Investigating risks to Australia of emergent henipaviruses

Abstract

To investigate the risks to Australia of emergent henipaviruses, studies were conducted with the aim of: (1) determining the occurrence of henipaviruses in targeted fruit bat populations in northern Australia, Papua New Guinea, East Timor and Indonesia; (2) investigating the extent and nature of contact between fruit bat populations in Australia with populations in Papua New Guinea, East Timor and Indonesia using satellite telemetry; and (3) investigating the dynamics of henipavirus infection in a fruit bat population. The risk of Nipah virus (NiV) becoming established in Australian fruit bat populations was assessed utilising data generated during this project, together with available literature and expert opinion. Targeted sampling of fruit bat populations found strong evidence for the presence of NiV in Pteropus vampyrus and Rousettus amplexicaudatus in East Timor. Evidence was also found for the presence of henipaviruses that were not Hendra virus (HeV) nor NiV in Pteropus alecto and Acerodon celebensis on Sulawesi and P. alecto on Sumba. The closest to Australia that NiV had previously been detected is Singapore although serological evidence of a henipavirus more closely related to NiV than Hendra virus (HeV) has also been reported from Sumatra and Java. Studies using satellite telemetry showed P. alecto to be capable of crossing Torres Strait (between Papua New Guinea and Australia) thus indicating potential close contact among the populations of this species in northern Queensland, south-western Papua New Guinea and south-eastern Papua, Indonesia. In contrast, an individual P. alecto tracked on Sumba, Indonesia, did not leave this island and two P. vampyrus tracked on Timor remained on that Island although one of these bats moved to Indonesian West Timor. The infection dynamics of HeV in a Pteropus conspicillatus population in northern Queensland was investigated using a serial cross-sectional study over a 25 month period. The results suggested that infection was endemic in the population over the study period and indicated that age, pregnancy and lactation were significant risk factors for a detectable neutralising antibody response. An endemic infection pattern for HeV in this bat population is contrary to a previously proposed episodic infection pattern, but is consistent with recent studies on henipavirus infection in captive bats in Malaysia and free-living bats in Equatorial Guinea. From the current targeted sampling of fruit bat populations and satellite telemetry studies, it appears that an incursion into Australia of a fruit bat infected with NiV is a plausible event, although the probability and consequences of this occurring are not clear. The infection dynamics study has implications for understanding the process of spillover of infection of HeV from bats to horses in Australia, as well as for the consequence of the incursion of a NiV infected bat (or bats) to Australia. Taking the above information into account, a qualitative risk assessment was undertaken to assess the risk of NiV becoming established in Australian fruit bats as a result of movements from nearby regions. Where data gaps were identified, expert opinion was used in this assessment via an Expert Opinion Workshop. The risk was estimated to be very low but was associated with a high level of uncertainty. Different methods for combining expert opinion were also used in the risk assessment model to determine their influence on risk outcome. In addition to estimating the risk of NiV establishing in Australian fruit bats, this study provides a framework for periodically reassessing the risk as further information becomes available. It also identified key areas of data sparsity where future research could be directed.