Evidence review: efficacy

Limited research has examined rifaximin for the treatment of pouchitis. Four relevant studies that have been published in full were identified: 1 small placebo-controlled randomised controlled trial (RCT) of rifaximin for the treatment of active pouchitis (n=18); 2 non-comparative observational studies of rifaximin plus ciprofloxacin for the treatment of chronic pouchitis refractory to previous treatment (n=18 and n=8); and 1 non-comparative observational study of rifaximin maintenance therapy after antibiotic-induced remission (n=51).

Randomised controlled trial

Isaacs et al. (2007) conducted a double-blind, placebo-controlled trial at 10 US sites between May 2003 and April 2004; the trial included 18 people with active pouchitis (acute or chronic).

The study recruited adults (aged 18 years and older) with a history of a total colectomy with ileal-pouch anal anastomosis (IPAA) for ulcerative colitis, and who had active pouchitis (Pouchitis Disease Activity Index [PDAI] score of 7–18 points). Pouchitis was defined as acute if symptoms had not been treated with antibiotics or other medical therapy within the previous 30 days, and chronic if antibiotics or other medical therapy for pouchitis had been given within the previous 30 days. Extensive exclusion criteria included infectious, ischaemic, or immunological diseases with gastrointestinal involvement; significant hepatic or renal disease; or unstable cardiovascular or pulmonary disease. Concomitant treatment with antibiotics, probiotics, sulfasalazine, olsalazine, balsalazide, mesalamine, rectal short-chain fatty acids, rectal glutamine, loperamide or diphenoxylate was not permitted after study enrolment. Concomitant therapy with rectal or oral corticosteroids was not permitted within 2 weeks of enrolment, and concomitant therapy with immune-modifying drugs was not permitted within 3 months of enrolment.

Participants were randomised to 4 weeks' treatment with rifaximin 400 mg 3 times daily (n=8) or placebo 3 times daily (n=10). It is unclear if allocation was concealed.

The primary outcome was clinical remission at 4 weeks (PDAI score of less than 7 points and a decrease from baseline PDAI score of 3 points). Secondary outcomes included complete remission (PDAI score of 0), symptomatic improvement (decrease from baseline PDAI clinical sub-score of 2 points), endoscopic improvement (decrease from baseline PDAI endoscopic subscore of 2 points), and health-related quality of life as measured by the Inflammatory Bowel Disease Questionnaire (IBDQ) score.

Analysis was based on the modified intention-to-treat population (all randomised participants who had a baseline efficacy evaluation, took at least 1 dose of the study medication, and had at least 1 post-baseline efficacy evaluation). This excluded 1 person from the placebo group who withdrew without having any post-baseline assessments. Among the remaining 17 people (mean age 41 years), there was no significant difference in baseline characteristics between the rifaximin and placebo groups. The mean PDAI score at baseline was 10.6 in the rifaximin group and 9.2 in the placebo group. All people in the rifaximin group and nearly all (8/9) of those in the placebo group had received previous antibiotic treatment. Seven people in each group had received 5-aminosalicylates, 4 of the 8 people in the rifaximin group had received non-steroidal anti-inflammatory drugs (1/9 in the placebo group), and 2 of the 8 people had received probiotics (1/9 in the placebo group).

Two people in the placebo group withdrew before 4 weeks because of lack of efficacy and their treatment was counted as ineffective, using a 'last observation was carried forward' approach. Clinical remission was obtained in 2 of the 8 people in the rifaximin group, both of whom had had acute pouchitis, compared with none of the placebo group. However, this difference was not statistically significant (p=0.211). No one in the study obtained complete remission. There were also no statistically significant differences in any of the secondary outcomes.

Observational studies

Rifaximin plus ciprofloxacin for chronic pouchitis

Gionchetti et al. (1999) reported the experience of 18 adults with chronic treatment-resistant (refractory) pouchitis who were treated with the combination of ciprofloxacin and rifaximin at 1 institution in Bologna, Italy.

Chronic pouchitis was defined as the presence of symptoms for more than 4 weeks and the need for treatment with antibiotics or anti-inflammatory drugs for more than 15 days each month to control symptoms. Treatment resistance was defined as non-response to at least 4 weeks' treatment with antibiotics: metronidazole 1.2 g once daily (12 people), amoxicillin/clavulanic acid 1 g twice daily (4 people) or ciprofloxacin 500 mg twice daily (2 people).

The 18 people (10 male, median age 33 years) had a median PDAI score of 11 (range 9–17). The reasons why they had had IPAA surgery were not given. The first episode of pouchitis had been a median of 27 months before the study (range 4–72 months), and the current episode of pouchitis had lasted for a median of 2 months (range 0.5–4.5 months).

All participants in the study received combination treatment with rifaximin 1 g twice daily plus ciprofloxacin 500 mg twice daily for 15 days. By the end of this period, 6 of the 18 people had obtained remission of pouchitis (PDAI score of 0), and a further 10 had an improvement in PDAI score of at least 3 points. The authors stated that the 2 people whose condition did not respond were not the 2 who had previously not obtained a beneficial effect from ciprofloxacin treatment.

Abdelrazeq et al. (2005) reported the experience of 8 people who underwent IPAA at York Hospital, UK, between 1988 and 2003 and who developed chronic refractory pouchitis that was treated with a combination of rifaximin and ciprofloxacin.

The 8 adults (aged 20–50 years, 6 male) had a median PDAI score was 12 (range 9–18). The reasons why they had had IPAA surgery were not given. They all had chronic active pouchitis (symptom duration of more than 4 weeks, or pouchitis that required more than 2 weeks' treatment each month to control symptoms) that was refractory (non-responsive to at least 4 weeks' treatment with metronidazole or ciprofloxacin or immediately relapsing upon stopping or reducing antibiotics). They were treated with rifaximin 1 g twice daily plus ciprofloxacin 500 mg twice daily for 2 weeks.

After treatment, 5 of the 8 people remained in remission for at least 6 months (PDAI score of 0) and a further 2 of the 8 people had an improvement in PDAI score of at least 3 points. In 2 of these 7 people who obtained a benefit from rifaximin, pouchitis recurred at 6 and 8 months, but responded again to further courses of the same treatment combination. At median 30-month follow-up, all 7 people had satisfactory pouch function.

The 1 person whose pouchitis did not respond to the treatment combination was also reported to have had no response to systemic steroids, immunosuppressive therapy and hyperbaric oxygen therapy. Their pouch was subsequently removed surgically.

Rifaximin as maintenance therapy

Shen et al. (2008) reported the experience of adults with antibiotic-dependent pouchitis who had undergone IPAA for ulcerative colitis and who attended the Pouchitis Clinic of The Cleveland Clinic, Ohio, USA between July 2004 and June 2006.

The 53 adults (mean age 46–47 years) all had current symptoms and met both of the following criteria:

  • Antibiotic-dependent pouchitis, defined as 4 or more episodes per year, each of which responded to a 2-week course of ciprofloxacin or metronidazole but recurred soon after treatment ended.

  • Frequent episodes of pouchitis needing long-term treatment (at least 16 weeks) for remission maintenance with continuous, low-dose antibiotics or frequent pulse therapy with antibiotics.

People with antibiotic-refractory pouchitis (pouchitis not responsive to a 2–4‑week course of ciprofloxacin or metronidazole) or with concurrent cuffitis (inflammation of the rectal cuff), irritable pouch syndrome or Crohn's disease of the pouch were not included in this group.

Active pouchitis was defined as a modified PDAI (mPDAI) score of 6 or more (mPDAI includes symptomatic and endoscopic, but not histological, criteria; total possible score 12). To induce remission, participants were given either single or combination treatment (at the clinician's discretion) with: ciprofloxacin 1000 mg daily; metronidazole 1000 mg or 1500 mg daily; tinidazole 1000 mg daily; or rifaximin 600 mg, 800 mg or 1200 mg daily for 2 weeks.

Of the 53 people who received induction with either single or combination antibiotics, 51 obtained remission after 2 weeks (mPDAI score less than 6) and began maintenance treatment with rifaximin for up to 24 months at a starting dose of 200 mg daily. Symptoms, adverse events and treatment compliance were assessed every 1–3 months. If there was partial response, the rifaximin dose was increased up to a maximum 1800 mg daily. Treatment was discontinued if active pouchitis recurred despite dose increases, or if the person chose to stop treatment.

After 3 months, 33 of the 51 people (65%) were still in remission, and the remainder (18/51) had relapsed. The 33 people in whom remission was maintained for 3 months had no symptomatic or endoscopic evidence of relapse between the end of induction and the 3-month assessment (0 point change on both symptom and endoscopy scores of the mPDAI). People who relapsed had a median increase of 3 points in both symptom and endoscopy scores between the end of induction and 3 months.

Four of the 33 people in whom remission was maintained had symptom recurrence after 3–12 months, but 26 of the 33 people continued rifaximin for at least 6 months, 19 for at least 12 months, 5 for at least 18 months and 2 for at least 24 months.

The majority of these people (23/33) received rifaximin 200 mg daily for the entire maintenance period. Ten people needed dose escalation during the maintenance period to 400 mg (3/33), 600 mg (3/33), 800 mg (2/33), 1200 mg (1/33) or 1800 mg daily (1/33). Similarly, the majority of those who relapsed had received 200 mg daily (11/18). Antibiotic induction regimen, symptom score after induction and rifaximin dose were not predictive of maintaining remission.