In France, laboratory surveillance of typhoid fever infections is performed by the FNRC-Salm through its network of ≈1,500 hospital and private clinical laboratories. Almost all Salmonella Typhi isolates in France are referred to the FNRC-Salm, and almost all are acquired abroad, mainly in Africa and Asia. Until 2009, Cip^DS Salmonella Typhi was monitored with the 30-μg Nal screening test. A total of 685 Salmonella Typhi isolates collected during 1997–2009 were reanalyzed to identify Nal^S–Cip^DS Salmonella Typhi isolates. The scattergram correlating the zone diameters around the 5-μg ciprofloxacin disk with those of the 30-μg Nal disk showed 4 subpopulations, which were labeled A (554 isolates), B (11 isolates), C (119 isolates), and D (1 isolate) (Figure 1). The characteristics of these populations are shown in Table 1 and Table 2. The QRDRs of gyrA, gyrB, parC, and parE genes were studied on 133 isolates selected to represent diversity in terms of year of isolation, geographic origin, and MICs. To analyze the isolate characteristics, we used the following approaches: sequencing (5), denaturing high performance liquid chromatography (4), and Luminex-based genotyping assays (12). QRDR DNA sequences were compared with those of Salmonella Typhi strain Ty2 (GenBank accession no. AE014613). In subpopulation A, 75 isolates had wild-type QRDR sequences, whereas 2 isolates had a gyrB mutation at codon 465 leading to amino acid substitution Gln to Leu. Their Nal MICs were 2 and 4 μg/mL, respectively, and those of Cip were 0.04 μg/mL and 0.08 μg/mL, respectively. Notably, both isolates were acquired in Mexico during 1998 and 2009, respectively. In subpopulation C, the lowest MIC values for Cip (0.06 μg/mL) were associated with a mutation at codon 87 of the gyrA gene, whereas MICs did not increase with the additional mutation in the parE gene. Subpopulation D consisted of 1 isolate, highly resistant to ciprofloxacin, which was acquired by a traveler in India in 2004. This isolate contained 2 NS mutations in the gyrA gene and 1 in the parC gene.

Eleven isolates of subpopulation B were categorized as susceptible to Nal by determining MICs and by using CLSI breakpoints (susceptibility, <16 μg/mL; resistance, >32 μg/mL). Of the 11 isolates, 8 (from 7 patients) had a ciprofloxacin MIC >0.125 μg/mL and were thus classified as Cip^DS isolates. We were able to review the medical records of 2 patients infected with a Nal^S–Cip^DS isolate. One patient (isolates 08-7675 and 09-1986) relapsed 15 days after completion of the treatment (oral ofloxacin, 200 mg 2×/d for 8 days) (13). The second patient (isolate 05-2556) was treated with extended-spectrum cephalosporins, and no fluoroquinolones. Regarding the resistance mechanisms the plasmid-mediated quinolone resistance–conferring genes qnr (qnrA, B, S, D), qepA, and aac(6′)-Ib-cr were not detected by PCR (5,14). The QRDRs of gyrA, parC, and parE genes were of a wild type, whereas an NS mutation was found in gyrB for all but 1 isolate. However, only the 8 isolates with mutations at codon 464 were Nal^S–Cip^DS. To assess whether these isolates were genetically related, haplotyping (4) and XbaI-pulsed-field gel electrophoresis (PFGE) subtyping (5) were performed. On the strength of the results, we concluded that the gyrB mutation was acquired independently by strains belonging to different PFGE types (Figure 2). According to a newly developed single nucleotide polymorphism assay (Y.S.), 2 of these strains belong to the current emerging H58 Asian population (4), whereas the others do not (Table 2). In our study, the Nal^S–Cip^DS isolates with gyrB mutations at codon 464 were most often non–multidrug-resistant and acquired mainly in India. Our first Nal^S–Cip^DS isolate was isolated 13 years ago, and since is rare (prevalence ≈1%.). Although Cooke et al. (10) did not characterize isolates for their resistance mechanisms, they reported that Nal^S–Cip^DS represented 11.6% (49/421) of Salmonella Typhi isolated in England, Scotland, and Wales during 1999–2003, while Lynch et al. (11) reported that such isolates were 4.6% (36/770) of Salmonella Typhi isolates identified in the United States during 1999–2006. Epidemiologic data were available for 39 isolates in the British study, 18 of which were acquired in India, 8 in Pakistan, and 4 in Bangladesh (10). The 10-fold difference in the prevalence observed between our study and that of Cooke et al. are probably related to the historical links and the subsequent population flow between the United Kingdom and the Indian subcontinent.

Nal^S–Cip^DS Salmonella Typhi isolates originating from Asia comprise ≈1% of Salmonella Typhi isolates in France but are more prevalent in the United States and the United Kingdom. The NS gyrB mutation at codon 464 was found exclusively in Nal^S–Cip^DS isolates; however, the effects of this mutation need to be formally demonstrated by site-directed mutagenesis. Furthermore, the involvement of an efflux system, such as AcrAB-TolC and OqxA, or the qnrC gene, have not been investigated and cannot be excluded.

Whatever the molecular mechanism of resistance of such strains, the main concern is detection of such isolates in clinical practice to prevent fluoroquinolone treatment failures. Consequently, the Nal^R screening test should no longer be recommended and ciprofloxacin drug MICs should be determined for all Salmonella Typhi isolates instead. There is also a clear need to reevaluate the clinical breakpoints for this pathogen.