Alimentary Pharmacology and Therapeutics

Randomised clinical trial: colestyramine vs. hydroxypropyl

cellulose in patients with functional chronic watery diarrhoea
nares*,†, M. Rosinach*,†, M. Piqueras‡, A. Ruiz-Cerulla§, I. Modolell‡, Y. Zabana*,†,
F. Fern
andez-Ba~
§
J. Guardiola & M. Esteve*,†

*Department of Gastroenterology, SUMMARY

Hospital Universitari Mutua Terrassa,
Terrassa, Spain.
†
Centro de Investigaci
on Biom
edica
Background
en Red de Enfermedades Hep
aticas y Idiopathic bile acid malabsorption (BAM) has been suggested as a cause of
Digestivas (CIBERehd), Barcelona, chronic watery diarrhoea, with a response to colestyramine in 70% of
Spain. patients. However, the efficacy of this drug has never been investigated in
‡
Department of Gastroenterology, placebo-controlled trials.
Consorci Sanitari Terrassa, Barcelona,
Spain.
§
Department of Gastroenterology, Aim
Hospital de Bellvitge, Barcelona, To evaluate the efficacy of colestyramine as compared with hydroxypropyl
Spain. cellulose in the treatment of functional chronic watery diarrhoea.

Methods
Correspondence to:
Dr F. Fern
andez-Ba~
nares, Department
Patients with chronic watery diarrhoea were randomly assigned to groups
of Gastroenterology, Hospital given colestyramine sachets 4 g twice daily (n = 13) or identical hydroxy-
Universitari Mutua Terrassa, Placßa Dr propyl cellulose sachets (n = 13) for 8 weeks. The primary end-point was
Robert 5, 08221 Terrassa (Barcelona), clinical remission defined as a mean of 3 or fewer stools per day during the
Spain. week before the visit, with less than 1 watery stool per day. A secondary
E-mail: ffbanares@mutuaterrassa.es
end-point was the reduction in daily watery stool number. SeHCAT test
was performed in all patients, but an abnormal test was not a prerequisite
Publication data to be included.
Submitted 5 December 2014
First decision 12 February 2015 Results
Resubmitted 24 February 2015 All included patients had a SeHCAT 7-day retention ≤20%. There were no
Resubmitted 19 March 2015
statistical differences in the percentage of patients in clinical remission at
Accepted 19 March 2015
EV Pub Online 10 April 2015 week 8 between colestyramine and hydroxypropyl cellulose with either
intention-to-treat (53.8% vs. 38.4%; P = 0.43) or per-protocol (63.6% vs.
This article was accepted for publication 38.4%; P = 0.22) analyses. However, the mean per cent decrease in watery
after full peer-review. stool number was significantly higher with colestyramine than with hy-
droxypropyl cellulose ( 92.4  3.5% vs. 75.8  7.1%; P = 0.048). The
rate of adverse events related to study drugs did not differ between groups.

Conclusions
Colestyramine (4 g twice daily) is effective and safe for short-term treat-
ment of patients with chronic watery diarrhoea presumably secondary to
BAM. Clinical Trials Register number EudraCT 2009-011149-14.

Aliment Pharmacol Ther 2015; 41: 1132–1140

1132 ª 2015 John Wiley & Sons Ltd

doi:10.1111/apt.13193
 Randomised clinical trial: colestyramine in chronic watery diarrhoea

INTRODUCTION METHODS
Idiopathic bile acid malabsorption (BAM) has been
suggested as a cause of watery diarrhoea in patients Study design and setting
diagnosed with irritable bowel syndrome (IBS-D). Since The study was designed as a double-blind, randomised
the 1980s, the use of the 75Selenium homocholic acid placebo-controlled, phase IV, noncommercial, indepen-
taurine (SeHCAT) test as a screening test for BAM dent, comparative clinical trial conducted in three cen-
has revealed that idiopathic (type 2) BAM is not a tres in Spain. The study was funded by a Spanish health
rare entity. A systematic review of studies investigating ministry grant on Independent Drug Research (grant
BAM in patients presenting with IBS-D symptoms number EC08/00085). Patients were included from July
showed that 10% of these patients had severe 2010 to September 2014. The study protocol was con-
BAM (SeHCAT < 5%), and that 32% had moderate- ducted in accordance with the International Conference
to-severe BAM (SeHCAT ≤ 10%).1 Colestyramine on Harmonisation Guideline for Good Clinical Practice
seems to be a successful therapy with improvement or and in full conformity with relevant regulations. It was
cessation of diarrhoea in 70% (range, 63–100%) of approved by the ethical committees of the three partici-
patients who were able to tolerate it.1, 2 However, due pating centres. The study protocol was registered at
to a lack of randomised controlled trials, the scientific www.clinicaltrialsregister.eu (EudraCT 2009-011149-14).
evidence for the use of colestyramine in this setting is All authors had access to the study data and reviewed
scarce.3 and approved the manuscript.
In a previous study by our group, diarrhoea second-
ary to BAM, diagnosed on the basis of both an abnor- Study population
mal SeHCAT test and a long-term response to Men and women aged 18 years or more were eligible for
colestyramine, was the cause of diarrhoea in most randomisation if they met the following inclusion crite-
patients suffering from chronic watery diarrhoea of ria: (i) presence of chronic watery diarrhoea defined as
functional characteristics when microscopic colitis and >2 watery stools per day (type 6–7 in the Bristol Stool
gluten-sensitive enteropathy had been ruled out.4 How- Scale); (ii) fulfilling Rome III criteria of IBS-D or func-
ever, the sensitivity of the SeHCAT test in detecting tional diarrhoea; (iii) normal physical examination and
BAM is not well established, and no standardisation of blood analysis, including blood biochemistry (fluid/elec-
the definition of a positive SeHCAT test exists.3 Like- trolyte status; kidney and liver function; prothrombin
wise, it has been suggested that response rate to coles- time; ferritin; folate; serum protein/globulins; cholesterol
tyramine is not dependent on the severity of BAM, and triglycerides; calcium; magnesium; C-reactive pro-
with a treatment success in 67% of patients with <5% tein), complete blood count, ESR, serum T4-TSH and
retention in the SeHCAT test, 73% of patients with <8– serum IgA-human anti-tissue transglutaminase antibod-
12% retention and 59% of those with <15% retention.2 ies; (iv) negative faecal bacterial cultures and exam for
For all these reasons, the necessity of performing a ova and parasites; (v) women of child-bearing potential
SeHCAT test to select patients with chronic watery were required to use appropriate contraceptive methods;
diarrhoea to be treated with colestyramine is not well and (vi) all participants provided written informed con-
documented. sent.
The aim of the present study was to evaluate the effi- A SeHCAT test was performed on all participants at
cacy and tolerability of short-term treatment with coles- baseline to quantify the extent of BAM. Patients were
tyramine as compared with hydroxypropyl cellulose in included if they had an abnormal SeHCAT test (≤10%
patients presenting with either functional chronic diar- 7-day retention) or if after an extensive diagnostic work-
rhoea or IBS-D symptoms in a double-blind, rando- up, there was no aetiology of the diarrhoea (idiopathic
mised, controlled trial. Hydroxypropyl cellulose was chronic watery diarrhoea). Exclusion criteria for partici-
chosen as a placebo although, as discussed below, we pation included previous treatment with colestyramine;
subsequently became aware that it may be an active drug significant ileal and colonic diseases (i.e. microscopic
in treating BAM. colitis – normal multiple colonic biopsies were required,
The presence of BAM (as assessed by a SeHCAT test polyps >2 cm, ulcerative colitis, Crohn’s disease, ischae-
≤10% 7-day retention) was not a prerequisite for inclu- mic colitis); ileal and/or colonic resection, vagotomy,
sion in the present trial. cholecystectomy; other intestinal diseases with structural

Aliment Pharmacol Ther 2015; 41: 1132–1140 1133

ª 2015 John Wiley & Sons Ltd
F. Fern
andez-Ban
~ares et al.

damage; infectious diarrhoea; coeliac disease [negative Clinical outcomes evaluation

coeliac serology and normal (Marsh 0) duodenal biop- Our primary end-point was clinical remission at 8 weeks,
sies, if positive HLA-DQ2 and/or HLA-DQ8, were defined as a mean of 3 or fewer stools per day during
required]; malignant disease; severe co-morbidity; abnor- the week before the visit, with less than 1 watery stool
mal hepatic function or liver cirrhosis; bile conduct per day.5 A secondary end-point was the reduction in
obstruction; renal insufficiency; either alcohol or drug number of watery stools (type 6–7 in Bristol Stool Scale)
abuse; chronic intake of NSAIDs or olmesartan; partici- per day. Additional end-points were changes in health-
pation in other randomised controlled trial in the previ- related quality of life as assessed with the Gastrointesti-
ous 4 weeks; pregnancy or breast-feeding. nal Quality of Life Index (GIQLI),6 and tolerability and
As inclusion/exclusion criteria refer to patients with- safety.
out clinical history, signs or symptoms, or analytical
abnormalities supporting either pancreatic insufficiency Safety evaluation
or bacterial overgrowth as the cause of the chronic diar- At baseline and final visit, patients underwent physical
rhoea, tests to rule out these entities were no routinely examination and general laboratory tests. In addition, at
indicated. each appointment, previous (at baseline) and concomi-
tant medications were recorded, and adverse events were
Treatment allocation noted.
A computer-generated list of random numbers was used
for allocation of participants. This list was prepared by Statistical analysis
the Pharmacy Department of Hospital Universitari Mutua Sample size was calculated assuming rates of clinical
Terrassa without any clinical involvement in the trial. Eli- remission of 70% in the colestyramine group.1, 2 As
gible patients were randomly assigned to one of two treat- there were no previous data about placebo response in
ment groups. The study medication was packed in boxes, this setting, we select a arbitrarily 20% response rate for
and consecutively numbered for each patient according to the placebo comparison group. With an alpha value of
the randomisation schedule. The investigators at the cen- 5% and an 80% statistical power to yield a statistically
tres enrolled the patients and dispensed the study medica- significant result, the calculated sample size was 12
tion according to the randomisation schedule. Patients patients per group.
received either 4 g colestyramine sachets twice daily or Efficacy was analysed for the intention-to-treat (ITT)
identical placebo sachets for 8 weeks in a double-blind population with a sensitive analysis for the per-protocol
fashion. It was very difficult to select a placebo of similar (PP) population. Patients with lack of compliance, viola-
organoleptic properties than colestyramine. Hydroxypro- tion of eligibility criteria or early discontinuation due to
pyl cellulose was chosen for comparison although, as dis- adverse event without causal relationship with study
cussed below, we subsequently became aware that it may drug were excluded from PP population. Safety analysis
be an active drug in treating BAM. was performed descriptively for the safety population.
Colestyramine sachets were re-packed for masking Results are given as mean  S.E.M. or percentages.
from Resincolestiramina sachets (Rubi
o Laboratories, Fisher’s exact test and paired or unpaired t-test were
Barcelona, Spain), by an independent company that also used for the statistical analysis, and type 1 error rate was
prepared the identical hydroxypropyl cellulose sachets. two-sided with alpha = 0.05. All statistical analyses were
Resincolestiramina sachets contain strawberry flavoured conducted using the SPSS for Windows statistical package
colestyramine (4 g) powder to be dissolved in water, and (SPSS Inc., Chicago, IL, USA).
as excipient, it includes carboxymethyl cellulose. Rubi
o
Laboratories did not have any role in the present inde- RESULTS
pendent, noncommercial research.
Interim visits were made at weeks 1, 2 and 4. Patients Patient population
nonresponsive after 2 weeks were allowed to increase We randomised a total of 26 patients (colestyramine, 13;
the study drug dose to three sachets daily. Adherence hydroxypropyl cellulose, 13) eligible for ITT analysis.
to the study treatment was monitored by sachet count at The flow of patients throughout the study is described in
each study visit and with the patient diary. During the Figure 1. A total of two patients in the colestyramine
entire study period, the use of anti-diarrhoeals, spasmolytics group had adverse events that prompted their with-
and other drugs causing constipation was not permitted. drawal, but there was no causal relationship with the

1134 Aliment Pharmacol Ther 2015; 41: 1132–1140

ª 2015 John Wiley & Sons Ltd
 Randomised clinical trial: colestyramine in chronic watery diarrhoea

Randomised, n = 26

Colestyramine, n = 13 Placebo, n = 13

Safety/ITT population, n = 13 Safety/ITT population, n = 13

PP population, n = 11
- 2 premature discontinuation by adverse PP population, n = 13
events not related to the study drug

Completed study, n = 10 Completed study, n = 8

Discontinuted (n = 3) Discontinuted (n = 5)
- 1 lack of efficacy - 4 lack of efficacy
- 2 adverse events - 1 violation of protocol

Figure 1 | Study flow of randomised patients.

Table 1 | Baseline demographic and clinical characteristics of each group

Characteristics Colestyramine (n = 13) Hydroxypropyl cellulose (n = 13) P

Age (years)* 50.0  4.9 49.3  4.5 NS
Sex (% women) 7/13 (53.8) 11/13 (84.6) NS
Duration of diarrhoea (months)* 69.7  23.6 80.1  29.8 NS
Daily stool number* 5.0  0.46 4.9  0.43 NS
IBS (%) 8/13 (61.5) 10/13 (76.9) NS
Functional diarrhoea (%) 5/13 (38.5) 3/13 (23.1) NS
%SeHCAT 7-day retention
≤20 13/13 (100%) 13/13 (100%) NS
≤15 10/13 (77%) 12/13 (92.3%)
≤12 10/13 (77%) 11/13 (84.6%)
≤10 10/13 (77%) 7/13 (53.8%)
≤5 7/13 (53.8%) 7/13 (53.8%)
GIQLI* 119.7  6.2 116.1  8.9 NS
NS, nonsignificant.
* Mean  S.E.M.

study medication (see below), leaving 24 patients for the groups had a SeHCAT 7-day retention ≤20% (Figure 2).
PP analysis. The baseline demographic and clinical char- Adherence to study drug was excellent, higher than 80%
acteristics of the ITT population were similar for the two in all patients and without differences between study
treatment groups (Table 1). All included patients in both groups.

Aliment Pharmacol Ther 2015; 41: 1132–1140 1135

ª 2015 John Wiley & Sons Ltd
F. Fern
andez-Ban
~ares et al.

In these subgroups, a trend was observed but the small

20 sample size precluded to achieve statistical significance.
17
The effect of study drugs on the daily stool number
throughout the trial is shown in Figures 3 and 4. Both
14 colestyramine and hydroxypropyl cellulose significantly
%SeHCAT 7 day

11
reduced the mean number of watery stools per day from
5.0  0.46 to 0.18  0.20 (P < 0.0005), and from
8 4.9  0.43 to 1.29  0.42 (P < 0.0005) respectively.
5
However, the mean per cent decrease was significantly
greater with colestyramine than with hydroxypropyl cel-
2 lulose ( 92.4  3.5% vs 75.8  7.1%; P = 0.048). The
–1
effect depending on the SeHCAT 7-day retention is
CLT HPC described in Figure 5a,b. The mean per cent decrease
Treatment group
was greater with colestyramine than with hydroxypropyl
cellulose for all SeHCAT 7-day retention values, achiev-
Figure 2 | Box plots describing the basal SeHCAT 7-day ing statistical significance in the ≤10% cut-off value for
retention in both study drug groups. There were no
both daily total stool number and daily liquid stool num-
significant differences between groups.
ber, and in the ≤20% cut-off value for the daily liquid
stool number.
The daily dose of colestyramine was increased to
Clinical efficacy 12 g/day in four nonresponsive patients after 2 weeks,
Regarding the primary end-point, the proportion of two of them achieving clinical remission. Eight nonre-
patients in clinical remission at week 8 was higher with sponsive patients required a dose increase to three daily
colestyramine than with hydroxypropyl cellulose, but the sachets in the hydroxypropyl cellulose group, being effec-
difference did not reach statistical significance either with tive in two of them.
the ITT or the PP analysis (Table 2). The differences
persisted nonsignificant for the different cut-offs of SeH- Health-related quality of life
CAT test, but they were more marked in the cut-offs Both colestyramine and hydroxypropyl cellulose
used in literature to define both moderate-to-severe mal- improved health-related quality of life as assessed with
absorption (SeHCAT 7-day retention ≤10%) and severe the GIQLI, but the difference was more marked with
malabsorption (SeHCAT 7-day retention ≤5%) (Table 2). colestyramine (Table 3). In the colestyramine group,

Table 2 | Clinical remission rates at week 8 by ITT and PP depending on the %SeHCAT 7-day retention. All included
patients had a SeHCAT 7-day retention ≤20%; therefore, this value corresponds to the whole series of patients

%SeHCAT 7-day retention Colestyramine (n = 13) Hydroxypropyl cellulose (n = 13) P value

≤20
ITT (n = 26) 7/13 (53.8%) 5/13 (38.5%) 0.43
PP (n = 24) 7/11 (63.6%) 5/13 (38.5%) 0.42
≤15
ITT (n = 22) 4/10 (40%) 5/12 (41.6%) 0.93
PP (n = 20) 4/8 (50%) 5/12 (41.6%) 0.71
≤12
ITT (n = 21) 4/10 (40%) 5/11 (45.4%) 0.80
PP (n = 19) 4/8 (50%) 5/11 (45.4%) 0.84
≤10
ITT (n = 18) 4/10 (40%) 2/8 (25%) 0.50
PP (n = 16) 4/8 (50%) 2/8 (25%) 0.30
≤5
ITT (n = 13) 3/6 (50%) 2/7 (28.6%) 0.43
PP (n = 12) 3/5 (60%) 2/7 (28.6%) 0.28

1136 Aliment Pharmacol Ther 2015; 41: 1132–1140

ª 2015 John Wiley & Sons Ltd
 Randomised clinical trial: colestyramine in chronic watery diarrhoea

8 10
P < 0.0005 Initial vs final P < 0.02 Initial vs final
9
7
8
6

Daily total stool number

Daily total stool number
7
5
6

4 5

4
3
3
Figure 3 | Evolution of the 2
2
daily total stool number
1
throughout the trial by patient 1
in each study group. Each 0 0
point represents the mean

8
e
1

8
e

in
in

k
k

ee

ee

ee

ee
el
ee

ee

ee

ee
el
value of the previous week at

s
s

W
W

Ba
Ba
every visit. HYDROXYPROPYL CELLULOSE
CHOLESTYRAMINE

10 10

9 P < 0.0005 Initial vs final 9 P < 0.0005 Initial vs final

8 8

Daily watery stool number

Daily watery stool number

7 7

6 6

5 5

4 4

3 3
Figure 4 | Evolution of the
2 2
daily watery stool number
throughout the trial by patient 1 1
in each study group. Each 0 0
point represents the mean
e
e

8
1

lin
lin

k
k

ee

ee

ee

ee
ee

ee

ee

ee

value of the previous week at

se
se

W
W

Ba
Ba

every visit. CHOLESTYRAMINE HYDROXYPROPYL CELLULOSE

there were significant improvements in symptoms, physi- colestyramine (abdominal distension and dyspepsia). No
cal and emotional scales; in the hydroxypropyl cellulose patients experienced serious adverse events.
group, the significant differences were in the symptom
and social dysfunction scales (Table 3). There were no DISCUSSION
significant differences between patients with moderate- This is the first controlled trial of the efficacy of colestyr-
to-severe and severe BAM (data not shown). amine in patients with chronic watery diarrhoea and
BAM. Though initially the trial was designed as a pla-
Safety cebo-controlled RCT, we realised that hydroxypropyl cel-
The rate of adverse events was higher in the colestyr- lulose may be an active drug in treating BAM. In fact,
amine than in the hydroxypropyl cellulose group: eight hydroxypropyl cellulose is a food additive that acts as a
adverse events (in six patients) (61.5%) vs. two adverse thickener and emulsifier, and conceivably might have
events (15.4%) (P = 0.041) (Table 4). However, only two some bulking effect. Moreover, it has been previously
adverse events had a probable causal relationship with shown in a proof-of-concept study published only in

Aliment Pharmacol Ther 2015; 41: 1132–1140 1137

ª 2015 John Wiley & Sons Ltd
F. Fern
andez-Ban
~ares et al.

%SeHCAT 7 day retention

Table 3 | Effect of study drugs on GIQLI scores in the
(a) <20% <15% <12% <10% <5%
(n = 24) (n = 20) (n = 19) (n = 16) (n = 12) PP population
0
CLT Colestyramine Hydroxypropyl
Mean percent decrease

–10 PLA *P = 0.047 (n = 11) cellulose (n = 10)*

–20
GIQLI
–30 Initial 119.7  6.2 116.1  8.9
–40 Final 146.5  4.9 135.9  5.9
* P value <0.0005 0.014
–50 Symptoms
–60 Initial 3.33  0.16 3.27  0.21
Final 4.10  0.14 3.81  0.12
–70
P value <0.0005 0.014
<20% <15% <12% <10% <5%
Physical
(b)
(n = 24) (n = 20) (n = 19) (n = 16) (n = 12) Initial 2.84  0.25 2.96  0.39
–50 Final 3.78  0.18 3.42  0.33
CLT P value <0.0005 0.07
*P = 0.045
Mean percent decrease

–60 PLA Emotional

Initial 3.64  0.22 3.28  0.21
–70 Final 4.04  0.19 3.73  0.24
P value 0.011 0.076
Social
–80
Initial 3.66  0.21 3.53  0.32
* Final 4.09  0.22 4.13  0.15
*
–90 P value 0.068 0.015

–100 * The final GIQLI questionnaires of three patients were not

valid.

Figure 5 | Effect of study drugs on the daily stool

number depending on the SeHCAT 7-day retention: (a)
Daily total stool number; (b) Daily watery stool differences between groups, with also an alpha value of
number. 5% and a statistical power of 80%, if the sample size is
calculated on the basis of a reduction on daily stool
number, with a hypothesised difference between groups
abstract form, to be effective in reducing symptoms in of 15% and the observed variance.
bile acid-induced watery diarrhoea.7 Regrettably, we do Colestyramine was significantly superior to hydroxy-
not become aware of that study until finishing the pres- propyl cellulose in reducing the daily stool number,
ent trial. This has to be taken into account when inter- mainly the watery stool number. This parameter is prob-
preting the results of the present trial. ably more accurate than stool frequency alone to differ-
The results showed that colestyramine was better than entiate between active intervention and placebo in
hydroxypropyl cellulose, but the differences only reached patients with chronic watery diarrhoea, similar to what
statistical significance for the secondary end-point reduc- occurred in a recent trial in patients with collagenous
tion in liquid stools, and not for our primary end-point colitis.5 In microscopic colitis, watery stools are a major
of remission rate. The remission rate after colestyramine determinant of quality of life.8 Likewise, the improve-
observed in the PP population was similar to that ment in both GIQLI and symptom scale in the present
reported in systematic reviews.2, 3 However, we failed to study was more marked after colestyramine than
note a statistically significant difference due to the previ- placebo.
ously unexpectedly high ‘placebo’ response rate. With SeHCAT 7-day retention is a continuous variable, and
the present response rates of 64% and 38.5% for the the debate has been whether the cut-offs of 10% or 15%
colestyramine and hydroxypropyl cellulose groups, (or even 20%) should be used in the prediction of
respectively, a sample of 60 patients per group would be response to bile acid sequestrants.2, 3, 9 All patients
required to yield a statistically significant result with an included in the present trial had low SeHCAT values
alpha value of 5% and a statistical power of 80%. How- (≤20%), lower than those previously described in healthy
ever, the present sample size is enough to find statistical controls (30–40% 7-day retention).10, 11 However, it has

1138 Aliment Pharmacol Ther 2015; 41: 1132–1140

ª 2015 John Wiley & Sons Ltd
 Randomised clinical trial: colestyramine in chronic watery diarrhoea

shown to preferentially bind dihydroxy bile acids and

Table 4 | Adverse events for each group
reduce the secondary bile acid pool, but in addition, it
Colestyramine Hydroxypropyl elevates the ratio of glycine- vs. taurine-conjugated bile
Adverse event (n = 13) cellulose (n = 13) acids in bile. Because glycine conjugates are more subject
Total 8 (61.5%) 2 (15.4%)* to passive absorption, this may also contribute to reduce
Causal relationship 2 (15.4%) 0 malabsorption.17 Besides, colestyramine is a strong
with study drug anion-exchange resin which can bind other compounds
Headache 3 0
Muscle pain 0 1
in the colon lumen, like bacterial toxins and mycotox-
Acute gastroenteritis 1 0 ins.18–21
Nasopharyngitis 1 1 Our study confirms the safety of short-term colestyr-
Abdominal distension 1 0 amine treatment, with only minor adverse events with a
Dyspepsia (heartburn) 1 0
causal relationship with the study drug. Reported side
Nausea/vomiting† 1 0
effects in the literature included abdominal bloating and
* P = 0.041.
pain, dyspepsia, nausea and vomiting, flatulence, bor-
† After withdrawal from the study, this patient tolerated coles- borygmi, abdominal distension, constipation and diar-
tyramine well in an open-label form, and thus this adverse
event was considered to be unrelated to the drug. rhoea increasing in severity.2 The present study reveals
that minor adverse events associated with colestyramine
use appeared in only 15% of patients.
In summary, our study confirms that colestyramine is
been suggested that diarrhoea per se might induce mal- effective and safe for short-term treatment of patients
absorption of orally ingested (exogenous) radiolabelled with functional watery chronic diarrhoea. Besides,
bile acid (14C-taurocholate) through less efficient ileal the high rate of response to hydroxypropyl cellulose sug-
absorption.12 In this sense, diarrhoea induced in healthy gests that this compound is also active in bile acid
subjects by intake of a polyethylene glycol-containing diarrhoea.
electrolyte solution produced a decrease in 7-day SeH-
CAT retention.10 Thus, BAM may be a secondary effect AUTHORSHIP
of accelerated transit by diarrhoea itself rather than the Guarantor of the article: Fern
andez-Ba~
nares, Esteve.
cause of diarrhoea in many instances. Although BAM is Author contributions: Fern
andez-Ba~ nares, Esteve: study
a known cause of diarrhoea, only aqueous concentrations concept and design. Fern
andez-Ba~ nares, Rosinach, Piqu-
of dihydroxy bile acids in the colon greater than 3 mM eras, Ruiz-Cerulla, Modolell, Zabana, Guardiola, Esteve:
can induce it.13 Several mechanisms including water and acquisition of data. Fern
andez-Ba~ nares, Esteve: analysis
sodium secretion, and increases of both epithelial perme- and interpretation of data. Fern
andez-Ba~ nares, Esteve:
ability and intestinal motility have been involved.14, 15 drafting of the manuscript. Fern
andez-Ba~ nares, Rosinach,
Measurement of faecal bile acid concentrations requires, Piqueras, Ruiz-Cerulla, Modolell, Zabana, Guardiola,
unfortunately, expensive analytical techniques and con- Esteve: critical revision of the manuscript for important
siderable expertise, which limits its use. Thus, it has been intellectual content. Fern
andez-Ba~
nares: statistical analy-
stated that the contribution of BAM to diarrhoea in a sis. Fern
andez-Ba~ nares: study supervision. All authors
given patient would be determined by a therapeutic trial have approved the final version of the manuscript.
of colestyramine.16 In the present trial, colestyramine
induced diarrhoea cessation in 64% of patients who tol- ACKNOWLEDGEMENTS
erate it, but its efficacy was similar in those patients with The ‘Centro de Investigaci
on Biom
edica en Red de En-
SeHCAT 7-day retention cut-offs of 5% or 10% than in fermedades Hep
aticas y Digestivas’ (CIBERehd) is an ini-
the whole series. These data are in agreement with a sys- tiative of the Instituto de Salud Carlos III, Madrid,
tematic review and with a recent study evaluating the Spain. The authors are very grateful to Dr Julian R.
effect of colestipol in bile acid-induced diarrhoea both Walters (Imperial College Healthcare NHS Trust, Lon-
suggesting that response rate to bile acid sequestrants is don) by his expert review and comments. Also we are
not dependent on the severity of BAM.2, 9 Thus, the very grateful for helpful assistance during the trial from
effect of colestyramine in improving diarrhoea could be Dr Susana Redondo (Pharmacy Department, Hospital
by another mechanism other than by binding malab- Universitary Mutua Terrassa).
sorbed bile acids. In this sense, colestyramine has been Declaration of personal interests: None.

Aliment Pharmacol Ther 2015; 41: 1132–1140 1139

ª 2015 John Wiley & Sons Ltd
F. Fern
andez-Ban
~ares et al.

Declaration of funding interests: This study was funded This sponsor had no role in the study design, acquisi-
by a grant from the Spanish Ministry of Health for Inde- tion, analysis or interpretation of the data, or the report
pendent Drug Research (grant number: EC08/00085). writing.

REFERENCES
1. Wedlake L, A’Hern R, Russell D, acid induced watery diarrhoea. Gut patogenesis and therapy. Nat Rev
Thomas K, Walters JR, Andreyev HJ. 2003; 52(Suppl. 1): A116. Gastroenterol Hepatol 2014; 11: 426–34.
Systematic review: the prevalence of 8. Hjortswang H, Tysk C, Bohr J, et al. 16. Fromm H, Thomas PJ, Hofmann AF.
idiopathic bile acid malabsorption as Health related quality of life is impaired Sensitivity and specificity in tests of
diagnosed by SeHCAT scanning in in active collagenous colitis. Dig Liv Dis distal ileal function: prospective
patients with diarrhoea-predominant 2011; 43: 102–109. comparison of bile acid and vitamin
irritable bowel syndrome. Aliment 9. Bajor A, T€ ornblom H, Rudling M, Ung B12 absorption in ileal resection
Pharmacol Ther 2009; 30: 707–17. KA, Simr
en M. Increased colonic bile patients. Gastroenterology 1973; 64:
2. Wilcox C, Turner J, Green J. Systematic acid exposure: a relevant factor for 1077–90.
review: the management of chronic symptoms and treatment in IBS. Gut 17. Garbutt JT, Kenney TJ. Effect of
diarrhoea due to bile acid 2015; 64: 84–92. colestyramine on bile acid metabolism
malabsorption. Aliment Pharmacol Ther 10. Fellous K, Jian R, Haniche M, et al. in normal man. J Clin Invest 1972; 51:
2014; 39: 923–39. Mesure de l’absorption il
eale des sels 2781–9.
3. Riemsma R, Al M, Corro Ramos I, biliaires par le test a l’homotaurocholate 18. Weiss K. Toxin-binding treatment
et al. SeHCAT [tauroselcholic marqu
e au s
el
enium 75. Validation et for Clostridium difficile: a review
(selenium-75) acid] for the investigation signification clinique. Gastroenterol Clin including reports of studies with
of bile acid malabsorption and Biol 1994; 18: 865–72. tolevamer. Int J Antimicrob Agents
measurement of bile acid pool loss: a 11. Merrick MV, Eastwood MA, Ford MJ. 2009; 33: 4–7.
systematic review and cost-effectiveness Is bile acid malabsorption 19. Mullan NA, Burgess MN, Bywater RJ,
analysis. Health Technol Assess 2013; underdiagnosed? An evaluation of Newsome PM. The ability of
17:1–236. accuracy of diagnosis by measurement colestyramine resin and other
4. Fern
andez-Ba~ nares F, Esteve M, Salas of SeHCAT retention. BMJ 1985; 290: adsorbents to bind Escherichia coli
A, et al. Systematic evaluation of the 665–8. enterotoxins. J Med Microbiol 1979; 12:
causes of chronic watery diarrhea with 12. Schiller LR, Hogan RB, Morawski SG, 487–96.
functional characteristics. Am J et al. Studies of the prevalence and 20. Solfrizzo M, Visconti A, Avantaggiato
Gastroenterol 2007; 102: 2520–8. significance of radiolabeled bile acid G, Torres A, Chulze S. In vitro and
5. Miehlke S, Madisch A, Kupcinskas L, malabsorption in a group of patients in vivo studies to assess the
et al. ; BUC-60/COC Study Group. with idiopathic chronic diarrhea. effectiveness of colestyramine as a
Budesonide is more effective than Gastroenterology 1987; 92:151–60. binding agent for fumonisins.
mesalamine or placebo in short-term 13. Fromm H, Malavolti M. Bile acid Mycopathologia 2001; 151:
treatment of collagenous colitis. induced diarrhea. Clin Gastroenterol 147–53.
Gastroenterology 2014; 146: 1222–30. 1986; 15: 567–82. 21. Hope JH, Hope BE. A review of the
6. Quintana JM, Cabriada J, de L
opez 14. Camilleri M. Advances in diagnosis and treatment of Ochratoxin
Tejada I, et al. Translation and understanding of bile acid diarrhea. A inhalational exposure associated with
validation of the gastrointestinal quality Expert Rev Gastroenterol Hepatol 2014; human illness and kidney disease
of life index (GIQLI). Rev Esp Enferm 8: 49–61. including focal segmental
Dig 2001; 93: 693–706. 15. Walters JR. Bile acid diarrhoea and glomerolosclerosis. J Environ Public
7. Brydon G, Ganguly R, Ghosh S. The FGF19: new views on diagnosis, Health 2012; 2012: 835059.
effect of hydroxypropyl cellulose on bile

1140 Aliment Pharmacol Ther 2015; 41: 1132–1140

ª 2015 John Wiley & Sons Ltd