Ciprofloxacin-Resistant Shigella sonnei among Men Who Have Sex with Men, Canada, 2010

In 2010, we observed isolates with matching pulsed-field gel electrophoresis patterns from 13 cases of ciprofloxacin-resistant Shigella sonnei in Montréal. We report on the emergence of this resistance type and a study of resistance mechanisms. The investigation suggested local transmission among men who have sex with men associated with sex venues.

In 2010, we observed isolates with matching pulsedfi eld gel electrophoresis patterns from 13 cases of ciprofl oxacin-resistant Shigella sonnei in Montréal. We report on the emergence of this resistance type and a study of resistance mechanisms. The investigation suggested local transmission among men who have sex with men associated with sex venues.
S higella spp. are enteropathogen bacteria that are transmitted person-to-person and require a low infectious inoculum (1). Fluoroquinolones are among the fi rst-choice antimicrobial drugs for treatment of Shigella spp. infections in adults (1), but resistance to these agents has been documented, primarily in Asia (2). Among men who have sex with men (MSM), Shigella spp. infection is, in most cases, sexually transmitted, and clusters are regularly reported (3)(4)(5). We investigated an outbreak of ciprofl oxacin-resistant Shigella sonnei among MSM and studied its resistance mechanisms.

The Study
Laboratories report shigellosis to the Montreal public health department (Quebéc, Canada). When a cluster is suspected, isolates are sent to the provincial laboratory to conduct pulsed-fi eld gel electrophoresis (PFGE) to identify links between patients.
In July 2010, microbiology services at the Hôpital Saint-Luc alerted the public health department to S. sonnei resistant to ciprofl oxacin and trimethoprim/ sulfamethoxazole and susceptible to ampicillin. The S. sonnei had been isolated 2 days apart from stool cultures of 2 HIV-positive MSM.
Public health offi cials sent a notice to physicians, clinics, and laboratories in Montréal to report the presence of ciprofl oxacin-resistant S. sonnei among MSM and to describe the antimicrobial treatment with ampicillin or azithromycin, procedures for case reporting, and preventive measures (6). Confi rmed cases were defi ned as infection by S. sonnei with resistance to ciprofl oxacin and trimethoprim/ sulfamethoxazole and susceptibility to ampicillin (later specifi ed as pulsovar 72). Probable cases were defi ned as infection by S. sonnei with a resistance profi le identical to that of confi rmed cases but where PFGE was not conducted.
Retrospective searching of the notifi able disease database found ciprofl oxacin-resistant S. sonnei with a different PFGE pattern that had been isolated in February 2010 from a female patient who had traveled to a country where shigellosis is highly prevalent. Hence, this case was not from this outbreak. The provincial laboratory searched their records to identify cases elsewhere in Québec. During June-October 2010, nine confi rmed cases and 4 probable cases were identifi ed in Montréal and the surrounding regions (Table 1). Most patients had an onset date from the end of June to mid-July 2010 ( Figure 1). All 13 patients were interviewed. Most patients were men (11/13; 85%) with a mean age of 40 years (range 20-65 years). All male patients were MSM, and 4 (36%) of 11 reported being HIV positive. Travel to a European country during August 2010 was mentioned by 1 MSM patient. Eight (73%) of 11 MSM patients mentioned participation in anal sex or contact during the exposure period. The use of sex venues was indicated by 4/11 MSM patients, and 3 mentioned a common sex venue. In addition, 1 other MSM patient reported that his sex partners frequented the common sex venue. This suggests that unprotected anal sex, associated with local sex venues, was the primary mode of transmission.
Two female patients (45 and 50 years of age) were reported. S. sonnei was detected in a food sample from a restaurant where 1 female patient ate during the exposure period, but the isolate did not match the outbreak PFGE pattern and was not related to any known human patients. No epidemiologic links between the female and male patients could be identifi ed.
The public health interventions included a weekly analysis of incident shigellosis infections, resistance profi les, and risk factors. Given the preponderance of infections among MSM visiting sex venues, kits of condoms, soap, and information on prevention were distributed at sex venues in August 2010 (3,6). Community-based organizations that work with MSM living with HIV/AIDS were contacted to disseminate information on preventive measures. As a potential effect, few cases were declared in September, although sporadic cases continued to appear until October 2010.
The resistance profi le investigation identifi ed 14 S. sonnei isolates from 13 patients by using commercial biochemical kit tests. Identifi cation of S. sonnei from 9 patients was confi rmed at the provincial laboratory. Antimicrobial susceptibility testing was done by agar dilution or disk diffusion method (7), Vitek 2 (bioMérieux, Marcy l'Étoile, France), or Etest (AB Biodisk, Solna, Sweden) (ampicillin, trimethoprim/sulfamethoxazole, and ciprofl oxacin) for 14 isolates and by Etest (AB Biodisk) (azithromycin, cefotaxime, and tetracycline) and with nalidixic acid (30-μg disk) for 7 or 8 isolates. The susceptibility of S. sonnei isolates to antimicrobial agents is reported in Tables 1 and 2. PFGE was done by the provincial laboratory according to international standards set by the US Centers for Disease Control and Prevention (8). The XbaI and BlnI patterns were interpreted using the standards of Tenover et al. (9). The Salmonella enterica serotype Braenderup strain (H9812) was used as the size marker in each gel (10). Band position tolerances and optimization values of 1% were used for all analyses. Similarity coeffi cient was obtained within BioNumerics (www.applied-maths. com/bionumerics/bionumerics.htm) by calculating Dice coeffi cients. Cluster analysis was done by using with the unweighted pair group method with arithmetic averages. The S. sonnei isolates from 9 patients for whom typing was done were indistinguishable for the 2 enzymes ( Figure 2). PulseNet Canada accession numbers for the isolate from our study are SSOXAI.0067 and SSOBNI.0040 for the XbaI and BlnI patterns, respectively.
For the study of the mechanisms of drug resistance, bacterial DNA was extracted using MasterPure Complete DNA Purifi cation Kit (Epicenter Biotechnologies, Madison, WI, USA). The gyrA and parC genes were analyzed by direct DNA sequencing procedures as described (11)   Biosystems, Foster City, CA, USA). The DNA sequences were converted into amino acid sequences by using the EMBOSS Transeq tool (European Molecular Biology-European Bioinformatics Institute), aligned by using ClustalW (DNAStar, Madison, WI, USA), and compared with that of the reference quinolone-susceptible strain (GenBank accession no. NC_008258). The 8 S. sonnei strains from 7 patients harbored the same nonsynonymous substitutions in comparison with the quinolone-susceptible reference strain: S83L and D87G for gyrA and S80I for parC. These amino acid substitutions have been previously associated with ciprofl oxacin resistance in Escherichia coli (11) and in S. dysenteriae, S. fl exneri, and S. boydii (2) but not in S. sonnei isolates. Blood in stools or fever was reported by 9 (69%) of 13 patients (Table 1). Of the known treatment outcomes, 2 of the 4 patients treated with oral ampicillin had a negative stool culture 48 hours and 72 hours after completion. One of the 2 patients treated with oral amoxicillin experienced a clinical and microbiologic treatment failure 48 hours after completion, but a clinical and microbiologic cure was achieved after treatment with oral azithromycin. Two other patients were treated with azithromycin and 1 other with ciprofl oxacin.
We report the suspected transmission of ciprofl oxacinresistant S. sonnei, among MSM in Montreal, Québec. Some authors suggest the antimicrobial drug treatment of all patients infected with Shigella spp. (1), but others disagree with this recommendation (14). It is essential that physicians request bacterial stool cultures when Shigella spp. enteric infection is suspected in MSM even without blood in stools or fever.