In September 2013, a 41-year-old man residing in Barcelona visited the emergency room of the Vall d’Hebron Hospital, reporting 10 days of fever and pain in the right upper abdominal quadrant, preceded by 3 days of diarrhea. At the time of his hospital visit, the patient had been back in Spain for 4 days after a 5-month stay in rural Guatemala. Physical examination revealed a temperature of 37.8°C and unremarkable blood pressure and heart rate results. Abdominal examination showed tenderness of the right upper and lower quadrants. General biochemistry values and blood cell counts were within reference ranges, but C-reactive protein level and erythrocyte sedimentation rate were elevated. Results of initial blood and stool culture testing were negative. Because of the patient’s persistent abdominal pain and newly documented fever, computed tomographic scan of the abdomen was performed; results showed thickening of the terminal ileum, adjacent fat tissue stranding, and regional lymphadenopathy. Antimicrobial drug treatment with amoxicillin-clavulanate acid was begun. New blood cultures, performed because of the patient’s persistent fever, yielded S. enterica ser. Typhi, and the strain produced an ESBL. Subsequently, intravenous ertapenem (1 g/d) was administered for 14 days; the patient experienced complete clinical recovery, and subsequent blood and stool culture results were negative.

The isolate was identified by using the VITEK2 system (bioMérieux, Marcy l’Etoile, France). Serotyping by slide agglutination by using commercial antisera according to the Kauffmann-White scheme yielded the antigenic formula 9,12,[Vi]:d:-. Multilocus sequence typing (MLST), which was done using primers described elsewhere (8) with allele sequences and allelic profiles verified in the MLST database (http://mlst.ucc.ie/mlst/dbs/Senterica), showed that the strain belonged to sequence type (ST) 2. This ST is one of the most prevalent among S. enterica ser. Typhi; isolates of this ST have been detected in Asia, Africa, and South America (9).

Antimicrobial susceptibility to β-lactams was assessed by disk diffusion following Clinical Laboratory Standards Institute recommendations (http://www.clsi.org). Suggestive evidence of ESBL production was observed as synergy between amoxicillin/clavulanate and >1 of the following: cefotaxime, ceftazidime, aztreonam, and cefepime. In addition, MICs of β-lactams, quinolones, trimethoprim/sulfamethoxazole, chloramphenicol, and azithromycin were determined by E-test (bioMérieux) (Table). According to Clinical Laboratory Standards Institute interpretative criteria, the isolate was resistant to all β-lactams evaluated except amoxicillin/clavulanate, piperacillin/tazobactam, cefoxitin, and carbapenems and was susceptible to chloramphenicol, trimethoprim/sulfamethoxazole, nalidixic acid, ciprofloxacin, and gentamicin. The MIC of azithromycin was 4 µg/mL; according to the European Committee on Antimicrobial Susceptibility Testing (http://www.eucast.org), isolates with an MIC ≤16 μg/mL of this drug should be considered wild-type organisms that are expected to respond to treatment (10).

To screen for TEM and SHV β-lactamases, CTX-M ESBL genes, and genes encoding resistance to quinolones (qnrA, qnrB, qnrS, qepA, oqxAB, and acc(6′)-Ib-cr), we used PCR as described previously (11,12). Mutations in aac(6′)-Ib that confer quinolone resistance and in the quinolone resistance–determining regions of gyrA, gyrB, parC, and parE, were studied by sequencing (12,13). These experiments showed that the S. enterica ser. Typhi isolate possessed bla_TEM-1, bla_CTX-M-15, and acc(6′)-Ib but none of the studied quinolone resistance genes or detectable quinolone resistance–determining region mutations.

ISEcp1, IS26, and orf477 are elements that previously have been identified in the genetic environment surrounding bla_CTX-M-15. The presence of such elements was investigated by PCR mapping and sequencing in combination with bla_CTX-M-15–specific primers, as reported previously (14).These results showed that the ESBL gene was situated between a nontruncated ISEcp1 and orf477, as noted previously in other Enterobacteriaceae.

Identification and characterization of the location of bla_CTX-M-15 was carried out by conjugation, using a nalidixic acid–resistant derivative of Escherichia coli HB101 as recipient, PFGE of total DNA of donor and transconjugant digested with S1-nuclease, Southern hybridization with specific probes, and PCR-based replicon typing, as described previously (14). These studies showed that the S. enterica ser. Typhi bla_CTX-M-15 was located in an IncL/M self-transferrable plasmid of 65 kb that also carried the bla_TEM-1 and acc(6′)-Ib genes. MICs to antimicrobial agents for donor, recipient, and transconjugant are shown in the Table.

We describe an ESBL-producing S. enterica ser. Typhi strain isolated from a man in Spain who had traveled to Guatemala. It is well documented that ESBL-producing Enterobacteriaceae are spreading worldwide. CTX-M-15 is one of the most commonly identified ESBLs; its high prevalence has been driven mainly by the pandemic spread and expansion of the ST131 E. coli clonal group. Extended-spectrum cephalosporin resistance has increased during the past few years in nontyphoidal S. enterica, principally in developing countries, where such resistance appears to be endemic in some areas (15).

To our knowledge, 5 cases of S. enterica ser. Typhi resistant to β-lactams by ESBL production have been reported, all with Asian origins (Bangladesh, the Philippines, Iraq, and India) (4–7). For the isolate originating in the Philippines, SHV-12 was the enzyme responsible for ESBL resistance (7); for the other 4 isolates, a CTX-M enzyme was detected, and in 3 of those, the CTX-M-15 variant was identified (4–6). In our case, however, we found a different genetic environment of bla_CTX-M-15. Specifically, we found a nontruncated ISEcp1 upstream of bla_CTX-M-15, whereas in the previously reported Iraq-origin strain, ISEcp1 was truncated by IS26 (5). Genetic environments of bla_CTX-M-15–producing S. enterica ser. Typhi isolates from India are not clear, but according to the methods used by the authors (4), a truncated ISEcp1 or a different structure may have been involved. Our results also confirm that the plasmid harboring bla_CTX-M-15, which also carries bla_TEM-1 and acc(6′)-Ib, carries an IncL/M replicon.

In summary, we report a case of typhoid fever caused by an ESBL-producing S. enterica ser. Typhi isolate from a traveler returning to Spain from Guatemala. This case represents the acquisition of an ESBL-producing S. enterica ser. Typhi strains in the Americas. Because typhoid fever is a serious public health issue, meticulous microbiological and epidemiologic investigation of strains of this sort are necessary to prevent further spread of this disease.