Biofilms and-lactam activity

JAC

Journal of Antimicrobial Chemotherapy (1998) 41, 571–578

Correspondence Biofilms and -lactam activity J Antimicrob Chemother 1998; 41: 571–572 P. Gilberta* and M. R. W. Brownb a

School of Pharmacy and Pharmaceutical Science, University of Manchester, Manchester M13 9PL; b School of Pharmaceutical and Biological Sciences, Aston University, Birmingham B4 7ET, UK *Tel: +44-(0)161-2752361; Fax: +44-(0)161-2752396. Sir, We have read with interest the recent correspondence in the Journal from Budhani & Struthers,1 in which they commend the Sorbarod method of biofilm culture as a means of assessing the in-vitro susceptibilities of biofilm bacteria to antibiotics. As originators of the Sorbarod methodology,2 we concur with the authors’ plea for susceptibility testing to be made more relevant; at present, it serves only to distinguish low-, medium- and high-level resistance associated with changes in genotype. There are, however, a number of points raised in the article which require clarification. While it is widely accepted that biofilm bacteria are markedly less susceptible to antibiotics than their more rapidly growing planktonic counterparts, the authors’ suggestion that this is invariably attributable to the failure of antibiotics to penetrate the glycocalyx is misleading. The mechanisms by which biofilms contribute to reduced susceptibility remain unclear, but a number of different explanations have been proposed.3 For chemically reactive biocides, such as chlorine, iodine and peroxygens,4 and for highly charged antibiotics, such as the glycopeptides, 5 the glycocalyx does indeed greatly affect the ability of the antimicrobial to reach those cells that are deep within the biofilm. On the other hand, for relatively unreactive, uncharged agents, such as the -lactams, such reaction-diffusion limitation is unlikely to occur.6 The glycocalyx may, however, contribute to reduced susceptibility to -lactams if the antibiotic is susceptible to inactivation by -lactamases and if the -lactamase is derepressed while the bacterium is in the biofilm mode of growth.7 In such cases, the enzyme is concentrated within the extracellular polymer matrix and hydrolyses the drug as it penetrates.8 Similar mechanisms whereby susceptibility has been compromised have been demonstrated with other agents, e.g., overproduction of formaldehyde lyase and formaldehyde

dehydrogenase by biofilm organisms leads to a reduction in susceptibility to formaldehyde.9 Reduced susceptibility to -lactams amongst biofilm bacteria is more likely to be a function of a diminished growth rate within the deeper recesses of the biofilm which causes the expression of penicillin-binding proteins that are unrepresentative of those normally targeted by these antibiotics.10,11 Retarded growth also affects the bactericidal action of the -lactams because transpeptidase inhibition, which induces cellular injury, is directly related to growth rate. The growth rate achieved by Budhani & Struthers1 in Sorbarod biofilms (an approximately 10 h generation time) was indeed much reduced compared with that (approximately 2–3 h) of the same organism (a strain of Streptococcus pneumoniae) grown conventionally in the same medium (Brain Heart Infusion broth), but was still greater than that (20–30 h) which is likely to be observed in vivo. Because incubation in the presence of the antibiotics took place over 18 h, at least two generation times would have elapsed following exposure—a period that is sufficiently long to bring about a lethal effect. Under these circumstances, bacteria growing in vitro are unlikely to reflect the antibiotic susceptibility profile of those growing in vivo and the use of a biofilm model might not distinguish between MICs/MBCs determined conventionally and biofilm eradication concentrations (BECs); this is particularly likely when the medium contains a relatively rich source of nutrients. Therefore, whilst we are broadly in agreement with the views expressed by Budhani & Struthers,1 we suggest that, had they studied other antibiotics, e.g., the quinolones, vancomycin, teicoplanin, gentamicin or erythromycin, the outcome of their investigation would have been markedly different and the benefits of employing susceptibility testing techniques such as the biofilm method to antibiotic prescribing would have been unequivocal.

References 1. Budhani, R. K. & Struthers, J. K. (1997). The use of Sorbarod biofilms to study the antimicrobial susceptibility of a strain of Streptococcus pneumoniae. Journal of Antimicrobial Chemotherapy 40, 601–2. 2. Hodgson, A. E., Nelson, S. M., Brown, M. R. W. & Gilbert, P. (1995). A simple in-vitro model for growth control of bacterial biofilms. Journal of Applied Bacteriology 79, 87–93. 3. Gilbert, P., Das, J. & Foley, I. (1997). Biofilm susceptibility to antimicrobials. Advances in Dental Research 11, 160–7. 4. Huang, C. T., Yu, F. P., McFeters, G. A. & Stewart, P. S. (1995).

571 © 1998 The British Society for Antimicrobial Chemotherapy

Correspondence Nonuniform spatial patterns of respiratory activity within biofilms during disinfection. Applied and Environmental Microbiology 61, 2252–6. 5. Hoyle, B. D., Wong, C. K. & Costerton, J. W. (1992). Disparate efficacy of tobramycin on Ca(2+)-, Mg(2+)-, and HEPES-treated Pseudomonas aeruginosa biofilms. Canadian Journal of Microbiology 38, 1214–8. 6. Nichols, W. W., Evans, M. J., Slack, M. P. E. & Walmsley, H. L. (1989). The penetration of antibiotics into aggregates of mucoid and non-mucoid Pseudomonas aeruginosa. Journal of General Microbiology 135, 1291–303. 7. Giwercman, B., Jensen, E. T., Hoiby, N., Kharazmi, A. & Costerton, J. W. (1991). Induction of -lactamase production in Pseudomonas aeruginosa biofilm. Antimicrobial Agents and Chemotherapy 35, 1008–10. 8. Dibdin, G. H. (1997). Mathematical modeling of biofilms. Advances in Dental Research 11, 127–32. 9. Sondossi, M., Rossmore, H. W. & Wireman, J. W. (1985). Observation of resistance and cross-resistance to formaldehyde and a formaldehyde condensate biocide in Pseudomonas aeruginosa. International Biodeterioration 21, 105–6. 10. Brown, M. R. W., Collier, P. J. & Gilbert, P. (1990). Influence of growth rate on susceptibility to antimicrobial agents: modification of the cell envelope and batch and continuous culture studies. Antimicrobial Agents and Chemotherapy 34, 1623–8. 11. Cozens, R. M., Tuomanen, E., Tosch, W., Zak, O., Suter, J. & Tomasz, A. (1986). Evaluation of the bactericidal activity of -lactam antibiotics upon slowly growing bacteria cultured in the chemostat. Antimicrobial Agents and Chemotherapy 29, 797–802.

Nucleotide sequences of inhibitor-resistant TEM -lactamases J Antimicrob Chemother 1998; 41: 572–573 C. J. Thomson*, P. M. A. Shanahan and S. G. B. Amyes Department of Medical Microbiology, University of Edinburgh, Teviot Place, Edinburgh EH8 9AG, UK *Tel: +44-(0)131-6503127; Fax: +44-(0)131-6506531; E-mail: [email protected] Sir, In 1992, the first TEM-derived -lactamase resistant to -lactamase inhibitors such as clavulanic acid (TRC-1) was identified in a clinical isolate.1 Since then, a number of other inhibitor-resistant TEM (IRT) enzymes produced by clinical isolates of Escherichia coli have been characterized, a development that has demonstrated the diversity of the genes encoding this resistance. DNA sequence analyses have shown that the IRT enzymes differ from the TEM-1 progenitor by as many as three

amino acid substitutions located predominantly at positions Met69, Trp165, Met182, Arg244, Arg275 or Asn276.2,3 The first report of the presence of these enzymes in Enterobacteriaceae other than E. coli was that of Lemozy et al.4 who identified two IRT -lactamases in strains of Klebsiella pneumoniae. More recently, Bret et al.5 described an IRT -lactamase derived from TEM-2 that was produced by strains of Proteus mirabilis. We have previously shown that an IRT -lactamase can be selected directly from TEM-1 in vitro by subculturing a strain of E. coli containing a plasmid which encodes TEM1 in the presence of a combination of amoxycillin and clavulanic acid.6 We have also demonstrated in vitro the ability to select back-mutations to the TEM-1 -lactamase from TRC-1, thereby restoring susceptibility to the clavulanic acid.7 As a result of these findings, we speculated that the presence of such -lactamases in clinical isolates may be a consequence of a two-way flux, with mutations in both forward and backward directions.7 The nucleotide sequences of these two selected lactamases have now been determined and compared with the DNA sequence of the original TRC-1 -lactamase.1 Direct DNA sequencing of PCR products was performed as described previously.8 Two oligonucleotide primers, 5 -CTCTCTAGAAAAAGGAAGAGTATGATT-3 and 5 -CTCGCATGCGTAAACTTGGTCTGCCAGTTA-3 , were used to amplify a fragment (c. 900 bp) of the TEM-1 -lactamase gene with a Taq polymerase kit obtained from Life Technologies (Paisley, UK). Single-stranded DNA from the generated PCR product was prepared from Dynabeads M-280 (Dynal A.S., Oslo, Norway) and processed with a Sequenase version 2.0 kit (Amersham Pharmacia Biotech UK Ltd). DNA sequencing analysis demonstrated that TRC-1 differed from TEM-1 by an amino acid substitution at position 244 (CGC (Arg) AGC (Ser)). This substitution has been described in another IRT enzyme, IRT-2 (TEM30).9 Arg244 is positioned at the beginning of strand 4 where it contributes to the active binding of substrates and clavulanic acid and plays a specific role in the inactivation process. The replacement of Arg244 with an amino acid residue with a shorter, uncharged side-chain, such as serine, accounts for the loss of -lactamase inhibitor activity as it affects binding and catalysis by preventing the interaction of the amino acid at position 244 with an essential water molecule (reviewed by NicolasChanoine10). It is noteworthy that a different amino acid substitution at position 244 (CGC (Arg) GGC (Gly)) was identified in the IRT variant generated in vitro from TEM-1. The effect is the same as that described above and, phenotypically, this enzyme appears identical to TRC-1 (TEM-30).6 DNA sequencing of the putative TEM-1 revertant gene generated from TRC-1 revealed the presence of arginine at position 244. Interestingly, this amino acid was represented by a codon (AGG) that differed from the codon at position 244 described by

572

Correspondence Table. Amino acid substitutions in IRT -lactamases Amino acid substitution (codon) at position -Lactamase TEM-1 TRC-1 clinical isolate TEM-1 revertant from TRC-1 IRT variant of TEM-1

69

165

182

244

275

276

Reference

Met (ATG) Trp (TGG)Met (ATG) Arg (CGC) Arg (CGA) Asn (AAT) Ser (AGC) Arg (AGG) Gly (GGC)

Sutcliffe (CGC).11 It should be remembered, however, that this TEM-1-like sequence has been selected back from TRC-1, in which the amino acid at position 244 was Ser (AGC), not Arg (CGC). The various amino acid substitutions are shown in the Table. The findings of this study confirm at the DNA level that forward mutations from TEM-1 to an IRT variant and back mutations from TRC-1 to TEM-1 are possible in vitro and suggest that such mutations are also likely to occur in vivo as the selective pressures of -lactamase inhibitors are applied and removed.

10 1 7 6

Salmonella typhi in India. Journal of Antimicrobial Chemotherapy 37, 891–900. 9. Belaaouaj, A., Lapoumeroulie, C., Caniça, M. M., Vedel, G., Nevot, R., Krishnamoorthy, R. et al. (1994). Nucleotide sequences of the genes coding for the TEM-like -lactamases IRT-1 and IRT-2 (formerly called TRI-1 and TRI-2). FEMS Microbiology Letters 120, 75–80. 10. Nicolas-Chanoine, M. H. (1997). Inhibitor-resistant lactamases. Journal of Antimicrobial Chemotherapy 40, 1–3.

-

11. Sutcliffe, J. G. (1978). Nucleotide sequence of the ampicillin resistance gene of Escherichia coli plasmid pBR322. Proceedings of the National Academy of Science of the USA 75, 3737–41.

References 1. Thomson, C. J. & Amyes, S. G. B. (1992). TRC-1: emergence of a clavulanic acid-resistant TEM -lactamase in a clinical strain. FEMS Microbiology Letters 91, 113–8. 2. Henqell, C., Chanal, C., Sirot, D., Labia, R. & Sirot, J. (1995). Molecular characterization of nine different types of mutants among 107 inhibitor-resistant TEM -lactamases from clinical isolates of Escherichia coli. Antimicrobial Agents and Chemotherapy 39, 427–30. 3. Stapleton, P., Wu, P.-J., King, A., Shannon, K., French, G. & Phillips, I. (1995). Incidence and mechanisms of resistance to the combination of amoxicillin and clavulanic acid in Escherichia coli. Antimicrobial Agents and Chemotherapy 39, 2478–83. 4. Lemozy, J., Sirot, D., Chanal, C., Huc, C., Labia, R., Dabernat, H. et al. (1995). First characterization of inhibitor-resistant TEM (IRT) -lactamases in Klebsiella pneumoniae strains. Antimicrobial Agents and Chemotherapy 39, 2580–3. 5. Bret, L., Chanal, C., Sirot, D., Labia, R. & Sirot, J. (1996). Characterization of an inhibitor-resistant enzyme IRT-2 derived from TEM-2 -lactamase produced by Proteus mirabilis strains. Journal of Antimicrobial Chemotherapy 38, 183–91. 6. Thomson, C. J. & Amyes, S. G. B. (1993). Selection of variants of the TEM-1 -lactamase, encoded by a plasmid of clinical origin, with increased resistance to -lactamase inhibitors. Journal of Antimicrobial Chemotherapy 37, 655–64. 7. Thomson, C. J. & Amyes, S. G. B. (1995). Back mutations to the TEM-1 -lactamase from TRC-1 lead to restored sensitivity to clavulanic acid. Journal of Medical Microbiology 42, 429–32. 8. Brown, J. C., Shanahan, P. M. A., Jesudason, M. V., Thomson, C. J. & Amyes, S. G. B. (1996). Mutations responsible for reduced susceptibility to 4-quinolones in clinical isolates of multi-resistant

High frequency of mutations at position 2144 of the 23S rRNA gene in clarithromycin-resistant Helicobacter pylori strains isolated in Spain J Antimicrob Chemother 1998 41: 573–574 D. Domingo*, T. Alarcón, J. C. Sanz, I. Sánchez and M. López-Brea Department of Microbiology, Hospital Universitario de la Princesa, C/Diego de León 62, Madrid 28006, Spain * Tel: +34-1-520-2317; Fax: +34-1-309-0047. Sir, The presence of Helicobacter pylori in the gastric or duodenal mucosa is strongly associated with the development of several disease processes, including gastritis, gastric and duodenal ulcers, gastric cancer and mucosaassociated lymphoid tissue lymphoma (MALT lymphoma).1 Several regimens have been proposed for the eradication of H. pylori and these usually comprise a combination of one or more antibiotics and a drug that modifies the gastric pH. One of the antibiotics that is prescribed most frequently for this purpose is clarithromycin. However, the use of this agent has led to the emergence of resistant strains, the rate of resistance

573

Correspondence varying from country to country; for example, we recently reported that 3.5% of H. pylori strains isolated in Spain during 1991–5 were resistant to clarithromycin.2 Standard techniques, such as the agar diffusion, agar dilution and broth microdilution methods and the Etest, are widely employed to determine the in-vitro susceptibilities of H. pylori isolates to clarithromycin. More recently, molecular biology techniques have been used to detect resistance to this agent3–5 which has been attributed to mutations (adenine to guanine or cytosine) at positions 2143 or 2144 in the 23S rRNA gene of the bacterium. Versalovic et al.6 developed a PCR technique which successfully identified the A G mutations; a 1.4 kb sequence of the 23S rRNA gene was amplified and the PCR products obtained following digestion with BsaI or MboII were analysed by electrophoresis. Of 59 clarithromycin-resistant clinical isolates of H. pylori, 54 (91.5%) had either the A2143 G or the A2144 G mutation, the former being detected more frequently than the latter (52.5% vs 39%). These investigators also observed that 83.9% (26 of 31) of isolates with the A2143 G mutation exhibited high-level resistance to clarithromycin (MIC 64 mg/L) and that 65% of isolates with MICs 64 mg/L contained the A2143 G mutation. On the other hand, most (58%) of the isolates with MICs 64 mg/L contained the A2144 G mutation. We used the technique of Versalovic et al.6 to investigate the mechanism of resistance to clarithromycin in 27 strains of H. pylori isolated from patients in Spain. The A2143 G or A2144 G mutation was detected in 22 (81.5%) of the isolates; in a further five (18.5%), the mechanism of resistance was not determined. In contrast to Versalovic et al., we identified the A2144 G mutation in a higher percentage of strains (55.6%) than the A2143 G mutation (25.9%). We also determined the MICs of clarithromycin for the 27 isolates by the agar dilution method. Overall, these were lower than the MICs reported by Versalovic et al., although, like the latter investigators, we observed a correlation between the A2143 G mutation and the magnitude of the MICs, i.e. the MICs for all seven strains with this mutation were 16 mg/L. On the other hand, the MICs for the 15 strains with the A2144 G mutation were lower, i.e. 16 mg/L for 11 (73.3%) strains and 16 mg/L (either 16 or 32 mg/L) for only four. For the five isolates in which neither mutation was detected, the MICs were 8, 16 and 32 mg/L for three strains and 64 mg/L for two. The availability of new technologies and their applications to the evaluation of the in-vitro activities of antibiotics against fastidious bacteria such as H. pylori, for which conventional methods of susceptibility testing are problematic, time-consuming and sometimes yield results that are difficult to interpret, has been an important development. These techniques have demonstrated the genetic diversity of H. pylori strains, an observation that can be accounted for by the variability in the frequency of

mutations in the 23S rRNA gene that are associated with resistance to clarithromycin; this variability is typified by the results of the present study and those of another recently carried out in the USA.6 It would now be of considerable interest to attempt to identify these or other mutations in H. pylori strains that are susceptible to clarithromycin according to standard susceptibility testing methods, but that are not eradicated by regimens that include clarithromycin.

References 1. Dunn, B. E., Cohen, H. & Blaser, M. J. (1997). Helicobacter pylori. Clinical Microbiology Reviews 10, 720–41. 2. López-Brea, M., Domingo, D., Sánchez, I. & Alarcón, T. (1997). Evolution of resistance to metronidazole and clarithromycin in Helicobacter pylori clinical isolates from Spain. Journal of Antimicrobial Chemotherapy 40, 279–81. 3. Versalovic, J., Shortridge, D., Kibler, K., Griffy, M. V., Beyer, J., Flamm, R. et al. (1996). Mutations in 23S rRNA are associated with clarithromycin resistance in Helicobacter pylori. Antimicrobial Agents and Chemotherapy 40, 477–80. 4. Szczebara, F., Dhaenens, L., Vincent, P. & Husson, M. O. (1997). Evaluation of rapid molecular methods for detection of clarithromycin resistance in Helicobacter pylori. European Journal of Clinical Microbiology and Infectious Diseases 16, 162–4. 5. Stone, G. G., Shortridge, D., Versalovic, J., Beyer, J., Flamm, R. K., Graham, D. Y. et al. (1997). A PCR-oligonucleotide ligation assay to determine the prevalence of 23S rRNA gene mutations in clarithromycin-resistant Helicobacter pylori. Antimicrobial Agents and Chemotherapy 41, 712–4. 6. Versalovic, J., Osato, M. S., Spakovsky, K., Pina Dore, M., Reddy, R., Stone, G. G. et al . (1997). Point mutations in the 23S rRNA gene of Helicobacter pylori associated with different levels of clarithromycin resistance. Journal of Antimicrobial Chemotherapy 40, 283–6.

Acinetobacter spp. isolates with reduced susceptibilities to carbapenems in a UK burns unit J Antimicrob Chemother 1998; 41: 574–576 M. J. Weinbrena*, A. P. Johnsonb, M. E. Kaufmannc and D. M. Livermoreb a

Department of Microbiology, Queen Mary’s University Hospital, Roehampton Lane, London SW15 5PN; bAntibiotic Reference and c Epidemiological Typing Units, Laboratory of Hospital Infection, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT, UK *Tel: +44-(0)181-7896611.

574

Correspondence Sir, Acinetobacter spp. are important opportunistic pathogens in burns patients, particularly those with large burns, in whom they are associated with local infection, graft failure, pneumonia and septicaemia.1 The susceptibility patterns of these bacteria are highly variable, but isolates frequently exhibit resistance to multiple antibiotics.2 Strains that are resistant to penicillins, cephalosporins, quinolones and aminoglycosides and susceptible only to carbapenems are endemic in many centres, including the Burns Unit at Queen Mary’s University Hospital (QMUH). Imipenem therefore became our treatment of choice for patients with infections caused by Acinetobacter spp., but was replaced with meropenem in August 1996; both carbapenems have similar in-vitro activities against Acinetobacter spp., including strains resistant to other -lactams.3 However, not long after the change to meropenem was implemented, an Acinetobacter sp. strain that was resistant to meropenem but susceptible to imipenem, according to the disc diffusion method,4 was isolated. The MICs of meropenem and imipenem for this strain (determined by the Etest method) were 2 mg/L and 0.5 mg/L respectively—in both cases, below the recommended MIC breakpoint for susceptibility (4 mg/L),4 but four- to eight-fold greater than the MICs for most Acinetobacter spp. isolates.3 Strains demonstrating a similar pattern of susceptibility were recovered during subsequent months and retesting of isolates obtained during the first half of 1996 (i.e., before meropenem was introduced) revealed several more that exhibited the same pattern. In the course of determining the susceptibility of one isolate by the disc diffusion method, synergy was observed between meropenem (a 10 g disc) and the adjacent piperacillin/tazobactam (a 75 g/10 g disc), the

latter potentiating the activity of the former. It was also noted that colonies in proximity to the two drugs were highly mucoid, an observation that was subsequently made in respect of other isolates. The detection of synergy between meropenem and tazobactam suggested that the reduced susceptibility to the carbapenems was -lactamase-mediated. To investigate this link, we carried out further studies with three representative isolates exhibiting reduced susceptibilities, as well as a fully susceptible control strain; the results are shown in the Table. Pulsed-field gel electrophoresis of ApaI-digested chromosomal DNA extracted from the organisms demonstrated that those with reduced susceptibilities were indistinguishable from each other, but differed from the control. Crude sonicates of cell suspensions were assayed by UV spectrophotometry with 0.1 mM solutions of each carbapenem in 0.1 M phosphate buffer, pH 7.0. Imipenem and meropenem were hydrolysed, at similar rates, by the extracts of isolates with reduced susceptibilities, but not by the extract of the control strain, thereby confirming the presence of carbapenemase activity in the former but not the latter. The activities of the extracts of the strains exhibiting reduced susceptibilities were approximately 50% lower when they were incubated for 15 min in the presence of 5 M tazobactam before the addition of the carbapenem, an observation that is in accordance with the results of the synergy studies. Finally, isoelectric focusing 5 revealed that the isoelectric points (pIs) of the -lactamases of the isolates with reduced susceptibilities to carbapenems were 9.0, although two strains also possessed -lactamases with pIs of 7.0. The contributions of the individual enzymes to carbapenem hydrolysis are currently under investigation.

Table. Characteristics of the strains of Acinetobacter spp. studied Isolate no. Characteristic

S233

1734

BR

Date isolated (month/year) MIC (mg/L) imipenem meropenem cefotaxime ceftazidime piperacillin piperacillin/tazobactam ciprofloxacin amikacin Hydrolysis of carbapenems by -lactamases (nmol/min/mL extract) imipenem meropenem

2/96

6/96

4/97

250.5 252 256 256 256 264 258 258

251 252 256 256 256 258 258 254

250.5 252 256 256 256 128 258 254

251.8 251.2

254.7 253.2

251.2 250.7

575

B1939 (control) 4/97 2 0.12 2 0.5 216 2 4 216 2 8 2 0.25 2 0.5

0.1 0.1

Correspondence Whether carbapenemase-producing strains of Acineto bacter spp. are unique to the QMUH Burns Unit or are more widespread throughout the UK is not currently known, although data from abroad suggest that they are becoming more prevalent.6 What is not in doubt is that diagnostic laboratories should be aware of the difficulties in identifying them. With regard to our isolates, the disc diffusion method detected reduced susceptibility to meropenem, but not to imipenem. The efficacy of carbapenems as treatment of patients with infections caused by these bacteria is also unclear. Although the MICs of both agents, and particularly those of imipenem, are below the susceptibility breakpoint, we have adopted a cautious approach to the use of these drugs and would welcome data regarding clinical response rates in other patient populations. For the treatment of patients infected with such organisms, the therapeutic options at QMUH include sulbactam/ampicillin and polymyxin, these being the only antibiotics to which the isolates are consistently susceptible.

References 1. Tilley, P. A. G. & Roberts, F. J. (1994). Bacteremia with Acinetobacter species: risk factors and prognosis in different clinical settings. Clinical Infectious Diseases 18, 896–900. 2. Bergogne-Berezin, E. & Towner, K. J. (1996). Acinetobacter spp. as nosocomial pathogens: microbiological, clinical, and epidemiological features. Clinical Microbiology Reviews 9, 148–65. 3. Edwards, J. R., Turner, P. J., Wannop, C., Withnell, E. S., Grindey, A. J. & Nairn, K. (1989). In vitro antibacterial activity of SM7338, a carbapenem antibiotic with stability to dehyrodropeptidase 1. Antimicrobial Agents and Chemotherapy 33, 215–22. 4. Working Party of the British Society for Antimicrobial Chemotherapy. (1991). A guide to sensitivity testing. Journal of Antimicrobial Chemotherapy 27, Suppl. D, 1–50. 5. Matthew, A., Harris, A. M., Marshall, M. G. & Ross, G. W. (1995). The use of analytical isoelectric focusing for detection and identification of -lactamases. Journal of General Microbiology 88, 169–78. 6. Afzal-Shah, M. & Livermore, D. M. (1998). Worldwide emergence of carbapenem-resistant Acinetobacter spp. Journal of Antimicrobial Chemotherapy 41, 576–577.

Worldwide emergence of carbapenem-resistant Acinetobacter spp. J Antimicrob Chemother 1998; 41: 576–577 Mariya Afzal-Shah and David M. Livermore* Antibiotic Reference Unit, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT, UK *Corresponding author. Tel: 44-(0)181-2004400; Fax: 44-(0)181-2007449; E-mail: [email protected]

Sir, Acinetobacter spp., especially Acinetobacter baumannii, are common opportunistic pathogens in immunocompromised patients. Historically, they were susceptible to many penicillins, cephalosporins, aminoglycosides and quinolones but, more recently, resistance to multiple antibiotics has been recognized increasingly.1 Imipenem and meropenem have retained in-vitro activities that are superior to those of other antimicrobials and, in many centres, they are the drugs of choice for patients with infections caused by Acinetobacter spp.1 Carbapenemresistant strains have been identified, but infrequently. One such strain, isolated in Scotland in 1985, produced ARI-1, a plasmid-mediated carbapenemase,2 while others, isolated in a New York medical centre, owed their resistance to target inaccessibility, rather than to drug inactivation by -lactamases.3 During the past 2 years, we have actively sought carbapenem-resistant Acinetobacter spp. worldwide and have received representative strains from Argentina, Belgium, Hong Kong, Kuwait, Singapore and Spain (Table). MICs for these isolates were determined by the agar dilution method with IsoSensitest agar and inocula of 104 cfu per spot. Many of the isolates exhibited low-level resistance, with the MICs of imipenem and meropenem ranging from 2 to 8 mg/L, compared with MICs of 0.12–0.25 mg/L for historical isolates and relative to a MIC breakpoint of 4 mg/L. Other strains were more resistant, with carbapenem MICs of 16– 128 mg/L (Table). Amongst all these isolates, cross-resistance to penicillins, cephalosporins, aminoglycosides and quinolones was virtually complete. -Lactamase extracts of the referred strains, which were prepared by sonicating cells harvested from nutrient agar cultures in 0.1 M phosphate buffer pH 7.0, were subjected to isoelectric focusing. In addition, their activities against imipenem and meropenem, at concentrations of 20 mg/L, were evaluated by a biological assay. Each antibiotic and an extract were incubated together for 1 h before being transferred to wells cut into a bioassay plate containing

576

Correspondence Table. Properties of referred and control Acinetobacter spp. isolates MICs (mg/L) Source Argentina Belgium Spain Hong Kong Kuwait Singapore Controlb

No. of isolates imipenem 2 3 15 4 1 1 1 3 40

4 64–128 64 32 64 4 4 8–16 0.12

Hydrolysis by cell-free extracts of:

meropenem

imipenem

> 2 >128 >128 >128 >128 > 16 > 8 >128 > 0.25

meropenem –

– –

–

– – – –

pIs of -lactamases 9.5, 7.5, 6.9a, 6.0 8.0, 8.2, 7.9a 8.0a, 9.2, 7.6 7.4, 7.6 7.6, 8.0a, 5.7 none detected 9.5, 8.0, 7.0a, 6.3 6.0, 6.8a variable

a

Enzyme purified and shown to have carbapenemase activity. Mode data for isolates collected at the Royal London Hospital during 1982–4. , hydrolysed; –, not hydrolysed.

b

Mueller–Hinton agar, the surface of which was inoculated with Escherichia coli NCTC 10418 as the indicator organism. The extracts of all of the strains, with the exception of one from Hong Kong, hydrolysed one or both carbapenems (Table), whereas extracts of the carbapenem-susceptible control strains (MICs 0.25 mg/L) did not do so. The isoelectric points (pIs) of the -lactamases were highly variable, with most extracts yielding multiple bands and with no single band common to all. In several cases, we isolated individual -lactamases by ion-exchange chromatography and identified those with carbapenemase activities. The biochemical characteristics of these enzymes, which also varied from isolate to isolate, will be described in a separate report. All of those that have been purified to date (indicated by a in the Table) were inhibited by tazobactam, but not by EDTA, thereby indicating that, unlike most carbapenemases, they are not zinc-dependent.4,5 This account of the worldwide emergence of carbapenem resistance amongst Acinetobacter spp. reinforces the concerns expressed by Weinbren et al.6 who described the persistence of such bacteria over a 1 year period in a UK burns unit. Others, from Argentina and South Africa, have also described carbapenem-resistant isolates.7,8 Many of these strains were susceptible only to sulbactam,3,6 which is not available as a single agent, and to polymyxin, the in-vivo efficacy of which remains unconfirmed.

Acknowledgements We are grateful to G. Claeys, J. M. Ling, P. J. Turner, H. Villar and P. West for providing the isolates.

References 1. Bergogne-Berezin, E. & Towner, K. J. (1996). Acinetobacter spp. as nosocomial pathogens: microbiological, clinical, and epidemiological features. Clinical Microbiology Reviews 9, 148–65. 2. Paton, R., Miles, R. S., Hood, J. & Amyes, S. G. B. (1993). ARI-1: -lactamase-mediated imipenem resistance in Acinetobacter baumannii. International Journal of Antimicrobial Agents 2, 81–8. 3. Urban, C., Go, E., Meyer, K. S., Mariano, N. & Rahal, J. J. (1995). Interactions of sulbactam, clavulanic acid and tazobactam with penicillin-binding proteins of imipenem-resistant and susceptible Acinetobacter baumannii. FEMS Microbiology Letters 125, 193–7. 4. Rasmussen, B. A. & Bush, K. (1997). Carbapenem-hydrolysing -lactamases. Antimicrobial Agents and Chemotherapy 41, 223–32. 5. Livermore, D. M. & Williams, J. D. (1996). -Lactams: mode of action and mechanisms of bacterial resistance. In Antibiotics in Laboratory Medicine, 4th edn (Lorian, V., Ed.), pp. 502–78. Williams & Wilkins, Baltimore, MD. 6. Weinbren, M. J., Johnson, A. P., Kaufmann, M. E. & Livermore, D. M. (1998). Acinetobacter spp. isolates with reduced susceptibilities to carbapenems in a UK burns unit. Journal of Antimicrobial Chemotherapy 41, 574–576. 7. Brown, S., Bantar, C., Young, H. K. & Amyes, S. (1997). ARI-2: a plasmid-mediated carbapenemase in Acinetobacter spp. In Programme and Abstracts of the Twentieth International Congress of Chemotherapy, Sydney, 1997. Abstract 5264, p. 216. International Society of Chemotherapy. 8. Liebowitz, L. L. D., Capper, T., Richards, G. & Koornhof, H. (1997). Low level carbapenemase production by clinical isolates of Acinetobacter spp. In Programme and Abstracts of the Twentieth International Congress of Chemotherapy, Sydney, 1997. Abstract 3108, p. 77. International Society of Chemotherapy.

577

Correspondence

Reduced susceptibility to teicoplanin in a methicillin-resistant strain of Staphylococcus aureus J Antimicrob Chemother 1998; 41: 578 Lorane Fitcha* and Alan P. Johnsonb a

Department of Microbiology, Bedford Hospital, Bedford MK42 9DJ; bAntibiotic Reference Unit, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT, UK *Tel: +44-(0)1234-792097; Fax: +44-(0)1234-792161. Sir, The glycopeptides, teicoplanin and vancomycin, have acquired increasing importance in the treatment of patients with staphylococcal infections, particularly those caused by methicillin-resistant Staphylococcus aureus (MRSA). We report here the isolation of a strain of MRSA with reduced susceptibility to teicoplanin from a patient who had received prolonged therapy with this agent. A 46 year old man with longstanding insulin-dependent diabetes mellitus was admitted to his local hospital for treatment of a foot ulcer infection, with possible underlying osteomyelitis, caused by MRSA. Before admission, he had been treated in another hospital with broadspectrum intravenous antibiotics, including vancomycin 750 mg bd, for 7 weeks, followed by treatment at home with intramuscular teicoplanin 200 mg od for 7 months. A swab of the ulcer taken at the time of the most recent admission yielded a mixed growth of organisms, with a strain of MRSA predominating. On the basis of the disc diffusion (Stokes’) method with Sensitivity Test agar (LabM) and S. aureus NCTC 6571 as the control, the isolate was susceptible to teicoplanin (8.0 mm radius with a 30 g disc, compared with 10.5 mm for the control) and vancomycin (9.0 mm and 6.5 mm radii with 30 g and 5 g discs respectively, compared with 11.0 mm and 8.0 mm radii with 30 g and 5 g discs respectively for the control). However, the patient’s poor clinical response to treatment aroused suspicion and the isolate was referred to the Antibiotic Reference Unit of the Central Public Health Laboratory for MIC determinations. The agar dilution method, with DST agar supplemented with 5% lysed horse blood, demonstrated the MIC of teicoplanin to be 16 mg/L and that of vancomycin to be 4 mg/L; the MIC breakpoint for each agent is 4 mg/L.1

Others have reported reduced susceptibility to teicoplanin in an MRSA isolate that was not detected by the disc diffusion method, but became evident only when the MIC (8 mg/L) was determined.2 Similar discrepancies have been described with coagulase-negative staphylococci3,4 and the disc diffusion method also failed to detect reduced susceptibilities to vancomycin in MRSA strains isolated in Japan5 and the USA.6 In both the present case and the original report from Japan of reduced susceptibility to vancomycin in a MRSA strain,5 the patients from whom the strains were isolated had received prolonged therapy with glycopeptides. (The patient described here also received a low dosage of teicoplanin, i.e. 200 mg od, which may have selected out a strain with reduced susceptibility.) This underscores the need for patients who require such treatment to be monitored carefully. Moreover, the in-vitro susceptibilities of staphylococci isolated from patients with serious infections who have failed to respond to therapy with vancomycin or teicoplanin should be based on MIC determinations rather than the disc diffusion method. The age of resistance to glycopeptides in strains of S. aureus, including MRSA, has arrived.

References 1. Working Party of the British Society for Antimicrobial Chemotherapy. (1991). A guide to sensitivity testing. Journal of Antimicrobial Chemotherapy 27, Suppl. D, 1–50. 2. Brunet, F., Vedel, G., Dreyfus, F., Vaxelaire, J. F., Giraud, T., Schremmer, B. et al . (1990). Failure of teicoplanin therapy in two neutropenic patients with staphylococcal septicemia who recovered after administration of vancomycin. European Journal of Clinical Microbiology and Infectious Diseases 9, 145–7. 3. Chomarat, M., Espinouse, D. & Flandrois, J.-P. (1991). Coagulase-negative staphylococci emerging during teicoplanin therapy and problems in the determination of their sensitivity. Journal of Antimicrobial Chemotherapy 27, 475–80. 4. Kenny, M. T., Mayer, G. D., Dulworth, J. K., Brackman, M. A. & Farrar, K. (1992). Evaluation of the teicoplanin broth microdilution and disk diffusion susceptibility tests and recommended interpretive criteria. Diagnostic Microbiology and Infectious Disease 15, 609–12. 5. Hiramatsu, K., Hanaki, H., Ino, T., Yabuta, K., Oguri, T. & Tenover, F. C. (1997). Methicillin-resistant Staphylococcus aureus clinical strain with reduced vancomycin susceptibility. Journal of Antimicrobial Chemotherapy 40, 135–6. 6. Centers for Disease Control. (1997). Staphylococcus aureus with reduced susceptibility to vancomycin—United States, 1997. Morbidity and Mortality Weekly Report 46, 765–6.

578