Abstract
The New Delhi metallo-β-lactamase gene (blaNDM-1) has emerged as a worldwide concern among isolates of Enterobacteriaceae. Its epidemiology is been strongly associated with travel and healthcare on the Indian Subcontinent. We report two cases of urinary tract infection with Enterobacteriaceae harboring a blaNDM-1. Both cases presented as infection in community-dwelling individuals in Australia and were associated with travel to the Indian Subcontinent. One isolate of Escherichia coli harbored the previously undescribed enzyme variant blaNDM-3, differing from blaNDM-1 by a single nonsynonymous SNP conferring a putative peptide sequence change at the 95th position (ASP→ASN). The second was an Enterobacter cloacae harboring blaNDM-1. Further genetic characterization included identification of additional β-lactamase and aminoglycoside resistance genes. Legacy antimicrobials were used for treatment. Oral therapy with nitrofurantoin was successful in one case, while combination of colistin and rifampicin was required in the second patient. Such infection, due to extensively drug-resistant pathogens, poses significant challenges in balancing the efficacy and toxicity of potential antimicrobial therapies.
Introduction
Case Histories
We report two contrasting cases of successful treatment of blaNDM-harboring Enterobacteriaceae causing urinary tract infection (UTI), illustrative of a spectrum of complexities in acquisition and management.
Case 1: Community-acquired uncomplicated cystitis
In December 2010, a 28-year-old previously well woman presented to her local doctor in Queensland, Australia, 7 days after return from travel to India. She reported symptoms of cystitis, including urinary frequency and dysuria. She had no history of UTI, renal tract abnormality, or significant medical illness. She was of Indian descent, and resided in Australia. The patient travelled in India for 2 months to visit friends and relatives. The majority of the trip was spent in metropolitan areas in the states of Gujarat and Madhya Pradesh. While in India, the patient underwent tooth extractions with a local dentist, including prescription of an unknown oral antimicrobial. She had no other healthcare contact, except a brief visit to a relative in hospital.
On presentation, she had no fever or suggestion of upper renal tract involvement. Midstream urine culture was sent to a private diagnostic laboratory. The urine yielded a pure growth of Escherichia coli (>108 organisms/L) on culture. Antimicrobial susceptibility testing by the private diagnostic laboratory indicated resistance to all agents tested (ampicillin, amoxicillin/clavulanate, piperacillin/tazobactam, cephalexin, ceftriaxone, ceftazidime, meropenem ciprofloxacin, trimethoprim, gentamicin, amikacin), with the exception of nitrofurantoin. The isolate was referred to our research laboratory for further investigation. The patient was treated with a 7-day course of nitrofurantoin with rapid resolution of all clinical symptoms.
Case 2: Complicated healthcare-associated cystitis
In January 2011, a 78-year-old man presented to a physician in Tasmania, Australia, with several weeks of mild dysuria and urinary frequency. Approximately 12 weeks earlier, he had been discharged from a hospital in Calcutta, West Bengal, India, after a 10-day stay for management of hematuria. During this admission, he underwent surgical management, including urethrotomy, transurethral resection of residual prostatic tissue, and evacuation of intravesical thrombus. After discharge from hospital, an in-dwelling catheter remained in situ for a further 14 days.
The patient was of Indian descent. He spends 5 months each year in India and the remaining 7 months of the year in Australia, residing with family in each location. He had returned to Australia ∼6 weeks before this presentation. Relevant history included a previous transurethral resection of prostate, hypertension, and cardiovascular and peripheral vascular disease.
On presentation in Australia, he was systemically well without signs of upper UTI or sepsis. A mid-stream urine culture demonstrated 1,000×106/L leukocytes and growth of both E. cloacae >108 CFU/L and Pseudomonas aeruginosa >108/L. An ultrasonography of the renal tract showed no abnormality. Both organisms were noted to be antimicrobial resistant. In particular, the E. cloacae was resistant to carbapenems. After confirmation of antimicrobial susceptibilities, the patient was admitted to hospital and treated with intravenous colistin methanesulfonate (and oral rifampicin), as no oral agents were available. Colistin methanesulfonate was commenced at a dose of 100 mg colistin base activity 12 hourly and oral rifampicin 300 mg 12 hourly. Therapy was complicated by deterioration in renal function on day 4, with serum creatinine rising from 77 to 146 μmol/L (reference range, 60–120 μmol/L).The colistin methane sulfonate dose was reduced to 75 mg colistin base activity 12 hourly in addition to increased hydration and withholding his regular angiotensin II receptor antagonist. He completed 14 days of combined therapy with resolution of all symptoms and demonstration of a negative urine culture 3 months after therapy.
Methods and Results
Phenotypic analysis
Phenotypic antimicrobial susceptibility was undertaken by disc diffusion as per CLSI methods 2 and Etest minimum inhibitory concentration (MIC) (bioMérieux, Marcy l'Etoile, France). Results for E. coli and E. cloacae are listed in Table 1. Phenotypic MBL activity was confirmed with the use of EDTA inhibition in both isolates. 3 Neither demonstrated synergy with clavulanic acid. The P. aeruginosa was susceptible to meropenem and colistin, while resistant to ceftazidime, piperacillin/tazobactam, ciprofloxacin, and gentamicin. No phenotypic MBL activity was detected.
Performed by disk as per CLSI standards or Etest, where an MIC is listed. Where no break point has been established for a given agent, only the MIC is provided.
Genotypic analysis
Antimicrobial resistance genes were investigated by PCR as described previously. 20 Genes investigated included β-lactamases and aminoglycoside resistance genes; Ambler class A ESBLs (blaCTX-M, blaTEM and blaSHV); class B metallo-β-lactamases (blaNDM, blaVIM, and blaIMP); class C extended-spectrum cephalosporinases (blaDHA and blaCMY); and the aminoglycoside resistance genes aac-6′-Ib and methylase genes. Results are in Table 1. The presence of blaNDM was detected using forward primer (5′-GGGCCGTATGAGTGATTGC-3′) and reverse primer (5′-GAAGCTGAGCACCGCATTAG-3′), producing a 758-bp product in E. coli and E. cloacae. In addition, a pair of flanking primers was used to amplify and sequence the entire gene, NDM F (5′-CTGGGTCGAGGTCAGGATAG-3′) and NDM R (5′-TCGCCCCATATTTTTGCTAC-3′). Amplicons were sequenced in forward and reverse direction using an Applied Biosystems 3730xl platform and compared to published sequences in GenBank (www.ncbi.nlm.nih.gov/genbank). P. aeruginosa did not harbor a detectable blaNDM or other MBL genes, and is not discussed further.
The E. coli blaNDM sequence contained a nonsynonymous single-nucleotide polymorphism at the 283rd position (G→A), conferring a putative peptide sequence change at the 95th position (ASP→ASN). Plasmid transformation studies reported elsewhere show similar carbapenem MICs in the parent and transformant, confirming the carbapenemase activity of this variant. 21 This variant has been designated blaNDM-3 (www.lahey.org/Studies/) and submitted to Genbank (accession: NDM-3 JQ734687). Further characterization is ongoing. The sequence obtained from E. cloacae demonstrated 100% concordance with the previously published sequences of blaNDM-1.
Discussion
We present two cases of travel-associated UTI with highly resistant organisms harboring the New Delhi metallo-β-lactamase enzyme.
The first case demonstrates importation of community-acquired blaNDM harboring E. coli. While less common than healthcare-associated infection, such acquisition has been reported in several other countries,1,7 although not Australia. Key epidemiological factors in this case include travel and antimicrobial use. 6 Travel per se is associated with UTI in females. 18 Travel to countries with a high incidence of antimicrobial resistance, in particular the Indian Subcontinent, is associated with gastrointestinal carriage of antimicrobial-resistant organisms on return. Antimicrobial use while travelling was an independent risk for acquisition and prolonged carriage of such organisms.6,17
The second case highlights the increasing frequency and complexity of the intercountry patient. This patient could be described as an informal medical tourist seeking healthcare in multiple nations for social and familial reasons. 16 While we cannot determine the exact timing of infection, the clinical picture is likely of healthcare associated if not nosocomial infection. The organism was most likely acquired in a hospital in Calcutta or after discharge while catheterized in the community.
In infection due to highly resistant pathogens, antimicrobial choice is limited, and requires a considered analysis of risk and benefit of available agents. The case reports described illustrate an unusual clinical situation, as the patients were systemically well, at most requiring oral therapy. However, the extreme resistance of the isolates dictated that the only options for therapy were limited to infrequently used and legacy antimicrobials.
Clinical studies of cystitis have observed natural resolution in most women, although delayed when compared with susceptible antimicrobial therapy. 10 In this setting of a highly resistant pathogen, however, withholding therapy must be weighed against a potential risk of upper UTI and subsequent sepsis with difficult to treat organism. In the second case, additional factors, including a concern about infection of the prostatic surgical bed, the patient's surgically altered urinary tract, and his underlying multiple medical comorbidities, also influenced the decision to treat and the duration of therapy.
As yet, there have been few descriptions in the literature and no prospective studies of oral therapy for blaNDM-harboring organisms. Virtually, all described isolates are resistant in vitro to commonly used non-β-lactams, such as trimethoprim/sulfamethoxazole and fluoroquinolones. Susceptibility testing to legacy oral antimicrobials such as nitrofurantoin, fosfomycin, and chloramphenicol is not included in all reports of blaNDM-harboring isolates. In a single study, one-third of urinary tract E. coli harboring blaNDM-1 remained nitrofurantoin susceptible. 19 Due to poor tissue penetration, this agent only has utility in cystitis, however. Fosfomycin has been used successfully in the therapy of cystitis of other resistance phenotypes, including extended-spectrum β-lactamases, and has good reported penetration into prostate tissue. 11 Susceptibility of NDM-harboring Enterobacteriaceae to fosfomycin has been reported,12–15 although with no clear denominator of isolates to determine a true rate of resistance. It may have been a suitable alternative in the first case presented, although unfortunately is not easily available in Australia. Susceptibility to other potentially useful legacy oral agents, including chloramphenicol8,9 and tetracycline, 15 is also occasionally reported, again without a clear denominator.
Conclusion
We present two cases of multidrug-resistant New Delhi metallo-β-lactamase-producing Enterobacteriaceae illustrating a spectrum of acquisition and treatment complexity. Both were treated with legacy antimicrobials, one case with significant complications. While such agents have gone out of favor due to potential side effects, tolerability issues, or absence of availability (in the case of fosfomycin in Australia), at times they may be the best option available. In the setting where an isolate can be confirmed as susceptible and the clinical scenario is appropriate, such as cystitis, we suggest consideration of oral legacy antimicrobials to treat NDM-1-harboring infections.
Footnotes
Acknowledgments
We would like to thank Dr. Sharma for her provision of some clinical details and the staff of Hobart Pathology, Microbiology Department, and QML Pathology for their identification and provision of the bacterial isolates.
Disclosure Statement
B.R., H.S., A.S., T.A., S.P., and J.L. nothing to disclose. D.P. has previously received research funding from Merck and AstraZeneca and has been a consultant to Merck, AstraZeneca, Johnson & Johnson and Leo Pharmaceuticals.