TYPE Original Research PUBLISHED 20 October 2022 DOI 10.3389/fvets.2022.1011975 Senecavirus A seroprevalence OPEN ACCESS and risk factors in United States pig farms EDITED BY Enric M. Mateu, Universitat Autònoma de Barcelona, Spain REVIEWED BY Guilherme Preis1 , Juan M. Sanhueza1,2 , Carles Vilalta1,3,4 , Tereza Cristina Cardoso, Universidade Estadual de São Fabio A. Vannucci1,5 , Marie R. Culhane 1 and Cesar A. Corzo1* Paulo, Brazil 1 Jean-Pierre Frossard, Veterinary Population Medicine Department, College of Veterinary Medicine, University of Animal and Plant Health Agency, Minnesota, Saint Paul, MN, United States, 2 Departamento de Ciencias Veterinarias y Salud Pública, United Kingdom Facultad de Recursos Naturales, Universidad Católica de Temuco, Temuco, Chile, 3 Unitat Mixta Lok Raj Joshi, d’Investigació IRTA-UAB en Sanitat Animal, Centre de Recerca en Sanitat Animal, Campus de la Ragon Institute of MGH MIT and Universitat Autònoma de Barcelona, Bellaterra, Catalonia, Spain, 4 IRTA, Programa de Sanitat Animal, Harvard, United States Centre de Recerca en Sanitat Animal, Campus de la Universitat Autònoma de Barcelona, Bellaterra, Charles K. Nfon, Catalonia, Spain, 5 University of Minnesota Veterinary Diagnostic Laboratory, College of Veterinary National Centre for Foreign Animal Medicine, University of Minnesota, St. Paul, MN, United States Disease (NCFAD), Canada *CORRESPONDENCE Cesar A. Corzo Senecavirus A (SVA) is a non-enveloped, single-stranded, positive-sense RNA corzo@umn.edu virus belonging to the Picornaviridae family. Senecavirus A is constantly SPECIALTY SECTION associated with outbreaks of vesicular disease in pigs and has been reported This article was submitted to Veterinary Epidemiology and in several countries since its first large-scale outbreak in 2014. Senecavirus A’s Economics, clinical disease and lesions are indistinguishable from other vesicular foreign a section of the journal animal diseases (FAD). Therefore, an FAD investigation needs to be conducted Frontiers in Veterinary Science for every SVA case. For this reason, SVA has been attributed as the cause of RECEIVED 04 August 2022 ACCEPTED 30 September 2022 an alarming increase in the number of yearly FAD investigations performed PUBLISHED 20 October 2022 by the United States Department of Agriculture (USDA). The objectives of this CITATION study were to estimate the seroprevalence of SVA antibodies in breeding and Preis G, Sanhueza JM, Vilalta C, growing pig farms in the United States and to determine the farm-level risk Vannucci FA, Culhane MR and Corzo CA (2022) Senecavirus A factors associated with seropositivity. A total of 5,794 blood samples were seroprevalence and risk factors in collected from 98 and 95 breeding and growing pig farms in 17 states. A farm United States pig farms. characteristics questionnaire was sent to all farms, to which 80% responded. Front. Vet. Sci. 9:1011975. doi: 10.3389/fvets.2022.1011975 The responses were used to conduct logistic regression analyses to assess COPYRIGHT the risk factors associated with SVA seropositivity. The estimated farm-level © 2022 Preis, Sanhueza, Vilalta, seroprevalences were 17.3% and 7.4% in breeding and growing pig farms, Vannucci, Culhane and Corzo. This is respectively. Breeding farms had 2.64 times higher odds of SVA seropositivity an open-access article distributed under the terms of the Creative than growing pig farms. One key risk factor identified in breeding farms was Commons Attribution License (CC BY). the practice of rendering dead animal carcasses. However, the adoption of a The use, distribution or reproduction in other forums is permitted, provided higher number of farm biosecurity measures was associated with a protective the original author(s) and the copyright effect against SVA seropositivity in breeding farms. owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted KEYWORDS academic practice. No use, distribution senecavirus A, seroprevalence, risk factors, vesicular disease, swine or reproduction is permitted which does not comply with these terms. Introduction Senecavirus A (SVA), previously known as Seneca Valley Virus (1), is a non- enveloped, single-stranded, positive-sense RNA virus belonging to the Picornaviridae family and the only member of the genus Senecavirus (2). The virus was first isolated in 2002 as a contaminant from PER.C6 cell lines, presumably introduced via fetal bovine Frontiers in Veterinary Science 01 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 serum or porcine trypsin during the cell cultivation process Experimental design (2). Current data suggest that SVA existed in the United States (U.S.) swine population for at least three decades (3). However, A cross-sectional study was designed and conducted the presence of SVA in pigs with vesicular lesions had not to estimate the seroprevalence of SVA in U.S. pig farms. been reported until 2007 after a trailer transporting 187 market Participation in the study was voluntary. Major veterinary clinics hogs from the Canadian province of Manitoba arrived at a and production systems throughout the country were invited harvest facility in Minnesota, U.S. (4). Larger multi-state SVA to participate. After agreeing to participate, each production outbreaks of this vesicular disease in pigs were reported in system or veterinary clinic was asked to select breeding and Brazil in 2014 (5), in the U.S. (6, 7), and China (8) in 2015, growing pig farms for sample collection randomly. Both followed by Colombia (9), Thailand (10) in 2016, Vietnam (11) participating veterinarians and, in some cases, investigators in 2018, and Mexico (12) and Chile (13) in 2022. Characteristic collected the study samples. vesicular lesions usually start developing approximately 4 days post-infection, consisting of multiple-sized vesicles on the snout, oral cavity, and feet (i.e., coronary band and Sample size calculation interdigital space), which may lead to lameness and lethargy. Vesicles tend to rupture 5 days post-infection (14, 15) and Number of farms are clinically indistinguishable from high-consequence foreign To calculate the number of breeding and growing pig farms animal diseases (FAD) such as foot-and-mouth disease (FMD). to be included in the study, the following formula was used (17): In neonatal pigs, diarrhea and a sudden increase in pre-weaning Z2 pq piglet mortality have also been reported (5–7), contributing to N= production losses. L2 The clinical disease and lesions caused by SVA are where N = number of farms to be sampled, Z = 1.96 indistinguishable from other vesicular animal diseases: swine (Z-score value for 95% confidence), p = expected true farm-level vesicular disease, vesicular exanthema of swine, vesicular prevalence (50% was used as the default as data was not available stomatitis, and FMD. Therefore, an FAD investigation needs at the time of the study), q = 0.5 (1 – p), L = precision of the to be conducted by local government authorities for every estimate was set at 0.1. A total of 97 breeding and 97 growing SVA case. This practice, while prudent, has resulted in an pig farms were needed for this study, which brings to a total of increase in false alarms for FADs in the United States. The 194 farms. average yearly number of FAD investigations conducted by the United States Department of Agriculture (USDA) in all animal species in the U.S. between 2008 and 2014 was Number of pigs sampled per farm 487. This number increased almost four times between 2015 The number of samples needed within each farm to classify and 2020, with an average of 1,808 FAD investigations per the farm as either seropositive or seronegative was calculated year. Around 75% of FAD investigations were attributed to using the following formula (18): swine vesicular disease in pigs in the last 4 years of this log(1 − C) period (16). N= log (1 − TP) Despite the considerable number of SVA outbreak reports in swine farms across the past few years and the where N = number of animals to be sampled in each farm, associated problems, the epidemiology of this disease is C = 0.95 (confidence of 95%), TP = 0.1 (assuming that the poorly understood. Basic information such as prevalence expected true within-farm prevalence was 10%). Therefore, and risk factor studies are scarce. Therefore, the objectives 29 blood samples were necessary to reach a 95% confidence of this study were to (1) estimate the seroprevalence of level that at least one positive sample would be detected when SVA antibodies in breeding and growing pig farms in the the within-farm prevalence was at least 10%, assuming perfect U.S. and (2) determine the farm-level risk factors associated sensitivity and specificity. with seropositivity. Sample collection, handling, and testing Materials and methods In breeding farms, sampling was performed randomly across sow parities. In the case of growing pig farms, samples were The University of Minnesota (UMN) Institutional Animal collected from 20-weeks-old or older pigs to avoid the possible Care and Use Committee (IACUC) approved this study detection of maternal antibodies (19). Blood samples were (protocol 1804-35818A). collected, refrigerated, and shipped to the UMN Food Centric Frontiers in Veterinary Science 02 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 Corridor Laboratory. Blood samples were sorted, organized, Assessment of farm-level factors associated de-identified upon reception, and submitted to the UMN with seropositivity Veterinary Diagnostic Laboratory for testing. Senecavirus A IgG The risk factor analyses for breeding and growing-pig farms presence was tested through an immunofluorescent antibody were conducted separately. Univariable logistic regressions were test (IFA), which was reported to have 90% and 100% diagnostic fit to determine the unconditional associations between all risk sensitivity (Se) and specificity (Sp), respectively (20). Briefly, factors recorded in the questionnaires and the outcome (SVA NCI-H1299 (ATCC R CRL-5803TM ) cells were inoculated with farm seropositivity). SVA and fixed with cold acetone. Sera samples were screened Linearity between continuous variables and the outcome for SVA-specific IgG in two dilutions (1:40 and 1:80) using in the logit scale was assessed visually using scatterplots and PBS (Gibco). After incubating for 60–75 min, plates had the statistically. If the relationship between the continuous variable sera removed and washed with PBS. DyLight R 650 anti-pig and the outcome was not linear, continuous variables were IgG (Abcam, Cambridge, MA) was added to the wells, and categorized based on their median values (less than/equal to the plates were incubated for 60–75 min. Plates were washed with median vs. greater than the median). PBS and observed under fluorescence microscopy (20) by the Variables with multiple categories where all positive cases same laboratory technician. Fluorescence observed at a sample were within the same category or had categories with few dilution of 1:40 or 1:80 indicated that the serum sample was observations and no cases (indicating a lack of substantial positive for SVA IgG antibodies. variability for the analysis) were either excluded from the analysis or had their observations regrouped in a new two-factor categorical variable. Farm characteristics questionnaire A new variable was created to evaluate the association between biosecurity measures and SVA-seropositivity. The Two questionnaires—one for breeding and another list of biosecurity measures included in the survey was (1) for growing pig farms—were designed to capture general Visitor check-in required to enter the farm, (2) Shower in/out information such as farm type, farm size, personnel flow, animal procedures, (3) Danish bench-entry system is installed, (4) Use sourcing, and other details on biosecurity measures. The survey of farm-specific boots are required, (5) Use of farm-specific was electronically sent to all participating veterinarians to clothing is required, and (6) A downtime is required before answer on a per farm basis. Data obtained on the questionnaires entering the farm. Since all farms responded to either having were then transcribed to an electronic spreadsheet (Microsoft or not having these six different biosecurity measures in place, Excel 2016, Microsoft Corporation) for analysis. they were categorized as having “four or less” or “five or six” biosecurity measures in place if they responded to having any combination of ≤4 or ≥5 biosecurity measures, respectively. Unconditional associations between each predictor Data analysis variable and the outcome were tabulated. Only variables with associated p-values below 0.2 were selected for inclusion in Estimating farm-level SVA seroprevalence the multivariable analysis. A backward elimination process The proportions of seropositive breeding and growing pig was used to build the final multivariable logistic model. First, farms were estimated after calculating the cut-point number of a maximum model was fit using all the previously screened positive samples needed to classify a farm as being positive. variables. Variables were then removed one at a time, and the The cut-point number of positive samples was determined by likelihood ratio test was used to compare the nested models maximizing herd sensitivity (HSe) and herd specificity (HSp) until a model with a maximum likelihood was found. values (21), based on the SVA IgG IFA antibody test’s Se and All statistical analyses were performed using R statistical Sp (90% and 100%, respectively) (20). Clopper-Pearson 95% software (22). confidence intervals for the proportions of seropositive farms were also calculated. Results Association between farm type and SVA This study involved pig farms from nine production systems, seropositivity eight veterinary clinics, and two private practitioners. Thirty-six A chi-square test of independence was used to determine if swine veterinarians contributed to this study by collecting 5,794 there was a significant association between farm type (breeding blood samples from 193 farms. The overall survey response or grow-finishing pig farms) and SVA seropositivity. The odds rate was 80% and included data from 155 (77 breeding and 78 ratio and 95% CI for SVA seropositivity between farm types growing pig farms) out of 193 tested farms, including all positive were calculated using the unconditional maximum likelihood breeding farms and six out of the seven positive growing- estimation method (Wald). pig farms. Frontiers in Veterinary Science 03 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 TABLE 1 Demographic characteristics of the 193 United States (U.S.) the 1:80 dilution and 4.9% (13/268) were positive at the 1:40 pig farms participating in the study. dilution. Twenty-four out of 193 (12.4%) sampled farms had at Total Breeding Growing-pig least one seropositive serum sample. The median, mean, and samples farms farms standard deviation for the number of positive samples within n = 193 n = 98 n = 95 positive farms were 6.5, 11.2, and 10.1, respectively. Overall, the proportion of IFA-positive sera samples from Responded to survey 155 (80%) 77 (79%) 78 (82%) breeding farms was 5.9% (174/2,943). Seventeen out of 98 Farm sizea (17.3%, 95%CI: 10.4, 26.3%) breeding farms had at least one Range – 120–9,600 800–55,194 positive sample and were located in Illinois, Indiana, Kansas, (Minimum–Maximum) Minnesota, North Carolina, and Texas. The overall estimates Median – 2,752 3,600 of SVA farm level seroprevalence among breeding farms in Mean (S.D.) – 3,147 (1,884) 4,922 (6,518) different states or regions are shown in Figure 1. Among Companiesb seropositive breeding farms, the median, mean, and standard Number of participating 19 19 17 deviation of the number of positive samples were 4, 10.2, and companies 9.8, respectively (Figure 2). The average within-farm apparent Median number of 10 4 5 prevalence among seropositive breeding farms was 34% (95% sampled farms per CI: 17.8, 53.5%). company Seven out of 95 (7.4%, 95% CI: 3, 14.6%) growing pig farms Mean number of 10 (7) 5 (3) 6 (4) had at least one positive sample and these were detected in sampled farms per the states of Illinois, Indiana, Minnesota, North Carolina, and company (SD) Oklahoma. The proportion of IFA-positive sera samples from U.S. States growing-pig farms was 3.3% (94/2,851). The overall estimates Number of participating 17 16 11 of SVA farm level seroprevalence among growing pig farms states in different states or regions are shown in Figure 3. The Median number of 5 4 6 median, mean, and standard deviation of the number of positive sampled farms per Statee samples within positive growing pig farms were 8, 13.4, and Mean number of 10 (10) 5 (5) 7 (6) 11.3, respectively (Figure 2). The average within-farm apparent sampled farms per State prevalence among seropositive growing pig farms was 44.7% (SD) (95% CI: 26.6, 63.8%). a Datafrom 75 and 76 participating breeding and growing-pig farms, respectively. b Data from all participating production companies, veterinary clinics, and two private practitioners. Association between farm type and SVA seropositivity Blood samples were collected from 193 participating farms: 98 and 95 breeding and growing pig farms located in 17 different A significant association between farm type and SVA states (Table 1). Recruitment and sample collection at all farms seropositivity was detected (X 2 = 4.411, df = 1, p = 0.035). occurred between October 2018 and October 2019. Breeding farms had 2.64 (95% CI: 1.04, 6.69) times higher odds of SVA seropositivity when compared to growing pig farms. Classification of farm status The number of positive samples needed to classify a farm Risk factors associated with SVA as seropositive was 1. This cut-off value maximized the HSp seropositivity and HSe values, which reached 100% and 94%, respectively. Changing the cut-off value to 2 or 3 did not alter HSp, but HSe After tabulation of unconditional associations between the decreased to 77% and 51%, respectively. surveyed predictors and the outcome of SVA seropositivity, six breeding and six grow-finishing pig farm predictors were selected to enter the multivariable model selection process. Seroprevalence results The results from the univariable and multivariable logistic regression analysis for the breeding farms’ characteristics are The overall proportion of IFA-positive sera samples from shown in Table 2. The final multivariable logistic regression breeding and growing pig sites was 4.6% (268/5,794). Of the model showed that breeding farms that reported rendering 268 IFA-positive sera samples, 95.1% (255/268) were positive at dead animal carcasses were more likely to be SVA-seropositive Frontiers in Veterinary Science 04 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 FIGURE 1 Estimated proportions and 95% confidence intervals of SVA-seropositive breeding farms by state, region, and national estimate. (3) all trucks that arrive in the farm are cleaned and disinfected (Table 3). Discussion The present study demonstrates that SVA antibodies exist in the U.S. swine population. To the authors’ knowledge, FIGURE 2 this is the first national study designed and sampled to Box and whisker plot of the number of SVA IFA-positive samples estimate the seroprevalence of SVA. Despite the high by pig farm type in the U.S. incidence of swine-vesicular FAD investigations (16), the estimated farm level apparent seroprevalences of 17.3% (95%CI: 10.4, 26.3%) and 7.4% (95% CI: 3, 14.6%) among U.S. breeding and growing pig farms, respectively, were (OR = 9.2, CI: 2.5, 33.7), while farms that reported practicing relatively low. These proportions change slightly when five or six different biosecurity measures were less likely to be accounting for the imperfect HSe estimate of the applied SVA-seropositive (OR = 0.2, CI: 0.1, 0.99). A summary of the methodology. Considering the calculated HSe (94%) and HSp biosecurity measures and associations is shown in Table 2. (100%), breeding and growing pig farms had estimated true It was not possible to fit a multivariable logistic regression seroprevalences of 18.5% (95%CI: 11.1, 28%) and 7.8% (95%CI: model for the growing-pig farms due to the low number of 3.2, 15.5%). Although slight numerical increases are seen seropositive farms. The risk factors in growing-pig farms that when comparing the apparent and true prevalence estimates, appeared to be positively associated with SVA-seropositivity there are no significant changes due to the overlapping (p < 0.2) in the univariable analysis were (1) pigs are loaded into confidence intervals. trucks by an external pig-loading crew, and (2) More than one Currently, there is scarce information on the serological external crew is hired to perform jobs at the farm. Alternatively, response to SVA at a population level. In a recent study, SVA IgG the risk factors that appeared to be negatively associated with was detected in a cohort of 60 sows from a 6,000-sow farrow- SVA-seropositivity (p < 0.2) were (1) people that load pigs into to-wean farm that underwent an SVA outbreak for up to 13 trucks have direct access to pigs in the barn after loading a truck, months after the outbreak, using the same IFA procedure (23). (2) all pigs in the farm are sourced by a single breeding farm, and This suggests that antibodies can be detected for an extended Frontiers in Veterinary Science 05 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 FIGURE 3 Estimated proportions and 95% confidence intervals of SVA-seropositive growing-pig farms, by state, region, and national estimate. TABLE 2 Univariable and multivariable logistic regression analysis of the risk factors associated with SVA seropositivity in U.S. breeding farms. Univariable Multivariable Characteristic OR 95%CI p-values OR 95%CI p-values Carcass disposal method Composting, burying, or incinerating – – – – – – Rendering 7.9 2.4–26.7 <0.001 9.2 2.5–33.7 <0.001 Biosecurity measures in place* Four or less – – – – – – Five or six 0.3 0.1–1.1 0.06 0.2 0.1–0.99 0.49 At least one employee works on another farm No – – – – – – Yes 0.2 0.0–1.7 0.1 – – – Type of manure storage Uncovered lagoon – – – – – – Deep pits 0.2 0.1–0.8 0.02 – – – Water treatment No – – – – – – Yes 0.2 0.0–0.9 0.03 – – – Cull sows and weaned piglets use the same ramp when truck loading No – – – – – – Yes 0.4 0.1–1.4 0.16 – – – *Farms were categorized as having “four or less” or “five or six” biosecurity measures in place if they responded to having any combination of ≤4 or ≥5 biosecurity measures in place, respectively, from a list of six biosecurity measures included in the survey. period after exposure. Therefore, the IgG detection in this study It is currently unknown what may have caused the onset of is likely the result of naturally-infected breeding and growing pig large-scale SVA outbreaks after 2014–2015. The virus is likely to farms, even if exposure happened a long time before sampling, as have been circulating within and between U.S. pig farms since at maternally derived antibodies may be undetectable after 6 weeks least 1988, as was shown by the sequence analysis of picorna-like of age (19). viruses isolated from pigs in the U.S. (3). Conversely, another Frontiers in Veterinary Science 06 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 TABLE 3 Univariable logistic regression analysis of the risk factors clinics, regardless of their SVA or other infectious diseases associated with SVA seropositivity in U.S. growing pig farms. status. However, in the study by Houston et al. (25), the source Univariable population was pig farms conducting porcine reproductive and respiratory syndrome virus (PRRSv) monitoring at one Characteristic OR 95%CI p-values Veterinary Diagnostic Laboratory with no known history of SVA. It is currently unknown if the presence of other infectious External pig-loading crewa diseases (e.g., PRRS) could be associated with the presence No – – – of SVA, which may have introduced potential biases. It is Yes 9.3 1.03–84.9 0.047 possible that biosecurity failures in PRRSv-positive farms are Direct access to pigs in the barn after loading a truckb also responsible for the introduction of SVA; thus, assessing the No – – – prevalence of SVA exposure exclusively in farms monitoring Yes 0.1 0.01–0.8 0.035 for PRRSv is not appropriate. The parallel interpretation of Hires more than one external crewc two different serological tests with fair-to-moderate results No – – – agreement by Houston et al. (25) may have overestimated the Yes 3.3 0.6–19.2 0.188 proportion of positive farms, partially explaining the significant Single-sourced pigsd differences between both studies. No – – – Very little is known about how SVA transmits between Yes 0.3 0.05–1.6 0.158 farms. Senecavirus A-infected animals appear to develop a All trucks come clean and disinfectede short-term viremia for up to 10 days post-infection, and shed No – – – the virus for up to 28 days post-infection in oral/nasal secretions Yes 0.3 0.05–1.7 0.165 and feces (14). While this information can help us mitigate a Market pigs are loaded into trucks by an external pig-loading crew. transmission between animals on a farm, more information is b People who load pigs into trucks have direct access to pigs in the barn after loading needed to prevent the infection of pig farms in the first place. a truck. c More than one external crew is hired to work at the farm. To shed some light on this matter, we performed a risk-factor d All pigs on the farm are sourced from a single breeding farm. analysis to identify what farm characteristics might be associated e All trucks that arrive on the farm are cleaned and disinfected. with SVA exposure. Implementation of biosecurity measures in breeding farms yielded a sparing effect in this study. While not surprising, it retrospective study attempted to assess the presence of SVA does remind the industry of the importance and needs that in Brazil through the serological testing of samples collected most modern pig farms have when adopting such preventive between 2007 and 2016 (24). The authors concluded that SVA measures to avoid the introduction of new pathogens carried was likely absent in the major Brazilian pig-producing states by people themselves or the boots and clothes they are wearing before 2014. However, the reported results must be interpreted (26). However, rendering was another predictor in the model carefully since a low number of samples were tested and that was found to have a significant association with SVA collected from asymptomatic farms for other research purposes seropositivity. Breeding farms that reported disposing of the not related to vesicular diseases. Therefore, the study design carcasses of dead animals via rendering had 9.2 higher odds of likely introduced a selection bias that significantly reduced being seropositive compared to farms that either compost, bury, the probability of detecting SVA-exposed animals. A more or incinerate the dead animals. One possible causal pathway comprehensive study design is needed to rule out the presence of for this association is the indirect transmission of the pathogen SVA among Brazilian pig farms before 2014. It may be possible between farms through the trucks transporting the carcasses that SVA can remain present and undetected in pig populations since the truck may need to visit several farms before filling and until a formal vesicular disease investigation is conducted and returning to the rendering plant. Similar associations involving SVA is ruled out. the disposal of dead animals via rendering have been reported The results from this study differ significantly from the in other studies, such as with the increased risk of respiratory results of another seroprevalence study conducted in U.S. pig disease outbreaks in pig farms (27), PRRSv positivity (28), farms using samples collected in 2016 (25). The estimated farm- porcine epidemic diarrhea positivity (Morrison Swine Health level seroprevalences were 75.8% in breeding farms and 42.7% Monitoring Project science page, personal communication), in growing pig farms vs. 17.3% and 7.4% in this study. The H5N2 highly pathogenic avian influenza virus (29), and H7N2 discrepancies in the proportions reported in both studies may be avian influenza virus in commercial poultry farms in the explained by fundamental differences in the study designs, time United States (30). More studies are needed to understand periods when samples were collected, and interpretation of the whether other carcass disposal methods should be considered to serological assays. This study’s source population was U.S. pig reduce the probability of introducing SVA or any other pathogen farms from major swine-producing companies and veterinary to the farm. Frontiers in Veterinary Science 07 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 Due to the low number of positive observations, it was not These findings may be applied in pig farms to help reduce possible to build a multivariable logistic regression model for the risk of SVA exposure. Other carcass disposal methods the growing pig farms. The positive univariable associations could be considered, such as composting or incineration, or between “external pig loading crew” and “hires more than the dead-animal disposal areas should be located away from one external crew” with the outcome of seropositivity (Table 3) the farms, and the trucks used for carcass collection should highlight the potential role of people in the introduction be prohibited from coming close to the barns. Furthermore, of pathogens. As for the protective associations, it is not attention should be given to biosecurity measures to reduce the surprising that single-sourcing of weaned pigs and disinfecting risk of pathogen introduction through any incoming personnel all incoming trucks would decrease the odds of seropositivity or fomites. Although this is the first assessment of farm-level risk since such measures prevent the comingling of pigs from factors associated with SVA seropositivity, more studies need to negative and positive populations and the cross-contamination be specifically designed to understand these associations. between different batches of animals, respectively. However, the statistically significant association of being at lesser odds of positivity when people had direct contact with pigs in the Data availability statement barn after loading pig trucks is unexpected and challenging to The raw data supporting the conclusions of this article will explain. Upon further investigating this artifact of the analysis, be made available by the authors upon reasonable request. it was seen that the only farms where people went back into the barns and had contact with the remaining pigs were the ones that did not hire an external pig-loading crew. All farms that Ethics statement hired an external pig-loading crew reported that people left the farms after loading the trucks without contacting the remaining The animal study was reviewed and approved by pigs. Therefore, it is likely that this association is measuring a Institutional Animal Care and Use Committee, University similar effect as the association with the farms that hire external of Minnesota, Minneapolis, United States. Written informed pig-loading crews. consent was obtained from the owners for the participation of Although the current study design is not optimal their animals in this study. for estimating within-farm prevalences, a broad range of SVA-positive sera samples was detected. The range of positive samples per positive farm was somewhat similar between the Author contributions breeding and growing pig farms (Figure 2), with an average of 10.2 and 13.4 out of 30 tested samples in breeding and growing JS, CV, and CC wrote the research proposal. GP and CC pig farms, respectively (Figure 2). As reported in the results developed the farm-characteristics questionnaire and contacted section, the estimated within-farm prevalence yielded wide the veterinarians for sample collection and questionnaire confidence intervals due to the reduced sample size per farm. response. MC helped with the questionnaire preparation. GP However, these results may still provide helpful information for compiled all test results and questionnaire responses and further investigations. Nevertheless, interpreting these results analyzed the data with JS and CC. FV contributed to the is rather difficult in cross-sectional studies since there is no diagnostic data interpretation. GP wrote the initial draft of information about the previous SVA history on tested farms. the manuscript. All authors read, edited, and approved the finalized manuscript. Conclusion Funding This is the first study specifically designed to estimate the This research was funded by the National Pork seroprevalence of SVA at a national level, with a broad selection Board (NPB#18-041). of farms from producing companies and veterinary clinics as the source population. After sampling and testing 5,794 sera samples from 98 breeding and 95 growing pig farms, it was Acknowledgments shown that SVA antibodies are present among U.S. pig farms. Seroprevalence was higher in breeding farms than in growing We thank all participating swine veterinarians, farm staffs, pig sites. and swine-producing companies for participating in this project Key risk factors identified were the rendering of dead and their assistance in collecting the samples and responding animals and access of external working crews to the farms. to the questionnaires. We thank the UMN-VDL staff for their At the same time, the implementation of biosecurity measures expertise and assistance with serology. We thank the National seemed to have a protective effect against SVA seropositivity. Pork Board for providing the funds for this study. Frontiers in Veterinary Science 08 frontiersin.org Preis et al. 10.3389/fvets.2022.1011975 Conflict of interest Publisher’s note The authors declare that the research was conducted in All claims expressed in this article are solely those of the the absence of any commercial or financial relationships authors and do not necessarily represent those of their affiliated that could be construed as a potential conflict organizations, or those of the publisher, the editors and the of interest. reviewers. Any product that may be evaluated in this article, or The handling editor EM declared a shared affiliation with the claim that may be made by its manufacturer, is not guaranteed author CV at the time of review. or endorsed by the publisher. References 1. Adams MJ, Lefkowitz EJ, King AMQ, Bamford DH, Breitbart M, Davison 17. Dohoo I, Martin W, Stryhn H. Methods in Epidemiologic Research. Edited by AJ, et al. Ratification vote on taxonomic proposals to the International S. M. McPike. Charlottetown: VER Inc. (2012). Committee on Taxonomy of Viruses (2015). Arch Virol. (2015) 160:1837– 50. doi: 10.1007/s00705-015-2425-z 18. Christensen J, Gardner IA. Herd-level interpretation of test results for epidemiologic studies of animal diseases. Prev Vet Med. (2000) 45:83– 2. Hales LM, Knowles NJ, Reddy PS, Xu L, Hay C, Hallenbeck PL. 106. doi: 10.1016/S0167-5877(00)00118-5 Complete genome sequence analysis of Seneca Valley virus-001, a novel oncolytic picornavirus. J Gen Virol. (2008) 89:1265–75. doi: 10.1099/vir.0.83570-0 19. Gimenez-Lirola LG, Rademacher C, Linhares D, Harmon K, Rotolo M, Sun Y, et al. Serological and molecular detection of senecavirus A 3. Knowles NJ, Hales LM, Jones BH, Landgraf JG, House JA, Skele KL, et al. associated with an outbreak of swine idiopathic vesicular disease and Epidemiology of Seneca valley virus: identification and characterization of isolates neonatal mortality. J Clin Microbiol. (2016) 54:2082–9. doi: 10.1128/jcm.00 from pigs in the United States. In: abstr G2. EUROPIC 2006: XIVth Meet Eur Study 710-16 Group Mol Biol Picornaviruses, Saariselka, Inari, Finland. (2006). Available online at: https://www.aasv.org/SVV/SVVUS.pdf (accessed October 5, 2022). 20. Goolia M, Vannucci F, Yang M, Patnayak D, Babiuk S, Nfon CK. Validation of a competitive ELISA and a virus neutralization test for 4. Pasma T, Davidson S, Shaw SL. Idiopathic vesicular disease in swine in the detection and confirmation of antibodies to senecavirus A in swine Manitoba. Can Vet J. (2008) 49:84–5. sera. J Vet Diagnostic Investig. (2017) 29:250–3. doi: 10.1177/10406387166 5. Vannucci FA, Linhares DCL, Barcellos DESN, Lam HC, Collins J, Marthaler 83214 D. Identification and complete genome of Seneca valley virus in vesicular fluid and 21. Sergeant E. Epitools Epidemiological Calculators. Ausvet (2018). Available sera of pigs affected with idiopathic vesicular disease, Brazil. Transbound Emerg online at: http://epitools.ausvet.com.au (accessed October 5, 2022). Dis. (2015) 62:589–93. doi: 10.1111/tbed.12410 22. R Core Team. R: A Language and Environment for Statistical Computing. 6. Joshi LR, Mohr KA, Clement T, Hain KS, Myers B, Yaros J, et al. Detection (2022). Available online at: http://www.r-project.org/ (accessed October 5, 2022). of the emerging picornavirus senecavirus A in pigs, mice, and houseflies. J Clin Microbiol. (2016) 54:1536–45. doi: 10.1128/jcm.03390-15 23. Preis G, Benjamin N, Corzo C. Detection of senecavirus A and IgG antibodies over time after an outbreak in a breeding herd: a case report. 7. Canning P, Canon A, Bates JL, Gerardy K, Linhares DCL, Piñeyro PE, et al. In: 2020 Allen D. Leman Swine Research Abstracts. Saint Paul, MN (2020). Neonatal mortality, vesicular lesions and lameness associated with senecavirus A in Available online at: https://sites.google.com/a/umn.edu/leman-swine-conference/ a US sow farm. Transbound Emerg Dis. (2016) 63:373–8. doi: 10.1111/tbed.12516 home/2020/2020-research-abstracts 8. Wu Q, Zhao X, Bai Y, Sun B, Xie Q, Ma J. The first identification and complete 24. Saporiti V, Fritzen JTT, Feronato C, Leme RA, Lobato ZIP, Alfieri genome of senecavirus A affecting pig with idiopathic vesicular disease in China. AF, et al. A ten years (2007–2016) retrospective serological survey Transbound Emerg Dis. (2017) 64:1633–40. doi: 10.1111/tbed.12557 for Seneca Valley virus infection in major pig producing states of 9. Sun D, Vannucci F, Knutson TP, Corzo C, Marthaler DG. Emergence and Brazil. Vet Res Commun. (2017) 41:317–21. doi: 10.1007/s11259-017- whole-genome sequence of Senecavirus A in Colombia. Transbound Emerg Dis. 9697-6 (2017) 64:1346–9. doi: 10.1111/tbed.12669 25. Houston E, Giménez-Lirola LG, Magtoto R, Mora-Díaz JC, Baum D, 10. Saeng-chuto K, Rodtian P, Temeeyasen G, Wegner M, Nilubol D. The first Piñeyro PE. Seroprevalence of Senecavirus A in sows and grower-finisher pigs detection of Senecavirus A in pigs in Thailand, 2016. Transbound Emerg Dis. (2018) in major swine producing-states in the United States. Prev Vet Med. (2019) 65:285–8. doi: 10.1111/tbed.12654 165:1–7. doi: 10.1016/j.prevetmed.2019.01.012 11. Arzt J, Bertram MR, Vu LT, Pauszek SJ, Hartwig EJ, Smoliga GR, et al. First 26. Alarcón LV, Alberto AA, Mateu E. Biosecurity in pig farms: a detection and genome sequence of senecavirus A in Vietnam. Microbiol Resour review. Porc Heal Manag. (2021) 7:1–15. doi: 10.1186/s40813-020-0 Announc. (2019) 8:17–8. doi: 10.1128/mra.01247-18 0181-z 12. CPA/SENASICA Gobierno de Mexico. AVISE 21 Boletin Informativo de la 27. Rose N, Madec F. Occurrence of respiratory disease outbreaks in fattening CPA. Ciudad de Mexico, Mexico (2022). Available online at: https://dj.senasica. pigs: relation with the features of a densely and a sparsely populated pig area in gob.mx/Contenido/files/2022/marzo/FEBRERO2022_ff602385-52bf-485c-a39d- France. Vet Res. (2002) 33:179–90. doi: 10.1051/vetres 00c5dd9986b3.pdf (accessed October 5, 2022). 28. Lambert MÈ, Arsenault J, Poljak Z, D’Allaire S. Epidemiological 13. Ministerio de Agricultura/Gobierno de Chile. Informar Sobre Deteccion Virus investigations in regard to porcine reproductive and respiratory syndrome Seneca en Chile. Santiago, Chile (2022). Available online at: https://ceropapel.sag. (PRRS) in Quebec, Canada. Part 2: prevalence and risk factors in breeding gob.cl/validar/?key=121105740&hash=8cc10 (accessed October 5, 2022). sites. Prev Vet Med. (2012) 104:84–93. doi: 10.1016/j.prevetmed.2011. 11.002 14. Joshi LR, Fernandes MHV, Clement T, Lawson S, Pillatzki A, Resende TP, et al. Pathogenesis of senecavirus a infection in finishing pigs. J Gen Virol. (2016) 29. Garber L, Bjork K, Patyk K, Rawdon T, Antognoli M, Delgado A, et al. 97:3267–79. doi: 10.1099/jgv.0.000631 Factors associated with highly pathogenic avian influenza H5N2 infection on 15. Maggioli MF, Lawson S, de Lima M, Joshi LR, Faccin TC, Bauermann FV, table-egg layer farms in the Midwestern United States, 2015. Avian Dis. (2016) et al. Adaptive immune responses following senecavirus A infection in pigs. J Virol. 60:460–6. doi: 10.1637/11351-121715-Reg (2017) 92:1–17. doi: 10.1128/jvi.01717-17 30. McQuiston JH, Garber LP, Porter-Spalding BA, Hahn JW, Pierson 16. United States Department of Agriculture. Calendar Year 2020 Update: FAD FW, Wainwright SH, et al. Evaluation of risk factors for the spread of Investigation Report. Riverdale, MD (2021). Available online at: https://www. low pathogenicity H7N2 avian influenza virus among commercial poultry aphis.usda.gov/animal_health/emergency_management/downloads/summary- farms. J Am Vet Med Assoc. (2005) 226:767–72. doi: 10.2460/javma.2005. fad-investigations-cy2020.pdf (accessed October 5, 2022). 226.767 Frontiers in Veterinary Science 09 frontiersin.org