Evaluation of Enterococcus faecalis clinical isolates with 'penicillin-resistant, ampicillin-susceptible' phenotype...

It has been recently reported that ampicillin susceptibility cannot accurately predict piperacillin and imipenem susceptibilities in penicillin-resistant, ampicillin-susceptible (Pen-R, Amp-S) Enterococcus faecalis isolates, contrary to the current Clinical and Laboratory Standards Institute (CLSI) recommendations. This has important therapeutic implications. Such isolates were noted after the use of Vitek-2 Compact system AST-GP67 susceptibility cards in a Singapore general hospital and they were increasing in numbers. The primary aim of this study was to evaluate these clinical isolates against microbroth dilution (MBD) technique and other commonly used antimicrobial susceptibility test (AST) methods for penicillin and ampicillin. The secondary aim was to evaluate whether ampicillin susceptibility could indeed be a reliable surrogate marker for piperacillin and imipenem susceptibilities in E. faecalis isolates that were confirmed Pen-R, Amp-S.From 2009 to 2013, a total of 49 isolates (5%) of 983 non-duplicate E. faecalis tested by Vitek-2 displayed the 'Pen-R, Amp-S' phenotype in a general hospital in Singapore. These were tested against MBD which was the reference method, Etest and disc diffusion for penicillin and ampicillin. Susceptibilities to piperacillin and imipenem were also tested using MBD. In addition, β-lactamase production test was performed. Forty E. faecalis isolates with penicillin-susceptible, ampicillin-susceptible (Pen-S, Amp-S) phenotype were included for comparative purposes.The categorical agreement rate was 100% for all AST methods in ampicillin reporting for the 'Pen-R, Amp-S' group of E. faecalis isolates. However, a large number of isolates (46 isolates, 93.9%) fell into the major error category for penicillin testing by the Vitek-2 system. Penicillin minimum inhibitory concentrations (MICs) generated by the Vitek-2 system for the majority of these isolates were two doubling dilutions higher compared to those obtained by the reference test. The Etest method correlated well with the MBD method. Thirty-two isolates (65.3%) were in categorical agreement with the MBD method when tested by the disc diffusion method for penicillin. Only three E. faecalis isolates (6.1%) were confirmed to have the uncommon penicillin resistance phenotype, with two of them showing resistance to piperacillin and intermediate to imipenem. β-lactamase production test was negative for all isolates. Among the Pen-S, Amp-S E. faecalis isolates, the categorical agreement was 100% for penicillin and ampicillin in all the tested methods.Enterococcus faecalis with 'Pen-R, Amp-S' phenotype reported by the Vitek-2 system using AST-GP67 susceptibility cards must be confirmed with a reference test, the Etest method being a good alternative. The Vitek-2 system generated higher penicillin MIC readings compared to MBD in this study. The actual prevalence of this uncommon penicillin resistance phenotype in E. faecalis was found to be low in this institution. More studies are required to confirm the reliability of ampicillin as a surrogate marker for piperacillin and imipenem susceptibilities in these isolates.