Digestive Diseases and Sciences

, Volume 56, Issue 7, pp 2067–2072

Effects of Rifaximin Treatment and Retreatment in Nonconstipated IBS Subjects


    • GI Motility Program, Cedars-Sinai Medical Center
  • Walter Morales
    • GI Motility Program, Cedars-Sinai Medical Center
  • Kathleen Chua
    • GI Motility Program, Cedars-Sinai Medical Center
  • Gillian Barlow
    • GI Motility Program, Cedars-Sinai Medical Center
  • Stacy Weitsman
    • GI Motility Program, Cedars-Sinai Medical Center
  • Gene Kim
    • GI Motility Program, Cedars-Sinai Medical Center
  • Meridythe M. Amichai
    • GI Motility Program, Cedars-Sinai Medical Center
  • Venkata Pokkunuri
    • GI Motility Program, Cedars-Sinai Medical Center
  • Emily Rook
    • GI Motility Program, Cedars-Sinai Medical Center
  • Ruchi Mathur
    • GI Motility Program, Cedars-Sinai Medical Center
  • Zachary Marsh
    • GI Motility Program, Cedars-Sinai Medical Center
Original Article

DOI: 10.1007/s10620-011-1728-5

Cite this article as:
Pimentel, M., Morales, W., Chua, K. et al. Dig Dis Sci (2011) 56: 2067. doi:10.1007/s10620-011-1728-5


Recent evidence suggests a role for gut bacteria and antibiotics in the pathophysiology and treatment of irritable bowel syndrome (IBS), respectively. While the benefits of the antibiotic rifaximin have demonstrated efficacy and durable improvement in symptoms over 3 months, the long-term need for retreatment using this approach is mostly unknown. In this retrospective study, subjects with nonconstipated IBS who were retreated with rifaximin were examined.


Charts of patients who were seen at a tertiary care medical center between 2007 and 2011 were reviewed. After exclusion criteria were applied, subjects who had received rifaximin and were seen for retreatment were fully reviewed. During review, demographic information, duration of response, and success of treatment and retreatment were evaluated.


A total of 522 charts were reviewed. Of these 522 charts, 71 subjects were nonconstipated IBS subjects who had received at least one retreatment. Of these, 48 had a second, 22 had a third, 7 had a fourth, and 4 had a fifth treatment. More than 75% of subjects who initially responded to rifaximin also responded to any further retreatment, with no significant reduction in benefit for successive retreatments. Furthermore, there was no change in the duration of benefit (median time between treatments) for successive retreatments.


Retreatment with rifaximin for subjects with nonconstipated IBS in a real-world clinical practice was successful up to five times without decrease in duration or effect.




Irritable bowel syndrome (IBS) is a common chronic medical condition and is the most common disorder treated by gastroenterologists [13]. While the precise etiology of IBS remains unknown, recent evidence has suggested that gut flora have a pathological role in this condition. While difficult to identify using noninvasive means, one of the most common attributable bacterial concepts in IBS is small intestinal bacterial overgrowth (SIBO). In fact, evidence is mounting that a large proportion of IBS sufferers have SIBO by small bowel culture [4].

Based on the principle that a significant number of IBS patients may have gut flora alteration, a number of randomized controlled trials have been conducted examining the effect of antibiotics in IBS [58]. All of these controlled trials have demonstrated significant success with antibiotic therapy. The most recent phase III randomized controlled trial comparing the nonabsorbed antibiotic rifaximin demonstrated four important findings [8]. First, rifaximin produced a statistically greater improvement in the primary endpoint of global improvement compared with placebo. Second, the improvement of all major IBS symptoms was also significantly greater with rifaximin. Third, the drug had a durable effect long beyond cessation of therapy. Finally, the new Food and Drug Administration (FDA) interim endpoints of improvement of abdominal pain and stool consistency were also improved with the therapy.

While antibiotics appear effective in IBS, there has remained a concern about retreatment. If SIBO is a potential cause of gut flora changes in IBS, overgrowth can relapse. Thus, in some instances, retreatment may be necessary. While studies have demonstrated that rifaximin is unlike other conventional antibiotics [911], is nonabsorbed and clinical resistance is not seen on small-scale study [12], further work is needed to demonstrate the effectiveness of retreatment more convincingly. However, prospective study would be very difficult and lengthy, with the significant challenge of long-term compliance and follow-up. Furthermore, a prospective large-scale study may only capture one to two relapses over a 2-year study.

In this study, we perform a large-scale retrospective chart review to examine the effect of retreatment in a real-time clinical setting.


Subject Population

A large-scale retrospective chart review was conducted of all subjects seen at a tertiary care medical center with a diagnosis of irritable bowel syndrome (IBS) to identify subjects who had been treated with rifaximin. Subjects were identified by searching an electronic clinic scheduling database for the years from January 2007 through January 2011 for three physicians who see IBS subjects. In some cases, these patients had first rifaximin treatments as early as 2004. During the screening process, subjects were excluded if they had history of, or symptoms consistent with, constipation-predominant IBS, but were included if they were described as having diarrhea, or mixed IBS. The definition of constipation-predominant IBS was based on the first visit to the clinic, whereby they had been characterized as “constipated.” Subjects were also excluded if they had other gastrointestinal (GI) disorders, such as inflammatory bowel disease (IBD), celiac disease, peptic ulcer disease, diabetes or systemic disease-related dysmotility (e.g., scleroderma), GI surgery (other than cholecystectomy or appendectomy), known adhesions or missing charts. Subjects with history of current narcotic use were also excluded (Lomotil and Imodium were not excluded). Subjects with nonconstipated IBS for whom on the initial visit a more pressing medical matter was identified were also excluded (for example, a patient on initial visit was noted to have nonconstipation IBS but had a severe murmur which was investigated and found to be a life-threatening aneurysm of the aortic root requiring urgent cardiovascular surgery). Finally, subjects who had been treated with rifaximin concurrently with another antibiotic were also excluded. This study was approved by the Cedars Sinai Institutional Review Board.

Data Extraction

The purpose of this review is to characterize the treatment and response patterns of rifaximin. To do this, each patient’s chart was reviewed to evaluate the patient’s demographics (age and sex), bowel symptoms, and medical history as recorded from their initial consultation notes. The date of first rifaximin treatment and effect of first rifaximin treatment were recorded. Patients who had received initial rifaximin and subsequently returned due to symptom relapse for rifaximin retreatment were then further reviewed to examine the effects of this retreatment. The success of these retreatments was determined by examining the follow-up clinical notes and documentation of subjective improvement. The date of retreatment was noted. Since, in some cases, the patients described the relapse as gradual, a specific date of relapse could not be established. Thus, the date of retreatment represented the date the patient’s symptoms justified a repeat visit and retreatment. In many instances, subjects who had a response to rifaximin simply called into the office for refills. Since these data are not a clinical confirmation of relapse at the bedside, these data were not included in the analysis, as there was no official record of this interaction in the routine clinical charts.

Data Analysis

The data were analyzed to determine the number of retreatments with rifaximin, the success of each retreatment, and the duration of benefit (defined as the time in months between successive treatments with rifaximin). While this study is principally a descriptive study, statistical analyses were performed to compare age and sex distribution between subjects with and without IBS as well as those who did and did not receive retreatment with rifaximin. When comparing these quantitative and qualitative data, a t-test and chi-square were used, respectively. In addition, to determine if there was a deterioration of benefit from one treatment to the next, a paired t-test was used to compare subsequent treatment duration between relapse. Durations of benefit for rifaximin were expressed as a median, since the data were not normally distributed. In all instances, this provided a more conservative estimate. If the data were expressed as a mean, this would falsely increase the duration due to subjects with long duration of sustained benefit.


Study Subjects

A total of 522 patient charts were initially reviewed and screened for this study. The review and screening process is summarized in Fig. 1. Of these, 335 patients did not meet the inclusion criteria and were excluded from further analysis. The remaining 187 subjects (35.8%) had non-constipation-predominant IBS, of whom 169 had been treated with rifaximin. Of these, 71 subjects (44 female) had taken rifaximin and returned for a symptom relapse with documented rifaximin retreatment. The average age of this group was 47.0 ± 16.2 years (Table 1).
Fig. 1

Flowchart depicting the review and screening process for this study. *This group includes nine subjects who did not receive rifaximin alone and were thus excluded from the final rifaximin retreatment group.

Table 1




Non-C-IBS non-retreatment group

Non-C-IBS (rifaximin >1 time)





Female [n (%)]

234 (70.7)

71 (61.2)

44 (62.0)

Age (years)a

51.3 ± 17.7

48.3 ± 16.8

47.0 ± 16.2

aThere was no statistical difference in age between groups

Overall Effect of First Rifaximin Treatment

Among the entire nonconstipation IBS group (n = 187), 169 received rifaximin and 148 were seen in follow-up. Of these subjects, rifaximin produced a clinical improvement in 111 subjects (75%; Fig. 2).
Fig. 2

Non-C-IBS overall response to rifaximin. This figure depicts all non-C-IBS subjects and their response to the first encounter with rifaximin

Number of Retreatments

The number of subjects who had undergone one or more retreatments with rifaximin were then summarized for those subjects for whom these data were available. In total, 71 subjects had at least one retreatment with rifaximin; as shown in Fig. 3, 48 subjects had a second, 22 had a third retreatment, 9 had a fourth, and 4 had a fifth retreatment with rifaximin.
Fig. 3

Number of subjects undergoing retreatments with rifaximin. This figure demonstrates the breakdown of subjects at each retreatment. Note that, in some cases, subjects who did not respond to rifaximin may have been retreated, although these were few

The success rate of each retreatment with rifaximin was next determined, where success was defined as a patient declaration of overall improvement in subjective IBS symptoms. It was noted that more than 75% of subjects who initially responded to rifaximin experienced clinical benefit with any retreatment (Fig. 4). Furthermore, there appeared to be no waxing of benefit with increasing number of retreatments; i.e., the percentage of subjects who experienced improvement in their IBS symptoms did not decrease with increasing number of retreatments.
Fig. 4

Success rate of each retreatment with rifaximin. The percentage of subjects with notable improvement in their IBS symptoms is given for each retreatment with rifaximin. This was based on the number of subjects who experienced improvement among subjects for whom data were available for the preceding treatment. The x-axis shows the retreatment number (RTX) after the first rifaximin treatment. Only 75 had follow-up or evaluable information on success of retreatment

Interestingly, ten subjects had received rifaximin from an outside physician before being seen by the clinic. Despite a lack of response to this rifaximin, changing the dose of rifaximin to known effective doses resulted in improvement in nine of these subjects.

Duration of Benefit Between Successive Retreatments with Rifaximin

Lastly, the duration of benefit after each treatment with rifaximin was examined. It was noted that there was a wide range of relapse rates for all retreatments (Fig. 5). However, there was a notably long duration of benefit from rifaximin, the median being always a minimum of 4 months. In addition, there appeared to be no change in recurrence time with increasing treatment number (Fig. 5). In addition, in paired comparison over time, there was no shortening of the durability with subsequent treatment.
Fig. 5

Duration of benefit between successive retreatments with rifaximin. The time in months refers to the number of months of wellness/patient seeking physician for retreatment before the retreatment number (RTX) in each column. For example, for the first RTX, this is the number of months the patient was well after the first rifaximin administration and prior to the first RTX. Rarely, patients were missing accurate dates and thus were not included in these analyses. Median values are used, as these represent a more accurate depiction of the duration of benefit, as the ranges were wide


Over the last few years, data have accumulated to support the use of rifaximin in the treatment of nonconstipated IBS. However, data on retreatment of IBS using this approach have been based on one small study [12]. In this large-scale retrospective study, up to five treatments of rifaximin were examined in IBS subjects presenting in the absence of constipation and other bowel disorders. Interestingly, if the initial treatment with rifaximin was successful or not, retreatment with rifaximin was successful in more than 75% of subjects regardless of the number of treatments given. In other words, rifaximin had the same benefit when used for any number of retreatments. In addition, the time between treatments did not change with subsequent retreatment visits.

Recent evidence is mounting that gut flora play a role in IBS. One proposed explanation for this altered flora has been small intestinal bacterial overgrowth. While much of this is based on breath testing [13, 14], recent culture studies have demonstrated an increase in small bowel coliforms in IBS compared with controls [4, 15]. Based on this hypothesis, antibiotic studies have been conducted and have uniformly demonstrated the benefits of antibiotics in IBS [58]. The most recent antibiotic studies have focused on rifaximin as a treatment for nonconstipated IBS. In study, rifaximin appears to have few side-effects compared with placebo [8]. In addition to meeting the primary endpoints, the benefits of rifaximin were demonstrated across symptoms of abdominal pain, stool consistency, bloating, and overall IBS symptoms.

While rifaximin has no apparent side-effects or safety concerns, there remains some question about efficacy over time. In the recent TARGET1 and TARGET2 studies, rifaximin proved to be efficacious for a minimum of 3 months after only 14 days of treatment [8]. However, the study was not continued to determine the true time to patient relapse and/or the need for further treatment. In this study, we examine this effect in a real clinical setting.

The issue with studying relapse is how to define relapse. If patients’ symptoms are relatively well controlled but there are some breakthrough symptoms, patients may not find it necessary to contact the physician’s office. It is only when symptoms affect daily function or reach an individual tolerance threshold that they seek medical attention again. This is probably the most real-world gage of relapse. While the charts were selected between 2007 and 2011, some patients had visits dating back to 2004. Thus, this is a longitudinal evaluation of pattern of practice and relapse with rifaximin retreatment. In this analysis, rifaximin was given up to six total times. This is a remarkable number of times in that no prospective study could collect this number of retreatments as it would require many years to conduct. What was seen was that, irrespective of the number of times a patient received rifaximin, effectiveness was seen in greater than 75% of subjects with any subsequent retreatment (Fig. 4).

While this study did not test for development of resistant bacteria in stool of patients, clinical resistance could be inferred in two ways. First, if resistance developed, then it would be reasonable to assume that subsequent treatments would be less and less effective. This was not seen in this study, in that the percentage of subjects who responded with increasing number of treatments did not change. The second way resistance might be seen is if there was a shortening of time between visits for retreatment. To answer this, a paired comparison was performed, which showed that, even comparing the first retreatment to the fifth retreatment, there was no change in the duration of benefit. These two findings indirectly suggest that subjects are not developing resistance.

Interestingly, this is not the first study to examine retreatment. In a small-scale study, we recently demonstrated that subjects who required retreatment with rifaximin almost always responded with subsequent retreatment [12]. This study was conducted a number of years ago, and thus only a few subjects (N = 16) were eligible for analysis and only three total treatments were evaluable. This current large-scale multiyear study now confirms these results with a robust patient number and extended retreatments.

There are a number of pitfalls with this study. While the obvious pitfall of this study is that it is retrospective, this is balanced by other strengths. First, it would be impossible to conduct and retain study participants in a 4-year protocol. Secondly, this study represents a true clinical experience of how rifaximin is used and reused. In a prospective study, the definition of relapse would dictate retreatment, and this would not mimic the realities of a real patient-care setting. Thus, this chart review is more representative of what a clinical practice with rifaximin would look like.

Other pitfalls are more important. For example, the study could not account for subjects who responded to rifaximin but would call in intermittently and request a prescription by phone for retreatment. These subjects were excluded, since it is not documented whether the subsequent rifaximin course was initiated by the patient to try rifaximin again even if it did not work initially or due to a successful improvement in past with relapse. In addition, some subjects may have had a response to rifaximin but never needed retreatment or a revisit to the physician. The duration of benefit in this circumstance would not be counted, as there is no duration yet, due to the fact that there is no relapse. Another potential pitfall might be those subjects who are lost to follow-up. It is difficult to know if this is due to benefit or due to lack of benefit. While there are only few subjects with data on up to five total treatments, this may reflect a number of factors, the most important of which is time. There may not have been enough years of study to identify more subjects with five treatments.

In conclusion, for subjects returning to attempt retreatment with rifaximin, this real clinic chart review demonstrates that retreatment with rifaximin is highly likely to improve IBS, even when used up to five times in the absence of other GI disorders or constipation. Interestingly, the success of retreatment and lack of shortening of benefit suggest a lack of development of clinical resistance to rifaximin. Furthermore, the mean time to the request for retreatment by patients due to relapse appears to be a median of greater than 4 months, but varies from patient to patient.


This study was funded by the Beatrice and Samuel A. Seaver Foundation.

Copyright information

© Springer Science+Business Media, LLC 2011