Efficacy of Bacillus probiotics in prevention of antibiotic-associated diarrhoea : a randomized , double-blind , placebo-controlled clinical trial

Introduction: Antibiotic-associated diarrhoea (AAD) is one of the most common side effects of antibiotic therapy. The main mechanism associated with the development of AAD is significant changes in the composition and quantity of the gut microbiota during the treatment with antibiotics. Probiotic bacteria have been shown to stabilize the gut microbiota and can be used to prevent diarrhoea associated with antibiotic therapy.


Introduction
One of the most common side effects of antibiotic therapy is antibiotic-associated diarrhoea (AAD).The frequency of AAD depends on the antibiotic used and varies from 2 to 25 % (Bartlett, 2002) and can be as high as 44 % (Gao et al., 2010).The route of antibiotic administration (oral or parenteral) does not affect the rate of AAD (Bartlett, 2002), and no difference has been found in frequency of AAD with respect to age and gender (Wistrom et al., 2001).The severity of AAD may vary from uncomplicated diarrhoea to Clostridium difficile-associated pseudomembranous colitis.AAD may be caused by different enteric pathogens (e.g.Salmonella spp., Staphylococcus aureus, Candida albicans, Clostridium perfringens and Klebsiella spp.) (Bartlett, 2002;Gorkiewicz, 2009).The main mechanism for the development of AAD is significant changes in the composition and quantity of the gut microbiota during the treatment with antibiotics (Young and Schmidt, 2004).Probiotic bacteria have been shown to stabilize the gut microbiota and to prevent diarrhoea associated with antibiotic use (Friedman, 2012;Persborn et al., 2013).Bacillus bacteria have attracted the growing attention of researchers as effective probiotics for the prevention and treatment of enteric infections (Coppi et al., 1985;Canani et al., 2007;Mazza, 1994).Previously, we showed a high efficacy of the Bacillus probiotic Biosporin in the treatment of acute intestinal infections (Gracheva et al., 1996).Biosporin is a mix of two strains -B.subtilis 3 and B. licheniformis 31, with a predominant amount of B. subtilis 3 (50:1).These strains have been deposited in the Russian National Collection of Industrial Micro-organisms (VKPM) as B. subtilis B2335 and B. licheniformis B2336.Both Bacillus strains have been thoroughly tested in preclinical testing (Pinchuk et al., 2001;Sorokulova, 2008;Sorokulova et al., 1997), including safety evaluations (Sorokulova et al., 2008), and in clinical trials (Gracheva et al., 1996;Sorokulova et al., 2003).Biosporin is approved by the National Control Authority for Biological Products of Ukraine and Russia as a drug for adults and children.Further studies of this probiotic showed that B. subtilis 3 had a higher antagonistic activity against pathogens than B. licheniformis 31 (Pinchuk et al., 2001;Sorokulova et al., 1997).
Our hypothesis was that Bacillus probiotics would reduce the frequency of AAD in patients treated with antibiotics.We also postulated that the efficacy of one B. subtilis 3 strain with pronounced antagonistic activity against pathogens would be the same as the efficacy of a twoprobiotic culture (B. subtilis 3 and B. licheniformis 31).

Case report
The main objectives of this study were to analyse efficacy of Bacillus probiotics in the prevention of AAD and to compare the efficacy of two Bacillus probiotics in a singlecentre, randomized, double-blinded, placebo-controlled clinical trial.The study was conducted at the Medical Center Vitbiomed OOO in Moscow, Russia, from October 2011 to November 2012.The study protocol was approved by the internal review board of the centre.Written informed consent was obtained from each patient.Adult outpatients aged 45 years or over who were anticipated to take one or more oral or intravenous antibiotics for at least 5 days were recruited for this trial.
Exclusion criteria were existing diarrhoea, Clostridium difficile-associated diarrhoea (CDD) within the previous 3 months, use of probiotics in the previous 3 weeks, adverse reactions to microbial preparations in the past and immunocompromised patients.Patients were assigned to each group by simple randomization procedures (computergenerated random numbers).
Each patient received a coded reporting form to record clinical data.Patients were instructed to record daily the number of faecal outputs and their consistency, and the presence of vomiting, nausea, abdominal pain, bloating and any other adverse events.

Investigation and treatment
Patients were randomized into three groups: (i) probiotic group 1, who received a probiotic, containing strains B. subtilis 3 and B. licheniformis 31 (2610 9 c.f.u. per vial); (ii) probiotic group 2, who received a probiotic containing B. subtilis 3 (2610 9 c.f.u. per vial); and (iii) a placebo group, who received an inert composition in vials, formulated to be indistinguishable from the vials with probiotics.
Maltodextrin was used as an inert component in all vials (probiotics and placebo).The volume of all vials was 2 ml.The probiotic vials did not vary from placebo in appearance, colour, taste or size.All vials were prepared by Vitbiomed OOO by independent production staff who did not have patient contact or data management responsibilities.
Bacterial cultures were grown on agar medium and harvested with sterile PBS.The number of live bacteria (c.f.u.) was identified by plating of 10-fold dilutions of bacterial suspension on nutritient agar plates.The bacterial suspension was diluted with sterile PBS to obtain 1610 9 c.f.u.ml 21 .This suspension was dispensed into opaque plastic vials in 2 ml volumes to obtain 2610 9 c.f.u. in each vial.
The study was conducted as a double-blinded clinical trial.The allocation sequence was performed by independent personnel and was unavailable to members of the research team and patients until data analysis had been completed.Patients, clinical staff and biostatisticians were blinded to treatment allocation throughout the trial.
The assigned intervention started 1 day before the beginning of antibiotic therapy and continued for 7 days after discontinuation of antibiotics.Participants received one vial (2 ml dose) in the morning and one in the evening prior to their meals.Patients were followed up for 4 weeks after stopping the antibiotics, unless AAD occurred before that time.The primary outcomes were the incidence of AAD during the study period.The diarrhoea was defined as three or more loose or watery stools day 21 for at least 2 days.In the case of diarrhoea, a stool sample was collected and analysed for the presence of Clostridium difficile toxins by an immunoenzymatic assay (C.difficile TOX A/B II; TechLab).
Secondary outcomes included the occurrence of CDD, abdominal symptoms, adverse effects and the acceptability of the probiotics.
In order to achieve a statistical power of 85 % at a 5 % significance level, we estimated the sample size of this trial to be n5270 (90 per group), allowing for a maximum 10 % dropout rate in each group.A x 2 test was used to compare all three groups.The x 2 test with Bonferroni correction [a (0.05)/number of tests (n53)] was applied for further pairwise comparison.Additionally, we calculated the relative risk (RR) and odds ratio (OR) together with their 95 % confidence interval (CI) for each of the characteristics.We also calculated the absolute risk reduction (ARR) when comparing treatment groups.Continuous variables were summarized using (i) the number of observations; (ii) the median and interquartile range; or (iii) the mean and SD.

Outcomes
The patient flow is summarized in Fig. 1.In all, 574 patients were screened for participation and 308 were enrolled in the study.The major reasons for exclusion were unwillingness to participate (221, 38.5 %), existing diarrhoea (15, 2.6 %), use of probiotics in the previous 3 weeks (19, 3.3 %) and immunocompromised patients (11, 1.9 %).Follow-up was not completed for 37 (12 %) patients because of loss of contact (23, 7.5 %) or because they had withdrawn from the study (14, 4.5 %) (Fig. 1).Thus, analysis was performed for 271 patients (n591 for probiotic mix, n590 for probiotic and n590 for placebo).
There was no difference in baseline characteristics of patients in the three groups (Table 1).Patients in all three groups were matched for gender and age.The indications for treatment with antibiotics were similar in all groups with a prevalence of respiratory and urinary tract infections (Table 1).The antibiotic regimen exposure was similar among the three groups.
Both probiotics were well tolerated and no probioticrelated adverse events were noted.

Discussion
The results of our study showed the efficacy of Bacillus probiotics in prevention of AAD during treatment with antibiotics.Beneficial effects of probiotics were also found in significant mitigation of secondary outcomes: abdominal pain, nausea, bloating and vomiting.These findings are in accordance with our previous data about the high efficacy of Bacillus probiotic in the elimination of abdominal symptoms in patients with acute intestinal infections (Gracheva et al., 1996).This study demonstrated that the efficacy of a singlestrain probiotic (B.subtilis 3) was the same as a mix of two Bacillus strains (B.subtilis 3 and B. licheniformis 31).The therapeutic success of probiotics depends on a variety of beneficial effects that are unique for each probiotic strain.We can assume that the efficacy of both tested probiotics was determined by the high probiotic activity of B. subtilis 3 strain, as indicated in vitro (Pinchuk et al., 2001;Sorokulova et al., 1997).To the best of our knowledge, this is the first report on the clinical efficacy of B. subtilis probiotic in the prevention of AAD.Previously, the Bacillus clausii probiotic strain was found to be effective in the reduction of diarrhoea incidents related to anti-Helicobacter pylori antibiotic therapy (Nista et al., 2004).Another probiotic strain, Bacillus coagulans GBI-30, 6086, was effective in decreasing abdominal pain and bloating symptoms in patients with inflammatory bowel disease (Hun, 2009) and in reducing daily bowel movements in patients with irritable bowel syndrome (Dolin, 2009).In patients with post-prandial gas-related symptoms, B. coagulans GBI-30, 6086 probiotic strain was effective in improving the quality of life and reducing gastrointestinal symptoms (abdominal pain, abdominal distention and flatus) (Kalman et al., 2009).The mechanism of the beneficial effects of Bacillus probiotics on the gastrointestinal tract is not completely known, but some properties known for Bacillus bacteria may contribute to their efficacy.Bacillus bacteria are metabolically highly active and they produce anti-microbial substances, amino acids and vitamins (Sorokulova, 2008).They could support digestive function of the gut by producing essential enzymes (proteolytic, lipolytic or cellulolytic).Bacillus bacteria may provide the host with the ability to maintain intestinal homeostasis.The quorum-sensing pentapeptide of B. subtilis, competence and sporulation factor, activates key survival pathways in intestinal epithelial cells of the host (Fujiya et al., 2007).It was shown that this competence and sporulation factor induces heat-shock proteins, which protect intestinal epithelial cells against injury and loss of barrier function.Clinical efficacy of Bacillus bacteria in the treatment of gastrointestinal infections has been reported by authors from different counties.Mazza (1994) summarized the main results of these studies and concluded that B. subtilis is one of the most important micro-organisms for the therapy and prophylaxis of intestinal disorders in humans.In clinical trials, Bacillus probiotic was more effective in the treatment of acute intestinal infections than Lactobacillus probiotic (Gracheva et al., 1996).
Our study demonstrated that treatment with Bacillus probiotics during antibiotic therapy significantly decreased the incidence of AAD and adverse effects related to antibiotics.Both probiotics were well tolerated by the patients without side effects.No significant difference was found in the efficacy of one probiotic strain (B.subtilis 3) in comparison with a mix of two strains (B.subtilis 3 and B. licheniformis 31).

Table 1 .
Baseline characteristics of patients and outcomes