: No currently available treatment provides consistent relief of irritable bowel syndrome. Colonic sensory and motor function are modulated partly through 5HT3-receptors.
: No currently available treatment provides consistent relief of irritable bowel syndrome. Colonic sensory and motor function are modulated partly through 5HT3-receptors.
: To evaluate effects of the 5HT3-receptor antagonist, alosetron, in irritable bowel syndrome.
: Randomized, double-blind, placebo-controlled, dose-ranging (1, 2, 4, 8 mg b.d. alosetron), 12-week trial in 370 patients with diarrhoea-predominant or alternating constipation and diarrhoea irritable bowel syndrome. Weekly measurement of adequate relief was the key end-point; other irritable bowel syndrome symptoms were collected daily using an electronic phone system.
: Alosetron (1 mg or 2 mg b.d.) significantly (P < 0.05 vs. placebo) increased the proportion of females, but not males, reporting adequate relief. Stool consistency, frequency and percentage days with urgency improved over placebo (P < 0.05) within the first month with all doses of alosetron, and persisted throughout the trial with all doses in female patients. With 1 mg b.d. alosetron, females had improved stool consistency and urgency within the first week, and adequate relief and improved stool frequency within the first 2 weeks. There was no consistent improvement in bowel function among male patients.
: In female irritable bowel syndrome patients with predominant diarrhoea or alternating constipation and diarrhoea, alosetron is effective in treatment of abdominal pain and discomfort and bowel-related symptoms.
Irritable bowel syndrome is one of the most common gastrointestinal disorders in Western populations, with an adult prevalence of approximately 15%.1, 2 The cardinal features of irritable bowel syndrome are recurrent abdominal pain and discomfort, altered bowel habits, and a strong female predominance of 70%–75%.1–4 Although the reason for this female predominance is not clear, it is consistent with gender-based differences in the reporting of other gastrointestinal diseases.5 Women also appear to have more classical irritable bowel syndrome symptoms. For example, Smith et al.6 and Thompson7 have shown greater diagnostic utility of the Manning criteria in female irritable bowel syndrome patients compared with male patients. There may also be physiologic differences that predispose females to irritable bowel syndrome and these may represent important cofactors that modify responses to pharmacologic therapy. Using positron emission tomography, different patterns of cerebral blood flow have been reported in female compared with male irritable bowel syndrome patients.8 Other studies9–11 with non-irritable bowel syndrome patients have also documented gender differences in colonic motor function. Thus, it is important to evaluate therapeutic efficacy of novel agents for irritable bowel syndrome with respect to the potential differences in responses in males and females.
Over the last two decades, there has been compelling evidence that enhanced perception of visceral stimuli develops in irritable bowel syndrome patients.2, 3, 12–15 In balloon distention studies of the colon or rectum the threshold for sensation of pain is lower in irritable bowel syndrome patients compared with controls, and this has been proposed as a biological marker for irritable bowel syndrome.15 In view of the evidence for enhanced visceral perception in irritable bowel syndrome, and the frequent occurrence of pain as a key symptom, it is generally accepted that any agent considered to be of utility in the treatment of irritable bowel syndrome should demonstrate effectiveness in the relief of pain.
Of the classes of therapeutic agents that have been proposed for the treatment of abdominal pain in irritable bowel syndrome, several lines of evidence support a potential role for 5HT3 antagonists. In animal models, these agents have been shown to decrease visceral pain responses.16, 17 Furthermore, the 5HT3-receptor antagonist, ondansetron, has been shown to slow colonic transit in normal volunteers.18, 19 In diarrhoea-predominant irritable bowel syndrome patients, ondansetron increases rectal compliance and improves stool consistency.20–23 Ondansetron also inhibits the contractile response of the colon in healthy volunteers in the early postprandial period,24 the time when many irritable bowel syndrome patients experience symptoms such as urgency, cramping or diarrhoea. A second 5HT3-receptor antagonist, granisetron, has also been shown to decrease rectal sensitivity and to reduce postprandial rectal motor activity in irritable bowel syndrome patients.25 Thus, a rational basis exists for evaluation of 5HT3 antagonists in irritable bowel syndrome.
Alosetron is a potent and selective 5HT3-receptor antagonist and, in preliminary reports, it has been shown to improve abdominal pain26 and to slow colonic transit in irritable bowel syndrome patients.27 The purpose of this study was to evaluate the efficacy of alosetron in the treatment of irritable bowel syndrome in a large, multinational, randomized, double-blind, dose-ranging trial comparing four doses of alosetron with placebo. The key end-points of the study were adequate relief of irritable bowel syndrome pain and discomfort and changes in bowel functions.
Three hundred and seventy irritable bowel syndrome patients were randomized from 68 sites located within the United States (54 sites; 315 patients), United Kingdom (four sites; 14 patients), Canada (four sites; 26 patients), Germany (two sites; eight patients) and the Netherlands (four sites; seven patients). Patients were required to have symptoms which fulfilled the Rome Criteria for irritable bowel syndrome28 for at least 6 months. Because of the ability of 5HT3-receptor antagonists to slow colonic transit,18, 19 constipation-predominant irritable bowel syndrome patients were excluded from this study, and only patients with diarrhoea-predominant irritable bowel syndrome or alternating constipation and diarrhoea were included.
General inclusion criteria were: ability to give informed consent; at least 18 years of age; ambulatory out-patient; if female, following precautions to avoid pregnancy; and normal structural evaluation of the colon within the previous 2 years. General exclusion criteria were: unstable status of the cardiovascular, renal, hepatic, pulmonary, endocrine, metabolic, or haematologic systems; other gastrointestinal conditions; major psychiatric disorder or substance abuse within the previous 2 years; AST or ALT greater than 2.5 times the upper limit of normal; serum creatinine greater than 2.0 mg/dL; hyper- or hypothyroid; evidence of nonskin malignancy within the previous 5 years; use of an investigational drug within 30 days of the screening phase; pregnant or breast feeding; or use of a prohibited drug.
Daily and weekly symptom data were collected using a recently described electronic touch-tone phone based system.29 Patients underwent a 2-week screening period with no treatment for irritable bowel syndrome to ensure a sufficient baseline level of abdominal pain, as well as to document compliance with the data collection system. Pain was assessed daily on a 5 point scale (0, none; 1, mild; 2, moderate; 3, intense; 4, severe). Average baseline pain over the 2-week screening period was required to be between 1.5 and 3.3, inclusive, and at least 4 days with at least moderate pain was required for enrollment into the study. Stool consistency data were monitored daily (1, very hard; 2, hard; 3, formed; 4, loose; 5, watery). A score of 0 was assigned on days with no stool. During the screening period an average stool consistency score of 2.5 was required for entry into the study in order to exclude those with predominant constipation.
Following the screening period, eligible patients were randomized with equal allocation to 12 weeks of b.d. treatment with placebo or alosetron 1, 2, 4 or 8 mg taken prior to meals. Patients were followed for 2 weeks after the 12 weeks of treatment. During the screening period, treatment phase and follow-up period, patients were questioned daily about their irritable bowel syndrome symptoms. Once every 7 days during the treatment phase of the study, patients responded to an additional question as to whether they had obtained adequate relief of their irritable bowel syndrome pain and discomfort (adequate relief, hereafter) during the previous 7 days.
This was a randomized, double-blind, dose-ranging, parallel-group study comparing four doses of alosetron with placebo. The planned sample size of 350 male and female patients with 70 patients per treatment group provided 80% power to detect a 25% difference in the proportions between placebo and any dose of alosetron (e.g. 40% vs. 65%) at alpha=0.05 significance level (unadjusted for multiple comparisons).
Improvement in adequate relief with alosetron was the key end-point in this study. As part of the prospectively planned sub-group analyses and because of the aforementioned gender-based differences in irritable bowel syndrome, efficacy of alosetron was evaluated separately in females and males. For patients who completed all 12 weeks of the treatment phase, a responder was prospectively defined as a patient who responded positively to the weekly question of whether he/she had obtained adequate relief for at least 6 of the weeks; otherwise the patient was defined as a nonresponder. The adequate relief response has been shown to be strongly associated with improvement in abdominal pain, bowel function and quality of life.30 Among adequate relief responders, the strongest associations were with improvement in abdominal pain and urgency to defecate. The proportion of patients with adequate relief at each week and the proportion of responders for adequate relief were compared between treatment groups using a Mantel–Haenszel test stratified by centres to provide covariate adjustments for centres in the analyses.31 In addition, approximate (unstratified) 95% confidence intervals were presented for the treatment difference for the proportion of responders for adequate relief.
Daily stool consistency scores and daily number of bowel movements were averaged at baseline (for the 2-week screening period), for each week (for weeks 1–4 only), and monthly (i.e. month 1=weeks 1–4, month 2=weeks 5–8, month 3=weeks 9–12). Data were analysed during the first 4 weeks to identify the time taken for significant improvement in bowel functions. In addition, the percentage of days that patients experienced a sense of urgency was calculated for the same weekly and monthly intervals. For the weekly intervals, treatment groups were compared using van Elteren’s test32 stratified by centres. For the monthly intervals, treatment groups were compared for change from baseline using van Elteren’s test stratified by centres.
Analyses of bowel-related functions were performed with observed data.
Eight hundred and thirty-five patients were screened for eligibility to enter the treatment phase of the study. Four hundred and sixty-five patients failed screening and were ineligible for entry into the treatment phase. The predominant cause of failure in screening was insufficient pain score. Three hundred and seventy patients (44%) satisfied the entry criteria and were randomized to treatment. Eighty patients were randomized to treatment with placebo b.d., 72–1 mg b.d. alosetron, 74–2 mg b.d. alosetron, 76–4 mg b.d. alosetron, and 68–8 mg b.d. alosetron.
Sixty-eight of the randomized patients (13 in the placebo, 18 in the 1-mg, 14 in the 2-mg, 12 in the 4-mg, and 11 in the 8-mg group) did not meet the screening criteria for stool consistency (i.e. >2.5) due to a program miscalculation; these patients were excluded from the per protocol analysis but were included in the intention-to-treat (ITT) analysis.
No treatment by centre interactions was observed for any of the placebo vs. alosetron treatment group comparisons (all P values >0.14).
Examination of each dose of alosetron showed a greater percentage of responders for adequate relief compared with placebo in female irritable bowel syndrome patients (Tables 3 and 4). In the ITT population, 1 mg b.d. alosetron produced a 27% greater increase in adequate relief responders in female patients than was seen with placebo (P < 0.05). As shown in Tables 3 and 4, in female patients in the ITT population, all doses of alosetron produced substantial benefit over that seen with placebo. By contrast, in males, no substantial increase was seen with any dose of alosetron.
The per protocol evaluation of the female patients showed that the largest treatment effect was also observed with 1 mg b.d. alosetron. In this treatment group, there were 40% more responders compared with placebo (30% placebo vs. 70% alosetron; P=0.002; Figure 1A; Tables 3 and 4). A similar result was observed with 2 mg b.d. alosetron which showed 63% responders (P=0.007). With 4 mg b.d. and 8 mg b.d. alosetron, the proportions of adequate relief responders were 23% (P=0.07) and 20% (P=0.12) greater than seen with placebo, respectively. Similar efficacy results were observed with 1 mg b.d. alosetron in diarrhoea-predominant irritable bowel syndrome patients and those with alternating bowel function.
There was no improvement relative to placebo in the male irritable bowel syndrome patients with any dose of alosetron (Figure 1B; Tables 3 and 4), either in the ITT or in the per protocol analysis. Similarly, no improvement in bloating was observed in males or females with any dose of alosetron.
Weekly analysis of adequate relief was based on the per protocol cohort. Figure 2 shows the weekly analysis of the percentage of female patients with adequate relief during treatment with alosetron (1 mg b.d.) and placebo. Improvement was achieved after the second week of treatment with alosetron and persisted through week 12 of the study. In female patients, 1 mg b.d. alosetron produced greater benefit than higher alosetron doses on the weekly end-point, and no substantial improvement was seen with any dose of alosetron in males. In Figure 3, we illustrate the weekly adequate relief analyses of placebo-treated male and female patients. In contrast to the ‘responder’ definition of adequate relief (based on positive response in at least 6 of the 12 weeks), the weekly adequate relief data show that the placebo response for male patients was not greater than that for female patients.
In female patients, all doses of alosetron significantly decreased the percentage days with urgency (Figure 4), hardened stool consistency (Figure 5) and decreased stool frequency (Figure 6) compared with placebo. In Figures 45–6, the effects of alosetron on each of these parameters at baseline and during months 1, 2 or 3 of treatment are shown. For each of these bowel-related functions, statistically significant benefit over placebo was achieved within the first month of treatment. With 1 mg b.d. alosetron in female patients, significant improvement was observed within the first week of treatment for stool consistency and for the percentage days with urgency; similarly, stool frequency was significantly decreased within the first 2 weeks of treatment. Among males, no consistent improvement over placebo was seen in any bowel-related function. Evaluation of the placebo response for bowel functions in males and females showed near identical placebo responses for stool consistency and frequency.
Eighty-nine of the 370 patients withdrew prematurely from the study: 12 from the placebo group; 15 from the 1 mg b.d. alosetron; 22 from 2 mg b.d. alosetron; 20 from 4 mg b.d. alosetron; and 20 from 8 mg b.d. alosetron. The most common reason for withdrawal from all alosetron treatment arms was the development of constipation. Constipation was reported in five (6%) of the placebo-treated patients, and in 14 (20%), 19 (26%), 15 (20%) and 20 (29%) of patients receiving treatment with 1, 2, 4 or 8 mg b.d. alosetron, respectively. The next most common adverse event reported was headache, which occurred in 11%, 10%, 7% or 13% of patients taking 1, 2, 4 or 8 mg b.d. alosetron, respectively. However, a greater frequency of headache was noted with placebo (20%) than with any dose of alosetron. No other adverse event was noted to occur at a frequency greater than 10%.
Because of the marked difference in efficacy with alosetron between males and females, adverse events were also analysed by gender and are shown in Table 5. Constipation was noted to occur more frequently in female as compared with male patients, a finding consistent with the previously described preferential efficacy of alosetron on improvement of bowel function in female as compared with male patients.
The results of the present study show that the 5HT3-receptor antagonist, alosetron, produced improvement in both pain and bowel-related functions in female irritable bowel syndrome patients. Improvement in abdominal pain was assessed by the percentage of patients with adequate relief, an end-point that is strongly associated with improvement in multiple dimensions of irritable bowel syndrome, including pain, bowel functions and quality of life.30 Improvements were seen within the first 2 weeks of treatment for both pain and bowel functions and improvement persisted through week 12 of treatment in female patients.
The lack of effect of alosetron in male patients with irritable bowel syndrome is intriguing. The placebo response rate for the primary end-point, adequate relief for at least 6 weeks, was higher in males than females; however, there was no difference by gender in the weekly analysis of adequate relief, and the proportion of males with placebo-related adequate relief response merely reflects the number fulfilling the a priori definition. A virtually identical proportion of males responded to three of the four doses of alosetron. It is unclear whether the lack of response to alosetron in males is due to the relatively small sample size. However, the current data suggest that, using these end-points, one would require a very large sample size to assess the efficacy of alosetron in males with irritable bowel syndrome. The placebo response rate for adequate relief of pain/discomfort among males was greater than that among females with irritable bowel syndrome. However, the percentages of responders to placebo were within the range of proportions of overall relief responders reported in previous studies (range 13%–69%, mean 43%), as summarized in a recent review.2
In the present study, 1 mg b.d. alosetron was more effective in relieving abdominal pain in females than the higher doses of alosetron. In a preliminary report,23 2 mg b.d. alosetron was a more efficacious dose in comparison with either 0.1 mg b.d. or 0.5 mg b.d. alosetron. It appears that 1 mg b.d. alosetron is the best dose in the treatment of irritable bowel syndrome, as efficacy in the current study with 1 mg b.d. was greater than with 2 mg b.d. The observed lack of a strict dose–response relationship with 5HT3-receptor antagonists in irritable bowel syndrome has also been observed in animal models of visceral pain,16 but the precise reason for the lack of the dose–response relationship is not clear.
Improvement in irritable bowel syndrome symptoms with alosetron treatment was anticipated based upon the ability of 5HT3-receptor antagonists to modulate both visceral sensory and motor processes in the gastrointestinal tract. In animal models of visceral pain, the 5HT3-receptor antagonists granisetron16 and alosetron17 reduce vasoactive reflexes induced by distension of the gastrointestinal tract. In patients with irritable bowel syndrome, granisetron also produces a decrease in rectal sensitivity and motility.25 Similarly, ondansetron has been shown to affect both bowel-related parameters and pain in irritable bowel syndrome patients.21–23 Preliminary studies26, 27 have also suggested motor and sensory effects of alosetron in irritable bowel syndrome patients. Although many specifics of the precise role of 5HT3-receptors in the pain pathway have not been completely elucidated, 5HT3-receptors are present in both the peripheral and the central nervous systems,33, 34 and may impact upon the conduction, processing and/or modulation of the pain signal via activation of depolarizing currents in neurones.35 In the present study, it appears that the 8-mg dose of alosetron may be more effective in increasing stool consistency and reducing stool frequency. However, 8 mg alosetron was less effective in improving urgency or adequate relief of pain/discomfort. This observation suggests that the inhibitory secretomotor effects of such agents may not alone provide adequate relief in nonconstipated irritable bowel syndrome patients.
As a class effect, it is recognized that 5HT3-receptor antagonists slow colonic transit,18, 19 leading to a firming of stool consistency and decrease in stool frequency. Constipation-predominant irritable bowel syndrome patients were therefore excluded from the present study. Hence, a decrease in stool frequency and hardening of stool consistency indicate an improvement in bowel function. In female irritable bowel syndrome patients alosetron produced a significant reduction in the percentage of days with urgency, hardening of stool consistency and decrease in stool frequency. This study did not address the mechanisms by which 5HT3 antagonists may modulate gastrointestinal motor activity. However, two of the potential sites of action of alosetron include involvement of 5HT3-receptors in the modulation of the peristaltic36, 37 or in the regulation of the release of motilin.38
Analysis of the short-term safety profile of alosetron shows that alosetron is safe and well tolerated. The only side-effect of note was constipation, which represents a class effect with 5HT3-receptor antagonists, and was observed at a greater frequency in female as compared with male patients. This latter observation is consistent with the preferential efficacy of alosetron in improving bowel function in female irritable bowel syndrome patients. In summary, based upon the results of the present study, alosetron represents an effective and safe therapy for nonconstipated female irritable bowel syndrome patients; conversely, alosetron was ineffective in the cohort of male patients studied.
This study was funded by Glaxo-Wellcome. Drs Camilleri, Mayer and Drossman were consultants for Glaxo-Wellcome for this study.
We gratefully acknowledge the participating investigators:
United States: Aven, A; Baerg, R; Ballard, D; Beckman, J; Bedard, C; Bianchi, T; Bird, P; Bowles, M; Caldwell, J; Camilleri, M; Chen, S; Chen, Y; Collins, D; Fidelholtz, J; Fisher, R; Fulp, S; Gottesman, A; Graff, A; Grossman, M; Harris, S; Holderman, W; Holloway, R; Holmes, J; Kogut, D; Kohagen, K; Krause, R; Kurtz, M; Lebovitz, P; McDonald, T; Medoff, J; Pruitt, R; Rashid, S; Reddy, A; Reindollar, R; Robison, M; Silvers, D; Sperling, M; Sprague, R; Strain, M; Tepper, R; Wagner, J; Wilborn, S; Winston, B; Wright, R; Zuckerman, M.
Canada: Bailey, R; Dallaire, C; Davies, R; Pare, R; Pellicano, A.
United Kingdom: Gunn, s.d.; Westaby, D; Wilson, JA; Whorwell, PJ.
Holland: Dekkers, CPM; Engels, LJGB; Smout, AJPM; Stronkhorst, A.
Germany: Plass, H; Schuetz, E.
We gratefully acknowledge significant input to the project by R.S.B. Ehsanullah, Ph.D.; L.R. Hamm, R.N.; L. Jacques, Ph.D.; J.P.Harding, S. G. Wolfe, M.S.