The mechanisms of SIV introduction in farrow-to-finish pig farms in Thailand have not been well studied. In this study, we conducted a longitudinal surveillance in farrow-to-finish pig farms located in the central part of Thailand to develop a proper strategy for SIV surveillance. We found that young pigs, in particular, piglets at the age of 8 weeks or younger could be the target animals to isolate SIVs circulating in farms. Seroprevalence against SIVs in fattening pigs was evidence that SIV infection did occur within farms, while results of a phylogenetical analysis suggested that farm-to-farm transmission had occurred. In addition, a discrepancy between the HI test and ELISA suggested the possibility that the sub-lineages of H1 and H3 SIVs that have not yet been isolated may be circulating in the Thai pig population. Thus, information obtained in this study would be useful for conducting SIV surveillances in farrow-to-finish farms.
In this study, SIVs were most frequently isolated from weaned piglets aged 4 and 8 weeks. Previous findings also pointed out that the majority of SIV infections take place in piglets aged under 10 weeks . Weaned piglets are considered to be susceptible to SIVs because the concentration of the maternal antibodies against SIVs in the serum declines with age in piglets , and the half life of antibodies against H1 and H3 SIVs was estimated to be 12 days . The high density of pigs in piggeries and large size herds is a contributing factor to the high SIV prevalence rates in fattening pigs and sows [27, 28]. Thus, gathering of weaned piglets in one nursery together with other piglets farrowed from different sows is another likely factor contributing to the high isolation rate in weaned piglets in farrow-to-finish farms.
Dissemination of the SIV of a particular genotype was suggested based on the fact that the H1N1 SIVs isolated from Farms A and B shared a common ancestor. Pigs and other materials were not transferred between farms and moreover, the farms were separated geographically by more than 100 km, suggesting that SIVs have spread extensively in Thailand. The introduction of pigs carrying SIVs is one of the most likely factors for viral dissemination among farms . The isolation of Sara109713-36 from a pig introduced to Farm B may have originated from the farm from which this pig was introduced. At the same time, there remains a possibility that the pig was infected after being introduced to the farm, because the affected pig was introduced into Farm B 4 days prior to the sampling date. A period of 4 days is known to be enough for pigs to start virus shedding after experimental infection . The other pigs introduced earlier than the affected pig were also in the same quarantined piggery, although they were separated into different compartments. In addition, there were no regulations for the movement of humans between piggeries (quarantine piggery, breeding/farrowing sites, weaned sites and fattening sites) in that farm.
Serological analysis revealed that the detection of antibodies against SIVs in fattening pigs could be an indicator of SIV infection in a farrow-to-finish farm. Maternal antibodies declined in fattening pigs aged 3 to 4 months [26, 30]. In addition, fattening pigs were replaced with neonatal pigs at each sampling in this study. Thus, fluctuations in the seropositive rate observed in fattening pigs indicated that SIV infection occurred prior to each sampling. On the other hand, neither the antibodies found in serum of weaned piglets nor those in sows could be used as an indicator of the recent SIV occurrence in a farm. Serological tests cannot distinguish maternal antibodies from those due to SIV infection. Sows are kept in a farm for more than a few years and antibodies against classical H1 SIVs in a pig are known to last up to more than 1 year after the primary infection . Thus, detection of the antibodies cannot indicate a recent infection of the sows. In farms such as Farms B and C where gilts are frequently introduced, it is not clear whether the seropositive sow was infected with SIVs before or after it was introduced into the farm.
The presence of seropositive fattening pigs in Farms C and E suggested SIV infection, however, no SIV was isolated. Sero-positive reaction was always observed against H1 SIVs in fattening pigs in Farm C, and an apparent increase in the number of fattening pigs seropositive against H1 SIVs was observed in January 2009 in Farm E. The reason why SIVs were not isolated in these farms may be explained by the fact that SIVs could circulate continuously in those farms with a prevalence rate lower than the detection limit rate in our sampling numbers. Sows were suggested to be a reservoir for continuous circulation of some respiratory pathogens (eg. Actinobacillus pleuropneumoniae, Porcine Circovirus type-2, SIVs) in a farm. Antibodies against those pathogens were detected at high rates in the sow population in a farm . Indeed, sows showed the highest seropositive rate in most samplings among the 4 groups examined in this study, suggesting the possibility that sows are repeatedly infected with SIVs during their 4 to 5 years of stay on a farm. Frequent introduction of pigs into a farm, such as Farm C, could also allow viral entry into the fattening pig population. In addition, movement of people/materials between farms could also be a possible route of entry of SIVs as was the case in Farm E where pigs were seldom introduced. Thus, further investigation is necessary to elucidate the mechanism of SIV persistence in farms.
Serological analysis suggested that SIVs belonging to unidentified sub-lineages within classical H1 and human-like H3 viruses likely exist in the Thai pig population. HI tests using various antigens revealed that antigenicity of the antigens within the subtypes in both H1 and H3 can vary. IDEXX ELISA often detected antibodies in serum samples that showed up negative in the HI tests. The antigens selected for use in the HI tests of this study represented SIVs circulating in Thailand according to our previous study . The ELISA test appeared to detect antibodies that could not be detected by the HI tests with the antigens used. This suggests that viruses possessing antigenicities different from those of the SIVs used in this study may be circulating in the Thai pig population.
H1N1pdmv infection in pigs in Thailand was detected in Farm D during our longitudinal monitoring. Direct human to pig transmission was suspected as was the case in pig farms in other countries [13, 32], because the affected farm had not introduced pigs since 2004. Based on the serological results, there is the possibility that H1N1pdmv was first introduced into sows before November 2009, and it then spread to fattening pigs and piglets within the farm. In Thailand, the number of confirmed human cases of H1N1pdmv infection increased from 8,800 to 28,300 from the end of June to October 2009 . Until July 3, 2009, the affected farm was shown to be free of H1N1pdmv by retrospective serological analysis. Remarkably, no significant clinical symptom was observed in the piglets carrying the virus at the time of swab collection, which is unlike other H1N1pdmv infections in pigs that have been reported along with respiratory symptoms [13, 32]. Therefore, the actual number of H1N1pdmv cases in the pig population may be much higher than that reported worldwide. According to the OIE weekly diseases information, up to March 2011, H1N1pdmv infection in pigs had been reported in 21 countries/districts . Emergence of reassortants with H1N1pdmv and other SIVs in pigs could be a threat to public health [35, 36]. In addition, transmission of H1N1pdmv from humans to pigs could have caused the amino acid changes in the HA, NA, M and NP genes, suggesting the possibility of a significant impact on viral evolution . Thus, to minimize the risk of H1N1pdmv infection in pigs, extensive bio-security protocols for farms need to be considered.
Many researchers have pointed out the importance of monitoring SIVs because pigs have the potential role as a mixing vessel for influenza viruses [1, 4]. To date, highly pathogenic H5N1 avian influenza viruses have been isolated sporadically in China  and Indonesia . H9N2 viruses that infect not only poultry but also humans [40, 41] were isolated from pigs from 1998 to 2007 in China. Under such circumstances, it is important that knowledge on the occurrence of SIVs in farms be deepened. The information obtained in this study could be useful to develop a strategy for SIV surveillance not only in Thailand but also in other countries, since the farrow-to-finish production system is commonly conducted worldwide. Crucial factors that determine the persistence and infection of SIVs in farms remain unclear. Further studies on SIVs in farms are needed in order to prevent economical losses caused by these viruses, and to prevent the emergence of novel viruses with the potential to cause pandemics in humans.