Analysis of Salmonella spp. with resistance to extended-spectrum cephalosporins and fluoroquinolones isolated in North America and Latin America: report from the SENTRY Antimicrobial Surveillance Program (1997–2004)

Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21 www.elsevier.com/locate/diagmicrobio

Analysis of Salmonella spp. with resistance to extended-spectrum cephalosporins and fluoroquinolones isolated in North America and Latin America: report from the SENTRY Antimicrobial Surveillance Program (1997–2004) Douglas J. Biedenbacha,T, Mark Tolemanb, Timothy R. Walshb, Ronald N. Jonesa a

JMI Laboratories, 345 Beaver Kreek Centre, Suite A, North Liberty, IA 52317, USA BCARE, Department of Pathology and Microbiology, School of Medicine University of Bristol, Bristol, BS8 1TD UK Received 16 June 2005; accepted 16 June 2005

b

Abstract Emerging antimicrobial-resistant Salmonella spp. requires increased efforts to appropriately test susceptibility. The SENTRY Antimicrobial Surveillance Program monitored Salmonella spp. and detected nalidixic acid-resistant strains with elevated fluoroquinolone minimum inhibitory concentration (MIC) results and strains with extended-spectrum h-lactamase (ESBL) bphenotypesQ over the last 8 years. A total of 786 stool and bloodstream isolates from North American and Latin American medical centers (2001–2003) were tested by reference broth microdilution methods. Genetic analysis was used to further characterize the resistance mechanisms. Twenty-one sites forwarded 89 (11.3%) nalidixic acid-resistant (MIC, z 32 Ag/mL) strains. Nineteen of these isolates were studied to determine mutations in the quinolone resistance-determining region (QRDR). Among the nalidixic acid-resistant Salmonella spp. isolates, fluoroquinolone MIC values were also elevated (8- to 32-fold) compared with bwild-typeQ strains. Ciprofloxacin and gatifloxacin (MIC90, 0.5 Ag/mL) were more potent than levofloxacin and garenoxacin (1 Ag/mL) against nalidixic acid-resistant strains. Single gyrA mutations were responsible for elevated fluoroquinolone MIC values and included D87Y (5), S83F (7), D87N (5), and S83Y (2). During 2001, 9 sites contributed 11 (2.9%) strains that met ESBL screening criteria ( z 2 Ag/mL) for aztreonam or ceftazidime or ceftriaxone. ESBL confirmation was evaluated by Etest (AB BIODISK, Solna, Sweden) ESBL strips and the enzymes were characterized by polymerase chain reaction and gene sequencing. The ESBL phenotype isolates had the following MIC patterns: ceftazidime ( z 16 Ag/mL), aztreonam (4 to N 16 Ag/mL), and ceftriaxone (8–32 Ag/mL). All strains were susceptible to cefepime, carbapenems, gentamicin, and fluoroquinolones. No strains were inhibited by clavulanic acid consistent with all isolates producing the identified CMY-2, AmpC-like enzyme. Fluoroquinolones may be compromised among isolates with QRDR mutations detected using nalidixic acid as a screening agent. Salmonella spp. with ESBL phenotypes were likely to harbor CMY-2 (not an ESBL) and remain susceptible to cefepime, carbapenems, and fluoroquinolones, which can be used for serious invasive Salmonella spp. infections. Compared with the stool culture isolates, the blood culture isolates had higher QRDR mutations, but remained susceptible to the fluoroquinolones. The blood culture isolates were more susceptible to penicillins (ampicillin and ticarcillin) and not significantly different for ceftriaxone or trimethoprim/sulfamethoxazole susceptibility patterns. No QRDR trends over time were detected in North America, but increased resistance was observed in Latin America. D 2006 Elsevier Inc. All rights reserved. Keywords: Salmonella spp.; Fluoroquinolone; CMY-2

1. Introduction Millions of illnesses are said to be caused by Salmonella spp., which produce diarrheal symptoms and are capable of

T Corresponding author. Tel.: +1-319-665-3370; fax: +1-319-6653371. E-mail address: [email protected] (D.J. Biedenbach). 0732-8893/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2005.06.013

serious invasive infections that can have high mortality rates, particularly in immunocompromised patients such as the very young and the elderly. The SENTRY Antimicrobial Surveillance Program reported that Salmonella spp. were ranked as the 13th most common bloodstream infection pathogen during 1997–2001 with occurrence rates from 0.4% in North America to 2.3% in the Asia-Pacific region (Stephen et al., 2003). Although most Salmonella spp.

14

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

isolates remain susceptible to many antimicrobial classes including most h-lactams, fluoroquinolones, and aminoglycosides, reports have documented resistance mechanisms in these species, including the dissemination of S. enterica serovar Typhimurium definitive phage type 104 (DT104). This multidrug-resistant Salmonella spp. was first detected in the mid-1980s and has been consistently isolated from humans since the mid-1990s in the United States and Europe (Cormican et al., 1998; Ribot et al., 1985). The resistance genes associated with DT104 isolates are located on integrons, which are the likely elements responsible for the increase in this resistance phenotype over a relatively short period (2 decades). More recently, Salmonella spp. isolates with extended-spectrum h-lactamase (ESBL) phenotypes and the development of fluoroquinolone resistance have become important clinical concerns that could further compromise commonly prescribed empiric treatments with ceftriaxone and fluoroquinolones such as ciprofloxacin. This is particularly troublesome when treating at-risk patients including children who are more prone to sequela related to infections caused by Salmonella spp. The production of plasmid-mediated ESBL and other h-lactamase enzymes in Salmonella spp. is also an epidemiologic problem because resistance can be transferred between Enterobacteriaceae species including Salmonella spp. and clonal dissemination could cause large community outbreaks and the spread of resistant pathogens to the hospital setting. Fluoroquinolone-resistant Salmonella spp. and strains with reduced susceptibility to fluoroquinolones (nalidixic acid-resistant) are being reported more frequently, particularly in Europe, Asia, and India (Hakanen et al., 2001; Molbak et al., 2002; Threlfall et al., 1999, 2001). Clinical failure has been documented when using ciprofloxacin to treat salmonellosis caused by these strains especially with short courses or low-dose regimens (Casin et al., 2003; Kristiansen et al., 2003; Piddock et al., 1993; Rupali et al., 2004). Isolation of Salmonella spp. with ESBL enzymes has occurred in diverse geographic areas of the world. Recent reports have documented several regions that have discovered true ESBL enzymes, which include Europe (CTX-M-3, CTX-M-5, CTX-M-9, SHV-5, SHV-12), North America (SHV-2a), Africa (TEM-3, TEM-63, or TEM-131), Turkey (PER-1), and the Asia-Pacific (SHV-9, TEM-52) in both typhoidal and nontyphoidal species (AitMhand et al., 2002; Baraniak et al., 2002; Edelstein et al., 2004; Hanson et al., 2002; Kruger et al., 2004; Lee et al., 2003; Miriagou et al., 2002; Mulvey et al., 2003; Simarro et al., 2000; Tassios et al., 1999; Vahaboglu et al., 1996; Villa et al., 2002). However, Salmonella spp. strains with an ESBL bphenotypeQ often express CMY-2, an AmpC-like enzyme, which has been reported in Europe, the Asia-Pacific, and North America (Carattoli et al., 2002; Navarro et al., 2001; Yan et al., 2003). Salmonella spp. strains with CMY-2 have been isolated from both humans and animals, for example, ubiquitous in Enterobacteriaceae, and this mobilized

enzyme appears to be easily transferred between species and could be acquired from animal sources or multiple other reservoirs (Fey et al., 2000; Pitout et al., 2003; Rankin et al., 2002). The widespread use of fluoroquinolones has led to increased rates of resistance among enteric bacterial species capable of causing human disease including Salmonella spp. (Kristiansen et al., 2003). Although the development of high-level fluoroquinolone resistance in Salmonella spp. remains very rare, strains with elevated minimum inhibitory concentration (MIC) values have been associated with point mutations in the quinolone resistance-determining region (QRDR) and/or active efflux (Giraud et al., 2000; Guerra et al., 2003; Ling et al., 2003; Nakaya et al., 2003). It has been demonstrated that single gyrA mutations mediate resistance to nalidixic acid and increase the MIC to broader-spectrum fluoroquinolones and a further mutation in parC leads to strains highly resistant to fluoroquinolones (Heurtin-Le Corre et al., 1999; Ling et al., 2003; Nakaya et al., 2003; Renuka et al., 2004). It has also been observed that fluoroquinolone treatment for strains with elevated fluoroquinolone MIC values, although susceptible using current Clinical and Laboratory Standards Institute (CLSI, formerly the National Committee for Clinical Laboratory Standards [NCCLS]) breakpoint criteria, have led elevated rates of to clinical failure (CLSI, 2005; Kristiansen et al., 2003; Piddock et al., 1993). This has recently led to discussions regarding whether the current breakpoint criteria used for Salmonella spp. and fluoroquinolones remain appropriate (Aarestrup et al., 2003; Crump et al., 2003). The CLSI recently suggested that nalidixic acid could be used to test for reduced fluoroquinolone susceptibility, which has been studied and reported by several investigators (Albayrak et al., 2004; CLSI, 2005; Hakanen et al., 1999; Oteo et al., 2000). The SENTRY Program has documented the susceptibility profile of Salmonella spp. isolates, particularly invasive infections (bloodstream infections) since 1997. During enhanced surveillance of gastrointestinal pathogens in 2001, the number of Salmonella spp. (both invasive and noninvasive) available for study was increased significantly. This study will determine the frequency of Salmonella spp. isolates with resistance to extended-spectrum cephalosporins in North America and Latin America during the surveillance years. During that period, strains with an ESBL phenotype (CLSI, 2005) were analyzed genetically to determine the responsible resistance mechanism(s) and document the susceptibility patterns of these isolates. An analysis of the fluoroquinolone susceptibility among nalidixic acid-susceptible and nalidixic acid-resistant Salmonella spp. strains was also completed using a large collection of isolates cultured during 2001–2003. These results also provide an expanded understanding of the relationship between nalidixic acid susceptibility and the appropriate ciprofloxacin MIC breakpoints used to define Salmonella spp. that may not optimally respond to contemporary

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

15

fluoroquinolone therapy. A subset of isolates cultured with resistance to nalidixic acid and elevated ciprofloxacin MIC values above the bwild-typeQ MIC distribution had the QRDR region characterized for mutations by polymerase chain reaction (PCR) and sequencing.

Eleven strains had MIC results that were z 2 Ag/mL for ceftazidime or ceftriaxone or aztreonam, consistent with CLSI (2005) screening criteria for ESBL production. These isolates were further tested using ceftazidime and ceftriaxone Etest ESBL strips (AB BIODISK, Solna, Sweden) with and without clavulanic acid as a confirmatory test.

2. Materials and methods

2.3. Molecular methods

2.1. Bacterial strains

Nineteen strains isolated in 2001 with nalidixic acid and ciprofloxacin MIC values of z 32 Ag/mL and z 0.12 Ag/mL, respectively, were genetically analyzed to determine mutations in gyrA, gyrB, parC, and parE by PCR and sequencing. The QRDR genes parC, parE, gyrA, and gyrB were amplified by PCR using primers and cycling conditions previously described (Stephen et al., 2003). Sequencing was performed in both strands by the dideoxy-chain termination method with a Perkin Elmer Biosystems 377 DNA sequencer and sequence analysis performed using the Lasergene software package (DNAStar, Madison, WI). All strains with ESBL phenotypes were genetically analyzed to determine the enzyme(s) responsible for the resistance. All genes encoding h-lactamases also were aligned and generic primers designed for PCR analysis. In most cases, PCR annealing was carried out at 48 8C, but gradient PCR was also used over a range of 38– 60 8C. Sequencing of enzymes was carried out using DuPont Automated systems and analyzed using DNAStar. Molecular epidemiology analysis was performed by pulsed-field gel electrophoresis (PFGE) of SpeI digests using the GenePath System (Bio-Rad Laboratories, Hercules, CA). The PFGE patterns were compared by visual inspection with identical DNA banding patterns considered clonal. Potential clonally related and unrelated isolates differed by 1–3 bands and N3 bands, respectively.

Participating SENTRY Program sites in North America and Latin America were requested to forward 25 consecutive isolates of species considered producers of diarrheal disease in 2001 and 50 consecutive isolates in 2003 (Latin America only). Additional Salmonella spp. isolates were collected through routine surveillance of blood cultures, and others, which yielded 786 strains during 2001–2003. Salmonella spp. were collected from blood cultures (334 strains), stool cultures (428 strains), and other or unknown sites of infection (24 strains). Strains were identified locally and confirmed at the reference center (JMI Laboratories, North Liberty, IA) using the Vitek (bioMerieux, Hazelwood, MO), which identified the isolates as Salmonella spp., serology confirmation required. The Kauffmann–White serotyping scheme was used to detect the somatic (O) antigen group and the flagellar (H) antigens to determine serotype identifications for 32 strains with ESBL phenotypes or elevated fluoroquinolone MIC values. The O antigen groups detected were A (1 strain), B (5 strains), C1 (3 strains), C2 (8 strains), and D (15 strains). Final identifications were as follows: Salmonella serotypes Enteritidis (14 strains), Newport (6), Heidelberg (3), Virchow (2), Albany (1), Thompson (1), Paratyphi A (1), Typhi (one,) and Istanbul (1). In addition, 2 strains of Salmonella enterica serovar Typhimurium were isolated in the midwest of the United States. All but 2 strains of Salmonella serotype Enteritidis were isolated in Latin America, which was also where Salmonella serotypes Albany and Thompson were discovered. The remaining serotypes occurred in the United States and Canada. 2.2. Antimicrobial susceptibility testing Isolates were tested for susceptibility to numerous antimicrobial classes including nalidixic acid, fluoroquinolones, h-lactams, aminoglycosides, tetracycline, and trimethoprim/ sulfamethoxazole. Reference broth microdilution methods were used to determine the isolates susceptibility profile with dry-form 96-well panels manufactured by TREK Diagnostics (Cleveland, OH). All methods used were those recommended by the CLSI (NCCLS, 2003). Breakpoint criteria for nalidixic acid of V 16 Ag/mL for susceptible and z 32 Ag/mL for resistant were used to classify strains that could potentially lead to clinical failure using fluoroquinolone therapy (CLSI, 2005). Among the strains tested, 89 strains were resistant to nalidixic acid, which were isolated in Latin America and North America.

3. Results The activities of 4 fluoroquinolones tested against a large collection (786 strains) of Salmonella spp. isolated during 2001–2003 are shown in Table 1. Using current CLSI (2005) breakpoint criteria, all strains were susceptible (V 1 Ag/mL) to ciprofloxacin, which was the most potent fluoroquinolone tested overall. Among the nalidixic acid-susceptible strains, ciprofloxacin potency (MIC90, V 0.03 Ag/mL) was greater than that of levofloxacin or gatifloxacin (MIC90, 0.06 Ag/mL) and garenoxacin (MIC90, 0.12 Ag/mL). When tested against the nalidixic acid-resistant Salmonella spp. isolates, ciprofloxacin and gatifloxacin had similar activity (MIC90, 0.5 Ag/mL) which was greater than levofloxacin and garenoxacin (MIC90, 1 Ag/mL). Nalidixic acid-resistant strains were detected in diverse geographic regions including Argentina, Brazil, Chile, Mexico, and Venezuela among Latin American countries. Resistant strains were also detected throughout the United States (California, Colorado, Delaware, Iowa, Indiana, New York, North Carolina, and Utah) and Canadian provinces (Alberta and Ontario). Overall, resistance to

16

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

Table 1 Fluoroquinolone potencies among 786 Salmonella spp. isolates that were susceptible or resistant to nalidixic acid from North America and Latin America SENTRY Program participants (2001–2003) Nalidixic acid (n)a

Fluoroquinolone MIC (Ag/mL)

Susceptible Ciprofloxacin (697) Levofloxacin Gatifloxacin Garenoxacin Resistant Ciprofloxacin (89) Levofloxacin Gatifloxacin Garenoxacin

50%

90%

Range

V 0.03 V 0.03 V 0.03 – 0.25 V 0.03 0.06 V 0.03 – 0.5 V 0.03 0.06 V 0.03 – 0.5 0.06 0.12 V 0.03 – 4 0.25 0.5 0.06–1 0.25 1 0.12–2 0.25 0.5 0.06–2 0.5 1 0.12 to N 4

% Susceptibleb 100.0 100.0 100.0 –c 100.0 100.0 100.0 –

a Breakpoint criteria for nalidixic acid of V 16 Ag/mL (susceptible) and z 32 Ag/mL (resistant) were used to classify fluoroquinolone resistance with potential clinical failure M100-S15 (CLSI, 2005). b Breakpoint criteria for Enterobacteriaceae (CLSI, 2005). c No breakpoint criteria have been established.

nalidixic acid was higher in Latin America (15.0%) compared with that in North America (6.3%), and rates were quite variable between countries (data not shown). For example, the rate of nalidixic acid resistance was 50.0% in Mexico and 33.6% in Brazil compared with 1.8–5.6% in the other Latin American nations. The nalidixic acid resistance rates were 6% and 7% in the United States and Canada, respectively. Among the 4 geographically diverse medical centers that sent samples from Brazil, nalidixic acid resistance was observed to be ubiquitous with rates of 22.2% (Brasilia), 27.8% (Porto Alegre), 28.8% (Sao Paulo), and 46.3% (Florianopolis). Fig. 1 shows the distribution of ciprofloxacin MIC values compared with the nalidixic acid MIC tested against Salmonella spp. isolates. Using the current NCCLS nali-

Fig. 1. Distribution of ciprofloxacin MIC results for 786 Salmonella spp. isolates susceptible and resistant to nalidixic acid (CLSI, 2005) during the SENTRY Program 2001–2003 surveillance period. Solid horizontal and vertical lines represent current susceptible breakpoint criteria for ciprofloxacin and nalidixic acid, respectively. The broken horizontal line represents a lower ciprofloxacin breakpoint that separates strains with potential QRDR mutations (upper right quadrant) when compared with the wild-type population.

Fig. 2. Ciprofloxacin MIC values among nalidixic acid-susceptible and nalidixic acid-resistant Salmonella spp. Data above the nalidixic acidresistant bars represent the single-point mutations found in the gyrA gene of the QRDR.

dixic acid breakpoint criteria (solid vertical line) and an adjusted ciprofloxacin breakpoint (broken horizontal line), the Salmonella spp. isolates were divided into fairly distinct populations. Among the 89 nalidixic acid-resistant strains, 84 (94.4%) had ciprofloxacin MIC values at z 0.12 Ag/mL. The 5 isolates that were nalidixic acid-resistant and had ciprofloxacin MIC values of 0.06 Ag/mL were reproducible and within 1 dilution of either breakpoint shown (solid vertical and/or broken horizontal lines). All but 1 strain that was susceptible to nalidixic acid had a ciprofloxacin MIC value of V 0.06 Ag/mL (99.9%). During 2001, 19 strains isolated with ciprofloxacin MIC values of 0.12–0.5 Ag/mL and nalidixic acid MIC values z 32 Ag/mL were tested for QRDR mutations, which detected amino acid substitutions in gyrA only (Fig. 2). The mutations occurred at positions 83 and 87 include the following substitutions: Ser83Phe (7 strains) or Tyr (2 strains), and Asp87Tyr or Asn (5 strains each). The position 83 substitutions resulted in higher ciprofloxacin MIC values (0.12 Ag/mL, 1 strain; 0.25 Ag/mL, 7 strains; and 0.5 Ag/mL, 1 strain) compared with strains with substitutions at position 87 (0.12 Ag/mL). Demographic information provided by each medical center confirmed that the isolates with elevated ciprofloxacin MIC values and gyrA mutations were collected from individual patients with ages from 1 month to 68 years. It was detected (PFGE) that a large outbreak of Salmonella serotype Enteritidis with reduced fluoroquinolone susceptibility occurred in 1 of the Brazilian medical center during the 2001 surveillance year. Over half of the isolates referred by that center were resistant to nalidixic acid and most of these strains had the same gyrA mutation (Asp87Asn). The Salmonella spp. with gyrA mutations included the following serotypes: Enteritidis (14 strains from 5 medical centers), Virchow (2 strains from 2 medical centers), and

Country/ strain

Ceftriaxone

Aztreonam

Cefepime

Imipenem

Cefoxitin

Ciprofloxacin

Gentamicin

Tetracycline

Trimethoprim/ sulfamethoxazole

Clavulanic acid inhibitiona

h-Lactamase enzyme

PFGE

Ceftazidime

USA 272G 330G 701G 821G 1231G 1374G

N 16 N 16 N 16 N 16 16 16

32 32 16 16 16 8

N 16 16 16 16 8 4

0.5 0.5 0.5 1 0.25 0.25

0.25 0.12 0.25 0.25 0.25 0.25

N 32 N 32 N 32 N 32 32 32

V 0.03 V 0.03 V 0.03 V 0.03 V 0.03 V 0.03

V1 V1 V1 V1 V1 V1

N8 N8 N8 N8 N8 N8

V 0.5 V 0.5 V 0.5 V 0.5 V 0.5 V 0.5

Negative Negative Negative Negative Negative Negative

CMY-2 CMY-2 CMY-2 CMY-2 CMY-2 CMY-2

A A A A D E

Canada 554G 684G 1275G 1366G 1409G

N 16 N 16 16 N 16 16

32 16 8 32 8

16 N 16 8 16 16

1 1 0.5 1 0.5

0.5 0.25 0.5 0.25 0.25

N 32 N 32 32 N 32 32

V 0.03 V 0.03 V 0.03 V 0.03 V 0.03

V1 V1 V1 V1 V1

V4 N8 V4 N8 V4

N2 V 0.5 V 0.5 V 0.5 V 0.5

Negative Negative Negative Negative Negative

CMY-2 CMY-2 CMY-2 CMY-2 CMY-2

C B C B C

a

Antimicrobial agents MIC (Ag/mL)

Results of CLSI (2005) confirmatory test.

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

Table 2 Susceptibility characterization for Salmonella spp. with ESBL phenotypes (CLSI screen-positive) isolated in North America from SENTRY Program (2001) participant hospitals

17

18

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

Fig. 3. Pulse-field analysis of 11 Salmonella spp. isolates harboring a confirmed CMY-2 enzyme from North American SENTRY Program sites. The strain numbers are as follows: lane 1, 272G; lane 2, 330G; lane 3, 554G; lane 4, 684G; lane 5, 701G; lane 6, 821G; lane 7, 1231G; lane 8, 1275G; lane 9, 1366G; lane 10, 1374G and lane 11, 1409G.

1 strain each of Typhi, Paratyphi A, Thompson, Newport, Istanbul, and Albany, and no Typhimurium. The susceptibility profiles of 11 Salmonella spp. isolates from the 2001 collection that had ESBL phenotypes are shown in Table 2. All isolates were from North America and were intermediate or resistant to ceftazidime (MIC, z 16 Ag/mL) and most were intermediate or resistant to both ceftriaxone and aztreonam. Among the compounds

shown in the table, cefepime (MIC, 0.25–1 Ag/mL), imipenem (MIC, 0.12–0.5 Ag/mL), ciprofloxacin (MIC, V 0.03 Ag/mL), and gentamicin (MIC, V 1 Ag/mL) retained high potency against these ESBL screen-positive isolates. All 6 strains isolated in the United States and 2 strains isolated in Canada were resistant to tetracycline and only 1 strain was resistant to trimethoprim/sulfamethoxazole (Canada). The results of clavulanic acid inhibition testing showed no decrease in the MIC values and thus a negative ESBL confirmation test. Upon genetic analysis of these screen-positive ESBL-negative strains, the enzyme responsible was determined to be CMY-2, a mobile AmpC enzyme. The 6 US strains with CMY-2 enzymes were isolated in 6 different medical centers in geographically diverse regions. Four strains were identified as Salmonella serotype Newport and 2 strains as Salmonella serovar Typhimurium. Among the 5 Canadian isolates, 2 were determined to be Salmonella serotype Newport and 3 were Salmonella serotype Heidelberg. During the year 2003 surveillance, 4 Salmonella spp. isolates with the ESBL phenotype were noted and included 2 strains from Latin American countries (Brazil and Venezuela). The other 2 isolates were recovered from 2 medical centers in New York (data not shown). Strains with confirmed PCR-positive reactions and sequence results consistent with CMY-2 were analyzed by PFGE (Fig. 3). It was determined that the 4 strains of Salmonella serotype Newport isolated from medical centers in the midwest (2 centers, strains 272G and 701G), the west coast (strain 330G), and the southwest (strain 821G) areas of

Table 3 Regional variation of antimicrobial susceptibility among Salmonella spp. isolated during 3 study intervals from the SENTRY Program Region/antimicrobial agent

Year(s)/% susceptible/resistanta 1997–2000b

Blood culturesc

Stool culturesd

2001–2003

2001–2003

2004e

North America Ampicillin Ceftriaxone Nalidixic acid Ciprofloxacin Tetracycline Trimethoprim/sulfamethoxazole

76.7/23.3 99.1/0.0 (1.7)f 93.1/6.9 (94.0)g 74.1/25.0 94.8/5.2

81.7/18.3 96.9/0.0 (3.8) 87.8/12.2 (87.8) 83.2/16.8 93.9/6.1

79.7/20.3 96.4/2.1 (5.2) 97.4/2.6 (97.4) 75.0/25.0 96.4/3.6

87.1/12.9 96.8/3.2 (3.2) 96.8/3.2 (96.8) 93.5/6.5 100.0/0.0

Latin America Ampicillin Ceftriaxone Nalidixic acid Ciprofloxacin Tetracycline Trimethoprim/sulfamethoxazole

92.5/6.8 100.0/0.0 (0.0) 95.0/5.0 (95.0) 88.8/11.2 91.9/8.1

92.6/7.4 99.5/0.0 (1.0) 82.8/17.2 (83.7) 85.2/14.3 93.1/6.9

84.2/15.4 99.6/0.4 (0.4) 86.8/13.2 (97.6) 81.2/18.4 92.3/7.7

91.5/8.5 100.0/0.0 (0.0) 85.1/14.9 (83.0) 83.0/17.0 91.5/8.5

a

Breakpoint criteria based upon M100-S15 recommendations (CLSI, 2005). Isolates included 116 strains (96.6% blood cultures) from North America and 161 strains (91.9% blood cultures) from Latin America. c Isolates included 131 strains from North America and 203 strains from Latin America. d Isolates included 192 strains from North America and 234 strains from Latin America. e Isolates included 31 strains from North America (64.5% blood cultures and 35.5% stool cultures) and 47 strains from Latin America (70.2% blood cultures and 19.1% stool cultures). f Percentage in parenthesis indicates the proportion of strains consistent with an ESBL phenotype (MIC, z 2 Ag/mL). All such strains did not confirm by clavulanate inhibition and were sequenced as CMY-2 enzymes. g Percentages in parenthesis were based on V 0.06 Ag/mL = susceptible. b

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

the United States had an identical PFGE banding pattern (A). A subtype (A1) of this strain was isolated in a medical center in Canada (2 isolates, strains 684G and 1366G). Three strains of Salmonella serotype Heidelberg (strains 554G, 1275G, and 1409G) with identical PFGE patterns were discovered in 2 medical centers in eastern Canadian provinces. The 2 isolates of Salmonella serovar Typhimurium (strains 1231G and 1374G) were cultured in different medical centers in the midwestern United States and were epidemiologically unrelated to each other. A close look at patient demographic information provided by each medical center confirmed that different individuals had acquired a Salmonella spp. isolate with this resistance mechanism including 2 at-risk patients (ages, V 10 and N 80 years). The data in Table 3 list the susceptibility and resistance percentages for 6 antimicrobial agents tested against Salmonella spp. comparing blood culture and stool culture isolates from North America and Latin America (2001– 2003). The table also shows the susceptibility/resistance data from the 1997–2000 Salmonella spp. collection (z 92% blood culture isolates) and the most recently collected 2004 isolates, which were mostly isolated from blood cultures (64.5–70.2%). Resistance to ampicillin was much higher in North America (12.9–23.3%) compared with Latin America (6.8–15.4%) during all 3 time intervals. Susceptibility to ceftriaxone decreased slightly in North America after 2000 and isolates were mainly cultured from stool. Overall, nalidixic acid resistance (QRDR mutants, but ciprofloxacin-susceptible) was higher in Latin America (5.0 –17.2%) and was greatest among the blood culture isolates compared with stool cultures from both continents. In contrasts, bacteremia isolates from hospitalized salmonellosis patients were more susceptible to ampicillin and tetracycline, but not significantly different than stool isolates for ceftriaxone or trimethoprim/sulfamethoxazole. During 1997–2003, resistance to tetracycline was higher in North America (16.8–25.0%), which declined sharply in 2004 (6.5%). Conversely, trimethoprim/sulfamethoxazole resistance was higher in Latin America (6.9 – 8.5%) compared with North America (3.6–6.1% and 0.0% in 2004). No significant trends toward increasing resistance over time in North America were observed with any of the antimicrobial agents listed in the table. Modest trending to greater resistance in Latin America, however, was detected (Table 3). 4. Discussion The occurrence of antimicrobial-resistant isolates of Salmonella spp. persists and remains a public health care concern as a food-borne pathogen worldwide. This organism is capable of large community outbreaks of food-borne illness evidenced by an early report from the United States in which an estimated 3400 people may have been effected (Horwitz et al., 1977). Interstate and international salmonella outbreaks due to contaminated food products are also

19

regularly reported (Lyytikainen et al., 2000; Sivapalasingam et al., 2003; Van Beneden et al., 1999). This study reports the antimicrobial resistance rates of Salmonella spp. isolates collected from numerous medical centers in North America and Latin America focusing on strains with ESBL screenpositive phenotypes and those that could cause fluoroquinolone therapy failure. In the context of community outbreaks and the potential to cause severe and invasive disease, an increasing prevalence and dissemination of Salmonella spp. exhibiting either of these 2 resistance profiles would pose a therapeutic dilemma as expandedspectrum cephalosporins (ceftriaxone or cefotaxime), and fluoroquinolones are bdrugs of choiceQ for the treatment of invasive gastrointestinal infections. A worldwide collection of Salmonella spp. isolates forwarded to the SENTRY Program during 1997–2001 documented the susceptibility profile of 601 isolates from patient bacteremias (Stephen et al., 2003). These data showed that the tested strains were very susceptible to bthird-generationQ cephalosporins (99.5%) with an ESBL phenotype rate of only 1.0%. Nearly all isolates (99.3%) were also susceptible to ciprofloxacin, although 4.3% of the strains had a ciprofloxacin MIC value of z 0.25 Ag/mL. In this updated analysis, both invasive and noninvasive isolates of Salmonella spp. were analyzed for the 2001–2003 SENTRY surveillance years, and we report twice the ESBL phenotype rate (2.0%) for North America and Latin America only (combined stool and bacteremic strains), compared with the previous results. The current study also showed a higher percentage of strains from these 2 continents with ciprofloxacin MIC values z 0.25 Ag/mL at 6.6%. Using a more conservative susceptibility breakpoint criteria (z 0.12 Ag/mL; Fig. 1), which includes nearly all nalidixic acid-resistant isolates, the rate of strains with potential QRDR mutations increased to 10.8%. For the 2 continents in this study, nalidixic acid-resistant Salmonella spp. with reduced susceptibility to ciprofloxacin were more than 2-fold more prevalent in Latin America (14.1%) compared with North America (6.0%). Although European and Asia-Pacific data were not part of the scope of this study, for comparison purposes we noted that isolates of Salmonella spp. with ESBL phenotypes were recognized in Russia, France, Spain, Germany, Israel, Taiwan, and South Africa during the SENTRY 2001–2003 surveillance periods (data not shown). These 2 regions also had elevated rates (15–16%) of nalidixic acid-resistant Salmonella spp. isolates with reduced susceptibility to ciprofloxacin compared to the Americas. Eleven countries in Europe plus Israel and Turkey cultured strains with this phenotype, and high-level ciprofloxacin resistance (MIC, N1 Ag/mL) was detected in Taiwan, Philippines, Hong Kong, Australia, and Japan. Although numerous ESBL enzymes have been described in Salmonella spp., all of the isolates in this study with ESBL phenotypes that were selected for molecular characterization determined that resistance was mediated by CMY-2, a class C AmpC h-lactamase enzyme (not an

20

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21

ESBL). A previous study documented the emergence and spread of a CMY-2 h-lactamase in Salmonella spp. strains as early as 1996 in the United States (Carattoli et al., 2002). This enzyme has also been described in Salmonella spp. isolates in Europe and the Asia-Pacific region (Miriagou et al., 2002; Navarro et al., 2001; Yan et al., 2003). Although CMY-2 producing strains of salmonella appear to be restricted to isolated events, an increasing prevalence of strains having this resistance mechanism may be predicted due to the mobilization of the bla CMY-2 into a plasmid backbone. These mobile DNA elements will likely continue the transmission and dissemination of CMY-2 among Enterobacteriaceae in the United States and other geographic areas, and the larger reservoir of CMY-2 genes appears to be in E. coli where AmpC-mediated resistance genes are predominant among ESBL screen-positive human isolates (Alvarez et al., 2004). Perhaps an even more eminent threat for treatment of invasive salmonella disease is the increasing reports related to the reduced susceptibility to fluoroquinolones among this common bacterial pathogen. In this report, 10.8% of the isolates tested had ciprofloxacin MIC values at z 0.12 Ag/mL and 11.5% were resistant to nalidixic acid. Other studies have observed similar findings, which include a study in Finland that documented reduced susceptibility to ciprofloxacin at a rate of 6.4% for isolates collected during 1995–1999 using the same breakpoint criteria ( z 0.12 Ag/mL) (Hakanen et al., 2001). This same study noted that an annual increase in fluoroquinolone resistance during the study period was confirmed by logistic regression analysis. Another study in Denmark conducted during 1995–2000 showed similar findings with an increase in nalidixic acid resistance from 0.8% to 8.5% during the first and last study years (Molbak et al., 2002). The justification of the findings in this report and numerous others previously published comes from documented treatment failures and in vivo development of fluoroquinolone resistance, which is reducing the effectiveness of this class among the list of agents to be used for invasive salmonella infections (Kristiansen et al., 2003; Piddock et al., 1993; Rupali et al., 2004). Like other researchers publishing on the topic and the current CLSI (2005) recommendations, these data give strong credibility for using nalidixic acid resistance as a marker for detecting Salmonella spp. isolates with decreased susceptibility to the fluoroquinolone class (Albayrak et al., 2004; CLSI, 2005; Hakanen et al., 1999; Oteo et al., 2000). Again, the fluoroquinolone resistance rates in other human Enterobacteriaceae isolates often exceed that described for Salmonella spp. (Alvarez et al., 2004). Continued surveillance of the antimicrobial profiles of Salmonella spp., a prevalent pathogen once considered very susceptible to commonly applied antimicrobial agents, must continue because of the risk for multidrug-resistant strains detected several decades ago (e.g., DT104) and recent strains that have emerged (not epidemic or endemic) including those with commonly encountered h-lactamases

of enteric origin, as well as those with reduced susceptibility to the newest of fluoroquinolones. Acknowledgments The secretarial assistance of K Meyer is appreciated. We also recognize M Pentella and M Navidomskis (University of Iowa Hygenic Laboratory, Coralville, IA) for their efforts and support with the final identification of the isolates. The SENTRY Program was supported by an educational/research grant from Bristol-Myers Squibb (1999–2004). References Aarestrup FM, Wiuff C, Molbak K, Threlfall EJ (2003) Letters to the Editor. Is it time to change fluoroquinolone breakpoints for Salmonella spp.? Antimicrob Agents Chemother 47:827 – 829. AitMhand R, Soukri A, Moustaoui N, Amarouch H, ElMdaghri N, Sirot D, Benbachir M (2002) Plasmid-mediated TEM-3 extended-spectrum h-lactamase production in Salmonella typhimurium in Casablanca. J Antimicrob Chemother 49:169 – 172. Albayrak F, Cokca F, Erdem B, Aysev AD (2004) Predictive value of nalidixic acid resistance for detecting salmonellae with decreased ciprofloxacin susceptibility. Int J Antimicrob Agents 23:332 – 336. Alvarez M, Tran JH, Chow N, Jacoby GA (2004) Epidemiology of conjugative plasmid-mediated AmpC h-lactamases in the United States. Antimicrob Agents Chemother 48:533 – 537. Baraniak A, Sadowy E, Hryniewicz W, Gniadkowski M (2002) Two different extended-spectrum h-lactamases (ESBLs) in one of the first ESBL-producing salmonella isolates in Poland. J Clin Microbiol 40:1095 – 1097. Carattoli A, Tosini F, Giles WP, Rupp ME, Hinrichs SH, Angulo FJ, Barrett TJ, Fey PD (2002) Characterization of plasmids carrying CMY-2 from expanded-spectrum cephalosporin-resistant Salmonella strains isolated in the United States between 1996 and 1998. Antimicrob Agents Chemother 46:1269 – 1272. Casin I, Breuil J, Darchis JP, Guelpa C, Collatz E (2003) Fluoroquinolone resistance linked to GyrA, GyrB, and ParC mutations in Salmonella enterica Typhimurium isolates in humans. Emerg Infect Dis 9:1455 – 1457. Clinical and Laboratory Standards Institute (2005) Performance standard for antimicrobial susceptibility testing. Document M100-S15. Wayne (PA)7 CLSI. Cormican M, Butler C, Morris D, Corbett-Feeney G, Flynn J (1998) Antibiotic resistance amongst Salmonella enterica species isolated in the Republic of Ireland. J Antimicrob Chemother 42:116 – 118. Crump JA, Barrett TJ, Nelson JT, Angulo FJ (2003) Reevaluating fluoroquinolone breakpoints for Salmonella enterica serotype Typhi and for non-typhi salmonellae. Clin Infect Dis 37:75 – 81. Edelstein M, Pimkin M, Dmitrachenko T, Semenov V, Kozlova N, Gladin D, Baraniak A, Stratchounski L (2004) Multiple outbreaks of nosocomial salmonellosis in Russia and Belarus caused by a single clone of Salmonella enterica serovar Typhimurium producing an extended-spectrum h-lactamase. Antimicrob Agents Chemother 48:2808 – 2815. Fey PD, Safranek TR, Rupp ME, Dunne EF, Ribot E, Iwen PC, Bradford PA, Angulo FJ, Hinrichs SH (2000) Ceftriaxone-resistant salmonella infection acquired by a child from cattle. N Engl J Med 342:1242 – 1249. Giraud E, Cloeckaert A, Kerboeuf D, Chaslus-Dancla E (2000) Evidence for active efflux as the primary mechanism of resistance to ciprofloxacin in Salmonella enterica serovar Typhimurium. Antimicrob Agents Chemother 44:1223 – 1228.

D.J. Biedenbach et al. / Diagnostic Microbiology and Infectious Disease 54 (2006) 13 – 21 Guerra B, Malorny B, Schroeter A, Helmuth R (2003) Letter to the editor. Multiple resistance mechanisms in fluoroquinolone-resistant Salmonella isolates from Germany. Antimicrob Agents Chemother 47:2059. Hakanen A, Kotilainen P, Jalava J, Siitonen A, Huovinen P (1999) Detection of decreased fluoroquinolone susceptibility in salmonellas and validation of nalidixic acid screening test. J Clin Microbiol 37:2577 – 3572. Hakanen A, Kotilainen P, Huovinen P, Helenius H, Siitonen A (2001) Reduced fluoroquinolone susceptibility in Salmonella enterica serotypes in travelers returning from Southeast Asia. Emerg Infect Dis 7:996 – 1003. Hanson ND, Moland ES, Hossain A, Neville SA, Gosbell IB, Thomson KS (2002) Unusual Salmonella enterica serotype Typhimurium isolate producing CMY-7, SHV-9 and OXA-30 h-lactamases. J Antimicrob Chemother 49:1011 – 1014. Heurtin-Le Corre C, Donnio P-Y, Perrin M, Travert M-F, Avril J-L (1999) Increasing incidence and comparison of nalidixic acid-resistant Salmonella enterica subsp. enterica serotype Typhimurium isolates from humans and animals. J Clin Microbiol 37:266 – 269. Horwitz MA, Pollard RA, Merson MH, Martin SM (1977) A large outbreak of foodborne salmonellosis on the Navajo Nation India Reservation, epidemiology and secondary transmission. Am J Public Health 67: 1071 – 1076. Kristiansen MAM, Sandvang D, Rasmussen TB (2003) In vivo development of quinolone resistance in Salmonella enterica serotype Typhimurium DT104. J Clin Microbiol 41:4462 – 4464. Kruger T, Szabo D, Keddy KH, Deeley K, Marsh JW, Huger AM, Nonomo RA, Paterson DL (2004) Infections with nontyphoidal Salmonella species producing TEM-63 or a novel TEM enzyme, TEM-131, in South Africa. Antimicrob Agents Chemother 48:4263 – 4270. Lee K, Yong D, Yum JH, Kim HH, Chong Y (2003) Diversity of TEM-52 extended-spectrum h-lactamase–producing non-typhoidal Salmonella isolates in Korea. J Antimicrob Chemother 52:493 – 496. Ling JM, Chan EW, Lam AW, Cheng AF (2003) Mutations in topoisomerase genes of fluroqouinolone-resistant salmonellae in Hong Kong. Antimicrob Agents Chemother 47:3567 – 3573. Lyytikainen O, Koort J, Ward L, Schildt R, Ruutu P, Japisson E, Timonen M, Siitoner A (2000) Molecular epidemiology of an outbreak caused by Salmonella enterica server Newport in Finland and the United Kingdom. Epidemiol Infect 124:185 – 192. Miriagou V, Filip R, Coman G, Tzouvelekis LS (2002) Expanded-spectrum cephalosporin-resistant Salmonella strains in Romania. J Clin Microbiol 40:4334 – 4336. Molbak K, Gerner-Smidt P, Wegener HC (2002) Increasing quinolone resistance in Salmonella enterica serotype Enteritidis. Emerg Infect Dis 5:514 – 515. Mulvey MR, Soule G, Boyd D, Demczuk W, Ahmed R (2003) MultiProvincial Salmonella Typhimurim Case Control Study Group. Characterization of the first extended-spectrum h-lactamase–producing Salmonella isolate identified in Canada. J Clin Microbiol 41:460 – 462. Nakaya H, Yasuhara A, Yoshimura K, Oshihoi Y, Izumiya H, Watanabe H (2003) Life-threatening infantile diarrhea from fluoroquinolone-resistant Salmonella enterica Typhimurium with mutations in both gyrA and parC. Emerg Infect Dis 9:255 – 257. National Committee for Clinical Laboratory Standards (2003) Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standard M7-A6. Wayne (PA)7 NCCLS. Navarro F, Perez-Trallero E, Marimon JM, Aliaga R, Gomariz M, Mirelis B (2001) CMY-2–producing Salmonella enterica, Klebsiella pneumoniae, Klebsiella oxytoca, Proteus mirabilis and Escherichia coli strains isolated in Spain (October 1999–December 2000. J Antimicrob Chemother 48:383 – 389. Oteo J, Aracil B, Alos JI, Gomez-Garces JL (2000) High rate of resistance to nalidixic acid in Salmonella enterica: Its role as a marker of resistance to fluoroquinolones. Clin Microbiol Infect 6:265 – 282.

21

Piddock LJV, Griggs DJ, Hall MC, Jin YF (1993) Ciprofloxacin resistance in clinical isolates of Salmonella typhimurium obtained from two patients. Antimicrob Agents Chemother 37:662 – 666. Pitout JD, Reisbig MD, Mulvey M, Chui L, Louie M, Crowe L, Church DL, Elsayed S, Gregson D, Ahmed R, Tilley P, Hanson ND (2003) Association between handling of pet treats and infection with Salmonella enterica serotype Newport expressing the AmpC h-lactamase, CMY-2. J Clin Microbiol 41:4578 – 4582. Rankin SC, Aceto H, Cassidy J, Holt J, Young S, Love B, Tewari D, Munro DS, Benson CE (2002) Molecular characterization of cephalosporinresistant Salmonella enterica serotype Newport isolates from animals in Pennsylvania. J Clin Microbiol 40:4679 – 4684. Renuka K, Kapil A, Kabra SK, Wig N, Das BK, Prasad VVSP, Chaudhry R, Seth P (2004) Reduced susceptibility to ciprofloxacin and gyrA gene mutation in North Indian strains of Salmonella enterica serotype Typhi and serotype Pharatyphi A. Microb Drug Resist 10:146 – 153. Ribot EM, Wierzba RK, Angulo FJ, Barrett, TJ (2002) Salmonella enterica serotype Typhimurium DT104 isolated from humans, United States, 1985, 1990, and 1995. Emerg Infect Dis 8:387–391. Rupali P, Abraham OC, Jesudason MV, John TJ, Zachariah A, Sivaram S, Mathai D (2004) Treatment failure in typhoid fever with ciprofloxacin susceptible Salmonella enterica serotype Typhi. Diagn Microbiol Infect Dis 49:1 – 3. Simarro E, Navarro F, Ruiz J, Miro E, Gomez J, Mirelis B (2000) Salmonella enterica serovar Virchow with CTX-M–like h-lactamase in Spain. J Clin Microbiol 38:4676 – 4678. Sivapalasingam S, Barrett E, Kimura A, Van Duyne S, De Witt W, Ying M, Frisch A, Phan Q, Gould E, Shillam P, Reddy V, Cooper T, Hoekstra M, Higgins C, Sanders JP, Tauxe RV, Slutsker L (2003) A multistate outbreak of Salmonella enterica serotype Newport infection linked to mango consumption: Impact of water-dip disinfestations technology. Clin Infect Dis 37:1585 – 1590. Stephen JM, Toleman MA, Walsh TR, Jones RN (2003) SENTRY Program Participants Group. Salmonella bloodstream infections: Report from the SENTRY Antimicrobial Surveillance Program (1997–2001). Int J Antimicrob Agents 22:395 – 405. Tassios PT, Gazouli M, Tzelepi E, Milch H, Kozlova N, Sidorenko S, Legakis NJ, Tzouvelekis LS (1999) Spread of a Salmonella typhimurium clone resistant to expanded-spectrum cephalosporins in three European countries. J Clin Microbiol 37:3774 – 3777. Threlfall EJ, Ward LR, Rowe B (1999) Resistance to ciprofloxacin in nontyphoidal salmonellas from humans in England and Wales: The current situation. Clin Microbiol Infect 5:130 – 134. Threlfall EJ, Skinner JA, Ward LR (2001) Detection of decreased in vitro susceptibility to ciprofloxacin in Salmonella enterica serotypes Typhi and Paratyphi A. J Antimicrob Chemother 48:740 – 741. Vahaboglu H, Dodanli S, Eroglu C, Ozturk R, Soyletir G, Yildirim I, Avkan V (1996) Characterization of multiple-antibiotic-resistant Salmonella typhimurium strains: Molecular epidemiology of PER-1-producing isolates and evidence of nosocomial plasmid exchange by a clone. J Clin Microbiol 34:2942 – 2946. Van Beneden CA, Keene WE, Strang RA, Werker DH, King AS, Mahon B, Hedberg K, Bell A, Kelly MT, Balan VK, MacKenzie WR, Fleming D (1999) Multinational outbreak of Salmonella enterica serotype Newport infection due to contaminated alfalfa sprouts. JAMA 281: 158 – 162. Villa L, Mammina C, Miriagou V, Tzouvelekis LS, Tassios PT, Nastasi A, Carattoli A (2002) Multidrug and broad-spectrum cephalosporin resistance among Salmonella enterica serotype Enteritidis clinical isolates in southern Italy. J Clin Microbiol 40:2662 – 2665. Yan JJ, Ko WC, Chiu CH, Tsai SH, Wu HM, Wu JJ (2003) Emergence of ceftriaxone-resistant Salmonella isolates and rapid spread of plasmidencoded CMY-2-like cephalosprinase, Taiwan. Emerg Infect Dis 9:323 – 328.