SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

Background

Clostridium difficile can cause severe antibiotic-associated colitis. Conventional treatments with metronidazole and vancomycin improve symptoms, but after discontinuation of treatment, C. difficile infection (CDI) recurs in a number of patients. Rifaximin is a rifamycin-based non-systemic antibiotic that has effect against C. difficile.

Aim

To assess the effectiveness of rifaximin in recurrent C. difficile infection.

Methods

We retrospectively evaluated the records of 32 patients who were treated with rifaximin for recurrent C. difficile infection. The symptoms were evaluated 12 weeks after the start of treatment and patient records were followed up until 1 year after treatment.

Results

The mean age of the patients was 55 years (median 64, range: 19–84 years). Before the initiation of rifaximin therapy, the patients had undergone, on the average, 4.4 (range: 2–12) antimicrobial courses for C. difficile infection. C. difficile strain typing was performed in 27 patients. Eight (30%) patients had a strain with a DNA profile compatible with the BI/NAP1/027 ribotype. Antibiotic susceptibilities were determined of isolates from 22 patients. Most isolates (68%) had very low MIC-values for rifampin (<0.002 μg/mL) and the highest MIC value was 3.0 μg/mL. Isolates with a DNA profile compatible with the BI/NAP1/027 ribotype had, on the average, higher MICs of rifampin. After 12 weeks 17 (53%) patients had no relapse. The MIC value of rifampin seemed to predict the response to rifaximin treatment.

Conclusions

Rifaximin is a safe treatment for C. difficile infection. It has a reasonable effect in C. difficile infection and it can be considered as an optional treatment for recurrent C. difficile infection.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

Antibiotic treatments disturb the normal faecal microbiota. This can result in a longstanding and treatment-resistant colitis caused by Clostridium difficile, especially in elderly patients.[1, 2] Current treatments of CDI with metronidazole and vancomycin improve symptoms,[3] but after discontinuation of treatment, colitis recurs in a number of patients leading to repeated therapies.[4] For these reasons, new treatment options have been developed for CDI including faecal microbiota transplantation[5-7] and various forms of immunotherapy.[2, 8, 9] Also, other antibiotics have been introduced for the treatment of CDI including rifaximin and fidaxomicin.[10] Both of these are nonsystemic drugs effective against C. difficile and, like vancomycin, are practically not absorbed from the gut. Fidaxomycin is a relatively new drug and was approved for CDI by FDA in 2011. It has shown a better effect in preventing relapses than vancomycin in two randomized studies.[11, 12] Rifaximin was first introduced for the treatment of traveller's diarrhoea[13] and because it is minimally absorbed, well tolerated, and efficacious, it remains as an option for the treatment of uncomplicated traveller's diarrhoea.[14] Rifaximin has also been used to reduce the risk of hepatic encephalopathy in patients with advanced liver disease[15] and even for Crohn's disease.[16] Initially, case reports indicated that rifaximin had an effect against CDI.[17, 18] Since then, a randomized study showed that rifaximin given immediately after a standard treatment with metronidazole or vancomycin had a better effect than placebo.[19] We have used rifaximin in selected cases since 2007. We now report our experience with rifaximin in 32 patients with recurrent CDI.

Patients and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

This study was a review of patients with CDI treated by rifaximin (Rifacol; Prodotti Formenti Srl, Milano, Italia) in Helsinki University Central Hospital, from March 2007 through December 2011. All the patients who received rifaximin for CDI during this period were included. Rifaximin was given only for recurrent CDI. Information was gathered from electronic patient records centres that included patient history, laboratory findings, and official information on the survival of the patient. The study was approved by the Institutional Review Board of Helsinki University Central Hospital. CDI was defined as persisting diarrhoea with a positive culture and toxin test for C. difficile. Treatment failure was defined as persisting diarrhoea with a positive C. difficile toxin stool test.

Clostridium difficile isolation was performed by culturing fresh stool samples on C. difficile selective CCFA agar (cycloserine-cefoxitin-fructose-egg yolk agar) at 35°C for 42 h in anaerobic atmosphere. Colonies with typical morphology, fluorescence, and odour were identified as C. difficile. Toxin production was analysed directly from faecal samples with the Premier Toxins A&B-test kit (Meridian; Bioscience Inc., Cincinnati, OH, USA) during 2007–2010 and with VIDAS C. difficile Toxin A & B CDAB-system (BioMerieux, Marcy l'Etoile, France) from 2011 onwards according to the manufacturer′s instructions, or if the direct toxin test was negative, from C. difficile colonies. Strain typing was performed by DNA analysis using multiplex PCR.[20] Antibiotic susceptibilities for metronidazole, vancomycin, and rifampin (reflecting the susceptibility for rifaximin) were determined with E-tests (BioMerieux) on FAA-plates (Fastidious Anaerobe Agar; LabM, Lancashire, UK).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

Our patient series included altogether 32 patients who had received rifaximin for recurrent CDI. The mean age of the patients was 58 years (median 66, range: 19–88 years). The index infections and antibiotics that preceded the initial CDI are given in Table 1. One patient developed CDI without any known preceding infection or antibiotic therapy. All the other patients developed CDI during or after antibiotic therapy.

Table 1. Patient characteristics and outcome
Patient numberAgeGenderIndex infectionIndex antibioticaNo. of positive tests for C. difficileC. difficile ribotypeNo. of antibiotic courses for CDI before rifaximinDays from first CDI diagnosis before rifaximinOutcome at 12 weeks
  1. a

    Cefurox, Cefuroxime; Cephalex, Cephalexin; Amox, Amoxacillin; Clavul, Clavulanic acid; Clindam, Clindamycin; Cefriax, Cefriaxone; Levoflox, Levofloxacin; Ciproflox, Ciprofloxacin; Metro, Metronidazole; Piper, Piperacillin; Tazo, Tazobactam; Vancom, Vancomycin; Moxiflox, Moxifloxacin; Pen, Penicillin G.

181FemalePreop. prophylaxisCefurox4027 neg. 4 92Failure
249FemaleSinus pilonidalis and UTICephalex2027 neg. 3 81Failure
330FemaleChronic sinusitisCephalex3027 neg. 4158Resolution
433MaleChronic sinusitisAmox/Clavul2027 neg. 3 44Resolution
575FemaleDiverticulitis and UTISeveral4027 neg. 6141Resolution
677FemalePostop. wound infectionCephalex, Clindam5027 neg. 7139Failure
784MalePneumococcus sepsisCefriax, Levoflox5027 neg. 5132Failure
822FemaleDental infectionAmox/Clavul2027 pos. 3 52Failure
964FemaleColitisCiproflox4027 neg. 3237Resolution
1031FemaleEndometritisCephalex, Metro2027 neg. 2 23Failure
1143MaleSinusitisAmox/Clavul1NA 2 27Failure
1246MaleDiverticulitisCephalex/Metro2027 neg. 3139Resolution
1319MaleColitis ulcerosaCiproflox/Metro5027 neg. 4167Resolution
1436FemaleNo known infectionNo given antibiotics2027 neg. 2 71Failure
1571MaleEnterococcus sepsisCefurox/Metro, Piper/Tazo3NA 4 99Resolution
1631MaleCerebral toxoplasmosisSeveral2027 neg. 2 39Resolution
1780FemaleErysipelasCefurox3027 pos. 5138Resolution
1857FemaleUTICiproflox4027 neg. 5 95Resolution
1966MaleErysipelasCefurox5027 neg. 5548Resolution
2075FemaleStaphylococcus sepsisVancom6NA 6184Resolution
2119MalePreop. prophylaxisCefurox5NA 5155Resolution
2220MaleSalmonella enteritisCiproflox3NA 3 68Resolution
2372MalePneumoniaCefurox, Levoflox6027 pos.12295Failure
2480FemaleOtitis mediaAmox, Cefurox, Cefriax3027 pos. 3 86Resolution
2550FemaleFever and neutropeniaCefriax, Levoflox, Cephalex4027 neg. 5133Failure
2681FemaleBronchitis and COPDMoxiflox, Levoflox2027 pos. 5 63Failure
2773FemalePurulent knee arthritisCefriax, Clindam, Pen5027 neg. 5141Resolution
2877FemalePyelonefritisCefurox3027 pos. 5111Failure
2966FemaleUndefined infectionCefriax, Levoflox4027 neg. 5150Resolution
3088FemalePneumoniaCefurox3027 pos. 4 68Failure
3178MalePyelonefritisCiproflox, Cefurox4027 pos. 4 90Failure
3286MaleUndefined infectionCefurox, Amox/Clavul4027 neg. 5 79Failure

Before rifaximin treatment, the patients had had, on the average, 3.5 positive stool tests for C. difficile (median 3, range: 1–6) and they had undergone, on the average, 4.3 (median 4, range: 2–12) metronidazole or vancomycin courses for CDIs. C. difficile strain typing was performed in 27 patients. Eight (30%) patients had a DNA profile compatible with the BI/NAP1/027 ribotype (Table 1).

Antibiotic susceptibilities were determined of isolates from 22 patients. We estimated the susceptibility to rifaximin by measuring the susceptibility to rifampin. Most isolates (15 of 22, 68%) had very low MIC-values for rifampin (<0.002 μg/mL). The highest MIC of rifampin was 3.0 μg/mL. The mean MIC for vancomycin was 0.90 μg/mL (range: 0.38–1.50 μg/mL) and that for metronidazole 0.21 μg/mL (range: 0.06–1.0 μg/mL).

Those strains with the DNA profile compatible with the virulent BI/NAP1/027 ribotype had, on the average, more than 10-fold higher MIC for rifampin than strains that did not have the BI/NAP1/027 ribotype DNA profile (Table 2). Also, the average MIC of metronidazole was approximately 2-fold higher in BI/NAP1/027 strains than in non-BI/NAP1/027 strains (Table 2). There was no evident difference in the average susceptibilities to vancomycin between BI/NAP1/027 and non-BI/NAP1/027 strains (Table 2).

Table 2. Susceptibilities Clostridium difficile isolates by strain type and antibiotic. Mann–Whitney P-values are of comparisons between BI/NAP1/027 and non-BI/NAP1/027 strain types
  N Geometric Mean MIC (μg/mL)MIC Range (μg/mL)Mann–Whitney P-value
Rifampicin   0.002
BI/NAP1/02770.46>0.002–3.0 
Non-BI/NAP1/02713>0.002>0.002 
Metronidazole   0.027
BI/NAP1/02770.350.19–0.5 
Non-BI/NAP1/027130.180.06–1.0 
Vancomycin   0.751
BI/NAP1/02770.840.38–1.5 
Non-BI/NAP1/027130.940.5–1.5 

Our standard rifaximin treatment was oral rifaximin 400 mg twice a day for 14 days that was preceded by an oral course of vancomycin 125 mg four times a day for 14 days. This standard treatment was given to 25 patients. Patient number 9 got vancomycin tapering for 6 weeks before the oral course of rifaximin. Three patients, patients 1, 5 and 14, got metronidazole 400 mg three times a day for 14 days instead of vancomycin before rifaximin. Patient 28 got both metronidazole and vancomycin before the rifaximin therapy. Two patients, patients 12 and 13, got oral rifaximin 400 mg twice a day for 28 days without any immediately preceding courses of vancomycin or metronidazole.

All the patients received the planned rifaximin therapy. No obvious side effects were observed during the rifaximin therapy. After 12 weeks, 17 (53%) patients had no relapse. There was a trend that those patients (75%, 6 of 8) who had C. difficile strain type BI/NAP1/027 tended to have a relapse more frequently than those who had a non- BI/NAP1/027 strain (42%, 8 of 19), but this trend was not statistically significant (P = 0.11, Fisher exact test).

We then evaluated whether the MIC of rifampin would predict outcome among the 22 patients whose C. difficile isolates were evaluated for antibiotic susceptibility. The analysis revealed that the MIC value of rifampicin predicted the response to rifaximin treatment (P = 0.0461, Mann–Whitney U-test, Figure 1). All the 12 patients who were cured had isolates with rifampin MIC values below 0.3 μg/mL, whereas 5 of the 10 patients who failed had a MIC value above this value (P = 0.010, Fisher exact test). When the epidemiological cut-off value of 0.004 μg/mL was used, which is considered to distinguish wild type C. difficile isolates from isolates with reduced susceptibility to rifampin according to EUCAST, the European Committee on Antimicrobial Susceptibility Testing (http://www.eucast.org), we get an OR 4.62 (95% CI: 0.52–65.4, P = 0.172) of risk of failure.

image

Figure 1. Minimal inhibitory concentrations (MICs) of rifampin (a) and vancomycin (b) of Clostridium difficile strains isolated in patients who were cured and who failed rifaximin treatment. Vancomycin MICs are presented only of those strains isolated from patients who received vancomycin immediately before rifaximin treatment. Closed circles represent strains with a DNA profile compatible with the BI/NAP1/027 ribotype, open circles represent strains of non-BI/NAP1/027 ribotype, and crosses represent a strain that was not typed.

Download figure to PowerPoint

Statistical analysis revealed also that the MIC value of vancomycin was associated with the outcome among those who received vancomycin immediately before the rifaximin therapy (P = 0.0281, Mann–Whitney U-test, Figure 1). However, the absolute difference in the average MIC values between those who were cured and those who were not were not very remarkable (Figure 1).

During the subsequent year, none of the 17 patients who responded to rifaximin developed DCI. Of the 15 who had a failure after rifaximin therapy, 8 patients recovered after receiving faecal transplantation. Two patients recovered after receiving a course of metronidazole. Four patients recovered after receiving one or repeated courses of vancomycin. One patient ultimately recovered after receiving continuous vancomycin for 3 years.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

The outcome after rifaximin in patients with recurrent CDI was promising as 17 (53%) of the 32 patients responded favourably. Although the treatment in almost all patients was preceded by a course of vancomycin or metronidazole, rifaximin most likely contributed to the cure, as there was an association between the susceptibility of C. difficile isolates to rifampin and clinical outcome. Nevertheless, it should be noted that in our series, the course of metronidazole or vancomycin, that immediately preceded the administration of rifaximin, probably had an important role in the cure rate. Rifaximin therapy seemed to be safe as none of our patients experienced any notable side effects during rifaximin treatment.

Despite this, our results were not as effective as reported previously for rifaximin in CDI. In a prospective trial on patients with mild-to-moderate CDI, all sixteen patients (100%) who completed treatment with rifaximin recovered from the infection.[21] In another series of 8 patients with recurrent CDI who received rifaximin, 7 patients (88%) experienced no further diarrhoea recurrences and the patient who had a recurrence subsequently responded favourably to a second course of rifaximin treatment.[22] In yet another study of 6 patients with recurrent CDI who received rifaximin, 4 (66%) patients had no further diarrhoea episodes.[23] Finally, in a randomized placebo-controlled trial where rifaximin was given immediately after standard antimicrobial therapy for CDI, 26 (79%) of the 33 patients responded favourably.[19] Taken together, the overall cure rate of 85% of all the 63 patients in the above four studies is clearly higher than the cure rate of 53% of our 32 patients.

It is possible that our patients were on the average more prone to relapses, which might have contributed to a less favourable response as compared with the results reported previously. All our patients who received rifaximin had recurrent CDI. This was because our standard treatment for the first episode of CDI is metronidazole or vancomycin, which is consistent with current recommendations.[24, 25] There may have been a selection in our series for those patients who were more prone to have relapses. Increased age, initial disease severity, and hospital exposure have been predicted to predict CDI recurrence,[26, 27] the use of proton pump inhibitors have been implicated as a risk factor for the development of CDI,[28, 29] and factors such as increased age, low serum albumin level and high serum creatinine level have been associated with the severity of the disease.[30, 31] We attempted to analyse whether such factors would reveal any association with the success of rifaximin treatment, but our series was too small to address this.

We estimated the sensitivity of C. difficile isolates to rifaximin by measuring the sensitivity to rifampin by an E-test that has been reported to correlate well with rifaximin sensitivity tests.[32-34] Most isolates were seemingly sensitive to rifampin in vitro with a MIC below 0.002 μg/mL. We did not find any isolates that have been considered resistant to rifampicin with a MIC higher than 16 or 32 μg/mL or any highly resistant C. difficile isolates with a MIC higher than 256 μg/mL as reported in some earlier studies.[32-35] The highest MIC value detected in our material was 3 μg/mL, which could be considered as intermediately susceptible to rifampin as MICs ranging from 0.003 to 32 μg/mL have been considered as intermediate.[36] Despite the low MICs of rifampin of our C. difficile strains in vitro, we still had a relatively low cure rate, which is surprising as rifaximin may reach very high concentrations in the stools, on the average, up to 8000 μg/g on the first day after treatment.[37] It is possible that C. difficile organisms in the large bowel reside in biofilms in a rifaximin-resistant phenotype, as C. difficile spores, or in specific bowel microenvironments including intracellular locations where adequate concentrations of rifaximin may not be achieved.

Of note was that the MIC value of rifampin, although all the MIC values of our strains were rather low, seemed to predict the response to rifaximin treatment. This may imply that sensitivity testing may be used to select patients with extremely low MIC values for rifaximin treatments. On the other hand, there was a correlation between the BI/NAP1/027 strain type of C. difficile and an increased MIC value of rifampin as well as to metronidazole that have also been observed in earlier studies.[38-41] Thus, specific pathogenic factors of the BI/NAP1/027 isolates may be responsible for the difference in the cure rate and not necessarily the low MIC value. This ribotype has been suggested to cause a more severe disease than other ribotypes.[42] However, recent reports suggest that patient characteristics, such as white blood cell count and serum albumin level, are more important than the ribotype in predicting the severity of the disease and that other characteristics of the C. difficile genome than the plain ribotype can significantly influence virulence.[43] Nevertheless, rifaximin may be more efficacious against non-BI/NAP1/027 strains with extremely low MIC values of rifampin than against other strains.

Development of resistance against rifaximin is a potential concern. Resistance to rifamycins results from mutations in the rpoB gene that encodes the beta subunit of RNA polymerase. Several different rpoB gene mutations have been found in resistant C. difficile strains (MICs higher than 32 μg/mL).[32] Prior exposure to rifamycins has been reported to be a risk factor for rifampin-resistant C. difficile infection[36] and rifaximin-resistant C. difficile strains may emerge even during therapy.[23, 44] Although the rate of rifaximin resistance among C. difficile isolates has been as low as 2%,[45] the rate of resistance may reach 20%[33] and rifamycin-resistant C. difficile strains have been reported to cause hospital outbreaks.[36] Of concern are also reports that suggest that the antibiotic resistance of C. difficile may be changing to the worse.[35] These issues on resistance restrict the use of rifaximin in CDI and highlight the importance of sensitivity testing of C. difficile isolates.

Our results suggest that rifaximin can be used for the treatment of recurrent CDI. Despite the development of other therapeutic approaches including faecal microbiota transplantation,[5, 7] various forms of immunotherapy,[2] and new antibiotics,[10] treatment of CDI remains a challenge and rifaximin may prove to be a valuable addendum to the armament against CDI.

Authorship

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

Guarantor of the article: E. Mattila.

Author contributions: All authors participated in the design of the study, contributed to the writing of the manuscript and approved the final version of the manuscript. EM, V-J A, PSM and PA participated in the collection and analysis of data from the patient records. PT and ET participated in the analysis of stool samples, identification of C. difficile isolates, typing of the strains and analysis of the antibiotic susceptibility testing of the C. difficile isolates.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References

Declaration of personal interests: EM has received research funding from the Finnish Foundation for Gastroenterological Research and the Helsinki University Central Hospital Research Funds. V-J A has received research funding from Merck, Pfizer and GSK. PA has served as a speaker for Merck and Abbot. EM and V-J A have served as members of the Astellas Nordic Advisory Board.

Declaration of funding interests: None.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. Authorship
  8. Acknowledgements
  9. References