Foreign Travel and Decreased Ciprofloxacin Susceptibility in Salmonella enterica Infections

To determine antimicrobial drug resistance patterns, we characterized nontyphoidal Salmonella enterica strains isolated in Liverpool, UK, January 2003 through December 2009. Decreased susceptibility to ciprofloxacin was found in 103 (20.9%) of 492 isolates. The lower susceptibility was associated with ciprofloxacin treatment failures and with particular serovars and phage types often acquired during foreign travel.

To determine antimicrobial drug resistance patterns, we characterized nontyphoidal Salmonella enterica strains isolated in Liverpool, UK, January 2003 through December 2009. Decreased susceptibility to ciprofl oxacin was found in 103 (20.9%) of 492 isolates. The lower susceptibility was associated with ciprofl oxacin treatment failures and with particular serovars and phage types often acquired during foreign travel. N ontyphoidal Salmonella enterica (NTS) isolates produce a common food-related infection that causes mild and self-limiting diarrhea and, occasionally, a potentially fatal invasive disease with bacteremia and endovascular infection (1). Antimicrobial drug therapy, often with the fl uoroquinolone ciprofl oxacin, is required for treating invasive NTS infections and severe diarrhea in immunocompromised patients. Full resistance to fl uoroquinolones is relatively uncommon for NTS infections, but decreased susceptibility to ciprofl oxacin (DCS), defi ned as MIC 0.1-1.0 μg/mL, has become common (2). Resistance to nalidixic acid is often used as a marker for DCS, although the validity of this assumption has been debated (3), and some have suggested that the ciprofl oxacin disk susceptibility zone size breakpoints should be changed to detect such strains (2).
DCS is associated with ciprofl oxacin treatment failures in patients with typhoid fever (4) and probably with invasive NTS infection, although the clinical signifi cance of DCS in NTS has not been widely explored (5)(6)(7). Clinical clues to the presence of DCS in S. enterica infections could guide early empirical prescription of antimicrobial drug therapy. Links between DCS and foreign travel have been suggested in reports from Denmark (8) and Finland (9,10). In this study, we characterized the resistance patterns of NTS strains isolated in Liverpool and explored the value of nalidixic acid-resistance testing, the clinical signifi cance of DCS, and possible links between DCS and foreign travel.

The Study
We studied all isolates of NTS detected in samples submitted to the microbiology department at the Royal Liverpool University Hospital from patients in the Merseyside area from January 2003 through December 2009. Isolates were identifi ed by standard biochemical and serologic methods. Disk susceptibility testing was performed for ampicillin, trimethoprim, ciprofl oxacin, and ceftriaxone. Nalidixic acid testing was recommended but not consistently performed. The serovar, phage type, and susceptibility of each isolate were confi rmed by the Laboratory for Gastrointestinal Pathogens, Health Protection Agency Centre for Infections, Colindale, London. Isolates were stored in glycerol broth at −70°C and later subcultured for determination of MICs to ciprofl oxacin and nalidixic acid by using the agar dilution and the Etest strip methods. The control organisms Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 25923 were used. Breakpoints were defi ned according to Clinical Laboratory Standards Institute guidelines (11). DCS was defi ned as MIC 0.1-1.0 μg/mL and resistance MIC >1 μg/mL.
Laboratory data, including serovar and antimicrobial drug susceptibility patterns, were merged with statutory notifi cation of diseases data from the Cheshire and Merseyside Health Protection Unit of the Health Protection Agency. This second database contained the travel history (validated through an enhanced questionnaire) of the patients, detailing whether they had a relevant history of foreign travel and, if so, where they had traveled. The medical records of patients with bacteremia were retrieved and reviewed to establish whether ciprofl oxacin had been used for therapy and the clinical outcome.
A total of 492 unique NTS patient isolates, including isolates from 116 inpatients, were identifi ed during the study period. Diarrheal isolates accounted for 479 of the total; 11 were from blood, 1 was from a wrist aspirate, and 1 was from a pathology specimen of the large bowel. The number and proportion of isolates resistant to ampicillin were 39 (7.9%), to trimethoprim 22 (4.5%), to ciprofl oxacin 2 (0.4%), and to ceftriaxone 6 (1.2%). Resistance to nalidixic acid was determined at isolation for 281 (57.1%) of 492 isolates and was found for 53 (18.9%). When all 492 patient isolates were tested, 103 (20.9%) were resistant to nalidixic acid, including both isolates that were cip- Forty-eight named serovars of S. enterica were found among the 492 characterized isolates. The serovars most commonly isolated from patients with nalidixic acid resistance were Enteritidis phage type 1 (PT1), Virchow, Newport, and Enteritidis PT21; the fi rst 3 of these were also commonly associated with a history of foreign travel (Table). The higher levels of DCS in serovars Enteritidis and Virchow and lower levels in serovar Typhimurium are consistent with data from Europe, as is the association with particular serovar Enteritidis phage types (12,13 (8)(9)(10). High levels of DCS or resistance to ciprofl oxacin have been observed among isolates of NTS from Spain and Thailand (14,15). This study does not include rates of travel to various destinations, so the higher numbers of S. enterica isolates with DCS in travelers to some destinations may simply refl ect travel patterns.
In a multivariate analysis, after different serovars and phage types and history of foreign travel were adjusted for, DCS was independently associated with serovars Enteritidis PT1 (OR 14.42, CI 6.41-32.43; p<0.001), Enteritidis PT21 (OR 5.81, CI 2.1-16.08; p = 0.001), Newport (OR 9.38, CI 2.8-31.38; p<0.001), and Virchow (OR 62.33, CI 7.37-526.82; p<0.001) but not with a history of foreign travel (OR 1.54, CI 0.82-2.85; p = 0.178). This fi nding suggests that the association with particular serovars and phage types is greater than any association with foreign travel and that travel is a factor only because travel can facilitate importation of these serovars.
The clinical features of the 11 patients with bacteremia are summarized in the online Appendix Table (www.cdc. gov/EID/content/17/1/123-appT.htm); 10 either were immunosuppressed or had gall bladder disease. Five blood culture isolates were resistant to nalidixic acid with DCS. Ciprofl oxacin was the initial drug choice for 3 of the patients infected with a DCS isolate, but in each instance, the drug was changed to an alternative (ceftriaxone for 2 patients, ampicillin for 1) because of an unsatisfactory clinical response. The other 2 patients initially received a cephalosporin, and outcome was acceptable. Although extended-spectrum cephalosporins are the principal alternative antimicrobial drugs for treatment of bacteremic infections, resistance is also emerging (2,13). Six of the isolates in this study were resistant to ceftriaxone, including 1 of the isolates from a patient with bacteremia, which was susceptible to ciprofl oxacin.

Conclusions
Our data show that one fi fth of NTS isolates in Liverpool demonstrated nalidixic acid resistance and that this was a good marker for DCS. The data also suggest that DCS may compromise ciprofl oxacin therapy for invasive disease caused by NTS. Infection with particular serovars and phage type, frequently associated with foreign travel, were signifi cant risk factors for infection with an S. enterica isolate with DCS, and this information can help guide initial empirical antimicrobial drug choices. Early detection of DCS is essential, but nalidixic acid-resistance testing  was not always performed at the time of isolation. Revision of ciprofl oxacin disk susceptibility breakpoints would allow such isolates to be detected more easily.