Vaccination of Porcine Circovirus type 2 (PCV2)-infected Sows against Porcine Parvovirus (PPV) and Erysipelas: Effect on Post-weaning Multisystemic Wasting Syndrome (PMWS) and on PCV2 Genome Load in the Offspring

ARTICLE IN PRESS J. Comp. Path. 2007,Vol. 136,133^144

www.elsevier.com/locate/jcpa

Vaccination of Porcine Circovirus type 2 (PCV2)-infected Sows against Porcine Parvovirus (PPV) and Erysipelas: Effect on Post-weaning Multisystemic Wasting Syndrome (PMWS) and on PCV2 Genome Load in the Offspring N. Rose*, P. Blanchard*, R. Cariolet*, B. Grasland*, N. Amennay, A. Oger*, B. Durandz, M. Balaschy, A. Jestin* and F. Madec* *

AFSSA-LERAP, yLDA 22, ZoopoŒle, Ploufragan F22440, zAFSSA-LERPAZ, 23 Avenue du GeŁneŁral de Gaulle, Maisons Alfort cedex, F94706, France and yFort DodgeVeterinaria, Carretera Camprodon s/n, La Riba, 17813 Vali de Bianya (Girona), Spain

Summary The e¡ect of di¡erent Parvovirus+Erysipelas vaccination schemes in PCV2-infected sows on PMWS outcome in the o¡spring was investigated under experimental conditions. Six PCV2-free sows were ¢rst infected oro-nasally with PCV2 two months before insemination (D0; ‘‘Day 0’’) and then by the intra-uterine route at insemination (D62). On D21and D42, vaccinated sows received either the two commercial monovalent vaccines, A1(PPV) and A2(Erysipelas), or the bivalent vaccine B (PPV+Erysipelas). In addition, three SPF sows (foster-sows) were synchronized for farrowing dates to enable them to foster piglets born to infected sows and removed at birth before colostrum intake. A signi¢cantly higher proportion of mummi¢ed fetuses was obtained from PCV2-infected non-vaccinated sows than from vaccinated sows. Acute myocarditis lesions were found in their piglets, together with a high PCV2 genome load. The latter was signi¢cantly higher than in those born to PCV2-infected vaccinated sows. Sentinel PCV2-negative piglets, born to SPF foster-sows, seroconverted at almost the same time as piglets without PCV2 passive immunity and born to infected sows. Sixteen of the 84 liveborn piglets born to infected sows and fostersows were a¡ected by a syndrome possibly related to PMWS, as judged by clinical signs and histological lesions. Most were born to PCV2-infected non-vaccinated sows and 12/16 did not receive PCV2 passive immunity. The probability of PCV2 infection and the number of PCV2 genome copies per gram of tissue were signi¢cantly increased in piglets that did not receive PCV2 passive immunity. r 2007 Elsevier Ltd. All rights reserved. Keywords: erysipelas; pig; PMWS; porcine circovirus; porcine parvovirus; viral infection

Introduction Porcine circovirus type 2 (PCV2) is recognized as the aetiological agent of post-weaning multisystemic wasting syndrome (PMWS) in pigs (Allan and Ellis, 2000; Segales and Domingo, 2002). Several experimental Correspondence to: N. Rose, AFSSA-LERAP, ZoopoŒle, Ploufragan, B.P. 53, F22440, France. (e-mail: [email protected]) 0021-9975/$ - see front matter

doi:10.1016/j.jcpa.2007.01.006

models based on inoculation with PCV2 (Balasch et al.,1999; Magar et al., 2000; Allan et al., 2003; Okuda et al., 2003) produced a mild disease; Albina et al. (2001), however, produced a severe disease, similar to that seen in the ¢eld, by inoculating conventional growing pigs. It is recognized that additional infectious factors such as porcine parvovirus (PPV) (Allan et al.,1999; Kennedy et al., 2000; Krakowka et al., 2000) or porcine reproductive and respiratory syndrome (PRRS) virus r 2007 Elsevier Ltd. All rights reserved.

ARTICLE IN PRESS 134

N. Rose et al.

(Harms et al., 2001; Rovira et al., 2002), as well as enhancement of the immune system (Krakowka et al., 2001; Grasland et al., 2005), are supplementary triggering factors. On-farm observations and experimental models have demonstrated the possible role of PCV2 in reproductive disorders such as abortion, stillbirth and mummi¢cation (West et al., 1999; Meehan et al., 2001; Cariolet et al., 2002; Kim et al., 2004). These reproductive failures were mainly associated with multiplication of the virus in the cardiomyocytes after its introduction into the uterus of pregnant sows (Sanchez et al., 2001; Pensaert et al., 2004). The target cells, which consist mainly of cardiomyocytes in fetal life, change after birth to macrophages exclusively (Sanchez et al., 2003). Thus, infection during early pregnancy may be responsible for fetal mortality and for the birth of pre-infected piglets that may develop the clinical disease thereafter. Some transplacental infection may occur (Farnham et al., 2003; Pensaert et al., 2004; Park et al., 2005), but the typical lesions of myocarditis are mainly associated with intra-uterine infection, suggesting infection from the time of insemination with contaminated semen onwards (Larochelle et al., 2000; LeTallec et al., 2001; Kim et al., 2003; McIntosh et al., 2005). Epidemiological studies have shown that clinical outcome may be in£uenced by non-infectious factors such as maternal vaccination (Rose et al., 2003; LopezSoria et al., 2005) and maternal immunity (Allan et al., 2002; Calsamiglia et al., 2004; Rose et al., 2005). Rose et al. (2003) identi¢ed the use of separate, rather than combined, parvovirus and erysipelas vaccines in breeding sows as a risk factor. In the present study, the aim was to investigate the e¡ect of di¡erent parvovirus vaccination schemes in PCV2-infected sows on PMWS in the o¡spring.

Materials and Methods Experimental Animals

Eleven Large White speci¢c pathogen-free (SPF) sows were used, derived from the AFSSA (Agence Franc- aise de SeŁcuriteŁ Sanitaire des Aliments) SPF herd (Cariolet et al., 2004). Experimental Design (Fig. 1)

Eight sows (1^8) were housed in pairs in four isolated rooms in a level 3 air-¢ltered biosecurity facility. Each pair consisted of a primiparous and a multiparous sow. Room 1 contained sows 1 and 2 (controls), which were neither infected with PCV2 nor vaccinated. Room 2 contained sows 3 and 4, which were infected with PCV2 and vaccinated with vaccine B (combined PPV and Erysipelas). Room 3 contained sows 5 and 6, which

were PCV2-infected and vaccinated with vaccines A1 (PPV) and A2 (Erysipelas). Room 4 contained sows 7 and 8, which were PCV2-infected but not vaccinated. Vaccines A and B, which came from di¡erent manufacturers, were both inactivated but contained di¡erent strains of micro-organisms and di¡erent adjuvants. The vaccinations were performed according to the manufacturers’ recommendations, which included aseptic measures and the use of a fresh needle for each animal. In addition, sows 9, 10 and 11 were housed individually in three further isolated rooms to permit transfer of piglets from room 2, 3 and 4, respectively, these three sows having been synchronized appropriately for farrowing. At 113 days’ gestation, the sows were given an injection of 2 ml of Planates (cloprostenol; ScheringPlough, Levallois-Perret, France). At farrowing, 1 day later, all piglets were immediately removed from the dam before gaining access to colostrum. Cross-fostering of piglets from rooms 2, 3 and 4 to sows 9,10 and 11, respectively, was organized according to the numbers of live piglets per litter, the maximum number of randomly selected piglets transferred being six per treatment. The experiment was performed in accordance with EU and French regulations on animal experimentation. Infection and Vaccination

The infective inocula each consisted of 2 ml of a PCV2 suspension (104 TCID50/ml), prepared as described by Rovira et al. (2002). Sows were infected by oro-nasal inoculation 2 months before insemination (‘‘day 0’’; D0) and by intra-uterine inoculation at insemination on D62, by which time they were expected to be seropositive. Control sows received sterile phosphate-bu¡ered saline (PBS). Vaccines were administered intramuscularly as under ¢eld conditions, a ¢rst injection (given on D21, i.e., 3 weeks after housing within the facility) being followed 3 weeks later (D42) by a booster injection. Observations and Sampling

Sows were examined daily during pregnancy (rectal temperature, feed consumption). Blood samples were taken every 2 weeks until the time at which the sows were killed (i.e., after weaning of the piglets at the age of 28 days). The piglets were examined daily (clinical observations, including rectal temperature) and weighed weekly. Blood samples were taken every 2 weeks, the ¢rst being collected immediately after birth, i.e., from piglets that had (with the exception of one piglet from room 4) received no colostrum. Sequential killing was

ARTICLE IN PRESS 135

Parvovirus-Erysipelas Vaccination and PMWS Room 4

Room 7

SOW 7

SOW 11

SOW 8

PCV2 infection but no vaccination

Fostering for room 4

Room 3

Room 6

SOW 5

SOW 10

SOW 6

PCV2 infection + Vaccines A1 and A2

Fostering for room 3 Room 5

Room 2

SOW 9 SOW 4

SOW 3

PCV2 infection + Vaccine B

Fostering for room 2

Room 1

SOW 1

SOW 2

No infection or vaccination (controls)

D0: oronasal PCV2 infection

D21: First PPV/Ery vaccine injection

D42: Second PPV/Ery vaccine injection

D62: Insemination and Intra-uterine PCV2 infection

Fig. 1. Experimental design and chronology of events. PPV, porcine parvovirus; Ery, erysipelas; D, day.

carried out to maintain the required space per pig throughout the study. At 10 weeks of age only eight o¡spring were kept in each room. At 18 weeks of age two pigs from each room were killed, the remainder being kept until they were 27 weeks old. At each point of slaughter, two control o¡spring (born to the non-infected non-vaccinated sows in room 1) were killed for purposes of comparison. Euthanasia was carried out by anaesthesia (Nesdonals; Merial, Lyon, France) followed by exsanguinations. Necropsy and Samples

At necropsy, immediately after killing, all organs were examined macroscopically and the following samples

were taken for histopathological and virological examination: lung, tonsil, heart, kidney, thymus, ileum, and lymph nodes (inguinal, mesenteric, tracheo-bronchial). Samples of the genital tract were taken from sows. Histopathology and In-situ Hybridization

Histological examination was carried out on the di¡erent tissues after ¢xation with 10% formaldehyde. Sections (8 mm) were stained with haemalum-eosinsafranin and observed under a light microscope. In-situ hybridization was performed on sections (4 mm) of tissues which were washed, heated to 40 1C, deproteinized and then hybridized at 55 1C with the PCV2-speci¢c

ARTICLE IN PRESS 136

N. Rose et al.

probe 50 CCGTTGTCCCTGAGATCTA 30 conjugated at both ends with digoxigenin. After three washings and a single incubation at 40 1C, a biotinylated mouse monoclonal antibody to digoxigenin (Ventana Medical, Strasbourg, France), a peroxidase-avidin complex and 3-amino-9-ethyl-carbazole were applied in succession. The slides were ¢nally counterstained with haemalum and blueing reagent (Ventana). Serology

Serum samples were tested for PCV2 antibodies and titrated by an ELISA based on the recognition of a recombinant PCV2 capsid protein/GST fused protein and a GST (Glutathione S-transferase) protein (Blanchard et al., 2003). Samples with an OD (optical density) ratio higher than 1.5 were considered positive for PCV2 antibodies (sensitivity: 0.98 and speci¢city: 0.95, taking IPMA [Immunoperoxidase Monolayer Assay] as a reference). Quantification of PCV2 Genomes by Real-time Polymerase Chain Reaction (PCR)

Homogenates (31%) of tissues samples (tracheobronchial, inguinal and mesenteric lymph nodes, tonsil, lung, heart, kidney, ileum and thymus) were prepared in ice-cooled PBS by means of a mechanical homogenizer (Bioblock Scienti¢c, Illkirch, France). The suspensions were centrifuged at 2000 g for 15 min and the supernates stored at ^ 80 1C. DNA was extracted from tissue homogenates (80 ml), with the QIAamps DNA mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. Elution was performed with 400 ml of bu¡er AE (Qiagen); 5 ml of the extract, corresponding to 1 ml of tissue homogenate (0.31mg of tissue), were used as a template for PCV2 TaqMan PCR. During DNA extraction, a control, in which tissue homogenate was replaced by PBS, was included for every ¢ve samples, to check PCV2 contamination.The number of PCV2 genome copies was assessed by a real-time PCR based onTaqMan technology (Blanchard et al., 2004). Brie£y, the designed PCV2 speci¢c primers 50 -GGGAGCAGGGCCAGAATT-30 (410^427) and 50 -CGCTCTGTGCCCTTTGAATACT-30 (473^452) target the PCV2 ORF2 region (GenBank accession no. AF201311) and allow the ampli¢cation of a 64 bp fragment. The TaqMan probe 50 -ACCTTAACCTTTCTTATTCTG-30 (430^450) was labelled with the £uorescent reporter dye FAM (6-carboxy£uorescein) at the 50 end and with the non£uorescent quencher (NFQ) associated with the minor groove binder at the 30 end. A DNA solution of a plasmid carrying a single copy of the PCV2 genome was serially diluted and used to generate a standard curve of quanti¢cation. The reactions were performed on an

ABI Prism 7000 thermocycler (Applied Biosystems, Foster City, CA, USA). Statistical Analysis of Real-time PCR Results

Crude real-time PCR results were ¢rst log-transformed. The data had a speci¢c structure (repeated measures within pigs because several organs were tested) and a non-normal distribution with a high number of zeros and a continuous part which was left censored. The mean of the log-transformed number of genome copies was therefore a poor indicator of the average number of PCV2 genome copies per pig and could not be used to compare the PCV2 genome load between treatments. We therefore adapted the method described by Berk and Lachenbruch (2002), which associates a Bernouilli model for the zeros and a lognormal model for the positive values. Both models are analysed together by maximizing the joint likelihood of the binomial and loglinear model. Clustering of organs within pigs is taken into account by estimating a random e¡ect, assumed to be normal, with mean ¼ 0, constant variance t2, and with independence between subjects (see Appendix A for details of the model). Due to the sequential slaughtering, the di¡erent models were adjusted according to the age at slaughter. The model was programmed with the PROC NLMIXED of SAS 9.1 (SAS Institute Inc., 2000). Quantification of PCV2 Transmission Within Groups of Pigs, With and Without PCV2 Passive Immunity: Estimation of the Basic Reproduction Ratio (R0)

The basic reproduction ratio (R0) is de¢ned as the mean number of new infections that one typical infectious individual causes in a totally susceptible population. If R0 is smaller than one, only minor outbreaks will occur; if R0 is larger than one, a major outbreak is also possible. This indicator was used to compare PCV2 transmission within groups of pigs, with or without PCV2 passive immunity. The ¢nal size algorithm based on the basic SIR (susceptible-infectiousremoved) model was used as described previously (De Jong and Kimman, 1994; Kroese and De Jong, 2001). This algorithm enables a common R0 value and its con¢dence interval to be estimated from the results of several repeats of the same experiment. Pigs born to infected sows were considered as infectious animals and sentinel pigs born to SPF sows as susceptible pigs. The PCV2 status of susceptible animals after slaughter (end of the experiment) was used to determine the number of susceptible animals that became infected during the experiment. The algorithm was implemented in R statistical software (Ihaka and Gentleman, 1996).

ARTICLE IN PRESS 137

Parvovirus-Erysipelas Vaccination and PMWS

Farrowing Results

Results Serology in Sows

The control sows remained PCV2-seronegative throughout the study. Seroconversion occurred 35 days after oro-nasal inoculation in the infected sows, except for one PCV2-infected B-vaccinated sow from room 2, which remained seronegative (albeit doubtful at D132). All the other sows were seropositive at insemination and remained so until killed, except for two which were transiently seronegative (Fig. 2). Parvovirus serology was consistent with vaccination in the vaccinated sows (haemagglutination inhibition titres ranging from 320 to 640), whereas the non-vaccinated sows remained negative throughout the study.

2.9

PCV2 antibodies - OD ratio

Intra-uterine infection + A.I.

sow 2-control sow 1-control sow 4-PCV2+B sow 3-PCV2+B sow 6-PCV2+A sow 5-PCV2+A sow 8-PCV2 sow 7-PCV2

2.7 2.5 2.3 2.1 1.9

PCV2-infected but non-vaccinated sows (room 4) produced a high number of mummi¢ed piglets, ranging from 5 to 33 cm in length. Only one live piglet was born to sow 8 and only four were born to sow 7. The numbers of mummi¢ed piglets from vaccinated and non-vaccinated sows di¡ered signi¢cantly (Po0.001; Fisher’s exact test). There was no signi¢cant di¡erence between the two vaccines (Table 1). Piglets mummi¢ed at the end of gestation had typical lesions of myocarditis with severe mononuclear in¢ltration. All mummi¢ed piglets gave negative results in PCR test for PPV, PRSS and encephalo-myocarditis virus.

1.7 1.5 1.3 1.1 0.9

Oro-nasal infection

0.7 0.5

-7

14

27

40

54 91 105 Days post oro-nasal infection

118

132

146

166

Fig. 2. PCV2 serology of the sows before and during gestation (positive threshold:1.5). A.I., arti¢cial insemination.

Table 1 Farrowing results and numbers of liveborn, stillborn and mummi¢ed piglets in the various groups Number of piglets born to sows no. Type of piglet

1

2

3y

4y

5z

6z

7z

8z

9y

10y

11y

Liveborn Stillborn Mummi¢ed

12 0 2

10 1 0

7 1 4

11 1 0

10 0 3

12 0 4

4 2 10

1 0 14

11 3 0

11 0 0

17 0 2

 Controls (room 1) y

PCV2+vaccine B (room 2) PCV2+vaccines A1+A2 (room 3) z PCV2 only (room 4) y Foster sows. z

ARTICLE IN PRESS 138

N. Rose et al.

Bearing in mind the numbers of liveborn o¡spring per litter, cross-fostering was arranged as follows (see also Fig. 1). Before the intake of any colostrum, piglets were moved from each of three rooms as follows: six piglets from room 2 (PCV2+vaccine B) to sow 9 for fostering; six piglets from room 3 (PCV2+vaccines A1 and A2) to sow 10 for fostering; and four piglets from room 4 (PCV2 but no vaccination) to sow 11 for fostering. In addition, eight piglets born to sow 11 (SPF sow) were moved to sow 7 (PCV2 but no vaccination) for fostering and three and four piglets born to sows 9 and 10 respectively (SPF sows) were moved together to sow 8 (PCV2 but no vaccination) for fostering. This procedure resulted in nine groups of piglets as follows. (1) Piglets born to PCV2-infected/B-vaccinated sows and receiving colostrum from the natural mother (n ¼ 12). (2) Piglets born to PCV2-infected/B-vaccinated sows and receiving no colostrum from the natural mother but fostered by an SPF sow (n ¼ 6). (3) Piglets born to PCV2-infected/A1+A2-vaccinated sows and receiving colostrum from the natural mother (n ¼ 16). (4) Piglets born to PCV2-infected/A1+A2-vaccinated sows and receiving no colostrum from the natural mother but fostered by an SPF sow (n ¼ 6). (5) One piglet born to a PCV2-infected/non-vaccinated sow and receiving colostrum from the natural mother. (6) Piglets born to PCV2-infected/non-vaccinated sows and receiving no colostrum from the natural mother but fostered by an SPF sow (n ¼ 4). (7) ‘‘Sentinel’’piglets born to SPF sows and remaining with their mother (n ¼ 24). (8) Sentinel piglets born to SPF sows and moved to a PCV2-infected/non-vaccinated sow (i.e., joining group 5) (n ¼ 15). (9) Control piglets from non-infected/non-vaccinated sows (n ¼ 22).

PCV2 Serology in the Offspring (Fig. 3)

Piglets born to control sows (group 9) remained below the positive threshold throughout the study. All piglets tested were seronegative before intake of colostrum, even those born to PCV2-infected sows. Two categories of piglets gave clear evidence of seroconversion between 34 and 48 days of age, all remaining positive until 140 days of age. These two categories consisted of (1) piglets born to infected sows but deprived of maternal colostrum (i.e., groups 2, 4 and 6), and (2) piglets born to SPF foster sows but raised in contact with infected foster-piglets (i.e., group 7). Seroconversion seemed to occur almost simultaneously in the two categories. The remaining piglets (either born to infected sows and receiving colostrum from PCV2-positive sows [groups 1, 3 and 5] or born to PCV2-negative sows but moved to infected sows [group 8]) showed a sharp increase in PCV2 passively transferred antibody in the ¢rst week of life, followed by a progressive decrease up to 90 days of age; thereafter, they became seronegative and remained so until killed. PMWS Morbidity, Mortality and Lesions (Table 2)

No lesion was observed in any of the control piglets at necropsy. In the other piglets, however, di¡erent clinical groups could be de¢ned as follows: piglets dying early and suddenly with acute myocarditis; one pig with typical lymphoid lesions of PMWS; several pigs with typical clinical signs, typical lymphoid lesions and in£ammatory lesions of the lungs and kidneys; and piglets showing no clinical signs but with histopathological ¢ndings suggestive of PMWS. The piglets exhibiting typical lesions of PMWS were mainly those

14 Controls Born to PCV2 non-infected sow / PCV2PIBorn to PCV2 non-infected sow / PCV2PI+ Born to PCV2-infected sow / PCV2PI+ Born to PCV2-infected sow / PCV2PI-

PCV2 antibodies - OD ratio

12 10 8 6 4 2 0 0

7

20

27

34

48

64

77 90 Age in days

105

119

140

161

175

190

Fig. 3. PCV2 serology of the o¡spring, according to the infection status of the sow at parturition and to presence or absence of passive immunity (PCV2PI+ or PCV2PI-) (positive threshold:1.5).

ARTICLE IN PRESS 139

Parvovirus-Erysipelas Vaccination and PMWS Table 2 Age at death and number of piglets with clinical and/or histopathological ¢ndings in the various rooms. Number of piglets in room no. Clinical and histopathological categories (A^D) A

B C

D

Age (days) at death

1»

2f

3$

4h

5y

6z

7&

7 14 18 37 64y 70y 78y 104y 140y 139z 140z 146z 176z

y y y y y y y y y y y y y

y y y y y y y y y y y y y

y 1i y y y 1i y y y y y y y

1i y y y y y y y y y y y 1s

y y y y y y 1s 1s y 1i y y y

y y y 1i y y y y y y y 1s+1i y

y y 1i y 1i y 1s y 1s y 2s y y

Clinical categories: A, Laboured breathing, sudden death. Severe myocarditis; B, Laboured breathing, sudden death. Severe lymphoid depletion with histiocytosis, viral inclusions, giant cells, In Situ Hybridisation positive; C, pallor, rough hair coat, low weight gain after weaning with respiratory and/ or digestive signs (ileitis). Moderate to severe lymphoid depletion, interstitial nephritis and interstitial pneumonia; D, No apparent clinical ¢ndings. Moderate to severe lymphoid depletion, interstitial nephritis and interstitial pneumonia. » Controls. f PCV2 infection+vaccine B. $ PCV2 infection+vaccine A1/A2. h PCV2 only. y Fostering for room 2. z Fostering for room 3. & Fostering for room 4. i Born to a PCV2-infected sow. s Born to a non-infected sow.  Found dead. y Humanely killed for ethical reasons. z Humanely killed at the end of the study (no clinical signs).

deprived of PCV2 passive immunity (12/16).These were either born to, or in contact with, a PCV2-infected/ non-vaccinated sow. Piglets born to PCV2-infected/ vaccinated sows or in contact with them were less likely to be a¡ected, but there were no signi¢cant di¡erences between the two vaccines. Half of the diseased pigs were sentinels, born to SPF foster-sows and deprived of PCV2 passive immunity. Parvovirus/Erysipelas Vaccine Evaluation in Respect of the Presence of PCV1 and PCV2

Both commercial vaccine products (vaccine A1/A2 and vaccine B) were examined in respect of PCV1 (by PCR) and PCV2 (by real-time PCR). Both vaccines gave negative results for PCV1 and no genome copy was demonstrated for PCV2. Additional SPF piglets were given the vaccines used in the experiment and monitored serologically for PCV2 antibodies; no seroconversion could be observed 3 weeks after the second (booster) injection (data not shown).

PCV2 Genome Load Assessed with Real-time PCR (Table 3)

All control piglets gave negative results for PCV2. In other piglets, the PCV2 genome load was similar in males and females. Mummi¢ed fetuses were signi¢cantly more heavily infected than were liveborn or stillborn piglets. The PCV2 genome load was particularly high in piglets exhibiting histological lesions typical of PMWS. Piglets that seroconverted against PCV2 were more likely to be PCV2-infected and to have a signi¢cantly higher genome load than were those that remained seronegative. However, three piglets remained seronegative until the end of the experiment and at necropsy their organs gave positive results and real-time PCR showed a signi¢cant genome load. When all piglets (liveborn, stillborn, and mummi¢ed) were considered, the probability of being infected and also the genome load were found to be greater in piglets born to infected but non-vaccinated sows than for those born to infected vaccinated sows (regardless

ARTICLE IN PRESS 140

N. Rose et al.

Table 3 PCV2 genome load assessed with real-time PCR: comparison of the estimated probability of positive results and of the log-number of PCV2 genome equivalent copies per gram of tissue according to piglet characteristics and with the associated binomial and loglinear model Variable and categories

Gender Male Female Status at birth (controls excluded) Liveborn Stillborn Mummi¢ed Histological lesions (liveborn piglets only) Lesions consistent with PMWS No lesion PCV2 seroconversion Yes No Vaccine e¡ect (liveborn+stillborn+mummi¢ed) PCV2 but not vaccinated PCV2+Vaccine A PCV2+Vaccine B Vaccine e¡ect and passive immunity (liveborn) PCV2+Vaccine A with passive immunity PCV2+Vaccine A without passive immunity PCV2+Vaccine B with passive immunity PCV2+Vaccine B without passive immunity Sentinel pigs with passive immunity Sentinel pigs without passive immunity

Number of piglets

Probability of positive results (95% CL)

Log-number of PCV2 genome copies per g of tissue (95% CL)

59 60

0.25 (0.01^0.49) 0.09 (0^0.21)

7.3 (6.8^7.7) 6.8 (6.3^7.3)

82 5 35

0.18 (0.05^0.32)a 0.02 (0^0.09)a 0.97 (0.93^1)b

7.1 (6.9^7.4)a 7.4 (6.0^8.7)a 9.2 (8.8^9.5)b

13 71

0.74 (0.46^1) a 0.11 (0.03^0.18) b

7.9 (7.4^8.4)a 6.9 (6.5^7.2)b

28 69

0.55 (0.27^0.82) a 0.04 (0^0.09) b

7.6 (7.3^8.0)a 6.6 (6.1^7.0)b

30 27 24

0.99 (0.98^1)a 0.28 (0^0.76)b 0.14 (0^0.45)b

9.0 (8.5^9.5)a 7.3 (6.7^7.9)b 6.9 (6.3^7.6)b

12 6 10 6 14 17

0.16 (0.03^0.3)ab 0.74 (0.55^0.93)c 0.04 (0^0.09)a 0.40 (0.14^0.66)bc 0.12 (0.03^0.21)ab 0.67 (0.52^0.82)c

7.0 (6.5^7.4)ab 8.2 (7.8^8.6)d 6.0 (5.3^6.6)a 7.2 (6.7^7.7)bc 6.6 (6.1^7.0)ab 7.8 (7.5^8.1)cd

Estimates for a single variable and within a column with di¡erent letters are signi¢cantly di¡erent (Po0.05). CL, con¢dence limits.  The model is adjusted for slaughter age.

of the vaccine used). For investigation of the role of passive immunity, the piglets born to PCV2-infected but non-vaccinated sows were removed from the calculation because they were too few. Hence there were six categories of piglets, as follows: born to infected and A-vaccinated sows, with or without PCV2 passive immunity; born to infected and B-vaccinated sows, with or without PCV2 passive immunity; and contact pigs, with or without PCV2 passive immunity (Table 3). Maternal immunity appeared to exert a strong e¡ect, as the probability of being infected, as well as the genome load, were always signi¢cantly lower for piglets that received PCV2 passive immunity than for those that did not. Sentinel piglets (SPF pigs in contact with piglets born to PCV2-infected sows) resembled their corresponding group (i.e., corresponding as regards passive immunity status) of piglets born to infected sows, in respect of (1) probability of being infected, and (2) genome load. Within the group of piglets deprived of PCV2 passive immunity, those born to infected and A-vaccinated sows had a signi¢cantly heavier genome load than did those born to PCV2-infected and B-vaccinated sows.

Quantification of PCV2 Transmission within Groups of Pigs with or without PCV2 Passive Immunity: Estimation of the Basic Reproduction Ratio (R0) (Table 4)

The basic reproduction ratio was lower in the group of piglets that received PCV2 passive immunity than in the group deprived of maternal antibodies (1.5 versus 3.1, respectively). This di¡erence was insigni¢cant (P40.05) because of the lack of power in the estimation of the R0 within the group of piglets receiving PCV2 passive immunity, the number of repetitions being lower (two) than for the group deprived of maternal antibodies (three). However, the R0 was signi¢cantly higher than 1 in the group deprived of PCV2 maternal antibodies, indicating a more e⁄cient transmission of the virus in this group (Table 4).

Discussion The study was based on 11 SPF sows and their o¡spring (153 piglets), reared in a completely controlled environment. Statistical analyses of results obtained in the piglets (farrowing results, PCV2 genome load) took into

ARTICLE IN PRESS 141

Parvovirus-Erysipelas Vaccination and PMWS

Table 4 Comparison of the estimated basic reproduction ratio (R0) for PCV2 transmission, with and without PCV2 passive immunity Groups

Repetition (sow number)

a

S0

b

I0

c

x

Estimated R0

Con¢dence interval (a ¼ 0.05)

P value for R041

3.1

1.3^7.2

0.004

1.5

0.8^46.7

0.06

Without PCV2 passive immunity

1 (9) 2 (10) 3 (11)

6 6 6

6 6 4

4 6 6

With PCV2 passive immunity

1 (7) 2 (8)

8 7

1 2

5 4

a

S0: initial number of susceptible animals. I0: initial number of infectious animals. c x: number of susceptible infected animals during the experiment. b

account dam-associated features (PCV2 status, vaccine status and PCV2 passive immunity). The sows came from a protected SPF piggery known to be free from PCV2. The clinical status of the o¡spring was assessed by the classical criteria for PMWS (Sorden, 2000) and the viral status of piglets was assessed by real-time PCR to quantify the genome load per gram of tissue. This method is a useful tool for correlating PCV2 infection and the PMWS status of animals (Blanchard et al., 2004; Grasland et al., 2005). However, the interpretation of real-time PCR results is sometimes misleading when several samples are taken from each animal. The average number of PCV2 genome copies for the animal is a poor indicator of the genome load carried by the pig; this is because within an infected animal several organs may give negative test results, while signi¢cant titres may be demonstrated in other organs. The distribution of the log number of PCV2 genome copies is therefore bimodal, with a high number of zeros and a continuous part which is left censored. In this study, an adaptation of the method described by Berk and Lachenbruch (2002) was used, enabling the joint likelihood of the binomial and the loglinear model to be maximized. The probability of being positive and the estimate of the log number of genome copies were thus simultaneously compared between groups of piglets, to prevent misinterpretation of the results due to di¡erent distributions of positive results between groups. The experimental model in this study was based on SPF sows, initially free from PCV2, but infected oronasally; thus, they were PCV2-seropositive at the time of insemination, as is commonly observed in the ¢eld (Rose et al., 2003; Lopez-Soria et al., 2005). The colostrum from all the infected sows gave positive test results for PCV2 antibodies, unlike the colostrum from the controls and the foster-sows (data not shown). Severe reproductive disorders were observed in the two infected sows that were not vaccinated. These results

were based on only two sows, but the outcome was similar in both. Moreover, a previous study with PCV2-seronegative SPF sows infected by the intra-uterine route at the time of insemination gave evidence of similar severe reproductive disorders (abortion, mummi¢cation, stillbirth) (Cariolet et al., 2002); moreover the mummi¢ed o¡spring were heavily infected with PCV2. In seropositive sows in the present study, similar symptoms and fetal lesions (acute myocarditis) were observed in the PCV2-infected non-vaccinated group.These observations are also consistent with those of Sanchez et al. (2001), who inoculated fetuses in utero and demonstrated myocarditis when inoculations were given at 57 days of gestation. The present results, in addition to those of Cariolet et al. (2002), suggest that PCV2 can replicate within fetuses, mainly in the heart, even if it is introduced into the uterus as early as the time of insemination. This result should be born in mind in the context of mummi¢cations and post-farrowing mortality in ¢eld outbreaks of PMWS. A strong protective e¡ect of Parvovirus+Erysipelas vaccination against reproductive failure was observed, the numbers of mummi¢ed and stillborn piglets being similar in the two groups of vaccinated litters and in control SPF sows. This was con¢rmed by the amount of PCV2 replication, as the genome load was far higher in piglets born to PCV2-infected non-vaccinated sows than in piglets born to infected vaccinated sows. Both vaccines were tested for PCV1 and PCV2 and the results were negative. It could therefore be concluded that the parvovirus and Erysipelas vaccines used were not contaminated by PCV1 or PCV2. However, it cannot be ruled out that the results observed may have been a consequence of (1) non-speci¢c cross-protection due to the adjuvant in the vaccines, (2) a general immune response produced by the antigens included in the vaccines, or (3) cross-protection of PPV vaccines against PCV2 infection. Thus, further investigations are needed. However, this study emphasizes the need for

ARTICLE IN PRESS 142

N. Rose et al.

the farmers to comply with the recommended schedules and methods of vaccination. No di¡erences between the vaccines were demonstrated in respect of farrowing results or clinical or histopathological ¢ndings. However, a slight di¡erence was observed in the group of piglets deprived of PCV2 passive immunity, in that the number of PCV2 genome copies per gram of tissue was lower in piglets born to PCV2-infected Bvaccinated sows. Sixteen of the 84 piglets born live to infected sows and foster sows were a¡ected by a syndrome possibly related to PMWS. Mild, moderate and acute clinical signs were observed in these piglets, but all developed histopathological changes (lymphoid depletion and in¢ltration of lungs and kidneys with in£ammatory cells). These piglets had a signi¢cantly heavier PCV2 genome load than did those without histological changes. Horizontal transmission of the virus apparently occurred early in the experiment as attested by the seroconversion of sentinel pigs; this occurred almost simultaneously in all piglets born to infected sows but deprived of PCV2 passive immunity. This viral transmission led to a clinical syndrome in eight contact piglets, all but one of which had been deprived of PCV2 passive immunity. Transfer of passive immunity from the dam to the litter was previously found (Rose et al., 2005) to be a key management factor. The results from the present study con¢rmed the importance of the post-natal immune status of the piglets with regard to the virus. Passive immunity protected against the clinical symptoms and the probability of PCV2 infection; in addition, the numbers of PCV2 genome copies per gram of tissue were signi¢cantly higher for piglets without PCV2 passive immunity than for those with such immunity. This result was con¢rmed for all categories, regardless of whether piglets were born to infected sows or were horizontally infected (contact) piglets. Moreover, quanti¢cation of PCV2 transmission based on the R0 indicated a more e⁄cient transmission of the virus from infectious to susceptible pigs when animals were deprived of PCV2 passive immunity. Early cross-fostering in a farm context may therefore have detrimental consequences when the immune status of the sows is unknown and presumably heterogeneous.

Acknowledgments The authors are grateful to AndreŁ Keran£ec’h and GeŁrard Bennevent for excellent management of the experiment. They are also indebted to the EU (STREP 513928 ‘‘Control of Porcine Circovirus Diseases’’) for ¢nancial support.

Appendix A. Statistical model for real-time PCR results of comparison between treatments (Berk and Lachenbruch, 2002). Let yijk be an observation from the kth measurement on the jth subject in the ith group of subjects where yijk are all non-negative, and let ymin be the smallest positive value. Let Pijk be the Bernouilli probability of a non-zero value and let zijk be 1 for a zero value and 0 for a non-zero value. For the log odds of the Pijk we have the linear model:   P ijk ln ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b4 x4 þ dij 1  P ijk where xn are indicators (orthogonal polynoms) for the x variable with n+1 categories. The random e¡ect dij is assumed to be normal, mean 0, constant variance t2, with independence between subjects. There is a similar linear model for the mean mijk of 1n (yijk), mijk ¼ g0 þ g1 x1 þ g2 x2 þ g3 x3 þ g4 x4 þ c  dij Given the random e¡ect, the standard deviation of 1n (yijk) is assumed to be a constant s. Given the random e¡ects, the likelihood contribution for a single observation is    lnðymin Þ  mijk zijk f ðyijk =dij Þ ¼ ð1  pijk Þ þ pijk F s   1zijk lnðyijk Þ  mijk  pijk f s with f the standard normal density and F the standard normal cumulative distribution function. The conditional likelihood for a single individual is obtained as a product Y gðyij1 ; yij2 ; . . . ; yijk =dij Þ ¼ f ðyijk =dij Þ k

and letting q(dij) be the normal density of the random e¡ect, the marginal likelihood for a single individual is Z hðyij1 ; yij2 ; . . . ; yijk Þ ¼ gðyij1 ; yij2 ; . . . ; yijk =dij Þqðdij Þddij Finally, the likelihood to be maximized is Y L¼ hðyij1 ; yij2 ; . . . ; yijk Þ i;j

References Albina, E.,Truong, C., Hutet, E., Blanchard, P., Cariolet, R., L’Hospitalier, R., MaheŁ, D., AlleŁe, C., Morvan, H., Amenna, N., Le Dimna, M., Madec, F. and Jestin, A. (2001). An experimental model for post-weaning

ARTICLE IN PRESS Parvovirus-Erysipelas Vaccination and PMWS

multisystemic wasting syndrome (PMWS) in growing piglets. Journal of Comparative Pathology, 125, 292^303. Allan, G. M. and Ellis, J. A. (2000). Porcine circoviruses: a review. Journal ofVeterinary Diagnostic Investigation, 12, 3^14. Allan, G. M., Kennedy, S., McNeilly, F., Foster, J. C., Ellis, J. A., Krakowka, S. J., Meehan, B. M. and Adair, B. M. (1999). Experimental reproduction of severe wasting disease by co-infection of pigs with porcine circovirus and porcine parvovirus. Journal of Comparative Pathology, 121, 1^11. Allan, G. M., McNeilly, F., McNair, I., Meehan, B. M., Marshall, M., Ellis, J. A., Lasagna, C., Boriosi, G., Krakowka, S. J., Reynaud, G., Boeuf-Tedeschi, L., Bublot, M. and Charreyre, C. (2002). Passive transfer of maternal antibodies to PCV2 protects against development of PMWS: experimental infections and a ¢eld study. Pig Journal, 50, 59^67. Allan, G. M., McNeilly, F., Meehan, B. M., McNair, I., Ellis, J. A., Krakowka, S. J., Fossum, C.,Wattrang, E.,Wallgren, P. and Adair, B. M. (2003). Reproduction of postweaning multisystemic wasting syndrome in pigs experimentally inoculated with a Swedish porcine circovirus 2 isolate. Journal of Veterinary Diagnostic Investigation, 15, 553^560. Balasch, M., SegaleŁs, J., Rosell, C., Domingo, M., Mankertz, A., Urniza, A. and Plana-Duran, J. (1999). Experimental inoculation of conventional pigs with tissue homogenates from pigs with post-weaning multisystemic wasting syndrome. Journal of Comparative Pathology, 121,139^148. Berk, K. N. and Lachenbruch, P. A. (2002). Repeated measures with zeros. Statistical Methods in Medical Research, 11, 303^316. Blanchard, P., Loizel, C., Baudouard, M. A., Nignol, A. C., Grasland, B., Dory, D., Morvan, H., Cariolet, R. andJestin, A. (2004). Quanti¢cation du geŁnome du circovirus porcin de type 2 (PCV2) par PCR en temps reŁel et correŁlation avec la maladie de l’amaigrissement du porcelet (MAP). LesJourneŁes de la Recherche Porcine, 36, 327^332. Blanchard, P., MaheŁ, D., Cariolet, R.,Truong, C., Le Dimna, M., Arnauld, C., Rose, N., Eveno, E., Albina, E., Madec, F. and Jestin, A. (2003). An ORF2 protein-based ELISA for porcine circovirus type 2 (PCV2) antibodies in postweaning multisystemic wasting syndrome (PMWS).Veterinary Microbiology, 94,183^194. Calsamiglia, M., Segales, J., Fraile, L., Espinal, A., Seminati, C., Martin, M., Mateu, E. and Domingo, M. (2004). Sow e¡ect on litter mortality in a swine integration system experiencing postweaning multisystemic wasting syndrome (PMWS). In: Proceedings of the18th International PigVeterinary Society Congress, Hamburg, Germany, p.78. Cariolet, R., Blanchard, P., Le Dimna, M., Truong, C., Keran£ech, A., Beaurepaire, B., Jolly, J. P., Julou, P., de Boisseson, C., MaheŁ, D., Madec, F. andJestin, A. (2002). Etude de di¡eŁrentes modaliteŁs d’inoculation du circovirus porcin de type 2 (PCV2) aØ des truies EOPS. LesJourneŁes de la Recherche Porcine, 34, 317^323. Cariolet, R., Le Diguerher, G., Ecobichon, P., Julou, P., Jolly, J. P. and Madec, F. (2004). Production of long term, lowcost speci¢c pathogen free pigs. In: Proceedings of the Sympo-

143

sium of the International Society for Animal Hygiene, SaintMalo, France, p. 149. De Jong, M. C. and Kimman, T. G. (1994). Experimental quanti¢cation of vaccine-induced reduction in virus transmission.Vaccine, 12,761^766. Farnham, M. W., Choi, Y. K., Goyal, S. M. and Joo, H. S. (2003). Isolation and characterization of porcine circovirus type-2 from sera of stillborn fetuses. CanadianJournal of Veterinary Research, 67,108^113. Grasland, B., Loizel, C., Blanchard, P., Oger, A., Nignol, A. C., BigarreŁ, L., Morvan, H., Cariolet, R. and Jestin, A. (2005). Reproduction of PMWS in immunostimulated SPF piglets transfected with infectious cloned genomic DNA of type 2 porcine circovirus. Veterinary Research, 36, 685^697. Harms, P. A., Sorden, S. D., Halbur, P. G., Bolin, S. R., Lager, K. M., Morozov, I. and Paul, P. S. (2001). Experimental reproduction of severe disease in CD/CD pigs concurrently infected with type 2 porcine circovirus and porcine reproductive and respiratory syndrome virus.Veterinary Pathology, 38, 528^539. Ihaka, R. and Gentleman, R. (1996). R: a language for data analysis and graphics. Journal of Computational Graphical Statistics, 5, 299^314. Kennedy, S., Mo¡ett, D., McNeilly, F., Meehan, B. M., Ellis, J. A., Krakowka, S. J. and Allan, G. M. (2000). Reproduction of lesions of postweaning multisystemic wasting syndrome by infection of conventional pigs with porcine circovirus type 2 alone or in combination with porcine parvovirus. Journal of Comparative Pathology, 122, 9^24. Kim, J., Han, D. U., Choi, C. and Chae, C. (2003). Simultaneous detection and di¡erentiation between porcine circovirus and porcine parvovirus in boar semen by multiplex seminested polymerase chain reaction. Journal of Veterinary Medical Science, 65,741^744. Kim, J., Jung, K. and Chae, C. (2004). Prevalence of porcine circovirus type 2 in aborted fetuses and stillborn piglets. Veterinary Record, 155, 489^492. Krakowka, S. J., Ellis, J. A., McNeilly, F., Ringler, S., Rings, D. M. and Allan, G. M. (2001). Activation of the immune system is the pivotal event in the production of wasting disease in pigs infected with porcine circovirus-2 (PCV2).Veterinary Pathology, 38, 31^42. Krakowka, S., Ellis, J. A., Meehan, B. M., Kennedy, S., McNeilly, F. and Allan, G. M. (2000). Viral wasting syndrome of swine: experimental reproduction of postweaning multisystemic wasting syndrome in gnotobiotic swine by coinfection with porcine circovirus 2 and porcine parvovirus.Veterinary Pathology, 37, 254^263. Kroese, A. H. and DeJong, M. C. (2001). Design and analysis of transmission experiment. In: Proceedings of the Annual Meeting of the Society forVeterinary Epidemiology and Preventive Medicine, Noordwijkerhout,The Netherlands, pp. 21^37. Larochelle, R., Bielanski, A., Muºller, P. and Magar, R. (2000). PCR detection and evidence of shedding of porcine circovirus type 2 in boar semen. Journal of Clinical Microbiology, 38, 4629^4632. Le Tallec, B., Pozzi, N., Blanchard, P., MaheŁ, D., Jestin, A. and GueŁrin, B. (2001). Longitudinal study of boars

ARTICLE IN PRESS 144

N. Rose et al.

naturally infected by PCV2. In: Proceedings of ‘‘ssDNA Viruses of Plants, Birds, Pigs and Primates’’, Saint-Malo, France, September 24^27, p. 120. Lopez-Soria, S., SegaleŁs, J., Rose, N., Vinas, M. J., Blanchard, P., Madec, F., Jestin, A., Casal, J. and Domingo, M. (2005). An exploratory study on risk factors for postweaning multisystemic wasting syndrome (PMWS) in Spain. PreventiveVeterinary Medicine, 69, 97^107. Magar, R., Larochelle, R.,Thibault, S. and Lamontagne, L. (2000). Experimental transmission of porcine circovirus type 2 (PCV2) in weaned pigs: a sequential study. Journal of Comparative Pathology, 123, 258^269. McIntosh, K. A., Harding, J. C. S., Ellis, J. A. and Appleyard, G. D. (2005). Nested PCR detection and duration of porcine circovirus type 2 in semen collected from naturally infected boars. In: Proceedings of the International Conference on Animal Circoviruses and Associated Diseases, Belfast, September,11^13, p. 93. Meehan, B. M., McNeilly, F., McNair, I., Walker, I., Ellis, J. A., Krakowka, S. J. and Allan, G. M. (2001). Isolation and characterization of porcine circovirus 2 from cases of sow abortion and porcine dermatitis and nephropathy syndrome. Archives of Virology, 146, 835^842. Okuda, Y., Ono, M., Yazawa, S. and Shibata, I. (2003). Experimental reproduction of postweaning multisystemic wasting syndrome in caesarean-derived, colostrum-deprived piglets inoculated with porcine circovirus type 2 (PCV2): investigation of quantitative PCV2 distribution and antibody responses. Journal of Veterinary Diagnostic Investigation, 15,107^114. Park, J. S., Kim, J., Ha, Y., Jung, K., Choi, C., Lim, J. K., Kim, S. H. and Chae, C. (2005). Birth abnormalities in pregnant sows infected intranasally with porcine circovirus 2. Journal of Comparative Pathology, 132, 139^144. Pensaert, M. B., Sanchez, R. E., LadekjaerMikkelsen, A. S., Allan, G. M. and Nauwynck, H. J. (2004).Viremia and effect of fetal infection with porcine viruses with special reference to porcine circovirus 2 infection. Veterinary Microbiology, 98,175^183. Rose, N., AbherveŁ-GueŁguen, A., Le Diguerher, G., Eveno, E., Jolly, J. P., Blanchard, P., Oger, A., Jestin, A. and Madec, F. (2005). E¡ect of the Pietrain breed used as terminal

boar on post-weaning multisystemic wasting syndrome (PMWS) in the o¡spring in four PMWS-a¡ected farms. Livestock Production Science, 95,177^186. Rose, N., Larour, G., Le Diguerher, G., Eveno, E., Jolly, J. P., Blanchard, P., Oger, A., Le Dimna, M., Jestin, A. and Madec, F. (2003). Risk factors for porcine post-weaning multisystemic wasting syndrome (PMWS) in 149 French farrow-to-¢nish herds. Preventive Veterinary Medicine, 61, 209^225. Rovira, A., Balasch, M., Segales, J., Garcia, L., PlanaDuran, J., Rosell, C., Ellerbrok, H., Mankertz, A. and Domingo, M. (2002). Experimental inoculation of conventional pigs with porcine reproductive and respiratory syndrome virus and porcine circovirus 2. Journal of Virology, 76, 3232^3239. Sanchez, R. E., Meerts, P., Nauwynck, H. J. and Pensaert, M. B. (2003). Change of porcine circovirus 2 target cells in pigs during development from fetal to early postnatal life. Veterinary Microbiology, 95,15^25. Sanchez, R. E., Nauwynck, H. J., McNeilly, F., Allan, G. M. and Pensaert, M. B. (2001). Porcine circovirus 2 infection in swine foetuses inoculated at di¡erent stages of gestation.Veterinary Microbiology, 83,169^176. SAS Institute Inc. (2000). SAS/STAT User’s Guide, SAS Institute, Cary, NC, USA. Segales, J. and Domingo, M. (2002). Postweaning multisystemic wasting syndrome (PMWS) in pigs. A review.Veterinary Quarterly, 24,109^124. Sorden, S. D. (2000). Update on porcine circovirus and postweaning multisystemic wasting syndrome (PMWS). Swine Health and Production, 8,133^136. West, K. H., Bystrom, J. M.,Wojnarowicz, C., Shantz, N., Jacobson, M., Allan, G. M., Haines, D. M., Clark, E. G., Krakowka, S. J., McNeilly, F., Konoby, C., Martin, K. and Ellis, J. A. (1999). Myocarditis and abortion associated with intrauterine infection of sows with porcine circovirus 2. Journal of Veterinary Diagnostic Investigation, 11, 530^532. 

Received, April 12th, 2006 Accepted, January 10th, 2007