Disease severity declines over time after a wild boar population has been affected by classical swine fever--legend or actual epidemiological process?

Classical swine fever (CSF) is a severe multi-systemic disease that can affect both domestic pigs and wild boar. Past outbreaks in European wild boar involved high-virulent CSF virus (CSFV) strains and were mostly self-limiting. In these cases, morbidity and mortality rates were high in the affected regions. In contrast, endemic infections have been observed in several European wild boar populations in recent decades. Morbidity and mortality rates were much lower despite the fact that outbreaks were still detected via diseased or fallen animals. The virus strains involved were mostly classified as genotype 2.3 strains of moderate virulence causing age-dependent disease outcomes. The mechanisms leading to the establishment and perpetuation of endemicity are still not fully understood, but the factor "moderate virulence" seems to be of considerable importance. In this study, we aim to clarify whether the perception of declined 'CSF severity' could hypothetically reflect the adaptation of an initially high-virulent virus or whether this might be better explained as a misinterpretation of observations. A mechanistic eco-epidemiological model was employed to follow up a highly virulent strain of CSFV introduced into large connected wild boar populations. In the model, the virulence of the CSF virus is represented by case mortality and life expectancy after lethal infection. Allowing for small stochastic variation, these two characteristics of the virus are passed on with every new simulated infection that occurs. Model analysis revealed a decrease from high to moderate case mortality within a few years of simulated perpetuation of the virus. The resulting mortality corresponded to the level where the population average of the infectious period and the basic reproduction number of the disease were maximal. This shift in virulence was sufficient to prolong virus circulation considerably beyond the epidemic phase of the simulated outbreaks. Alternative mechanistic explanations for the decrease in disease severity in a CSF-affected wild boar population were evaluated in the light of the simulation experiments and the available epidemiological or virological evidence. In conclusion, the current virus isolates of subgroup 2.3 might be the ideally adapted variants of the CSF virus for long-term perpetuation in wildlife and indeed may have evolved (once) during past outbreaks in large populations. A repeated perception of a declining severity of disease pattern during the course of a CSF outbreak, however, favours the explanation based on monitoring and detection biases rather than repeated observation of selection against highly virulent virus during the time of virus perpetuation.