Abstract
Introduction: Tenapanor is a first-in-class, minimally absorbed intestinal sodium/hydrogen exchanger isoform 3 inhibitor approved by the US Food and Drug Administration for adults with irritable bowel syndrome with constipation (IBS-C). Pooled data from the phase 2b (NCT01923428) and phase 3 T3MPO-1 (NCT02621892) and T3MPO-2 (NCT02686138) studies examined the effects of tenapanor on abdominal symptoms independent of tenapanor’s effect on complete spontaneous bowel movement (CSBM) frequency in adults with IBS-C. Methods: This post hoc analysis was performed for patients with no CSBMs in ≥6 of the first 12 weeks of treatment (no-CSBM subgroup). The three-item abdominal score (AS3; the average of weekly abdominal pain, bloating, and discomfort scores) measured abdominal symptom response in tenapanor versus placebo. The overall change from baseline and response rate (improvement of ≥2 points or a reduction of ≥30%) in AS3 and individual abdominal scores during the 12 weeks were assessed. Results: In the pooled safety analysis set (N = 1,382), 641 patients were classified as no-CSBM patients and 640 were included in the efficacy analysis. Among the no-CSBM subgroup, tenapanor-treated patients experienced a greater improvement in AS3 in week 12 versus placebo-treated patients (least squares mean change, −1.74 vs. −1.29; p = 0.007), and the AS3 responder rate was higher for tenapanor (40.2% vs. 29.6%; p = 0.008). Similar improvements were displayed across individual abdominal symptom scores. Diarrhea was the most common adverse event in tenapanor-treated patients. Conclusion: Tenapanor was observed to improve abdominal symptoms independent of its effect on bowel symptoms in adults with IBS-C.
Introduction
Irritable bowel syndrome (IBS) is a common disorder of gut-brain interaction characterized by abdominal pain and altered bowel movements [1, 2]. While the pathophysiology of IBS is not completely understood, motility disturbances, visceral hypersensitivity, and altered central nervous system processing have all been identified as possible disease drivers [3, 4]. IBS is most common in women under the age of 50 years [1, 2] and is classified into subtypes based on the predominant stool consistency: IBS with constipation (IBS-C), IBS with diarrhea, and IBS with mixed bowel habits [1, 2]. Applying the Rome diagnostic criteria, the estimated prevalence of IBS ranges from 4.8% (Rome IV) to 8.8% (Rome III) in the USA adult population, with IBS-C accounting for 16% to 29% of these cases [4].
Abdominal pain is more commonly reported to be bothersome and frequent, and to interfere with daily activities in patients with IBS-C compared with patients with other IBS subtypes [5]. Patients with IBS-C may also experience abdominal symptoms such as bloating, discomfort, cramping, and fullness [6, 7]. Abdominal pain and bloating have been linked to visceral hypersensitivity, which is a heightened sensation in response to physiological stimuli and activation of afferent visceral sensory nerves [8, 9]. Moreover, abdominal pain exacerbation has also been linked to increased intestinal permeability, potentially allowing for enhanced infiltration of macromolecules from the gut lumen into the deeper layers of the intestine (i.e., submucosa) [10‒13]. Abdominal symptoms associated with IBS-C can result in increased use of health care resources, decreased work productivity, and reduced health-related quality of life [14‒16]. While over-the-counter medications can improve stool consistency and frequency, these treatments primarily target the underlying bowel symptoms of IBS (i.e., constipation and/or diarrhea) and may not address abdominal symptoms [15, 17, 18].
The Diary for IBS Symptoms-Constipation (DIBSS-C) was developed by the IBS Working Group of the Critical Path Institute’s Patient-Reported Outcome Consortium with Food and Drug Administration (FDA) guidance [6, 19]. The DIBSS-C is comprised of a three-item abdominal score (AS3), which is a composite measure of three key abdominal symptom scores associated with IBS: pain, bloating, and discomfort [19]. The AS3 was previously used as the primary endpoint in a phase 3b clinical trial examining the efficacy of linaclotide in reducing combined symptoms of pain, bloating, and discomfort through 12 weeks of treatment in patients with IBS-C [20].
Tenapanor is a first-in-class, minimally absorbed, small-molecule inhibitor of intestinal sodium/hydrogen exchanger isoform 3 (NHE3) that has been approved by the FDA for the treatment of adults with IBS-C [21‒23]. NHE3 is expressed on the apical surface of the small intestine and colon and is primarily responsible for the absorption of dietary sodium [23‒25]. By inhibiting NHE3, tenapanor reduces sodium absorption, resulting in increased stool sodium and water content, which accelerates intestinal transit and softens stool [23, 25]. Nonclinical studies have demonstrated that tenapanor can reduce abdominal pain by increasing transepithelial electrical resistance, which reduces intestinal permeability to macromolecules [26, 27]. Furthermore, tenapanor has been found to alleviate abdominal pain by reducing visceral hypersensitivity through normalizing colonic sensory neuronal excitability and transient receptor potential cation channel subfamily V member 1 currents [26]. Therefore, tenapanor has the potential to alleviate both bowel and abdominal symptoms associated with IBS-C.
In a phase 2b and two phase 3 studies, T3MPO-1 and T3MPO-2, tenapanor 50 mg twice a day (bid) significantly increased complete spontaneous bowel movements (CSBMs) and reduced abdominal symptoms compared with placebo [28‒30]. The relationship between these improvements in abdominal symptoms and bowel movements remains undetermined. This post hoc analysis aimed to evaluate the effects of tenapanor on abdominal pain and other abdominal symptoms separately from its effect on CSBM frequency in adults with IBS-C.
Methods
Patients and Study Design
Study design and primary results of the phase 2b (NCT01923428), T3MPO-1 (NCT02621892), and T3MPO-2 (NCT02686138) studies have been previously published [28‒30]. Briefly, patients in the phase 2b study were randomized in an equal allocation ratio into four treatment groups: tenapanor 5 mg bid, 20 mg bid, 50 mg bid, or placebo bid [28]. In the two phase 3 studies, patients were randomized to receive 50 mg of tenapanor or placebo bid for 12 (T3MPO-1) or 26 (T3MPO-2) weeks [29, 30].
In this post hoc analysis, data derived from patients who received tenapanor 50 mg bid or placebo in the phase 2b trial were pooled into their respective groups with patients from the T3MPO-1 and T3MPO-2 trials (all received tenapanor 50 mg bid or placebo; online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000543166). Early study dropouts, as long as they had received one dose of study drug, were included in the pooled safety analysis set. Patients who were in the pooled safety analysis set and met the study entry criteria were included in the pooled intent-to-treat (ITT) analysis set. All patients enrolled in the three trials were adults who met Rome III criteria for IBS-C. A CSBM was defined as a spontaneous bowel movement (a bowel movement occurring without the use of a rescue medication within 24 h) for which the subject responded “yes” to the following question: “did you feel like you completely emptied your bowels?” Although the value of the CSBM frequency is an integer, the average weekly CSBMs are the total number of CSBMs on days with observed data in a week standardized to 7-day frequencies. Therefore, the value of the average weekly CSBMs can have a decimal point and the mean value of the average weekly CSBMs may be less than 1 for highly constipated patients with IBS-C, given that patients might report 0 or 1 CSBM in one analysis week.
Integrated efficacy analyses were independently performed on (a) patients in the pooled ITT analysis set with no CSBMs in at least six of the first 12 weeks of treatment (no-CSBM subgroup) and (b) patients in the pooled ITT analysis set who did not meet the no-CSBM definition (ITT excluding no-CSBM subgroup). By selecting patients with no CSBMs, the analyses could explore the effect of tenapanor on abdominal symptoms even when there were no changes in CSBMs. For consistency, efficacy data from the first 12 weeks of the 26-week T3MPO-2 study were pooled with data from the 12-week randomized treatment period of T3MPO-1 and the 12-week phase 2b study (Figure 1; see online suppl. Fig. 2 for individual study designs). Treatment-emergent adverse events (TEAEs) during the randomized treatment periods of all three clinical trials were summarized for the corresponding no-CSBM subgroup of the pooled safety analysis set and patients in the pooled safety analysis set who did not meet the no-CSBM definition (safety excluding no-CSBM subgroup).
Study designs of phase 2b, T3MPO-1, and T3MPO-2 studies. This post hoc analysis included pooled data from 3 studies of tenapanor 50 mg bid versus placebo. We included 12-week data from tenapanor 50 mg bid and placebo groups of the phase 2b study along with 12-week data from the T3MPO-1 and T3MPO-2 studies. T3MPO-1 included a 4-week withdrawal period, and T3MPO-2 included additional 14 treatment weeks, which are not included in this analysis. aThe phase 2b study was a multiple dose study (5 mg, 20 mg, and 50 mg) with a placebo arm. Only patients randomized to tenapanor 50 mg bid or placebo bid were included in the analysis population. bid, twice a day.
Study designs of phase 2b, T3MPO-1, and T3MPO-2 studies. This post hoc analysis included pooled data from 3 studies of tenapanor 50 mg bid versus placebo. We included 12-week data from tenapanor 50 mg bid and placebo groups of the phase 2b study along with 12-week data from the T3MPO-1 and T3MPO-2 studies. T3MPO-1 included a 4-week withdrawal period, and T3MPO-2 included additional 14 treatment weeks, which are not included in this analysis. aThe phase 2b study was a multiple dose study (5 mg, 20 mg, and 50 mg) with a placebo arm. Only patients randomized to tenapanor 50 mg bid or placebo bid were included in the analysis population. bid, twice a day.
Abdominal Symptom Scores, Endpoints
Patients reported daily assessments for abdominal pain, bloating, and discomfort on an 11-point scale (0–10) through an interactive voice response system telephone diary, where 0 represented no abdominal symptoms and 10 represented very severe abdominal symptoms (online suppl. Information). Changes from baseline in weekly abdominal symptom scores and AS3 were analyzed. The AS3 was retrospectively calculated as an average of the weekly abdominal pain, bloating, and discomfort scores. Changes in scores for abdominal pain, bloating, and discomfort were also assessed individually, per the Rome IV diagnostic criteria for IBS [31]. Weekly scores for each abdominal symptom were calculated as the average score for all days during a valid week (i.e., a week with ≥4 days of reporting of the given abdominal symptom) [19].
The changes from baseline in weekly AS3 and abdominal symptom scores were assessed using a mixed-effects model with repeated measures (MMRM). Each MMRM included treatment, week, and treatment-by-week as fixed effect factors; baseline weekly score and baseline-by-week as fixed effect covariates; and patient as a random effect. An unstructured covariance was employed in each MMRM analysis by default. When the unstructured model failed to converge, a first-order heterogeneous auto-regression covariance structure was employed.
The weekly response in AS3 and abdominal symptom scores were defined as achieving a reduction of ≥2 points in the corresponding weekly score for a given week [19]. The “30% weekly response” was defined as achieving a percent reduction of ≥30% in the corresponding weekly score for a given week. Patients with a missing weekly abdominal pain score or AS3 due to early withdrawal or an “invalid week” (<4 days of reporting of abdominal symptom[s]) were considered nonresponders for that week (i.e., worst-case imputation approach). The association between weekly response status and treatment was assessed using the Pearson’s chi-square test. Due to the nature of post hoc analyses, p values from MMRM for continuous variables and Pearson’s chi-square tests for categorical variables are displayed only to indicate statistical significance at the nominal level (i.e., p < 0.05) without adjusting for multiplicity.
Straining Score
Patients reported daily assessments of straining for each spontaneous bowel movement on a 5-point scale (1–5), where 1 = not at all, 2 = a little bit, 3 = a moderate amount, 4 = a great deal, and 5 = an extreme amount. Changes from baseline in weekly straining score were analyzed. Weekly straining scores were calculated as the average straining score for all valid spontaneous bowel movements based on observed cases without missing data imputations. The change from baseline in weekly straining score was assessed using an MMRM, with the same specifications of the MMRM for the changes from baseline in weekly AS3 and abdominal symptom scores.
Results
Patient Disposition, Demographics, and Baseline Characteristics
The pooled safety analysis set included 1,382 patients (tenapanor, n = 691; placebo, n = 691). The pooled ITT analysis set was a subset of the pooled safety analysis set and included 1,372 patients (tenapanor, n = 684; placebo, n = 688) who met the study entry criteria. In the pooled safety analysis set, 641 patients (tenapanor, n = 275; placebo, n = 366) who reported no CSBMs for at least six of the first 12 weeks of treatment were classified as no-CSBM patients, and 640 of them (tenapanor, n = 275; placebo, n = 365) were included in the pooled ITT analysis set. Demographic and baseline characteristics of the pooled ITT analysis set were similar between the tenapanor and placebo groups (Table 1). Patient demographics and baseline characteristics of the study-level ITT analysis set were also similar across studies (online suppl. Table S1).
Patient demographics and baseline characteristics (no-CSBM subgroup, ITT excluding no-CSBM subgroup, and pooled ITT analysis set)
. | No-CSBM subgroupa . | ITT excluding no-CSBM subgroupb . | Pooled ITT analysis set . | |||
---|---|---|---|---|---|---|
tenapanor 50 mg bid, N = 275 . | placebo bid, N = 365 . | tenapanor 50 mg bid, N = 409 . | placebo bid, N = 323 . | tenapanor 50 mg bid, N = 684 . | placebo bid, N = 688 . | |
Age, mean (SD), years | 46.6 (12.5) | 45.9 (13.3) | 45.0 (13.4) | 43.9 (13.5) | 45.7 (13.1) | 45.0 (13.5) |
BMI, mean (SD), kg/m2 | 29.6 (6.6) | 30.0 (6.8) | 30.2 (7.2) | 29.8 (6.8) | 30.0 (7.0) | 29.9 (6.8) |
Duration of IBS-C at baseline, mean (SD), years | n = 275 | n = 362 | n = 407 | n = 322 | n = 682 | n = 684 |
10.9 (11.9) | 11.9 (12.0) | 11.5 (11.4) | 11.3 (11.8) | 11.2 (11.6) | 11.6 (11.9) | |
Female, n (%) | 226 (82.2) | 300 (82.2) | 333 (81.4) | 272 (84.2) | 559 (81.7) | 572 (83.1) |
Race, n (%) | ||||||
American Indian or Alaskan Native | 0 | 0 | 1 (0.2) | 2 (0.6) | 1 (0.1) | 2 (0.3) |
Asian | 16 (5.8) | 7 (1.9) | 8 (2.0) | 7 (2.2) | 24 (3.5) | 14 (2.0) |
Black | 89 (32.4) | 114 (31.2) | 106 (25.9) | 100 (31.0) | 195 (28.5) | 214 (31.1) |
Multiple | 2 (0.7) | 5 (1.4) | 6 (1.5) | 5 (1.5) | 8 (1.2) | 10 (1.5) |
Other | 3 (1.1) | 5 (1.4) | 0 | 1 (0.3) | 3 (0.4) | 6 (0.9) |
White | 165 (60.0) | 234 (64.1) | 287 (70.2) | 208 (64.4) | 452 (66.1) | 442 (64.2) |
Prior IBS-C medication use, n (%) | 39 (14.2) | 82 (22.5) | 102 (24.9) | 71 (22.0) | 141 (20.6) | 153 (22.2) |
Baseline weekly measures, mean (SD) | ||||||
AS3 | 6.93 (1.56) | 6.61 (1.65) | 6.10 (1.61) | 6.22 (1.62) | 6.43 (1.64) | 6.43 (1.64) |
Abdominal pain | 6.73 (1.66) | 6.46 (1.63) | 5.91 (1.60) | 6.05 (1.66) | 6.24 (1.67) | 6.27 (1.65) |
Bloating | 7.14 (1.59) | 6.74 (1.86) | 6.25 (1.79) | 6.37 (1.75) | 6.61 (1.77) | 6.57 (1.82) |
Discomfort | 6.92 (1.59) | 6.62 (1.69) | 6.13 (1.65) | 6.26 (1.64) | 6.45 (1.67) | 6.45 (1.68) |
CSBMs/week | 0.06 (0.24) | 0.08 (0.26) | 0.22 (0.49) | 0.25 (0.50) | 0.16 (0.42) | 0.16 (0.40) |
Straining | n = 182 | n = 255 | n = 314 | n = 246 | n = 496 | n = 501 |
3.78 (0.91) | 3.69 (0.90) | 3.59 (0.91) | 3.59 (0.84) | 3.66 (0.92) | 3.64 (0.87) |
. | No-CSBM subgroupa . | ITT excluding no-CSBM subgroupb . | Pooled ITT analysis set . | |||
---|---|---|---|---|---|---|
tenapanor 50 mg bid, N = 275 . | placebo bid, N = 365 . | tenapanor 50 mg bid, N = 409 . | placebo bid, N = 323 . | tenapanor 50 mg bid, N = 684 . | placebo bid, N = 688 . | |
Age, mean (SD), years | 46.6 (12.5) | 45.9 (13.3) | 45.0 (13.4) | 43.9 (13.5) | 45.7 (13.1) | 45.0 (13.5) |
BMI, mean (SD), kg/m2 | 29.6 (6.6) | 30.0 (6.8) | 30.2 (7.2) | 29.8 (6.8) | 30.0 (7.0) | 29.9 (6.8) |
Duration of IBS-C at baseline, mean (SD), years | n = 275 | n = 362 | n = 407 | n = 322 | n = 682 | n = 684 |
10.9 (11.9) | 11.9 (12.0) | 11.5 (11.4) | 11.3 (11.8) | 11.2 (11.6) | 11.6 (11.9) | |
Female, n (%) | 226 (82.2) | 300 (82.2) | 333 (81.4) | 272 (84.2) | 559 (81.7) | 572 (83.1) |
Race, n (%) | ||||||
American Indian or Alaskan Native | 0 | 0 | 1 (0.2) | 2 (0.6) | 1 (0.1) | 2 (0.3) |
Asian | 16 (5.8) | 7 (1.9) | 8 (2.0) | 7 (2.2) | 24 (3.5) | 14 (2.0) |
Black | 89 (32.4) | 114 (31.2) | 106 (25.9) | 100 (31.0) | 195 (28.5) | 214 (31.1) |
Multiple | 2 (0.7) | 5 (1.4) | 6 (1.5) | 5 (1.5) | 8 (1.2) | 10 (1.5) |
Other | 3 (1.1) | 5 (1.4) | 0 | 1 (0.3) | 3 (0.4) | 6 (0.9) |
White | 165 (60.0) | 234 (64.1) | 287 (70.2) | 208 (64.4) | 452 (66.1) | 442 (64.2) |
Prior IBS-C medication use, n (%) | 39 (14.2) | 82 (22.5) | 102 (24.9) | 71 (22.0) | 141 (20.6) | 153 (22.2) |
Baseline weekly measures, mean (SD) | ||||||
AS3 | 6.93 (1.56) | 6.61 (1.65) | 6.10 (1.61) | 6.22 (1.62) | 6.43 (1.64) | 6.43 (1.64) |
Abdominal pain | 6.73 (1.66) | 6.46 (1.63) | 5.91 (1.60) | 6.05 (1.66) | 6.24 (1.67) | 6.27 (1.65) |
Bloating | 7.14 (1.59) | 6.74 (1.86) | 6.25 (1.79) | 6.37 (1.75) | 6.61 (1.77) | 6.57 (1.82) |
Discomfort | 6.92 (1.59) | 6.62 (1.69) | 6.13 (1.65) | 6.26 (1.64) | 6.45 (1.67) | 6.45 (1.68) |
CSBMs/week | 0.06 (0.24) | 0.08 (0.26) | 0.22 (0.49) | 0.25 (0.50) | 0.16 (0.42) | 0.16 (0.40) |
Straining | n = 182 | n = 255 | n = 314 | n = 246 | n = 496 | n = 501 |
3.78 (0.91) | 3.69 (0.90) | 3.59 (0.91) | 3.59 (0.84) | 3.66 (0.92) | 3.64 (0.87) |
AS3, 3-item abdominal score; bid, twice a day; BMI, body mass index; CSBM, complete spontaneous bowel movement; IBS-C, irritable bowel syndrome with constipation; ITT, intent-to-treat.
aDefined as patients with no CSBMs in ≥6 of the first 12 weeks of treatment.
bDefined as patients who did not meet the no-CSBM definition.
Analysis of Weekly Abdominal Symptoms
In the no-CSBM subgroup, tenapanor-treated patients experienced a greater mean reduction from baseline in AS3 than placebo-treated patients in week 12 (least squares mean [LSM] change, −1.74 vs. −1.29; LSM difference, −0.45; p = 0.007; Fig. 2a). In the ITT excluding no-CSBM subgroup, the mean reduction in AS3 for tenapanor-treated patients was also greater than in placebo-treated patients in week 12 (LSM change, −3.70 vs. −3.19; LSM difference, −0.51; p = 0.003; Fig. 2b).
LSM (±SE) change in AS3 in the no-CSBM subgroup (a) and the ITT excluding no-CSBM subgroup (b). The change from baseline in AS3 was assessed using a MMRM. Each MMRM included treatment, week, and treatment-by-week as fixed effect factors; baseline weekly score and baseline-by-week as fixed effect covariates; and patient as a random effect. An unstructured covariance was employed in each MMRM analysis. AS3, 3-item abdominal score; bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; LSM, least squares mean; MMRM, mixed-effects model with repeated measures; n, the number of patients with a valid week in the corresponding study week; SE, standard error.
LSM (±SE) change in AS3 in the no-CSBM subgroup (a) and the ITT excluding no-CSBM subgroup (b). The change from baseline in AS3 was assessed using a MMRM. Each MMRM included treatment, week, and treatment-by-week as fixed effect factors; baseline weekly score and baseline-by-week as fixed effect covariates; and patient as a random effect. An unstructured covariance was employed in each MMRM analysis. AS3, 3-item abdominal score; bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; LSM, least squares mean; MMRM, mixed-effects model with repeated measures; n, the number of patients with a valid week in the corresponding study week; SE, standard error.
In the no-CSBM subgroup, the weekly AS3 response rate in week 12 was 40.2% in tenapanor-treated patients compared with 29.6% in placebo-treated patients (Fig. 3a); the association between week 12 response status and treatment was statistically significant (p = 0.008). In the ITT excluding the no-CSBM subgroup, the weekly AS3 response rate in week 12 was 74.9% in tenapanor-treated patients compared with 64.9% in placebo-treated patients (Fig. 3b); the association between week 12 response status and treatment was statistically significant (p = 0.015).
Weekly AS3 response rates in the no-CSBM subgroup (a) and the ITT excluding no-CSBM subgroup (b). The weekly AS3 response was defined as achieving a reduction from baseline of ≥2 points in AS3 for a given week. Response rates were analyzed using Pearson’s chi-square tests with a worst-case imputation approach (patients with missing data were considered to not have achieved a response). AS3, 3-item abdominal score; bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; n, the number of patients with a valid week in the corresponding study week.
Weekly AS3 response rates in the no-CSBM subgroup (a) and the ITT excluding no-CSBM subgroup (b). The weekly AS3 response was defined as achieving a reduction from baseline of ≥2 points in AS3 for a given week. Response rates were analyzed using Pearson’s chi-square tests with a worst-case imputation approach (patients with missing data were considered to not have achieved a response). AS3, 3-item abdominal score; bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; n, the number of patients with a valid week in the corresponding study week.
The no-CSBM subgroup also demonstrated a greater mean reduction from baseline in weekly abdominal pain score (LSM change, −1.85 vs. −1.40; LSM difference, −0.45; p = 0.010; Fig. 4a). For the ITT excluding no-CSBM subgroup, the mean reduction in weekly abdominal pain score was also greater for tenapanor-treated patients versus placebo-treated patients in week 12 (LSM change, −3.84 vs. −3.28; LSM difference, −0.56; p = 0.001; Fig. 4a).
LSM (±SE) change in abdominal pain (a), bloating (b), and discomfort scores (c) in the no-CSBM subgroup (left) and the ITT excluding no-CSBM subgroup (right). The change from baseline in abdominal pain score was assessed using a MMRM. Each MMRM included treatment, week, and treatment-by-week as fixed effect factors; baseline weekly score and baseline-by-week as fixed effect covariates; and patient as a random effect. An unstructured covariance was employed in each MMRM analysis. bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; LS, least squares; LSM, least squares mean; MMRM, mixed-effects model with repeated measures; n, the number of patients with a valid week in the corresponding study week; SE, standard error.
LSM (±SE) change in abdominal pain (a), bloating (b), and discomfort scores (c) in the no-CSBM subgroup (left) and the ITT excluding no-CSBM subgroup (right). The change from baseline in abdominal pain score was assessed using a MMRM. Each MMRM included treatment, week, and treatment-by-week as fixed effect factors; baseline weekly score and baseline-by-week as fixed effect covariates; and patient as a random effect. An unstructured covariance was employed in each MMRM analysis. bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; LS, least squares; LSM, least squares mean; MMRM, mixed-effects model with repeated measures; n, the number of patients with a valid week in the corresponding study week; SE, standard error.
The weekly abdominal pain response rate in week 12 was 42.3% in tenapanor-treated patients compared with 32.1% in placebo-treated patients (Fig. 5a); the association between week 12 response status and treatment was statistically significant (p = 0.013). In the ITT excluding no-CSBM subgroup, the weekly abdominal pain response rate in week 12 was 77.5% in tenapanor-treated patients compared with 68.4% in placebo-treated patients (Fig. 5a); the association between week 12 response status and treatment was statistically significant (p = 0.022).
Weekly abdominal pain (a), bloating (b), and discomfort response rates (c) in the no-CSBM subgroup (left) and the ITT excluding no-CSBM subgroup (right). The weekly abdominal pain response was defined as achieving a reduction from baseline of ≥2 points in weekly abdominal pain score for a given week. Response rates were analyzed using Pearson’s chi-square tests with a worst-case imputation approach (patients with missing data were considered to not have achieved a response). bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; n, the number of patients with a valid week in the corresponding study week.
Weekly abdominal pain (a), bloating (b), and discomfort response rates (c) in the no-CSBM subgroup (left) and the ITT excluding no-CSBM subgroup (right). The weekly abdominal pain response was defined as achieving a reduction from baseline of ≥2 points in weekly abdominal pain score for a given week. Response rates were analyzed using Pearson’s chi-square tests with a worst-case imputation approach (patients with missing data were considered to not have achieved a response). bid, twice a day; CSBM, complete spontaneous bowel movement; ITT, intent-to-treat; n, the number of patients with a valid week in the corresponding study week.
There were also reductions from baseline in the weekly bloating and discomfort symptom scores for no-CSBM patients in week 12 (LSM difference −0.45 [p = 0.008] and −0.46 [p = 0.009], respectively) (Fig. 4b, c). For the ITT excluding no-CSBM subgroup, LSM difference between tenapanor and placebo groups was −0.42 (p = 0.026) and −0.55 (p = 0.002) for weekly bloating and discomfort scores, respectively (Fig. 4b, c). In the no-CSBM subgroup, the weekly response rates in week 12 for bloating and discomfort were greater for tenapanor-treated patients compared with placebo-treated patients (37.8% vs. 27.0% [association test p = 0.007] and 41.5% vs. 29.3% [association test p = 0.003], respectively; Fig. 5b, c). In the ITT excluding no-CSBM subgroup, the weekly response rates in week 12 for bloating and discomfort were greater for tenapanor-treated patients compared with placebo-treated patients (70.0% vs. 65.8% [association test p = 0.312] and 75.7% vs. 63.6% [association test p = 0.003], respectively; Fig. 5b, c).
A greater percentage of patients treated with tenapanor versus placebo achieved ≥30% reduction from baseline in weekly AS3 in both the no-CSBM subgroup and the ITT excluding no-CSBM subgroup in week 12 (41.9% vs. 31.5% [association test p = 0.011] and 80.1% vs. 71.0% [association test p = 0.017], respectively) (online suppl. Fig. 3a). Similarly, the percentage of patients achieving ≥30% reduction from baseline in weekly abdominal pain scores in week 12 was 44.3% in the tenapanor group versus 34.3% in the placebo group (association test p = 0.016), and 82.8% versus 72.3% (association test p = 0.005), respectively (online suppl. Fig. 3b). At least 30% reductions in weekly bloating and discomfort scores were also more common in tenapanor- versus placebo-treated patients for both subgroups in week 12 (online suppl. Fig. 3c, d).
Analysis of Weekly Straining
In the no-CSBM subgroup, tenapanor-treated patients experienced a greater mean reduction in weekly straining score than placebo-treated patients in week 12 (LSM change, −1.07 vs. −0.62; LSM difference, −0.45; p < 0.001; online suppl. Fig. 4a). In the ITT excluding no-CSBM subgroup, the mean reduction in weekly straining score for tenapanor-treated patients was also greater than in placebo-treated patients in week 12 (LSM change, −1.75 vs. −1.34; LSM difference, −0.42; p < 0.001; online suppl. Fig. 4b).
Adverse Events
The pooled safety analysis set included 1,382 patients (tenapanor, n = 691; placebo, n = 691). Among the no-CSBM subgroup in the pooled safety analysis set (n = 641), the most common TEAEs were diarrhea (tenapanor, 6.9% vs. placebo, 1.1%), nasopharyngitis (4.0% vs. 3.0%), urinary tract infection (2.5% vs. 2.2%), and flatulence (2.2% vs. 1.9%). In the pooled safety analysis excluding the no-CSBM subgroup (n = 741), the rates of TEAEs were similar to those in no-CSBM patients (Table 2). The incidence of diarrhea with tenapanor in the no-CSBM subgroup was lower than that in the pooled safety analysis excluding the no-CSBM subgroup and in the pooled safety analysis set (6.9% vs. 20% and 14.8%, respectively).
Most common TEAEs (no-CSBM subgroup, safety excluding no-CSBM subgroup, and pooled safety analysis set)
. | No-CSBM subgroupa . | Safety excluding no-CSBM subgroupb . | Pooled safety analysis set . | |||
---|---|---|---|---|---|---|
tenapanor 50 mg bid, N = 275 . | placebo bid, N = 366 . | tenapanor 50 mg bid, N = 416 . | placebo bid, N = 325 . | tenapanor 50 mg bid, N = 691 . | placebo bid, N = 691 . | |
Patients with any TEAE | 103 (37.5) | 125 (34.2) | 195 (46.9) | 111 (34.2) | 298 (43.1) | 236 (34.2) |
TEAE by preferred term reported by >2% of patients | ||||||
Diarrhea | 19 (6.9) | 4 (1.1) | 83 (20.0) | 12 (3.7) | 102 (14.8) | 16 (2.3) |
Nasopharyngitis | 11 (4.0) | 11 (3.0) | 10 (2.4) | 9 (2.8) | 21 (3.0) | 20 (2.9) |
Urinary tract infection | 7 (2.5) | 8 (2.2) | 7 (1.7) | 15 (4.6) | 14 (2.0) | 23 (3.3) |
Nausea | 4 (1.5) | 10 (2.7) | 15 (3.6) | 7 (2.2) | 19 (2.7) | 17 (2.5) |
Headache | 6 (2.2) | 8 (2.2) | 5 (1.2) | 9 (2.8) | 11 (1.6) | 17 (2.5) |
Flatulence | 6 (2.2) | 7 (1.9) | 11 (2.6) | 5 (1.5) | 17 (2.5) | 12 (1.7) |
Vomiting | 4 (1.5) | 9 (2.5) | 6 (1.4) | 4 (1.2) | 10 (1.4) | 13 (1.9) |
Sinusitis | 1 (0.4) | 10 (2.7) | 7 (1.7) | 2 (0.6) | 8 (1.2) | 12 (1.7) |
. | No-CSBM subgroupa . | Safety excluding no-CSBM subgroupb . | Pooled safety analysis set . | |||
---|---|---|---|---|---|---|
tenapanor 50 mg bid, N = 275 . | placebo bid, N = 366 . | tenapanor 50 mg bid, N = 416 . | placebo bid, N = 325 . | tenapanor 50 mg bid, N = 691 . | placebo bid, N = 691 . | |
Patients with any TEAE | 103 (37.5) | 125 (34.2) | 195 (46.9) | 111 (34.2) | 298 (43.1) | 236 (34.2) |
TEAE by preferred term reported by >2% of patients | ||||||
Diarrhea | 19 (6.9) | 4 (1.1) | 83 (20.0) | 12 (3.7) | 102 (14.8) | 16 (2.3) |
Nasopharyngitis | 11 (4.0) | 11 (3.0) | 10 (2.4) | 9 (2.8) | 21 (3.0) | 20 (2.9) |
Urinary tract infection | 7 (2.5) | 8 (2.2) | 7 (1.7) | 15 (4.6) | 14 (2.0) | 23 (3.3) |
Nausea | 4 (1.5) | 10 (2.7) | 15 (3.6) | 7 (2.2) | 19 (2.7) | 17 (2.5) |
Headache | 6 (2.2) | 8 (2.2) | 5 (1.2) | 9 (2.8) | 11 (1.6) | 17 (2.5) |
Flatulence | 6 (2.2) | 7 (1.9) | 11 (2.6) | 5 (1.5) | 17 (2.5) | 12 (1.7) |
Vomiting | 4 (1.5) | 9 (2.5) | 6 (1.4) | 4 (1.2) | 10 (1.4) | 13 (1.9) |
Sinusitis | 1 (0.4) | 10 (2.7) | 7 (1.7) | 2 (0.6) | 8 (1.2) | 12 (1.7) |
Data are n (%).
AS3, 3-item abdominal score; bid, twice a day; CSBM, complete spontaneous bowel movement; SD, standard deviation; TEAE, treatment-emergent adverse event.
aDefined as patients with no CSBMs in ≥6 of the first 12 weeks of treatment.
bDefined as patients not meeting the no-CSBM definition.
Discussion
In this post hoc analysis, we demonstrated that tenapanor improves abdominal symptoms irrespective of changes in CSBM frequency in adults with IBS-C. To assess clinically meaningful improvement in three key abdominal symptoms – pain, bloating, and discomfort – the AS3 was utilized. In the subgroup of patients with IBS-C who had zero CSBMs in ≥6 of the first 12 weeks of treatment, tenapanor treatment improved AS3. The mean reduction from baseline in AS3 was greater in week 12 for tenapanor-treated patients compared with placebo-treated patients in the no-CSBM subgroup. Furthermore, a higher percentage of tenapanor-treated patients achieved ≥2-point improvement in AS3 compared with placebo-treated patients. The mean change from baseline and responder rates in each of the individual weekly abdominal pain, bloating, and discomfort scores also showed clinically meaningful benefit in each week for tenapanor-treated patients compared with placebo-treated patients in the no-CSBM subgroup.
Although improvements in the no-CSBM subgroup in AS3 and each of the individual symptom scores were comparable to improvements observed in the ITT excluding no-CSBM subgroup, patients in the ITT excluding no-CSBM subgroup trended toward greater changes from baseline in abdominal symptoms and the separation between placebo and tenapanor was numerically greater in this subgroup. This observation suggests there is a synergistic effect of bowel movement response on pain, bloating, and discomfort; however, patients with zero CSBMs in ≥6 of the first 12 weeks of treatment can still have an AS3 response comparable to that of a broader IBS-C patient population, including responders.
When making a diagnosis of IBS-C, the presence of defecatory dysfunction should always be taken into consideration. To provide a more comprehensive assessment of symptoms of evacuatory function, we assessed changes in straining using pooled weekly straining scores. Results similar to those for abdominal symptoms were observed, demonstrating that tenapanor improves straining irrespective of changes in CSBM frequency.
Tenapanor demonstrated a consistent safety profile in phase 2 and 3 studies of patients with IBS-C [28‒30]. The most common adverse event in the no-CSBM subgroup was diarrhea, occurring in 6.9% of tenapanor-treated patients, which is consistent with previous studies [28‒30]. Diarrhea associated with tenapanor treatment is predominately mild to moderate in severity and generally short-lived, lasting up to 1 week [32]. The occurrence of diarrhea is also consistent with tenapanor’s mechanism of action since inhibition of NHE3 increases the excretion of sodium and fluid in stool [23, 25]. It should be noted that patients in the no-CSBM subgroup did not experience a complete absence of CSBMs, and the occurrence of diarrhea in the no-CSBM subgroup indicates that patients in this subgroup were still experiencing bowel evacuation during the treatment period. Therefore, some of the improvement in this subgroup’s abdominal symptoms may relate to the effect of tenapanor on patients’ bowel function.
Some IBS-C treatments improve constipation but have not demonstrated significant effects on abdominal symptoms [15, 17, 18]. Tenapanor was previously shown to alleviate both constipation and abdominal pain in adults with IBS-C [28‒30]. The findings of this post hoc analysis may reflect favorable effects of tenapanor on intestinal permeability and visceral hypersensitivity [26], as previously demonstrated in nonclinical studies. In studies using human colonic epithelial monolayers, tenapanor reduced intestinal cell permeability to macromolecules, and in animal models of visceral hypersensitivity, tenapanor normalized visceromotor responses [33]. These mechanisms are separate from tenapanor’s effect on bowel function and may explain improvement in abdominal symptoms independent of improvements in stool frequency/consistency.
A limitation of this post hoc analysis as with other IBS-C clinical trials is that the study population may not be entirely representative of real-world disease populations due to the stringent criteria used to enroll patients with IBS-C in the clinical trials [34‒36]. Interestingly, 14.2% of tenapanor-treated patients in the no-CSBM subgroup reported prior use of IBS-C medication, compared with 22.5% of those receiving placebo. In the ITT excluding no-CSBM subgroup, 24.9% of tenapanor-treated patients and 22.0% placebo-treated patients reported prior use of IBS-C medication. This might suggest that in these studies, tenapanor-treated patients in the no-CSBM subgroup were less treatment resistant than placebo-treated patients or patients in the ITT excluding no-CSBM subgroup. Additional prospective studies are needed to confirm the effect of tenapanor on abdominal symptoms in patients with IBS-C and examine long-term effects of tenapanor on abdominal symptoms. We did not evaluate quality of life in the no-CSBM subgroups that were the focus of this post hoc analysis. However, in the T3MPO-2 trial, tenapanor-treated patients demonstrated significantly greater improvements in overall health-related quality of life from baseline to week 26 versus placebo-treated patients [30].
In conclusion, this post hoc analysis demonstrates that tenapanor is effective in reducing both abdominal and bowel symptoms in adults with IBS-C. The decrease in abdominal pain and associated abdominal symptoms in these clinical trials appears to be separate from bowel symptoms and may be attributable to tenapanor’s additional mechanisms of action.
Acknowledgments
We would like to thank the patients and investigators who participated in the study. Medical writing support, under the direction of the authors, was provided by Ashfield MedComms, an Inizio company.
Statement of Ethics
The phase 2b study was approved by the Copernicus Group Institutional Review Board (#ARD1-13-256). T3MPO-1 was approved by Quorum Institutional Review Board (#30954). T3MPO-2 was approved by Quorum Institutional Review Board (#31080). All patients provided written informed consent before entering the phase 2b (NCT01923428), T3MPO-1 (NCT02621892), and T3MPO-2 (NCT02686138) studies, which were conducted in accordance with the Declaration of Helsinki, Good Clinical Practice Guidelines, and all applicable local laws and regulations.
Conflict of Interest Statement
Darren M. Brenner is a consultant, advisor, and/or speaker for Anji, Ardelyx, AbbVie, Alnylam, Salix, Ironwood, Takeda, Bayer, Gemelli, Redhill, Mahana, Laborie, Owlstone, Entrinsic Bioscience, Vibrant, CinPhloro. Dr. Reddy is a member of the Board of Directors of the International Foundation for GI Disorders (IFFGD). Gregory S. Sayuk is a consultant and speaker for Salix Pharmaceuticals, AbbVie/Ironwood Pharmaceuticals, Sanofi/Regeneron, Phathom, and Ardelyx, Inc., and is a consultant for the GI Health Foundation and Rome Foundation. Brooks D. Cash is a consultant and speaker for AbbVie, Inc, Alnylam Pharmaceuticals, Inc, Ardelyx, QOL Medical, Phathom Pharmaceuticals, Takeda Pharmaceuticals, and Salix Pharmaceuticals and has received research support from Napo Pharmaceuticals, Inc. Lucinda A. Harris has received financial support for consulting from Ardelyx, Inc., AbbVie, Alnylam, Ironwood, Gemelli Biotech, and Salix, and research support from Takeda. Nitin K. Ahuja has served as a consultant for Ardelyx and Haleon and received research support from Vanda Pharmaceuticals. Jill K. Deutsch has no relevant disclosures. Yang Yang, Suling Zhao, and David P. Rosenbaum are employees of Ardelyx, Inc. Anthony J. Lembo is a consultant for Allergan, Ardelyx, Atmo, Allakos, BioAmerica, AEON, Arena, Takeda, Evoke Pharma, Ironwood Pharmaceuticals, Gemelli, Alkermes, Pfizer, OrthoMed, and Vibrant and has stock with Johnson & Johnson, Bristol Myers Squibb, and Allurion.
Funding Sources
The study and medical writing support were funded by Ardelyx, Inc.
Author Contributions
Darren M. Brenner, Gregory S. Sayuk, Brooks D. Cash, Lucinda A. Harris, Nitin K. Ahuja, Jill K. Deutsch, and Anthony J. Lembo were responsible for data curation and investigation. David P. Rosenbaum, Yang Yang, and Suling Zhao conceptualized the study. Yang Yang and Suling Zhao were responsible for the formal analysis and reviewed the manuscript critically for statistical content. Darren M. Brenner, Gregory S. Sayuk, Brooks D. Cash, Lucinda A. Harris, Nitin K. Ahuja, Jill K. Deutsch, David P. Rosenbaum, and Anthony J. Lembo contributed to the critical revision of the manuscript for important intellectual content. All authors approved the final version of the manuscript for submission.
Additional Information
Clinical Trial Registrations. NCT01923428: The Efficacy of AZD1722 in Constipation Predominant Irritable Bowel Syndrome (IBS-C) was registered on July 14, 2013. NCT02621892: A 12-Week Study With a 4-Week Randomized Withdrawal Period to Evaluate the Efficacy and Safety of Tenapanor for the Treatment of IBS-C (T3MPO-1) was registered on July 2, 2015. NCT02686138: A 26-Week Study to Evaluate the Efficacy and Safety of Tenapanor in IBS-C (T3MPO-2) was registered on February 15, 2016.
Data Availability Statement
Ardelyx will consider reasonable requests for data sharing such as the study protocol, SAP, and ICF on a case-by-case basis based on data availability, burden, and data privacy issues. This will go into effect immediately after publication for a period of up to 1 year. Data will be shared to achieve aims in an investigator-submitted proposal, which has been approved by Ardelyx. Proposals should be directed to [email protected]. To gain access to data, requestors will need to sign a data access agreement.