High Incidence of New Delhi Metallo-Beta-Lactamase-Producing Klebsiella pneumoniae Isolates in Sharjah,United Arab Emirates

Introduction

Carbapenem-resistant Klebsiella pneumoniae (CRKP) is increasingly considered a leading cause of healthcare-associated infections worldwide.^13,21 Among the several carbapenemase determinants, New Delhi metallo-β-lactamase (NDM) is one of the latest, initially disseminating in the Indian sub-continent, and lately, globally.^11,13,22,23 In the Middle East reports on carbapenem-resistant Enterobacteriaceae are also fast emerging. ¹ The most common enzymes encountered in the region are the OXA48-like and NDM-type carbapenemases,^5,8,14,16,18 while lately the emergence of strains carrying the VIM-4 enzymes have also been noted in the United Arab Emirates (UAE) ¹⁹ and in Kuwait. ⁹ The high percentage of expatriate population mostly coming from countries with high levels of drug resistance and receiving health care, at least occasionally, there, and the extensive movement of people in these commercial and touristic hubs make the import of resistant organisms likely. Furthermore, the fact that, guidelines for the prudent use of antibiotics and for infection control policies are still under development in some hospitals of the region create a niche in the Gulf countries particularly permissive for the emergence, acquisition and dissemination of multidrug-resistant pathogens, like CRKP. ¹² The purpose of this study was to identify carbapenem-resistant strains among the K. pneumoniae isolates encountered in a tertiary care hospital of Sharjah Emirate, detect the carbapenemase genes carried, and reveal any clonal relationship between the strains.

Materials and Methods

Results

Using the disc diffusion screening, 14 of the 202 K. pneumoniae isolates (7%) were found to be resistant to meropenem. The characteristics of the isolates along with the data collected on the patients are listed in (Table 1). Of the 202 isolates, sixty-eight percent of the strains were isolated from intensive care unit (ICU) patients, 14% from internal medicine, and rest 18% from surgery wards, respectively. Fifty-four percent of the isolates were recovered from urine, 19% from respiratory tract, 16% from exudates, 7% from blood samples, and 4% from central venous catheters. The retrospective data collection of the patients with CRKP infection revealed that only 2 of the 14 patients had travel histories during the previous 6 months outside of the UAE, but none of them traveled to India. There was no death attributed to CRKP infection during the study period.

When the MIC testing was performed after recovering the isolates from the −80°C storage, only 13 of the original 14 strains exhibited MIC values above the resistance breakpoint (≥4 mg/L), while one strain (No. 6) showed susceptibility to meropenem (MIC=0.5 mg/L). The 13 meropenem-resistant strains were further tested with phenotypic carbapenemase detection. With the exception of isolate No. 5, all 13 strains exhibited positive MHT, and all showed >5 mm increase in the inhibition zone when meropenem was combined with EDTA. No inhibition was seen with PBA.

Screening for common carbapenemase genes by PCR detected bla_NDM in all 13 meropenem resistant strains with three isolates also positive for bla_OXA-48-like gene, while no carbapenemase gene was detected in the isolate converting to meropenem susceptible after storage. Sequencing of the entire bla_NDM gene revealed the bla_NDM-1 allele in all 13 isolates.

Plasmid analysis of the 13 strains revealed 2–6 plasmids in each isolates (data not shown). Plasmids of ∼170 kb or above hybridized with the NDM-1 probe in 10 of 13 the isolates. Isolate No. 13 contained two plasmids of ∼170 and 125 kb size and both gave a signal with the NDM probe. Two isolates with identical plasmid profile (No. 4 and 5) showed hybridization signal with a plasmid of approx.140 kb size (Table 1, Fig. 1). OXA-48 hybridization localized the gene in isolate No. 3, 10, and 14 on plasmids of identical size (<7.4 kb) (data not shown).

OPEN IN VIEWER

FIG. 1.

PFGE comparison of carbapenem resistant Klebsiella pneumoniae isolates. Boxes labeled A–C represent pulsotypes with >80% similarity. MMW, male medical ward; ICU, intensive care unit; URO, urology; FMW, female medical ward; CCU, coronary care unit; MSW, male surgical ward; PFGE, pulsed field gel electrophoresis.

Molecular fingerprinting grouped the 10 NDM-1 producer and the 3 NDM-1 and OXA-48-like producer isolates into three clusters (Fig. 1 and Table 1). The largest cluster (A) contained isolates carrying the bla_NDM-1 gene mostly on plasmids of 170 kb (Fig. 1 and Table 1). In the other two clusters, isolates, although their plasmid profiles were not always identical, carried the bla_NDM-1 on plasmids of the same size (i.e., 140 and 170 kb, respectively). Strains carrying both NDM-1 and OXA-48-like genes grouped together in cluster C with isolate No. 11 which was only positive for bla_NDM-1.

Discussion

In this retrospective study, conducted over a year period, we established that 7% of K. pneumoniae strains isolated in our hospital were carbapenem-resistant among hospitalized patients. It was noteworthy, that 13 of the 14 CRKP isolates (and all which retained its CR phenotype all through the study) were carrying the bla_NDM-1 gene. As originally identified as being resistant to meropenem, we surmise that the strain negative for all the carbapenemase genes but susceptible at the time of PCR might have lost its beta-lactamase-coding plasmid during storage. Theoretically, in this strain we cannot exclude an enzyme-independent resistance mechanisms (e.g., the loss of porins), but the resistance at the time of isolations and the subsequent full susceptibility favours the first possibility. Nevertheless, all attempts to recover a nonsusceptible form of strain No. 6 failed.

Interestingly, 3 of the 13 NDM-1 positive strains were additionally carrying the bla_OXA-48-like gene, as well. The carriage of more than one carbapenemase genes, although not common, is a well known phenomenon and has already been reported from the region. ⁵ An important finding of the study was that no bla_KPC carrying organism was detected. Although these strains are of serious concerns globally, to the best of our knowledge they have not been encountered in the Arabian Peninsula.

This is the first study reporting NDM-1 producing isolates from Sharjah Emirate of the UAE. Earlier observation from Abu Dhabi reported that a plasmid of ∼50 kb size, belonging to the incX3 incompatibility group, was carrying the bla_NDM-1 gene in several Enterobacteriaceae, albeit not K. pneumoniae, isolates. ¹⁸ Similar size plasmids carrying bla_NDM-1 were not encountered in our hospital, with the majority of isolates carrying the bla_NDM-1 on much larger, ∼170 kb, plasmids. In one isolate, the NDM probe hybridized with two plasmids, a phenomenon also reported earlier by Kumarasamy and colleagues. ¹¹ Further characterization of these plasmids was beyond the scope of this present study.

The high incidence of NDM-1 producer K. pneumoniae in the study period was not attributed to the spread of a single clone, as three clusters of strains could clearly be identified by PFGE. It was also interesting to note that, with the exception of isolates No.1 and 12, all CRKP with similar pulsotype were isolated from different wards of the hospital. While both strains No.1 and 12 were from Male Medical Ward, they were isolated 11 months apart. These observations suggest the possibility of multiple introductions of CRKP strains into our hospital, or, alternatively, an extended presence with occasional inter-ward transfers.

There are several potential limitations of the study. We targeted five key carbapenemase resistance genes, only, that is, IMP, VIM, OXA, KPC, and NDM. Although these genes are recognized as major resistance determinants, ¹⁷ several other, rare carbapenemase genes found in Klebsiella peumoniae, were not looked for by our study. While all the 202 isolates were from patients with active signs and symptoms of the disease present in them, a majority of the strains (73%) were obtained from urine and respiratory secretions, especially in ICU patients which can create some uncertainties in distinguishing critical colonization and infection. Furthermore, this is a retrospective, single hospital based study and does not allow us to draw a conclusion on the prevalence of NDM-1 positive CRKP in the UAE. Nevertheless, previous reports^18,19 suggest that other Emirates of the UAE may face similar problems.

In conclusion, detection of NDM-1 positive CRKP strains in our hospital, particularly in patients with no travel history, clearly supports the earlier suggestion that the Middle East serves as a secondary reservoir for carbapenemase producer Enterobacteriaceae^{5,8,9,13,14,16,18,19} and highlights the need for action to prevent the emergence and interrupt the spread of these strains in the hospitals of the UAE.