Is a diagnostic system based exclusively on agar gel immunodiffusion adequate for controlling the spread of equine infectious anaemia?

Abstract: To improve the efficiency of the National equine infectious anaemia (EIA) surveillance program in Italy, a three-tiered diagnostic system has been adopted. This procedure involves initial screening by ELISA (Tier 1) with test-positive samples confirmed by the agar gel immunodiffusion test (AGIDT) (Tier 2) and, in the case of ELISA positive/AGIDT negative results, final determination by immunoblot (IB) (Tier 3). During this evaluation, 74,880 samples, principally collected from two Regions of Central Italy (Latium and Abruzzo) were examined, with 44 identified as negative in AGIDT but positive in both ELISA and IB. As the majority of these reactions occurred in mules, an observational study was conducted in this hybrid equid species to investigate if there is a correlation between plasma-associated viral loads and serological reactivity, to test the hypothesis that false-negative or very weak positive AGIDT results are associated with elite control of EIA virus (EIAV) replication accompanied by reduced transmission risks. The study animals consisted of 5 mules with positive AGIDT readings, along with another 5 giving negative or very weak positive results in this test. All mules were seropositive in Elisa and IB. Samples were collected routinely during an initial 56-day observation period, prior to dexamethasone treatment lasting 10 days, to determine the effect of immune suppression (IS) on clinical, humoral and virological responses. All mules were monitored for a further 28 days from day 0 of IS. None of the animals experienced relevant clinical responses before IS and there were no significant changes in antibody levels in ELISA, IB or AGIDT. However, plasma-associated viral-RNA (vRNA) loads, as determined using TaqMan(®) based RT-PCR, showed unexpectedly high sample to sample variation in all mules, demonstrating host-mediated control of viral replication is not constant over time. Furthermore, there was no apparent correlation between vRNA loads and antibody reactivity in serological tests. Analysis of PCR products established all mules were infected with viruses possessing nucleotide sequence similarity, varying from 77 to 96%, to previously identified European EIAV strains. Following IS, all mules showed increases in plasma-associated vRNA loads, suggesting control of EIAV replication is mediated by immune responses in this hybrid species. However, only three mules showed anamnestic humoral responses to rises in viral loads, as defined by at least a four-fold increase in ELISA titre, while two remained AGIDT-negative. This study demonstrates that viral loads in equids with consistent ELISA/IB positive-AGIDT negative to very weak positive test results (Group N) can be equivalent to those that produce clearly positive results in all three serologic tests (Group P). Therefore, such animals do not pose inherently lower risks for the transmission of EIAV. Consequently, the exclusive use of the AGIDT, as prescribed by the World Organization of Animal Health (OIE) for diagnosis of EIA prior to the international movement of horses, can report as negative some EIAV-infected equids. These results dramatically underscore the necessity of combining the specificity of AGIDT with tests with higher sensitivity, such as the ELISA and the power of the IB to enhance the accuracy of EIA diagnosis.