Doi:10.1016/j.ijantimicag.2005.04.004

International Journal of Antimicrobial Agents 26 (2005) 69–74 Effects of probiotics on the composition of the intestinal microbiota Susan F. Plummer , Iveta Garaiova , Tinnu Sarvotham , Simon L. Cottrell , Stephanie Le Scouiller , Mark A. Weaver , James Tang , Philippa Dee , John Hunter a Cultech Biospeciality Products, Research Department, York Chambers, York Street, Swansea SA1 3NJ, UK b University of Wales, College of Medicine, Cardiff CF14 4XN, UK c Cardiff School of Biosciences, Cardiff University, Cardiff CF10 3XQ, UK d School of Construction Management and Engineering, The University of Reading, Reading RG6 6AH, UK e Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK Received 14 February 2005; accepted 6 April 2005 Abstract
The effects of probiotic supplementation on the intestinal re-growth microbiota following antibiotic therapy were studied in a double-blind placebo-controlled study. In the placebo group, numbers of facultative anaerobes and enterobacteria increased significantly, and at day 35 thenumbers were significantly higher in the placebo group than in the active group; in the active group, the numbers of bacteroides increasedsignificantly. Although the numbers of enterococci in both groups did not change, in the placebo group the number of patients harbouringantibiotic-resistant enterococci post therapy increased significantly. There was no change in the incidence rate of antibiotic resistance amongthe patients in the probiotic group.
2005 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
Keywords: Intestinal microbiota; Antibiotics; Probiotics; Antibiotic resistance 1. Introduction
Disturbance of the microbiota is frequently associated with diarrhoea, gastritis, glossitis and pruritus well as The bacterial flora of the gastrointestinal tract play a fungal infections. In addition, altered sensitivity to secondary major role in human physiology, modulating metabolic and infection can occur. A single oral dose of streptomycin can immunological processes and providing colonisation resis- enhance susceptibility of laboratory animals to challenge tance, which is the prevention of overgrowth of opportunistic by Salmonella spp. by at least 100 000-fold Another microorganisms. Administration of antimicrobial agents, important and growing area of concern is the effect of whether therapeutically or prophylactically, disturbs the eco- antibiotics on the colonisation resistance properties of the logical balance between the host and the normal microbiota indigenous microbiota resulting in the emergence and spread extent of the disturbance depends on the nature of the of resistant strains between patients and the dissemination antimicrobial agent, the absorption, the route of elimination of resistance determinants between microorganisms and any potential enzymatic degradation and/or binding to Reid and Friendship that in 1998 the World Health faecal material. However, predicting the effects of an antibi- Organization cited diarrhoeal diseases as the second most otic on the microbiota can be difficult due to the complex common cause of disability-adjusted life-years lost and of relationships among the components of the microbiota death (2.2 million). However, in many instances there is anessential requirement for the administration of antibiotics,and hence it is necessary to identify means of minimising the ∗ Corresponding author. Tel.: +44 1639 825100; fax: +44 1792 472466.
adverse effects of antibiotics whilst maximising their poten- E-mail address: [email protected] (S.F. Plummer).
tial benefits. One method is to select for antimicrobial agents 0924-8579/$ – see front matter 2005 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
doi:10.1016/j.ijantimicag.2005.04.004 S.F. Plummer et al. / International Journal of Antimicrobial Agents 26 (2005) 69–74 that do not disturb the microbial colonisation resistance, but lansoprazole (30 mg bd) for 7 days. For penicillin allergy, 400 mg metronidazole three times a day was substituted.
Beneficial effects have been observed when probiotics The probiotic product (Cultech Ltd., Port Talbot, UK) com- have been used for the prevention and treatment of gastroin- prised two strains of Lactobacillus acidophilus (CUL60 and testinal disturbance Trials have shown the potential CUL21) and two strains of Bifidobacterium spp. at a total for the use of probiotics in the treatment of rotavirus infec- of 2.5 × 1010 colony-forming units (CFU)/capsule, and the tions, antibiotic-associated diarrhoea, traveller’s diarrhoea, placebo comprised an inactive carrier (maltodextrin). Patients infantile diarrhoea, relapsing Clostridium difficile colitis, received one capsule daily. The probiotic strains used were inflammatory bowel disease, irritable bowel syndrome, sensitive to test antibiotics using the disk diffusion assay atopy in at-risk infants and chronic sinusitis according to National Committee for Clinical Laboratory For the purposes of this study, a cohort of Helicobacter pylori-infected patients receiving the triple therapy antibiotictreatment regimen was selected for investigation.
The aim was to determine the effects of probiotic sup- plementation during triple therapy on the composition of the Of the 162 patients recruited, 7 were excluded for failing intestinal re-growth population, looking both at numbers and to provide the samples. The 155 remaining patients were ran- types of microorganisms and on the incidence of antibiotic domly divided between the placebo group (79 patients) and resistance in the intestinal microbiota.
2.5. General microbiological screen 2. Materials and methods
Traditional microbiological methods were used to analyse the samples. On the basis of pilot screening studies (unpub-lished data), selective media and Gram staining of colony One hundred and sixty-two patients infected with H. types were used to enable enumeration and differentiation pylori were enrolled into a study at Addenbrooke’s Hospital, of the faecal microbiotas. An anaerobic dilution series of Cambridge, UK. The H. pylori infection was verified the faecal samples was set up in pre-reduced Maximum by positive serology and histology by the Public Health Recovery Diluent (MRD; Oxoid Ltd., Basingstoke, UK). A Laboratory Service at Addenbrooke’s Hospital. Patients modification of the Miles and Misra plate count technique provided written consent and had no other gastrointestinal as used to plate 10 × 10 ␮L of appropriate dilutions disorders apart from peptic ulcers thought to be related to onto the pre-reduced selective agars (all agars were obtained their H. pylori infection. None had received any antibiotics from Oxoid Ltd. unless otherwise stated): anaerobic blood or been subject to any dietary intervention in 6 weeks prior (total anaerobes; bioMerieux, Basingstoke, UK); blood agar to the study. Ethical approval was obtained from Cambridge (total facultative anaerobes; bioMerieux); Wilkins–Chalgren agar (Bacteroides spp.); MacConkey No. 3 agar (MAC; enter-obacteriaceae); Kanamycin Aesculin Azide agar (KAA; ente- rococci); Baird Parker agar (staphylococci); de Mann RogosaSharpe agar (MRS; Lactobacillus spp.); modified MRS The trial was a double-blind placebo-controlled study agar (0.3% (w/v) sodium propionate, 0.2% (w/v) lithium with all patients receiving antibiotics from days 1 to 7.
chloride, 0.05% (w/v) cysteine hydrochloride and 5% (v/v) One group of patients received the probiotic product (active defibrinated sheep blood included; Bifidobacterium spp.); group) from days 1 to 21 and the second group received the Rose Bengal Agar (yeasts); and ID2 Agar (Candida albicans; placebo product (placebo group) from days 1 to 21. Two consecutive faecal samples were tested prior to antibiotic Anaerobic plates were incubated at 37 ◦C for 72 h and therapy and statistical analysis indicated that the data could aerobic plates were incubated at 37 ◦C for 48 h. Organisms be pooled to provide day 1 results. A further sample was were identified by anaerobic/aerobic growth colony, Gram collected on day 7. Two consecutive faecal samples were stain and API biochemical identification strips (bioMerieux).
obtained 4 weeks after completion of antibiotic therapy and The results were expressed as the CFU per gram of dry weight these were pooled to provide day 35 results. The faecal sam- ples were sealed in anaerobic bags and stored at −70 ◦C untiltested.
2.6. Antibiotic resistance analysis The effects of the antibiotic therapy on the numbers of antibiotic-resistant enterococci and enterobacteriaceae in The patients received standard eradication therapy: amox- the faecal microbiotas pre and post antibiotic treatment were icillin (1 g twice a day (bd)), clarithromycin (500 mg bd) and chosen for assessment in this study. KAA agar or MAC S.F. Plummer et al. / International Journal of Antimicrobial Agents 26 (2005) 69–74 agar containing a range of concentrations of amoxicillin clinical significance (P < 0.05), and increased post therapy or clarithromycin (0, 0.015, 0.06, 0.5, 1.0, 4.0, 8.0, 16.0, (). There were no significant differences 32.0, 128.0 or 512.0 ␮g/mL) were used for enumeration of between the numbers at days 1 and 35.
enterococci and enterobacteriaceae, respectively.
The samples were plated out using the modified Miles 3.1. Changes in the numbers of enterobacteria and and Misra technique × 10 ␮L drops) and plates were incubated aerobically at 37 ◦C for 48 h (KAA agar) or 24 h(MAC agar). The breakpoints for enterococci/amoxicillin In the placebo group, the numbers of facultative anaer- at a minimum inhibitory concentration (MIC) >8 ␮g/mL, obes increased significantly between days 7 and 35, and the for enterobacteriaceae/amoxicillin at a MIC >32 ␮g/mL, numbers of enterobacteria were significantly higher at day 35 for enterococci/clarithromycin at a MIC >1 ␮g/mL and for than at day 1 (P < 0.05). No significant changes occurred in enterobacteriaceae/clarithromycin at a MIC >8 ␮g/mL rep- the numbers of enterococci and staphylococci in this group, although the numbers of enterococci decreased during antibi-otic therapy ( In the probiotic-supplemented group, the enterobacterial population decreased during therapy but then increased so Statistics were performed using the SPSS v11.5 program that at day 35 the numbers were not significantly different (SPSS, Chicago, IL, USA). Within the same treatment group, two related samples from days 1/2 and days 35/36 were com-pared using the Wilcoxon signed-rank test to ensure that 3.2. Effects on the bacteroides population pooling of the replicates was feasible. No significant dif-ferences were detected between these two sets of replicates The numbers of bacteroides in the placebo group (except staphylococci at days 1/2; The Wilcoxon decreased significantly between days 1 and 7 (P < 0.05), fol- signed-rank test was also used to compare related samples in lowed by a significant increase during the re-growth period each microbial population between days 1 and 7, days 7 and so that at day 35 the numbers were not significantly different 35, and days 1 and 35. The non-parametric Mann–Whitney U- from day 1. In contrast, in the active group the numbers of test was used to compare the unrelated median values (active bacteroides increased from days 7 to 35 (P < 0.01) so that the and placebo) for each microbial population. A P-value of less final numbers at day 35 were significantly higher than at day than 0.05 was considered statistically significant. The McNe- mar test was used to compare antibiotic resistance betweenany two time points (days 1, 7 and 35) and P ≤0.05 indicatesthat the proportions are not equal.
3.3. Changes in lactobacilli and bifidobacteria The bifidobacterial population in both groups decreased 3. Results
in response to antibiotic therapy (P < 0.0001) and, despiteincreasing significantly between days 7 and 35, the numbers In the placebo and active groups, the total bacterial for both groups at day 35 were significantly lower than those numbers decreased during antibiotic therapy, with a small Table 1Distribution of the intestinal microbiota in patients in the placebo 4.7 (<1.7–10.3), 4.0 (<1.7–10.0) Statistical analysis using SPSS v11.5 statistical package: the Wilcoxon signed-rank test comparison between sample collection days: *P ≤ 0.05, day 1 comparedwith day 7; **P ≤ 0.05, day 7 compared with day 35; †P ≤ 0.05, day 1 compared with day 35.
a Data given as median (minimum–maximum) log10 colony-forming units (CFU)/g dry weight of faeces.
b Number of patients harbouring microbial population.
c There were significant differences between the medians of two samples before therapy (day 1), therefore the results from these samples could not be merged S.F. Plummer et al. / International Journal of Antimicrobial Agents 26 (2005) 69–74 Table 2Distribution of the intestinal microbiota in patients in the active Statistical analysis using SPSS v11.5 statistical package: the Wilcoxon signed-rank test comparison between sample collection days: *P ≤ 0.05, day 1 comparedwith day 7; **P ≤ 0.05, day 7 compared with day 35; †P ≤ 0.05, day 1 compared with day 35.
a Data given as median (minimum–maximum) log10 colony-forming units (CFU)/g dry weight of faeces.
b Number of patients harbouring microbial population.
The lactobacillus population of the placebo group patients, which made it very difficult to make any assessment decreased significantly between days 1 and 7, but then of changes in the antibiotic resistance profiles. There was no increased (P < 0.01) so that at day 35 the numbers were decrease in resistance in response to probiotic supplement comparable with those at day 1 (The numbers of action, but the indigenous resistance levels were too high lactobacilli in the probiotic group decreased during antibi- to determine whether the probiotics had registered any otic therapy and increased again post treatment, but none of these changes was statistically significant The development of resistance to amoxicillin and clar- ithromycin between days 1 and 35 among the enterococcal 3.4. The effects of antibiotics on the yeast component of population of patients in the two groups is shown in vely. At day 35 in the placebo group,the number of patients expressing antibiotic resistance Although the numbers of yeast increased in both groups within the enterococcal population was significantly higher during antibiotic therapy, in the placebo group this was asso- (P ≤ 0.05) than the number in the initial population at all ciated with a significant increase in the number of C. albicans; antibiotic concentrations up to 32.0 ␮g/mL At the a similar increase did not occur among the patients receiving highest concentrations of amoxicillin (128 and 512 ␮g/mL), the probiotic supplement. At day 35, the numbers of C. albi- comparison of days 1 and 35 showed no significant cans in both groups were significantly higher than at day 1 differences in the levels of antibiotic resistance in the However, for the probiotic-supplemented group at all 3.5. Comparison of the components of the microbiotas amoxicillin concentrations, there was no significant increase in the number of patients carrying antibiotic-resistant ente-rococci between days 1 and 35.
When the microbiotas of the two groups were compared the numbers of total facultative anaerobes at day 35 in the active group were significantly lower than in the Comparison of microbial populations among the placebo and active groups placebo group (U = 1648; P = 0.031). Similarly, the numbers of enterobacteriaceae in the active group were significantly lower than in the placebo group (U = 1608; P = 0.014). The number of C. albicans after antibiotic therapy in the placebo group was significantly higher than in the active group (U = 1891; P = 0.049), but by day 35 the numbers of yeasts were comparable in both groups. There were no significant differences between the two treatment groups for any of the a Data given as median log10 colony-forming units (CFU)/g dry weight of A very high level of indigenous antibiotic resistance was found among the enterobacteriaceae in this cohort of b According to Mann–Whitney U-test.
S.F. Plummer et al. / International Journal of Antimicrobial Agents 26 (2005) 69–74 Madden et al. a significant increase in the fac- Number of patients developing amoxicillin resistance within the faecal ente- ultative anaerobe component of the microbiota between days rococcal population between days 1 and 35 1 and 27 in placebo group with amoxicillin, metronidazole and lansoprazole treatment. When probiotics were given after antibiotics, numbers decreased significantly between days 7 The eradication therapy did not significantly disrupt the total anaerobe population from days 1 to 35 which contrasts with the results of other studies where found that the total anaerobic microbiota was strongly sup- pressed in H. pylori patients (amoxicillin and metronidazole combination (OAM) group or clarithromycin and metronida- b The breakpoint for enterococci:amoxicillin at a minimum inhibitory con- zole (OCM) group), although the effect was most pronounced centration (MIC) >8 ␮g/mL represented antibiotic resistance.
in the OCM group. Amoxicillin as a single agent causes onlyminor disturbances, but in some studies the anaerobic micro- Table 5Number of patients developing clarithromycin resistance within the faecal biota has been found to be disrupted due to metronidazole enterococcal population between days 1 and 35 It is also interesting that despite the sensitivity of the probiotic organisms to antibiotics, no significant changes were observed for the total Lactobacillus numbers in the probiotic-supplemented group—an observation not recorded for the placebo group. However, the antibiotic sensitivity of the bifidobacteria was apparent in both groups, as also observed by Adamsson et al. Buhling et al. Although there was no significant change in total num- bers of yeast between days 1 and 35 in the placebo group, the number of C. albicans increased significantly (P < 0.01).
This finding contrasts with the study of Buhling et al. The breakpoint for enterococci:clarithromycin at a minimum inhibitory concentration (MIC) >1 ␮g/mL represented antibiotic resistance.
who found that the numbers both of yeast and C. albicansin patients with H. pylori returned back to the starting levels With clarithromycin, in the placebo group there was a significant development of resistance (P ≤ 0.001) at concen- The very high levels of antibiotic resistance among the trations near to the resistance breakpoint, which was not enterobacteriaceae in this cohort of patients made any assess- seen in the probiotic group. However, significant resistance ment of changes (increases) in resistance post therapy very (P = 0.001) developed in both groups to the same extent at difficult, but the extent of antibiotic resistance might have been related to the significantly lower numbers of enter-obacteria seen in the active group patients compared withthe placebo group at day 35. Working with a similar cohort, 4. Discussion
Stark et al. ed overgrowth by amoxicillin-resistantenterobacteria post antibiotic therapy.
Administration of antibiotics often causes disturbances Antibiotic resistance among the enterococci was signifi- in the normal intestinal microbiota In the present cantly higher in the placebo group than in the probiotic group study, the total bacterial and total facultative anaerobe pop- post therapy in this study, suggesting that the probiotics had ulation results indicate that despite the probiotic supplement in some way modulated the composition of the re-growth the microbiotas of both the placebo and active groups were population. It is known that bacteria have an energy require- susceptible to the effects of the antibiotics administered to ment to achieve antibiotic resistance either owing eradicate H. pylori. It appeared that there was recovery of to chromosomal alterations (e.g. target site alterations) or the majority of the components of the microbiota post antibi- owing to the use of accessory elements (such as enzymes otic therapy, with no significant difference between days 1 and antibiotic efflux pumps). Such energy requirements and 35. However, the noticeable difference occurred with the could affect the growth kinetics of the bacteria, but the enterobacterial component of the placebo group, which was antibiotic resistance provides a competitive advantage over subject to disturbance, suggesting that supplementation with the antibiotic-sensitive strains, enabling their survival. The probiotics had impacted on the intestinal microbiota, result- energy costs involved in the mechanisms of resistance for ing in less disruption of the compositional balance for the the bacteria in this study are unknown, but it is possible that the additional challenge to these ‘energy-depleted’ bacteria S.F. Plummer et al. / International Journal of Antimicrobial Agents 26 (2005) 69–74 caused by the daily supplement of probiotic bacteria could [10] Cremonini F, Di Caro S, Nista EC, et al. Meta-analysis: the effect be too great to enable their domination and hence this could of probiotic administration on antibiotic-associated diarrhea. Aliment account for the lower incidence of antibiotic resistance [11] Ishikawa H, Akedo I, Umesaki Y, Tanaka R, Imaoka A, Otani T.
Randomized controlled trial of the effect of bifidobacteria-fermented From this study, it would appear that daily supplementa- milk on ulcerative colitis. J Am Coll Nutr 2002;22:56–63.
tion with viable probiotic bacteria during and post antibiotic [12] Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, therapy reduces the extent of disruption to the intestinal Isolauri E. Probiotics in primary prevention of atopic disease: a ran- microbiota as well as the incidence and total numbers of domized placebo-controlled trial. Lancet 2001;357:1076–9.
[13] Hilton E, Kolalowoski P, Singer C, Smith M. Efficacy of Lacto- antibiotic-resistant strains in the re-growth population.
bacillus GG as a diarrhoeal preventive in travellers. J Travel Med1997;4:41–3.
[14] Reid G, Jass J, Sebulsky MT, McCormick JK. Potential uses of Acknowledgments
probiotics in clinical practice. Clin Microbiol Rev 2003;16:658–72.
[15] Marteau P, Lepage P, Mangin I, et al. Gut flora and inflammatory We would like to thank Dr Martin Day of Cardiff Uni- bowel disease. Aliment Pharmacol Ther 2004;20:18–23.
[16] National Committee for Clinical Laboratory Standards. Performance versity, UK, and Dr Asa Sullivan of Karolinska University standards for antimicrobial disk susceptibility tests: approved stan- Hospital, Huddinge, Sweden, for their expert advice and dard, 8th ed. Wayne, PA: NCCLS; 2000.
knowledge of antibiotic resistance. We would like to express [17] Miles AA, Misra SS. The estimation of the bactericidal power of our sincere thanks posthumously to Prof. Denver Russell of the blood. J Hygeine 1938;38:732–49.
Cardiff University, UK, for his support and guidance until his [18] Adamsson I, Nord CE, Lundquist P, Sj¨ostedt S, Edlund C. Compar- ative effects of omeprazole, amoxycillin plus metronidazole versus omeprazole, clarithromycin plus metronidazole on the oral, gastricand intestinal microflora in Helicobacter pylori-infected individuals.
J Antimicrob Chemother 1999;44:629–40.
References
[19] van der Waaij D, Nord CE. Development and persistence of multi-resistance to antibiotics in bacteria; an analysis and a new [1] Sullivan A, Edlund C, Nord CE. Effect of antimicrobial agents approach to this urgent problem. Int J Antimicrob Agents 2000;16: on the ecological balance of human microflora. Lancet Infect Dis [20] Lindberg E, Adlerberth I, Wold AE. Antibiotic resistance in Staphy- [2] Levy J. The effects of antibiotic use on gastrointestinal function. Am lococcus aureus colonising the intestines of Swedish infants. Clin J Gastroenterol 2000;95(Suppl. 1):S8-S10.
[3] Finegold SM. Interaction of antimicrobial therapy and intestinal flora.
[21] Madden JAJ, Plummer S, Plummer NT, Herbison M, Hunter JO. A double-blind, placebo controlled pilot study to assess the effects of a [4] Bohnhoff M, Miller CP, Martin WR. Resistance of the mouse’s probiotic on the response of the intestinal microflora to Helicobacter intestinal tract to experimental salmonella infection. I. Factors which pylori eradication therapy. Gut 2002;50:A22.
interfere with the initiation of infection by oral inoculation. J Exp [22] Stark CA, Adamsson I, Edlund C, et al. Effects of omeprazole and amoxycillin on the human oral and gastrointestinal microflora in [5] Reid G, Friendship R. Alternatives to antibiotic use: probiotics for patients with Helicobacter pylori infection. J Antimicrob Chemother the gut. Anim Biotechnol 2002;13:97–112.
[6] Sullivan A, Barkholt L, Nord CE. Lactobacillus acidophilus, [23] Adamsson I, Edlund C, Nord CE. Microbial ecology and treat- Bifidobacterium lactis and Lactobacillus F19 prevent antibiotic- ment of Helicobacter pylori infections. J Chemother 2000;12:5– associated ecological disturbances of Bacteroides fragilis in the intestine. J Antimicrob Chemother 2003;52:308–11.
[24] Edlund C, Nord CE. Ecological impact of antimicrobial agents on [7] Fern´andez MF, Boris S, Barb´es C. Probiotic properties of human human intestinal microflora. Alpe Adria Microbiol J 1993;2:137– lactobacilli strains to be used in the gastrointestinal tract. J Appl [25] Buhling A, Radun D, Muller WA, et al. Influence of anti- [8] Friˇc P. Probiotics in gastroenterology. Z Gastroenterol 2002;40: Helicobacter triple therapy with metronidazole, omeprazole and clarithromycin on intestinal microflora. Aliment Pharmacol Ther [9] Hamilton-Miller JMT. The role of probiotics in the treatment and prevention of Helicobacter pylori infection. Int J Antimicrob Agents [26] Andersson DI, Levin BR. The biological cost of antibiotic resistance.
Curr Opin Microbiol 1999;2:489–93.

Source: http://www.provenprobiotics.co.uk/creo_files/upload/default/effects_of_probiotics_on_the_composition_of_the_intestinal_microbiota_following_antibiotic_therapy_2005.pdf