­To the Editor: Extended-spectrum β-lactamase (ESBL)–producing organisms have become a common problem for patients in hospitals and other healthcare facilities (1). Community-onset ESBL infections have recently been described in Spain, the United Kingdom, Israel, and Canada (2,3). Typically, the infections are urinary tract infections (UTIs) with CTX-M–producing Escherichia coli. These organisms may be resistant to most or all antimicrobial agents commonly used to treat UTIs, such as ciprofloxacin, trimethoprim-sulfamethoxazole, gentamicin, and ceftriaxone.

Although CTX-M–producing E. coli have previously been found in the United States (4), clinical descriptions of community-acquired ESBL-producing E. coli infections have not been reported in this country. We describe 2 healthy young women in Pennsylvania in whom UTI with CTX-M-15–producing E. coli developed.

A 25-year-old woman was seen in October 2006 at the emergency department of a hospital in Pittsburgh reporting frequent urination, chills, and bilateral back pain. She had no relevant past medical history except for previous UTIs. Results of a physical examination were unremarkable. Urinalysis showed >20 leukocytes per high-powered microscopic field. Urine culture grew >100,000 colonies/mL E. coli, resistant to trimethoprim/sulfamethoxazole and ciprofloxacin. The organism was positive by phenotypic confirmatory tests for ESBL production. Molecular characterization showed a gene encoding the CTX-M-15 β-lactamase. The patient was treated empirically with oral ciprofloxacin, 500 mg every 12 hours for 10 days and was lost to further follow-up. She has not had urine or blood cultures collected through our healthcare system in the 3 months since the time of her UTI.

A 24-year-old woman visited a student health service in Pittsburgh in September 2006 with urinary frequency, nausea, and back pain. There was mild costovertebral angle tenderness, and a clinical diagnosis of early pyelonephritis was made. The patient had no relevant past medical history except for previous UTIs (3 in the last 12 months). She was treated empirically with oral ciprofloxacin, 500 mg every 12 hours for 7 days. Urine culture grew >100,000 colonies/mL Streptococcus agalactiae and 25,000 colonies/mL Klebsiella pneumoniae. The K. pneumoniae was not an ESBL producer and was resistant to ampicillin and susceptible to ciprofloxacin, trimethoprim/sulfamethoxazole, ceftriaxone, and tobramycin.

Ten days after the initial visit, the patient returned for further assessment, and a follow-up urine sample was collected. There was no pyuria, but urine culture grew 15,000 colonies/mL E. coli. The organism was an ESBL producer that was resistant to ciprofloxacin and tobramycin and susceptible to trimethoprim/sulfamethoxazole. Given the lack of pyuria and the low intensity of symptoms, the patient was not treated with antimicrobial agents. She has not returned for follow-up, and no urine or blood cultures have been collected through our healthcare system in the 3 months since the time of her E. coli UTI. Molecular characterization of the gene encoding the ESBL indicated that it encoded CTX-M-15.

To our knowledge, these 2 cases represent the first cases of community-acquired ESBL-producing E. coli known to have occurred in the United States. In 2003, Moland et al. detected CTX-M–like ESBLs in the United States (4). The 9 E. coli isolates they described were from patients in 5 states—Virginia, Idaho, Ohio, Washington, and Texas—which suggests that CTX-M producers are geographically widespread in this country. Although some isolates were from a urinary source, these isolates were from a hospital surveillance study and the authors were unable to determine if these organisms were from community-acquired infections.

These 2 patients did not appear to have substantial clinical effects from their infections. However, the potential importance of community-acquired ESBL-producing E. coli is that UTI may be associated with bloodstream infection. Empiric antimicrobial therapy of bloodstream infection presumed to be of urinary tract origin typically comprises use of fluoroquinolones, aminoglycosides, or ceftriaxone. ESBL-producing E. coli may be resistant to all of these antimicrobial agents. In the United Kingdom, 25 of the first 108 patients with documented community-onset ESBL-producing E. coli infections died (5). Frequent occurrence of ESBL-producing E. coli in the United States would be an important public health problem and may necessitate changes in empiric antimicrobial therapy.

In Europe, many community-onset infections with ESBL producers actually appear to be healthcare associated, rather than truly community acquired. Patients with these infections may have been hospitalized in the recent past or have had relevant underlying diseases (2,6). It is also possible that the 2 infections we observed were not actually community-acquired and that these 2 women had unrecognized exposure to hospitals or healthcare facilities. However, the CTX-M producers we have previously isolated from hospitals in Pittsburgh produced CTX-M-9, not CTX-M-15 (data not shown). Our healthcare system provides coverage for most people in Pittsburgh and surrounding regions, and we found no records of previous hospitalization or chronic healthcare contact for either of the 2 persons. As far as we can ascertain, the 2 cases we describe appear to be truly community acquired, as determined by previous definitions differentiating healthcare-associated from community-acquired infections (7).

If these infections were truly community-acquired, how and why did they arise? The CTX-M-15 ESBL has been found in many countries. We do not know the travel histories of these 2 patients. Thus, the organisms may have been acquired overseas. We assessed the genetic relatedness of these 2 strains by pulsed-field gel electrophoresis but found no evidence of clonality (data not shown). Another possibility is that food was the source of infections. CTX-M-15–producing E. coli have been detected in food-producing animals (8), and we have recently found CTX-M-15–producing E. coli in chicken sold at a Pittsburgh area supermarket. We are currently conducting ongoing surveillance for community-acquired ESBL producers in our region in E. coli isolates from both humans and from foodstuffs to determine the prevalence of CTX-M producers in the United States.

Top

Author affiliations: *University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA;

Related Links

• More Letter Articles