Sepsis in neonates due to imipenem-resistant Klebsiella pneumoniae producing NDM-1 in India

Sir, Treatment of neonatal infections is becoming increasingly difficult due to multidrug-resistant organisms.1 Carbapenems are the last resort for the treatment of severe infections, however, the emergence of carbapenemases in Enterobacteriaceae has left the clinician cornered with very few options. Recently a new carbapenemase designated New Delhi metallo-β-lactamase (NDM-1) was identified in Klebsiella pneumoniae.2,3 Most reported cases of infection with NDM-1 producers have involved adult patients. This communication reports the presence of blaNDM-1 in two isolates of K. pneumoniae from neonates admitted to a neonatal intensive care unit (NICU) at a tertiary care centre in India. The first neonate was clinically septic. Imipenem-resistant K. pneumoniae (Kp-1) was isolated from an endotracheal aspirate, but blood culture was negative. The patient was started on colistin, however, the parents removed the child from the hospital against medical advice and the patient was lost to follow-up. The other neonate was born at a different hospital and was admitted to the NICU with suspected sepsis. Blood culture yielded imipenem-resistant K. pneumoniae (Kp-2). A combination of colistin and minocycline was given for a period of 10 days, after which the patient recovered and was discharged. The identity of the isolates was confirmed by an ID 32 E kit (bioMérieux, Marcy l’Étoile, France). A disc diffusion test using antibiotics (BD Diagnostics, Franklin Lakes, NJ, USA) was performed according to CLSI guidelines.4 MICs were determined using Etest (AB Biodisk, Solna, Sweden). MICs were also determined with 40 mg/L phenylalanine arginine β-naphthylamide (PAβN) (Sigma-Aldrich, St Louis, MO, USA). Testing with the cephalosporin/clavulanic acid combination disc test,4 modified Hodge test and imipenem/EDTA double-disc combination tests was also undertaken. Both isolates were resistant to a range of antibiotics, including β-lactams, quinolones and aminoglycosides. Kp-1 was susceptible only to colistin and tigecycline, but Kp-2 was susceptible to colistin, tigecycline, tetracycline, minocycline and doxycycline (Table 1). Both isolates showed phenotypic evidence of carbapenemase, metallo-β-lactamase (MBL) and extended-spectrum β-lactamase (ESBL) production. Augmentation of the zone was noted in both isolates with ceftazidime, but not cefotaxime. Investigations of production of AmpC using cefoxitin, cefoxitin with 3-aminophenylboronic acid (APB) (Sigma) and cefoxitin plus APB containing EDTA (Sigma) were negative. PFGE performed following PulseNet standardized procedures with XbaI (http://www.cdc.gov/pulsenet/protocols.htm) demonstrated that the two isolates were clonally distinct (Figure S1, available as Supplementary data at JAC Online). MIC values, phenotypic and genotypic detection of β-lactamases and detection of porins in two K. pneumoniae isolates AMP, ampicillin; AMK, amikacin; GEN, gentamicin; CIP, ciprofloxacin; CPD, cefpodoxime; CRO, ceftriaxone; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; IPM, imipenem; PIP, piperacillin; ATM, aztreonam; SXT, trimethoprim/sulfamethoxazole; TZP, piperacillin/tazobactam; NET, netilmicin; SAM, ampicillin/sulbactam; MIN, minocycline; DOX, doxycycline; TGC, tigecycline; TET, tetracycline; CST, colistin. MIC values, phenotypic and genotypic detection of β-lactamases and detection of porins in two K. pneumoniae isolates AMP, ampicillin; AMK, amikacin; GEN, gentamicin; CIP, ciprofloxacin; CPD, cefpodoxime; CRO, ceftriaxone; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; IPM, imipenem; PIP, piperacillin; ATM, aztreonam; SXT, trimethoprim/sulfamethoxazole; TZP, piperacillin/tazobactam; NET, netilmicin; SAM, ampicillin/sulbactam; MIN, minocycline; DOX, doxycycline; TGC, tigecycline; TET, tetracycline; CST, colistin. The MICs of imipenem were >32 mg/L and not affected by the presence of PAβN, indicating that efflux did not contribute to carbapenem resistance. To elucidate the mechanism of carbapenem resistance, PCR for blaKPC,5blaVIM,IMP,SPM-1,GIM-1,SIM-1 2 and blaOXA-23,OXA-24,OXA-48,OXA-58 6 was carried out as previously described. Detection of NDM-1 was performed by PCR with primers NDM-1-F (5′-GTCTGGCAGCACACTTCCTA-3′) and NDM-1-R (5′-TAGTGCTCAGTGTCGGCATC-3′) under the following conditions: denaturation for 5 min at 95°C; 35 cycles of 1 min at 95°C, 1 min at 55°C and 1 min at 72°C; and a final extension step of 10 min at 72°C. PCRs for ESBL7 and AmpC8 genes were also carried out. Sequence analysis of PCR products was performed using the Lasergene DNAStar sequence analysis software (DNAStar, Madison, WI, USA). The deduced protein sequences were analysed with the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST). Among ESBLs, blaCTX-M-15, blaSHV-1, blaTEM-1 and blaOXA-1 were detected in Kp-1, but only blaSHV-11 was found in Kp-2. Although blaSHV-11 is not an ESBL, the presence of rare types of ESBL that were not tested for may be responsible for the ESBL phenotype. ampC genes were not detected by PCR. Among the carbapenemases tested, only blaNDM-1 was detected in both isolates and confirmed by sequencing. Immunodetection of OmpA and porins were carried out using polyclonal anti-OmpA and polyclonal anti-OmpC/F antibodies separately. Isolates retained normal levels of OmpA, but one of the two major porins (OmpC or OmpF) was not detected in Kp-1 and Kp-2. We believe the lost porin to be OmpF, which is generally lost or has reduced expression in most ESBL-producing strains.9 Permeability change and expression of ESBL, as seen in our isolates, generally confers low-level resistance to carbapenems, and the elevated imipenem MICs seen with the present isolates was probably due to the presence of blaNDM-1. The presence of plasmids was determined and their approximate sizes were estimated using logarithmic plots, generated with plasmids of known molecular mass. Large plasmids approximately >212 kb and approximately 174 kb in Kp-1 and approximately >212 kb, approximately 158 kb and approximately 43 kb in Kp-2 were detected. However, attempts to show conjugal transfer of blaNDM-1 to the sodium azide-resistant recipient Escherichia coli strain J53 by a broth mating assay using Luria-Bertani agar plates containing cefotaxime (4 mg/L) and sodium azide (100 mg/L) were negative. A similar failure to transfer imipenem resistance was reported previously.3 In summary, this report describes neonatal infections due to two distinct strains of multiresistant K. pneumoniae producing NDM-1. The emergence of such strains raises important issues regarding the treatment of neonates. The focus has to be on prevention, as treatment options are becoming limited. The study was supported by institutional funding. The Collaborative Research Centre of Okayama University for Infectious Diseases in India gave a fellowship to S. R., and R. V. was a recipient of the Women Scientist Scholarship Scheme for Societal Programmes (WOS-B), Department of Science & Technology, Government of India. None to declare. We thank Dr Guillaume Arlet (Service de Bacteriologie, Hospital Tenon, Paris, France) for the control DNA of CTX-M major groups, George A. Jacoby (Lahey Clinic, Massachusetts) for providing E. coli J53 and Heinz Schwarz (Max-Planck-Institute of Tübingen, Germany) for the antibodies.