Evidence review: efficacy
No randomised controlled trials (RCTs) were identified that compared colesevelam with placebo or other treatments in people with bile acid malabsorption.
One published RCT was identified that compared colesevelam with placebo in 24 women with diarrhoea-predominant irritable bowel syndrome, 4 of whom had evidence of bile acid malabsorption.
Two case series were identified that reported on the efficacy of colesevelam for treating bile acid malabsorption. One included people who developed bile acid malabsorption after cancer therapy; the other included people with bile acid malabsorption who could not tolerate colestyramine.
Odunsi-Shiyanbade et al. (2010) conducted a small double-blind RCT that assessed the effect of colesevelam on gastrointestinal and colonic transit, bowel function and colonic permeability in 24 women with diarrhoea-predominant irritable bowel syndrome. The women were randomised to receive 1.875 g colesevelam or matching placebo twice daily for 12–14 days.
Three women in the placebo group and 1 in the colesevelam group had evidence of bile acid malabsorption (fasting serum 7-hydroxy-4-cholesten-3-one [7C4] 61 nanograms/ml or more [an indirect measure of bile acid synthesis that is used to screen for bile acid malabsorption]). However, the results for these women were not reported separately.
The mean age of the women was 43 years in the placebo group and 42 years in the colesevelam group. Mean body mass index was 29 kg/m2 in the placebo group and 32 kg/m2 in the colesevelam group. Baseline colonic geometric centre at 24 hours (a measure of colonic transit, with a higher geometric centre implying a faster colonic transit) was 3.4 in both groups. Mean fasting serum 7C4 was 38.2 nanograms/ml in the placebo group and 31.8 nanograms/ml in the colesevelam group. It is unclear whether differences between the baseline characteristics of the groups were statistically significant. Allocation was concealed, all women completed the study and adherence was 100%.
The primary end points were colonic geometric centre at 24 hours and ascending colon emptying half-life (the time taken for 50% of a radiolabeled meal to empty from the colon). Other outcomes included bowel function (for example, daily bowel frequency and consistency) and colonic mucosal permeability.
After treatment, there was no statistically significant difference in gastric, small bowel and overall colonic transit, or ascending colonic emptying with colesevelam compared with placebo.
Compared with placebo, colonic geometric centre was lower at 4 hours (0.39 lower on average; statistical significance of difference not reported) and 24 hours (0.62 lower on average, p=0.18) in women receiving colesevelam, suggesting slower colonic transit. However, the difference between the groups at 24 hours, the primary outcome, was not statistically significant and at 48 hours the results were similar in both groups (placebo 4.47 compared with colesevelam 4.6; statistical significance not reported). Gastric emptying half-life was, on average, 36.5 minutes longer in women receiving colesevelam, compared with placebo. However, this result did not reach statistical significance (p=0.14). Similarly, ascending colon half-life was, on average, 3.95 hours longer in women receiving colesevelam compared with women receiving placebo. The statistical significance of this difference was not reported.
There was no difference between the colesevelam and placebo groups in the number of bowel movements (mean 2 per day in both groups; statistical significance not reported) or in stool consistency (Bristol Stool Form Scale, mean score 4.57 with placebo and 3.78 with colesevelam; p=0.12). Stool passage was statistically significantly easier in women who received colesevelam, compared with women who received placebo (4.18 with colesevelam compared with 4.39 with placebo, on a scale of 1 to 7, with 1 representing manual disimpaction and 7 incontinence; p=0.047). However, it is unclear whether this difference is of clinical importance. There was no significant difference in colonic mucosal permeability between the colesevelam and placebo groups.
Higher baseline fasting serum levels of 7C4 were significantly associated with higher ascending colon half-lives (a marker for reduced colonic transit times; p<0.001) in women who received colesevelam. The authors suggested that 7C4 levels may predict responsiveness to colesevelam in women with diarrhoea-predominant irritable bowel syndrome.
Wedlake et al. (2009) reported a retrospective chart review of 45 people who had completed cancer treatment or were undergoing treatment for ongoing or relapsed cancer, and who had chronic symptoms of bile acid malabsorption for 3 months or more for which they had been prescribed colesevelam (1.25 g 3 times a day with or after food) between 2004 and 2007.
Hospital electronic patient records were reviewed retrospectively for all 45 people to assess the effects of 2−6 weeks of colesevelam treatment on 6 gastrointestinal symptoms, which had been recorded prospectively in the patient record. Outcomes were also assessed by a patient questionnaire, which was sent to 36 of the 45 people included in the case series (excluding those who were terminally ill). Out of 36 people, 30 (83%) responded. The median treatment period in these people was 27 months. When data from the electronic patient records and questionnaires were combined, it was found that 30 people (67%) were still taking colesevelam at their last follow-up (up to 4 years).
Of the 45 people, 37 were women and 8 were men. Their median age was 58 years. Probable causes of bile acid malabsorption were pelvic radiotherapy (n=29), right hemicolectomy or small-bowel resection (n=12), upper gastrointestinal surgery (n=2), high-dose chemotherapy (n=1) and new-onset Crohn's disease (n=1).
Bile acid malabsorption was confirmed by SeHCAT scan in 27 people (7‑day absorption 10% or less). SeHCAT scan was not undertaken in the remaining 18 people because they were very ill, could not attend appointments for a scan, or developed loose stools immediately after ileal resection.
Symptoms had not responded to prior colestyramine treatment in 30 (67%) of the 45 people. The other 15 people were prescribed colesevelam as a first treatment either because colestyramine was considered unsuitable or because they requested a tablet rather than a powder.
In people who experienced each symptom, the following were improved after colesevelam treatment (proportions of people):
diarrhoea (record review 88%; questionnaire 80%)
frequency of defecation (record review 77%; questionnaire 83%)
steatorrhoea (record review 76%; questionnaire 80%)
urgency of defecation (record review 76%; questionnaire 80%)
abdominal pain (record review 74%; questionnaire 58%)
faecal incontinence (record review 69%; questionnaire 74%).
Based on the review of patient records, of the 30 people who did not respond to prior colestyramine treatment, colesevelam improved diarrhoea in 83%, urgency of defecation in 74%, frequency of defecation in 72%, steatorrhoea in 71%, abdominal pain in 68% and faecal incontinence in 62%.
Puleston et al. (2005) reported on the use of colesevelam in 5 people with bile acid malabsorption who could not tolerate colestyramine.
Bile acid malabsorption was caused by right hemicolectomy in 2 people, radiation enteritis in 1 person, and both radiation enteritis and right hemicolectomy in 1 person. In the fifth person, bile acid malabsorption was idiopathic. Two people were male and 3 were female. Their average age was 47 years.
Colesevelam resolved diarrhoea with no adverse effects at doses between 1.25 g per day and 3.75 g per day for 2–7 months. It also resolved steatorrhoea and lethargy in 1 person (these outcomes were not reported on for the other 4 people).