Community-Onset Genitourinary Tract Infection due to CTX-M-15–Producing Escherichia coli among Travelers to the Indian Subcontinent in New Zealand

Extended-spectrum β-lactamase-producing Escherichia coli (ESBL-EC) is widely recognized as a cause of nosocomial infection and is an emerging cause of community-onset urinary tract infection (UTI), particularly in India [1]. UTIs are the most common bacterial infections affecting adults in the community and are usually readily managed in outpatient settings by general practitioners. The emergence of ESBL-EC as a cause of community-onset UTI has profound therapeutic implications, because many ESBL-EC isolates are resistant to all available oral antimicrobial agents.

Although ESBL-EC has been increasingly observed to cause nosocomial infections in New Zealand, it has remained a highly unusual cause of community-onset UTI, particularly in patients without previous exposure to health care facilities [2]. However, during the period 2004–2006, one of the authors reviewed 3 cases in female patients from the community who presented with UTI due to ESBL-EC. Remarkably, all 3 women had a history of recent emigration from or travel to the Indian subcontinent. Furthermore, all had previously been healthy, with no history of hospital admission or UTI. They were referred to Auckland City Hospital (Auckland, New Zealand), a 900-bed teaching hospital that serves a population of ∼420,000. The referral was made in each case because of the unavailability or failure of oral antimicrobial agents to treat these infections in the community. This observation prompted an investigation at Auckland City Hospital to examine whether community-onset UTI due to ESBL-EC was associated with recent overseas travel or immigration.

Methods. Patients with community-onset UTI due to ESBL-EC were defined as those who were referred from the community to the hospital for assessment and had symptomatic genitourinary tract infection, with isolation of ESBL-EC from a urine specimen within 48 h of referral. A database of all patients with newly isolated ESBL-producing organisms is maintained by the Auckland City Hospital microbiology laboratory. Using this record, we reviewed the clinical notes for all patients who had ESBL-EC isolated from a urine specimen obtained within 48 h after hospital admission during a 3-year period (1 January 2004–31 December 2006). During this time, it was routine laboratory practice to screen all clinical isolates of E. coli for ESBL production.

PFGE typing was performed on XbaI-restricted genomic DNA of selected ESBL-EC isolates. PFGE profiles were analyzed with BioNumerics software, version 4.61 (Applied Maths). ESBL genes were identified using PCR and sequencing, with use of methods described elsewhere [3, 4].

Results. During the review period, 5936 isolates of E. coli were cultured from urine specimens, and 66 (1.1%) were confirmed to be ESBL-EC. Of 28 patients with community-onset ESBL-EC bacteriuria, 27 (96.4%) had symptomatic UTI. Of these 27 patients, 14 (51.9%) had a history of admission to a hospital or a residential care facility in New Zealand during the previous 6 months. All of the remaining 13 patients (48.1%) had a history of overseas travel or immigration to New Zealand. Two travelers from New Zealand had a prolonged hospital visit while overseas (1 in China and 1 in the United States), and a third patient was a visitor to New Zealand from Europe with no documented history of hospital admission. The largest group was formed by the remaining 10 patients, all of whom had a history of emigration from or recent travel to the Indian subcontinent (9 had traveled to or emigrated from India and 1 had emigrated from Bangladesh), without a history of previous hospital admission.

Of the 10 patients who had contact with the Indian subcontinent, 9 had traveled within 6 months before presentation. Five patients had immigrated permanently to New Zealand, 2 had recently returned after visiting family in India, 2 were temporary visitors to New Zealand, and 1 was a New Zealand resident who had visited India (table 1).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Demographic and clinical characteristics of patients with community-onset urinary tract infection (UTI) due to extended-spectrum β-lactamase producing Escherichia coli (ESBL) and Indian subcontinent contact.

None of the 10 patients had been previously admitted to a hospital, either in New Zealand or overseas, and none resided in residential care facilities. Furthermore, 7 patients had no hospital exposure within 12 months before presentation. Of the remaining 3 patients, 1 had attended a single hospital outpatient clinic in New Zealand within the preceding 12 months, 1 had attended multiple outpatient clinics in India for blood transfusions, and 1 was a health care worker in a New Zealand hospital (table 1).

Eight of the 10 patients were treated with carbapenems and/or aminoglycosides. Four patients with cystitis and 1 patient with epididymo-orchitis had received ineffective courses of treatment in the community. In 2 patients, treatment failed despite the in vitro susceptibility of their isolates to the agents used (nitrofurantoin and doxycycline). Three patients experienced relapse of infection after initial carbapenem treatment at Auckland City Hospital and required additional prolonged intravenous treatment. Isolates from all 10 patients were resistant to ciprofloxacin, and 8 of 10 isolates and 9 of 10 isolates were resistant to gentamicin and trimethoprim-sulfamethoxazole, respectively. Two of the isolates were resistant to all oral agents tested, leaving no options available for oral therapy. None of the isolates demonstrated cefoxitin resistance in addition to the ESBL phenotype, and all isolates were susceptible to meropenem, ertapenem, and amikacin.

PFGE typing and identification of the ESBL gene was performed on 9 of the 10 isolates (table 1). All isolates had distinct PFGE profiles and shared 64%–90% similarity with each other. Eight of the 9 isolates carried the bla_CTX-M-15 gene. The remaining isolate carried the bla_SHV-12 ESBL gene. All 9 isolates had remarkably similar antibiotic susceptibility patterns, suggesting that the 9 different strains carried plasmids with a similar range of resistance determinants.

Discussion. It is striking that, in the absence of prior hospitalization as a risk factor, 10 of 11 patients with community-onset UTI due to ESBL-EC had a history of travel to or from the Indian subcontinent. During the 3-year review period, Indian nationals comprised only 4% of new permanent immigrants to New Zealand [5], but 5 (38%) of 13 individuals with travel-related community-onset UTI due to ESBL-EC presenting to Auckland City Hospital were from this group.

The bla_CTX-M-15 ESBL gene was detected in 8 of the 9 isolates tested. This gene was first described in India in 2000 [6] and has since been observed worldwide, in association with well-described outbreaks in the United Kingdom, France, and Canada [7–9]. A recently published survey of ESBL-producing clinical isolates collected from hospitals throughout India demonstrated the presence of bla_CTX-M-15 in 73% of the ESBL-EC studied, and this gene was the only bla_CTX-M subtype identified in any of the 115 isolates [10]. In addition, in a survey performed in North India, 52% of patients who presented to the hospital with a symptomatic infection due to ESBL-EC had no history of hospital admissions at any time [11].

These reports demonstrate that strains of bla_CTX-M-15 E. coli are widespread in India and are commonly acquired within the community. This is supported by a study performed in Mumbai in 2004 that found that at least 7% of healthy executives with no history of hospitalization had ESBL-producing organisms isolated from stool specimens [12]. A high prevalence of gastrointestinal carriage of ESBL-EC in India presumably explains the association that we have observed between community-onset UTI due to ESBL-EC and contact with the Indian subcontinent. A number of factors may possibly contribute to this high rate of carriage, including “over the counter” availability of antimicrobials, poor sanitation, and agricultural antibiotic use. Two of our patients were tourists to India who had recently returned, which suggests that even brief exposure to a community with a high prevalence of infection may allow colonization and subsequent infection with ESBL-EC. Furthermore, the lack of predisposing risk factors for UTI in these patients suggests that many circulating ESBL-EC strains in India may have a high propensity for causing UTI. Recent studies demonstrated that geographically diverse isolates of CTX-M-15-producing E. coli, with as little as 68% similarity by PFGE, may share highly similar genotypic virulence profiles and a common sequence type with use of multilocus sequence typing [13, 14]. Therefore, although PFGE of our isolates demonstrated 9 distinct profiles sharing as little as 64% similarity, it remains possible that these isolates have a common underlying virulence gene profile and multilocus sequence type.

Community-onset UTI due to ESBL-EC in Auckland, New Zealand, is associated with travel to or emigration from the Indian subcontinent. Travel to areas with a high prevalence of ESBL-EC gastrointestinal tract colonization in the community is a potential risk factor for UTI due to ESBL-EC in areas with a low prevalence of such colonization, even in the absence of prior hospitalization. Because 3 of the 10 patients who were otherwise healthy experienced relapse of infection after 7- to 10-day treatment courses with a carbapenem, there is added concern regarding the therapeutic and economic implications of these observations.

• Received March 10, 2008.
• Accepted May 7, 2008.

• © 2008 by the Infectious Diseases Society of America