bs_bs_banner
Equine Veterinary Journal ISSN 0425-1644 DOI: 10.1111/evj.12234
Rapid Salmonella detection in experimentally inoculated equine faecal and veterinary hospital environmental samples using commercially available lateral flow immunoassays B. A. BURGESS, N. R. NOYES, D. S. BOLTE, D. R. HYATT, D. C. VAN METRE and P. S. MORLEY* Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, USA. *Corresponding author email: [email protected]; Received: 12.04.13; Accepted: 26.01.14
Summary Reasons for performing study: Salmonella enterica is the most commonly reported cause of outbreaks of nosocomial infections in large animal veterinary teaching hospitals and the closure of equine hospitals. Rapid detection may facilitate effective control practices in equine populations. Shipping and laboratory testing typically require ≥48 h to obtain results. Lateral flow immunoassays developed for use in food-safety microbiology provide an alternative that has not been evaluated for use with faeces or environmental samples. Objectives: We aimed to identify enrichment methods that would allow commercially available rapid Salmonella detection systems (lateral flow immunoassays) to be used in clinical practice with equine faecal and environmental samples, providing test results in 18–24 h. Study design: In vitro experiment. Methods: Equine faecal and environmental samples were inoculated with known quantities of S. enterica serotype Typhimurium and cultured using 2 different enrichment techniques for faeces and 4 enrichment techniques for environmental samples. Samples were tested blindly using 2 different lateral flow immunoassays and plated on agar media for confirmatory testing. Results: In general, commercial lateral flow immunoassays resulted in fewer false-negative test results with enrichment of 1 g faecal samples in tetrathionate for 18 h, while all environmental sample enrichment techniques resulted in similar detection rates. The limit of detection from spiked samples, ∼4 colony-forming units/g, was similar for all methods evaluated. Conclusions: The lateral flow immunoassays evaluated could reliably detect S. enterica within 18 h, indicating that they may be useful for rapid point-of-care testing in equine practice applications. Additional evaluation is needed using samples from naturally infected cases and the environment to gain an accurate estimate of test sensitivity and specificity and to substantiate further the true value of these tests in clinical practice. Keywords: horse; faeces; culture; lateral flow immunoassay; Salmonella
Introduction Salmonella enterica is the most commonly reported cause of outbreaks of nosocomial infections in large animal veterinary teaching hospitals and the most common cause of closure of equine hospitals at these facilities [1]. Congregating horses from multiple sources, as is common at breeding farms, racetracks or equestrian events, is also associated with increased risks for spread of contagious microorganisms, such as Salmonella. Additionally, horses returning home from veterinary hospitals or other facilities can serve as a source of infection for others. Control of Salmonella is further complicated by the fact that subclinical infection and shedding are much more common than clinical infections, and horses can shed infectious doses of Salmonella in the absence of disease [2]. Furthermore, significant environmental contamination is inevitably present when Salmonella spreads between horses, whether as a cause or an effect, and it is well documented that environments in equine facilities that appear clean can still be contaminated with Salmonella [3–5]. The development of methods for point-of-care testing and performance of objective comparisons of Salmonella detection methodologies were recently identified as critical needs for infection control in equine populations by an international panel of infection control experts [6]. Rapid and reliable testing methods for S. enterica in environmental and faecal samples are considered essential for facilitating effective infection control in horse populations. Enriched aerobic culture and polymerase chain reaction (PCR) are currently the most common detection methods employed by veterinarians to identify Salmonella. These typically require 48–96 and 24–72 h from submission to reporting, respectively, in addition to the time needed for transportation of samples to the laboratory. Unfortunately, there are several overarching problems for these tests when used to detect Salmonella, including poorly standardised and laborious testing methodologies, poor sensitivity, the lengthy time needed to obtain results, limited availability of tests that can be used in practice settings and the costs for testing. Recently, commercial tests have been developed for use in food-safety microbiology that may provide a practical and useful alternative to Equine Veterinary Journal •• (2014) ••–•• © 2014 EVJ Ltd
traditional culture and PCR methods when practitioners are screening for Salmonella. Two of these commercial rapid diagnostic tests kits for S. enterica detection are marketed for point-of-use testing of a variety of food items including, but not limited to, raw ground beef and chicken, beef and pork skin, chicken rinse water and poultry feed [7,8]. These rapid tests are immunoassays that use antibodies specific for surface antigens of Salmonella and colloidal gold-antibody conjugates incorporated into a lateral flow test strip. They are commercially available from the manufacturers for relatively low cost, simple to use and require minimal equipment. In theory, these lateral flow immunoassays (LFIs) could be employed economically in point-of-care testing to provide results with reasonable sensitivity in 24 h for faecal samples and within 24–48 h for environmental samples. This would allow more extensive yet less expensive use of screening to control Salmonella in veterinary settings. Cultures would still be needed to provide antimicrobial susceptibility information and for follow-up testing (e.g. genetic characterisation) to allow identification of nosocomial transmission, but this could be limited to culture of samples found to be test positive with the rapid test strip. Use of a rapid Salmonella detection system represents a saving not only in cost but, just as importantly, in time. Early detection of animal shedding or environmental contamination would allow facility managers to implement measures more effectively to decrease animal and zoonotic infections. The objectives of this study were to evaluate different enrichment methods for use with 2 commercially available LFIs for use with equine faecal and environmental samples. The protocols that we evaluated were designed so that they might be easily implemented at veterinary practices to obtain test results within 24–48 h of sample collection.
Materials and methods Faecal samples A composite pool of faeces was used in each round of this study to provide a uniform sample matrix and microbiome. This was created by collecting
1
Rapid Salmonella detection
faeces from each of 5 horses considered to have a very low risk for Salmonella infection because they were members of a university-owned, isolated herd, which is tested periodically and was repeatedly culturenegative prior to study initiation. An S. enterica serotype Typhimurium isolate that had been previously recovered from an equine patient at the Colorado State University was used for inoculating faecal samples. A stock culture was made by inoculating this isolate into 10 ml of tryptic soy brotha and incubating overnight at 43°C. The bacterial concentration of the overnight culture was determined by plating 10-fold dilutions on blood agara, incubating overnight at 43°C, and calculating an estimate for the original culture concentration (∼4 × 108 colony-forming units [cfu]/ml). Tenfold dilutions of overnight broth cultures were made, and 1 ml/g were used to inoculate pooled 1 and 10 g faecal samples to achieve concentrations of ∼4 × 100 to 4 × 104 cfu/g of faeces; uninoculated faecal samples served as negative controls. Ten uninoculated control samples, 10 samples inoculated with 100, 101 and 102 cfu/g (30 total), and 5 samples inoculated with 103 and 104 cfu/g (10 total) were evaluated. All faecal samples were cultured using 2 different enrichment methods that were designed to detect Salmonella within 24 h when testing with LFIs. In the first method, 10 g faecal samples were enriched for 6 h in 100 ml of buffered peptone waterb (BPW) at 43°C, then 0.1 ml was passed into 10 ml of Rappaport-Vassiliadisc (RV) broth and incubated for 18 h at 43°C. In the second method, 1 g faecal samples were incubated in 10 ml of tetrathionateb (TET) for 18 h at 43°C. For purposes of comparison, samples from broth enrichment cultures were also streaked for isolation on xylose-lysine-tergitol 4d agar and incubated for an additional 24 h at 43°C. A positive plate had at least one colony with characteristic morphology. In general, higher inoculum concentrations resulted in greater numbers of characteristic colonies. One isolate from each plate was selected for serogroup confirmation using commercially available grouping antisera. The study was approved by the Colorado State University Institutional Animal Care and Use Committee. Consent from the Colorado State University Veterinary Teaching Hospital (CSU-VTH) was obtained for animal use in this study.
Environmental samples Environmental samples used in this experiment were collected from a hightraffic linoleum floor surface in an educational building adjacent to the CSUVTH that was previously found to have a very low frequency of S. enterica contamination (1.8 m × 6.7 m area per each sample) using an electrostatic dust wipee (Swiffer®) [9]. The same S. enterica serotype Typhimurium isolate was used for inoculating environmental samples. A stock culture was made by inoculating 10 ml of tryptic soy broth and incubating overnight at 43°C. The bacterial concentration of the overnight culture was determined by plating 10-fold dilutions on blood agar, incubating overnight at 43°C, and calculating the original culture concentration (∼4 × 108 cfu/ml). Tenfold dilutions of overnight broth culture were used to inoculate environmental samples to achieve concentrations of ∼4 × 100 to 4 × 103 cfu/ml BPW; uninoculated environmental samples served as controls. Five uninoculated control samples and 5 samples each inoculated with ∼4 × 100, ∼4 × 101, ∼4 × 102 and ∼4 × 103 cfu/ml (20 total) were evaluated. Four different enrichment methods were used prior to testing with LFIs. In the first method, collection wipes were enriched for 6 h in 100 ml of BPW at 43°C. In the second method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 0.1 ml was passed into 10 ml of RV broth and incubated for 18 h at 43°C. In the third method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 1 ml was passed into 9 ml TET for 18 h at 43°C. In the fourth method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 1 ml was passed into 9 ml TET for 18 h at 43°C, then 0.1 ml was passed into 10 ml RV broth for 18 h at 43°C. For purposes of comparison, samples from broth enrichment cultures were also streaked for isolation on xylose-lysine-tergitol agar and incubated for an additional 24 h at 43°C.
Sample testing Two different LFIs (Reveal® 2.0 [LFI-A] and RapidChek® Select™ [LFI-B])f,g were used to detect S. enterica in broth cultures of faecal samples and environmental samples as previously described. A positive test was defined as one where the test line had an equal or greater intensity as the
2
B. A. Burgess et al.
TABLE 1: Results of testing faeces experimentally inoculated with Salmonella enterica serotype Typhimurium using commercial immunoassays and aerobic culture
Enrichment method 10 g in BPW-RVa
1 g in TETb
Inoculum concentration (cfu/g)
Total tests
LFI-Ac,f
LFI-Bd,f
Culturee,g
Control ∼4 × 100 ∼4 × 101 ∼4 × 102 ∼4 × 103 ∼4 × 104 Control ∼4 × 100 ∼4 × 101 ∼4 × 102 ∼4 × 103 ∼4 × 104
10 10 10 10 5 5 10 10 10 10 5 5
0 0 0 2 4 4 0 6 6 10 5 5
0 0 0 3 3 4 0 6 6 10 5 5
0 2 6 5 5 5 0 6 6 10 5 5
Positive tests
a
Ten grams faecal samples enriched for 6 h in 100 ml of BPW at 43°C, then 0.1 ml was passed into 10 ml of RV broth and incubated for 18 h. bOne gram faecal samples incubated in 9 ml of TET for 18 h. cReveal® 2.0. dRapidChek® Select™. eOvernight aerobic culture on XLT4 agar at 43°C. fProportion of test positive results was significantly greater for 1 g samples enriched in TET compared to 10 g samples enriched in BPW-RV (P
Equine Veterinary Journal ISSN 0425-1644 DOI: 10.1111/evj.12234
Rapid Salmonella detection in experimentally inoculated equine faecal and veterinary hospital environmental samples using commercially available lateral flow immunoassays B. A. BURGESS, N. R. NOYES, D. S. BOLTE, D. R. HYATT, D. C. VAN METRE and P. S. MORLEY* Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, USA. *Corresponding author email: [email protected]; Received: 12.04.13; Accepted: 26.01.14
Summary Reasons for performing study: Salmonella enterica is the most commonly reported cause of outbreaks of nosocomial infections in large animal veterinary teaching hospitals and the closure of equine hospitals. Rapid detection may facilitate effective control practices in equine populations. Shipping and laboratory testing typically require ≥48 h to obtain results. Lateral flow immunoassays developed for use in food-safety microbiology provide an alternative that has not been evaluated for use with faeces or environmental samples. Objectives: We aimed to identify enrichment methods that would allow commercially available rapid Salmonella detection systems (lateral flow immunoassays) to be used in clinical practice with equine faecal and environmental samples, providing test results in 18–24 h. Study design: In vitro experiment. Methods: Equine faecal and environmental samples were inoculated with known quantities of S. enterica serotype Typhimurium and cultured using 2 different enrichment techniques for faeces and 4 enrichment techniques for environmental samples. Samples were tested blindly using 2 different lateral flow immunoassays and plated on agar media for confirmatory testing. Results: In general, commercial lateral flow immunoassays resulted in fewer false-negative test results with enrichment of 1 g faecal samples in tetrathionate for 18 h, while all environmental sample enrichment techniques resulted in similar detection rates. The limit of detection from spiked samples, ∼4 colony-forming units/g, was similar for all methods evaluated. Conclusions: The lateral flow immunoassays evaluated could reliably detect S. enterica within 18 h, indicating that they may be useful for rapid point-of-care testing in equine practice applications. Additional evaluation is needed using samples from naturally infected cases and the environment to gain an accurate estimate of test sensitivity and specificity and to substantiate further the true value of these tests in clinical practice. Keywords: horse; faeces; culture; lateral flow immunoassay; Salmonella
Introduction Salmonella enterica is the most commonly reported cause of outbreaks of nosocomial infections in large animal veterinary teaching hospitals and the most common cause of closure of equine hospitals at these facilities [1]. Congregating horses from multiple sources, as is common at breeding farms, racetracks or equestrian events, is also associated with increased risks for spread of contagious microorganisms, such as Salmonella. Additionally, horses returning home from veterinary hospitals or other facilities can serve as a source of infection for others. Control of Salmonella is further complicated by the fact that subclinical infection and shedding are much more common than clinical infections, and horses can shed infectious doses of Salmonella in the absence of disease [2]. Furthermore, significant environmental contamination is inevitably present when Salmonella spreads between horses, whether as a cause or an effect, and it is well documented that environments in equine facilities that appear clean can still be contaminated with Salmonella [3–5]. The development of methods for point-of-care testing and performance of objective comparisons of Salmonella detection methodologies were recently identified as critical needs for infection control in equine populations by an international panel of infection control experts [6]. Rapid and reliable testing methods for S. enterica in environmental and faecal samples are considered essential for facilitating effective infection control in horse populations. Enriched aerobic culture and polymerase chain reaction (PCR) are currently the most common detection methods employed by veterinarians to identify Salmonella. These typically require 48–96 and 24–72 h from submission to reporting, respectively, in addition to the time needed for transportation of samples to the laboratory. Unfortunately, there are several overarching problems for these tests when used to detect Salmonella, including poorly standardised and laborious testing methodologies, poor sensitivity, the lengthy time needed to obtain results, limited availability of tests that can be used in practice settings and the costs for testing. Recently, commercial tests have been developed for use in food-safety microbiology that may provide a practical and useful alternative to Equine Veterinary Journal •• (2014) ••–•• © 2014 EVJ Ltd
traditional culture and PCR methods when practitioners are screening for Salmonella. Two of these commercial rapid diagnostic tests kits for S. enterica detection are marketed for point-of-use testing of a variety of food items including, but not limited to, raw ground beef and chicken, beef and pork skin, chicken rinse water and poultry feed [7,8]. These rapid tests are immunoassays that use antibodies specific for surface antigens of Salmonella and colloidal gold-antibody conjugates incorporated into a lateral flow test strip. They are commercially available from the manufacturers for relatively low cost, simple to use and require minimal equipment. In theory, these lateral flow immunoassays (LFIs) could be employed economically in point-of-care testing to provide results with reasonable sensitivity in 24 h for faecal samples and within 24–48 h for environmental samples. This would allow more extensive yet less expensive use of screening to control Salmonella in veterinary settings. Cultures would still be needed to provide antimicrobial susceptibility information and for follow-up testing (e.g. genetic characterisation) to allow identification of nosocomial transmission, but this could be limited to culture of samples found to be test positive with the rapid test strip. Use of a rapid Salmonella detection system represents a saving not only in cost but, just as importantly, in time. Early detection of animal shedding or environmental contamination would allow facility managers to implement measures more effectively to decrease animal and zoonotic infections. The objectives of this study were to evaluate different enrichment methods for use with 2 commercially available LFIs for use with equine faecal and environmental samples. The protocols that we evaluated were designed so that they might be easily implemented at veterinary practices to obtain test results within 24–48 h of sample collection.
Materials and methods Faecal samples A composite pool of faeces was used in each round of this study to provide a uniform sample matrix and microbiome. This was created by collecting
1
Rapid Salmonella detection
faeces from each of 5 horses considered to have a very low risk for Salmonella infection because they were members of a university-owned, isolated herd, which is tested periodically and was repeatedly culturenegative prior to study initiation. An S. enterica serotype Typhimurium isolate that had been previously recovered from an equine patient at the Colorado State University was used for inoculating faecal samples. A stock culture was made by inoculating this isolate into 10 ml of tryptic soy brotha and incubating overnight at 43°C. The bacterial concentration of the overnight culture was determined by plating 10-fold dilutions on blood agara, incubating overnight at 43°C, and calculating an estimate for the original culture concentration (∼4 × 108 colony-forming units [cfu]/ml). Tenfold dilutions of overnight broth cultures were made, and 1 ml/g were used to inoculate pooled 1 and 10 g faecal samples to achieve concentrations of ∼4 × 100 to 4 × 104 cfu/g of faeces; uninoculated faecal samples served as negative controls. Ten uninoculated control samples, 10 samples inoculated with 100, 101 and 102 cfu/g (30 total), and 5 samples inoculated with 103 and 104 cfu/g (10 total) were evaluated. All faecal samples were cultured using 2 different enrichment methods that were designed to detect Salmonella within 24 h when testing with LFIs. In the first method, 10 g faecal samples were enriched for 6 h in 100 ml of buffered peptone waterb (BPW) at 43°C, then 0.1 ml was passed into 10 ml of Rappaport-Vassiliadisc (RV) broth and incubated for 18 h at 43°C. In the second method, 1 g faecal samples were incubated in 10 ml of tetrathionateb (TET) for 18 h at 43°C. For purposes of comparison, samples from broth enrichment cultures were also streaked for isolation on xylose-lysine-tergitol 4d agar and incubated for an additional 24 h at 43°C. A positive plate had at least one colony with characteristic morphology. In general, higher inoculum concentrations resulted in greater numbers of characteristic colonies. One isolate from each plate was selected for serogroup confirmation using commercially available grouping antisera. The study was approved by the Colorado State University Institutional Animal Care and Use Committee. Consent from the Colorado State University Veterinary Teaching Hospital (CSU-VTH) was obtained for animal use in this study.
Environmental samples Environmental samples used in this experiment were collected from a hightraffic linoleum floor surface in an educational building adjacent to the CSUVTH that was previously found to have a very low frequency of S. enterica contamination (1.8 m × 6.7 m area per each sample) using an electrostatic dust wipee (Swiffer®) [9]. The same S. enterica serotype Typhimurium isolate was used for inoculating environmental samples. A stock culture was made by inoculating 10 ml of tryptic soy broth and incubating overnight at 43°C. The bacterial concentration of the overnight culture was determined by plating 10-fold dilutions on blood agar, incubating overnight at 43°C, and calculating the original culture concentration (∼4 × 108 cfu/ml). Tenfold dilutions of overnight broth culture were used to inoculate environmental samples to achieve concentrations of ∼4 × 100 to 4 × 103 cfu/ml BPW; uninoculated environmental samples served as controls. Five uninoculated control samples and 5 samples each inoculated with ∼4 × 100, ∼4 × 101, ∼4 × 102 and ∼4 × 103 cfu/ml (20 total) were evaluated. Four different enrichment methods were used prior to testing with LFIs. In the first method, collection wipes were enriched for 6 h in 100 ml of BPW at 43°C. In the second method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 0.1 ml was passed into 10 ml of RV broth and incubated for 18 h at 43°C. In the third method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 1 ml was passed into 9 ml TET for 18 h at 43°C. In the fourth method, collection wipes were enriched for 6 h in 100 ml BPW at 43°C, then 1 ml was passed into 9 ml TET for 18 h at 43°C, then 0.1 ml was passed into 10 ml RV broth for 18 h at 43°C. For purposes of comparison, samples from broth enrichment cultures were also streaked for isolation on xylose-lysine-tergitol agar and incubated for an additional 24 h at 43°C.
Sample testing Two different LFIs (Reveal® 2.0 [LFI-A] and RapidChek® Select™ [LFI-B])f,g were used to detect S. enterica in broth cultures of faecal samples and environmental samples as previously described. A positive test was defined as one where the test line had an equal or greater intensity as the
2
B. A. Burgess et al.
TABLE 1: Results of testing faeces experimentally inoculated with Salmonella enterica serotype Typhimurium using commercial immunoassays and aerobic culture
Enrichment method 10 g in BPW-RVa
1 g in TETb
Inoculum concentration (cfu/g)
Total tests
LFI-Ac,f
LFI-Bd,f
Culturee,g
Control ∼4 × 100 ∼4 × 101 ∼4 × 102 ∼4 × 103 ∼4 × 104 Control ∼4 × 100 ∼4 × 101 ∼4 × 102 ∼4 × 103 ∼4 × 104
10 10 10 10 5 5 10 10 10 10 5 5
0 0 0 2 4 4 0 6 6 10 5 5
0 0 0 3 3 4 0 6 6 10 5 5
0 2 6 5 5 5 0 6 6 10 5 5
Positive tests
a
Ten grams faecal samples enriched for 6 h in 100 ml of BPW at 43°C, then 0.1 ml was passed into 10 ml of RV broth and incubated for 18 h. bOne gram faecal samples incubated in 9 ml of TET for 18 h. cReveal® 2.0. dRapidChek® Select™. eOvernight aerobic culture on XLT4 agar at 43°C. fProportion of test positive results was significantly greater for 1 g samples enriched in TET compared to 10 g samples enriched in BPW-RV (P
