sign in sign up
<<
Home , Metronidazole, Rifaximin

European Review for Medical and Pharmacological Sciences 2009; 13: 111-116 therapy in small intestinal bacterial overgrowth: versus

E.C. LAURITANO, M. GABRIELLI, E. SCARPELLINI, V. OJETTI, D. ROCCARINA, A. VILLITA, F. FIORE, R. FLORE, A. SANTOLIQUIDO, P. TONDI, G. GASBARRINI, G. GHIRLANDA, A. GASBARRINI

Internal Medicine Department, Catholic University of the Sacred Heart, Rome (Italy)

Abstract. – Background and Objectives: Few controlled trials on antibiotic therapy for GBT: glucose breath test small intestinal bacterial overgrowth are avail- H2: hydrogen able at present. Aim of the study was to assess IBS: efficacy, safety and tolerability of rifaximin with respect to metronidazole for the treatment of small intestinal bacterial overgrowth. Material and Methods: We enrolled 142 con- secutive patients with diagnosis of small intesti- Introduction nal bacterial overgrowth. Diagnosis of small in- testinal bacterial overgrowth based on the clini- Small intestinal bacterial overgrowth (SIBO) is cal history and the positivity of glucose breath a common clinical syndrome due to an increased test. Patients were randomised to two 7-day level of bacteria exceeding the presence of more treatment groups: rifaximin 1200 mg/day and than 105 CFU/mL of intestinal aspirate or of metronidazole 750 mg/day. Glucose breath test colonic-type species within the small bowel1,2. was reassessed 1 month after. Compliance and side-effect incidence were also evaluated. SIBO symptoms could be many and variably Results: One drop-out was observed in rifax- associated; abdominal pain or discomfort, bloat- imin group. Five drops-out occurred in metron- ing, diarrhoea and/or signs of malabsorption are idazole group. The glucose breath test normal- the most common1,2. Recent findings suggest that ization rate was significantly higher in the rifax- SIBO is highly prevalent in patients with IBS and imin with respect to the metronidazole group that SIBO decontamination is associated to a sig- (63.4% versus 43.7%; p<0.05; OR 1.50, 95% CI nificant improvement of IBS symptoms3-5. 1.14-4.38). The overall prevalence of adverse events was significantly lower in rifaximin with The culture of jejunal aspirates, regarded by respect to metronidazole group. many as the gold standard for the SIBO diagno- Discussion: Rifaximin showed an higher SIBO sis, has several limitations such as the potential decontamination rate than metronidazole at the for contamination by oropharyngeal bacteria dur- tested doses, both with a significant gain in ing intubation, and the fact that SIBO may be terms of tolerability. Either the present study or patchy and thus missed by a single aspiration. In recent evidencies suggest that rifaximin repre- sents a good choice for the management of pa- addition, it is too much invasive, expensive and tients affected by SIBO. difficult and too little reproducible to be pro- posed as a routine diagnostic test for SIBO in the Key Words: clinical practice, especially for patients with non- Bacterial overgrowth, Glucose breath test, Rifax- specific symptoms or those requiring repeated 1,2 imin, Metronidazole. testing . The GBT is considered a simple tool for SIBO diagnosis, since it is non invasive, highly repro- ducible and inexpensive when compared to the culture of jejunal aspirates. In addition, the speci- Abbreviatons ficity and the sensitivity of GBT are acceptable for screening studies (77-100% and 67-98% re- 6-8 SIBO: small intestinal bacterial overgrowth spectively) . The H2 produced in the human CFU: colony-forming unit body after glucose ingestion derives entirely

Corresponding Author: Antonio Gasbarrini, MD; e-mail: [email protected] 111 E.C. Lauritano, M. Gabrielli, E. Scarpellini, V. Ojetti, D. Roccarina, A. Villita, et al. from intestinal bacterial fermentation. The ap- Major organic gastrointestinal disorders were pearance of an early increase in breath H2 con- ruled out on the basis of: history collection; full centration suggests the presence of SIBO6-8. physical examination; laboratory tests (total An effective antibiotic decontamination regi- blood count, erythrocyte sedimentation rate, re- men should include one or more drugs with ac- active C protein, stool examination for occult tivity against both aerobic and anaerobic bacte- blood, ova and parasites, anti-transglutaminase ria since SIBO may occur either by a mix of aer- antibodies); abdominal ultrasonography and obic and anaerobic flora or by purely aerobic colonoscopy when alarm symptoms were present flora in a minority of cases1,2,9-11. Empirical (fever, gastrointestinal bleeding, weight loss, courses of broad-spectrum are widely anemia, abdominal mass). used at present for SIBO decontamination, since The Rome II criteria were used to verify the few well-conducted trials have been performed diagnosis of IBS or other functional bowel disor- up today to verify which is the best antibiotic ders19. regimen1,2,11,12. The exclusion criteria were: previous antibiot- Metronidazole is effective against Gram-nega- ic treatment associated to SIBO diagnosis; age tive and Gram-positive anaerobic bacteria such <18 years; use of antimicrobial agents within the as Bacterioides, and Peptostrep- previous 3 months; hypersensitivity to the antibi- tococci13. These characteristics make it potential- otics used in the present study; pregnancy or ly useful for the treatment of small bowel bacter- breast-feeding; evidence of major concomitant ial overgrowth as confirmed by literature data14. diseases (including tumours and hepatic and/or Rifaximin is a derivative with an- renal insufficiency). tibacterial activity caused by inhibition of bacter- Exclusion criteria were: age <18 years; hyper- ial synthesis of RNA15. It is active against gram- sensitivity to the antibiotics; pregnancy or breast- positive and gram-negative bacteria, including feeding; evidence of major concomitant diseases both aerobes and anaerobes15-17. Less than 0.1% (including tumours and hepatic and/or renal in- of the oral dose is absorbed16. Rifaximin at a sufficiency). Consecutive patients with positive dosage of 1200 mg per day for 1 week is associ- GBT were included in the present study after in- ated to a significant gain in terms of therapeutic formed consent. efficacy in SIBO contamination without increas- The procedures followed were in accordance ing the incidence of side-effects with respect to with with the Helsinki Declaration of the World lower dosages (600 and 800 mg per day for the Medical Association. same treatment period)18. The aim of the present study was to assess the Laboratory Parameters efficacy, safety and tolerability of the non-ab- Total blood cell count, and kidney func- sorbable antibiotic rifaximin with respect to the tion were assessed in all patients at enrolment systemic antibiotic metronidazole in patients af- and 3 days after the end of the treatment. fected by SIBO.

Breath H2 Testing GBT was performed under standard condi- tions. In the month preceding the test patients Material and Methods should not have received antibiotics or laxatives. To minimize basal H2 , patients were This prospective parallel-group randomized asked to follow a carbohydrate-restricted dinner trial was conducted between February 2005 and on the day before the test and to fast for at least August 2007 in consecutive out-patients from 12 hours. On the day of testing, patients did a the Gastroenterology and Internal Medicine De- mouthwash with 20 ml of 0.05%. partments of the Catholic University of Rome, Smoking and physical exercise were not allowed Italy. for 30 minutes before and during the test. End- alveolar breath samples were collected immedi- Inclusion/Exclusion Criteria ately before glucose ingestion. A dose of 50 g of Patients referring to our centre for the pres- glucose in the form of iso-osmotic solution was ence of gastrointestinal symptoms (, ab- administered and samples were taken every 10 dominal pain, and diarrhoea) since ≥ 6 min for 2 hours respectively using a two-bag months were evaluated. system. The two-bag system is a device consist-

112 Antibiotic therapy in small intestinal bacterial overgrowth: rifaximin versus metronidazole ing of a mouthpiece, a T valve and two collapsi- domly assigned to one of the two 7-day treatment ble bags, the first one collects dead space air, the groups: second one collects alveolar air. The breath sam- ple was aspirated from this bag into a 20 ml 1. Rifaximin 1200 mg/day (2 tablets tid; group 1: plastic syringe. Samples were analyzed immedi- n=71) ately using a model Quintron Gas Chromato- 2. Metronidazole 750 mg/day (1 tablet tid; group graph (Quintron Instrument Company, Milwau- 2: n=71). kee, WI). The test was considered as indicative of the Data Analysis presence of SIBO when the peak, that is the in- For the purpose of the analysis, the incidence crease over the baseline of H2 levels, was of side-effects was considered as a binomial vari- >12 p.p.m.7. able (present/absent). To detect differences in The GBT was repeated 1 month after the end GBT normalization rates and the incidence of of therapy to assess GBT normalization. side-effects, the Χ2 or Fisher’s exact tests were used. Odds Ratio (OR) for achieving GBT nor- Outcomes malization with 95% confidence intervals (95% The primary outcome of the present study was CI) was calculated. The statistical analysis was the GBT normalization rate using rifaximin or performed using STATA 6.0. metronidazole. Secondary outcomes were patient’s compli- ance and incidence of side-effects in the two therapeutic schemes. Results Compliance was assessed by an interview (ad- ministered by a trained physician) performed af- Patients Characteristics and ter the end of the therapy and by a pill count of Overall Compliance the drugs boxes returned at the same interview. A total of 142 patients were enrolled. Charac- Low compliance was defined as more than 20% teristics of the study groups are summarized in of pills returned. Table I. Groups were similar for sex, age, body At enrolment the patients were informed on mass index and the prevalence of irritable bowel the common side-effects expected from the syndrome (IBS) and of the other subtypes of studied therapies. Side effects were defined as functional bowel disorders. the occurrence of: a) abnormalities in the main The overall compliance of the present study haematochemical parameters considered; b) was excellent: 135 of 141 completed the studied “adverse experiences”, considered as clinical therapeutic regimens. However the incidence of findings or patient complaints that were not drop-outs was higher in the group of patients present in the 24 hours immediately before the enrolment in the trial. All patients were asked to fill in a validated questionnaire (modified DeBoer) in order to report therapy related ad- Table I. Demographic and clinical characteristics of SIBO verse experiences (diarrhoea, taste disturbance, positive patients included in the two treatment groups , bloating, loss of appetite, , ab- (group 1=rifaximin; group 2=metronidazole). BMI, body dominal pain, , headache, skin mass index; IBS, irritable bowel syndrome; FAD, functional abdominal disorder. rash)20. Each symptom was graded from absent (0) to severe (interruption of treatment, 3) Characteristics Group 1 Group 2 P based on the intensity. The questionnaire was (n = 71) (n = 71) administered at enrolment and the patients were asked to complete diary cards in the same Age (years) 34 ± 9 35 ± 11 ns Males (%) 35% 40% ns format (Likert scales) during the treatment pe- BMI (kg/m2) 22 ± 6 22 ± 7 ns riod and to return them at the post-therapy in- IBS 45% 39% ns terview. Functional abdominal 21% 25% ns bloating Randomization Functional abdominal 15% 14% ns pain Using a computer-generated number sequence, Unspecified FAD 19% 22% ns generated by a statistician, the patients were ran-

113 E.C. Lauritano, M. Gabrielli, E. Scarpellini, V. Ojetti, D. Roccarina, A. Villita, et al. treated with metronidazole with respect to those Discussion treated by rifaximin. In fact, 5 drop-outs occurred in the metronidazole group (one for the inability Empirical courses of broad-spectrum antibi- to maintain appointments and 4 for the occur- otics are widely used at present in the clinical rence of side-effects), with respect to 1 in the ri- setting for SIBO decontamination1,2,10-12.This faximin group (for the inability to maintain ap- occurs for several reasons. Few literature data pointments). are available on the bacterial population conta- minating the small bowel and its antibiotic sen- GBT Normalization Rate sitivity patterns1,2,9. On the other hand, although The glucose breath test normalization rate was ideally the choice of agent should significantly higher in the rifaximin with respect reflect susceptibility testing, this usual- to the metronidazole group in intention-to-treat ly is impractical in the case of SIBO because analysis (63.4%, 45/71 versus 43.7%, 31/71; many different bacterial species, with different p<0.05; OR 1.50, 95% CI 1.14-4.38). No signifi- antibiotic sensitivities, typically coexist1,2,9. In cant differences were found between groups in addition, few well-conducted clinical trials has per protocol analysis. been performed in order to assess the most ef- fective and safe antibiotic regimen for SIBO Side-Effects Profile decontamination1,2. No abnormalities in the tested laboratory para- Metronidazole may be suitable for SIBO treat- meters were observed in the two groups at the ment since it is effective against Gram-negative control performed three days after the end of the and Gram-positive anaerobic bacteria such as treatment. Details on the incidence of adverse , Fusobacterium and Peptostrepto- events during the study period are reported in cocci13,14. Castiglione et al14showed a good thera- Table II. The overall incidence of adverse events peutic efficacy of both metronidazole and was 15.5% (22/142). The overall incidence of ad- in terms of breath test normaliza- verse events was significantly higher in the tion rate in patients affected by Crohn’s disease metronidazole with respect to the rifaximin and evidence of SIBO. group (22.5%, 16/71 versus 8.5%, 6/71; OR However, all systemic antibiotics, also if en- 1.59, 95% CI 1.15-8.61). dowed with a satisfactory efficacy profile, are as- In the metronidazole group, 3 patients reported sociated to several side-effects such as diarrhoea, adverse events graduated as moderate and 4 pa- constipation, dizziness, weakness, skin rush and tients abandoned the study because of the occur- dyspepsia. The safety and tolerability of an an- rence of severe adverse events. The 6 adverse ex- tibiotic treatment are as important as its efficacy, periences observed in the rifaximin group were especially in a disease as SIBO that is character- all mild. ized by high recurrence rate and necessity of re- peated antibiotic courses. Non-absorbable antibiotics such as and rifaximin, both able to act against bacteria topically within the gut lumen, have been pro- Table II. Adverse events during the study period in the two posed for treatment of SIBO in order to minimize treatment groups (group 1=rifaximin; group 2=metronida- zole). the potential side-effects of systemic antibiotics. The neomycin treatment achieved the normal- Adverse events Group 1 Group 2 ization of breath test in 20% of patients (n=71) (n=71) carrying SIBO with respect to 2% in the placebo group; no relevant side-effects and no drop-out Skin rush 0 1 were observed during the study21. The high bind- Taste disturbance 0 3 Bloating 2 2 ing (about 90%) of neomycin with faeces could Abdominal pain 0 2 explain the limited in vivo activity21. Nausea/vomiting 1 3 Rifaximin has a broad-spectrum antibiotic effi- Diarrhoea 0 2 cacy, especially against anaerobic intestinal bac- Constipation 3 1 teria, such as Bacteroides, Lactobacilli and Weakness 0 0 Loss of appetite 0 2 Clostridia, bacteria frequently responsible for Others 0 0 metabolic alterations observed in SIBO pa- Overall 6 16 tients15-17. Its toxicity is very low since it is not

114 Antibiotic therapy in small intestinal bacterial overgrowth: rifaximin versus metronidazole absorbed by the gut16. In a double-blind con- ritable bowel syndrome. Aliment Pharmacol Ther trolled trial Di Stefano et al. compared the effica- 2005; 22: 1157-1160. cy of rifaximin (1200 mg/die) with respect to 4) PIMENTEL M, CHOW EJ, LIN HC. Eradication of small in the short-term treatment of intestinal bacterial overgrowth reduces symptoms SIBO. GBT normalized in 70% of patients treat- of irritable bowel syndrome. Am J Gastroenterol 2000; 95: 3503-3506. ed with rifaximin with respect to 27% of chlorte- tracycline group. No side-effect occurred, thus 5) CARRARA M, DESIDERI S, AZZURRO M, BULIGHIN GM, DI confirming that rifaximin is a safe drug for SIBO PIRAMO D, LOMONACO L, ADAMO S. 22 bacterial overgrowth in patients with irritable bow- treatment . In a recent study by our group, high- el syndrome. Eur Rev Med Pharmacol Sci 2008; er doses of rifaximin (1200 mg/day) led to a sig- 12: 197-202. nificantly higher therapeutic efficacy in terms of 6) KERLIN P. Glucose-H2 breath test for small intesti- SIBO contamination with respect to doses of 600 nal bacterial overgrowth. Gastroenterology 1990; and 800 mg per day. Moreover, at the tested dos- 98: 253-254. es, rifaximin was associated to uncommon, mild, 7) ROMAGNUOLO J, SCHILLER D, BAILEY RJ. Using breath transient side-effects and no drop-out was regis- tests wisely in a gastroenterology practice: an ev- tered18. Another advantage of rifaximin concerns idence-based review of indications and pitfalls in antibiotic resistance: it has been demonstrated interpretation. Am J Gastroenterol 2002; 97: that resistant strains rapidly disappear from the 1113-1126. gut thus allowing cyclic administration of rifax- 8) KERLIN P, W ONG L. Breath hydrogen testing in bac- imin15,16. terial overgrowth of the small intestine. Gastroen- In the present study we tested efficacy, safety terology 1988; 95: 982-988. and tolerability of rifaximin with respect to a sys- 9) BOUHNIK Y, A LAIN S, ATTAR A, FLOURIÉ B, RASKINE L, temic antibiotics such as metronidazole for SIBO SANSON-LE PORS MJ, RAMBAUD JC. Bacterial popula- tions contaminating the upper gut in patients with decontamination. Rifaximin showed an higher small intestinal bacterial overgrowth syndrome. decontamination rate compared to the absorbable Am J Gastroenterol 1999; 94: 1327-1331. antibiotic metronidazole, both with a significant 10) TOSKES PP, KUMAR A. Enteric bacterial flora and bac- gain in terms of tolerability. terial overgrowth syndrome. In: Felman, M, In conclusion, the present data, both with Sleisenger, MH, Scharschmidt, BF, eds. Sleisen- available previous literature evidencies, suggest ger and Fordtran’s Gastrointestinal and Liver Dis- that rifaximin may represent a good option for ease: pathophysiology, diagnosis, management. Philadelphia, USA: WB Saunders, 1998: pp. SIBO decontamination in consideration of its 1523-1535. good GBT normalization rate, null toxicity, high tolerability. Future studies should be addressed to 11) VAN CITTERS GW, LIN HC. Management of small in- testinal bacterial overgrowth. Curr Gastroenterol the management of SIBO patients refractory to Rep 2005; 7: 317-320. the current rifaximin decontamination scheme 12) DI STEFANO M, MICELI E, MISSANELLI A, CORAZZA GR. and to verify its efficacy in the re-treatment of Treatment of small intestine bacterial overgrowth. patients with SIBO recurrence. Eur Rev Med Pharmacol Sci 2005; 9: 217-222.

13) FREEMAN CD, KLUTMAN NE, LAMP KC. Metronidazole. A therapeutic review and update. Drugs 1997; 54: 679-708.

14) CASTIGLIONE F, R ISPO A, DI GIROLAMO E, COZZOLINO A, References MANGUSO F, G RASSIA R, MAZZACCA C. Antibiotic treat- ment of small bowel bacterial overgrowth in pa- 1) SINGH VV, TOSKES PP. Small bowel bacterial over- tients with Crohn’s disease. Aliment Pharmacol growth: presentation, diagnosis, and treatment. Ther 2003; 18: 1107-1112. Curr Treat Options Gastroenterol 2004; 7: 19-28. 15) HOOVER WW, GERLACH EH, HOBAN DJ, ELIOPOULOS 2) QUIGLEY E, QUERA R. Small intestinal bacterial GM, PFALLER MA, JONES RN. Antimicrobial activity overgrowth: roles of antibiotics, prebiotics, and and spectrum of rifaximin, a new topical rifamycin probiotics. Gastroenterology 2006; 130(2 Suppl. derivative. Diagn Microbiol Infect Dis 1993; 16: 1): 78-90. 111-118.

3) LUPASCU A, GABRIELLI M, LAURITANO EC, SCARPELLINI E, 16) GILLIS JC, BROGDEN RN. Rifaximin. A review of its SANTOLIQUIDO A, CAMMAROTA G, FLORE R, TONDI P, P O- antibacterial activity, pharmacokinetic properties LA P, G ASBARRINI G, GASBARRINI A. Hydrogen glucose and therapeutic potential in conditions mediated breath test to detect small intestinal bacterial by gastrointestinal bacteria. Drugs 1995; 49: 467- overgrowth: a prevalence case-control study in ir- 484.

115 E.C. Lauritano, M. Gabrielli, E. Scarpellini, V. Ojetti, D. Roccarina, A. Villita, et al.

17) GIONCHETTI P, R IZZELLO F, V ENTURI A, UGOLINI F, R OSSI 20) DE BOER WA, THYS JC, BORODY TJ, GRAHAM DY, M, BRIGIDI P, J OHANSSON R, FERRIERI A, POGGIOLI G, O’MORAIN C, TYTGAT GN. Proposal for use of a CAMPIERI M. Review–antibiotic treatment in inflam- standard side effect scoring system in studies matory bowel disease: rifaximin, a new possible exploring treatment regi- approach. Eur Rev Med Pharmacol Sci 1999; 3: mens. Eur J Gastroenterol Hepatol 1996; 8: 27-30. 641-643.

18) LAURITANO EC, GABRIELLI M, LUPASCU A, SANTOLIQUIDO 21) PIMENTEL M, CHOW EJ, LIN HC. Normalization of A, NUCERA G, SCARPELLINI E, VINCENTI F, C AMMAROTA lactulose breath testing correlates with symp- G, FLORE R, POLA P, G ASBARRINI G, GASBARRINI A. Ri- tom improvement in irritable bowel syndrome. a faximin dose-finding study for the treatment of double-blind, randomized, placebo controlled small intestinal bacterial overgrowth. Aliment study. Am J Gastroenterol 2003; 98: 412-419. Pharmacol Ther 2005; 22: 31-35. 22) DI STEFANO M, MALSERVISI S, VENETO G, FERRIERI A, 19) THOMPSON WG, LONGSTRETH GF, DROSSMAN DA, CORAZZA GR. Rifaximin versus chlortetracycline in HEATON KW, IRVINE EJ, MÜLLER-LISSNER SA. Function- the short term treatment of small intestinal bacter- al bowel disorders and functional abdominal pain. ial overgrowth. Aliment Pharmacol Ther 2000; 14: Gut 1999; 45(Suppl. 2): 43-47. 551-556.

–––––––––––––––––––– Acknowledgements

This work was supported by an unrestricted grant pro- vided by Fondazione Ricerca in Medicina.

116