Oral fluids - PRRS diagnostics

Oral fluids were first reported as a diagnostic specimen for PRRSV in 1997. This review is limited to PRRSV, but it is noteworthy that oral fluids samples have diagnostic value for many diseases. A review by Prickett et al. noted that work has been conducted with oral fluids and classical swine fever, vesicular stomatitis virus, foot and mouth virus, Actinobacillus pleuropneumoniae, and PCV2. 
 

Collecting oral fluids

Quick guide to fluid sample collection in pigs

More info

Detection of PRRSv-specific antibodies in serum is the most common method for diagnosis of PRRS, although oral fluid samples or muscle transudates can also be used.  Indeed, in the US, oral fluid analysis has increased more than 10-fold since 2010 (Figure 1).

Figure1. The number of oral fluid tests at Iowa State University, IA, US 

Oral fluids are a mixture of saliva and mucosal transudate. Saliva is mostly water containing enzymes that begin food digestion. The transudate from capillaries in the oral mucosa and gingiva mirrors the presence and concentration of serum antibodies, hormones, drugs, and viruses. Passive transudation of serum components into oral fluids is extremely rapid resulting in a strong correlation between time to appearance of virus and antibodies in serum and oral fluids.



Not all specimens collected from the oral cavity are equal. The composition is influenced by the method and site of collection. This review defines oral fluids as the fluid in the oral cavity collected by use of an absorptive device. In pigs this is most easily achieved by providing a cotton rope for them to chew on. This approach is simple, rapid, inexpensive and non-invasive. In this review buccal and tonsilar swabs are not considered oral fluids.

All classes of pigs will readily donate oral fluids via ropes. Procedures have been refined to reflect the physical, behavioral, and husbandry differences between age groups  Ropes are usually hung for 20-30 minutes, which is long enough to ensure many pigs interact with the rope and not so long that the oral fluids evaporate. In general, the rope and pig size are correlated with a recommended rope diameter of 1.3 cm for nursery pigs and 1.6 cm for grow-finish pigs and boars. Ropes should be hung at shoulder height. By forcing pigs to reach up slightly, sample contamination is minimized. Unpublished observations by Dr. Jeffery Zimmerman note that ropes can be used to collect oral fluids samples from suckling piglets but the rope should be hung where the sow can touch the rope to increase piglet interest and willingness to chew. The number of pigs that interact with the rope affects the sensitivity and probability of virus detection in a low prevalence population. A flavoring solution (unsweetened apple juice with sucrose) was used, which did not appear to affect sensitivity but was not formally examined.

Boar stud sampling protocols should consider the risk of never sampling boars that fail to participate; traditional collection methods may still be required on these animals. Kittawornrat et al also determined that infection with a mild strain of PRRSV did not affect the success rate of oral fluid collection but did lower the amount of fluid collected. An average of 17.6 ml (range 4 to 38 ml) was collected pre-infection and an average of 15.7 ml (range 1 to 37 ml) post infection. Given that an optimal amount of oral fluid is 4 ml, and all collections reportedly produced sufficient oral fluids for PRRSV qRT-PCR testing, it appears that mild PRRSV infections have little likelihood of resulting in insufficient sample to test. Presumably severe infections would be identified by clinical signs rather than routine monitoring.

Collection ropes should be cotton because it is highly absorbent. Ropes reportedly must not be bleached. It is unclear how failing to follow these recommendations would affect diagnostic test performance but ropes are readily available to producers and it is possible that changes to the type of rope could cause false negative test results without raising suspicion. To avoid this problem, clear communication between producers, veterinarians and veterinary diagnostic laboratories will be required. Oral fluids samples with visual particulates should be centrifuged prior to submission and testing. Diagnosticians have commented that oral fluid samples appear highly variable; some being very clear and others flocculent. This may be exacerbated by inconsistent centrifuging of samples. The effect of sample quality or consistency on sensitivity has not been reported. This information would be valuable for estimating reliability. Further research on the effect of rope type, rope treatment, and centrifuging on test results would be of value.

Once the rope is removed from the pen, it is wrung into a plastic bag and the fluid drained into a tube (Falcon 2054 or equivalent). Without careful handling, these steps could cause cross contamination of specimens. This is not a concern for herd level diagnosis but could falsely elevate the within herd prevalence estimate. Fluids that will be analyzed within 24 hours should be chilled, while those with delayed testing should be frozen. Over a 12 day observation period, the highest virus concentrations and SP ratios were found in samples held below 10 ºC, an intermediate result at +20 ºC and the poorest results at +30 ºC. Between -20 ºC and +10 ºC no significant difference were found.

PRRSV Direct Tests
The viral load in oral fluids is correlated with, but lower than, in serum. Lower viral concentrations were found in oral fluids at days 0, similar levels at 14, and higher levels in oral fluids at 21 dpi compared to serum. In the field, additional variation may be induced by partial immunity, age, multiple viral strains and variable collection times post infection. From one perspective it could be argued that more research is required because failing to account for this variation might overinflate sensitivity estimates. On the other hand, natural variation has always been present in our diagnostics and we have not previously delayed access to diagnostic tests until all field variation is explained. Oral fluids have been promoted as a screening tool to identify early PRRSV infection. This means sensitivity is affected by both the lower concentration of virus in oral fluids and potential dilution of a few positive animals in a sampling pool of 20 or more pigs.

The likelihood of early PRRSV detection should be confirmed through observational trials because the findings of Wang et al. were not fully corroborated in an on-farm trial. Three herds were studied by collecting 1 oral fluid and 5 blood samples from 6 pens at 3 week intervals. In two of three herds, PRRSV was detected in the majority of pens by serum and oral fluids on the same sampling visit. But, in one of the 3 herds the first visit with PRRSV detected was characterized by all 6 pens positive on serum qRT-PCR and all 6 negative on oral fluids. Time to detection of PRRSV in oral fluids appears to be similar in oral fluids and serum. Virus was detectable in 10% of boars within 1 dpi, 76% within 2 dpi and 94% within 3 dpi which was comparable to the 100% positive serum samples at 3 dpi (n = 70). Interestingly, significantly higher proportions of oral fluids samples remained positive at 14 and 21 days than serum samples suggesting that the window for detection may be slightly longer for oral fluids than serum (average over trials: Oral fluids 94% versus serum 80% at 14 days and oral fluids 90% vs. serum 79% at 21 days). As noted under filter discs, this finding may be dependent on how the sample is tested as oral fluids on filter discs were unreliable by 19 dpi.

PRRSV Indirect Tests
Evidence supporting ELISA testing for PRRSV antibodies in oral fluids is limited and preliminary. Prickett et al reported no significant difference in ELISA values between oral fluid samples collected from experimentally inoculated pigs and controls. More recently three conference proceedings involving some of the same researchers have indicated ELISA tests can be used. One noted that the IDEXX 2XR Ab Test can be adapted to oral fluids to achieve a diagnostic sensitivity of ~80% and diagnostic specificity of ~100%. A second reported that PRRS-specific IgG, IgM, and IgA isotopes in oral fluids can be identified. The third described samples tested at the Iowa State University Veterinary Diagnostic Laboratory. Samples were collected from two “high-prevalence” and two “lowprevalence” herds, where prevalence was undefined. In each herd, 3 age-groups had 6 pens sampled each. In the high-prevalence herds where the ELISA was reported, a diagnosis based on 5 serum samples per pen or one oral fluid sample were correlated between 50 and 100%. The average serum SP, average rope SP, and pen correlation all increased with age from 6 to 8 to 10 weeks. Few positives were reported in the low prevalence sites.