As the global spread of ASF continues, the situation for pig holders and nature conservationists has never been as tense as it is now. Millions of pigs are at risk, representing the livelihoods of farmers, and entire species of certain wild suids, e.g. bearded pigs, that are now threatened with extinction (Ewers, Nathan et al. 2021). Against this background, a vaccine is urgently needed to complement available control measures. In this context, we may not be in the position to wait for a perfect vaccine candidate and should rather stress the application of a practicable solution as fast as reasonably possible. On the other hand, we cannot afford to compromise on vaccine safety, or as Gavier-Widen, Stahl et al. (2020) put it, allow hasty solutions. Experiences in Spain and Portugal from the last century using attenuated field isolates (Petisca 1965), and possibly very recently in China (https://www.reuters.com/article/us-china-swinefever-vaccines-insight-idUSKBN29R00X, visited May 15th 2022), show us that premature field testing of live vaccines can cause prolonged forms of ASF with extended shedding and delayed clinical characteristics, often below the detection limit. The use of vaccine viruses with unacceptable residual virulence or that revert to virulence can lead to an iatrogenic, self-sustaining infection cycle with increasing virulence. This would further complicate eradication efforts and these scenarios must be avoided under all circumstances.
Consequently, we took one of the few fully efficacious and possibly licensable vaccine candidates, “ΔMGF”, and examined its safety profile in terms of genetic stability and reversion to virulence under a worst-case scenario.
In our traditional reversion to virulence study over five animal passages, we observed the occurrence of a virus variant in one animal of passage one that subsequently overgrew the wild-type MSV. This variant was genetically characterized by a large deletion at the 5’-end of the genome and an accompanied duplication at the 3’-end. Clinically, the variant was associated with a slightly increased virulence, e.g. induction of a short episode of fever in most animals in the later passages. However, all animals even in the final 5th passage were clinically inapparent at the end of the experiment, showed no evidence of incipient, chronic infection, and showed little or no vaccine virus in the tissues tested.
For further characterization of the “ΔMGFnV”, we conducted comparative in vitro growth kinetics of both viruses in primary macrophages, revealing no indications for a significant advantage in in vitro replication of the variant virus in this setup. Clear limits in explanatory power should be considered, however, since modifications in the MGF regions are known to have effects on interferon expression (Wang, Kang et al. 2021), and the full consequences of such changes are probably only observable in vivo. The underlying factors causing the in vivo replication advantages of ΔMGFnV remain therefore unanswered, stressing that many of these questions can only be addressed by further animal experiments due to the highly complex virus-host interactions of ASFV.
Whether our findings showing genetic changes and a slight rise in virulence after in vivo passaging disqualify the vaccine candidate is a matter of critical debate, since the mode of transmission is highly artificial (selecting particularly positive samples for further passaging and intramuscular injection of tissue homogenates), and there is no evidence of reversion to the original levels of high virulence of ASFV “Georgia07”. In this context, it could be discussed whether a brief period of fever can be tolerable for an efficacious and therefore otherwise practicable first generation ASFV vaccine. A prerequisite for this assumption would be that the novel variant is genetically stable and does not mark the beginning of a maintained process of further genetic adaption, only mirrored by this first mutation.
The mutant detection qPCR described here was also used retrospectively for representative samples from several efficacy tests with the ΔMGF vaccine candidate, with clearly negative results (data not further shown).
It should be noted, however, that in this case a viral variant with altered geno- and phenotypic properties has already emerged in the first passage, i.e. after application of the MSV. This phenomenon did not occur in any of the previous studies, but it is relevant because it could also happen in the field during intensive use. It is also remarkable that the mechanism of a large deletion complemented by reorganization of genomic regions has been observed for an ASFV strain under natural circumstances before (Zani, Forth et al. 2018). We may have unraveled a common mechanism of ASFV for genetic adaption when a certain selection pressure is applied.
For a fact-based benefit-risk assessment, further studies with the evolved virus variant are needed, which should address excretion, long-term effects, and transmission to naive contact animals.
In general, our study confirms that even the most promising ASF live vaccine candidates require very comprehensive safety testing (Gavier-Widen et al., 2020). However, it also provides a first indication of what an attenuated ASF vaccine virus would need to do to increase its replication efficacy in the animal or to compensate for deletions in the MGF region. This knowledge can be deepened and used to devise strategies to make these changes even more difficult for the virus.
If field application is considered after benefit-risk-assessment, one should apply genetic tools to differentiate infected from vaccinated animals (DIVA). The PCR described here could aid such approaches. Moreover, conditional licensing under controlled conditions could be a solution to obtain field data for final decisions on the use of the vaccine to complement national control measures.