Received: 13 September 2021 Revised: 18 February 2022 Accepted: 4 March 2022 DOI: 10.1002/vro2.34 O R I G I NA L R E S E A RC H The use of a whole inactivated PRRS virus vaccine administered in sows and impact on maternally derived immunity and timing of PRRS virus infection in piglets Gerard Eduard Martín-Valls Preben Mortensen Hepzibar Clilvert Yanli Li Martí Cortey Melanie Sno Timea Barna Marisa Terré Nicolas Guerra Enric Mateu, 1 Departament de Sanitat i Anatomia Animals, Abstract Facultat de Veterinària, Travessera dels Turons s/n, Universitat Autònoma de Barcelona, Barcelona, Background: Porcine reproductive and respiratory syndrome virus (PRRSV) vaccina- Spain tion is usually based on administering periodically PRRS modified live virus (MLV) in 2 Ceva Santé Animale, Libourne, France sows throughout their life. Using this schedule, transfer of maternally derived antibodies 3 to the offspring is limited. The aim of the present study was to test the concept of prim- Ceva Animal Health, Ceva Phylaxia Veterinary Biologicals Co. Ltd., Budapest, Hungary ing with an MLV and boosting with a commercial inactivated virus vaccine in sows to 4 Cooperativa d’Artesa de Segre, Artesa de Segre, reduce PRRSV incidence and improve productivity. Lleida, Spain Methods: On two farms, all the sows were vaccinated with a MLV vaccine at week 8 of 5 C, ReSA-IRTA-UAB, Campus UAB, Cerdanyola gestation. Then two groups were designated, one group was re-vaccinated in the third del Vallès, Barcelona, Spain week prior to farrowing and using a commercial inactivated vaccine (the PG group). The second group was the control group (C). Assays for PRRSV infection and productive Correspondence parameters were evaluated. Gerard Eduard Martín-Valls, Departament de Sanitat i Anatomia Animals, Facultat de Results: For both farms, the incidence of PRRSV was lower at 6 weeks of age in PG than Veterinària,Travessera dels Turons s/n, Universitat in C (p < 0.05). At weaning the proportion of PRRSV seropositive piglets was higher Autònoma de Barcelona, 08193 Cerdanyola del for PG as well (p < 0.05). The litters from C sows from both farms showed a higher Vallès, Barcelona, Spain. Email: gerard.martin@uab.cat pre-weaning mortality (odds ratio, C vs. PG = 1.18 ± 0.09; p < 0.05). Conclusions: Administration of the vaccine in sows before farrowing was safe and asso- ciated with reduced incidence of PRRSV in piglets up to 6 weeks of age. INTRODUCTION for reducing PRRSV transmission.13,14 However, vaccination of piglets interferes with monitoring and requires additional Porcine reproductive and respiratory syndrome (PRRS) is measures to avoid recombination between field and vaccine characterised by abortion, respiratory disease, increased piglet strains. mortality1–4 and secondary infections.5–7 In a previous study,8 Whole PRRSV inactivated vaccines (IV) can be useful to PRRS cost about US$ 664 million in the USA; weaners and boost previous immunity. This approach is appealing when growers accounted for 55% of the cost. In Europe, the aver- aimed at increasing the levels of maternally derived antibod- age cost during an outbreak has been estimated at €126 per ies (MDA). Passively transferred neutralising antibodies (NA) sow.9 may protect against the development of viraemia.15 Strate- Transmission of PRRSV may occur by several routes gies resulting in increased MDA transfer may contribute to including direct contact with fomites.10 Current vaccines do delaying the circulation of PRRSV and minimise the impact not produce sterilising immunity and transmission to or from of infection and disease. The present study explores this con- vaccinated pigs is possible. Controlling PRRS without vac- cept by using an IV before farrowing, after priming sows in cines is difficult. On most farms, the goal is to stop virus cir- the eighth week of gestation with a MLV. The efficacy of this culation among breeding pigs (stabilisation).11 Such control vaccination scheme was assessed by comparing incidences, programmes may start with administration of a modified live antibody levels and production parameters of piglets born vaccine (MLV) for a primary immunisation.12 Once stabilisa- from sows vaccinated with the combination of an IV with an tion is achieved, eliminating PRRSV in the piglet nursery is MLV compared with piglets born from sows receiving only the feasible by depopulation. Vaccination of piglets is an option MLV. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 The Authors. Veterinary Record Open published by John Wiley & Sons Ltd on behalf of British Veterinary Association. Vet Rec Open. 2022;9:e34. wileyonlinelibrary.com/journal/vro2  of  https://doi.org/10.1002/vro2.34  of  Veterinary Record Open Gilts (pre-mating) Adult sows Modified live vaccine (at least 8 weeks pre-mating) Blanket vaccination (2 months before the trials) Progressis (at least 4 weeks after MLV) Insemination P Treatment: PG r Control: C e 60 days: g 60 days: MLV n MLV a 90 days: n PRRSV dynamics Productive data Progressis c y • % positive litters Farrowing • Total newborns • Incidence at weaning O O • Stillbirths • Seropositive piglets at weaning f f • Mummies • Neutralizing antibodies at 2w ofage f f • Weaned piglets s s p Weaning p r r i i • Incidences at 6 and 9 weeks • % ofpost-weaning deaths n n g g FIGURE  Graphical description of the study design MATERIALS AND METHODS 1, only the gilts in the sixth batch in the PG group received a dose of IV before the first insemination; that is, 4 weeks after Selection of the farms and design of the study the last MLV dose in quarantine and, at least, 4 weeks before insemination. On Farm 2, all gilts in the PG group received the Two farrow-to-wean farms where PRRSV circulated in the IV before the first insemination, similarly to the sixth batch of farrowing units were included in the study. The circulating Farm 1. PRRSV strains were sequenced (whole genome; accession MZ318698 and MZ318699). Farm 1 (300 sows) was run on a 3-week batch basis while Farm 2 (1700 sows) operated on a Sampling weekly basis. Both farms obtained PRRSV-free gilts of about 6 months of age that were quarantined until the first service. Umbilical cords (UC) were collected at birth from Gilts were immunised with two doses (4 weeks apart) of a 6 piglets/litter (7–8 for the sixth batch of Farm 1) that PRRS MLV-Porcilis PRRSV (MSD) for Farm 1 and Pyrsvac- were ear-tagged. To avoid bias, cross-fostering was limited 183 (Syva) for Farm 2 before the first insemination. These to the minimum possible to assure survival of piglets (only farms had two physically separated farrowing units (≥14 sows) during the first 24 h of life). The animals that were cross- and this enabled at least two separated flows of weaners (all- fostered were identified by an ear-tag and were excluded. in/all-out, no shared airspace). Farm 1 weaned pigs at 28 days Animal movements between farrowing crates were recorded. and Farm 2 at 21 days. On both farms, sows were blanket vac- On Farm 1, 155 litters were examined (77 PG and 78 C sows, cinated with the MLV vaccine three times a year. with 785 newborn piglets, 473 PG and 480 C followed for the Before the beginning of the trial, all sows were vaccinated whole study period). On Farm 2, 148 litters were examined once with the regularly used MLV in the farms to provide a (77 PG and 71 C sows, with 826 newborn piglets, 448 PG and baseline for their immune status. 418 C piglets followed). The design of the study (Figure 1) was to produce a treat- At weaning, ≥3 piglets/sow (among those where UC had ment (PG) and a control group (C). On both farms and for all been collected) were blood sampled in order to compare batches, piglets in half of the farrowing rooms were weaned ≥45 animals/group/batch (to evaluate a 50% reduction in in a nursery unit and those from the other half were weaned prevalence, considering a 50% basal prevalence per batch at in a separated unit. One half of the sows were allocated to weaning, with 95% confidence and 80% power). The hypoth- the treatment group (PG) and the other half to the control esis was that the proportion of PRRSV-infected animals in group (C). All sows were administered the MLV at the eighth the PG group would be half of the proportion in group C week of pregnancy and were either re-vaccinated with the IV (95% confidence). To evaluate the PRRSV incidence in nurs- (PROGRESSIS, group PG) 4 weeks before the expected far- eries at least 30 pigs/group among those sampled at wean- rowing date (around 90th day of gestation) or did not receive ing were sampled again at 6 and 9 weeks of age (WOA) (50% any other booster (group C). Parity was matched as much as reduction of positive animals per batch, 95% confidence, 80% possible between PG and C sows. Six batches of piglets were power, considering an initial prevalence of 70% at 6 WOA). followed on Farm 1 (representing 77 PG sows and 78 C sows) Table 1 summarises the number of animals followed up per and three batches on Farm 2 (77 PG and 71 C sows). On Farm age, group, batch and farm. Veterinary Record Open  of  TA B L E  Samples collected per farm, batch and age/type Number of piglets Farm number Batch number Umbilical cord  weeks  weeks – weeks Farm 1 #1 PG 96/16 51 43 43 C 96/16 53 31 31 #2 PG 78/13 50 30 30 C 84/14 50 29 29 #3 PG 84/14 47 46 46 C 84/14 49 49 49 #4 PG 84/14 65 32 31 C 84/14 50 32 32 #5 PG 84/14 48 46 42 C 84/14 48 44 40 #6 PG 47/6 45 44 44 C 48/6 47 46 46 Total PG 473/77 308 241 236 C 480/78 295 231 227 Farm 2 #1 PG 126/21 83 83 82 C 126/21 85 85 83 #2 PG 126/21 85 85 87 C 126/21 85 85 75 #3 PG 196/35 100 99 87 C 166/29 87 85 75 Total PG 448/77 268 267 254 C 418/78 257 255 243 Note: PG offspring of sows receiving the whole virus inactivated vaccine. C offspring of control sows. The column ‘umbilical cord’ shows the number of samples collected/number litters examined. Analysis of the samples The results obtained at birth were compared with the appear- ance of RT-PCR-positive pigs at weaning to evaluate if trans- Umbilical cords were homogenised and pooled in pairs. For mission chains had occurred. positive pools, individual samples were examined afterwards. Sera were selected and evaluated by the viral neutralisa- Sera were analysed individually. tion test (VNT) using the MLV of each farm as antigen.16 On RNA was extracted with the kit MagMax Core nucleic Farm 1, at least 23 sera/group were collected before weaning acid purification kit (Life Technologies). RT-qPCR for the in batches 1–4. In batch #6, a further 23 extra piglets/group detection of PRRSV was performed using the LSI VetMax were randomly selected. For Farm 2, 49 and 52 sera from PRRSVEU/NA v2.0 kit (Life Technologies). According to PG and C groups were randomly selected from batches manufacturer instructions, Ct values <40 were considered 1 and 2. positive. Antibodies were analysed with a commercial ELISA (Idexx PRRSV X3). Animals were analysed at weaning and 9 WOA. Reproductive data Animals that tested negative by RT-qPCR throughout the study and were positive by ELISA at 9 WOA were considered The following data were collected. (1) Total piglets born per to have been infected. litter, the number of born alive piglets per litter, stillbirths per A vertical transmission event was considered to have hap- litter, mummified per litter and weaned piglets per litter and pened when at least one UC per litter tested positive by RT- (2) mortality in the nurseries. These data were obtained for qPCR. At weaning, 6 and 9 WOA, the proportion of positive all sows in each batch (n = 89 for PG and n = 72 for C sows animals was calculated based on the RT-qPCR results. Inci- and their respective progenies on Farm 1; n = 131 for PG and dence was calculated considering the number of susceptible n = 135 for C sows on Farm 2) and not only for the ones animals (S) and new cases (NC) for each sampling period. included in the PRRSV follow-up. Cumulative incidence was calculated by aggregating the data between sampling periods (weaning to 6 weeks, 6–9 weeks). An animal found to be infected in two consecutive samplings Statistical analysis was considered as a NC only in the first positive sampling time. Animals seroconverting at 9 WOA were added as NC for this To assess the effect of vaccination on PRRSV incidence (wean- sampling time-point. Only the animals that could be followed ing to 6 weeks and 6–9 weeks) a generalised linear mixed- throughout the study were considered for these calculations. effects model (GLMM) in Rstudio Cloud (glmer function;  of  Veterinary Record Open emmeans, Matrix and lme4 libraries) was used. Treatment be viraemic than piglets in the PG group. A trend for lower Ct (IV vaccination vs. non-vaccination) and farm were consid- values in infected piglets of group PG was observed (29.7 ± 5.8 ered fixed effects; the batch was considered as a random effect for C vs. 32.6 ± 4.8 for PG, p = 0.0503). The maximum inci- nested on the farm. Cumulative incidences from 4 to 6 and dence observed was at 9 WOA in both groups. Figure 2 shows from 6 to 9 WOA and the proportion of seropositive animals the evolution of the cumulative incidence of each group. at weaning were aggregated per batch and compared between vaccination groups per farm using a chi-square test with Yates correction. Relative risk was calculated using the Koopman’s Detailed PRRSV dynamics for Farm  likelihood-based approximation. Neutralisation titres (log2 converted for normalisation of For Farm 2, circulation of PRRSV was also detected at wean- the data) were compared between PG and C groups by the ing before the beginning of the trial. When examining UC, ANOVA and Kruskal–Wallis tests (non-normally distributed 14.3% of the PG litters (11/77; CI95% : 7.3%−24.1%) and 8.4% variables) on StatsDirect 3.2.10. (6/71; CI95% : 3.2%−17.5%) of the C litters contained PRRSV- For reproductive data, a GLMM was used considering the positive UC (non-significant). The PG and C UC samples data as binomial. In this model, pre-weaning mortality in a yielded similar Ct values (37.9 vs. 37.5). On this farm, 1/11 litter could be influenced (fixed effects) by the treatment, the litters in the PG group had RT-qPCR-positive piglets at number of live piglets that were born and the parity of the weaning. sow. The interaction between the treatment and the number On this farm, the cumulative incidences at 3, 6 and 9 WOA of piglets born alive was included since vaccination could pre- were similar in all groups (see Figure 2). These results were vent vertical transmission. Data were transformed logarithmi- affected by increased circulation of the virus at 6 WOA in the cally. Batch was used as a random effect. Since GLMM does third batch (>97% incidence) in both PG and C piglets. Since not calculate the p-values, these were obtained using the Wald the behaviour of the infection in the third batch differed sub- chi-square test (car library) and likelihood ratio tests (drop1 stantially from that of previous batches in this farm or Farm function). 1, this batch was considered to be an outlier. Considering only the first two batches, the cumulative incidence at 6 WOA was significantly lower in PG group (11.3%; CI95% : 6.9%−17.1% vs. RESULTS 22.4%; CI95% : 16.1–28.6 of incidence; p < 0.05) resulting in a relative risk of 1.9 (CI95% : 1.2–3.3; p < 0.01) for C versus PG. Dynamics of PRRSV circulation as assessed by The Ct values were similar in both groups (32.7 ± 3.2 for C vs. RT-PCR 34.3 ± 2.9 for PG). Figure 2 shows the incidences for Farms 1 and 2 for the aggregated batches. Supporting Information S1 Before the trial, PRRSV was detected at weaning by RT-qPCR, shows the individual results per batch. confirming that the farm was infected. In the following text, the results are firstly summarised per farm and batch and then are aggregated in the GLMM. ELISA and viral neutralisation test Farm 1 Comparison of PRRSV incidences between PG and C groups On Farm 1, the proportion of seropositive (ELISA) piglets at weaning in the offspring of PG sows was 94.1% versus 86.5% Generalised linear mixed-effects model showed that, for both in C piglets (p < 0.01). This corresponded to a relative risk farms, from 4 to 6 WOA, the PRRSV incidence was lower in of 2.4 (CI95% : 1.3–4.4; p < 0.01) indicating that PG animals the offspring of PG sows compared to the C group (p = 0.014). were more likely to have anti-PRRSV MDA at that age. When From 6 to 9 weeks, after fading of MDA, the incidence was comparing the average S/P ratios for the whole population higher in the offspring of PG than in C (p = 0.044) (see or only for the seropositive piglets, no significant differences Table 2). were observed (1.08 ± 0.45 vs. 1.04 ± 0.50 in PG and C for all animals; 1.13 ± 0.42 for PG vs. 1.16 ± 0.45 for C if only seropositive animals were considered). When the S/P ratios Detailed PRRSV dynamics for Farm  were examined with regards to the parity of the sow, the oldest PG sows tended to have higher S/P ratios than C sows, except Regarding the UC results, 12.7% (8/63; confidence interval those of parities 3 and 4 (Figure 3). 95% (CI95% ): 5.6%–23.5%) of the PG litters had positive results Regarding the VNT before weaning, the average log2 titre with an average Ct of 34.5 ± 3.4. In C, 10.9% of the litters for PG was 3.6 ± 0.2 versus 3.5 ± 0.4 for C group (non- had positive RT-qPCR results (7/64, CI95% : 4.5%–21.2%, non- significant). The dispersion of titres was different in the vacci- significant) with an average Ct of 36.5 ± 3.1. Only 2/15 lit- nated and control groups (Figure 4). Titres >5log2 were only ters tested positive by RT-qPCR at weaning (one PG and observed in the offspring of PG sows (Supporting Information one C). S2). For Farm 1, 467 pigs were followed from birth until 9 WOA Gilts from the sixth batch were boosted with the IV after (236 PG and 231 C). Cumulative incidence until 6 WOA was priming and before mating. They then received a second IV 25.5% in C group and 14.8% of piglets in group PG tested dose before farrowing. Comparison of the MDA levels in the positive (p < 0.05), resulting in a relative risk of 1.57 (CI95% : offspring showed that S/P ratios and NA titres tended to be 1.07–2.29). Piglets in group C were 1.57 times more likely to higher in the PG group (p = 0.08 and 0.03, respectively) Veterinary Record Open  of  TA B L E  Output of the generalised linear mixed models used in the present work for the comparison of productive data and incidences Standard Evaluated variables Estimate error z-Value Pr(>|z|) F-value p-Value (chi) Mortality All batches included Intercept −5.51 0.670 −8.225 <2e-16 GROUPV −0.73 0.940 −0.773 0.440 log(ALIVE) 1.45890 0.250 5.834 5.41e-09 log(PARITY) 0.00191 0.047 0.040 0.968 GROUPV:log(ALIVE) 0.28830 0.353 0.816 0.414 Batch 5 from Farm 1 excluded Intercept −6.116 0.739 −8.268 <2e-16 GROUPV −3.003 1.070 −2.805 −0.005 0.136 0.005 log(ALIVE) 1.706 0.274 6.205 5.45e-10 121.72 5.453e-10 log(PARITY) −0.024 0.051 −0.483 0.628 0.083 0.628 GROUPV:log(ALIVE) 1.097 0.400 2.741 0.006 7.504 0.006 PRRSV incidence Incidence 4–6 weeks (both farms) Intercept −2.09089 1.09 −2.662 0.007 GROUPV −1.6307 0.66 −2.446 0.014 5.774 0.0144 Incidence 6–9 weeks (both farms) Intercept −1.086 0.691 −1.572 0.115 GROUPV 0.314 0.156 2.017 0.043 4.076 0.0436 Note: GROUPV represents PG group, log(ALIVE) represents born alive, log(PARITY) represents the parity of sows, and GROUPV:log(ALIVE) represents the interaction between both variables. For productive data, only when batch #5 from Farm 1 was excluded it was possible to observe an effect of the vaccination group, which resulted in a reduction of the mortality for PG group (GROUPV). In the case of the incidences, the estimates indicate that that the PRRSV incidence was lower for PG group (negative value for GROUPV) before 6 weeks of age, and vice versa from 6 to 9 weeks. F I G U R E  Cumulative incidence at weaning (3–4 weeks), 6 and 9 weeks of age observed per batch for Farms 1 and 2. Grey bars correspond to PG group and empty bars to C group. *Significative differences at 6 weeks of age, where PG group showed lower incidence from 4 to 6 weeks of age  of  Veterinary Record Open Distribution of neutralization titres (log2) PG C 0 1 2 3 4 5 6 7 min < LQ < median > UQ > max, fences (1.5 & 3.0 IQR) Distribution of neutralization titres (log2) PG C 0 1 2 3 4 5 6 7 8 min < 1 SD < mean > 1 SD > max, fences (1.96 SD, 2.58 SD) F I G U R E  Distribution of S/P ratios by parity of the sow in Farm 1 (batches 1–4 and 6) and Farm 2 (batches 1 and 2). The whiskers and boxes F I G U R E  Distribution of the neutralisation titres in the pre-weaning depict, from left to right: minimum, lower quartile, median (diamond), piglets per group in Farm 1 (upper) and Farm 2 (lower). For Farm 1, batches upper quartile and maximum. The Y-axis indicate the parity of the sows in 1–4 are shown (no pre-mating vaccination in primiparous sows) and for PG and C groups. Statistically significant differences (p < 0.05) are shown by Farm 2 batches 1 and 2. Whiskers and boxes show minimum, lower quartile different letters between groups within each farm. Asterisks indicate a (25%), median, upper quartile (75%) and maximum. Titres are shown as p-value between 0.1 (*) and 0.05 (**) log2 values (Figure 5). The distribution of the NA titres in the PG group was more homogeneous than in the C group. for PG animals was 6.0 ± 1.7 versus 5.5 ± 1.9 in C animals (p = 0.08) (Figure 4). When the titres were considered as ‘high’ (≥1:32) or ‘low’ (1:4 to 1:16) some 83.0% of the PG piglets had Farm 2 high titres while 63.3% of the C piglets had them (χ2 = 5.1; p < 0.05). When considering only the two first batches, at weaning 158/168 (94.1%; CI95% : 90.5%–97.1%) animals were positive in the PG group and 129/170 (75.9%; CI95% : 69.5%–82.3%) were Reproductive data positive in the C group (p < 0.0001). Considering all batches, the proportion of seropositive piglets in the offspring of PG On Farm 1, suckling piglets in batch #5 suffered an outbreak sows was higher (74.9% vs. 59.9%, for PG and C, respectively, of diarrhoea. The outbreak started in a room where PG sows p = 0.02). No differences were observed when comparing S/P were present. During the outbreak mortality reached 13% and ratios for positive animals considering either batches #1 and 24% for C and PG groups, respectively. The RT-qPCR anal- #2 or the three batches. When the distribution of S/P ratios in ysis of batch #5 piglets (2 and 4 WOA) showed that PRRSV- positive piglets only was examined, regarding the parity of the positive animals belonged to three litters in each group. None sow, differences were clear (Figure 3). Thus, in the PG group, of the PCR-positive piglets of batch #5 died before weaning. S/P ratios were similar in all piglets regardless of the parity Since the outbreak started in the PG group and was associated of their mothers. While in the C group the S/P ratios of the with higher mortality in that group, this batch was biased by offspring of young sows (parities 1 and 2) were lower than the this event. As the present trial was designed as an intention-to- average S/P ratios of older sows (parity ≥7) or than the S/P treat study and batch #5 was biased by a cause external to the ratios of piglets born from PG sows of any parity (p < 0.05). treatment affecting some of the assessed outcomes, batch #5 The VNT was carried out only with seropositive animals violated the protocol and was removed from the calculations. (ELISA) in batches 1 and 2. The average log2 titre of MLV NA Even so, in terms of MDA and PRRSV incidence at 6 and 9 Veterinary Record Open  of  F I G U R E  S/P ratios (a) and neutralisation titres (log2 ) (b) S/P values ofpiglets(ELISA) of1st parity sows-Batch6 per group (PG and C) in the offspring of parity 1 (P1) sows of Farm 1. Whiskers and boxes show minimum, lower quartile (25%), P=0.07 median, upper quartile (75%) and maximum. PG piglets tended to have higher S/P ratios at weaning than C sows (p = 0.09) P1-2-PG P1-2-C 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 min < LQ < median > UQ > max, fences(1.5 & 3.0 IQR) Distribution ofneutralization titres in 1st parity sows- Batch 6 P=0.03 GroupPGG GroupC 1 2 3 4 5 6 min < LQ < median > UQ > max, fences (1.5 & 3.0 IQR) WOA, batch #5 performed as the other five batches. Results nation programmes for sows, although other systems such as for batch #5 are presented in Supporting Information S3. herd closure may be applied.17,18 Nevertheless, on some farms, No differences were observed regarding the number of new- intensive vaccination programmes for sows are not sufficient. born, stillborn or mummified piglets per litter. The GLMM In such instances additional actions such as vaccination of analysis including all batches on both farms did not show a piglets may potentially reduce transmission of PRRSV1.13,14 significant impact on the survival of suckling piglets. When Although MDA may partially block vaccination.19,20 batch #5 was removed (because of the diarrhoea outbreak) Another strategy would be to increase the NA titres against the results suggested that vaccination had a significant effect PRRSV by passive transfer. From previous studies,15 1:8 to on reducing mortality (see Table 2). The effect was estimated 1:16 NA titres are reported to protect piglets in the homol- with an odds ratio of 1.18 ± 0.09 (p = 0.0405) (C vs. PG). ogous model challenge. Therefore, theoretically increasing Mortality was also related to the number of born alive piglets anti-PRRSV MDA might help to decrease the incidence of and its interaction with the treatment. This interaction was an PRRSV at earlier ages. It could be contented that a recall pre- essential variable of the GLMM as calculated with the likeli- farrowing vaccination would be helpful for that purpose. The hood ratio test (p = 0.00661). Parity had no significant effect. IV produces immunity when administered repeatedly.12 The Supporting Information S3 shows the reproductive data. No safety of such vaccines makes them candidates for a MDA significant differences were observed regarding mortality in boosting strategy. nurseries (Supporting Information S4). In our study on both farms the offspring of PG sows had a higher proportion of ELISA-positive animals, suggesting that anti-PRRSV humoral immunity was boosted. The strat- DISCUSSION egy resulted in the homogenising of the antibody values in the progeny of treated sows. On Farm 1, the offspring of primi- The impact of PRRS virus-associated disease in piglets will parous PG sows tended to have a NA titre that was 1log2 higher depend on several factors including the flow of viraemic than the control gilts. On Farm 2, these differences could be piglets from farrowing units to nurseries and this can result observed for all parities and the number of piglets showing in higher costs. Therefore, reducing the transmission of titres over 5log2 was significantly higher in the PG group. PRRSV in the farrowing units and nurseries is crucial for The effect of vaccination on the antibody response seemed controlling the disease. Most often this is intended by vacci- to be higher on Farm 2 than Farm 1. This may have  of  Veterinary Record Open several explanations. The MLVs used on Farms 1 and 2 age of seropositive piglets at weaning dropped. One possible and the circulating strains were different. This could lead explanation is a problem related to the vaccination although to different interactions with the IV, resulting in different we have no evidence. Sows were vaccinated with the MLV boosting of the humoral response. It is known that neutral- in mid-July and with PG in mid-August, with environmen- ising cross-reactivity is extremely diverse.21 It would have tal temperatures reaching as high as 34◦ C–39◦ C on the days been interesting to evaluate the impact of the vaccination of vaccination. Ventilation was managed by manually open- strategy on the potential neutralisation of the field strains; ing windows and we cannot discard some effect from thermal unfortunately, it had not been possible to adapt those strains stress. to MARC-145 cells, therefore the MLV strains were used. Also With current knowledge, it is not possible to know if pre- the management on Farms 1 and 2 was different, which could farrowing vaccination with an MLV would have produced lead to different policies in the farrowing units with different similar results to the ones obtained with the IV. However, the colostrum intake and consequently different antibody titres. use of a MLV at 90 days of gestation may raise safety concerns It is worth noting that in most, but not all, of the exam- (e.g. the vertical transmission of the MLV or a decrease in the ined batches, the incidence of PRRSV at 4 or 6 WOA was number of piglets born alive).24 lower in the PG pigs. This would agree with the previous It is important to comment that since the field strains could observation of a higher proportion of seropositive animals at not be adapted to culture with MARC-145, despite being easily weaning and suggests a beneficial effect of the pre-farrowing isolated in macrophages, the neutralisation results were pro- vaccination with an IV in delaying the spread of the virus duced using the resident MLV strain. Hence, we cannot be at earlier ages. After the fade out of the MDA, animals were sure how effective the antibodies were for neutralising the spe- infected at a similar rate. Thus, when MDA disappeared cific farm strain. Even so, it can be considered that the higher the incidence increased in the PG group. This observa- the NA titres the higher the probability of heterologous neu- tion is interesting because it is thought that the older the tralisation. animal, the lower the impact of PRRSV infection.22 Even The present study suggests that IV can be used for boosting so, no differences in weaner mortality were observed. For previous immunity in sows and can be helpful to homogenise future studies, it would be interesting to extend the obser- the transfer of PRRSV-specific MDA. A larger study would be vation period in nurseries and add parameters other than needed to assess the precise productive and economic impact mortality. under different circumstances. In the present study, no differences were observed in the litter productivity or the number of the PRRSV-positive litters ACKNOWLED GEMENT S at birth. In a previous study,23 it was reported that Ct values The authors thank the staff members of farms ‘Castilla i >30 most likely represented environmental contamination Dorado’ and ‘Fages’ and workers and veterinarians of Cooper- from facilities, sow secretions or infected littermates. In the ativa d’Artesa de Segre for their collaboration. Also Jordi Soto present study, the average Ct values of UC were >34, indicat- for his help in the first sampling. ing probable contamination from the environment. Only one positive litter at birth per group was related to the presence CONFLICTS OF INTEREST of positive animals at weaning would strengthen this hypoth- Preben Mortensen, Nicolas Guerra, Melanie Sno and Timea esis. Another consideration is that, given similar infectious Barna work for Ceva Santé Animale. The other authors have pressures in PG and C groups at birth, the transmission was not declared any conflicts of interest. less efficient in the PG group than in the C group, suggesting that protection after birth was improved in PG. It is worth F U N D I N G I N F O R M AT I O N noting that in both farms the circulating virus was always CEVA Santé Animale. identified as wild type. In the present work, sows received the IV only in one gesta- E T H I C S S TAT E M E N T tion. Given the results, it is reasonable to think that repeated All the procedures were approved by the Ethics Committee in cycles of vaccination may have increased the observed differ- Animal and Human Research of the Universitat Autònoma de ences. Barcelona (approval numbers 3221-CEEA-UAB and CEEAH- The scheme of vaccination in the present study was a mod- 5691 for the project and the procedure, respectively). ification of the recommended use of the IV in which the con- cept of initial vaccination with a MLV was introduced. Both DATA AVA I L A B I L I T Y S TAT E M E N T farms maintained the vaccination scheme after the study. At The datasets used and/or analysed during the current study present, Farm 1 remains positive in the nurseries while on are available from the corresponding author on reasonable Farm 2 PRRSV1 is no longer detected. request. It is important to comment on the exclusion of batch #5 in Farm 1 because of a diarrhoea outbreak unrelated to PRRSV. REFERENCES Since the impact of the outbreak was different in the PG and 1. Terpstra C, Wensvoort G, Pol JM. Experimental reproduction of porcine C groups we considered excluding this batch. It is not possible epidemic abortion and respiratory syndrome (mystery swine disease) to report the levels of PRRSV antibodies of the dead animals. by infection with Lelystad virus: Koch’s postulates fulfilled. Vet Q. Therefore, we thought that the most reasonable approach was 1991;13:131–6. 2. Pol JM, van Dijk JE, Wensvoort G, Terpstra C. Pathological, ultrastruc- to exclude it from the analysis and to show the results in Sup- tural, and immunohistochemical changes caused by Lelystad virus in porting Information. experimentally induced infections of mystery swine disease (synonym: In the case of Farm 2, in batch #3 an increase in the inci- porcine epidemic abortion and respiratory syndrome (PEARS)). Vet Q. dence of PRRSV was noticed in both groups and the percent- 1991;13:137–43. Veterinary Record Open  of  3. Christianson WT, Collins JE, Benfield DA, Harris L, Gorcyca DE, 16. Yoon IJ, Joo HS, Goyal SM, Molitor TW. A modified serum neutral- Chladek DW, et al. Experimental reproduction of swine infertility ization test for the detection of antibody to porcine reproductive and and respiratory syndrome in pregnant sows. Am J Vet Res. 1992;53: respiratory syndrome virus in swine sera. J Vet Diagn Invest. 1994;6: 485–8. 289–92. 4. Plana J, Vayreda M, Vilarrasa J, Bastons M, Rosell R, Martinez M, et al. 17. Corzo CA, Mondaca E, Wayne S, Torremorell M, Dee S, Davies P, et al. Porcine epidemic abortion and respiratory syndrome (mystery swine Control and elimination of porcine reproductive and respiratory syn- disease). Isolation in Spain of the causative agent and experimental drome virus. Virus Res. 2010;154:185–92. reproduction of the disease. Vet Microbiol. 1992;33:203–11. 18. Berton P, Normand V, Martineau GP, Bouchet F, Lebret A, Waret- 5. Galina L, Pijoan C, Sitjar M, Christianson WT, Rossow K, Collins JE. Szkuta A. Evaluation of porcine reproductive and respiratory syndrome Interaction between Streptococcus suis serotype 2 and porcine repro- stabilization protocols in 23 French Farrow-to-finish farms located in a ductive and respiratory syndrome virus in specific pathogen-free piglets. high-density swine area. Porcine Health Manag. 2017;3:11. Vet Rec. 1994;134:60–4. 19. Fablet C, Renson P, Eono F, Mahé S, Eveno E, Le Dimna M, et al. 6. Brockmeier SL, Palmer MV, Bolin SR. Effects of intranasal inoculation of Maternally-derived antibodies (MDAs) impair piglets’ humoral and cel- porcine reproductive and respiratory syndrome virus, Bordetella bron- lular immune responses to vaccination against porcine reproductive and chiseptica, or a combination of both organisms in pigs. Am J Vet Res. respiratory syndrome (PRRS). Vet Microbiol. 2016;192:175–80. 2000;61:892–9. 20. Renson P, Fablet C, Andraud M, Normand V, Lebret A, Paboeuf F, 7. Yu J, Wu J, Zhang Y, Guo L, Cong X, Du Y, et al. Concurrent highly et al. Maternally-derived neutralizing antibodies reduce vaccine efficacy pathogenic porcine reproductive and respiratory syndrome virus infec- against porcine reproductive and respiratory syndrome virus infection. tion accelerates Haemophilus parasuis infection in conventional pigs. Vet Vaccine 2019;37:4318–24. Microbiol. 2012;158:316–21. 21. Martínez-Lobo FJ, Díez-Fuertes F, Simarro I, Castro JM, Prieto C. 8. Holtkamp DJ, Kliebenstein JB, Neumann EJ, Zimmerman JJ, Rotto HF, Porcine reproductive and respiratory syndrome virus isolates differ in Yoder TK, et al. Assessment of the economic impact of porcine reproduc- their susceptibility to neutralization. Vaccine 2011;29:6928–40. tive and respiratory syndrome virus on United States pork producers. J 22. Martín-Valls GE, Hidalgo M, Cano E, Mateu E. Testing of umbilical Swine Health Prod. 2013;21:72–84 cords by real time PCR is suitable for assessing vertical transmission of 9. Nieuwenhuis N, Duinhof TF, van Nes A. Economic analysis of outbreaks porcine reproductive and respiratory syndrome virus under field condi- of porcine reproductive and respiratory syndrome virus in nine sow tions. Vet J. 2018;4:27–9. herds. Vet Rec. 2012;170:225. 23. Klinge KL, Vaughn EM, Roof BR, Bautista EM, Murtaugh MP. Age- 10. Pileri E, Mateu E. Review on the transmission porcine reproductive and dependent resistance to porcine reproductive and respiratory syndrome respiratory syndrome virus between pigs and farms and impact on vac- virus replication in swine. Virol J. 2009;6:177. cination. Vet Res. 2016;47:108. 24. Scortti M, Prieto C, Martínez-Lobo FJ, Simarro I, Castro JM. Effects 11. Holtkamp DJ, Polson DD, Torremorell M, Morrison B, Classen DM, of two commercial European modified-live vaccines against porcine Becton L, et al. Terminology for classifying swine herds by porcine reproductive and respiratory syndrome viruses in pregnant gilts. Vet J. reproductive and respiratory syndrome virus status. Swine Heal Prod. 2006;172:506–14. 2011;19:44–56. 12. Díaz I, Gimeno M, Callén A, Pujols J, López S, Charreyre C, et al. Com- parison of different vaccination schedules for sustaining the immune response against porcine reproductive and respiratory syndrome virus. S U P P O R T I N G I N F O R M AT I O N Vet J. 2013;197:438–44. Additional supporting information may be found in the 13. Pileri E, Gibert E, Soldevila F, García-Saenz A, Pujols J, Diaz I, et al. Vac- online version of the article at the publisher’s website. cination with a genotype 1 modified live vaccine against porcine repro- ductive and respiratory syndrome virus significantly reduces viremia, viral shedding and transmission of the virus in a quasi-natural exper- imental model. Vet Microbiol. 2015;175:7–16. 14. Rose N, Renson P, Andraud M, Paboeuf F, Le Potier MF, Bourry How to cite this article: Martín-Valls GE, Mortensen O. Porcine reproductive and respiratory syndrome virus (PRRSv) P, Clilvert H, Li Y, Cortey M, Sno M, et al. The use of a modified-live vaccine reduces virus transmission in experimental con- whole inactivated PRRS virus vaccine administered in ditions. Vaccine 2015;33:2493–9. sows and impact on maternally derived immunity and 15. Lopez OJ, Oliveira MF, Garcia EA, Kwon BJ, Doster A, Osorio FA. Pro- timing of PRRS virus infection in piglets. Vet Rec tection against porcine reproductive and respiratory syndrome virus (PRRSV) infection through passive transfer of PRRSV-neutralizing anti- Open. 2022;9:e34. https://doi.org/10.1002/vro2.34 bodies is dose dependent. Clin Vaccine Immunol. 2007;14:269–75.