Journal of Antimicrobial Chemotherapy (2004) 54, 791 797 DOI: 10.1093/jac/dkh406 Advance Access publication 25 August 2004 Effect of Lactobacillus F19 on the emergence of antibioticresistant microorganisms in the intestinal microflora Åsa Sullivan 1, Anita Johansson 2, Bo Svenungsson 2,3 and Carl Erik Nord 1 * JAC Divisions of 1 Clinical Bacteriology, F82, and 2 Infectious Diseases, Karolinska University Hospital Huddinge, SE-141 86 Stockholm, Sweden; 3 Division of Infectious Diseases, Karolinska University Hospital Solna, Karolinska Institutet, Stockholm, Sweden Received 16 March 2004; returned 13 June 2004; revised 30 June 2004; accepted 20 July 2004 Objectives: Probiotic lactic-acid-producing bacteria have been used for prevention of gastrointestinal diseases. The aim of the present study was to examine whether Lactobacillus F19 in conjunction with treatment with penicillin, ciprofloxacin or norfloxacin prevents establishment of resistant bacteria in the gastrointestinal tract. Methods: Twenty patients admitted to hospital due for treatment with penicillin and 16 patients due for treatment with ciprofloxacin or norfloxacin were included in the study. In either group, the patients were randomized into two groups, receiving placebo or an active probiotic product. Faecal samples were collected before treatment, on day 10 and 1 month after the start of the treatment. Isolates of enterococci, enterobacteria and Bacteroides fragilis species were screened for resistance to penicillin and ciprofloxacin, respectively. Results: Administration of penicillin did not influence resistance in enterococci while quinolone resistance increased during quinolone treatment. Susceptibility to ampicillin and piperacillin/tazobactam decreased in enterobacteria during penicillin treatment and ciprofloxacin resistance increased in the quinolone group. Penicillin and quinolones did not influence the resistance rates of Bacteroides isolates. No major differences were observed between the probiotic- and placebo-supplemented groups. Conclusions: There was a limited effect of Lactobacillus F19 on the emergence of resistant isolates during treatment with penicillin and quinolones. Keywords: probiotics, resistance, antimicrobial agents Introduction An increasing problem in clinical medicine is the development of bacterial strains resistant to antibacterial agents. Today it is generally accepted that the extensive use of antibiotics is the main reason for the development of the increased bacterial resistance. 1 The human gastrointestinal tract is the natural habitat of microorganisms usually encountered in infections resistant to antimicrobials. The majority of resistant strains are found in patients in hospitals, mainly from patients in intensive care units. 2 Among the aerobic microorganisms, enterococci are regarded as one of the most common causes of nosocomial infection. The majority of enterococcal infections are caused by Enterococcus faecalis followed by Enterococcus faecium. 3 Enterococci are intrinsically resistant to several antibacterial agents, i.e. low-level resistance to cell-wall-active agents and aminoglycosides. 4 However, in the Nordic countries the level of resistance in enterococci is still low. 5 Other aerobic microorganisms frequently encountered among clinical isolates are members of the family Enterobacteriaceae. 6 It has been shown that long-term hospitalized elderly patients are at increased risk of being infected with Gram-negative bacilli resistant to ampicillin. 7 A problem with regard to Enterobacteriaceae, since the introduction of the third-generation cephalosporins, is the increasing number of strains expressing extended-spectrum b-lactamases (ESBLs). Almost 17% of Klebsiella pneumoniae strains in a European multi-centre study were found to be potential carriers of ESBL enzymes. 6 Anaerobic bacteria outnumber aerobic bacteria in the gastrointestinal tract, and among the anaerobic microorganisms, Bacteroides spp. are the most frequent pathogenic bacteria isolated. 8... *Corresponding author. Tel: +46-8-58587838; Fax: +46-8-7113918; E-mail: carl.erik.nord@labmed.ki.se... 791 JAC vol.54 no.4 q The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.
Å. Sullivan et al. Clinically important species are the Bacteroides fragilis group. The majority of the B. fragilis group is resistant to penicillins and cephalosporins. Resistance to clindamycin is increasing while resistance to carbapenems and nitroimidazoles is still rare. 8 The gastrointestinal microbial ecosystem is relatively stable but quantitative and qualitative disturbances are seen after oral administration of antibiotics. 9 The normal flora limits the concentration of potentially pathogenic microorganisms, which can reach high numbers in connection with intake of antimicrobial agents. 10 The use of probiotics for prevention of gastrointestinal diseases is well established. 11 However, no studies have been performed on the role of probiotics in the prevention of emergence of resistant isolates. By competition for receptors and nutrition and by production of lactic acid and antimicrobial substances, probiotic strains could be a means by which colonization with resistant strains could be prevented. The aim of the present study was to examine whether administration of Lactobacillus F19 in conjunction with treatment with penicillin or quinolones (ciprofloxacin or norfloxacin) prevents the emergence of resistant isolates of enterococci, enterobacteria and B. fragilis spp. in the intestinal tract. Material and methods Patients Forty-one patients admitted to the Division of Infectious Diseases, Karolinska University Hospital Huddinge, Stockholm, Sweden receiving treatment with penicillin G followed by penicillin V, and 23 patients receiving treatment with ciprofloxacin or norfloxacin were included in the study. None of the patients had any history of gastrointestinal, liver or kidney diseases, and had not been treated with antimicrobial agents within 3 months preceding the study. The two treatment groups were randomized into one placebo and one active group regarding the probiotic supplement in a double-blind fashion. The study was approved by the local ethics committee of Karolinska University Hospital, Karolinska Institutet, Stockholm, Sweden. Administration of antimicrobial agents Patients in the penicillin group received penicillin G (Bensylpenicillin; Astra Läkemedel, Södertälje, Sweden) 1 or 3 g three times daily followed by penicillin V (Kåvepenin; Astra Läkemedel) 1 g three times daily and patients in the quinolone group received either ciprofloxacin (Ciproxin; Bayer AB, Göteborg, Sweden) 500 or 750 mg twice per day or norfloxacin (Lexinor; Astra Läkemedel) 400 mg twice daily. Compliance Signed diary cards that were returned at the end of the study checked compliance in respect of the probiotic product. Assay of the probiotic microorganism in faecal samples was a further verification of compliance. Sampling procedures Stool specimens were collected before the administration of antimicrobial agents and probiotic/placebo products (day 1), on day 10 and 30 days after the start of administration. The samples were collected in sterile plastic containers, which were frozen at 708C until processed. Microbiological procedures Stool specimens were suspended in pre-reduced peptone yeast extract medium, diluted to 10 7 and inoculated on the following selective media: Enterococcosel agar (BBL, Cockeysville, MD, USA) for detection of enterococci, CLED agar (Merck, Darmstadt, Germany) for detection of Enterobacteriaceae, BKV (kanamycin vancomycin blood) agar for detection of B. fragilis group, Rogosa agar (Difco, Detroit, MI, USA) for cultivation of lactobacilli and TCCFA agar (Pepton from Casein/Proteose PeptonIII 40 mg/ml, sodium hydrogen phosphate 5 mg/ml, potassium dihydrogen phosphate 1 mg/l, sodium chloride 2 mg/ml, magnesium sulphate 0.2 mg/ml, Bacto agar/agar agar 20 mg/l, taurocholic acid 1 mg/ml, neutral red 0.03 mg/ml, fructose 15%, Clostridium difficile-supplemented D-cycloserine, cefoxitin) for cultivation of C. difficile. Blood agar (LabM Kemila, Bury, UK) was used for total counts of aerobic and anaerobic bacteria. ASM agar (PDM; Biodisk, Solna, Sweden) supplemented with 5% blood and 8 mg/l penicillin G for aerobic microorganisms and 2 mg/l penicillin G for anaerobic microorganisms or ciprofloxacin 4 mg/l (used for aerobic and anaerobic bacteria) was used for assessing the total number of penicillin- or quinolone-resistant microorganisms, respectively. The aerobic agar plates were incubated for 24 h at 378C and the anaerobic plates for 48 h at 378C in anaerobic jars (GasPak; BBL). After incubation different colony types were counted, isolated in pure cultures and identified to genus level. All isolates were analysed according to Gram s stain and colony morphology, followed by biochemical tests. The API 20E kit (biomérieux SA, Marcy l Étoile, France) was used for identification of Enterobacteriaceae. Anaerobic microorganisms were identified to genus level by gas liquid chromatography of metabolites from glucose. The lower limit of detection was 10 2 microorganisms per gram faeces. Isolated strains of C. difficile were tested for cytotoxin production on McCoy cells with an antitoxin kit for detection of toxin B (Tech- Lab, Blacksburg, VA, USA). C. difficile CCUG 19126 was used as the reference strain. Administration of probiotic and placebo products The probiotic product consisted of powdered milk (10 g) and 10 10 colony forming units (cfu) of freeze-dried Lactobacillus paracasei ssp. paracasei F19 (Arla Foods, Stockholm, Sweden). A similar product was given to patients in the placebo groups but with no added microorganisms. The products were prepared in connection with the start of the study and were stored at 708C. After inclusion of a patient, the product was stored at 88C. The products were dissolved in 100 ml water immediately before use. The products were administered twice daily for 14 days. The cell viability was checked after storage under similar conditions. Antibiotic susceptibility tests Up to five colonies of enterococci, Enterobacteriaceae and B. fragilis group spp. were isolated from each patient on each sampling occasion. Depending on which antimicrobial agent had been used as treatment, the MIC of penicillin or ciprofloxacin was determined by the agar dilution method, according to the NCCLS M7-A6 document for aerobic bacteria 12 and the NCCLS M11-A5 document for anaerobic bacteria. 13 Species intrinsically resistant and those found to be resistant on susceptibility testing were further tested for susceptibility to several different antimicrobial agents, i.e. isolates of enterococci were tested against ampicillin, gentamicin, linezolid 792
Probiotics and resistance and vancomycin, Enterobacteriaceae were tested against ampicillin, piperacillin/tazobactam, cefotaxime and gentamicin, and B. fragilis group isolates were tested against ampicillin/clavulanic acid, cefoperazone, cefoxitin, imipenem, clindamycin and metronidazole. The control strains used were E. faecalis ATCC 29212, Escherichia coli ATCC 25922, B. fragilis ATCC 25285 and Bacteroides thetaiotaomicron ATCC 29741. The final inoculum for aerobic bacteria was 1 10 4 cfu/spot and for anaerobic bacteria 1 10 5 cfu/ spot. Inoculated plates were incubated for 18 h (aerobic bacteria) and 48 h (anaerobic bacteria) at 378C. The breakpoint values used were according to NCCLS recommendations. Detection of ESBLs Isolates of Enterobacteriaceae intermediately resistant or resistant to cefotaxime were tested for production of ESBLs using the Etest (AB Biodisk). K. pneumoniae ATCC 700603 and E. coli ATCC 35218 were used as positive and negative controls, respectively. Identification of Lactobacillus F19 strains Colony forming units exhibiting colony and Gram s stain morphology similar to the probiotic strain Lactobacillus F19 were analysed by randomly amplified polymorphic DNA-PCR (RAPD-PCR) to verify the identification. The primer Lbc-19 (5 0 -AGTAGCCAC- 3 0 ) was used in screening and OPA-02 (5 0 -TGCCGAGCTC-3 0 ) for identification of Lactobacillus F19. 14 Statistical analysis The quantitative alterations were compared within groups between pretreatment (day 1) and day 10 and between pretreatment and after treatment (day 30) and were analysed by Wilcoxon s signed rank test for paired samples. The MIC values for each species were compared within groups between pretreatment and day 10 and between pretreatment and day 30 by the Mann Whitney U test in order to detect significant decreases in susceptibility during and after the administration period. Differences in MIC value between the groups were likewise analysed by the Mann Whitney U test. P values <0.05 were considered statistically significant and were adjusted for the multiple analysis. Results Patients: penicillin group Of 41 randomized patients, 20 (49%) completed the study. Seventeen patients were treated for pneumonia and three for streptococcal infections (tonsillitis and erysipelas). The mean age was 59 years (range 18 89 years), the mean length of hospital stay was 5 days (range 2 11 days), the mean amount of penicillin G was 11 g (range 3 30 g) and the mean total length of antimicrobial treatment (penicillin G followed by penicillin V) was 14 days (range 11 17 days). There were no statistically significant differences between the patients regarding the above parameters in the active and placebo groups. Twenty-one patients did not complete the study, of whom 13 received new antimicrobial agents, three patients were nauseated by the ingestion of the probiotic/placebo product, four patients felt too ill to be able to complete the study, and in, one case, faecal samples were not received. Patients: quinolone group Of 23 randomized patients, 16 (70%) completed the study. Fourteen patients had urinary tract infections and two patients had fever of unknown origin and cholecystitis. Nine patients were treated with ciprofloxacin 500 mg twice daily and one patient received 750 mg twice daily for 3 days followed by norfloxacin 400 mg twice daily for 14 days and 200 mg twice daily for 14 days. The remaining six patients were treated with norfloxacin 400 mg twice daily. The mean age was 61 years (range 28 86 years), the mean duration of hospital stay was 4 days (range 2 8 days), and the total length of antimicrobial treatment was 12 days (range 7 31 days) (15 patients <_15 days). There were no statistically significant differences between the patients concerning these parameters in the active and placebo groups. Seven patients did not complete the study. In three cases there was a change of medication or the medication was withdrawn, one patient felt nauseous when drinking the supplement and three subjects failed to produce faecal samples. Compliance Four patients in the penicillin group receiving the active probiotic product returned completed diary cards. In six cases, information was lacking concerning the intake of the probiotic product. In the placebo group, there were six completed diary cards and four incomplete. In the quinolone-treated patients in the active and placebo groups, there were six and seven completed cards and two and one uncompleted cards, respectively. Impact of antimicrobial agents and Lactobacillus F19 on numbers of microorganisms The numbers of enterococci, enterobacteria and Bacteroides species in patients receiving penicillin showed no major changes apart from increased numbers of enterococci in the active group (increase from day 1 to day 30 P < 0.05). The total numbers of penicillin-resistant microorganisms on the penicillin-containing non-selective agar remained stable during the study period. In the patients treated with quinolones, there was a statistically significant increase in the numbers of resistant isolates on the ciprofloxacin-containing non-selective agar (P < 0.05) between day 1 and 10 and numbers of Enterobacteriaceae (P < 0.05) between day 1 and day 30 in the microflora of patients in the placebo group. In the placebo group, the numbers of B. fragilis spp. (P < 0.05), the total number of anaerobic microorganisms (P < 0.001) as well as the total numbers of resistant strains (P < 0.05) increased from day 1 to day 30. No statistically significant changes were seen in the intestinal microflora in the active group. C. difficile was isolated in four samples of four patients treated with quinolones, in one patient on day 10 and in three patients on day 30. Two patients belonged to the placebo and two to the active group. None of these isolates produced toxins. Prevalence of Lactobacillus F19 Lactobacillus F19 was recovered in three samples from 10 patients in the active group of penicillin-treated individuals (log 10 cfu range 3.3 4.7) on day 10. In the active group of quinolone-treated patients, Lactobacillus F19 was isolated in three samples of eight patients (log 10 cfu 3.3 5.2) on day 10, and one 793
Å. Sullivan et al. patient still harboured the probiotic strain on day 30 (log 10 cfu 2.3). Five of the six patients positive for Lactobacillus F19 returned their diary cards completed. In vitro activity of antimicrobial agents against enterococci One isolate from one patient in the active penicillin group was resistant to penicillin, ampicillin and ciprofloxacin on day 10, while the same isolate was sensitive to vancomycin, linezolid and gentamicin. The numbers of ciprofloxacin-resistant enterococci in quinolone-treated patients increased in the placebo group from three of 32 (9%) isolates on day 1 to 21 of 25 (84%) isolates on day 10. Thirty days after the start of treatment, six of 46 (13%) isolates were resistant to ciprofloxacin. In the active group, the corresponding figures were seven of 26 (27%) isolates on day 1, 12 of 20 (60%) isolates on day 10 and 17 of 39 (44%) isolates on day 30. Isolates resistant to ampicillin were detected in one patient in the placebo group and in one patient in the active group in samples from days 10 and 30. Gentamicin-resistant isolates were also detected in all samples from one patient in the active group. All isolates were sensitive to vancomycin and linezolid. In vitro activity of antimicrobial agents against Enterobacteriaceae Before administration of antimicrobial agents, isolates of enterobacteria with decreased susceptibility to ampicillin and piperacillin/tazobactam were detected in patients treated with penicillin (Table 1). The MIC values increased for isolates from sample 2 and decreased again on day 30 in both the placebo and active groups. There was a statistically significant difference between the groups concerning the susceptibility to piperacillin/tazobactam also existing before the start of the study. Exclusion of patients with piperacillin/tazobactam-resistant isolates before treatment revealed the emergence of resistant or intermediately resistant isolates in four patients in the placebo group and one in the active group. In 23 of 135 (17%) isolates in the placebo group, reduced susceptibility to both ampicillin and piperacillin/ tazobactam was observed. The corresponding figure for the active group was nine of 130 (7%) isolates. In both the active and placebo groups, all isolates were susceptible to ciprofloxacin. In the placebo group, four isolates from two patients with decreased susceptibility to cefotaxime were detected on days 10 and 30. Two isolates were intermediately resistant to gentamicin on day 10. Isolates of enterobacteria from patients treated with quinolones were all susceptible to ampicillin and piperacillin/tazobactam. Isolates from one patient in the placebo group and from three patients in the active group were resistant to ciprofloxacin before treatment. All resistant isolates on day 10 were found in samples from these patients. On day 30, resistant isolates were isolated from one additional patient in the placebo group and from two in the active group. Exclusion of patients with resistant isolates before treatment rendered the material too small for statistical analysis. Isolates with reduced susceptibility to cefotaxime and gentamicin were detected in all samples from one patient in the placebo group and from two patients in the active group. Another patient in the active group harboured gentamicin-resistant isolates on day 30. ESBLs One patient treated with penicillin was found to harbour ESBLpositive strains of Klebsiella oxytoca on day 10. The patient belonged to the placebo group. Amongst the quinolone-treated patients, ESBL-positive strains of E. coli was identified in two subjects, one from the placebo and one from the active group. In both cases the strains were found before treatment and on day 10 and in one patient also on day 30. In vitro activity of antimicrobial agents against B. fragilis spp In the penicillin-treated group, isolates of B. fragilis spp. had reduced susceptibility to ampicillin/clavulanic acid, cefoxitin, cefoperazone and clindamycin before, during and after the administration of penicillin (Table 2). There were no Table 1. In vitro activity of antimicrobial agents against isolates of Enterobacteriaceae in 20 patients treated with penicillin and placebo or active probiotic supplement Antimicrobial agent Day of sampling n MIC 50 MIC 90 MIC range S a (%) R a (%) Placebo group ampicillin 1 50 4.0 >128.0 <2.0 >128.0 56 42 10 40 >128.0 >128.0 <2.0 >128.0 33 67 30 45 4.0 >128.0 <2.0 >128.0 62 38 piperacillin/tazobactam 1 50 <4.0/4.0 64.0/4.0 <4.0/4.0 256.0/4.0 90 8 10 40 <4.0/4.0 256.0/4.0 <4.0/4.0 >512.0/4.0 58 25 30 45 <4.0/4.0 64.0/4.0 <4.0/4.0 256.0/4.0 87 2 Active group ampicillin 1 45 4.0 >128.0 <2.0 >128.0 60 40 10 40 >128.0 >128.0 <2.0 >128.0 40 55 30 45 4.0 >128.0 <2.0 >128.0 53 47 piperacillin/tazobactam 1 45 <4.0/4.0 <4.0/4.0 <4.0/4.0 <4.0 4.0 100 10 40 <4.0/4.0 64.0/4.0 <4.0/4.0 64.0/4.0 88 30 45 <4.0/4.0 <4.0/4.0 <4.0/4.0 64.0/4.0 91 a S, susceptible; R, resistant; the percentage of intermediately susceptible isolates is not shown. 794
Probiotics and resistance Table 2. In vitro activity of antimicrobial agents against isolates of B. fragilis spp. in 20 patients treated with penicillin and placebo or active probiotic supplement Antimicrobial agent Day of sampling n MIC 50 MIC 90 MIC range S a (%) R a (%) Placebo group ampicillin/clavulanic acid 1 49 2.0/1.0 16.0/8.0 <1.0/0.5 32.0/16.0 90 2 10 50 2.0/1.0 16.0/8.0 <1.0/0.5 >64.0/32.0 80 6 30 50 <1.0/0.5 8.0/4.0 <1.0/0.5 16.0/8.0 98 cefoxitin 1 49 16.0 64.0 <4.0 64.0 61 27 10 50 32.0 64.0 <4.0 >256.0 48 18 30 50 32.0 64.0 <4.0 >256.0 50 24 cefoperazone 1 49 32.0 256.0 <4.0 >256.0 27 45 10 50 64.0 >256.0 16.0 >256.0 16 58 30 50 64.0 >256.0 8.0 >256.0 34 64 clindamycin 1 49 2.0 8.0 <0.5 >32.0 65 20 10 50 1.0 8.0 >0.5 >32.0 84 10 30 50 1.0 4.0 <0.5 >32.0 78 8 Active group ampicillin/clavulanic acid 1 50 <1.0/0.5 8.0/4.0 <1.0/0.5 32.0/16.0 94 2 10 45 <1.0/0.5 8.0/4.0 <1.0/0.5 32.0/16.0 91 4 30 50 2.0/1.0 8.0/4.0 <1.0/0.5 16.0/8.0 94 cefoxitin 1 50 16.0 64.0 <4.0 128.0 60 10 10 45 16.0 128.0 <4.0 128.0 60 16 30 50 16.0 128.0 <4.0 256.0 70 14 cefoperazone 1 50 64.0 256.0 <4.0 >256.0 26 64 10 45 128.0 >256.0 <4.0 >256.0 20 64 30 50 64.0 >256.0 8.0 256.0 20 62 clindamycin 1 50 2.0 >32.0 <0.5 >32.0 68 14 10 45 2.0 >32.0 <0.5 >32.0 71 22 30 50 <0.5 >32.0 <0.5 >32.0 78 18 a S, susceptible; R, resistant; the percentage of intermediately susceptible isolates is not shown. differences between the active and placebo groups. Over 90% of all isolates were resistant to penicillin. Two isolates from two patients in the placebo group were resistant to ampicillin/clavulanic acid, cefoxitin, cefoperazone, imipenem and metronidazole on day 10. All other isolates were susceptible to imipenem and metronidazole. Isolates of B. fragilis spp. from quinolone-treated patients showed a similar pattern to that found in the penicillin-treated patients as regards susceptibility to cefoxitin, cefoperazone and clindamycin whilst the majority of isolates were susceptible to ampicillin/clavulanic acid (Table 3). All isolates from one patient in the placebo group had reduced susceptibility to metronidazole (intermediately resistant on days 1 and 30, resistant on day 10). All isolates were susceptible to imipenem. Side effects No adverse events were reported. Eight patients treated with penicillin reported looser stools, four in the active and four in the placebo group. In the quinolone-treated patients there were two individuals who experienced looser stools, one from each of the placebo and active groups. One of the patients also complained of nausea during treatment. Discussion To a minor extent, effects on the normal microflora and development of resistance occurred more in patients treated with penicillin than in quinolone-treated patients. In the penicillintreated group, the impact of treatment on enterococci was seen mainly as increased numbers during the study. However, the numbers before treatment were rather low (median approximately 3.0 log 10 cfu/g faeces in both groups). Penicillin treatment did not affect resistance in enterococci, and in enterobacteria there was an increase in resistance to ampicillin and piperacillin/tazobactam. Approximately 40% of enterobacterial isolates in penicillin-treated patients were resistant to ampicillin before the study. There were differences between patients receiving placebo and active probiotic supplement. There were a few enterobacterial isolates in the placebo group resistant to cefotaxime and gentamicin. The majority of B. fragilis spp. were susceptible to ampicillin/clavulanic acid whilst susceptibility to cefoperazone was generally rather low. Reduced susceptibility to cefoxitin and clindamycin was also observed on all sampling occasions. In the quinolone-treated group the numbers of enterobacteria increased in the placebo group. The increased numbers of enterobacteria partly explain the increase in total numbers of quinolone-resistant aerobic isolates in this group but also reflect the increased numbers of resistant enterococci. Use of fluoroquinolone and length of hospital stay have been shown to be associated with ampicillin resistance in enterococci. 15 Only two patients harboured enterococci resistant to ampicillin and also to ciprofloxacin. Four patients were already colonized with resistant enterobacteria before the treatment. Intake of quinolones during the month preceding admission to hospital has 795
Å. Sullivan et al. Table 3. In vitro activity of antimicrobial agents against isolates of B. fragilis spp. in 16 patients treated with quinolones and placebo or active probiotic supplement Antimicrobial agent Day of sampling n MIC 50 MIC 90 MIC range S a (%) R a (%) Placebo group ampicillin/clavulanic acid 1 40 <0.5/0.25 16.0/8.0 <1.0/0.5 32.0/16.0 88 5 10 40 <0.5/0.25 8.0/4.0 <1.0/0.5 16.0/8.0 92 30 39 <0.5/0.25 32.0/16.0 <1.0/0.5 32.0/16.0 82 13 cefoxitin 1 40 8.0 32.0 <4.0 64.0 80 2 10 40 8.0 128.0 <4.0 64.0 75 12 30 39 16.0 64.0 <4.0 >256.0 80 15 cefoperazone 1 40 32.0 128.0 <4.0 >256.0 10 47 10 40 32.0 256.0 <4.0 >256.0 18 37 30 39 64.0 >256.0 <4.0 >256.0 23 56 clindamycin 1 40 1.0 32.0 <0.5 >32.0 88 10 10 40 <0.5 2.0 >0.5 >32.0 92 5 30 39 1.0 2.0 <0.5 >32.0 92 5 Active group ampicillin/clavulanic acid 1 34 <1.0/0.5 4.0/2.0 <1.0/0.5 16.0/8.0 94 10 35 <1.0/0.5 4.0/2.0 <1.0/0.5 4.0/2.0 100 30 35 <1.0/0.5 4.0/2.0 <1.0/0.5 8.0/4.0 100 cefoxitin 1 34 8.0 32.0 <4.0 64.0 82 9 10 35 8.0 32.0 <4.0 64.0 83 3 30 35 8.0 32.0 <4.0 64.0 80 3 cefoperazone 1 34 <4.0 64.0 <4.0 >256.0 9 62 10 35 <4.0 32.0 <4.0 256.0 23 43 30 35 32.0 128.0 <4.0 256.0 8 46 clindamycin 1 34 <0.5 2.0 <0.5 64.0 91 6 10 35 <0.5 4.0 <0.5 64.0 86 3 30 35 1.0 32.0 <0.5 32.0 86 11 a S, susceptible; R, resistant; the percentage of intermediately susceptible isolates is not shown. been shown to be a risk factor for the emergence of quinoloneresistant intestinal enterobacteria. 16 Selection of quinoloneresistant E. coli has also been shown to occur after short exposure to quinolones in the intestinal microflora of cancer patients. 17 None of the patients in this study had received antimicrobial agents or been admitted to hospital within the 3 months preceding the study. The resistance rates in B. fragilis group spp. in quinolone-treated patients followed the same pattern as for penicillin-treated patients. The overall difference between placebo and active supplemented groups was rather limited in this study. A plausible explanation could be poor compliance, which is indicated by the low rate of completed diary cards, in particular in the penicillin group. Some patients were old and weak and several subjects complained that they did not enjoy the taste and the consistency of the products. The recovery rate of the probiotic strain was also limited. Lactobacillus F19 was only detected in three of 10 in the penicillin-treated group and in three of eight of the quinolone-treated patients. The antimicrobial agents may also have influenced the probiotic strain: the MICs of penicillin and ciprofloxacin are 0.5 and 1.5 mg/l, respectively. Furthermore, the proportions of the patient groups may have been another factor of importance. The exclusion criteria used restricted the selection, and in particular the quinolone-treated group was rather small. The effect of probiotics is strain dependent and the strain Lactobacillus F19 in combination with Lactobacillus acidophilus and Bifidobacterium lactis has been shown to have a normalizing effect on the microflora of healthy subjects during administration of clindamycin. 18 In that study, which was commenced after the present one, the probiotic was delivered in a fermented milk product that might have been more agreeable to the patients. In conclusion, the effect of Lactobacillus F19 supplement on the emergence of resistance in enterococci, enterobacteria and B. fragilis group during treatment with penicillin or quinolones was limited. References 1. Cristino, J. M. (1999). Correlation between consumption of antimicrobials in humans and development of resistance in bacteria. International Journal of Antimicrobial Agents 12, 199 202. 2. Elliot, T. S. & Lambert, P. A. (1999). Antibacterial resistance in the intensive care unit: mechanisms and management. British Medical Bulletin 5, 259 76. 3. European Antimicrobial Resistance Surveillance System (EARSS). Annual Report 2002. [Online.] www.earss.rivm.nl (9 January 2004 date last accessed.) 4. Witte, W. (1999). Antibiotic resistance in Gram-positive bacteria: epidemiological aspects. Journal of Antimicrobial Chemotherapy 44, Suppl. A, 1 9. 5. Simonsen, G. S., Smabrekke, L., Monnet, D. L. et al. (2003). Prevalence of resistance to ampicillin, gentamicin and vancomycin in Enterococcus faecalis and Enterococcus faecium isolates from clinical 796
Probiotics and resistance specimens and use of antimicrobials in five Nordic hospitals. Journal of Antimicrobial Chemotherapy 51, 323 31. 6. Fluit, A. C., Jones, M. E., Schmitz, F. J. et al. (2000). Antimicrobial susceptibility and frequency of occurrence of clinical blood isolates in Europe from the SENTRY antimicrobial surveillance program, 1997 and 1998. Clinical Infectious Diseases 30, 454 60. 7. Leistevuo, T., Toivonen, P., Osterblad, M. et al. (1996). Problem of antimicrobial resistance of fecal aerobic Gram-negative bacilli in the elderly. Antimicrobial Agents and Chemotherapy 40, 2399 403. 8. Fang, H., Edlund, C., Hedberg, M. et al. (2002). New findings in b-lactam and metronidazole resistant Bacteroides fragilis group. International Journal of Antimicrobial Agents 19, 361 70. 9. Andremont, A. (2003). Commensal flora may play key role in spreading antibiotic resistance. ASM News 69, 601 7. 10. Vollaard, E. J. & Clasener, H. A. (1994). Colonization resistance. Antimicrobial Agents and Chemotherapy 38, 409 14. 11. Sullivan, A. & Nord, C. E. (2002). The place of probiotics in human intestinal infections. International Journal of Antimicrobial Agents 20, 313 9. 12. National Committee for Clinical Laboratory Standards. (2003). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Sixth Edition: Approved Standard M7-A6. NCCLS, Wayne, PA, USA. 13. National Committee for Clinical Laboratory Standards. (2001). Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria Fifth Edition: Approved Standard M11-A5. NCCLS, Wayne, PA, USA. 14. Björneholm, S., Eklöw, A., Saarela, M. et al. (2002). Enumeration and identification of Lactobacillus paracasei subsp. paracasei F19. Microbial Ecology in Health and Disease Suppl. 3, 7 13. 15. Torell, E., Cars, O., Olsson-Liljequist, B. et al. (1999). Near absence of vancomycin-resistant enterococci but high carriage rates of quinolone-resistant ampicillin-resistant enterococci among hospitalized patients and nonhospitalized individuals in Sweden. Journal of Clinical Microbiology 37, 3509 13. 16. Richard, P., Delangle, M. H., Raffi, F. et al. (2001). Impact of fluoroquinolone administration on the emergence of fluoroquinoloneresistant Gram-negative bacilli from gastrointestinal flora. Clinical Infectious Diseases 32, 162 6. 17. Perea, S., Hidalgo, M., Arcediano, A. et al. (1999). Incidence and clinical impact of fluoroquinolone-resistant Escherichia coli in the faecal flora of cancer patients treated with high dose chemotherapy and ciprofloxacin prophylaxis. Journal of Antimicrobial Chemotherapy 44, 117 20. 18. Sullivan, A., Barkholt, L. & Nord, C. E. (2003). Lactobacillus acidophilus, Bifidobacterium lactis and Lactobacillus F19 prevent antibiotic-associated ecological disturbances of Bacteroides fragilis in the intestine. Journal of Antimicrobial Chemotherapy 52, 308 11. 797