Trends in Antimicrobial Resistance of Salmonella Isolated from Animals and Animal Sources in Canada

Table of Contents

1. Abstract
2. Introduction
3. Materials and Methods
4. Results and Discussion
5. Acknowledgements

Abstract

The purpose of our study was to determine the occurrence, magnitude, diagnostic features, genetic aspects, trends and relationships regarding antibiotic resistance of Salmonella isolated from animals, animal food products and the environment of animals. We thereto examined 621 strains of 67 different serovars isolated in 1994, 721 strains of 75 different serovars isolated in 1995, 1219 strains of 83 different serovars isolated in 1996, and 1336 Salmonella strains isolated from 92 different serovars isolated in 1997, for resistance to 17 antibiotics at 1-3 different concentrations with the agar dilution method. The overall resistance magnitude regressed from 9.2% in 1994 to 8.2 % in 1997. Resistance to Streptomycin (30.6% of 3897 strains), tetracycline (27.3%), and sulfisoxazole (23.6%) was highest. Resistance to streptomycin, tetracycline, kanamycin and gentamicin declined whilst resistance to ampicillin, chloramphenicol and neomycin increased during the 4 year period. None of the strains were resistant to amikacin. None of the isolates were resistant to ciprofloxacin at 1, 2 and 4µg/ml. Resistance to nalidixic acid correlated significantly with decreased sensitivity to ciprofloxacin. Resistance of S. typhimurium to each of the seven antibiotics Amp, Chl, Kan, Neo, Str, Sul, and Tet increased persistently during each of the years 1994-97, but none showed decreased sensitivity to ciprofloxacin.

Introduction

Infection of humans (Gold and Moellering, 1996) and animals (Wray et al., 1993) with antimicrobial resistant bacteria and contamination of food (Spika et al., 1987), drink (Ryan et al., 1987) and the environment (Waters and Davies, 1997) with resistant bacteria has become a significant problem in recent years (Tenover and McGowan, 1996). Antimicrobials used for treatment or growth promotion in animals are also used for disease control in humans (Threlfall et al., 1993), and may select for cross-resistance in bacteria to antimicrobials used in human medicine (Wells and James, 1973; Gast and Stephens, 1988; Gast et al., 1988). Microbiological and clinical evidence is mounting that resistant bacteria or resistance determinants might be passed from animals to humans, resulting in infections that are more difficult to treat (Anon, 1997). An increase in resistance of zoonotic pathogens such as certain non-typhoid Salmonella serovars (Trelfall et al., 1996) and Campylobacter spp. (Endtz et al., 1991), infecting humans via the food chain (Davies et al., 1996) or by direct contact with infected animals (Wall et al. 1994) has been observed.

The main purpose of monitoring programmes for antimicrobial resistance in bacteria like non-typhoid Salmonella is to systematically collect and evaluate microbiological and epidemiological information pertinent to effective control and containment of resistant bacteria that could be transmitted from animals to humans (Anon, 1997; Aarestrup et al., 1998). This information may then lead to the detection and measures to prevent transmission of resistant bacteria and resistance determinants from animals to humans, and promote the prudent use of antibiotics in food animals and humans (Anon, 1997). The aim of the present study was to determine the antibiotic resistance of Salmonella isolated from animal, animal food products and the environment of animal production, to record the associated epidemiological information about the animal host or animal product from which the Salmonella were isolated, and to determine trends regarding antibiotic resistance of Salmonella isolated from animal sources in Canada during the 1994-97 period.

Materials and Methods

Isolates: Salmonella strains isolated during the years 1994-97 in Canada from animals, animal food products, and the environment of animals were serotyped at the Health Canada, OIÉ Reference Laboratory for Salmonellosis in Guelph. All S. enteritidis, S. hadar, S. heidelberg and S. typhimurium isolates were phagetyped. Previous studies have shown that, although shifts in resistance to antibiotics occur (Wray et al., 1998), resistance of Salmonella strains is often highly associated with the serovar (Sojka et al., 1986; Poppe et al., 1995) and, for certain serovars such as S. typhimurium and S. enteritidis with the phage type (Duck et al., 1978; Frost et al. 1989) of the strain. Therefore, Salmonella strains were selected for determination of antibiotic resistance in the following manner and order: a) all clinical and non-clinical or survey Salmonella isolates that were of a different serovar and phage type from one source or submission, were examined for resistance to antibiotics provided that, as a minimum, information regarding the date, the province and the species from which the sample was collected were known; b) each fourth Salmonella isolate from a submission or source consisting of more than four isolates was examined for resistance to antibiotics even if belonging to the same serovar. Other information regarding the isolate, including place where the strain was isolated, the animal species or animal food product and the type of sample from which the isolate was made, was recorded. Repetitive submissions consisting of large numbers of Salmonella isolates cultured from broilers, broiler carcass rinses, turkey parts, and mechanically separated meat destined for export, were excluded from the study. The latter isolates were cultured from samples obtained at two large poultry slaughtering plants in Ontario primarily during 1994.

Resistance to antimicrobial agents: Antibiotic-susceptibility tests of the Salmonella strains isolated in 1994-96 were carried out with the Cathra Replicator (Automed, Shoreview, Minn.) using Mueller Hinton agar plates (Oxoid, Nepean, Ont.) containing the following drugs: amikacin (Amk) at 16, 32 and 64 µg/mL, ampicillin (A) at 8, 16 and 64 µg/mL, ceftriaxone (Cef) at 8, 16 and 32 µg/mL, cephalothin (Clt) at 8, 16 and 32 µgmL, chloramphenicol (C) at 8, 16 and 32 µg/mL, ciprofloxacin (Cip) at 1, 2 and 4 µg/mL, cotrimoxazole (Cot) (trimethoprim at levels of 0.5, 2 and 4 µg/mL with sulfamethoxazole at levels of 9.5, 38 and 76 µg/mL, respectively), gentamicin (G) at 4, 8 and 16 µg/mL, kanamycin (K) at 16, 32 and 64 µg/mL, nalidixic acid (Nal) at 32 µg/mL, neomycin (Neo) at 4, 8 and 16 µg/mL, nitrofurantoin (Nit) at 64 µg/mL, polymyxin B (Pol) at levels of 2, 8, and 16 µg/mL, streptomycin (S) at 8, 16 and 32 µg/mL, sulfisoxazole (Su) at 256 µg/mL, tetracycline (T) at 4, 8, 16 µg/mL and tobramycin (Tob) at 4, 8 and 16 µg/mL. Salmonella strains isolated in 1997 were examined as indicated above except that reduced sensitivity to ciprofloxacin was tested at 0.125 µg/mL only, to polymyxin B at 16 µg/mL only, and to streptomycin at 32 µg/mL only. All 1994-96 strains resistant to nalidixic acid were examined additionally for resistance to ciprofloxacin at 0.125 µg/mL. All 1997 strains resistant to nalidixic acid or showing reduced sensitivity to ciprofloxacin at 0.125 µg/mL were also tested for growth on agar plates containing ciprofloxacin at 1 µg/mL. Antibiotic susceptibility testing was done in accordance with the National Committee for Clinical Laboratory Standards (NCCLS) guidelines (1994).

Results and Discussion

Six-hundred-and-twenty-one Salmonella strains of 67 different serovars isolated in 1994, 721 Salmonella strains of 75 different serovars isolated in 1995, 1,219 Salmonella strains of 83 different serovars isolated in 1996, and 1,336 Salmonella isolates of 92 different serovars isolated in 1997, in total 3,897 strains, were examined for resistance to antimicrobial agents. The number of clinical versus non-clinical or survey isolates examined for antibiotic resistance were 316 and 305 isolates, respectively, of the 1994 isolates; 457 and 264, respectively, of the 1995 isolates, 436 and 783, respectively, of the 1996 isolates; and 464 and 872, respectively, of the 1997 isolates.

Table 1. Percentages of Salmonella strains resistant to antimicrobial agents

Antimicrobial	µg/mL	% of 621	% of 721	% of 1219	% of 1336
		1994	1995	1996	1997
Amikacin	³64	0.0	0.0	0.0	0.0
Ampicillin	³64	8.4	10.8	11.9	15.9
Ceftriaxone	³32	0.0	0.3	0.1	0.1
Cephalothin	³32	0.6	1.9	0.9	2.0
Chloramphenicol	³32	4.3	7.3	7.6	8.7
Ciprofloxacin	³4	0.0	0.0	0.0	NDc
Ciprofloxacin	³0.125	ND	ND	ND	1.5d
Cotrimoxazolee	³80	2.3	0.8	1.9	2.2
Gentamicin	³16	12.1	10.8	8.1	6.8
Kanamycin	³64	12.7	11.9	8.0	9.2
Nalidixic acid	³32	6.1	4.4	2.9	1.6
Neomycin	³16	5.5	6.0	5.7	7.8
Nitrofurantoin	³64	7.1	10.3	5.8	5.1
Polymixin B	³16	0.0	0.0	0.1	0.1
Streptomycin	³32	36.7	29.8	31.4	26.6
Sulfisoxazole	³256	23.5	23.3	26.1	22.0
Tetracycline	³16	29.4	30.0	26.8	25.5
Tobramycin	³16	7.9	6.5	2.5	2.9

aHighest concentration of antibiotic employed
bPercentage of total number of strains examined that were resistant
cND = not done
dThese strains showed decreased sensitivity to ciprofloxacin and grew on agar containing 0.125 mg/mL but were considered sensitive to ciprofloxacin since they did not grow on plates containing 1 mg/mL
eCotrimoxazole = combination of trimethoprim (4 mg/mL) and sulfamethoxazole (76 mg/mL)

The percentage of strains resistant to the highest level of the antibiotics in the testing panel is shown in Table 1. The strains were either resistant to the highest level of the antibiotics used or were sensitive to antibiotics at all levels employed; a low percentage of strains (2.5% or less) grew on plates containing intermediate levels of antibiotics. The exception was resistance to streptomycin; 3.6% of the 1994 strains and 8.5% of the 1995 strains were resistant at the intermediate level of 16 µg/mL. The average percentages of Salmonella isolates resistant to each of the antibiotics for the 1994-97 period is shown in Figure 1. Resistance of isolates to streptomycin was highest followed by resistance to tetracycline, sulfisoxazole, ampicillin, kanamycin, gentamicin, nitrofurantoin, chloramphenicol, neomycin, tobramycin, nalidixic acid, cotrimoxazole and cephalothin. Only 4 strains were resistant to ceftriaxone; none among the 1994, 2 of the 1995, and 1 each among the 1996 and the 1997 isolates. Only 2 strains, 1 each of the 1996 and the 1997 isolates were resistant to polymyxin B. None of the strains were resistant to amikacin at 64 µg/mL, neither did any of the strains grow on agar plates containing 16 and 32 µg/mL amikacin. The latter results are similar to previous studies in which we examined 457, 1,159 and 2,690 Salmonella strains isolated from laying hens, broilers and turkeys, respectively, but found no resistance to amikacin (Poppe et al., 1995; Poppe et al., 1996).

None of the 1994-96 isolates were resistant to ciprofloxacin at 4 µg/mL; neither did any of them grow on agar plates with 2 or 1 µg/mL. One-and-a-half percent of the 1997 strains showed reduced sensitivity to ciprofloxacin at the level of 0.125 µg/mL but none of these strains grew on agar containing 1 µg/mL. Resistance to nalidixic acid correlated significantly with lowered sensitivity to ciprofloxacin. One-hundred-and-twenty-two of 127 (96%) strains resistant to nalidixic acid at 32 µg/mL grew on agar plates containing ciprofloxacin at 0.125 µg/mL but not on those containing 1 µg/mL. Eighty-nine percent of the 1994-97 isolates of S. bredeney and 23% and 19% of the isolates of S. senftenberg and S. ohio, respectively, isolated during the 1994-97 period, showed resistance to nalidixic acid at 32 µg/mL and decreased sensitivity to ciprofloxacin. Resistance of S. typhimurium to nalidixic acid and ciprofloxacin may reside on the same site of the gyrA gene resulting in high level resistance to nalidixic acid and decreased sensitivity to ciprofloxacin with a minimal inhibitory concentration (MIC) of 0.25 to 1 µg/mL (Heurtin-Le Corre et al., 1999). When resistance to nalidixic acid and ciprofloxacin resistance resides at different sites of the gyrA gene, the MIC for ciprofloxacin was lower and ranged between 0.03 µg/mL to 0.25 µg/mL (Heurtin-Le Corre et al., 1999).

Resistance of strains to ampicillin, chloramphenicol and neomycin increased during the 1994-97 years (Figure 2). Resistance to ampicillin rose from 8.4% of the 1994 isolates to 10.8% of the 1995, to 12.0% of the 1996, and to 15.9% of the 1997 isolates. Similarly, the percentages of strains resistant to chloramphenicol rose consistently from 4.3% of the 1994 isolates to 8.7% of the 1997 isolates, respectively. Resistance to neomycin rose from 5.5% among the 1994 isolates to 7.7% of the 1997 isolates.

A large percentage of the increased resistances could be attributed to the increased percentage of S. typhimurium strains that were resistant to these antibiotics (Figure 3), and, to a lesser degree, to the rise from 11.4% in 1994 to 17.7% in 1997 of S. typhimurium isolates among the Salmonella isolates in the collection of strains examined for antimicrobial resistance. Resistance to ampicillin among the S. typhimurium strains rose consistently from 14.1% in 1994 to 47.7% in 1997, to chloramphenicol it rose from 14.1% in 1994 to 44.3% in 1997, and to neomycin it rose from 11.3% to 20%. The overall increased resistance of Salmonella strains to ampicillin and chloramphenicol may in part be attributed to the increased isolation rates during the study period of S. typhimurium DT104 since 85.6% of the DT104 strains were resistant to ampicillin, chloramphenicol, streptomycin, sulfisoxazole and tetracycline. The overall increased resistance to neomycin may in part be attributed to the finding that 39.6% of the S. typhimurium DT104 strains were resistant to neomycin and kanamycin in addition to being resistant to A, C, S, Su and T. Also, the percentage of S. typhimurium DT104 isolates among the S. typhimurium strains derived from cattle rose from 19.2% in 1994 to 76.7% in 1997, and among the S. typhimurium strains from pigs the percentage of DT104 strains rose from 3.4% during the 1994-95 period to 21% during the 1996-97 period. Increased resistance of S. typhimurium during each of the years of the study period to sulfisoxazole, tetracycline and, to a lesser degree to streptomycin, may also be contributed to the increased percentages of S. typhimurium DT104 among the S. typhimurium isolates; it rose from 12.7% in 1994 to 51.5% in 1997.

Although the overall resistances to ampicillin, chloramphenicol and neomycin increased, the overall resistance magnitude to all antibiotics in the testing panel regressed from 9.1% in 1994 to 8.0% in 1997. This may be attributed to an overall decline in resistance of Salmonella strains to streptomycin and tetracycline, and to a lesser degree, to decreased resistance to kanamycin, gentamicin, nitrofurantoin, tobramycin and nalidixic acid.

The degree of resistance of strains to antibiotics showed marked differences depending on the serovar of the isolate. For example, S. bredeney strains were on average resistant to 6.6 (38.6%) of the 17 antibiotics in the test panel but S. brandenburg strains were resistant to only 1.2% of the antibiotics.

Strains of Salmonella typhimurium, S. heidelberg and S. hadar were most numerous among those submitted for serotyping and examined for drug resistance. These serovars and S. enteritidis are also the most common serovars isolated from humans (Khakhria et al., 1997). Strains of S. typhimurium were most commonly isolated from cattle and pigs, whilst strains of S. heidelberg and S. hadar were most frequently isolated from chickens and turkeys. The average percentage resistance of S. typhimurium, S. heidelberg and S. hadar strains was 14.9%, 7.1%, and 13.8%, respectively. The percentage resistance to antibiotics of S. typhimurium strains rose annually from 7.7% in 1994 to 17.7% in 1997 and for S. hadar it rose from 12.8% in 1994 to 15% in 1997 (Figure 3). For Salmonella enteritidis, which was the seventh most commonly serovar among strains submitted for serotyping during the 1994-97 period, the average percentage resistance of strains was 2.4%.

The Salmonella serovars showed a high degree of specificity for certain drug resistance patterns. Thus, a high percentage of the S. bredeney were resistant to the aminoglycosides gentamicin, kanamycin, streptomycin and tobramycin, and to nalidixic acid, sulfisoxazole and tetracycline, and a high percentage of S. derby strains were resistant to streptomycin, sulfisoxazole and tetracycline, but not to other antibiotics. Salmonella hadar strains were mainly resistant to streptomycin and tetracycline but not to the sulfonamides, whereas about 25% of S. heidelberg strains were resistant to gentamicin, streptomycin, tetracycline and sulfisoxazole.

Acknowledgements

We wish to thank Dr. J. Greenfield, Dr. M. Ayroud, Dr. G. Ollis, Dr. M. Chirino-Trejo, Dr. N. Smart, Dr. H. Hariharan, Dr. H. Whitney and many others at provincial, federal and private laboratories across Canada for submitting Salmonella strains for serotyping and phagetyping and for helpful discussions regarding antimicrobial resistance. We thank Ms. Carla Duncan, research assistant, and Hilary Davies, Kevin Graham, Jennifer Becker, Steven Moore and Kate Daniel, Co-op students, for excellent technical assistance.

 

 

Figure 1. Percentages of Salmonella isolates resistant to antimicrobial agents (1994 - 1997)

Figure 2. Percentages of Salmonella isolates resistant to antimicrobial agents by year (1994 - 1997)

Figure 3. Average percent resistance of Salmonella serovars to 17 antimicrobial agents by year (1994 - 1997)