Ciprofloxacin-Resistant Salmonella enterica Serotype Typhi, United States, 1999–2008

We report 9 ciprofloxacin-resistant Salmonella enterica serotype Typhi isolates submitted to the US National Antimicrobial Resistance Monitoring System during 1999–2008. The first 2 had indistinguishable pulsed-field gel electrophoresis patterns and identical gyrA and parC mutations. Eight of the 9 patients had traveled to India within 30 days before illness onset.

T yphoid fever, caused by Salmonella enterica serotype Typhi, is a systemic bacterial illness that has been rare in the United States in the era of modern sanitation (1,2). However, typhoid fever remains common in many developing countries. In the United States, 72%-81% of patients with typhoid fever report international travel in the month before illness onset (1,(3)(4)(5). Highest risk has been associated with travel to southern Asia (1)(2)(3)(4)(5).

The Cases
State public health laboratories receive Salmonella isolates from clinical diagnostic laboratories as part of routine surveillance. State and local health department offi cials report demographic, clinical, and travel information about laboratory-confi rmed typhoid fever on a standard form to the Centers for Disease Control and Prevention (CDC, Atlanta, GA, USA). Participating states began submitting all S. enterica serotype Typhi isolates to NARMS in 1999; since 2003, all state public health laboratories have participated. Isolates were tested for susceptibility by using broth microdilution (Sensititre: Trek Diagnostics, Westlake, OH, USA). MICs were determined for 15 antimicrobial agents and interpreted by using Clinical and Laboratory Standards Institute (CLSI) criteria when available (Table  1) (7,13). For ciprofl oxacin-resistant isolates, subtyping by pulsed-fi eld gel electrophoresis (PFGE) was performed by using the protocol established by the National Molecular Subtyping Network for Foodborne Disease Surveillance (PulseNet) (14). PFGE pattern similarity was assessed by cluster analysis (Dice, UPGMA [unweighted pair group method using arithmetic averages]) and band-matching applications of BioNumerics software (Applied Maths, Sint-Martens-Latem, Belgium) and confi rmed by visual comparison (Figure). For ciprofl oxacin-resistant isolates detected for 1999-2005, sequencing of the quinolone resistance-determining region (QRDR; defi ned as amino acids 67-106 for gyrA) was performed according to the methods described by Crump et al. (6), and additional patient information (e.g., antimicrobial drug treatment) was requested by using a questionnaire with institutional review board approval.
In 2003, a 1-year-old girl had onset of fever 1 day before arriving in the United States from India. A blood specimen collected 3 days after fever onset yielded S. enterica serotype Typhi. Diarrhea or vomiting at time of specimen collection was not reported. Information about antimicrobial drug treatment was not available. The child was hospitalized for 14 days.
In 2005, a 2-year-old girl had onset of diarrhea, which was treated with ofl oxacin, 2 days before she arrived in the United States from India. Seven days later, she continued to have diarrhea, and fever, vomiting, and abdominal cramps developed. She was hospitalized and treated with antimicrobial agents, including ciprofl oxacin. Blood and fecal specimens collected 3 weeks after illness onset yielded S. enterica serotype Typhi. The patient was discharged after 14 days of hospitalization. She had lived in India for 6 months before traveling to the United States.
The S. enterica serotype Typhi isolates were resistant to ciprofl oxacin (Tables 1, 2) and had indistinguishable PFGE patterns when restriction enzymes XbaI and BlnI were used: PulseNet-designated XbaI pattern JPPX01.0026 and BlnI pattern JPPA26.0110 (Table 2; Figure). QRDR sequencing showed gyrA mutations resulting in a serine to tyrosine substitution at codon 83 and an aspartic acid to asparagine substitution at codon 87, and a parC mutation conferring a serine to isoleucine substitution at codon 80.
Seven (0.6%) ciprofl oxacin-resistant infections were detected among patients from whom 1,131 S. enterica serotype Typhi isolates were submitted during 2006-2008 ( Table 2). The 7 cases occurred in 2006 and 2007. Patients were a median of 22 years of age (range 5-48 years); 5 (71%) were male. All 6 patients with known travel histories reported travel to India in the 30 days before illness onset. In addition to XbaI JPPX01.0026 and BlnI JPPA26.0110, 3 different XbaI and BlnI pattern combinations were detected in the 7 isolates (Table 2; Figure).
Six additional patients, who were detected in 2006 and 2007, also reported travel to India. Travel to the Indian subcontinent has been associated with nalidixic acidresistant S. enterica serotype Typhi infection; however, ciprofl oxacin-resistant infections are rarely reported by using current CLSI criteria (4,5,11). Other resistance patterns were fi rst described in southern Asia, where the incidence of typhoid fever is high and antimicrobial agents are widely available without prescription, providing the opportunity for the development and selection of resistant strains (8). Other than reports by 8 patients of travel to India, we have no information about possible shared exposures, such as specifi c locations visited, sources of food or water, or contact with carriers of S. enterica serotype Typhi.
However, the indistinguishable PFGE XbaI and BlnI patterns and identical gyrA and parC mutations of isolates from the fi rst 2 patients suggest that, although typhoid fever occurred nearly 2 years apart, the same ciprofl oxacinresistant strain is likely to have been involved. After 2005, different XbaI and BlnI patterns have been identifi ed in ciprofl oxacin-resistant isolates, indicating independent selection of ciprofl oxacin resistance in different strains.
The gyrA and parC mutations of isolates from the fi rst 2 patients were reported in ciprofl oxacin-resistant S. enterica serotype Typhi in India (11). The 2 gyrA mutations are well characterized and known to be associated with quinolone resistance; 2 point mutations in gyrA and 1 in parC confer fl uoroquinolone resistance (8,(10)(11)(12). Further studies, including characterization of other resistance mechanisms,  are needed to track the evolution of fl uoroquinoloneresistant S. enterica serotype Typhi. Although the ciprofl oxacin resistance we detected using current CLSI criteria is rare in S. enterica serotype Typhi, nalidixic acid resistance, which correlates with decreased susceptibility to ciprofl oxacin, has increased (7). Clinicians should be aware that infection with Salmonella spp. with decreased susceptibility to ciprofl oxacin may not respond satisfactorily to this agent (6,8,9,13,15). In addition, identifi cation of ciprofl oxacin-resistant cases has been increasing. In the presence of quinolone resistance, third-generation cephalosporins, such as ceftriaxone, can be used (2,6,8,15). Recent clinical trials suggest that azithromycin might be useful for treating uncomplicated typhoid fever (2,8,9,15). Recommendations for empiric treatment of typhoid fever in the United States are best developed by using information about antimicrobial drug resistance trends in isolates from countries where the infection was acquired.