Journal of Medical Microbiology (0), 60, 7 76 DOI 0.099/jmm.0.07573-0 Comparison of minimum inhibitory concentration by broth microdilution testing versus standard disc diffusion testing in the detection of penicillin, erythromycin and ciprofloxacin resistance in viridans group streptococci Yasunori Maeda,, Colin E. Goldsmith, Wilson A. Coulter, 3 Charlene Mason, 3 James S. G. Dooley, Colm J. Lowery, B. Cherie Millar and John E. Moore, Correspondence John E. Moore jemoore@niphl.dnet.co.uk Northern Ireland Public Health Laboratory, Department of Bacteriology, Belfast City Hospital, Lisburn Road, Belfast, Northern Ireland BT9 7AD, UK School of Biomedical Sciences, University of Ulster, Coleraine, Northern Ireland BT5 SA, UK 3 School of Dentistry, Queen s University of Belfast, Royal Group of Hospitals, Belfast, Northern Ireland, UK Received 30 October 00 Accepted 5 August 0 The aim of this study was to investigate the reliability of disc diffusion testing with penicillin, erythromycin and ciprofloxacin within the viridans group streptococci (VGS). In total, the antibiotic susceptibilities of 67 VGS isolates were compared by standard disc diffusion and broth microdilution methods, and these phenotypic data were compared to the carriage of the respective gene resistance determinants [ermb and mefa/e (macrolides); QRDR, gyra, gyrb, parc and pare (quinolones)]. Overall, there were 35 discrepancies [resistant by MIC and susceptible by zone diameter (.0 %)] between MIC and disc diameter when penicillin susceptibility was interpreted by Clinical and Laboratory Standards Institute criteria. Scattergrams showed a bimodal distribution between non-susceptible and susceptible strains when erythromycin susceptibility was tested by both methods. Thirty-four (0. %) isolates were categorized as resistant by MIC breakpoints, while disc diameter defined these as having intermediate resistance. With ciprofloxacin, three isolates (. %) showed minor discrepancies between MIC breakpoints and disc diameter. Isolates nonsusceptible to all three antimicrobial agents tested were reliably distinguished from susceptible isolates by disc diffusion testing, except for the detection of low-level resistance to penicillin, where broth microdilution or an alternative quantitative MIC method should be used. Otherwise, we conclude that disc diffusion testing is a reliable method to detect strains of VGS non-susceptible to penicillin, erythromycin and ciprofloxacin, as demonstrated with their concordance to their gene resistance characteristics. INTRODUCTION Viridans group streptococci (VGS) are causative agents of infective endocarditis (IE) and bacteraemia (Razonable et al., 00). In neutropenic patients and cancer patients, VGS cause viridans-related septic shock syndrome, which results in high mortality (0 00 %) (Bruckner & Gigliotti, 006). The emergence of antimicrobial-resistant VGS has been reported worldwide, although the resistant rates are variable due to differences in the sampling targets in each study. Resistance rates are as follows: penicillin resistance Abbreviations: CLSI, Clinical and Laboratory Standards Institute; IE, infective endocarditis; QRDR, quinolone resistance-determining region; VGS, viridans group streptococci. rate ranges from.3 to 5.0 % (Finland, Korea), erythromycin resistance rate ranges from 7.0 to 79.3 % (Sweden, Spain) and quinolone resistance rate ranges from.6 to 7.0 % (UK, USA) (Lindgren et al., 007; Uh et al., 00; Westling et al., 006; Cerdá Zolezzi et al., 00; Kerr et al., 999; Han et al., 006). To determine antimicrobial susceptibility, disc diffusion and broth microdilution methods are widely employed. In particular, the standardized disc diffusion test is the most frequently used methodology in the clinical microbiology laboratory, because of its relatively low cost and its wide application in most clinical microbiology analyses globally. The American Proficiency Institute and the World Health Organization have reported a wide diversity in results from 7 07573 G 0 SGM Printed in Great Britain
MIC versus disc testing in viridans streptococci laboratories that use disc diffusion testing to determine susceptibilities of Streptococcus pneumoniae (Tenover et al., 00; Edson et al., 006). Therefore, it is important to have reliable interpretive standards for disc diffusion testing with VGS organisms, as well as MIC breakpoints. Currently, such criteria for VGS are incomplete relating to Clinical and Laboratory Standards Institute (CLSI) standard methodology, for example, penicillin disc diffusion testing is commented on as not being reliable and ciprofloxacin is not included in the interpretive standards (CLSI, 005). In this study, we wished to evaluate the CLSI criteria for erythromycin, and to look into the possibility of developing interpretive criteria for susceptibility testing for penicillin and ciprofloxacin with VGS organisms. METHODS Bacterial isolates. VGS (n567), were obtained from the culture collection of the Northern Ireland Public Health Laboratory, Belfast City Hospital. These isolates had originally been collected from respiratory specimens, and had been obtained by conventional microbiological culture techniques for the presence of VGS organisms and pneumococci. All isolates were recovered on Columbia Blood agar (Oxoid) supplemented with 5 % (v/v) defibrinated horse blood and incubated for h at 37 uc, under microaerophilic conditions, with 5 % (v/v) CO. The identity of each VGS isolate employed in this study was determined molecularly by the sequencing of five gene loci, namely rnpb, 6S rrna, 6S 3S rrna, soda and dnaj (unpublished data). Characteristics of macrolide resistance determinant genes and quinolone resistance-determining regions (QRDRs) are summarized in Tables and. Determination of antimicrobial susceptibility. Antibiotic susceptibility tests for penicillin, erythromycin and ciprofloxacin were performed on all VGS isolates and pneumococci by standard disc susceptibility analysis, as well as by determination of the MIC value. Standardized routine disc susceptibility tests were performed employing an inoculum density of each isolate at a McFarland standard. The MIC for the same antibiotics was determined by broth microdilution according to the CLSI standard method (CLSI, 005). Susceptibility was interpreted using the criteria for Streptococcus spp. other than Streptococcus pneumoniae. In order to estimate ciprofloxacin susceptibility, interpretive criteria for ofloxacin were used. Susceptibility test discs and lyophilized antimicrobial tablets were purchased from Becton Dickinson and MAST, respectively. Molecular characterization of macrolide- and fluoroquinoloneresistance determinants. The QRDRs, gyra, gyrb, parc and pare gene loci, were analysed through sequence typing for all VGS isolates examined (Maeda et al., 0), as well as carrying out detection and characterization of the macrolide resistance determinants, the ermb and mefa/e gene loci (Tazumi et al., 009). These methods employed the sequencing of known gene resistance determinants associated with macrolide resistance, as well as fluoroquinolone resistance. RESULTS The MICs from broth microdilution tests and zone diameters of disc diffusion tests are shown in Tables and. Overall, there were 35 major discrepancies [i.e. resistant by MIC and susceptible by zone diameter (NCCLS, 99) (.0 %)] between MIC and disc diameter when penicillin susceptibility was interpreted by CLSI criteria (Fig. a). Scattergrams showed a bimodal distribution of organisms between non-susceptible and susceptible strains when erythromycin susceptibility was tested by both methods (Fig. b). Thirty-four (0. %) isolates were categorized as resistant by MIC breakpoints, while disc diameter defined these as intermediate (minor discrepancies). With ciprofloxacin, three isolates (. %) showed minor discrepancies, where two were resistant and intermediate by MIC breakpoints and disc diameter, and the reverse for the one remaining isolate (Fig. c). DISCUSSION Taking the broth microdilution method as the gold standard, we demonstrated that 35 penicillin-resistant strains were incorrectly identified as penicillin susceptible by disc diffusion testing using 0 unit penicillin discs, which corroborates a previous statement from the CLSI regarding CLSI interpretive standards, that penicillin, ampicillin and oxacillin disc diffusion testing is not reliable with VGS organisms. Although resistant strains (MIC mg ml ) were correctly identified by disc diffusion testing, from our results, it appears to be necessary to employ MIC methods to detect the intermediate susceptibility/resistance of VGS organisms. The clinical microbiologist ought to be aware Table. Patterns of macrolide resistance determinant genes of VGS strains used in this study, and range of zone diameter and MIC No. of isolates (n567) ermb Genotype mefa/e Zone diameter (mm) ERY MIC (mg ml ) + 6 0 6 3 6 + 0 0 6 + + 6 3 6 5 35 0.03 ERY, Erythromycin. http://jmm.sgmjournals.org 73
Y. Maeda and others Table. Patterns of efflux activity and substitutions at the QRDRs of VGS strains used in this study, and range of zone diameter and MIC No. of isolates (n567) Efflux Amino acid substitution Zone diameter (mm) GyrA GyrB ParC ParE CIP MIC (mg ml ) 70 +* 3 6 + S66YD + S79I/Y/R/C, D3G/Y/A 6 3 3 + SFd Q06V S79I K07Q 6 D35N SF D3Y 6 6 3 SL S66Y S79R, Q96D 6 6 7 6 30 CIP, Ciprofloxacin. *Increased efflux was detected by 5 mg reserpine ml. DAmino acid substitution before and after at position shown. dmutation position corresponds to S. pneumoniae R6. of the clinical significance of such a finding on the choice of antibiotic agent and the dosing regimen indicated, where the dose may be altered (e.g. increased in the case of b-lactam agents to ensure adequate concentrations above the organism s MIC value). The American Heart Association has proposed antibiotic guidelines for the treatment of IE caused by VGS (Baddour et al., 005). In these guidelines (Baddour et al., 005), penicillin is the recommended first line antimicrobial agent to treat IE and the recommended regimen is guided by the MIC of infecting VGS isolates, where the breakpoints are different from the stated CLSI standard, namely: resistant,. mg ml ; intermediate,.0. mg ml and mg ml ; and susceptible, 0. mg ml. In such cases, the full MIC should be determined rather than relying solely on disc diffusion testing. In the case of erythromycin susceptibility testing, results obtained by disc testing were considered valid, as we could positively match these against molecular characterization of the ermb and mefa/e gene determinant loci. Those isolates that possessed the ermb gene, which confers highlevel macrolide resistance, were categorized as resistant, while all strains assigned as intermediate possessed the mefa/e gene, which confers low-level resistance. Furthermore, none of the susceptible strains harboured either of these genes. Therefore, disc diffusion testing could be used to detect erythromycin resistant and intermediate strains reliably. Ciprofloxacin susceptibility was found to have the best concordance between disc diffusion and broth microdilution methods, even though ofloxacin interpretive standards were applied. Again, we were able to compare the phenotypic susceptibility data with the molecular QRDR characteristics of VGS organisms, which showed a good correlation between methods. None of the isolates having QRDR mutations were assigned as susceptible but were assigned as intermediate or resistant, whereas all other isolates with no QRDR mutation were categorized as susceptible, with a few exceptional isolates resistant via an active efflux pump. This mechanism confers low-level resistance and the feature is inhibited by reserpine. On account of this feature, it is easily monitored during MIC tests. Jumbe et al. (006) have suggested that active efflux pumps play a key role during the development of fluoroquinolone resistance. The low-level resistance caused by efflux pumps can increase the number of opportunities to obtain target mutations. Although it used to be considered that there was only one efflux system for fluoroquinolone resistance, putative ATPdependent efflux proteins PatA and PatB were described recently. The difference of this efflux system and the PmrA efflux system is its sensitivity to reserpine. PatA and PatB efflux systems are not influenced by reserpine (Marrer et al., 006). These results suggest that interpreting ciprofloxacin disc diameter with ofloxacin criteria can detect ciprofloxacin non-susceptible isolates, and potentially distinguish resistant and intermediate isolates. The ability to detect low-level resistance is important, as this can be caused by first step mutations in the QRDR, which will elevate the potential of the development of double mutations conferring high-level resistance (Schurek et al., 006). In conclusion, isolates non-susceptible to all three antimicrobial agents tested were reliably distinguished from susceptible isolates by disc diffusion testing, except for the detection of low-level resistance to penicillin, where broth microdilution or an alternative quantitative MIC method should be used. Otherwise, we conclude that disc diffusion testing is a reliable method to detect strains of VGS non-susceptible to penicillin, erythromycin and ciprofloxacin. 7 Journal of Medical Microbiology 60
MIC versus disc testing in viridans streptococci (a) 6 0.06 0.03 0.05 6 0 6 30 3 3 36 3 0 6 MIC (mg ml ) (b) 3 6 0.06 0.03 3 5 6 7 6 0 6 0 6 30 3 3 (c) 6 3 6 3 5 6 7 9 3 6 0 6 0 6 30 3 5 6 7 9 0 Fig.. Scattergram of MIC and zone diameter of (a) penicillin, (b) erythromycin and (c) ciprofloxacin. Horizontal lines represent MIC breakpoints and vertical lines indicate the interpretive standard for zone diameter as defined by the CLSI. http://jmm.sgmjournals.org 75
Y. Maeda and others ACKNOWLEDGEMENTS This work was financially supported through an HSC R&D Office commissioned grant: Antimicrobial Resistance Action Plan (AMRAP) (COM/730/0). REFERENCES Baddour, L. M., Wilson, W. R., Bayer, A. S., Fowler, V. G., Jr, Bolger, A. F., Levison, M. E., Ferrieri, P., Gerber, M. A., Tani, L. Y. & other authors (005). Infective endocarditis: diagnosis, antimicrobial therapy, and management of complications: a statement for healthcare professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, and the Councils on Clinical Cardiology, Stroke, and Cardiovascular Surgery and Anesthesia, American Heart Association: endorsed by the Infectious Diseases Society of America. Circulation, e39 e3. Bruckner, L. & Gigliotti, F. (006). Viridans group streptococcal infections among children with cancer and the importance of emerging antibiotic resistance. Semin Pediatr Infect Dis 7, 53 60. Cerdá Zolezzi, P., Rubio Calvo, M. C., Millán, L., Goñi, P., Canales, M., Capilla, S., Durán, E. & Gómez-Lus, R. (00). Macrolide resistance phenotypes of commensal viridans group streptococci and Gemella spp. and PCR detection of resistance genes. Int J Antimicrob Agents 3, 5 59. CLSI (005). Performance Standards for Antimicrobial Susceptibility Testing, document M00 S5. Wayne, PA: Clinical and Laboratory Standards Institute. Edson, D. C., Glick, T. & Massey, L. D. (006). Susceptibility testing practices for Streptococcus pneumoniae: results of a proficiency testing survey of clinical laboratories. Diagn Microbiol Infect Dis 55, 5 30. Han, X. Y., Kamana, M. & Rolston, K. V. (006). Viridans streptococci isolated by culture from blood of cancer patients: clinical and microbiologic analysis of 50 cases. J Clin Microbiol, 60 65. Jumbe, N. L., Louie, A., Miller, M. H., Liu, W., Deziel, M. R., Tam, V. H., Bachhawat, R. & Drusano, G. L. (006). Quinolone efflux pumps play a central role in emergence of fluoroquinolone resistance in Streptococcus pneumoniae. Antimicrob Agents Chemother 50, 30 37. Kerr, K. G., Armitage, H. T. & McWhinney, P. H. (999). Activity of quinolones against viridans group streptococci isolated from blood cultures of patients with haematological malignancy. Support Care Cancer 7, 30. Lindgren, M., Jalava, J., Rantakokko-Jalava, K. & Meurman, O. (007). In vitro susceptibility of viridans group streptococci isolated from blood in southwest Finland in 993 00. Scand J Infect Dis 39, 50 53. Maeda, Y., Murayama, M., Goldsmith, C. E., Coulter, W. A., Mason, C., Millar, B. C., Dooley, J. S. G., Lowery, C. J., Matsuda, M. & other authors (0). Molecular characterization and phylogenetic analysis of quinolone resistance-determining regions (QRDRs) of gyra, gyrb, parc and pare gene loci in viridans group streptococci isolated from adult patients with cystic fibrosis. J Antimicrob Chemother 66, 76 6. Marrer, E., Schad, K., Satoh, A. T., Page, M. G., Johnson, M. M. & Piddock, L. J. (006). Involvement of the putative ATP-dependent efflux proteins PatA and PatB in fluoroquinolone resistance of a multidrug-resistant mutant of Streptococcus pneumoniae. Antimicrob Agents Chemother 50, 65 693. NCCLS (99). Development of In Vitro Susceptibility Testing Criteria and Quality Control Parameters, approved standard M3-A. Wayne, PA: National Committee for Clinical Laboratory Standards. Razonable, R. R., Litzow, M. R., Khaliq, Y., Piper, K. E., Rouse, M. S. & Patel, R. (00). Bacteremia due to viridans group streptococci with diminished susceptibility to levofloxacin among neutropenic patients receiving levofloxacin prophylaxis. Clin Infect Dis 3, 69 7. Schurek, K. N., Adam, H. J., Hoban, D. J. & Zhanel, G. G. (006). Call for the international adoption of microbiological breakpoints for fluoroquinolones and Streptococcus pneumoniae. Int J Antimicrob Agents, 66 69. Tazumi, A., Maeda, Y., Goldsmith, C. E., Coulter, W. A., Mason, C., Millar, B. C., McCalmont, M., Rendall, J., Elborn, J. S. & other authors (009). Molecular characterization of macrolide resistance determinants [erm(b) and mef(a)] in Streptococcus pneumoniae and viridans group streptococci (VGS) isolated from adult patients with cystic fibrosis (CF). J Antimicrob Chemother 6, 50 506. Tenover, F. C., Mohammed, M. J., Stelling, J., O Brien, T. & Williams, R. (00). Ability of laboratories to detect emerging antimicrobial resistance: proficiency testing and quality control results from the World Health Organization s external quality assurance system for antimicrobial susceptibility testing. J Clin Microbiol 39, 50. Uh, Y., Shin, D. H., Jang, I. H., Hwang, G. Y., Lee, M. K., Yoon, K. J. & Kim, H. Y. (00). Antimicrobial susceptibility patterns and macrolide resistance genes of viridans group streptococci from blood cultures in Korea. J Antimicrob Chemother 53, 095 097. Westling, K., Julander, I., Ljungman, P., Jalal, S., Nord, C. E. & Wretlind, B. (006). Viridans group streptococci in blood culture isolates in a Swedish university hospital: antibiotic susceptibility and identification of erythromycin resistance genes. Int J Antimicrob Agents, 9 96. 76 Journal of Medical Microbiology 60