Laboratory diagnosis of porcine infertility in the UK
By C. Bidwell and S. Williamson, VLA, Bury St Edmunds and S. H. Done VLA, Thirsk.Summary

Investigation of any of the so-called infectious causes of reproductive failure calls for a complete history of the problem. Detailed records must be compiled and should include all clinical signs, their chronological sequence, together with their magnitude. A proper combination of appropriate diagnostic tests must be undertaken, requiring the submission of the right material to the laboratory. Help with the correct interpretation of such results is often necessary. All these factors are looked at in detail in this paper, together with useful advice from the authors to the practising veterinarian.
Introduction
If Veterinary Investigation Diagnosis Analysis (VIDA) data is examined, it is evident that a low number of diagnoses of pig reproductive disease are shown each year. This is shown, for 2003, in Fig. 1 and compared with cattle and sheep. The reasons include:
Laboratory investigations are geared to diagnosing infectious disease. For pigs, as in cattle, where diagnosis is only established in about 20% of submissions for reproductive disease, non-infectious disease and management problems are prominent amongst the causes of infertility. These can only be assessed by on-farm investigations, with laboratory submissions mainly used to rule out infectious disease. Diagnosis of some diseases is problematic. Even where infectious disease is involved, some organisms are very fragile (e.g. leptospires) and tests for them may lack sensitivity, while some tests are expensive.
- Abortions/stillbirths may not be a main feature of the infertility problem. Where returning or not-in-pig sows are the main clinical presentation, the only samples available are bloods and genital tracts. These can be useful, together with other information, in supporting a suspected diagnosis; but are less likely to allow laboratory confirmation of an infectious disease.
- Even where abortions do occur, aborted material may not be found, especially in outdoor herds.

Fig. 1 – Diagnosis reached, and not reached, for samples submitted to investigate reproductive disease in 2003
Although the same low proportion of diagnoses are made for bovine reproductive disease as for pigs, with more submissions being received at VLA and SAC laboratories, it is still possible to examine epidemiological factors and look at trends over time for particular diseases. For pigs, too few submissions are received to make a realistic epidemiological analysis. In contrast, a diagnosis is made in over 50% of submissions for ovine reproductive disease, as this more regularly presents as abortion due to pathogens which are also readily identified.
Diagnostic Serology for Reproductive Disease – General Guidelines
Some general principles apply whatever the test being used. It is important that affected sows are bled as soon as possible after return to oestrus, abortion or being found not-in-pig. Paired serology (acute and convalescent sera) is indicated if acute disease is apparent in sows, for example, pyrexia, respiratory signs, or skin lesions, as in erysipelas or swine influenza. Paired serology can also be useful where there is no apparent disease in sows, but there is an ongoing problem with high morbidity, either ‘before’ and ‘after’ samples are collected. As sows are often already seropositive by the time reproductive disease manifests itself, groups of unaffected sows should be bled at a convienient time, e.g. weaning and sera stored. Any sows later affected with reproductive disease are re-bled for paired serology to establish if there is any connection between seroconversion and disease. This is more informative than single serology which only indicates exposure. For some tests, e.g. leptospira MAT, high antibody titres suggest recent infection and for others, e.g. parvovirus, negative results rule out involvement. Group serology is also sometimes used as a partial substitute for paired serology or to give a ‘snapshot’ of the serological status of sows/gilts at different stages of production and integration. This is used more for monitoring than diagnostic purposes.
Serology – the numbers game
The number of sows to bleed when investigating disease depends on many factors. To give an idea of how confident one can be in interpreting serological results, two examples are given below. These do not represent a protocol - logistic, welfare and time issues may prevent large numbers of sows being bled, especially outdoors. The examples serve to illustrate that, when limited numbers of sows are bled, one must be cautious about reading too much into the results.
Example 1 – Aim is to detect the presence of disease X:
Here, the aim is to detect whether or not there is any evidence of exposure to disease X, so finding just one positive is sufficient. In a 400 sow herd with overall 10% prevalence of disease X, to be 95% confident of detecting at least one positive sow, one needs to randomly sample 28 sows. In reality, one selects a biased group; clinically affected sows. If 20 sows have returned and it is estimated that half of these actually returned due to disease X (i.e. 50% prevalence of disease X) 5 sows need to be sampled to be 95% confident of detecting at least one of these as positive for disease X. Here, the low sample numbers mean that the level of disease cannot be assessed with confidence and groups cannot be compared.
Example 2 – Aim is to estimate prevalence of disease X:
Here, the aim is to measure how much disease X is present. In the same scenario with a 400 sow herd with overall 10% prevalence of disease X, to be 95% confident of detecting this prevalence (accepting +/- 10% error), 32 sows should be sampled. Again, in reality, clinically affected sows are probably sampled and if 20 sows returned, half of which were due to disease X, 17 of the 20 sows need to be sampled to be 95% confident of detecting this prevalence (+/- 10% error).
Diagnosis of bacterial causes of porcine reproductive disease
This section covers the diagnosis of leptospirosis (mainly due to L. bratislava), erysipelas, brucellosis and other incidental bacterial infections, including chlamydial infection in pigs.
L. bratislava infection
The importance of L. bratislava as a cause of reproductive disease in pigs remains uncertain. Past papers in The Pig Journal have described leptospirosis in pigs in more detail (Ellis, 1989; Williamson et al, 2004). There is equivocal evidence from experimental infections, however, that the organism has been identified in aborted pig foetuses and in aborting and infertile sows (Ellis et al, 1986a; Ellis et al, 1986b; Power, 1991). There is also serological evidence of sow exposure in some herds with reproductive disease, in which disease has resolved following antibiotic treatment, or more specifically, vaccination against L. bratislava.
Overall, evidence from elsewhere indicates that L. bratislava can cause reproductive disease and circumstantial evidence from England and Wales indicates that it may be involved. As for other leptospires, sporadic abortion storms, with stillbirths and weak and dying neonatal piglets, certainly can be caused. Barlow (2004) described such an outbreak due to L. mozdoc. At present, L. pomona, L. tarassovi and L. grippotyphosa, which can cause serious reproductive disease, are not present in the UK pig population.
The diagnostic tests currently available for diagnosing leptospirosis are:
Foetuses/piglets: (detection of organism)
- PCR - not species specific
- FAT - Northern Ireland, not VLA
- Culture - slow and technically demanding
The best material for diagnosis is aborted/stillborn/weak piglets and placentas with sera from aborting sows. Whole litters should be submitted, more than one litter if possible. Prompt submission is important to minimise autolysis - leptospires are fragile organisms and histopathology may also need to be undertaken.
Information yielded by leptospira serology
- The MAT mainly detects IgM antibodies and titres rise rapidly after infection. Titres of 1/800 or more are suggestive of recent infection.
- Paired serology helps interpretation. If sera are collected at the time of abortion and two weeks later, it may be possible to demonstrate seroconversion or rising titres. However, often by the time disease manifests, sows are already seropositive.
- High titres (1/800+) to other serovars are suspicious.
- Antibodies to exotic serovars should not be detected (except as low titre cross-reactions to endemic serovars).
- Serological results assist management and treatment decisions, e.g. whether to medicate, vaccinate.
The ideal sample for diagnosing abortions and stillbirths due to erysipelas are foetuses and placentas submitted fresh for culture and isolation of the organism.
If no aborted material is available and disease is manifesting in sows, paired maternal serology can also confirm the diagnosis, collecting acute and convalescent sera about 2 weeks apart. Single serology rules out erysipelas if sera from recovered sows are negative for antibody; but positive single serology only indicates exposure to the organism at some time in the past. This could represent field infection alone, vaccination alone or vaccination and challenge.
Brucella suis infection
Brucella suis infection is not currently present in the UK. As well as import/export testing, DEFRA funds ongoing surveillance for infection on samples submitted to VLA Regional Laboratories. This involves microscopy (MZN-stained smears) and selective cultures for brucella organisms on aborted and stillborn material, and cultures on male and female reproductive tracts and suspected infectious arthritis lesions in breeding stock. Surveillance by serology is problematic in pigs due to the high number of serological cross-reactions due to exposure to Yersinia species.
Chlamydia species infection
Reports of involvement in reproductive disease – Switzerland, Germany
Detect by microscopy and PCR, not serology
N.B. Both implicated in ORCHITIS in boars
Other bacteria
Abortions due to infection with Pasteurella species, Klebsiella species and various streptococci are regularly diagnosed by culture of foetal stomach contents and viscera. Although these feature quite often in VIDA diagnoses, this is because they are readily diagnosed compared to some other causes, e.g. PRRSV infection and, overall, they are of minor importance. In some cases, they may occur secondary to infection in the sow. In others, they may be opportunistic infections by commensal organisms in the sow’s vagina, establishing infection in the foetuses after the cervix has opened and the pregnancy failed for some other reason.
Infertility: | Vaginal discharges – timing, duration, amount, effect. Urogenital tracts and kidneys, not vaginal swabs. Collect tract when observe discharge. Secondary to pre-disposing factors – hygiene, housing age, early weaning N.B. Actinobaculum suis |
Viral causes of porcine reproductive disease
‘SMEDI’ is a syndrome of stillbirths, mummification, embryonic deaths and infertility. The syndrome does not include abortions. It is helpful to establish whether the material under investigation includes abortions and whether foetal death has occurred pre-, peri- or post-partum. Fig. 2 gives an algorithm to assist with these decisions. There will be variation especially in the degree of autolysis, which can make if difficult to determine if deaths are peri-or post-partum. Further details and pictures of specific lesions can be found in Barlow (1998).

Fig. 2 - Investigation of stillborn and mummified foetuses
A number of infectious agents result in ‘SMEDI’ and the following list is not exhaustive. It is of note that abortion is not usually a feature of parvovirus or enterovirus infection:
- Swine fevers + abortion
- Aujeszky’s Disease + abortion
- Porcine enterovirus
- Swine influenza + abortion
- Porcine parvovirus
- PRRS + abortion
- The Swine Fevers (ASF and CSF) are indistinguishable in the field. Both may produce all the signs of the haemorrhagic diatheses; but in the chronic forms just a few teratogenic or reproductive effects such as trembling pigs may be seen. With the acute forms with high temperature there may be abortion, mummification and stillbirths. A variety of antigen techniques are available (PCR, RT-PCR, IHC, ISH,) and also antigen-capture ELISA and FAB for antigen. There are also a wide variety of serological tests for recovered animals.
- Aujeszky’s Disease Virus. This may cause abortion 10-20 days after clinical illness at any time during gestation; but especially during the first 2 months of pregnancy. Sows may show pyrexia, anorexia and depression. Histology with inclusions and IHC are most useful. A variety of ELISA tests are available.
Full investigation of reproductive disease is expensive (up to £1250/litter; but the charge to the practice may be as low as £40/litter). Therefore, the pathologist needs a rationale and justification for determining the tests undertaken. This will include a detailed history and full investigation is not justified if only one pig is affected. The history should include the group size and the number of pigs affected, herd disease status, dam parity, whether disease is acute onset or insidious, the disease presentation including whether dams are malaised and any vaccinal history.
Porcine parvovirus (PPV)
The only clinical response to infection is maternal reproductive failure. The virus causes damage to a variety of foetal organs and the placenta, i.e. reproductive failure, is a direct effect of the virus on the conceptus.
The consequences of the infection vary with the stage of gestation. By about 70 days of gestation, the foetus can mount its own protective immune response.
A method of gauging the approximate gestational age (days) is that it equals 21+ (3 x crown-rump length in cm). At 70 days, the crown-rump length is approximately 16cm and the foetal weight 190g (Christensen, 1994).
The consequences of PPV infection are shown in Table 1.
Table 1 - Results and consequences of infection with PPV (Mengeling, 1999)
Stage of gestation (days) assuming transplacental infection 10-14 days after maternal infection. | Consequences of infection | |
Infection of dam
|
Infection of conceptus
|
|
≤56
|
10 – 30
|
Death, resorption, small litters
|
30 – 70
|
Death, mummification
|
|
>56
|
70 – term
|
Immune response and survival
|
Intra-uterine spread of virus can result in a range of findings in the same litter.
The clinical signs of porcine parvovirus infection
Often a sporadic problem because of varying herd immunity.
Signs include:
- often only affecting gilts not sows
- returns to oestrus
- failure to farrow despite anoestrus (uteri can be checked at abattoir for foetal remains although total absorption is possible)
- small litters
- mummified foetuses (often of variable size)
- few or no abortions
- lack of maternal illness
Maternal
Serology (ELISA) for detection of antibody
If dam is sero-negative rules out PPV
If dam is sero-positive may be detecting vaccinal titres, previous field exposure, vaccinal and field challenge (maternally derived antibody usually cannot be detected after six months).
Paired samples
Of no use at the time of the incident as viral insult is usually more than one month prior to reproductive disease. They can be useful if sero-conversion coincides with reproductive failure.
Herd sero-profiling of gilts through to older sows
This may detect suspicious gaps in immunity; but requires further back-up testing to reach a diagnosis.
Foetus
Detection of antigen - PCR or Antigen ELISA
(Sample foetal liver, lung and foetal fluid, with tissues from mummies preferred).
Detection of antibody - HIT or ELISA*
(Sample foetal fluid)
The detection of antibody depends on the ability of the foetus to mount an immune response.
*The VLA is currently validating the antibody ELISA for foetal fluid.
Fig. 3 below shows a comparison of the HIT titres with the ELISA titres for sera.

Fig. 3 - PPV serology (Banks, 2005)
Interpretation of the PPV ELISA test
These viruses, which have also been referred to as ‘SMEDI viruses,’ are excreted in pig faeces and therefore widespread in the environment. By breeding age, most pigs are immune and therefore reproductive disease, which presents as for parvovirus, is presumed uncommon. Standard VLA investigation of porcine infertility does not include tests for enterovirus. Virus isolation can be undertaken on foetal lung, intestine and brain (preferably fairly fresh) but the virus is difficult to propagate.

Fig. 4 – PPV serology; comparison of hit titres with ELISA blocking%
PCV2 AND PRRS
What is the evidence that PCV2 is associated with reproductive failure in sows? The evidence is collected in Table 2. The laboratory tests are shown in Table 3.
Table 2 – Evidence that PCV2 is associated with reproductive failure in pigs
- PCV2 WAS ASSOCIATED WITH ABORTIONS IN WESTERN CANADA WHEN PCV2 FIRST DESCRIBED (West
PCV2 WAS ANTIGEN RECORDED IN ABORTED FOETUSES (Ellis et al, 1998).
PCV2 ISOLATED FROM ABORTED FOETAL TISSUES (most of the herds were PRRS negative) (Harding et al, 1998).
IN WESTERN CANADA, 60% HAD PRRS (Sorden, 1998) AND 20% PPV (Ellis et al, 2000).
MANY ABORTED FOETUSES HAD MYOCARDIAL DILATATION WITH MYOCARDITIS AND ABUNDANT PCV2 ANTIGEN IN CARDIAC MYOCYTES.
GILTS DELIVERED MUMMIFIED FOETUSES. THORACIC FLUID PCR POSITIVE AND AFFECTED GILTS SHOWED RISING TITRES TO PCV2. Table 3 - Diagnostic tests available for the diagnosis of PCV2
- ANTIBODY
Not particularly useful as nearly all pigs have responded to the ubiquitous infection.
ANTIGEN DETECTION
Immunohistochemistry is most useful as it can be combined with histology to report on pathology (Allan
Antigen detection using the PCR is very successful. It may detect some very small amounts of antigen which have been incorporated into the host’s genome and which do not reflect the presence of PCV2-associated disease (Tischer et al, 1986). Samples to take:
Foetal tissues
RECENT COMMENTS ON PCV2
Reproductive disease associated with PCV2 may or may not be a significant problem. It may be principally a function of the immune status of the host at the time of the PCV2 exposure. It is yet unknown whether it causes congenital tremor or not. Stevenson says it does (Stevenson et al, 2001), but others say there is no evidence (Ha et al, 2005; Kennedy et al, 2003).
Oronasal infection of gilts with PCV2 leads to long-lasting viraemia, which is often cell-associated, despite the onset of the immune response. The zona pellucida is a barrier to the entry of the PCV2; but, based on the size of the channels in the zona pellucida, there is the possibility that the virus can gain entry through these.
The virus replicates in zona pellucida-free embryos, e.g. the morula and the blastocysts. The susceptibility increases with development. In foetuses, the extent of the virus replication and gross pathology and clinical outcome of infection depend on the age at infection. When the infection occurs in the middle third of gestation (57-75 days), extensive virus replication takes place in different organs, leading to foetal death (mummified and stillborn) or weak-born piglets at term. The rapid increase in the cardiomyocytes mid-term to birth appears to be the favourite replication site. Death due to high virus replication does not necessarily interrupt gestation. After the middle third of pregnancy, the foetus appears to be resistant to infection. Greater than 75 days of age, the immuno-competent foetus usually produces a virus positive and antibody positive little animal, which is not grossly affected. Foetuses infected in utero can carry the virus post-natally in the absence of disease. In late gestation, the lymphoid cells are rapidly increasing and infected lymphoid organs can occur without subsequent wasting disease or lymphoid depletion.
PRRS
What is the evidence that reproductive failure occurs with PRRS? This is seen in Table 4. Table 5 shows the samples that need to be taken for diagnosis.
Table 4 – Reasons for believing PRRS is a cause of reproductive failure
- LESIONS IN UTERI AND PLACENTAE HAVE BEEN REPORTED IN GERMANY.
LESIONS IN UMBILICAL CORDS HAVE BEEN REPORTED IN THE USA.
EXPERIMENTAL PRODUCTION OF ABORTION AT AROUND 109 DAYS OF AGE.
VIRUSES ISOLATED FROM ABORTING SOWS, FOETUSES, AND PAIRED SERUM SAMPLES SHOW RISING TITRES.
CLINICALLY, ROLLING INAPPETANCE AND EXCESSIVE RETURNS.
VIRUS NOT FOUND IN OVARY DIRECTLY, BUT GETS CARRIED IN BY MACROPHAGES.
EARLY INFECTION CAUSES FOETAL DEATH.
NO PATHOGNONOMIC GROSS LESIONS.
NO PATHOGNONOMIC HISTOLOGY LESIONS.
IT IS ENDEMIC IN MANY HERDS.
NEEDS TO BE CONFIRMED BY ANTIGEN DETECTION.
LESIONS FOUND IN LIVEBORN, CONGENITALLY–INFECTED PIGS MAY BE SUGGESTIVE OF MYOCARDITIS, INTERSTITIAL PNEUMONIA AND ENLARGED LYMPH NODES.
FOETAL TISSUES | ANTIGEN DETECTION, PCR, IHC VIRUS ISOLATION ANTIBODY IN BLOOD ANTIBODY IN BODY FLUIDS, e.g.THORACIC |
FOETAL NEONATAL GROWER ADULT |
FROZEN TISSUES FOR FLUORESCENT ANTIBODIES AND IHC FRESH TISSUES AT P.M; - LUNG, LIVER, LYMPH NODE, KIDNEY, SPLEEN, THYMUS, TONSIL FOR VIRUS ISOLATION, ISH, IHC, PCR |
SEMEN | PCR IS THE BEST FOR DETECTION IN SEMEN (DUE TO ANTI-VIRAL PROPERTIES OF SEMEN). (Christopher-Hennings et al, 1995). |
VIRUS ISOLATION | (Lelystad virus still grows best in alveolar macrophages, but it also grows in cell lines). All viruses do not grow equally in all systems. |
ANTIGEN DETECTION | PCR-BASED. Usually ORF6, or 7, but sometimes 5. Gives an early diagnosis. May not indicate a replicating virus, but does usually. RT-PCR is specific and sensitive (Wagstrom, 2000). Nested-PCR; the RNA is extracted from the sample. RT converts RNA to DNA and this is amplified by the PCR. It takes 1-2 days, has high specificity, high sensitivity and rapid turn-around. May not detect field virus if wide discrepancies between field virus and primers used. May detect bits of virus as well as replicating virus (Egli et al, 2001). Restriction length fragment polymorphism (RFLP) can then be used to cut PCR products (3 enzymes). Can track introduction of new viruses. Gives no indication of virus virulence. Cannot be used to make a vaccine selection. (Wesley et al, 1998). |
VIRUS SEQUENCE ANALYSIS | This may show additions, deletions and mutations. If there is more than 1% difference to reference strain, then the viruses are not related. It does not help in assessing virulence or biological properties. Dendograms show ancestral relationships of random mutations, but not recombinants. (Andreyev et al, 1997; Murtaugh et al, 2001). |
SEROLOGY INTERPRETAION
The testing methods used for PRRS are shown in Table 6.
A single sample is no good. Because PRRS is so common - it means absolutely nothing. The ELISA can indicate if a herd is positive or negative. However, the wide antigenic variation complicates the tests. If there is a wide variation, then the field virus may not show up as it is far removed from the original test viruses e.g. Lelystad and USA strain 2332 (Christopher-Hennings et al, 2002; Yoon et al, 1995).
It is impossible to predict shedding or carrier state on the basis of serological assays.
The tests cannot tell the vaccine strains from the field virus.
The PRRS virus is unique in that it divides the immunological responses so that both Th-1 and Th-2 cells are stimulated, i.e. both the cellular and humoral immunities are stimulated. This is detrimental in itself because stimulation of one has an inhibitory effect on the other.
When monitoring negative herds, use both PCR and serology. Positive serology needs to show a 4-fold increase to be positive.
Negative PRRS serology could be due to a) not infected, b) infected, but not seroconverted, c) pigs having been infected but have become seronegative, d) having a non-clearance of a very low sensitivity virus to the techniques being used.
Virus neutralisation is not used very much because the AB take so long to develop (>21 days), possibly at low titres and probably not in all pigs. It is also technically difficult, time-consuming and is used more as a research tool (Nelson et al, 1994). Western immunoblotting is also mainly a research tool in which proteins are separated by electrophoresis and then transferred to membranes. Antibody binding to specific proteins is then visualised by conjugated enzymes and substrates. It is expensive and time-consuming.
For the tests used in the UK, the ELISA is 100% sensitive and has a specificity of around 95%. Results are produced as an S:P ratio and 0.4 is considered positive.
The IFA test is of good specificity, but of variable sensitivity. The magnitude of the titre can be measured and it will reliably detect AB for 3 months post-infection.
PRRS: A FEW THINGS TO REMEMBER
Piglets in uteri are not susceptible to PRRS before placentation at 3 weeks. The foetal immune response occurs from around day 48 and in all pigs, by 70 days. Boars with PRRS infection will transmit virus; but it is not known when. PCR techniques will detect virus antigen in semen; but not necessarily replicating virus. Boar semen itself is toxic to viruses and needs to be fractionated after collection, if it needs to be examined for PRRS.
Remember the cardinal sign of PRRS when it first occurred in the early 1990s - it was the sheer variability of the signs. Even with the mild strains of virus in Europe, there may be no clinical signs on the one hand, or reproductive failure, respiratory signs, subcutaneous oedema, late mortality in finishing pigs, failure to grow or even just conjunctivitis, on the other.
In the reproductive area, there may be early farrowings, with pigs brown or macerated and looking as though they had been dead for 5-7 days, mummification (may rise by 20-30%), stillbirth (may rise by 50%), weak pigs born, decrease in farrowing rate by 20-40% over a 2-3 month period, with rise in death from 3-5%.
Repeat breeders may also increase and the rate of not-in-pig sows.
In chronic cases of PRRS, there may be increase in stillbirths, pre-weaning mortality, higher post-weaning mortality and irregular returns to oestrus.
Control of PRRS requires correct sampling techniques and sample handling for sufficient members of the population, with an intensive diagnostic assessment. Knowledge of the tests and their significance and their limitations is essential. A reliable set of results on which to base decisions is thus necessary.
Summary: making sense of it all
- To investigate any of these so-called infectious causes of reproductive failure, a complete history is required. If this suggests notifiable disease, telephone the DVM.
- Proper records are essential.
- Clinical signs, their order of occurrence and magnitude involves examination, not just inspection.
- A combination of appropriate diagnostic tests is needed. Submit all aborted and stillborn piglets and placenta, if possible, to the laboratory. Submit more than one litter, if possible, and sera from the aborting sows.
- In all probability, help will be needed with the interpretation of these tests or guidance with their appropriate use and limitations.
- In many instances, the art and skill of veterinary science is required to sort out these problems, i.e. experience and common sense, but mostly thought and time.
Acknowledgments
Thanks are due to Cornelia and Susanna for involvement in the production of this paper. Alex Barlow, Malcolm Banks and Charlotte Featherstone have also been most helpful, together with Mrs Maureen Sunderland. VLA-library provided support in the furnishing of useful documents.
References
Allan, G.M. et al (1996).Veterinary Immunology and Immunopathology, 43, 357-371.Pig Journal - February 2006