African swine fever (ASF) diagnosis, an essential tool in the epidemiological investigation

Introduction

African swine fever (ASF) is one of the most complex infectious swine diseases. Its notification to the World Organization for Animal Health (OIE) is mandatory due to the high mortality it causes, its efficient transmission rate and the great sanitary and socioeconomic impact that it produces on international trade of pigs and pork products. The African swine fever virus (ASFV) is a large, enveloped double-stranded DNA virus, which is the only member of the Asfarviridae family (Dixon et al., 2005).

Endemic in more than 20 sub-Saharan African countries (Mulumba-Mfumu et al., 2019) and in the island of Sardinia (Italy) since the last century (Cappai et al., 2018; Jurado et al., 2018a; Laddomada et al., 2019), ASF arrived at a Black Sea harbor in Georgia in 2007 (Rowlands et al., 2008), from where the disease spread quickly to other neighboring countries, reaching the European Union (EU) in 2014. The first cases of infected wild boar were reported in January and February 2014 from Lithuania and Poland, respectively, followed by Latvia in June and Estonia in September (Gallardo et al., 2014). In 2017 ASF spread to Romania and Czech Republic, although in the latter was declared as resolved in January 2019. The latest countries affected have been Bulgaria, Hungary and Belgium in 2018, and Serbia and Slovakia in summer 2019. In the EU scenario, the wild boar has been the most severely affected host, being responsible of more than 85% of the reported cases in all countries except in Romania (Cwynar et al., 2019; EFSA 2018a,b; EC 2019; Iglesias et al., 2017; Jurado et al., 2018b).

In August 2018, ASFV demonstrated its huge capacity for transboundary and transcontinental spread jumping to China, several hundreds of kilometers away from previously known infected regions. There, it rapidly spread with 165 ASF outbreaks confirmed in 35 provinces and the culling of more than 1 million pigs by 6 February 2020 (FAO 2020). The continuous spread of ASF to other Asian countries, with confirmed detections in Viet Nam, Mongolia, Cambodia, Democratic People’s Republic of Korea, Myanmar, The Philippines, Republic of Korea, Timor-Leste, and Indonesia, makes controlling the spread even harder.

Although ASF was first described almost a century ago, controlling the disease has proven to be a challenge, in particular, because no vaccine or treatment are available. Spread of ASF can only be prevented by early detection and the application of strict compliance of classical disease control methods, including surveillance, epidemiological investigation, tracing of pigs, stamping out in infected holdings, strict quarantine and biosecurity measures and animal movement control. Surveillance, to be successful, must have adequate laboratory support for a rapid diagnosis, which in combination with the information coming from disease epidemiology, scenario, and the clinical signs, will allow the early detection of the disease and therefore reduce/prevent ASFV spreading (EC 2013).

ASFV genetic diversity and clinical presentations in affected areas of Central-Eastern Europe and Asia

The molecular phylogeny of the virus is firstly investigated by sequencing the 3’end of the VP72 coding gene, which differentiates up to 24 distinct genotypes (Bastos et al., 2003; Quembo et al., 2018). All the ASFV strains circulating in Europe (except in Sardinia) and in Asia belong to the p72 genotype II (fig. 1) (Gallardo et al., 2014; Garigliany et al., 2019; Ge et al., 2018; Kim et al., 2020; Malogolovkin et al., 2012; Le et al., 2019; Rowlands et al., 2008). Despite different variants have been identified within genotype II strains throughout the analysis of small genome regions (Fraczyk et al., 2016a; Gallardo et al., 2014; Mazur-Panasiuk and Woźniakowski, 2019a; Nieto et al. 2016), the full genome sequence of 17 European genotype II viruses shows a low mutation rate and high genetic stability with a homology of more than 99.9% (Bao et al., 2019; Cano-Gómez et al., 2018; Forth et al., 2019; Mazur‑Panasiuk et al., 2019b). These results hinder the definition of reliable genetic markers associated to virulence. Therefore, the current approach to identify changes in virulence and pathogenesis mechanisms is still based on classical experimental infections and field observations.

Figure 1. Distribution of the ASFV genotypes.

Clinical signs associated with ASFV infection are highly variable depending on various factors: virus virulence, swine breed affected, route of exposure, infectious dose, and endemicity status in the area. According to their virulence, ASFVs are classified as highly, moderately or low virulent (Blome, et al., 2013; Beltrán-Alcrudo, et al., 2017; Sánchez-Vizcaino et al., 2015). Highly virulent strains are usually responsible for the peracute and acute forms that give rise to high mortality rates that may reach 100% within 4–9 days post-infection. In peracute ASF, affected animals can die suddenly 1–4 days after the onset of clinical signs with no evident lesions in organs. Pigs showing the acute forms of the disease display mainly a febrile syndrome with erythema and cyanosis of the skin. Internal lesions are mainly characterized by hyperaemic splenomegaly and haemorrhages in organs, particularly in the visceral lymph nodes, with fluids in body cavities and fibrin strands on organ surfaces. The distribution and frequency of these lesions are variable and most are seen in other swine diseases such as classical swine fever (CSF). Moderately virulent viruses lead to the appearance of acute, subacute and chronic forms. Pigs with subacute infection may have persistent or fluctuating temperature responses for up to 20 days, during which time some pigs stay in good condition, while others display the symptoms described above for the acute process form (but less severely) with mortality rates in the range 30–70%, usually after 20 dpi. In the chronic form of ASF, clinical signs and lesions are not specific and may persist for several months, giving rise to a range of illnesses, with symptoms such as skin ulcers and arthritis, stunted delayed growth, emaciation, lameness, pneumonia, and abortion, but with low mortality rates (Moulton and Coggins, 1968; Mebus and Dardiri 1980; Leitao, et al., 2001; Sánchez-Vizcaíno, et al., 2015).

From the published data, most of the genotype II isolates of the “Georgia 2007 type” that are currently circulating in Eastern and Central Europe and, now in Asia, are highly virulent and cause very high mortality rates of 91–100%. After intramuscular inoculation, the animals, regardless of the host, became infected after an average of 4.4±1.2 days and did not survive for more than 11 days (Gallardo et al., 2017, 2018a, 2019b; Pikalo et al., 2019; Zhao et al., 2019). Both intranasal and oronasal routes are similarly lethal although the nasal route resulted in higher ASF incidence than the oral route when using a lower infectious dose (Guinat et al., 2016). Once infected, the animals developed acute clinical signs between 3.5 to 14 (average) days with the death of 91 up to 100% of the infected animals between 7 to 21 days after the first case (Gabriel et al., 2011; Nurmoja et al., 2016; Olensen et al., 2017; Pietschmann et al., 2015; Vlasova et al., 2015). A similar picture has been observed in pigs exposed to the virus by direct contact with infected animals. The exposed animals developed similar acute clinical signs, which resulted in death between 11 to 25 days post exposure (Blome et al., 2013; Guinat et al., 2014; Gallardo et al., 2017; 2019b; Olesen et al., 2017).

From these experimental data, doubts remained about a potential reduction in the virulence of genotype II ASFV strains circulating in Europe and Asia, and the possibility that domestic pigs or wild boar may develop chronic infections and thus may recover and become carriers. However, the early identification of ASF outbreaks/cases in certain areas of Europe, as the Baltic countries and Poland were the disease has become endemic, has been hampered by the inherent difficulties to recognize the initial signs of infection (Gallardo et al., 2015a; 2018; Nurmoja et al., 2017). It is important to point out that, when introduced into a region or a domestic pig population, ASF is typically associated with high mortality rates and a rapid spread of outbreaks (Sanchez-Vizcaino et al., 2015; FAO 2020). However, even in acute infections, a 2-10% of the infected animals can recover. These survivors may establish a persistent infection in some tissues and, under certain natural or induced conditions (transport, underfeeding, immunosuppression, etc.) may reactivate the virus, thereby facilitating its transmission. Furthermore, these animals are protected to a secondary ASFV infection, remaining sub-clinically infected, acting as a potential source of infection for the environment and for healthy animals as they could show low levels of viremia (wild boar and domestic pigs). This explains the natural evolution of the ASFVs including the emergence of less virulent forms over time, as occurred in different geographic regions where ASF has been present for a long time (Africa, Iberian Peninsula and Sardinia) (Arias and Sanchez-Vizcaino 2002, 2012; Arias et al., 2018; Gallardo et al., 2015b, 2015c, 2018; 2019b).

Data obtained from the field in areas where the genotype II strains are circulating suggest the evolution of the ASFVs towards less virulence forms. Field epidemiological investigations conducted in Armenia and Estonia described the presence of atypical clinical forms of ASF coexisting with acute typical forms, suggesting the co-circulation of strains of different virulence in the countries (Nurmoja et al., 2017; Sargsyan et al., 2018; Zani et al., 2018). The work developed by Gallardo et al., (2018) with different EU ASFV genotype II strains confirmed the presence of virus of moderate virulence circulating among the wild boar population in Estonia in 2015 and in 2016 able to induce variable clinical signs in the infected domestic pigs ranging from acute, subacute to chronic ASF. Finally, two non-haemadsorbing (non-HAD) genotype II ASFVs were isolated from hunted wild boar in Latvia in February and November 2017 (Gallardo et al., 2019). Domestic pigs infected with these strains developed a non-specific or subclinical form of the disease and were protected from a re-infection with an HAD virulent Latvia strain. Non-HAD virus isolates were previously isolated in the past from regions where ASFV has been maintained for a long time. During the extended epizooty of the last century in the Iberian Peninsula, non-pathogenic and non-HAD viruses were collected in Portugal from pigs and ticks in field areas where most of the herds with seropositive pigs were detected (Boinas et al., 2004). Non-HAD strains were also reported in Spain, with a total of 206 non-HAD isolates obtained in the period between 1965 and the first semester of 1974 (Boinas et al., 200 4). Non-HAD isolates have been isolated also in Africa (Gonzague et al., 2001; Pini & Wagenaar, 1974; Thomson et al., 1979).

These data provide evidence of the natural evolution of the genotype II ASFVs in Central-Eastern Europe, from virulent to attenuated strains, able to induce different ASF clinical forms from acute to subclinical infections, which are coexisting in the field, in more or less proportion, depending of the affected region (tab. 1).

Table 1. Biological characterization of African swine fever virus (ASFV) through in vivo studies done at the EURL in domestic pigs with ASFV genotype II strains isolated from wild boar (WB) or domestic pig (DP) within the EU countries.

Biological characterization of African swine fever virus (ASFV) through in vivo studies done at the EURL in domestic pigs with ASFV genotype II strains isolated from wild boar (WB) or domestic pig (DP) within the EU countries.

Available ASF diagnostic tests

Given the demonstrated clinical evolution of the disease in some affected areas in Europe, mainly where the ASFV persists within the wild boar population, every country should update the contingency plan and the early warning system in place to prevent the ASFV entry in free areas, its spread, and subsequent maintenance. Any delay in outbreak response and implementation of control measures can result in greater viral contamination of the environment and disease spread (Bellini et al., 2016). Highly virulent ASFV isolates are associated with evident clinical forms and should therefore be easier to detect by passive surveillance. However, the demonstrated presence of animals that survive from sub-acute infections or even are subclinically infected, makes the passive surveillance not sufficient for early disease detection in the case of infection with moderately virulent or attenuated ASFV isolates. This is particularly important in high-risk areas. In case of wildlife infected animals, a combination of passive surveillance of dead wild boar and active surveillance in areas at highest risk should be considered (Arias et al., 2018). The active surveillance will also provide very valuable data on the evolution of the disease and guidance on the assessment of the effectiveness of the control measures. A surveillance system, to be successful, must have adequate laboratory support for a rapid diagnosis, being a key step to design effective control and eradication programs.

On the international level, laboratory methods, as well as sampling and shipping guidelines, can be found in the OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (Chapter 3.8.1, 2019 edition) and the respective EU Diagnostic Manual (European Commission Decision EC 2003/422/EC). A wide spectrum of accurate ASF diagnostic tests is available, most of them successfully employed in surveillance, control, and eradication programs (see Tab. 2).

The OIE-recommended tests for virus detection include both real-time and conventional PCR assays (Agüero et al., 2003; King et al., 2003; Fernández-Pinero et al., 2013), virus isolation and direct immunofluorescence tests (DIF) (OIE 2019). Virus isolation (VI) and identification by HAD tests, a characteristic feature of the ASFV-infected cells, are recommended as reference confirmatory tests in the event of a primary outbreak or a case of ASF (EC 2003). However, it is not likely to be the most fruitful approach for an effective ASF diagnosis since it is more expensive than other techniques, requires both specialized facilities and training, is time consuming and cannot be adapted to high throughput. In addition, attempts to isolate infectious virus from field-derived samples provide irregular results. The reason lies in the poor state of samples received, which affects the virus viability, especially on samples obtained from dead or hunted animals, such as wild boar (Gallardo et al., 2015a, 2019b). Other assays such as antigen detection ELISA, which allows for large-scale testing of samples, are also available but exhibiting lower diagnostic sensitivity than PCR tests. The use of DIF or antigen ELISA is only recommended as a herd assay and should be combined with some other virological and/or serological tests due to the lack of sensitivity in case of seropositive animals. In conclusion, the real-time PCR is considered the ‘gold standard’ virus detection test for early detection of the disease and for surveillance. This is due to its superior sensitivity, specificity, robustness and high-throughput application to detect the ASFV genome in any kind of clinical samples from domestic pigs, wild boar and ticks (Gallardo et al., 2015a, 2019; Oura et al., 2013). The use of the real-time PCR as the preferred virological method for routine diagnosis and the need of high-throughput application, added to its worldwide demand, has led to the emergence of a considerable number of commercial kits in recent years. The diagnostic performance of PCR kits should be ensured before being implemented in the lab (table 2).

Serological assays are the most commonly used diagnostic tests due to their simplicity, relatively low cost and need for slight specialized devices or facilities. For ASF diagnosis, the antibody detection is particularly relevant given that no vaccine is available, which means that the presence of anti-ASFV antibodies always indicates infection. Furthermore, anti-ASFV antibodies appear soon after infection and persist for up to several months or even years (Arias and Sánchez-Vizcaino 2002, 2012). Antibody-based surveillance is therefore essential for the detection of surviving animals, to elucidate the epidemiological characteristics of the epidemics, i.e., time since the virus introduction into a farm, and for detecting incursions involving low virulence ASFV isolates (Arias et al., 2018; Gallardo et al., 2015a, b, 2019b; Laddomada et al., 2019). The use of antibody detection assays was also crucial for successful eradication programs in the past (Arias and Sánchez-Vizcaíno 2002, 2012). Current ASFV antibody-based tests approved by the OIE involve the use of an ELISA for antibody screening, backed up by Immunoblotting (IB), Indirect Immunofluorescence (IIF) or the Indirect immunoperoxidase tests (IPT) as confirmatory methods (OIE 2019). Furthermore, the IPT has been proved as the best test for ASF serological diagnosis due its superior sensitivity, but moreover, its performance to test any kind of porcine material such as blood, exudate tissues or body fluids (Gallardo et al., 2015a). This is particularly relevant for wild boar surveillance and control programs.

Finally, the use of pen-side tests offers a first-line diagnosis that can be useful for rapid application in case of sanitary emergency. The time elapsed between the clinical suspicion and laboratory confirmation used to be relatively long due to the logistics of sending samples to official laboratories. On the other hand, in many countries, regional laboratories do not have the expertise, equipment and/or facilities to diagnose exotic diseases such as ASF. This can be sorted out by the use of pen-side tests for a first front-line diagnosis under field conditions, giving real-time data on the animal’s infection status. Two different lateral flow devices (LFDs) for the detection of antibodies or the viral antigen in blood are commercially available by INGENASA so far, and it is expected several others come in the near future. However, from the published data, the LFDs should not be used alone due to a very limited sensitivity compared to the gold standard methods, overall in the case of the LFD for the detection of antigens (Sastre et al., 2016; Gallardo et al., 2019). It is important to point out that antibody LFD penside tests are not useful for the detection of acute forms of the disease since cannot be detected before 12-14 days post-infection. Then, these tests need to be accompanied by virus/antigen detection techniques. The analysis of suspicious samples by both virus and antibody detection techniques/penside tests will give us an initial picture of what is going on. Test sensitivity needs to be high, whereas specificity is less critical, since any positive result will need to be verified by the competent National reference laboratory. The use of pen-side tests for on-farm or field screening requires to be restricted to official veterinarians or regional laboratories with limited resources and should be used taking into consideration a specific circumstances.

Table 2. African swine fever validated diagnostic tests.

The starting point for any laboratory investigation of ASF is sample collection. An important consideration is the purpose of the investigation, for example disease diagnosis, disease surveillance, or health certification. Which animals to sample will depend on the objective of the sampling. For example, when investigating an outbreak (passive surveillance), sick and dead animals should be targeted, while the oldest animals should be sampled when checking if animals have been exposed to the disease (active surveillance) (Beltrán-Alcrudo et al., 2017). To be effective, proper samples combined with the selection of diagnostic methods, is of fundamental importance in order to make a rapid and reliable diagnosis. Nevertheless, in certain situations, sampling can be the bottleneck of swine fever diagnosis, especially in the case of wild boar, but also in remote areas. For this reason, alternative sampling strategies and sample matrices (oral fluids, dried blood-spots on filter papers and swabs) have been tested for ASF (often combined with Classical swine fever sampling) especially for wildlife specimens and under rural conditions. However, most of them are not yet in routine use and need further validation (Blome et al., 2014; Braae et al., 2015; Davies et al., 2017; Grau et al., 2015; Michaud et al., 2007; Mur et al., 2013; Petrov et al., 2014; Randriamparany et al., 2016).

The ASF diagnostic interpretation

As in any other disease, there is not a single test being 100% reliable (sensitive and specific). For this reason, final diagnosis should be based on the interpretation of the results derived from the use of a number of validated tests, in combination with the information coming from disease epidemiology, scenario, and the clinical signs. A detailed understanding of the time course of viremia and antibody seroconversion during the ASFV infection is a prerequisite to obtain relevant information about the dynamic of the infection in affected areas and to support control and eradication programs. An appropriate diagnosis therefore should involve the detection and identification of ASFV-specific antigens or DNA and antibodies.

On the basis of experimental data gained at the EURL, upon infection with virulent ASFV strains, an starting weak viremia (Cycle threshold Ct > 35) can be detectable by real-time PCR at on average of 3.75 ± 1.4 days, two days before the onset of the clinical signs. No antibodies are developed at this early time. Therefore, a weak PCR positive result on a field blood sample in absence of antibodies could be an indicator of an early phase of the infection (< 1week). Spleen, lymph nodes, liver, and lungs are the sites of secondary viral growth and, after 24-30 hours of the primary viremia, all tissues contain the virus, reaching the maximum titres around 7-8 days after being infected. The ASFV is therefore easily detected in any kind of porcine sample by real-time PCR and even using the antigen detection techniques (DIF or ELISA). A weak antibody response can be detected as early as 7-8 days by IPT in sera and in tissues exudate, mainly spleen and lung, 2-3 days before being rarely detectable by ELISA. In this scenario, the mortality can reach up 92% to 100% within 4 (peracute form) to 12 (acute form) days after the infection.

Since acute forms are predominant at the beginning of outbreaks, the measures taken in free areas bordering infected areas are based on a risk assessment and on enhanced passive surveillance and PCR testing (SANTE/7113/2015 – Rev 11). However, false positive PCR results, although rare, can occur (e.g., due to lab contamination or other factors). It is unlikely that a primary outbreak (or case) of ASF will occur only based on a PCR positive result. The most conclusive evidence of infection is the isolation of ASFV, but several situations limit or even prevent its application. Virus isolation is time-consuming, requires specialized laboratory and staff and, in comparison with the PCR, has limited sensitivity, particularly on samples obtained from dead or hunted wild boar, or in weak positive PCR samples (Gallardo et al., 2015, 2019). The occurrence of ASFV infection can be confirmed if clinical signs or lesions of disease have been detected in the pigs in question and at least two distinct virus or antibody detection tests have given a positive result on samples taken from the same-suspected pig. In wild boar, if virus isolation is not possible, a primary case of ASF can be confirmed when at least two virus or antibody detection tests have given a positive result (EC 2003).

Where ASF becomes endemic, increased numbers of subacute, chronic, and subclinical infections occur and mortality rates decline over time. In clinical terms, subacute ASF develops over a 10-20-day period and the mortality rate ranges from 30 to 70% after 20 days post infection (Arias and Sánchez-Vizcaino, 2012; Beltrán-Alcrudo et al., 2017). Viremia can be detected by real-time PCR on average of 8.5 ± 3.6 days and antibodies by ELISA and IPT from 10 days, reaching mean antibody titres of 1:20,000 from the third week. All tissues obtained from animals that succumb within the first month to the infection are positive by PCR and in IPT (Gallardo et al., 2018a, b). The presence of ASF is therefore easily confirmed combining both virus and antibody detection tests, in either blood, sera or tissue samples.

In recovered pigs surviving acute or subacute infections, the viremia, clearly detectable within the first month, declines over the time and only weak PCR results (Ct > 35) can be sporadically detected for up to 78 days (Gallardo, et al., 2018). The common feature in these survivors is the presence of high antibody titer in either blood, sera or tissue samples for the entire life of the animal. Nevertheless, the detection of antibodies in a field sample that resulted PCR weak or negative, is not only an indicator of the presence of survivors from acute or subacute infections. Animals infected with non-HAD and low virulence strains seroconvert after the first week of the infection even in absence of clinical signs or viremia. The antibodies are easily detected by IPT and ELISA, reaching antibody levels > 1:160,000 after one month and are maintained over the time (Gallardo et al., 2015c, 2018, 2019a; Leitao et al., 2001; Sánchez-Cordon et al., 2017; Sánchez-Vizcaíno et al., 2015). The relative low percentage of non-HAD viruses isolated within the EU could come to the fact that these non-HAD viruses are more difficult to isolate than HAD viruses since the viraemia they cause is sporadic and virus is isolated in small amounts from the organs. Infection of wild boar with these attenuated non-HAD isolates may account for the seropositive animals detected in the field in absence of clinical signs and viremia. These data emphasize the fact that early detection based only on clinical signs and ASFV genome detection is not an efficient approach for the control of ASF in the current epidemiological situation in Europe. It is likely that the European wild boar is being endemically infected in certain regions becoming a recurrent source of infection to other wild boar, but also to domestic pigs. Since each animal could be at a different stage of the infection, sub-clinically ASFV infected wild boar or domestic pigs can be detectable if both virus and antibody detection tests are used, for examining transient viremia and the presence of anti-ASFV-specific antibodies. A positive test for the presence of the virus indicates that the tested animal was undergoing infection at the time of sampling. On the other hand, a positive ASFV antibody test indicates an ongoing or past infection, where the animals have recovered (and may remain seropositive for life; tab. 3).

Table 3. Interpretation of the ASF diagnostic results.

Final remarks

In conclusion, control-eradication programs in areas with a clear endemic tendency should be reviewed and updated and should include parallel routine laboratory monitoring together with the regular clinical inspection. The use of the most fitting diagnostic tools combining both ASF virus and antibody detection will improve the efficacy of disease control measures, regardless of the nature of the circulating ASFV strains (Arias and Sánchez-Vizcaíno 2002, 2012; Gallardo et al., 2015a, 2019b). An accurate evaluation of the results of the serological and virological tests must be carried out, considering all the clinical and epidemiological findings, in the framework of the enquiry to be carried out in case of suspicion or confirmation of ASF.

Published in the proceedings of the International Pig Veterinary Society Congress – IPVS2020. For information on the event, past and future editions, check out https://ipvs2022.com/en.