Aliment Pharmacol Ther 2003; 17: 895–904.

doi: 10.1046/j.0269-2813.2003.01543.x

A randomized controlled trial of a probiotic, VSL#3, on gut transit and symptoms in diarrhoea-predominant irritable bowel syndrome
H. J. K IM, M. CAM ILLERI, S. M C KINZIE, M. B. LEMPKE, D. D. BURTO N, G. M. THOMFORDE & A. R. ZINSMEISTER Clinical Enteric Neuroscience Translational & Epidemiological Research Program, Mayo Clinic and Mayo Foundation, Rochester, MN, USA
Accepted for publication 3 February 2003

SUMMARY

Aim: To investigate the effects of a probiotic formulation, VSL#3, on gastrointestinal transit and symptoms of patients with Rome II irritable bowel syndrome with predominant diarrhoea. Methods: Twenty-five patients with diarrhoea-predominant irritable bowel syndrome were randomly assigned to receive VSL#3 powder (450 billion lyophilized bacteria/day) or matching placebo twice daily for 8 weeks after a 2-week run-in period. Pre- and posttreatment gastrointestinal transit measurements were performed in all patients. Patients recorded their bowel function and symptoms daily in a diary during the 10-week study, which was powered to detect a 50% change in the primary colonic transit end-point. Results: There were no significant differences in mean gastrointestinal transit measurements, bowel function

scores or satisfactory global symptom relief between the two treatment groups, pre- or post-therapy. Differences in abdominal bloating scores between treatments were borderline significant (P ¼ 0.09, analysis of covariance). Further analysis revealed that abdominal bloating was reduced (P ¼ 0.046) with VSL#3 [mean post- minus pre-treatment score, ) 13.7; 95% confidence interval (CI), ) 2.5 to ) 24.9], but not with placebo (P ¼ 0.54) (mean post- minus pre-treatment score, ) 1.7; 95% CI, 7.1 to ) 10.4). With the exception of changes in abdominal bloating, VSL#3 had no effect on other individual symptoms: abdominal pain, gas and urgency. All patients tolerated VSL#3 well. Conclusion: VSL#3 appears to be promising in the relief of abdominal bloating in patients with diarrhoeapredominant irritable bowel syndrome. This is unrelated to an alteration in gastrointestinal or colonic transit.

INTRODUCTION

Probiotics are viable micro-organisms that have a beneficial effect in the prevention of recurrent pouchitis and Clostridium difficile diarrhoea.1 A relatively large amount of clinical experience and some literature data support the use of probiotics to prevent or treat intestinal disorders. However, the scientific basis for
Correspondence to: Dr M. Camilleri, Mayo Clinic, Charlton 8–110, 200 First St. S.W., Rochester, MN 55905, USA. E-mail: camilleri.michael@mayo.edu Ó 2003 Blackwell Publishing Ltd

probiotic use is less firmly established, and sound clinical and mechanistic studies are required. Organisms used as probiotics in humans include Lactobacillus sp., Bifidobacterium sp., Escherichia coli, Streptococcus sp., Enterococcus sp., Bacteroides sp., Bacillus sp., Propionibacterium sp. and various fungi. Some probiotic preparations contain mixtures of more than one bacterial strain. Probiotics appear to be useful in the prevention or treatment of several gastrointestinal disorders, including infectious diarrhoea, antibiotic diarrhoea and traveller’s diarrhoea.1 There is anecdotal clinical
895

896

H. J. KIM et al.

evidence, based on open trials, suggesting that the symptoms of patients with diarrhoea-predominant irritable bowel syndrome improve with probiotic therapy; however, no formal randomized trial has yet demonstrated an overall efficacy in irritable bowel syndrome or identified the mechanism of action of the probiotic. In a 4-week randomized controlled trial, 5 · 107 colonyforming units/mL of Lactobacillus plantarum and 3.6 g of oat flour (fibre) were compared with placebo (rose hip syrup) in 60 patients with irritable bowel syndrome; flatulence was significantly lower in the Lactobacillustreated group than in the placebo-treated group. Abdominal pain was lower than at baseline in both groups, but no difference between the two treatments was observed. Overall gastrointestinal function at 1 year was reported to be significantly better in the active (probiotic) treatment group.2 The clinical efficacy reported in diarrhoea-predominant irritable bowel syndrome and in acute diarrhoea, such as infectious and travellers’ diarrhoea,3–8 suggests that these agents may retard colonic transit and facilitate fluid and electrolyte re-absorption, thereby relieving diarrhoea. Symptoms of abdominal pain, bloating, flatulence and diarrhoea are commonly seen in patients with irritable bowel syndrome, and abdominal pain and bloating are difficult to treat and have a significant impact on patients’ lives. Hahn et al. examined symptom frequency, duration and severity in 122 adult patients with irritable bowel syndrome; on average, patients reported moderate to severe pain/discomfort on 33% of days, and severe bloating on 28% of days. In contrast, altered stool form was recorded on 25% of days. The duration of each episode of bloating and abdominal pain was 5 days on average, with 1–4 episodes occurring per month.9 These results confirm the relevance of symptoms such as pain and bloating, which present an unmet clinical need. We undertook a placebo-controlled study to determine the efficacy of VSL#3, a patented probiotic preparation, on symptoms and gastrointestinal and colonic transit in patients with diarrhoea-predominant irritable bowel syndrome.

child-bearing potential with negative pregnancy blood test within 48 h of each transit study; present or previous chronic gastrointestinal symptoms suggestive of irritable bowel syndrome using the Rome II criteria; diarrhoea-predominant symptoms; no previous abdominal surgery except appendectomy, caesarean section, tubal ligation, abdominal wall hernia repair or laparoscopic cholecystectomy; no history of non-functional bowel disease; no current use of medications that may alter gastrointestinal motility; no over-the-counter medication use (other than those specifically screened by the investigator) within 3 days of transit studies; no antibiotic use within 2 weeks prior to the 2-week runin period; previous colonic transit measured at Mayo Clinic Rochester of a geometric centre of ‡ 2.67 at 24 h or ‡ 3.90 at 48 h, which are the mean values of colonic transit in healthy volunteers (n ¼ 37).10 All participants provided written informed consent according to institutional guidelines, and the protocol was approved by the Mayo Clinic’s Institutional Review Board. Twenty-nine participants with diarrhoea-predominant irritable bowel syndrome were recruited from the clinical practice of the principal investigator or by letter. The latter patients have been evaluated previously at the Mayo Clinic and are thus included in an administrative database of patients with irritable bowel syndrome residing within a 200-ml distance from the Mayo Clinic Rochester. After indicating their willingness to participate, they filled out a bowel disease questionnaire11 to confirm the persistence of chronic gastrointestinal symptoms suggestive of diarrhoea-predominant irritable bowel syndrome. Experimental design We performed a parallel-group, double-blind, placebocontrolled study with a total of 25 subjects randomized to receive either placebo (n ¼ 13) or VSL#3 (n ¼ 12). A 2-week run-in observation period was followed by an 8-week treatment period. During this entire 10-week period, participants were required to record a weekly response to a question on the satisfactory relief of irritable bowel syndrome symptoms and a daily diary of bowel function [frequency, consistency (using a validated Bristol stool form scale12–15) and ease of passage (seven-point adjectival scale from manual disimpaction to incontinence16)]. A daily diary
Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

METHODS

Eligibility The eligibility criteria were as follows: age of 18– 75 years, both males and females; females of

VSL#3 IN PATIENTS WITH DIARRHOEA-PREDOMINANT IBS

897

was also used to appraise other symptom scores (abdominal pain, bloating, flatulence and urgency, using a 100-mm visual analogue scale). Participants received either the study agent (VSL#3) or placebo in powder form (see components below) twice daily for 8 weeks. Participants whose baseline transit study showed a geometric centre of less than 2.67 at 24 h and less than 3.90 at 48 h were excluded. These values correspond to the mean values of colonic transit in healthy controls. Treatment plan with VSL#3 or placebo Randomization was developed and maintained by the research pharmacist (MBL) with all investigators, including the biostatistician (ARZ), blind to the details. Patients were assigned to either VSL#3 or placebo prior to performing the baseline transit study. VSL#3 is a composite probiotic containing multiple strains of three viable lyophilized bacteria species: three strains of Bifidobacterium (B. longum, B. infantis and B. breve); four strains of Lactobacillus (L. acidophilus, L. casei, L. bulgaricus and L. plantarum); and one strain of Streptococcus (S. salivarius subspecies thermophilus). Each VSL#3 packet contained 225 billion lyophilized bacteria in a powder form miscible in yoghurt or soluble in water. Each participant in the VSL#3 treatment group received one packet orally, twice daily (450 billion lyophilized bacteria per day), for 8 weeks. Placebo powder containing starch, which looked identical to the study agent, was supplied by VSL Pharmaceuticals, Inc., Fort Lauderdale, FL, USA. Each participant in the placebo group received one packet of placebo powder, twice daily, for 8 weeks. Assessment of symptoms of irritable bowel syndrome Participants filled in a daily diary to evaluate the symptoms of irritable bowel syndrome (bowel function and symptom scores) and to track the intake of all medications during the trial. Participants also reported, on a weekly basis, their response to the following question: ‘During the past 7 days, have you experienced satisfactory relief of your irritable bowel syndrome symptoms?’. This question was posed at the end of each week during the 2-week run-in observation period, during which no treatment was administered, and during the 8-week treatment
Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

period, during which double-blind treatment was received. A participant who reported at least four of the eight weekly assessments (during the treatment period) with satisfactory relief was declared a responder, as in recent phase II and III trials of new pharmaceutical agents in irritable bowel syndrome.17, 18 Assessment of gastrointestinal and colonic transit An established scintigraphic method was used.19–22 Briefly, 111In adsorbed on activated charcoal particles was delivered to the colon by a methacrylate-coated, delayed-release capsule.23 Simultaneously, 99mTc-sulphur colloid was used to label two scrambled eggs, which were ingested with one slice of whole-wheat bread and one glass of whole milk (300 kcal) to facilitate the measurement of gastric and small bowel transit. Subjects ingested standardized meals for lunch and dinner. A variable region of interest program was used to measure transit as in previous studies. We obtained abdominal images every hour for the first 4 h, and scans at 6, 8, 24, 32 and 48 h. The primary summaries for the comparison of transit profiles were: gastric residual at 2 and 4 h; colonic filling at 6 h; and colonic geometric centre at 24 and 48 h. The geometric centre is the weighted average of counts in the different colonic regions [ascending (AC), transverse (TC), descending (DC), rectosigmoid (RS)] and stool (weighting factor, 1–5, respectively). Thus, at any time, the proportion of colonic counts in each colonic region was multiplied by its weighting factor as follows: ð%AC  1 þ %TC  2 þ %DC  3 þ %RS  4 þ %stool  5Þ=100 ¼ geometric centre

Concomitant medications The following medications were permitted during the course of the study, as long as they were used at constant dosage and had been commenced at least 1 month prior to the start of the 2-week run-in observation period: birth control pill or depot intramuscular contraceptive preparation; oestrogen–progesterone replacement therapy; l-thyroxine; low-dose antidepressants (up to 25 mg/day of amitriptyline or selective serotonin re-uptake inhibitor); antihypertensives in the diuretic; angiotensin-converting enzyme inhibitor or angiotensin II inhibitor classes.

898

H. J. KIM et al.

DATA ANALYSIS

Symptoms The primary analysis for the assessment of the satisfactory relief of irritable bowel syndrome was a comparison of the proportion of responders treated with placebo vs. VSL#3. Responders were defined as individuals whose response to the weekly satisfactory relief question was yes on at least four of the eight weekly assessments of the treatment phase. The secondary analyses compared the mean scores of bowel function and symptoms. The daily diary was used to appraise the mean bowel function scores (frequency, consistency and ease of passage) and the severity of individual symptoms (abdominal pain, bloating, flatulence and urgency). The severity of individual symptoms was assessed using a 100-mm visual analogue scale. The adjectival scales for stool consistency and ease of passage were recorded as numerical values. Transit The primary end-point of the study for gastrointestinal transit was the geometric centre of colonic counts at 24 h, with secondary analyses of interest being gastric emptying at 2 and 4 h, colonic filling at 6 h and the colonic geometric centre at 48 h. Statistical analysis Our study was designed to include 24 patients. With a two-sided alpha level of 0.05, the study had an estimated 80% power to detect a change of approximately 50% in the geometric centre measurement at 24 h and a change of 80% in colonic filling at 6 h. All eligible patients were included in the analyses of gastrointestinal transit, satisfactory relief and stool and symptom scores according to the intention-to-treat principle. The data on the satisfactory relief of irritable bowel syndrome, bowel function scores and visual analogue scale scores of individual symptoms are summarized as the means and standard errors in each treatment group. The visual analogue scale symptom scores are rounded to the nearest integer in millimetres. Post-treatment colonic transit at 24 and 48 h and symptom and bowel function scores were compared between the two treatment groups using an analysis of covariance (ancova), with the baseline geometric centre at 24 and 48 h and symptom and bowel function scores, respectively, as covariates. ancova was used to

account for the potential effects of differences in the mean baseline measurements and scores between the two treatment groups. The remaining transit summaries were compared between the two treatment groups using a two-sample t-test or Wilcoxon rank sum test, as warranted. The proportions of responders were compared by Fisher’s exact test. Pre- vs. post-treatment changes in bowel function and symptom scores for each treatment group were assessed using the signed rank test with Pratt’s modification for observed differences of zero.24

RESULTS

Demographic characteristics Between 1 July 2001 and 1 December 2001, 29 patients were screened. Three patients (14%) were ineligible as they did not meet the baseline colonic transit criteria, and one patient withdrew consent. Of the remaining 25 patients, 12 were randomly assigned to VSL#3 and 13 to placebo. Demographic factors (Table 1) were similar between the two groups; patients in the VSL#3-treated group were slightly older than those in the placebo group. There were 10 females in the VSL#3 group and eight females in the placebo group (P ¼ N.S.). The baseline geometric centres at 24 and 48 h were greater than 2.67 and 3.90, respectively, in all participants, in association with their diarrhoeapredominant irritable bowel syndrome. The median duration of diarrhoea-predominant irritable bowel syndrome was 8 years (range, 2–41 years) in the VSL#3treated group and 6 years (range, 1–22 years) in the placebo group (Table 1). Treatment: compliance and concomitant medications Of the 12 participants assigned to the VSL#3 group, all completed treatment as planned. One participant, eventually shown to have been assigned to placebo, withdrew from the study due to an exacerbation of typical abdominal pain associated with irritable bowel syndrome. This patient was included in the efficacy analysis based on the intention-to-treat principle; responder status was imputed as ‘no relief’ and the bowel and symptom scores were imputed using the mean value of all study participants with corresponding data values. A total of 24 participants (12 in the VSL#3 group and 12 in the placebo group) completed the study.
Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

VSL#3 IN PATIENTS WITH DIARRHOEA-PREDOMINANT IBS
Table 1. Characteristics of the patients and the baseline gastrointestinal transit, bowel function scores and symptom scores*

899

Characteristic Age (years) Mean Range Females (n) Duration of IBS (years) Baseline colonic transit GC24 (mean) GC48 (mean) Bowel function score Frequency (mean) Consistency Ease of passageà Symptom score (mm) Abdominal pain (mean) Bloating (mean) Flatulence (mean) Faecal urgency (mean)

VSL#3 group (n ¼ 12) 48 ± 5.7 19–70 10 8 (range, 2–41) 3.4 ± 0.4 4.7 ± 0.1 2.4 ± 0.3 4.0 ± 0.4 4.5 ± 0.2 34 37 36 46 ± ± ± ± 7 8 7 7

Placebo group (n ¼ 13) 38 ± 3.4 19–59 8 6 (range, 1–22) 3.4 ± 0.2 4.7 ± 0.1 2.4 ± 0.4 3.4 ± 0.4 4.1 ± 0.2 24 30 31 30 ± ± ± ± 6 7 5 5

P value N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S.

GC24/48, geometric centre at 24 and 48 h; IBS, irritable bowel syndrome; N.S., not significant. * Age, baseline transit, bowel function and symptom scores are summarized as means ± standard error in each treatment group. Used a scale of 1–7: 1, watery stool; 7, hard, lumpy stool (Bristol stool form scale). à Used a scale of 1–7: 1, faecal incontinence; 7, manual disimpaction.

Five of the 24 participants required antibiotic therapy during the study period for infection (upper respiratory infection, subacute bacterial endocarditis [SBE] prophylaxis for dental work and cellulitis). Four were in the placebo group and one in the VSL#3 group. Patients were asked to stop their study agent and to record their bowel function and symptom scores during antibiotic treatment, plus 7 days after treatment, to allow for a sufficient antibiotic wash-out period. They then resumed the study agent and daily diaries to fulfil a total of 8 weeks of treatment (not including the time for antibiotic treatment). Five of 24 patients used a loperamide tablet as predefined rescue medication for severe diarrhoea during the 10-week trial (two from the VSL#3 group and three from the placebo group). The average use of loperamide in the two groups was two tablets (range, 1–3). Other medications used by the patients during the 10-week trial were: metamucil (one patient for 1 day); rofecoxib and acetaminophen with codeine phosphate (one patient for 2 days); contrast medium for computed tomography scan (one patient); prednisone for the exacerbation of asthma (one patient for 5 days); bismuth subsalicylate (one patient for 1 day); milk of magnesia (one patient for 1 day); triamcinolone acetoÓ 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

nide inhaler, benzonatate perles and guaifenesin cough syrup (one patient for 5 days); extra strength acetaminophen and promethazine (1 patient for 2 days). Toxicity There were no adverse events noted with VSL#3 treatment. Gastrointestinal transit The mean baseline geometric centres at 24 and 48 h were similar in the two groups (P ¼ N.S. for both; Table 1). The post-treatment gastric, small bowel and colonic transit measurements were not significantly different between the two treatment groups (Table 2). There was an unexpected trend for the placebo group to decelerate colonic transit at 48 h (P ¼ 0.07) (mean geometric centre: pre-treatment, 4.7; post-treatment, 4.4). Satisfactory relief (Figure 1) The proportions of responders (satisfactory relief of overall irritable bowel syndrome symptoms on at least four of the eight weekly assessments) were 38% (five of

900

H. J. KIM et al.

Table 2. Post-treatment gastrointestinal transit measurements*
0.6

Run-in Weekly proportion of responders, ITT

Randomized treatment VSL#3

Gastrointestinal transit

VSL#3 group Placebo group (n ¼ 12) (n ¼ 13) P value 5 2 6 0.4 0.3 61 97 67 3.2 4.4 ± ± ± ± ± 6 2 5 0.3 0.2 0.41 0.86 0.62 0.70 0.99

0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7

PLACEBO

GE 2 h (% emptied) 54 ± GE 4 h (% emptied) 97 ± Colonic filling 6 h (%) 72 ± 3.4 ± Post-treatment GC24 4.5 ± Post-treatment GC48

GC24/48, geometric centre at 24 and 48 h; GE, gastric emptying. * Gastrointestinal transit measurements are summarized as means ± standard error in each treatment group. An analysis of covariance (ancova), with baseline GC24 and GC48 as covariates, was used to compare the post-treatment GC24 and GC48 values between the two groups. Analysis was based on the intentionto-treat principle. The other transit summaries were compared using a two-sample t-test or Wilcoxon rank sum test, as warranted.

8

9

10

Weeks

Figure 2. Plot showing the weekly proportions of responders amongst all eligible patients using the intention-to-treat (ITT) principle, according to treatment group assignment. The difference in weekly proportions of responders between the two groups was not significant.

Proportion of patients (%) with satisfactory relief (responder) of IBS, ITT

100

75

P = 1.00
50

overall bowel function aggregate score gave P ¼ 0.59 and P ¼ 0.68 for placebo and VSL#3, respectively, for within-group pre- vs. post-treatment changes (Table 3). Other symptom assessments

25

0 VSL#3 Placebo

Figure 1. Proportion of responders [satisfactory relief of overall irritable bowel syndrome (IBS) symptoms on at least four of the eight weekly assessments] amongst all eligible patients using the intention-to-treat (ITT) principle, according to treatment group assignment. Thirty-three per cent (four of 12) and 38% (five of 13) of patients in the VSL#3 and placebo groups, respectively, were responders. The difference in the proportion of responders between the two groups was not significant (P ¼ 1.00; Fisher’s exact test).

13) in the placebo group and 33% (four of 12) in the VSL#3 group (P ¼ 1.00 based on intention-to-treat analysis). Weekly assessment of the proportion of responders between the two groups also failed to detect any significant differences (Figure 2). Bowel function The mean post-treatment stool frequency, consistency and ease of passage for the VSL#3 and placebo groups were similar (P ¼ N.S.; ancova; Table 3). Moreover, no significant within-group changes were indicated. The

At baseline, the VSL#3-treated patients tended to have higher symptom scores (abdominal pain, bloating, flatulence and faecal urgency) than those in the placebo group (Table 1); therefore, baseline scores were incorporated as covariates to compare the post-treatment symptoms between the two groups using ancova. The adjusted mean post-treatment scores for VSL#3 and placebo are shown in Table 3. The comparison between the two groups using ancova indicated a borderline statistically significant improvement with VSL#3 treatment for abdominal bloating (P ¼ 0.09). Abdominal bloating was reduced (P ¼ 0.046) in the VSL#3 group [mean post- minus pre-treatment score, ) 13.7; 95% confidence interval (CI), ) 2.5 to ) 24.9], but not in the placebo group (P ¼ 0.54) (mean postminus pre-treatment score, ) 1.7; 95% CI, 7.1 to ) 10.4) (Figure 3). Within-group changes for urgency were borderline different from zero for the VSL#3 group (P ¼ 0.056), but not for the placebo group (P ¼ 0.12). The scores for abdominal pain and flatulence were not significantly reduced in patients treated with VSL#3. The overall symptom scores for within-group pre- vs. post-treatment changes gave P values of 0.31 and 0.08 for placebo and VSL#3, respectively (Table 3).
Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

VSL#3 IN PATIENTS WITH DIARRHOEA-PREDOMINANT IBS
Table 3. Comparison of bowel function and symptom scores Between groups (post-treatment) Adjusted mean (± S.E.) Pre- vs. post- changes (within groups) Placebo Median D Mean D (95% CI) ) 0.1 ) 0.1 () 0.0 ) 0.1 () 0.0 ) 0.1 () ) 0.4 ) 0.3 () ) 2.0 ) 2.0 ) 1.0 ) 1.0 ) 4.0 ) 0.0 ) 6.0 ) 4.0 ) 10.0 ) 7.0 VSL#3 Median D Mean D (95% CI) 0.0 0.3 0.1 0.1 0.2 0.1 0.0 ) 0.5 ) ) ) ) ) ) 8.0 ) 14.0 ) 5.0 ) 6.0 ) 6.0 ) 8.0 ) 4.0 ) 8.0 ) 27.0 ) 35.0

901

Placebo Bowel function Frequency Consistencyà Ease of passage§ Aggregate score Symptoms– Bloating Flatulence Pain Urgency Overall score 2.3 ± 0.2 3.5 ± 0.2 4.0 ± 0.1 9.8 ± 0.4

VSL#3 2.1 ± 0.2 3.7 ± 0.2 4.3 ± 0.1 10.2 ± 0.4

P* 0.45 0.30 0.15 0.51

P 0.65

P 0.34

0.6, 0.4) 0.89 0.5, 0.4) 0.50 0.6, 0.3) 0.59 1.5, 0.9) 0.54

() 0.8, 0.2) 0.85 () 0.7, 0.5) 0.62 () 0.5, 0.3) 0.68 () 1.8, 0.8) 0.05 () 25, ) 2) 0.30 () 14, 3) 0.34 () 20, 4) 0.06 () 15, ) 2) 0.08 () 62, 2)

30 ± 4 32 ± 4 26 ± 4 33 ± 3 121 ± 13

21 ± 4 28 ± 4 23 ± 4 30 ± 3 103 ± 14

0.09 0.51 0.60 0.60 0.37

() 10, 7) 0.74 () 10, 8) 0.45 () 9, 8) 0.12 () 11, 3) 0.31 () 38, 24)

CI, confidence interval; S.E., standard error; D, change. * From analysis of covariance (ancova) adjusting for pre-treatment values as the covariate; analysis based on intention-to-treat principle. Wilcoxon signed rank test. à Used a scale of 1–7: 1, watery stool; 7, hard, lumpy stool (Bristol stool form scale). § Used a scale of 1–7: 1, faecal incontinence; 7, manual disimpaction. – Used a visual analogue scale (100 mm).

100 80 VAS scores (mm) 60 40 20 0

VSL#3 (n=12)
P = 0.046

Placebo (n=13)
P = 0.54

Analyses excluding five patients with antibiotic use As five patients received antibiotics for other clinical indications, we took two precautions to evaluate the effect of VSL#3 independent of any confounding effect of antibiotics. First, the data presented above excluded symptom ratings on days during which antibiotics were taken and for a 7-day ‘wash-out’ period after the cessation of antibiotic. Secondly, we performed further analyses excluding all these patients (four in the placebo group and one in the VSL#3 group), and the results were essentially unchanged from the analyses of all participants as shown above (Table 4). There was no demonstrable effect on transit, bowel function, global symptom relief or individual symptoms, other than abdominal bloating (P ¼ 0.05).

Pre-

Post-

Pre-

Post-

Figure 3. Plot of the individual abdominal bloating scores using a visual analogue scale (mm), according to treatment group assignment. Note the significant improvement in the VSL#3 group (post- vs. pre-VSL#3; P ¼ 0.046), but not in the placebo group (post- vs. pre-placebo; P ¼ 0.54).

Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

902

H. J. KIM et al.

Table 4. Comparison of bowel function and symptom scores Between groups (post-treatment) Adjusted mean (± S.E.) Pre- vs. post- changes (within groups) Placebo (n = 9) Median D Mean D (95% CI) ) 0.2 ) 0.1 () 0.0 ) 0.1 () ) 0.1 ) 0.2 () ) 2.0 ) 1.4 () ) 3.0 ) 2.3 ) 4.0 ) 0.6 ) 3.0 ) 0.8 ) 4.0 ) 2.3 ) 10.0 ) 6.4 VSL#3 (n = 11) Median D Mean D (95% CI) 0.1 0.3 0.0 0.0 0.1 0.1 0.0 ) 0.5 ) 7.0 ) 13.8 ) 5.0 ) 5.3 ) 5.0 ) 6.7 ) 3.0 ) 7.0 ) 12.0 ) 32.8 ) ) ) ) ) )

Placebo Bowel function Frequency Consistencyà Ease of passage§ Aggregate score Symptoms– Bloating Flatulence Pain Urgency Overall score 2.6 ± 0.2 3.5 ± 0.2 4.0 ± 0.2 11.1 ± 0.5

VSL#3 2.3 ± 0.2 3.9 ± 0.2 4.4 ± 0.2 11.2 ± 0.5

P* 0.33 0.22 0.12 0.87

P 0.82

P 0.24

0.9, 0.7) 0.91 0.8, 0.5) 0.43 0.9, 0.5) 0.36 4.2, 1.5) 0.52

() 0.9, 0.2) 0.99 () 0.7, 0.6) 0.76 () 0.5, 0.4) 0.90 () 1.9, 1.0) 0.05 () 26.3, ) 1.4) 0.35 () 14.3, 3.8) 0.54 () 19.7, 6.2) 0.11 () 14.0, 0.0) 0.15 () 69.4, 3.8)

32.2 ± 4.8 34.5 ± 5.0 28.3 ± 4.7 33.9 ± 4.1 128.7 ± 17.6

23.5 ± 4.4 30.6 ± 4.5 26.0 ± 4.3 32.3 ± 3.7 113.9 ± 15.9

0.20 0.57 0.72 0.79 0.55

() 15.7, 11.0) 0.76 () 13.2, 12.1) 0.44 () 12.2, 10.6) 0.30 () 12.6, 8.0) 0.38 () 52.4, 39.5)

CI, confidence interval; S.E., standard error; D, change. * From analysis of covariance (ancova) adjusting for pre-treatment values as the covariate; analysis based on intention-to-treat principle. Wilcoxon signed rank test. à Used a scale of 1–7: 1, watery stool; 7, hard, lumpy stool (Bristol stool form scale). § Used a scale of 1–7: 1, faecal incontinence; 7, manual disimpaction. – Used a visual analogue scale (100 mm).

DISCUSSION

Our results demonstrate that, in diarrhoea-predominant irritable bowel syndrome, VSL#3 has no effect on small bowel or colonic transit, does not increase the proportion of patients achieving the satisfactory relief of irritable bowel syndrome and has no effect on bowel dysfunction, abdominal pain, flatulence or urgency relative to placebo. However, the administration of VSL#3 appears to reduce abdominal bloating relative to placebo in patients with diarrhoea-predominant irritable bowel syndrome. This is particularly evident in patients with higher bloating scores at baseline, suggesting that further studies in patients with high bloating scores should be performed. Controlled studies of the effects of probiotics in irritable bowel syndrome are limited. There has been only one published controlled clinical study evaluating the effect

of a probiotic (Lactobacillus plantarum) in patients with irritable bowel syndrome. This trial showed a significant improvement of flatulence in the probiotic-treated group compared with those given placebo: 44% of patients in the probiotic group reported a reduction of greater than 50% in flatulence, compared with only 18% in the placebo group. There were no significant improvements in other symptoms (abdominal pain, defecation function and overall gastrointestinal function) with probiotic therapy in this study.2 The study did not evaluate the effect on abdominal bloating.2 In the past, beneficial effects have also been attributed to probiotic treatment in uncontrolled trials in irritable bowel syndrome patients.25 The mechanism of the possible beneficial effect of probiotics in irritable bowel syndrome patients is unknown. Johansson et al. demonstrated that the administration and colonization of Lactobacillus plantaÓ 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

VSL#3 IN PATIENTS WITH DIARRHOEA-PREDOMINANT IBS

903

rum in the colon resulted in a decrease in the bacterial groups with gas-producing ability, such as Clostridium sp., in the colonic mucosa.26 Furthermore, King et al. showed that colonic gas (particularly hydrogen) production was greater in patients with irritable bowel syndrome than in controls, and that both symptoms and gas production were reduced by an exclusion diet.27 These observations suggest a role for gut bacteria in the symptomatology of irritable bowel syndrome patients, and the potential of a ‘more favourable’ intestinal milieu to relieve irritable bowel syndrome symptoms. Many patients with irritable bowel syndrome report prominent bloating and flatulence, and a treatment which relieves bloating or flatulence would fit a currently unmet need. The study by Lembo et al. demonstrated that 60% of 443 irritable bowel syndrome patients referred to a tertiary centre rated bloating as their most bothersome symptom, whereas only 29% felt that abdominal pain was most severe; furthermore, 66% of patients reported flatulence as one of their symptoms.28 Amongst 122 adult irritable bowel syndrome patients, moderate to severe bloating was reported on 28% of days and abdominal pain on 33% of days.9 Impaired intestinal gas transit has been demonstrated in irritable bowel syndrome.29 In these experiments, delayed transit of gas infused into the small intestine was consistent with an abnormality of the motility of the bowel. Others have also shown impaired transit of solids or liquids in patients with irritable bowel syndrome.30, 31 Our data suggest that a probiotic is unlikely to retard small bowel or colonic transit in patients with diarrhoea-predominant irritable bowel syndrome; however, further studies in patients with constipation- or bloating-predominant irritable bowel syndrome are needed before the exclusion of a potential role of probiotics in altering intestinal transit. Constipation-predominant irritable bowel syndrome patients tend to be more bloated than those with diarrhoea.32 Tegaserod, a partial 5-hydroxytryptamine-4 agonist, accelerates small bowel and colonic transit in these patients,33 and may have a potential role in relieving the symptom of abdominal bloating, which has been reported to be reduced in irritable bowel syndrome patients treated with tegaserod in a large phase III trial.34 The sample size in the current study was selected to demonstrate a clinically meaningful effect on gastrointestinal transit. Hence, the small sample size of this study resulted in a lack of power to detect a
Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904

significant symptom improvement with VSL#3 treatment compared with placebo. Nevertheless, the effect on bloating appears to be considerable given the trend towards the improvement of this symptom with VSL#3 therapy. Recent studies have suggested that VSL#3 may play a role in restoring pouchitis to normal or, at least, in the relief of symptoms of pouchitis, either as single therapy35, 36 or as an adjuvant to immunosuppressive agents.37 Probiotics need to be evaluated further to determine specifically which irritable bowel syndrome sub-groups and symptoms will respond to probiotic therapy, the best probiotic preparation and the optimal regimen in terms of duration and dosage.
ACKNOWLEDGEMENTS

This study was supported by VSL Pharmaceuticals, Inc., Fort Lauderdale, FL, USA, and by R01 grant #DK54681 (MC) and General Clinical Research Center grant #RR00585 from the National Institutes of Health. We thank Mrs Cindy Stanislav for excellent secretarial assistance.
REFERENCES
1 Rolfe RD. The role of probiotic cultures in the control of gastrointestinal health. J Nutr 2000; 130(2S Suppl.): 396S–402S. 2 Nobaek S, Johansson ML, Molin G, Ahrne S, Jeppsson B. Alteration of intestinal microflora is associated with reduction in abdominal bloating and pain in patients with irritable bowel syndrome. Am J Gastroenterol 2000; 95(5): 1231–8. 3 Barone C, Pettinato R, Avola E, et al. Comparison of three probiotics in the treatment of acute diarrhea in mentally retarded children. Minerva Pediatr 2000; 52(3): 161–5. 4 Gomez-Gil B, Roque A, Turnbull JF, Inglis V. A review on the use of microorganisms as probiotics. Rev Latinoam Microbiol 1998; 40(3–4): 166–72. 5 Guandalini S, Pensabene L, Zikri MA, et al. Lactobacillus GG administered in oral rehydration solution to children with acute diarrhea: a multicenter European trial. J Pediatr Gastro enterol Nutr 2000; 30(1): 54–60. 6 Vanderhoof JA, Young RJ. Use of probiotics in childhood gastrointestinal disorders. J Pediatr Gastroenterol Nutr 1998; 27(3): 323–32. 7 Gorbach SL. Probiotics and gastrointestinal health. Am J Gastroenterol 2000; 95(1 Suppl.): S2–4. 8 Saavedra J. Probiotics and infectious diarrhea. Am J Gastroenterol 2000; 95(1 Suppl.): S16–8. 9 Hahn B, Watson M, Yan S, Gunput D, Heuijerjans J. Irritable bowel syndrome symptom patterns: frequency, duration, and severity. Dig Dis Sci 1998; 43(12): 2715–8.

904

H. J. KIM et al.
24 Pratt JW. Remarks on zeros and ties in the Wilcoxon signed rank procedures. J Am Statist Assoc 1959; 54(287): 655–67. 25 Hamilton-Miller JMT. Probiotics in the management of irritable bowel syndrome: a review of clinical trials. Microbial Ecol Health Dis 2001; 13: 212–6. 26 Johansson ML, Molin G, Jeppsson B, Nobaek S, Ahrne S, Bengmark S. Administration of different Lactobacillus strains in fermented oatmeal soup: in vivo colonization of human intestinal mucosa and effect on the indigenous flora. Appl Environ Microbiol 1993; 59(1): 15–20. 27 King TS, Elia M, Hunter JO. Abnormal colonic fermentation in irritable bowel syndrome. Lancet 1998; 352(9135): 1187–9. 28 Lembo T, Naliboff B, Munakata J, et al. Symptoms and visceral perception in patients with pain-predominant irritable bowel syndrome. Am J Gastroenterol 1999; 94(5): 1320–6. 29 Serra J, Azpiroz F, Malagelada JR. Impaired transit and tolerance of intestinal gas in the irritable bowel syndrome. Gut 2001; 48(1): 14–9. 30 Stivland T, Camilleri M, Vassallo M, et al. Scintigraphic measurement of regional gut transit in idiopathic constipation. Gastroenterology 1991; 101(1): 107–15. 31 Cann PA, Read NW, Brown C, Hobson N, Holdsworth CD. Irritable bowel syndrome: relationship of disorders in the transit of a single solid meal to symptom patterns. Gut 1983; 24(5): 405–11. 32 Schmulson M, Lee OY, Chang L, Naliboff B, Mayer EA. Symptom differences in moderate to severe IBS patients based on predominant bowel habit. Am J Gastroenterol 1999; 94(10): 2929–35. 33 Prather CM, Camilleri M, Zinsmeister AR, McKinzie S, Thomforde G. Tegaserod accelerates orocecal transit in patients with constipation-predominant irritable bowel syndrome. Gastroenterology 2000; 118(3): 463–8. 34 Muller-Lissner SA, Fumagalli I, Bardhan KD, et al. Tegaserod, a 5-HT(4) receptor partial agonist, relieves symptoms in irritable bowel syndrome patients with abdominal pain, bloating and constipation. Aliment Pharmacol Ther 2001; 15(10): 1655–66. 35 Gionchetti P, Rizzello F, Venturi A, et al. Oral bacteriotherapy as maintenance treatment in patients with chronic pouchitis: a double-blind, placebo-controlled trial. Gastroenterology 2000; 119(2): 305–9. 36 Mimura T, Fernando R, Schreiber S, et al. Once daily high dose probiotic therapy maintains remission and improved quality of life in patients with recurrent or refractory pouchitis: a randomized, placebo-controlled, double-blind trial. Gastroenterology 2002; 122: A81. 37 Lammers K, Vergopoulos A, Babel N, et al. Probiotic therapy as a prophylaxis in ileo-anal anastomosis patients decreases interferon gamma, interleukin 1 beta and interleukin 8 mRNA expression. Gastroenterology 2002; 122: A274.

10 Cremonini F. Performance characteristics of scintigraphic gastrointestinal and colonic transit measurements in health: normal values and reliability. Gastroenterology 2002; 122: A1061(Abstract). 11 Talley NJ, Phillips SF, Wiltgen CM, Zinsmeister AR, Melton LJ 3rd. Assessment of functional gastrointestinal disease: the bowel disease questionnaire. Mayo Clin Proc 1990; 65(11): 1456–79. 12 Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol 1997; 32(9): 920–4. 13 Heaton KW, O’Donnell LJ. An office guide to whole-gut transit time. Patients’ recollection of their stool form. J Clin Gastroenterol 1994; 19(1): 28–30. 14 Heaton KW, Ghosh S, Braddon FE. How bad are the symptoms and bowel dysfunction of patients with the irritable bowel syndrome? A prospective, controlled study with emphasis on stool form. Gut 1991; 32(1): 73–9. 15 Heaton KW, Radvan J, Cripps H, Mountford RA, Braddon FE, Hughes AO. Defecation frequency and timing, and stool form in the general population: a prospective study. Gut 1992; 33(6): 818–24. 16 Coulie B, Szarka LA, Camilleri M, et al. Recombinant human neurotrophic factors accelerate colonic transit and relieve constipation in humans. Gastroenterology 2000; 119(1): 41–50. 17 Camilleri M, Mayer EA, Drossman DA, et al. Improvement in pain and bowel function in female irritable bowel patients with alosetron, a 5-HT3 receptor antagonist. Aliment Pharmacol Ther 1999; 13(9): 1149–59. 18 Camilleri M, Northcutt AR, Kong S, Dukes GE, McSorley D, Mangel AW. Efficacy and safety of alosetron in women with irritable bowel syndrome: a randomised, placebo-controlled trial. Lancet 2000; 355(9209): 1035–40. 19 Camilleri M, Colemont LJ, Phillips SF, et al. Human gastric emptying and colonic filling of solids characterized by a new method. Am J Physiol 1989; 257(2 Part 1): G284–90. 20 Proano M, Camilleri M, Phillips SF, Brown ML, Thomforde GM. Transit of solids through the human colon: regional quantification in the unprepared bowel. Am J Physiol 1990; 258(6 Part 1): G856–62. 21 Camilleri M, Zinsmeister AR, Greydanus MP, Brown ML, Proano M. Towards a less costly but accurate test of gastric emptying and small bowel transit. Dig Dis Sci 1991; 36(5): 609–15. 22 Camilleri M, Zinsmeister AR. Towards a relatively inexpensive, noninvasive, accurate test for colonic motility disorders. Gastroenterology 1992; 103(1): 36–42. 23 Burton DD, Camilleri M, Mullan BP, Forstrom LA, Hung JC. Colonic transit scintigraphy with labeled activated charcoal compared with ion exchange pellets. J Nucl Med 1997; 38(11): 1807–10.

Ó 2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 17, 895–904