Typhoid fever is a systemic infection that causes bacteraemia and inflammatory destruction of the intestine and other organs. Salmonella enterica serovar Typhi (S. Typhi) is the causative agent and is transmitted from human to human via food or drinking water; therefore, hygiene and sanitary conditions mainly determine its spread [Reference Bhan, Bahl and Bhatnagar1, Reference Connor and Schwartz2]. Until the 1960s in Japan, outbreaks of typhoid fever were associated with the ingestion of contaminated food or well water. Thereafter, as a result of improved public sanitation, most of the cases have been sporadic and have come from abroad. Fluoroquinolones have been used for the treatment of typhoid fever as the first drug of choice following the emergence of multidrug-resistant (MDR; resistant to ampicillin, chloramphenicol, and trimethoprim–sulfamethoxazole) S. Typhi strains [Reference Bhan, Bahl and Bhatnagar1, Reference Asperilla, Smego and Scott3–Reference Eykyn and Williams5]. However, the frequency of isolates resistant to nalidixic acid, which exhibit reduced susceptibility to other fluoroquinolones has increased. There have been several clinical treatment failures following the administration of ciprofloxacin and other fluoroquinolones to patients with typhoid fever due to these resistant strains [Reference Rowe, Ward and Threlfall6, Reference Threlfall7]. More recently, high-level fluoroquinolone-resistant S. Typhi isolates have also been reported [Reference Gaind8]. In this study, we examined S. Typhi isolates collected over 5 years from imported cases by Vi-phage typing and determination of antimicrobial susceptibility. This surveillance revealed increases in the frequencies of drug resistance yielded first isolation of high-level fluoroquinolone-resistant S. Typhi in Japan. Isolates were also characterized by molecular typing and identification of mutations conferring quinolone resistance.

We assembled 226 clinical isolates of S. Typhi between 2001 and 2006 from patients with a history of foreign travel. All isolates were collected from regional public health centres and sent to the Department of Bacteriology, National Institute of Infectious Diseases. Isolates were phage typed by the standard technique with the phage set kindly provided by the Health Protection Agency, London, UK [Reference Bernstein and Wilson9]. The minimum inhibitory concentrations (MICs) of 15 antimicrobials for all isolates were determined using E-tests (AB Biodisk, Sweden) according to the manufacturer's instructions. The antimicrobials were ampicillin, cefotaxime, ceftriaxone, imipenem, aztreonam, kanamycin, gentamicin, tetracycline, fosfomycin, chloramphenicol, nalidixic acid, norfloxacin, ofloxacin, ciprofloxacin and trimethoprim–sulfamethoxazole. Escherichia coli ATCC25922 was included in each test as quality control. Inhibition zones were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) method [10]. Pulsed-field gel electrophoresis (PFGE) of chromosomal DNA digested with XbaI was performed as previously described using the S. enterica serovar Braenderup H9812 as the standard strain [Reference Hunter11]. DNA sequences of the quinolone resistance-determining regions (QRDRs) of the gyrA, gyrB, parC, and parE genes were determined as described previously [Reference Hirose12].

With the exception of 2005, there were around 50 cases of S. Typhi infection in Japan annually, and 226/306 patients reported during the past 6 years had a history of foreign travel before onset of typhoid fever. Of these, 212 cases (93·8%) had a history of travel to Asian countries, with South Asia deemed to be a particularly high-risk travel destination for typhoid fever (Table 1). Of the 226 isolates from the imported cases, 140 (61·9%) were from males, and 86 (38·1%) were from females. The patients ranged in age from 1 to 67 years (median age 25 years) and there was no report of death. The bacterial strains were isolated from blood (181 isolates, 80·1%), stool (39 isolates, 17·3%), bile (one isolate, 0·4%), and urine (one isolate, 0·4%). The information on sources was not available for six isolates. Twenty-one phage-type patterns were identified among all and strains of phage type E1 were the most frequent (30·5%) followed by UVS (15·9%), B1 (9·3%), E9 (8·8%), D2 (8·0%), and A (5·3%); seven strains had unique phage types (Table 1).

Table 1. Presumed region of infection and phage type of isolates in Japan from 2001 to 2006

* Country visited is not noted.

† Patients in whom travel history was not available or without recent foreign travel.

Table 2 shows that the MIC₉₀ of ampicillin, chloramphenicol, trimethoprim–sulfamethoxazole, and tetracycline for S. Typhi exceeded the highest concentration tested. This underlines the finding that agents traditionally used for first-line treatment for typhoid fever are no longer effective for this purpose, at least in Japan. The incidence of MDR strains of S. Typhi in 2001 was 21·1%, 15·8% in 2002, 13·2% in 2003, 22·2% in 2004, 28·6% in 2005, and 37·0% in 2006. Of the 52 MDR strains identified, 48 were recovered from travellers to South Asia. The most predominant phage types among MDR strains were E1 (27 strains), followed by E9 (14 strains).

Table 2. Antibiotic susceptibilities of S. Typhi from imported cases (n=226)

MIC, Minimum inhibitory concentration.

* Resistance based on CLSI breakpoint.

Fluoroquinolones and third-generation cephalosporins were the most effective against S. Typhi in vitro. However, reduced susceptibility to fluoroquinolones must currently be considered for treatment of S. Typhi infection as we have previously observed nalidixic acid-resistant S. Typhi strains, which had reduced susceptibility to fluoroquinolones [Reference Hirose13]. The incidence of nalidixic-acid resistance of S. Typhi was 28·9% in 2001, 28·9% in 2002, 39·5% in 2003, 66·7% in 2004, 47·6% in 2005, and 69·6% in 2006; of the 109 nalidixic acid-resistant strains, 96 were imported from South Asian countries with E1 (47·7%), E9 (15·6%), and UVS (14·7%) being the most common phage types. Forty-eight nalidixic acid-resistant strains also exhibited multidrug resistance.

In 2006, two high-level fluoroquinolone-resistant strains of phage type UVS were identified for the first time in Japan; these were isolated from independent travellers to the Indian subcontinent. Their common visited country was India. Both patients were successfully treated with appropriate antibiotics according to the resistant profiles of isolated S. Typhi strains. As resistance to fluoroquinolones in Enterobacteriaceae are mostly attributed to mutations in the genes encoding DNA gyrase (gyrA and gyrB) and topoisomerase IV (parC and parE) [Reference Hopkins, Davies and Threlfall14], the nucleotide sequences of the QRDRs were determined [Reference Hirose12]. Both strains, showed three identical mutations, two within gyrA at codons 83 and 87, and one mutation in parC at codon 80. For gyrA, TCC of Ser⁸³ codon and GAC of Asp⁸⁷ changed to TTC (Phe) and AAC (Asn), respectively. For parC, the mutation was a change of AGC (Ser) to ATC (Ile). No alteration in the QRDRs of gyrB and parE was found.

PFGE genotyping exhibited unique profiles of two high-level fluoroquinolone-resistant strains of serovar Typhi, but the profiles were quite similar to each other (data not shown). Both the fluoroquinolone-resistant strains gave highly related DNA profiles suggesting clonal identity.

Typhoid fever has become a predominantly travel-associated disease in developed countries, and the emergence of strains resistant to and with reduced susceptibility for fluoroquinolones is a matter of grave concern. From 2001 to 2006 in Japan, 73·9% of cases were related to foreign travel and 48·2% of isolates showed reduced susceptibility to fluoroquinolones. Further, we demonstrated the presence of S. Typhi resistant to fluoroquinolones, including norfloxacin, ofloxacin, and ciprofloxacin. All isolates investigated here were susceptible to third-generation cephalosporins (cefotaxime and ceftriaxone), which might indicate that these antibiotics could still provide an appropriate therapy for typhoid fever. Indeed, the use of ceftriaxone is currently recommended as the first-line therapy [Reference Connor and Schwartz2, Reference Frenck15]; however, we should be cautious of the real possibility that strains resistant to third-generation cephalosporins will emerge in the future.