syndrome-Clinical Evaluation of Drugs for Treatment ( (2012).
17. Camilleri M, Chey WY, Mayer EA, et al. A randomized controlled clinical trial of the serotonin type 3 receptor antagonist alosetron in women with diarrhea-predominant irritable bowel syndrome. Arch Intern Med 2001;161:1733–40.
18. Lembo AJ, Olden KW, Ameen VZ, et al. Effect of alosetron on bowel urgency and global symptoms in women with severe, diarrhea-predominant irritable bowel syndrome: Analysis of two controlled trials. Clin Gastroenterol Hepatol 2004;2:675–82.
19. Pimentel M, Lembo A, Chey WD, et al. Rifaximin therapy for patients with irritable bowel syndrome without constipation. N Engl J Med 2011;364:22–32.
20. Lembo AJ, Lacy BE, Zuckerman MJ, et al. Eluxadoline for irritable bowel syndrome with diarrhea. N Eng J Med 2016;374:242–53.
21. Hod K, Ringel-Kulka T, Martin CF, et al. High sensitive C-reactive protein as a marker for inflammation in irritable bowel syndrome. J Clin Gastroenterol 2016;50:227–32.

Supplemental Digital Content

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of The American College of Gastroenterology