Phenotypic and molecular characterisation of metallo-beta lactamases (MBLs) producing bacteria at a tertiary care hospital

Abstract

<h3>Introduction</h3> In 2005–2006 at the department of Microbiology, Nepean Hospital, multiresistant Gram negative rods were increasingly encountered at the sensitivity bench. They carried phenotypic markers of resistance to cefotaxime and ceftazidime and gentamicin and were not extended spectrum beta lactamases (ESBLs). Isolates originated mainly from the neonatal intensive care unit (NICU) and were identified as carrying IMP-4 gene. These organisms present a challenge to the routine microbiology laboratory because of variable phenotypic expression, the range of organisms involved and the need for rapid results. As this genetic element was new, a greater understanding of the problem in the local environment was required and this prompted us to look at improved screening for these organisms using a mixture of phenotypic and molecular methods. <h3>Aims</h3> (1) To establish rapid and low cost PCR to detect IMP-4 gene and low cost phenotypic test/s to screen for metallo-beta lactamases (MBLs). (2) To evaluate the performances of phenotypic tests. <h3>Methods</h3> A total of 276 Gram negative clinical isolates recovered from the sensitivity bench were studied. The 276 isolates comprised of 100 isolates sensitive to third generation cephalosporins and aminoglycosides, 83 stored historical isolates resistant to either/both drugs and 93 isolates from the surface swabs and urine of NICU babies where it was thought an ongoing outbreak of MBL positive bacteria was occurring. The bacterial isolates were identified and tested using VITEK and CLSI disk diffusion. Phenotypic testing for MBL detection was performed using EDTA and 2-Mercaptopropionic acid (2-MPA) as described in the literature. Real time PCR was performed for IMP-4 gene. For the isolates with discrepant phenotypic and molecular tests, a multiplex PCR was performed to detect the majority of described MBLs. <h3>Results</h3> IMP-4 gene was detected in 20.2% of the study population. EDTA has sensitivity of 95% and specificity of 89% and 2-MPA has sensitivity of 95% and specificity of 70% for MBL detection. Sixty- seven false positives were detected in 2-MPA test and 26 in EDTA test. No MBLs other than IMP-4 were detected by multiplex PCR. <h3>Conclusions</h3> Phenotypic tests are useful as they are cheap, quick and easy to perform in a routine microbiology laboratory. They are highly sensitive but the higher number of false positives among non-fermenters is a cause for concern. MBL gene detection by real time PCR is highly sensitive and specific, results can be obtained within a few hours, and they are therefore extremely useful in outbreak situations to initiate infection control measures. However, geographically prevalent MBL genes should be targeted in a multiplex PCR.