Irritable bowel syndrome is a common gastrointestinal disorder characterized by abdominal pain and discomfort and altered bowel habit. Antagonism at the 5-HT3 receptor may be of benefit in the treatment of irritable bowel syndrome.
Irritable bowel syndrome is a common gastrointestinal disorder characterized by abdominal pain and discomfort and altered bowel habit. Antagonism at the 5-HT3 receptor may be of benefit in the treatment of irritable bowel syndrome.
To evaluate the effect of 12 weeks of treatment with alosetron, a 5-HT3 receptor antagonist at doses of 0.1 mg b.d., 0.5 mg b.d. and 2 mg b.d. in irritable bowel syndrome patients.
A double-blind, placebo-controlled, parallel-group study with a 2-week screening and a 12-week treatment period was conducted. A total of 462 patients (335 female) recorded details of the severity of their abdominal pain, and bowel function daily on a diary card throughout the study. At monthly clinic visits patients recorded the severity of their abdominal pain/discomfort and diarrhoea on a visual analogue scale.
In the total population and in the female subpopulation (but not in males) alosetron 2 mg b.d. significantly increased the proportion of pain-free days and decreased the visual analogue scale score for diarrhoea compared with placebo. Alosetron at doses of 0.5 mg b.d. and 2 mg b.d. led to a significant hardening of stool, and a reduction in stool frequency in the total population.
Alosetron at a dose of 2 mg b.d. is an effective treatment for female patients with irritable bowel syndrome.
Irritable bowel syndrome is a common chronic gastrointestinal disorder characterized by lower abdominal pain or discomfort associated with altered bowel habit. It is estimated that between 8% and 19% of the adult western populations have symptoms compatible with irritable bowel syndrome.1, 2 There is a strong female predominance with 70–75% of irritable bowel syndrome patients being female.3–5 Although the pathophysiology of irritable bowel syndrome is poorly understood, abnormalities in either motor function or sensory function (visceral hypersensitivity) may contribute to its pathogenesis.6
The multiple symptoms that are so characteristic of irritable bowel syndrome have proven to be a hurdle in the management of the condition. Current therapies are of limited efficacy as they are aimed at alleviating the most predominant symptom and therefore are only moderately effective overall.7, 8 Usual treatments include increased dietary fibre for constipation, loperamide for diarrhoea and antispasmodics or low dose antidepressants for pain.9 However, there are few, if any, rigorous therapeutics trials of any of these agents.
Serotonin (5-hydroxytryptamine [5-HT]) is an important transmitter within the enteric and autonomic nervous system. Many subtypes of the 5-HT receptor have been identified, with most actions of serotonin in the gastrointestinal tract appearing to be mediated by 5-HT1, 5-HT3 and 5-HT4 receptors. 5-HT3 receptor antagonists increase sensory thresholds to balloon distension of the rectum, either by a direct effect on afferent pain perception, or via an increase in rectal compliance.10–13 In addition, 5-HT3 receptor antagonists have been shown to slow colonic transit in man.14, 15 This pharmacological profile suggests potential efficacy in the treatment of irritable bowel syndrome.
Alosetron is a potent, highly selective 5-HT3 receptor antagonist.16 The present study was dose-ranging and exploratory in nature and was designed to evaluate the potential use of alosetron in alleviating irritable bowel syndrome associated abdominal pain/discomfort and in normalizing bowel function in irritable bowel syndrome patients. In addition the effect of gender on treatment response was also assessed.
Patients, aged 18–74 years, were recruited from 43 centres (including hospital out-patient clinics, primary care practice and private physician clinics) in nine countries. Patients were required to have at least a 6-month history of irritable bowel syndrome with abdominal pain or discomfort occurring on at least 3 days in each of the 4 weeks prior to the study. Irritable bowel syndrome was diagnosed in accordance with the Rome criteria.17
Inclusion criteria were: ability to give informed consent; normal structural evaluation of the colon within 6 weeks of the screening visit; female patients were required to use adequate contraceptive precautions. Exclusion criteria included significant gastrointestinal disorders other than irritable bowel syndrome; severe constipation (defined as one or less bowel movement per week); symptoms related to or exacerbated by milk products or by any known specific food intolerance; presence of unstable cardiovascular, pulmonary, metabolic, haematological, endocrine disorder; recent history of drug or alcohol abuse; AST/ALT greater than twice the upper limit of normal; serum creatinine > 140 μmol/L; significantly raised erythrocyte sedimentation rate; use of an investigational drug within 1 month of the screening visit; pregnancy or breast feeding; use of a prohibited drug; for patients presenting with diarrhoea, evidence of parasitic or bacterial infection on stool examination. All drug therapy for irritable bowel syndrome was halted before the start of the run-in period, with the exception of regular use of bulking agents which patients were allowed to continue. Other prohibited medications included chronic use of drugs that may modify gastrointestinal function such as prokinetic agents, antispasmodics, or anticholinergics.
This was a randomized, double-blind, dose-ranging parallel-group study. Patients who met the inclusion/exclusion criteria entered a 2 week screening phase during which they completed a daily diary card recording severity of abdominal pain (on a four point scale 0 = none, 1 = mild, 2 = moderate, 3 = severe), number of bowel movements and stool consistency (on a six point scale 0 = no stool, 1 = hard, 2 = firm, 3 = soft, 4 = loose, 5 = watery). Patients were then randomized according to a predetermined randomization schedule at the week 0 visit to receive either alosetron 0.1 mg, 0.5 mg, 2 mg or placebo twice daily. The patients were randomly assigned with equal allocation to the treatment groups as determined by the treatment number. Treatment numbers were assigned consecutively in ascending order, starting with the lowest available treatment number. The block size was 12. Screening diary card data were not used to determine eligibility but were used as baseline data for the analysis. Patients received study medication for 12 weeks during which they continued to complete daily diary cards. Patients attended the clinic at the start of screening, at the randomization visit, after 2, 4, 8, and 12 weeks of treatment, and at 14–21 days after stopping treatment. No other treatment for irritable bowel syndrome was permitted throughout the study. At the randomization visit and at the clinic visits after 4, 8, and 12 weeks of treatment, patients were asked to complete questions relating to irritable bowel syndrome symptoms such as abdominal pain/discomfort, diarrhoea and constipation. Questions took the form of a 100-mm visual analogue scale censored as not at all/worst possible for symptoms.
Details of adverse events were recorded at each clinic visit. Blood and urine samples were collected at each clinic visit for routine laboratory analyses.
Analyses were performed on a variety of endpoints at weeks 2, 4, 8, and 12 to identify the treatment dose and duration, as well as key efficacy measures, for progression into subsequent confirmatory trials. Consequently, treatment comparisons focused on the pair-wise comparisons of each alosetron dose vs. placebo. Statistical significance was not the only determinant of success; persistence and magnitude of treatment effect across endpoints were factors as well.
Investigation of the effects of various risk factors, including gender, on the improvement of abdominal pain and discomfort was originally planned as a secondary analysis. However, due to initial findings of these secondary analyses, as well as results from another dose-ranging study of alosetron in irritable bowel syndrome, the analyses of efficacy parameters among genders were refined and emphasized retrospectively.18 Statistical testing was carried out for the total population and separately for the male and female sub-groups; as this was a dose-ranging study to generate hypotheses, and the sub-group analyses were exploratory in nature, no adjustment for multiple comparisons was made. Such analyses within sub-groups may lack sufficient power for adequate inferential assessment; this is especially true in the male sub-group because the study population reflects the predominantly female irritable bowel syndrome patient population (70–75%).
The number of patients (male and female) in each treatment group who had at least one efficacy assessment (n = 117 for placebo, n = 113 for alosetron 0.1 mg b.d., n = 111 for alosetron 0.5 mg b.d., n = 113 for alosetron 2 mg b.d.) provided at least 80% power to detect a 20% difference in proportions between any dose of alosetron and placebo at an alpha = 0.05 significance level (unadjusted for multiple comparisons).
The primary population for analyses included the available data for all patients who had taken study medication and provided at least one efficacy assessment. The key endpoints were the proportion of pain-free days and the proportion of patients with a ≥ 10% improvement in visual analogue scale pain/discomfort score. The following bowel functions were also evaluated: hardening of stool, number of bowel actions, and improvement in diarrhoea. For analyses of visual analogue scale scores, missing data (due to dropouts or otherwise) were handled via the last observation carried forward principle whereby missing values were replaced with the last available non-missing value. All other analyses utilized observed data.
The percentage of pain-free days was computed for the 2-week screening phase, weeks 1–2, 5–8, and 9–12; pairwise comparisons of each alosetron dose group vs. placebo were made using the van Elteren method of the Wilcoxon rank-sum test with stratification for clusters of centres.19 Proportions of patients showing an improvement in the visual analogue scale score at each time-point were compared between treatment groups using the Mantel–Haenszel χ2-test without continuity correction, stratified by cluster of centres and the mean baseline (weeks −2 to 0) stool consistency score as recorded in the diary card.20 Two strata of mean baseline stool consistency scores were defined: 1.0–2.5, and > 2.5–5.0. Furthermore, approximate (unstratified) 95% confidence intervals were calculated for the treatment effect on the pain parameters at each time-point.
The proportion of patients reporting hardening of stool was compared between treatment groups using the Mantel–Haenszel χ2-test without continuity correction, stratified by clusters of centres.20 The number of bowel actions was compared between treatment groups using the van Elteren method of the Wilcoxon rank-sum test with stratification for clusters of centres. The proportion of patients reporting improvement in diarrhoea (i.e. a positive improvement of 10% or more on the visual analogue scale) was compared between treatment groups using the Mantel–Haenszel χ2-test without continuity correction, stratified by clusters of centres and the mean baseline stool consistency score.
The study was approved by local Ethics Committees and each patient gave written informed consent. The study was conducted in accordance with the Declaration of Helsinki.
A total of 462 patients with irritable bowel syndrome were randomized to treatment and took at least one dose of study medication. Of these, eight patients provided no efficacy data: therefore 454 patients were available for analysis of efficacy. (see Figure 1).
Demographic variables were similar across treatment groups and were similar for the male and female sub-groups (Table 1). The majority of patients were female (73%) and the average age was 43 years (range 18–74 years). The median time since diagnosis of irritable bowel syndrome was 5 years, although approximately 7% of patients were diagnosed with irritable bowel syndrome within 1 year of entry to the study and 30% were first diagnosed 10 years or more before entry. More than half of the patients reported at least one concurrent medical condition. The incidence was similar across all treatment groups and the most frequently recorded conditions were gastrointestinal, musculoskeletal, cardiovascular and reproductive. Twenty per cent of patients used concomitant medication during the study, of which the most frequently recorded were for nervous system disorders (7% of patients), alimentary system disorders (5%) and anti-infective agents (4%). The number of patients using concurrent medications and the types of medications used were evenly distributed across the treatment groups. Of the patients, 36% reported taking medication for their irritable bowel syndrome symptoms immediately prior to entry to the study.
Diary card measurement. The mean proportion of pain-free days increased in all treatment groups between the screening phase and weeks 9–12 ( Figure 2A). The proportion of pain-free days during the periods weeks 5–8 and weeks 9–12 was significantly greater for patients treated with alosetron 2 mg b.d. than for the placebo group. The results for females and males are shown in Figure 2(B) and (C), respectively. Treatment differences and the 95% confidence interval around the treatment differences at week 12 are shown in Table 2. The results in females were similar to the total population with the alosetron 2 mg b.d. group showing significant benefit over placebo at the 9–12 week interval. In males there was no significant benefit over placebo with any dose of alosetron.
The mean baseline visual analogue scale score for pain/discomfort was between 53 and 56 mm for each of the four groups. There was a large placebo response with about half of the patients reporting an improvement in abdominal pain or discomfort within 4 weeks. However, treatment with alosetron 2 mg b.d. was associated with a statistically significant increase over the placebo group in the proportion of patients with an improvement in abdominal pain/discomfort after 12 weeks of treatment (63% vs. 51%, P = 0.04) ( Figure 3). Significant benefit was seen in patients with a baseline stool consistency of > 2.5–5 (i.e. firm to loose/watery) but not in patients with a baseline stool consistency of 1.0–2.5 (i.e. hard/firm stool). Analysis by gender showed that there was a non-significant increase in the number of patients in the alosetron 2 mg b.d. group who improved for both males and females compared with placebo ( Figure 3). The treatment difference against placebo and the 95% confidence intervals around the difference at week 12 are shown in Table 3.
There was a statistically significant increase in the proportion of patients reporting a hardening of stool consistency and a reduction in the frequency of bowel actions in patients treated with alosetron 0.5 mg b.d. and alosetron 2 mg b.d. These differences were apparent within the first 2 weeks of treatment ( Figures 4 and 5). The effect on hardening of stool consistency was significant (P < 0.05) in both females (38% improvement over placebo at week 12) and males (50% improvement over placebo at week 12) ( Figure 4B and C). An observed reduction in stool frequency in both male and female sub-groups was not statistically significant at weeks 9–12 ( Figure 5B and 5 . The number of bowel actions per week in the total (A), female (B) and male (C) population. *P < 0.05C).
The effects of alosetron on bowel function were also apparent from analysis of visual analogue scale scores for diarrhoea. After 12 weeks 50% of patients receiving alosetron 2 mg b.d. reported an improvement in diarrhoea (defined as a decrease of ≥ 10% on the visual analogue scale) compared to 36% of those on placebo (P = 0.03). In females, 51% of those receiving alosetron 2 mg b.d. reported an improvement in the visual analogue scale score for diarrhoea compared with 34% of those on placebo (P = 0.033). In males there was no significant improvement over placebo (41% patients compared with 46% of patients on alosetron 2 mg b.d.).
The overall incidence of adverse events was similar in all treatment groups including placebo (Table 4). The incidence of constipation was significantly higher during treatment with alosetron 0.5 mg b.d. and alosetron 2 mg b.d. than with placebo; for eight of the 39 patients who reported constipation as an adverse event (six receiving alosetron 2 mg b.d.) this was recorded as an exacerbation of a pre-existing symptom. Seventeen patients (14 female) withdrew from the study because of constipation (placebo, one; alosetron 0.1 mg b.d., two [both female]; alosetron 0.5 mg b.d., six [five female and one male]; alosetron 2 mg b.d., eight [seven female]). Of these 17 patients, 11 had been classified as constipation-predominant at the pre-trial visit. A total of 35 other patients were withdrawn due to a variety of adverse events, most of which were assessed as drug related, but which showed no trends towards increasing incidence with dose of alosetron. In total, 139 patients withdrew prematurely from the study, 33 from the placebo group, 33 from the alosetron 0.1 mg b.d. group, 41 from the alosetron 0.5 mg b.d. group and 32 from the alosetron 2 mg b.d. group. Apart from adverse events the other reasons for premature withdrawal included lack of efficacy/did not wish to continue (48 patients), lost to follow-up (18 patients) and other reasons (21 patients).
There were no changes in laboratory values associated with treatment with alosetron.
Few studies have provided convincing evidence for efficacy of any treatment for patients with irritable bowel syndrome. There are no objective markers of improvement and the large placebo response confounds interpretation of potential clinical benefit. The present study sought to address some of the issues that have limited evaluation of previous data and to examine the effects of alosetron on symptoms in irritable bowel syndrome patients. In addition, since irritable bowel syndrome sufferers are predominantly female, analyses on male and female sub-populations were also carried out.
The patient population was clearly defined, and a parallel-group study design of 12 weeks’ duration was employed to assess whether any treatment effects were sustained. Detailed symptom assessments were made with emphasis on patient self-assessments. It is acknowledged that the placebo response rate was high. Approximately 50% of placebo treated patients had an improvement in the visual analogue scale score for pain. However, this degree of placebo response is consistent with placebo response rates of up to 70% that are cited in the literature.21, 22 Despite the large placebo response rates, under these rigorous conditions alosetron 2 mg b.d. was associated with a statistically significant increase in the proportion of patients reporting an improvement in their abdominal pain/discomfort.
Irritable bowel syndrome is a female predominant condition. The reasons for this female predominance are not fully understood. There may be physiological differences that predispose females to irritable bowel syndrome. Smith and Thompson have shown the Manning Criteria to be of predictive value for diagnosing irritable bowel syndrome in females but not in male patients.23, 24 Using positron emission tomography Silverman has shown different patterns of cerebral blood flow during colo-rectal distension in male and female irritable bowel syndrome patients.25 Therefore it may be important to assess gender effects when assessing response to treatment of a novel therapy for irritable bowel syndrome.
The effect of alosetron was analysed separately in males and females. Whereas females showed a consistent response (alosetron 2 mg b.d. significantly improved pain/discomfort, stool consistency and diarrhoea and there was a trend towards improvement in stool frequency), convincing improvements were not seen in males. The only significant benefit in males was seen in a hardening of stool with the 2 mg b.d. dose. No significant or consistent benefits in pain/discomfort scores (analysed either as proportion of pain/discomfort-free days or visual analogue scale score) were seen in male patients with any dose of alosetron compared with placebo. In fact, the dose–response in males for their response for proportion of pain-free days appeared opposite to that for the visual analogue scale score for pain. The apparent lack of effect in males is surprising. The placebo responses were similar in the male and female sub-populations. Also, baseline stool consistency scores were similar in males and females so this cannot explain the difference in response to alosetron treatment. However, care must be taken in interpreting the findings in males. Lack of demonstrable efficacy may be partly explained by the low patient numbers and the resultant lack of power to detect a treatment difference. It is notable however, that a second dose-ranging study has also shown that alosetron is effective in female but not male irritable bowel syndrome patients; the number of male patients included in that study was also low.18
There is no universally acceptable efficacy end-point for clinical studies in irritable bowel syndrome patients. One of the aims of this study was to assess a suitable end-point for future studies. The definition of improvement in visual analogue scale scores used in this study was a ≥ 10% shift from baseline score and there was clear evidence that alosetron 2 mg b.d. was associated with an improvement in the patients’ perception of the severity of their condition. A 10% benefit was chosen prior to the study, after discussion with physicians with an interest in irritable bowel syndrome, as a benefit that would probably be clinically meaningful. A patient may feel benefit if symptoms improve even though symptoms may not be completely relieved. A visual analogue scale score allows the patient to register this. It is interesting that alosetron at a dose of 2 mg b.d. led to improvement across a number of end-points (abdominal pain/discomfort and bowel function parameters), particularly in patients with normal to loose/watery stool at baseline. These results are very encouraging that alosetron may be the first therapy that improves most irritable bowel syndrome symptoms in non-constipated patients.
5-HT3 receptor antagonists have been shown to slow colonic transit4, 5 and the current finding of a dose-related hardening of stools and reduction of bowel frequency was not unexpected. For many patients this represents a therapeutic benefit but for a few, constipation constituted an adverse event. However, the number of withdrawals due to constipation was low. In some cases constipation was recorded as an exacerbation of a pre-existing condition, and it would seem sensible to exclude patients with constipation during baseline screening from future studies with 5-HT3 receptor antagonists.
This study has shown that alosetron at a dose of 2 mg b.d. significantly improves abdominal pain/discomfort in irritable bowel syndrome patients as well as normalizing bowel function (as shown by the hardening of stool consistency, reduced stool number per day and improvement in visual analogue scale score for diarrhoea). While benefit was observed in patients with normal to loose/watery stools at baseline, no benefit was observed in patients whose baseline stool consistency was described as hard to firm.
In females, alosetron 2 mg b.d. significantly improved the proportion of pain-free days, the visual analogue scale score for diarrhoea, led to a significant hardening of stool and tended to reduce stool frequency. In males results were less convincing. The number of male patients was low and thus lack of demonstrable efficacy may be related to lack of power. However, the gender effect seen in this exploratory study is interesting and should be addressed in further alosetron efficacy studies.
In conclusion, alosetron 2 mg b.d. is an effective treatment for non-constipated female irritable bowel syndrome sufferers. Further work on the apparent difference in response between genders is required.
This study was funded by Glaxo Wellcome Research and Development, Protocol No S3BP12.
A preliminary report of these data was published as an abstract in Gastroenterology 1996; 110: A630.
The authors would like to thank all members of the study group and their staff for their contribution to the study.
Belgium: Prof. G. Coremans, Dr V. Gillard, Dr G. Ligny, Dr H. Naessens. Canada: Dr R. J. Bailey, Dr C. Dallaire, Dr R. Davies, Dr P. Parê, Dr A. Pellicano. Denmark: Dr B. N. Andersen, Dr P. Bech-Jansen, Dr L. Hendel, Dr P. J. Ranloev, Dr S. N. Rasmussen, Prof O. B. Schaffalitzky de Muckadell. France: Dr A. Blanchi, Prof. G. Bommelaer, Dr M. Evreux. Germany: Dr H. Baumann, Prof. W. Rösch, Dr E. Schütz, Dr A. Sommer, Dr H. Stock. The Netherlands: Dr S. Y. De Boer, Dr C. P. M. Dekkers, Dr H. Geldof, Dr C. J. J. Mulder, Dr J. A. J. Staal, Prof. G. P. Van Berge Henegouwen. Poland: Prof. E. Butruk, Prof. K. Marlicz, Prof. A. Nowak. Sweden: Dr D.-A. Hallbäck, Dr L. Lööf, Dr H. Nyhlin, Dr L. Weywadt, Dr G. Bodemar. UK: Dr K. D. Bardhan, Dr A. D. Beattie, Dr J. Collins, Dr A. S. Hungin, Dr R. E. Pounder, Dr G. Tildesley.