Discussion
The concept of FMD as a fulminant, outbreak-driven animal emergency is appropriate for considering the disease within FMD-free countries. Useful as it is, the concept has subsequently been “imported” into countries where FMDV is widespread and endemic, and where virus presence is known to exist without ongoing clinical outbreaks (Farooq et al., 2018; de Carvalho Ferreira et al. 2017). Although outbreaks occur frequently in many endemic regions, they are not sufficient to explain the high seroprevalence in many regions where no recent clinical outbreak has been reported (van Andel, 2020b).
We found that 2.3% of healthy, commercially-slaughtered large ruminants from an endemic region had detectable FMDV RNA within the oral cavity, and that 1.5% had FMD RNA in the nasal cavity. The results were similar for both Laos and Myanmar, suggesting this might be a pattern which is common throughout the region. This is the first study to sample the oral and nasal mucosal surface of a healthy population of large ruminants within an endemic region, and adds to a small but growing literature examining the state of FMDV infection in animals from endemic regions. By comparison, in naïve cattle, oral and nasal shedding following experimental infection sharply declines over 21 days (Parthiban et al. 2015) with extinction of virus in the oral or nasal mucosa regardless of whether animals clear infection or become chronic carriers (Stenfeldt et al., 2016). By contrast, the presence of FMDV RNA within the dorsal nasopharynx, has been well-established by multiple other studies (Stenfeldt et al., 2016; Barend et al 2016; among others) which have defined the role of that region in the chronic carrier state. Although FMDV detection in the nasopharynx is typically demonstrated by collection of oropharyngeal fluid from live animals with a probang cup, the current study demonstrates the utility of detection by swabbing the nasopharynx in the post-mortem context.
The animals we sampled were chosen by convenience, and nothing was known about their previous exposure to FMDV except that they originated from an FMD-endemic region. None of the animals in this study had visible scars, lesions or other clinical signs (e.g. salivation, lameness) suggestive of current or recent FMD. We used NSP ELISA positivity as an imperfect proxy for prior exposure, as a way of judging whether any viral excretion might be related to pre-clinical acute FMD infection. Vaccination with whole vaccines (which could cause NSP-positive reactors) is thought to be low in both Laos and Myanmar, and was not considered likely to influence NSP ELISA results. Animals with detectable mucosal viral RNA at any site were significantly more likely to be seropositive than the sampled population, but not all RNA-positive animals had antibodies. Since development of antibodies postdates initial infection and early replication, we hypothesized that animals with detectable FMDV RNA but no antibodies might have been in the early (neoteric) stages of infection (Stenfeldt and Arzt, 2020). Six seropositive animals had detectable viral RNA in the oral and nasal cavity. Since NSP serology is not serotype specific, one possible reason for shedding is that these animals did not have protective antibodies to the same virus that was detected by RT-PCR.
One benefit of slaughterhouses is that they aggregate animals and present a single site for sampling many animals with minimal handling complications compared to field sampling. Others have used slaughterhouse sampling as a way of accessing FMDV samples, including in Pakistan and Kenya: Navid et al (2019) found that 11% of buffalo tested at Pakistani slaughterhouses carried detectable FMDV antigen in their probang sample. Anderson et al. (1974) found that over 3% of Kenyan cattle at slaughter had viable virus in pharyngeal scrapings. However, animals sent for slaughter might not be indicative of a population at large, and more work remains to be done before it is clear whether our results can be extrapolated to the general large ruminant population of Laos and Myanmar. In addition, animals transported for slaughter are more likely to have undergone stress and mixing, with potential new infection or reactivation of viral activity. The potential role of stress in FMDV reactivation has been an area of interest, although experiments using immunomodulatory drugs have been unable to effect active shedding of virus in chronic carriers (reviewed more thoroughly in Stenfeldt and Arzt, 2020).
Outbreaks of active FMD in Myanmar and Laos have been attributed to incursions of new strains of FMDV (Bo et al, 2019; Khounsey et al. 2008). Our results suggest the interesting possibility that in addition to these virus “incursions”, the epidemiology of FMDV spread within endemic contexts could include subclinical viral infection and shedding within populations of apparently healthy animals. Dissecting the pathophysiology of infection in each animal in our study was not attempted, since too much was unknown about previous exposure and infection. Our intent was to evaluate a cross-section of healthy animals for a possible role in viral shedding, to inform a regional epidemiological understanding. Descriptions of viral location in chronic carrier cattle and buffalo focus heavily on nasopharyngeal virus detection (Stenfeldt et al., 2016; Moonen et al. 2004; de Carvalho Ferreira et al. 2017). Our results suggest that perhaps more attention should be paid to other mucosal surfaces, and that the relationship between FMDV and large ruminants in endemic settings may differ from that predicted experimentally. If small numbers of cattle and buffalo contain viable virus on environmentally-accessible mucosal surfaces such as the nose and mouth, there is the potential for low-level transmission of virus on a regular basis between animals in close contact. This might help explain the phenomena seen in FMDV-enzootic contexts, where seroprevalence rises without disease (Bertram et al, 2018), and where outbreak reporting imperfectly predicts seroprevalence (van Andel 2020b). The observation that healthy dairy cattle and buffalo shed virus in their milk (Nawaz, 2019; Ahmed 2017), raises the possibility that in endemic countries, some cattle might be exposed from birth to both virus and protective maternal antibody. Additionally, intensive sampling of buffalo in a dairy colony in Pakistan demonstrated that within one year numerous serotypes and strains could move through a herd without any clinical signs of infection (DOI: 10.1111/tbed.12963).
A limitation of this study is that we cannot be sure the presence of viral RNA is indicative of the presence of viable FMDV. In addition, although environmental control swabs were negative, the possibility of cross contamination within different anatomical sites or between animals cannot be discounted. Positive results however were widespread throughout the sampling period and locations, suggesting contamination was not a prominent cause of positivity. Dry swabs were employed for sampling rather than metal cuvettes used at slaughter by Anderson et al. (1974); this was to prevent possible contamination between animals. Swabs were also considered an affordable, useful surveillance tool that might enable use of the technique by agencies seeking to increase their PCP-FMD surveillance options. Because we chose not to obtain probang samples prior to slaughter in order to prevent possible contamination of sites, our swab-based dorsal nasopharyngeal results are not comparable with results from other studies. Future studies should make use of virus isolation to determine presence of viable virus in animals with detectable FMDV RNA, and should work to establish the relative sensitivity of the post-mortem swab-based technique with more established techniques such as the probang.
In summary, we found that 5.3% of healthy, asymptomatic cattle and buffalo sampled during slaughter in Laos and Myanmar had detectable FMDV RNA in their dorsal nasopharynx, and 3.4% had detectable FMDV RNA in their oral and/or nasal cavities. Although the viability of this viral RNA is unclear, our results suggest the possibility that FMDV present on the oral and nasal mucosa of asymptomatic large ruminants could play a role in the epidemiology of FMD in Southeast Asia. It is known that surveillance approaches focussing solely on clinical outbreaks of FMD may not be sufficient to detect all circulating virus in populations of large ruminants. Yet, field sampling of live animals is highly labor-intensive and often logistically prohibitive. The current study demonstrates the utility of detection of FMDV RNA by swabbing of the nasopharyngeal, oral and nasal mucosa in a post-mortem context, and may constitute a useful complementary tool for surveillance in endemic regions.