- Email: [email protected]
Indirect estimation of porcine parvovirus maternal immunity decay in free-living wild boar (Sus scrofa) piglets by capture–recapture data
Available online at www.sciencedirect.com
The Veterinary Journal The Veterinary Journal 180 (2009) 262–264 www.elsevier.com/locate/tvjl
Short Communication
Indirect estimation of porcine parvovirus maternal immunity decay in free-living wild boar (Sus scrofa) piglets by capture–recapture data M. Fenati *, E. Armaroli, R. Corrain, V. Guberti Istituto Nazionale Fauna Selvatica, Ozzano Emilia, BO, Italy Accepted 17 December 2007
Abstract Linear mixed regression and logspline density estimation were performed to estimate the survival curve and half life of passively acquired antibodies against porcine parvovirus (PPV) in 44 wild boar piglets captured in the Northern Apennines, Italy. One piglet had no detectable maternal antibodies at 2.5 months post partum and no antibodies were detected in any of the remaining piglets by 4 months of age. Fitted survival curves indicated that maternal antibodies were undetectable from 2.5 to 6 months of age, with a median value at 3 months and a low probability of persistent maternal antibodies after 4 months of age. The estimated half life was 23 days (95% confidence interval 22–26 days). The results agree with previous data for decay of maternally acquired antibodies against PPV in the domestic pig and indicate the value of capture–recapture analysis for the estimation of infection parameters in free-living animals. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Wild boar; Porcine parvovirus; Passive immunity; Half life; Survival curve
Wild animals are involved in the epidemiology of many transmissible diseases common to both domestic animals and humans, including rabies and bovine tuberculosis. Disease management requires knowledge of infection parameters, but estimation of these parameters is complicated by the difficulty in performing longitudinal studies in wild populations. In this study, capture–recapture data were used to investigate the decay (half life and survival curve) of passively acquired antibodies against porcine parvovirus (PPV) in wild boars. The study was conducted from June 2003 to October 2004 in a hilly area of about 13 km2 located in Bologna Province, Northern Apennines, Italy. Infection with PPV is endemic and widespread in this region (55% seroprevalence; M. Fenati et al., unpublished data). In accordance with Italian legislation (L.157/92), the study was approved by the Istituto Nazionale Fauna Selvatica (Prot. 5092/ T.D2/2000). Forty-four wild boar piglets (49% males,
*
Corresponding author. Tel.: +39 0516512245; fax: +39 051796628. E-mail address: [email protected] (M. Fenati).
1090-0233/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tvjl.2007.12.009
51% females) were captured and released, including 34 more than once and some up to five times. All animals were ear-tagged, blood samples were collected and age, gender, weight and body length were recorded. The initial cohort was retested in successive recaptures at irregular intervals. The presence of antibodies against PPV in serum samples was determined by indirect enzyme-linked immunosorbent assay (ELISA) (Ingezim-PPV, Ingenasa) using a peroxidase-labelled anti-porcine immunoglobulin G (IgG). IgG is the main isotype in circulating antibodies of piglets and in sow colostrum (Paul et al., 1982). Animals passively immunised with colostral antibodies were identified by a change from positive to negative serological status in the ELISA. Data with no repeated measures were only used in the case of animals captured once and belonging to the same litter as passively immunised pigs, as identified by direct observations and radiotelemetry. The probability function for the detection of animals with passive immunity was estimated by fitting a logspline survival curve (Kooperberg and Stone, 1992), recommended for interval-censored data (Lindsey and Ryan, 1998). The half life was estimated by solving Eq. (1) under
M. Fenati et al. / The Veterinary Journal 180 (2009) 262–264
263
the assumption of exponential decay of maternal immunity, where IOD is the fixed value of the initial optical density (OD), COD is the log of the OD cut-off and B is the fixed value of the regression coefficient: OD Þ log 1 ðI OD2IC OD ð1Þ Half life ¼ B A mixed-effect linear regression model was performed to estimate the B value using Eq. (2) (Laird and Ware, 1982), where the fixed coefficients (Aij and Bij) explain the average effect across all the subjects, while the random effects (aij and bij) describe the variation between subjects, respectively, for intercept and slope and related to the ith observation from the jth subject. B ¼ logðODij Þ ¼ ðAij þ aij Þ þ ðBij þ bij Þ time þ eij
Fig. 2. Logspline estimates. (a) Survival function. (b) Density.
ð2Þ
The random measurements were assumed to be normally distributed and an auto-correlation function was included in eij. The litter level was not considered in the model because of the excessive number of litters with respect to the sample size, but their possible effect was integrated in the random effect. Model selection was based on both log-likelihood ratio test and Akaike information criterion (AIC), while goodness-of-fit was tested by log-likelihood-ratio and Wald-statistic. Statistical analysis was performed with R software (R Development Core Team, 2006). One piglet had no detectable maternal antibodies at 2.5 months post partum and no antibodies were detected in any of the remaining piglets by 4 months of age (Fig. 1). The estimated survival function, the probability of encountering animals with a undetectable level of colostral antibodies, had a median value (50%) at 3 months of age, 93% at 4 months of age and 100% at 6 months of age (Fig. 2a). In domestic pigs, passively acquired antibodies against PPV were undetectable by 3–6 months of age due
to dilution in the growing piglet and biological degradation over time (Paul et al., 1982). The logspline density was bimodal and the greatest peak was near the median value of survival function as an S-shaped curve (Fig. 2b). The second peak was unclear and perhaps overestimated due to a large interval of censoring or small sample size (Kooperberg and Stone, 1992). Regression analysis provided a good fit for the decay of passive immunity and the final model included fixed and random effects for both intercept and slope. The average PPV half life estimated from the final model was 23 days (95% confidence interval 22–26 days), which agrees with the estimate of 20 days for passively acquired antibodies against PPV in domestic pigs (Paul et al., 1982) and is also similar to an estimate of 21 days for antibodies against pseudorabies virus in wild boar (Muller et al., 2005). The mixed model highlights great variability across the subjects in the initial levels of passively acquired antibodies, which could be explained by the presence of animals belonging to several litters. There is a strong correlation between passively acquired antibody concentrations in piglets and the serological status of the dam (Vigre et al., 2003), although large differences in serum antibody titres against PPV may occur in piglets from the same litter (De Passille´ et al., 1988; Damm et al., 2002). The observed variability in the persistence of passively acquired antibodies (slope random effect) could be explained by the positive relationship between antibody persistence and the initial level of acquired immunity (Vigre et al., 2003). When appropriate statistical tools are applied, capture– recapture analysis of free-living animal data, despite being incomplete or censored, is a practical way of estimating otherwise unknown infectious disease parameters and is useful in understanding the dynamics of infection in wildlife populations. Appendix A. Supplementary data
Fig. 1. Optical density (OD) values plotted against age of piglets. The dashed line indicates the cut-off of the ELISA test.
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.tvjl.2007.12.009.
264
M. Fenati et al. / The Veterinary Journal 180 (2009) 262–264
References Damm, B.I., Friggens, N.C., Nielsen, J., Ingvartsen, K.L., Pedersen, L.J., 2002. Factors affecting the transfer of porcine parvovirus antibodies from sows to piglets. Journal of Veterinary Medicine Series A 49, 487– 495. De Passille´, A.M.B., Rushem, J., Pelletier, G., 1988. Sucking behaviour and serum immunoglobulin levels in neonatal piglets. Animal Production 47, 447–456. Kooperberg, C., Stone, C.J., 1992. Logspline density estimation for censored data. Journal of Computational and Graphical Statistics 1, 301–328. Laird, N.M., Ware, J.H., 1982. Random-effects models for longitudinal data. Biometrics 46, 963–974. Lindsey, J.C., Ryan, L.M., 1998. Tutorial in biostatistics methods for interval-censored data. Statistics in Medicine 17, 219–238.
Muller, T., Teuffert, J., Staubach, C., Selhorst, T., Depner, K.R., 2005. Long-term studies on maternal immunity for Aujeszky’s disease and classical swine fever in wild boar piglets. Journal of Veterinary Medicine Series B 52, 432–436. Paul, P.S., Mengeling, W.L., Pirtle, E.C., 1982. Duration and biological half-life of passively acquired colostral antibodies to porcine parvovirus. American Journal of Veterinary Research 43, 1376–1379. R Development Core Team, 2006. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0.. (accessed 25.04.2006). Vigre, H., Ersboll, A.K., Sorensen, V., 2003. Decay of acquired colostral antibodies to Actinobacillus pleuropneumoniae in pigs. Journal of Veterinary Medicine Series B 50, 430–435.
The Veterinary Journal The Veterinary Journal 180 (2009) 262–264 www.elsevier.com/locate/tvjl
Short Communication
Indirect estimation of porcine parvovirus maternal immunity decay in free-living wild boar (Sus scrofa) piglets by capture–recapture data M. Fenati *, E. Armaroli, R. Corrain, V. Guberti Istituto Nazionale Fauna Selvatica, Ozzano Emilia, BO, Italy Accepted 17 December 2007
Abstract Linear mixed regression and logspline density estimation were performed to estimate the survival curve and half life of passively acquired antibodies against porcine parvovirus (PPV) in 44 wild boar piglets captured in the Northern Apennines, Italy. One piglet had no detectable maternal antibodies at 2.5 months post partum and no antibodies were detected in any of the remaining piglets by 4 months of age. Fitted survival curves indicated that maternal antibodies were undetectable from 2.5 to 6 months of age, with a median value at 3 months and a low probability of persistent maternal antibodies after 4 months of age. The estimated half life was 23 days (95% confidence interval 22–26 days). The results agree with previous data for decay of maternally acquired antibodies against PPV in the domestic pig and indicate the value of capture–recapture analysis for the estimation of infection parameters in free-living animals. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Wild boar; Porcine parvovirus; Passive immunity; Half life; Survival curve
Wild animals are involved in the epidemiology of many transmissible diseases common to both domestic animals and humans, including rabies and bovine tuberculosis. Disease management requires knowledge of infection parameters, but estimation of these parameters is complicated by the difficulty in performing longitudinal studies in wild populations. In this study, capture–recapture data were used to investigate the decay (half life and survival curve) of passively acquired antibodies against porcine parvovirus (PPV) in wild boars. The study was conducted from June 2003 to October 2004 in a hilly area of about 13 km2 located in Bologna Province, Northern Apennines, Italy. Infection with PPV is endemic and widespread in this region (55% seroprevalence; M. Fenati et al., unpublished data). In accordance with Italian legislation (L.157/92), the study was approved by the Istituto Nazionale Fauna Selvatica (Prot. 5092/ T.D2/2000). Forty-four wild boar piglets (49% males,
*
Corresponding author. Tel.: +39 0516512245; fax: +39 051796628. E-mail address: [email protected] (M. Fenati).
1090-0233/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tvjl.2007.12.009
51% females) were captured and released, including 34 more than once and some up to five times. All animals were ear-tagged, blood samples were collected and age, gender, weight and body length were recorded. The initial cohort was retested in successive recaptures at irregular intervals. The presence of antibodies against PPV in serum samples was determined by indirect enzyme-linked immunosorbent assay (ELISA) (Ingezim-PPV, Ingenasa) using a peroxidase-labelled anti-porcine immunoglobulin G (IgG). IgG is the main isotype in circulating antibodies of piglets and in sow colostrum (Paul et al., 1982). Animals passively immunised with colostral antibodies were identified by a change from positive to negative serological status in the ELISA. Data with no repeated measures were only used in the case of animals captured once and belonging to the same litter as passively immunised pigs, as identified by direct observations and radiotelemetry. The probability function for the detection of animals with passive immunity was estimated by fitting a logspline survival curve (Kooperberg and Stone, 1992), recommended for interval-censored data (Lindsey and Ryan, 1998). The half life was estimated by solving Eq. (1) under
M. Fenati et al. / The Veterinary Journal 180 (2009) 262–264
263
the assumption of exponential decay of maternal immunity, where IOD is the fixed value of the initial optical density (OD), COD is the log of the OD cut-off and B is the fixed value of the regression coefficient: OD Þ log 1 ðI OD2IC OD ð1Þ Half life ¼ B A mixed-effect linear regression model was performed to estimate the B value using Eq. (2) (Laird and Ware, 1982), where the fixed coefficients (Aij and Bij) explain the average effect across all the subjects, while the random effects (aij and bij) describe the variation between subjects, respectively, for intercept and slope and related to the ith observation from the jth subject. B ¼ logðODij Þ ¼ ðAij þ aij Þ þ ðBij þ bij Þ time þ eij
Fig. 2. Logspline estimates. (a) Survival function. (b) Density.
ð2Þ
The random measurements were assumed to be normally distributed and an auto-correlation function was included in eij. The litter level was not considered in the model because of the excessive number of litters with respect to the sample size, but their possible effect was integrated in the random effect. Model selection was based on both log-likelihood ratio test and Akaike information criterion (AIC), while goodness-of-fit was tested by log-likelihood-ratio and Wald-statistic. Statistical analysis was performed with R software (R Development Core Team, 2006). One piglet had no detectable maternal antibodies at 2.5 months post partum and no antibodies were detected in any of the remaining piglets by 4 months of age (Fig. 1). The estimated survival function, the probability of encountering animals with a undetectable level of colostral antibodies, had a median value (50%) at 3 months of age, 93% at 4 months of age and 100% at 6 months of age (Fig. 2a). In domestic pigs, passively acquired antibodies against PPV were undetectable by 3–6 months of age due
to dilution in the growing piglet and biological degradation over time (Paul et al., 1982). The logspline density was bimodal and the greatest peak was near the median value of survival function as an S-shaped curve (Fig. 2b). The second peak was unclear and perhaps overestimated due to a large interval of censoring or small sample size (Kooperberg and Stone, 1992). Regression analysis provided a good fit for the decay of passive immunity and the final model included fixed and random effects for both intercept and slope. The average PPV half life estimated from the final model was 23 days (95% confidence interval 22–26 days), which agrees with the estimate of 20 days for passively acquired antibodies against PPV in domestic pigs (Paul et al., 1982) and is also similar to an estimate of 21 days for antibodies against pseudorabies virus in wild boar (Muller et al., 2005). The mixed model highlights great variability across the subjects in the initial levels of passively acquired antibodies, which could be explained by the presence of animals belonging to several litters. There is a strong correlation between passively acquired antibody concentrations in piglets and the serological status of the dam (Vigre et al., 2003), although large differences in serum antibody titres against PPV may occur in piglets from the same litter (De Passille´ et al., 1988; Damm et al., 2002). The observed variability in the persistence of passively acquired antibodies (slope random effect) could be explained by the positive relationship between antibody persistence and the initial level of acquired immunity (Vigre et al., 2003). When appropriate statistical tools are applied, capture– recapture analysis of free-living animal data, despite being incomplete or censored, is a practical way of estimating otherwise unknown infectious disease parameters and is useful in understanding the dynamics of infection in wildlife populations. Appendix A. Supplementary data
Fig. 1. Optical density (OD) values plotted against age of piglets. The dashed line indicates the cut-off of the ELISA test.
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.tvjl.2007.12.009.
264
M. Fenati et al. / The Veterinary Journal 180 (2009) 262–264
References Damm, B.I., Friggens, N.C., Nielsen, J., Ingvartsen, K.L., Pedersen, L.J., 2002. Factors affecting the transfer of porcine parvovirus antibodies from sows to piglets. Journal of Veterinary Medicine Series A 49, 487– 495. De Passille´, A.M.B., Rushem, J., Pelletier, G., 1988. Sucking behaviour and serum immunoglobulin levels in neonatal piglets. Animal Production 47, 447–456. Kooperberg, C., Stone, C.J., 1992. Logspline density estimation for censored data. Journal of Computational and Graphical Statistics 1, 301–328. Laird, N.M., Ware, J.H., 1982. Random-effects models for longitudinal data. Biometrics 46, 963–974. Lindsey, J.C., Ryan, L.M., 1998. Tutorial in biostatistics methods for interval-censored data. Statistics in Medicine 17, 219–238.
Muller, T., Teuffert, J., Staubach, C., Selhorst, T., Depner, K.R., 2005. Long-term studies on maternal immunity for Aujeszky’s disease and classical swine fever in wild boar piglets. Journal of Veterinary Medicine Series B 52, 432–436. Paul, P.S., Mengeling, W.L., Pirtle, E.C., 1982. Duration and biological half-life of passively acquired colostral antibodies to porcine parvovirus. American Journal of Veterinary Research 43, 1376–1379. R Development Core Team, 2006. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0.