Probiotics for pouchitis

Bandolier, August, 2005

Pouchitis is a non-specific acute inflammation within an ileal reservoir. It leads to increased frequency of loose stool, and abdominal cramping. About half of people having surgery, usually for ulcerative colitis, have at least one episode of pouchitis over about 10 years. Treatment is usually with antibiotics, but a minority of patients (perhaps 10%) experience refractory or frequently recurrent pouchitis, usually regarded as two episodes or more in a year. Even so, antibiotics can be effective.

The problem is in maintaining a pouchitis-free state thereafter. Probiotics have been suggested as being helpful, and this brie(liandolier review examines the evidence.

Search

Bandolier sought randomised trials for the use of probiotics in pouchitis. There was no prior intent to differentiate between different probiotics, for any specific outcome in trials, or type of pouchitis.

Results

Four randomised trials were found. Three concerned VSL#3, containing four strains of lactobacilli, three strains of bifidobacteria, and one strain of streptococcus salivarius. This was presented in bags with about 300 billion bacteria per gram. One trial concerned use of lactobacillus GG, in a formulation of about 10 billion bacteria per capsule. These were analysed separately.

VSL#3

The trial details are shown in Table 1. Two of them were for prophylaxis against pouchitis in patients with established refractory or frequently recurrent pouchitis, and one was in patients without pouchitis, as prophylaxis against it developing.

[FIGURE 1 OMITTED]

The duration of trials was nine or 12 months. The three trials were of good reporting quality (generally scores of 3 to 5 out of 5 on a popular scoring system), but were small, with about 40 patients in each. Patients had 600 to 900 billion bacteria daily.

Results were good, with low rates of pouchitis with VSL#3, but high rates with placebo, the common comparator (Figure 1). Overall there was a big reduction with VSL#3, with relative risk of developing pouchitis of 0.2 (95% confidence interval 0.1 to 0.3). The number needed to treat to prevent pouchitis was 1.5 (1.3 to 2.0).

In those with refractory or frequently recurrent pouchitis, placebo pouchitis rates were almost 100%, but were about 10% with VSL#3. The number needed to treat to prevent pouchitis over about a year was 1.2 (1.0 to 1.4).

One adverse event was noted on VSL#3. One patient had abdominal cramps, vomiting, and diarrhoea on VSL#3, recurring on repeat challenge.

Lactobacillus GG

One small trial involving 20 patients [1] treated previously for clinical symptoms of pouchitis randomised them into two groups, treated with either four capsules of lactobacillus GG a day (40 billion bacteria) or placebo for three months. It showed changed pouch intestinal bacterial flora in four subjects, but few other changes. Though most patients had pouchitis, only one patient in the lactobacillus group may have improved.

Comment

Numbers needed to treat below 2 are rare, so that the NNTs to prevent pouchitis with VSL#3 therapy should be taken seriously. Though numbers of patients were small, results were consistent in valid studies of reasonable reporting quality Although acute pouchitis was the outcome concentrated on here, other useful results of interest to gastroenterologists were reported in the trials. It is an interesting example of small numbers of small trials properly done, where a big effect can be taken seriously.

Probiotic use in gastrointestinal diseases has been reviewed recently [2], though not with a systematic search. It suggests that treatment of acute diarrhoea in children and prevention of antibiotic associated diarrhoea are the only two areas with any likely utility. A review of the latter (Bandolier 104) suggested an NNT of only about 10, making the results for pouchitis much more impressive. One issue, though, is the range of bacteria, and the very high dose in the VSL#3 studies. Not all probiotics are the same.

References:

[1] J Kuisma et al. Effect of lactobacillus rhamnosus GG on ileal pouch inflammation and microbial flora. Alimentary Pharmacology and Therapeutics 200317: 509-515.

[2] A Sullivan, CE Nord. Probiotics and gastrointestinal diseases. Journal of Internal Medicine 2005 257: 78-92.

Table 1: Randomised trials of VSL#3 versus placebo for pouchitis

Reference                  Design

Gionchetti et al.          Randomised, double blind,
Gastroenterology 2000      placebo controlled, parallel
119: 305-309               group comparison of VSL#3
                           with placebo over 9 months; 20
                           patients in each group
                           Oral 3 g bags twice daily

Gionchetti et al.          Randomised, double blind,
Gastroenterology 2003      placebo controlled, parallel
124:1202-1209              group comparison of VSL#3
                           with placebo over 12 months;
                           20 patients in each group
                           Oral 3 g bag once daily

Mimura et al. Gut 2004     Randomised, double blind,
53: 108-114                placebo controlled, parallel
                           group comparison of VSL#3
                           with placebo over 12 months;
                           20 patients on VSL#3 and 16
                           on placebo
                           Oral 6 g bags once daily

Reference                  Patients

Gionchetti et al.          Adults (18-65 years)
Gastroenterology 2000      with chronic relapsing
119: 305-309               pouchitis of at least 3
                           relapses per year, in
                           clinical and endoscopic
                           remission

Gionchetti et al.          Adults (18-65 years).
Gastroenterology 2003      Randomisation was
124:1202-1209              within 1 week of
                           ileostomy closure

Mimura et al. Gut 2004     Patients with active
53: 108-114                refractory or recurrent
                           pouchitis (disease
                           score of 7 or above) in
                           remission after antibiotic
                           therapy

Reference                  Outcomes

Gionchetti et al.          Relapse, defined as at
Gastroenterology 2000      least 2 point increase in
119: 305-309               clinical portion of specific
                           disease activity index,
                           confirmed by endoscopy
                           or histology

Gionchetti et al.          Episodes of acute
Gastroenterology 2003      pouchitis, defined as
124:1202-1209              total disease activity
                           score of 7/18 or more

Mimura et al. Gut 2004     Relapse, defined as at
53: 108-114                least 2 point increase in
                           clinical portion of specific
                           disease activity index,
                           confirmed by endoscopy
                           or histology

Reference                  Results

Gionchetti et al.          Patients were well matched at baseline
Gastroenterology 2000      Relapses in:
119: 305-309               20/20 on placebo (all by 4 months)
                           3/20 on VSL#3 (all relapsed within 4
                           months of end of therapy)
                           No adverse events noted

Gionchetti et al.          Patients were well matched at baseline
Gastroenterology 2003      Acute pouchitis:
124:1202-1209              8/20 on placebo
                           2/20 on VSL#3
                           No adverse events noted

Mimura et al. Gut 2004     Patients were well matched at baseline
53: 108-114                Relapses in:
                           15/16 on placebo (all by 4 months)
                           2/20 on VSL#3 (all relapsed within 4
                           months of end of therapy)
                           One patient had abdominal cramps,
                           vomiting, and diarrhoea on VSL#3,
                           recurring on repeat challenge