In vitro susceptibility of Bartonella species to 17 antimicrobial compounds: comparison of Etest and agar dilution

Transcription

1 Journal of Antimicrobial Chemotherapy (2006) 58, doi: /jac/dkl341 Advance Access publication 17 August 2006 In vitro susceptibility of Bartonella species to 17 antimicrobial compounds: comparison of Etest and agar dilution Christina Dörbecker 1 *, Anna Sander 1, Karin Oberle 1 and Tanja Schülin-Casonato 2 1 Institute for Medical Microbiology and Hygiene, University of Freiburg, Germany; 2 Department of Medical Microbiology, UMC St Radboud, Radboud University Nijmegen, The Netherlands Received 16 June 2006; returned 28 June 2006; revised 27 July 2006; accepted 30 July 2006 Objectives: In vitro susceptibility testing of 31 Bartonella spp. strains including 21 Bartonella henselae isolates was performed for 17 antimicrobial agents (telithromycin, four macrolides, five fluoroquinolones, five aminoglycosides, doxycycline and rifampicin). Methods: MICs were determined by agar dilution and Etest using chocolate agar containing 5% defibrinated sheep blood as assay medium. Longer incubation periods of 3 5 days in a humid atmosphere with 5% CO 2 were required until bacterial growth became visible and MICs could be read. Results: The ketolide telithromycin was the most active agent exhibiting the lowest MICs. The Bartonella spp. were also highly susceptible to macrolides, particularly clarithromycin, and to doxycycline and rifampicin, with MICs of 0.12 mg/l. Gatifloxacin, gemifloxacin and moxifloxacin were the most potent fluoroquinolones, with MICs ranging from 0.06 to 2 mg/l. Netilmicin was the most active agent among the aminoglycosides. Etest MICs correlated well with MICs determined by agar dilution. Conclusions: Telithromycin, macrolides, doxycycline and rifampicin were the most effective agents against Bartonella spp. Our data confirm that Etest may be a reliable method for determining susceptibility of Bartonella spp. Keywords: MICs, telithromycin, macrolides, fluoroquinolones, aminoglycosides Introduction In the last few years the number of identified Bartonella spp. has increased rapidly and the Bartonellaceae family currently contains 22 species and three subspecies. 1,2 At least eight of these species are associated with human infections including cat scratch disease (Bartonella henselae), bacillary angiomatosis and bacillary peliosis hepatis (B. henselae, Bartonella quintana), bacteraemia, endocarditis, Carrion s disease (Bartonella bacilliformis) and trench fever (B. quintana). 3 Laboratory diagnosis is usually confirmed by serological or PCR-based methods, because isolation of Bartonella spp. from clinical specimens requires long incubation times and special growth conditions and is rarely possible. 4,5 Treatment of Bartonella infections depends on clinical disease, status of the host immune system and severity of illness. Cat scratch disease is the most common clinical presentation of Bartonella infections and is usually self-limiting. In immunocompromised patients (especially HIV patients) infections due to Bartonella spp. may persist for a longer time and lifelong treatment, or prophylaxis may be necessary. Current recommendations for antibiotic treatment are based on a few case reports and very limited data from a few clinical studies. 6 Macrolides and tetracyclines are currently used as first-line antibiotics for the treatment of the majority of diseases caused by Bartonella spp. Treatment with other antibiotics such as aminoglycosides, fluoroquinolones, rifampicin, ceftriaxone and chloramphenicol has also been reported to be successful, and these agents have been recommended for the treatment of some clinical manifestations of Bartonella infections. 6,7 Owing to the fastidious nature of these microorganisms and the limited number of strains that have been isolated worldwide, only sparse information on in vitro susceptibility is available. In this study we determined the MICs of 17 antimicrobial agents (telithromycin, four macrolides, five fluoroquinolones, five aminoglycosides, doxycycline and rifampicin) of 31 Bartonella spp. strains by two different methods agar dilution and Etest.... *Correspondence address. Institute for Medical Microbiology, Immunology and Hygiene, University of Cologne, Goldenfelsstrasse 19 21, Cologne, Germany. Tel: / ; Fax: / ; christina.doerbecker@uk-koeln.de Ó The Author Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 In vitro susceptibility of Bartonella spp. Materials and methods Bacterial strains Thirty-one strains of Bartonella spp. were tested including 21 strains of B. henselae that were isolated from domestic cats in Germany, 8,9 two strains of B. quintana (CIP , B. quintana München) 10 and one strain each of Bartonella elizabethae (ATCC 35685), Bartonella tribocorum (CIP ), Bartonella alsatica (CIP ), Bartonella vinsonii subsp. vinsonii (ATCC VR-152), B. vinsonii subsp. arupensis (ATCC ), Bartonella schoenbuchii (R4), 11 Bartonella doshiae (ATCC ) and Bartonella grahamii (ATCC ). s Antimicrobial substances and their sources were as follows: telithromycin, erythromycin A, roxithromycin, clarithromycin, levofloxacin and rifampicin were gifts from Sanofi-Aventis (formerly Hoechst Marion Roussel, Romainville, France); azithromycin was a gift from Pfizer, Karlsruhe, Germany; ciprofloxacin and moxifloxacin were gifts from Bayer, Wuppertal, Germany; and gemifloxacin was a gift of SmithKline Beecham Pharmaceuticals, Essex, United Kingdom. Standard clinical preparations were purchased for doxycycline (Vibravenös Ò, 20 mg/ml, Pfizer), gentamicin (Refobacin Ò, 40 mg/ml, Merck, Darmstadt, Germany), tobramycin (Gernebcin Ò, 40 mg/ml, Lilly, Bad Homburg, Germany), amikacin (Biklin Ò, 40 mg/ml, Bristol-Myers Squibb, München, Germany), streptomycin (Streptomycin Grünenthal Ò, 1 g, Grünenthal, Aachen, Germany) and netilmicin (Certomycin Ò, 100 mg/ml, Essex Pharma, München, Germany). Antimicrobial powders were diluted as recommended by the manufacturer. Etest strips for erythromycin, roxithromycin, clarithromycin, azithromycin, ciprofloxacin, levofloxacin, moxifloxacin, gatifloxacin, gentamicin, tobramycin, amikacin, streptomycin, netilmicin, doxycycline and rifampicin were manufactured by AB Biodisk, Solna, Sweden. Gemifloxacin and telithromycin MICs were determined by agar dilution only, because Etests for these agents are not commercially obtainable. Susceptibility testing for gatifloxacin was performed by Etest only, because no other gatifloxacin formulation suitable for testing was available at the time of our study. Susceptibility testing Bartonella strains were cultured on chocolate agar containing 5% defibrinated sheep blood in a humid atmosphere with 5% CO 2 at 36 C and harvested after 5 days when bacterial growth was sufficient. Susceptibility testing was carried out with the same agar medium as described above. A final inoculum of 10 6 cfu/spot was used for agar dilution. MICs were determined after 3 5 days of incubation at 36 C with 5% CO 2. Susceptibility testing was performed twice for each strain and antimicrobial agent. Bacterial growth was verified on drug-free agar controls. Agar plates were inoculated with a 5.0 McFarland standard bacterial suspension prepared in Mueller Hinton broth for Etest susceptibility testing. MICs were read after 3 5 days of incubation. Etest MICs were rounded up to the next highest doubling dilution value for comparison with agar dilution results. Controls Staphylococcus aureus ATCC and Escherichia coli ATCC were used as controls. Agar dilution for control strains was performed according to CLSI (formerly NCCLS) guidelines except that chocolate agar was used as test medium. Etests were performed according to the manufacturers instructions. Results fell into normal ranges. Results MIC distributions determined by agar dilution and Etest are shown in Tables 1 and 2, respectively. Telithromycin was the most active agent in our study. All strains of Bartonella spp. except one were inhibited at the lowest concentration of telithromycin tested, mg/l. The isolates were also highly susceptible to the macrolides, doxycycline and rifampicin with MICs 0.12 mg/l. Among the macrolides, clarithromycin was the most effective agent with MICs 0.03 mg/l. Fluoroquinolone MICs ranged between 0.06 and 2 mg/l. Gatifloxacin, gemifloxacin and moxifloxacin were the most potent fluoroquinolones tested. The aminoglycosides were the least effective agents with MICs of up to 16 mg/l. Netilmicin was the most active aminoglycoside, and amikacin the least active. No significant difference in susceptibility between the B. henselae strains and the non-henselae Bartonella strains was observed (data not shown). Table 3 illustrates the agreement between Etest MICs and agar dilution MICs using agar dilution as reference method. Etest results correlated well with results determined by agar dilution. Agreement within 1 log 2 dilution for antimicrobial agents ranged from 43% for rifampicin to 97% for clarithromycin; agreement within 2 log 2 dilutions ranged from 83% to 100%, respectively. However, larger variations of up to 4 log 2 dilutions between MICs determined by the different methods were observed. Six per cent of the MICs showed variations between the two methods of 3 log 2 dilutions and more; 1.3% showed variations of 4 log 2 dilutions. MICs determined by Etest were overall slightly lower than agar dilution MICs, with 43% (164) being lower and 21% (82) being higher than the respective agar dilution MIC. Discussion In the present study antibiotic susceptibilities of 31 Bartonella spp. isolates against 17 antimicrobial compounds were determined by agar dilution and Etest. Owing to the fastidious nature of Bartonella spp., standardized susceptibility testing guidelines are not recommended, as of yet. 3 We performed susceptibility testing on chocolate agar, because growth of Bartonella spp. on Mueller Hinton agar supplemented with 5% sheep blood was insufficient and would have led to falsely low MICs. Longer incubation periods (3 5 days) in a humid atmosphere containing 5% CO 2 were required until bacterial growth became visible and MICs could be determined. Our results showed a high in vitro susceptibility of Bartonella spp. to the ketolide telithromycin. Telithromycin MICs were <0.002 mg/l for all strains except for the isolate B. quintana München with an MIC of mg/l. High activity of telithromycin against Bartonella spp. 12 and other intracellular bacteria including Chlamydia pneumoniae, 13 Legionella spp., 14 Rickettsia spp., Coxiella burnetii 12 and Francisella tularensis 15 has been documented previously. However, clinical trials to 785

3 Dörbecker et al. Table 1. MIC distributions determined by agar dilution on chocolate agar containing 5% defibrinated sheep blood for Bartonella spp. (number of strains) >16 Telithromycin (29) 28 1 Macrolides erythromycin (26) roxithromycin (30) clarithromycin (31) azithromycin (30) Aminoglycosides gentamicin (25) tobramycin (26) amikacin (29) streptomycin (26) netilmicin (26) Quinolones ciprofloxacin (27) levofloxacin (24) moxifloxacin (29) gemifloxacin (28) Doxycycline (28) Rifampicin (30) Some strains were excluded from analysis owing to insufficient bacterial growth in the agar dilution method. Table 2. MIC distributions determined by Etest on chocolate agar containing 5% defibrinated sheep blood for 31 Bartonella spp. strains >16 Macrolides erythromycin roxithromycin clarithromycin 27 4 azithromycin Aminoglycosides gentamicin tobramycin amikacin streptomycin netilmicin Quinolones ciprofloxacin levofloxacin moxifloxacin gatifloxacin Doxycycline >16 Rifampicin

4 In vitro susceptibility of Bartonella spp. Table 3. Comparison of MICs for Bartonella spp. determined by Etest and agar dilution (number of strains) Number of Etest MICs within indicated log 2 dilution of agar dilution MICs Agreement 1 log 2 a Agreement 2 log 2 a Erythromycin (26) (81%) 25 (96%) Roxithromycin (29) (72%) 26 (90%) Clarithromycin (31) (97%) 31 (100%) Azithromycin (30) (93%) 29 (97%) Gentamicin (25) (60%) 23 (92%) Tobramycin (26) (85%) 25 (96%) Amikacin (29) (76%) 28 (97%) Streptomycin (26) (62%) 23 (88%) Netilmicin (26) (58%) 23 (88%) Ciprofloxacin (27) (78%) 26 (96%) Levofloxacin (24) (54%) 23 (96%) Moxifloxacin (29) (69%) 29 (100%) Doxycycline (27) (78%) 26 (96%) Rifampicin (30) (43%) 25 (83%) a Numbers and percentages of strains with Etest MICs within 1 and 2 log 2 dilutions of agar dilution MICs. determine the in vivo activity of telithromycin against Bartonella species are warranted. The Bartonella spp. strains were also highly susceptible to the macrolides, doxycycline and rifampicin, with MICs 0.12 mg/l. These data confirm the results of previous studies. 8,16 23 Among the macrolides, clarithromycin was the most active agent in our study, followed by azithromycin, whereas previously published results showed comparable activity of clarithromycin and azithromycin. 21,22 Only a few cases of Bartonella infections treated with clarithromycin have been reported in the literature. 24,25 For the treatment of Bartonella infections with macrolides, azithromycin and erythromycin are usually preferred because better data from clinical trials are available for these drugs. 6 In a prospective, randomized, double-blind and placebo-controlled study on the treatment of cat scratch disease performed by Bass et al., 26 azithromycin taken orally for 5 days was shown to be effective within the first 4 weeks of treatment. However, newly enlarged lymph nodes or lymph nodes increasing in size did appear in some study subjects despite therapy. In both the azithromycin-treated and the placebo-treated group, the lymph nodes disappeared within 4 months. Since some Bartonella infections (e.g. cat scratch disease) are self-limiting diseases, the efficacy of an antimicrobial therapy remains difficult to assess. The aminoglycosides were the least active agents against Bartonella spp. in our study. Gentamicin MICs ranged from 0.12 to 8 mg/l. In previous studies aminoglycoside MICs were slightly lower with gentamicin MICs ranging from 0.12 to 2 mg/l, to 1 mg/l 27 and <0.125 to 1 mg/l 18 for Bartonella spp. strains determined by agar dilution, and from to 1 mg/l for B. henselae strains tested by Etest. 23 This discrepancy between our study and previous studies may be due to the different strains that were tested, and to differences in susceptibility testing techniques (e.g. different agar media and variations in inoculum size). However, interpretation of susceptibility results obtained for aminoglycosides against Bartonella spp. using agar dilution or Etest methodology is difficult, because the incubation in a CO 2 -enriched atmosphere can lead to a reduction of antimicrobial activity of aminoglycosides and may result in falsely high MICs. 28,29 Rolain et al. 16 determined the antibiotic susceptibility of B. quintana in a new cell culture model with human erythrocytes, and this approach showed that gentamicin was bactericidal at 4 mg/l after prolonged incubation. However, the bactericidal effect of gentamicin is limited to Bartonellaceae that are extraerythrocytic; intra-erythrocytic Bartonellaceae are protected from gentamicin. 16,30 Bactericidal activity of aminoglycosides has also been demonstrated against Bartonella spp. grown in liquid medium 27 and in endothelial cells. 31 Owing to the in vitro bactericidal effect of aminoglycosides against Bartonella spp. and their efficacy in the treatment of endocarditis 32 and B. quintana bacteraemia, 33,34 aminoglycosides are recommended for combination therapy in the treatment of these infections. A few case reports have described the effective treatment of some Bartonella infections with fluoroquinolones, 7 and their ability to achieve high intracellular concentrations may contribute to efficacy. However, MICs of fluoroquinolones in our series were high compared with those of macrolides or telithromycin. Fluoroquinolone MICs were lowest for the newer agents gatifloxacin, gemifloxacin and moxifloxacin, but their potential role in the therapy of Bartonella infections remains unclear. In vitro susceptibility testing of B. henselae using Etest was previously performed by Wolfson et al. 19 and Pendle et al. 23 Results of both studies were comparable with formerly published results determined by agar dilution. In our study, we performed both Etest and agar dilution, and we tested B. henselae and nonhenselae Bartonella strains. Etest MICs correlated well with MICs determined by agar dilution for all antimicrobial agents tested. Larger variations between MICs determined by Etest compared with agar dilution of some strains may be due to the fastidious 787

5 Dörbecker et al. nature of Bartonella spp. and to the different nature of the assays. Our results confirm that in vitro susceptibility testing of Bartonella spp. by Etest is a simple, reproducible and reliable method for diagnostic laboratories. Although the clinical value of in vitro susceptibility testing for Bartonella spp. remains uncertain, these data may be useful for therapeutic interventions in diseases due to Bartonella spp. as well as for monitoring development of resistance in some isolates. Transparency declarations None to declare. References 1. Maillard R, Riegel P, Barrat F et al. Bartonella chomelii sp. nov., isolated from French domestic cattle (Bos taurus). Int J Syst Evol Microbiol 2004; 54: Gundi VA, Davoust B, Khamis A et al. Isolation of Bartonella rattimassiliensis sp. nov. and Bartonella phoceensis sp. nov. from European Rattus norvegicus. J Clin Microbiol 2004; 42: Sander A. Bartonellosis. In: Cimolai N, ed. Laboratory Diagnosis of Bacterial Infections. New York: Marcel Decker, 2001; Dehio C, Sander A. Bartonella as emerging pathogens. Trends Microbiol 1999; 7: La Scola B, Raoult D. Culture of Bartonella quintana and Bartonella henselae from human samples: a 5-year experience ( ). J Clin Microbiol 1999; 37: Rolain JM, Brouqui P, Koehler JE et al. Recommendations for treatment of human infections caused by Bartonella species. Antimicrob Agents Chemother 2004; 48: Margileth AM. Antibiotic therapy for cat-scratch disease: clinical study of therapeutic outcome in 268 patients and a review of the literature. Pediatr Infect Dis J 1992; 11: Sander A, Bühler C, Pelz K et al. Detection and identification of two Bartonella henselae variants in domestic cats in Germany. J Clin Microbiol 1997; 35: Sander A, Ruess M, Bereswill S et al. Comparison of different DNA fingerprint techniques for molecular typing of Bartonella henselae isolates. J Clin Microbiol 1998; 36: Schmidt HU, Kaliebe T, Poppinger J et al. Isolation of Bartonella quintana from an HIV-positive patient with bacillary angiomatosis. Eur J Clin Microbiol Infect Dis 1996; 15: Dehio C, Lanz C, Pohl R et al. Bartonella schoenbuchii sp. nov., isolated from the blood of wild roe deer. Int J Syst Evol Microbiol 2001; 51: Rolain JM, Maurin M, Bryskier A et al. In vitro activities of telithromycin (HMR 3647) against Rickettsia rickettsii, Rickettsia conorii, Rickettsia africae, Rickettsia typhi, Rickettsia prowazekii, Coxiella burnetii, Bartonella henselae, Bartonella quintana, Bartonella bacilliformis, and Ehrlichia chaffeensis. Antimicrob Agents Chemother 2000; 44: Roblin PM, Hammerschlag MR. In vitro activity of a new ketolide antibiotic, HMR 3647, against Chlamydia pneumoniae. Antimicrob Agents Chemother 1998; 42: Schülin T, Wennersten CB, Ferraro MJ et al. Susceptibilities of Legionella spp. to newer antimicrobials in vitro. Antimicrob Agents Chemother 1998; 42: Maurin M, Mersali NF, Raoult D. Bactericidal activities of antibiotics against intracellular Francisella tularensis. Antimicrob Agents Chemother 2000; 44: Rolain JM, Maurin M, Mallet MN et al. Culture and antibiotic susceptibility of Bartonella quintana in human erythrocytes. Antimicrob Agents Chemother 2003; 47: Myers WF, Grossman DM, Wisseman CL Jr. Antibiotic susceptibility patterns in Rochalimaea quintana, the agent of trench fever. Antimicrob Agents Chemother 1984; 25: Maurin M, Raoult D. Antimicrobial susceptibility of Rochalimaea quintana, Rochalimaea vinsonii, and the newly recognized Rochalimaea henselae. J Antimicrob Chemother 1993; 32: Wolfson C, Branley J, Gottlieb T. The Etest for antimicrobial susceptibility testing of Bartonella henselae. J Antimicrob Chemother 1996; 38: Sobraques M, Maurin M, Birtles RJ et al. In vitro susceptibilities of four Bartonella bacilliformis strains to 30 antibiotic compounds. Antimicrob Agents Chemother 1999; 43: Maurin M, Gasquet S, Ducco C et al. MICs of 28 antibiotic compounds for 14 Bartonella (formerly Rochalimaea) isolates. Antimicrob Agents Chemother 1995; 39: Ives TJ, Manzewitsch P, Regnery RL et al. In vitro susceptibilities of Bartonella henselae, B. quintana, B. elizabethae, Rickettsia rickettsii, R. conorii, R. akari, and R. prowazekii to macrolide antibiotics as determined by immunofluorescent-antibody analysis of infected Vero cell monolayers. Antimicrob Agents Chemother 1997; 41: Pendle S, Ginn A, Iredell J. Antimicrobial susceptibility of Bartonella henselae using Etest methodology. J Antimicrob Chemother 2006; 57: Krause R, Wenisch C, Fladerer P et al. Osteomyelitis of the hip joint associated with systemic cat-scratch disease in an adult. Eur J Clin Microbiol Infect Dis 2000; 19: Gazineo JL, Trope BM, Maceira JP et al. Bacillary angiomatosis: description of 13 cases reported in five reference centers for AIDS treatment in Rio de Janeiro, Brazil. Rev Inst Med Trop Sao Paulo 2001; 43: Bass JW, Freitas BC, Freitas AD et al. Prospective randomized double blind placebo-controlled evaluation of azithromycin for treatment of cat-scratch disease. Pediatr Infect Dis J 1998; 17: Rolain JM, Maurin M, Raoult D. Bactericidal effect of antibiotics on Bartonella and Brucella spp.: clinical implications. J Antimicrob Chemother 2000; 46: Reynolds AV, Hamilton-Miller JM, Brumfitt W. Diminished effect of gentamicin under anaerobic or hypercapnic conditions. Lancet 1976; 1: Traub WH, Leonhard B. Antibiotic susceptibility tests with fastidious and nonfastidious bacterial reference strains: effects of aerobic versus hypercapnic incubation. Chemotherapy 1995; 41: Schülein R, Seubert A, Gille C et al. Invasion and persistent intracellular colonization of erythrocytes. A unique parasitic strategy of the emerging pathogen Bartonella. J Exp Med 2001; 193: Musso D, Drancourt M, Raoult D. Lack of bactericidal effect of antibiotics except aminoglycosides on Bartonella (Rochalimaea) henselae. J Antimicrob Chemother 1995; 36: Raoult D, Fournier P-E, Vandenesch F et al. Outcome and treatment of Bartonella Endocarditis. Arch Intern Med 2003; 163: Foucault C, Raoult D, Brouqui P. Randomized open trial of gentamicin and doxycycline for eradication of Bartonella quintana from blood in patients with chronic bacteremia. Antimicrob Agents Chemother 2003; 47: Foucault C, Barrau K, Brouqui P et al. Bartonella quintana bacteremia among homeless people. Clin Infect Dis 2002; 35: