In vitro activity of temocillin against multidrug-resistant clinical isolates of Escherichia coli, Klebsiella spp. and Enterobacter spp., and evaluation of high-level temocillin resistance as a diagnostic marker for OXA-48 carbapenemase

Abstract

Sir, Temocillin is a narrow-spectrum penicillin active primarily against Enterobacteriaceae and resistant to hydrolysis by penicillinases, extended-spectrum b-lactamases (ESBLs), and AmpC enzymes. It has been shown also to retain some activity in vitro against Enterobacteriaceae with KPC-type carbapenemases, although not against bacteria with OXA-48-like non-metalloenzymes or the metalloenzymes (including IMP, NDM and VIM types). The emergence and spread of bacteria with acquired carbapenemases raises public health concerns, and there is a clear need for reliable diagnostic tests in routine bacteriology laboratories. Simple phenotypic tests with b-lactam/b-lactamase inhibitor combinations can aid the identification of isolates with KPC-type or metallo-carbapenemases, but cannot detect OXA-48 carbapenemases, for which there is currently no good inhibitor. Molecular tests could overcome this shortcoming but are not an option in many laboratories. The Carba NP acido/colorimetric phenotypic test detects carbapenem hydrolysis per se, but its sensitivity is reportedly lower for OXA-48 producers than for isolates with other carbapenemases. High-level temocillin resistance is a characteristic of many OXA-48 producers and has been included in algorithms to aid their recognition. – 8 We investigated the in vitro activity of temocillin against 2280 clinical isolates of Escherichia coli, Klebsiella spp. and Enterobacter spp. submitted to Public Health England’s Antimicrobial Resistance and Healthcare Associated Infections (AMRHAI) Reference Unit for the investigation of unusual resistance between January 2012 and April 2013. They included 1029 producers of KPC, OXA-48 or IMP, NDM and VIM carbapenemases. We also evaluated the diagnostic potential of high-level temocillin resistance (MIC ≥128 mg/L) to predict the production of an OXA-48 carbapenemase. Temocillin (Eumedica, Brussels, Belgium) was tested in the range 1–128 mg/L using agar dilution methodology; MICs were interpreted using BSAC breakpoints (susceptibility: systemic infections, MICs≤8 mg/L; urinary infections, MICs≤32 mg/L). Isolates were also tested against AMRHAI’s standard Gram-negative antibiotic panel, which includes ertapenem, imipenem (tested+EDTA to indicate likely metallo-carbapenemase producers) and meropenem. Carbapenemase production was sought using in-house PCRs and/or a commercial microarray (Check-MDR CT102; Check-Points, Wageningen, The Netherlands). The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of temocillin MIC ≥128 mg/L as an indicator of OXA-48 carbapenemase production were determined using an online tool (http://www. medcalc.org/calc/diagnostic_test.php). Temocillin MIC distributions for the 2280 isolates are shown in Table 1. These were submissions to a reference laboratory so were biased towards multiresistance, with proportions of ‘unusual’ isolates far higher than would be observed in most diagnostic laboratories. Allowing for the fact that a larger proportion of Klebsiella spp. isolates produced carbapenemases (760/1133; 67.1%) than did isolates of E. coli (206/701; 29.4%) or Enterobacter spp. (63/446; 14.1%), no major differences were observed in the MIC distributions of temocillin for the three genera. Overall, 920 (40.4%) isolates were susceptible to temocillin at the systemic breakpoint (MICs ≤8 mg/L); those considered resistant included 329/669 (49.2%) KPC producers, 671/1251 (53.6%) isolates without carbapenemases, and all 360 (100%) isolates with other types of carbapenemase (OXA-48, IMP, NDM or VIM). At the higher, urinary breakpoint, 1758 (77.1%) isolates were susceptible to temocillin (MICs ≤32 mg/L), including 628 (93.4%) KPC producers and 1098 (87.8%) isolates without carbapenemases. However, 328 (91.4%) isolates with other carbapenemases (OXA-48, IMP, NDM or VIM) remained resistant with MICs .32 mg/L. Temocillin MICs ≥128 mg/L were observed for 355 (15.6%) isolates overall, most (292; 82.3%) of which had a carbapenemase. However, 63 isolates (17.7%; 30 E. coli, 22 Klebsiella spp. and 11 Enterobacter spp.) lacked these mechanisms and the nature of their high-level temocillin resistance warrants further investigation. Many of them had complex b-lactam antibiograms—for example, 57 were also non-susceptible to ertapenem (MICs .0.5 mg/L)—consistent with ESBL/AmpC production combined with impermeability (data not shown). High-level temocillin resistance (MICs ≥128 mg/L) was recorded for 98/108 (90.7%) OXA-48 producers compared with 257/2172 (11.8%) isolates lacking OXA-48 or with 72/1920 (3.8%) isolates lacking a metallo-carbapenemase. Based on these proportions, the detection of high-level resistance was not a sufficiently robust marker when used as the sole criterion to predict the presence of an OXA-48-like carbapenemase; the PPV of the test was only 27.6% (Table 2). However, if combined with an absence of imipenem/EDTA synergy (i.e. when there was no phenotypic evidence of metallo-carbapenemase production), the PPV of high-level temocillin resistance predicting OXA-48