|
| |||||||||||||||||||||||||||||||||||||
Background Salmonella enterica serotype choleraesuis is a cause of serious systemic infections. Because fluoroquinolones are the drug of choice for the treatment of severe salmonella infections, the emergence and dissemination of fluoroquinolone-resistant S. enterica serotype choleraesuis have clinical consequences.
Methods In Taiwan, a hospital-based surveillance system has been in place since 1987 to monitor the incidence of S. enterica serotype choleraesuis infections and the antimicrobial susceptibility of the isolates. We investigated the rapid emergence of fluoroquinolone resistance in this serotype in 2000 and 2001. Pigs in Taiwan were evaluated as a potential source of the resistant salmonella.
Results A total of 501 clinical isolates of S. enterica serotype choleraesuis were recovered in our hospital from 1987 through 2000. The proportion of total salmonella isolates made up by S. enterica serotype choleraesuis decreased from an average of 8.4 percent before 1995 to 2.7 percent in 1996 through 1998. During 1999 and 2000, this proportion increased significantly, to an average of 5.0 percent. Ciprofloxacin resistance in S. enterica serotype choleraesuis has been observed since 2000. In the third quarter of 2001, 60 percent of isolates were resistant to ciprofloxacin. Molecular typing indicated that the primary source of S. enterica serotype choleraesuis isolates was herds of swine. All the resistant isolates from humans and swine had mutations that led to the substitution of phenylalanine for serine at position 83 and asparagine for aspartic acid at position 87 in the gene for DNA gyrase A.
Conclusions This investigation in Taiwan indicates that fluoroquinolone-resistant S. enterica serotype choleraesuis can spread from swine to humans. The use of fluoroquinolones in food animals should be prohibited.
The first recognized outbreak of fluoroquinolone-resistant salmonella infection in the United States occurred in 1997.8 Increasing resistance to fluoroquinolones has been detected in many other countries,9,10,11,12 and it has also emerged in the multidrug-resistant clone of S. enterica serotype typhimurium definitive type 104 (DT104).12,13 Most of the resistant salmonella strains belong to serotypes typhimurium, Hadar, Schwarzengrund, and enteritidis,8,9,10,11,12,13 which usually cause gastroenteritis and sometimes cause extraintestinal infections.4,14 The development of fluoroquinolone resistance among serotypes that often cause systemic infections would have serious implications for public health, as well as implications for the care of individual patients. We report the emergence in Taiwan of fluoroquinolone resistance in S. enterica serotype choleraesuis. Molecular epidemiologic investigation indicated that swine served as a reservoir for the resistant bacteria.
Methods
Hospital-Based Surveillance and Identification of Isolates
Since 1987, infections caused by salmonella in Taiwan have been monitored by review of the records of the isolation of salmonella in the Clinical Microbiology Laboratory, Department of Clinical Pathology, Chang Gung Memorial Hospital and Chang Gung Children's Hospital. The data, including dates and sites of isolation and information on antimicrobial susceptibility, have been incorporated into computerized data bases and shared in a timely fashion with physicians. We analyzed data on salmonella isolates and investigated resistance to fluoroquinolones in S. enterica serotype choleraesuis isolates from 2000 and 2001. All the isolates described in this report came from different patients or, rarely, from unrelated episodes in the same patient. Any additional isolates from the same patient during a single episode were excluded. The Chang Gung Memorial Hospital is a 3500-bed, university-affiliated medical center, and the Chang Gung Children's Hospital is a 450-bed hospital for children. Both are located in northern Taiwan, but the patients came from throughout Taiwan, including the scattered islands.
All isolates were cultured and identified according to standard methods,15 with no major changes over time in the policy for the identification of salmonella. All isolates were serotyped by the slide agglutination test with the use of O antiserum to detect O antigen and by the tube agglutination test with the use of H antiserum to detect H antigen (Difco). The identification of S. enterica serotype choleraesuis followed the KauffmanWhite scheme.
The antimicrobial susceptibility of all clinical isolates of salmonella was investigated by the standard disk-diffusion method for nonblood isolates and by the microbroth-dilution method for blood isolates.16,17 The antimicrobial agents examined included ampicillin, chloramphenicol, trimethoprimsulfamethoxazole, cefixime, ceftazidime, ceftizoxime, ceftriaxone, cefepime, and ciprofloxacin. Ciprofloxacin was first used at the hospitals in October 1996. Susceptible and resistant isolates were defined according to the criteria suggested by the National Committee for Clinical Laboratory Standards.16,17
Microbiologic Examination
Forty-eight clinical isolates of S. enterica serotype choleraesuis collected between January 2000 and June 2001 were subjected to molecular microbiologic examination. In addition, 17 clinical isolates collected in 199718 and another 26 recovered from pigs with scours (dysentery) in 2000 and 2001 were tested. The minimal inhibitory concentration (MIC) of various antibiotics was determined in these isolates by the E test (AB Biodisk) in accordance with guidelines of the National Committee for Clinical Laboratory Standards.17 The plasmid profiles of the isolates were determined by the method of Kado and Liu.19 A pair of oligonucleotide primers synthesized according to the published DNA sequence of spvC, a conserved gene located on the salmonella virulence plasmid, was used to detect the plasmid in the isolates.20,21 These isolates were also genotyped by infrequent-restriction-sitepolymerase-chain-reaction (IRS-PCR) analysis. This method followed the procedures described previously,22 except that 15 µl of adapter-ligated DNA templates was used in the final amplification. Each isolate was analyzed at least twice to ensure the reproducibility of the result. The criteria proposed by Tenover et al.23 were used to analyze the DNA fingerprints generated by IRS-PCR.
A 313-bp DNA fragment was amplified with the primers P1 (5'TACCGTCATAGTTATCCACGA) and P2 (5'GTACTTTACGCCATGAACGT),12,13 which correspond to nucleotides 434 to 454 and 142 to 161 of the gene for DNA gyrase A (gyrA), respectively. The amplified fragment contains the quinolone-resistancedetermining region of gyrA.12 PCR products obtained after amplification were purified with use of the Wizard PCR Preps kit (Promega) and were sequenced with an automatic sequencer (ABI 373A, PerkinElmer, Applied Biosystems). The sequences obtained were analyzed by PCGene software (IntelliGenetics). The search for homologous sequences was performed in the GenBank data base with use of FASTA software.
To examine whether the resistance was mediated by a plasmid, plasmids of the resistant isolates underwent electrophoresis on agarose gels and were then transferred onto Zeta-Probe membranes (Bio-Rad). To prepare a probe, the purified PCR product was labeled with [
-32P]2'-deoxycytidine-5'-triphosphate with use of the Random Primers DNA Labeling System (GIBCOBRL). DNADNA hybridization was performed as described earlier.18
Statistical Analysis
The chi-square test was used to determine the significance of differences. A difference was considered statistically significant if the P value was less than 0.05. All statistical analyses were performed with the use of Epi Info software (version 6.04).
Results
Surveillance of S. enterica Serotype Choleraesuis
A total of 8196 salmonella isolates were analyzed. The annual number of isolates increased from 232 in 1987 to 700 in 2000 (average, 585; maximum, 910 in 1995). As a proportion of all bacterial isolates tested in this laboratory (26,731 in 1987 and 49,778 in 2000), the proportion of salmonella isolates increased significantly (P<0.001), from 0.9 percent in 1987 to 1.4 percent in 2000 (average, 1.5 percent; maximum, 2.1 percent in 1995). A total of 501 separate clinical isolates of S. enterica serotype choleraesuis were recovered in our laboratory from 1987 through 2000. A total of 359 (72 percent) were isolated from blood. Other sources of isolates included urine, wound, bone, tissue from a mycotic aneurysm, and rarely, stool. As shown in Figure 1, the annual number of isolates increased gradually up to 1995, with a substantial decline in 1996 through 1998 and an increase thereafter. The proportion of S. enterica serotype choleraesuis isolates among all salmonella isolates also decreased from an average of 8.4 percent before 1995 to 2.7 percent in 1996 through 1998 (P<0.001). During 1999 and 2000, this proportion increased significantly to an average of 5.0 percent (P<0.001).
|
Emergence of Fluoroquinolone Resistance in S. enterica Serotype Choleraesuis
Before 1991, with rates of resistance below 40 percent, most S. enterica serotype choleraesuis isolates were susceptible to ampicillin, chloramphenicol, and trimethoprimsulfamethoxazole. Since then, the rates of resistance to these antibiotics have increased substantially (Figure 2). In 2000, resistance to at least one of the three antibiotics was found in approximately 90 percent of S. enterica serotype choleraesuis isolates (Figure 2). Seventy-eight percent of the isolates were resistant to all three antibiotics in 2000, when ciprofloxacin resistance began to emerge.
|
Microbiologic Examination
According to the criteria of Tenover et al.,23 four types (1 to 4) were identified among the S. enterica serotype choleraesuis isolates on IRS-PCR analysis. The four major types were further differentiated into 12 subtypes, with a difference in at least one band used to define a subtype (Figure 3). Seventy-one of the 91 isolates analyzed (78 percent) were of subtype 1a, indicating that there was an endemic strain of S. enterica serotype choleraesuis circulating in Taiwan (Table 1). The isolates recovered from different species (humans and swine) and in different years (1997 and 2000 to 2001) did not differ significantly in the distribution of their IRS-PCR patterns (Table 1). There was also no significant difference in the distribution of the patterns between ciprofloxacin-susceptible and ciprofloxacin-resistant isolates (Table 1).
|
|
DNA sequencing of the PCR-amplified gyrA region of 65 isolates from humans and 26 from swine identified a base substitution at codons 83 and 87 (according to Escherichia coli numbering12) in all ciprofloxacin-resistant isolates (Table 1). The substitutions, TCC
TTC at codon 83 and GAC
AAC at codon 87, caused amino acid changes of serine to phenylalanine and aspartic acid to asparagine, respectively. All resistant isolates from humans contained the two substitutions, as did resistant isolates from swine. Some ciprofloxacin-susceptible isolates possessed different nucleotide substitutions at the two positions, leading to different amino acid changes (Table 1). Most susceptible strains, including those isolated from humans in 1997, had a single amino acid change either Ser83Phe or Asp87Asn (Table 1). Thus, the two mutations are equally important in permitting S. enterica serotype choleraesuis to attain resistance to ciprofloxacin. The amplified PCR product of gyrA did not hybridize to any plasmid of the 35 resistant isolates examined in DNADNA hybridization.
Discussion
Infection with S. enterica serotype choleraesuis has become common in Taiwan.14,24 S. enterica serotype choleraesuis is the most frequently isolated serotype of S. enterica after typhimurium and Schwarzengrund.14,24 The number of S. enterica serotype choleraesuis infections decreased significantly from 1996 through 1998. In Taiwan in 1996, an epidemic of foot-and-mouth disease in swine resulted in the slaughter of many pigs across the island. Because humans acquire salmonella infections by eating infected animals, the decrease in S. enterica serotype choleraesuis infections in humans may have been the result of the extensive reduction in the pig population, which served as the reservoir of this serotype.
The proportion of clinical isolates of S. enterica serotype choleraesuis that were fluoroquinolone-resistant increased rapidly and dramatically from none through 1999 to 60 percent in the third quarter of 2001. All of the isolates from humans and swine that we examined had similar IRS-PCR and plasmid patterns, suggesting that human S. enterica serotype choleraesuis infections were acquired from pigs. It appears that there was an endemic strain of S. enterica serotype choleraesuis circulating in humans and pigs. The sequencing results support this hypothesis, since all resistant isolates, irrespective of source, had the same amino acid changes in gyrA. We previously demonstrated that S. enterica serotype choleraesuis isolates were resistant to ampicillin and sulfonamides as a result of the acquisition of a large R (resistance) plasmid, the recombination of the R plasmid with the virulence plasmid, or both.18 It appears, however, that the gene for resistance to fluoroquinolones in S. enterica serotype choleraesuis is located on the chromosome rather than on the plasmid. Our data suggest that the two mutations in gyrA gave rise to the resistance and that the rapid emergence of the highly resistant population was due mainly to the clonal spread of an endemic resistant strain.
The two most common mutations associated with fluoroquinolone resistance in S. enterica serotype typhimurium DT104 are at codons 83 and 87 and involve amino acid changes Ser83Phe and Asp87Asn, respectively.13 The same mutations were responsible for fluoroquinolone resistance in S. enterica serotype choleraesuis. We could not completely exclude the possibility that mutations in gyrB, parC, or parE might also contribute to the resistance phenotype; however, this is unlikely, since mutations in gyrB and parC are not normally associated with resistance to fluoroquinolones in salmonella,25,26 and strains with a single gyrA mutation were present among human isolates as early as 1997. This fact implies that if the antecedent strains with a single mutation had been detected earlier, by either phenotypic or molecular methods, the emergence of full resistance might have been predicted. Since resistant isolates probably spread from animals to humans, regular surveillance of salmonella isolates from animals could be helpful.27
There is increasing concern in the public health community about the possibility that antibiotics fed to animals that are then consumed by humans may contribute to resistance in human pathogens.1,3,10,13,27,28,29,30 Certain antibiotics are critical to human medicine because they are effective against pathogens that are resistant to other antibiotics. Fluoroquinolones are among these critical antibiotics, so the emergence of fluoroquinolone-resistant S. enterica serotype choleraesuis is potentially a serious problem. A recent survey by the National Health Research Institute of Taiwan found that five antibiotics important to human medicine, including a fluoroquinolone (enrofloxacin), have been widely added to animal feed for years.31 Half of the feed-mill operators surveyed said they added enrofloxacin to pig feed as a growth promoter.31 The nontherapeutic use of enrofloxacin in domestic pigs, therefore, can reasonably be expected to select for resistance to fluoroquinolones, including ciprofloxacin, in S. enterica serotype choleraesuis.
In 2000, the U.S. Food and Drug Administration announced plans for withdrawing two fluoroquinolones, sarafloxacin and enrofloxacin, from use to treat respiratory and diarrheal diseases in poultry. However, many experts within the agricultural and pharmaceutical industries opposed the proposed ban, arguing that improper use of such drugs in hospital settings represented the chief source of resistance. To explore this question, we reviewed the medical records of the patients infected by ciprofloxacin-resistant S. enterica serotype choleraesuis. None of these patients had received quinolone therapy before the onset of the infection (data not shown). Fluoroquinolone use in animals, rather than in humans, appears to account for most of the problem of resistance.
The emergence of fluoroquinolone resistance would change the policy for the treatment of S. enterica serotype choleraesuis infections. Because the majority of S. enterica serotype choleraesuis isolates are also resistant to ampicillin, chloramphenicol, and trimethoprimsulfamethoxazole, the third-generation cephalosporins are now the only antibiotics with reliable activity against this serotype in Taiwan. In view of the severe adverse consequences for human health of the use of fluoroquinolones in food animals, we suggest that such use should be prohibited.
Supported by grants (CMRP1197, to Dr. Chiu, and CMRP798-III, to Ms. Wu) from Chang Gung Memorial Hospital and Chang Gung Children's Hospital.
Source Information
From the Division of Pediatric Infectious Diseases, Department of Pediatrics, Chang Gung Children's Hospital, Taoyuan (C.-H.C., T.-Y.L.); the Department of Clinical Pathology, Chang Gung Memorial Hospital, Taoyuan (T.-L.W., L.-H.S., J.-H.C., A.-J.K.); the Department of Microbiology and Immunology, Chang Gung University College of Medicine, Taoyuan (C.C.); and the Graduate Institute of Veterinary Pathology, National Chung-Hsing University, Taichung (M.-S.C.) all in Taiwan.
Address reprint requests to Dr. Chiu at the Chang Gung Children's Hospital, 5 Fu-Hsin St., Kweishan 333, Taoyuan, Taiwan, or at chchiu{at}adm.cgmh.org.tw.
References
| |||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |