Elucidating the Local Transmission Dynamics of Highly Pathogenic Avian Influenza H5N6 in the Republic of Korea by Integrating Phylogenetic Information

Abstract

Highly pathogenic avian influenza (HPAI) virus is one of the most virulent and infectious pathogens of poultry. As a response to HPAI epidemics, veterinary authorities implement preemptive depopulation as a controlling strategy. However, mass culling within a uniform radius of the infection site can result in unnecessary depopulation. Therefore, it is useful to quantify the transmission distance from infected premises (IPs) before determining the optimal area for preemptive depopulation. Accordingly, we analyzed the transmission risk within spatiotemporal clusters of IPs using transmission kernel estimates derived from phylogenetic clustering information on 311 HPAI H5N6 IPs identified during the 2016-2017 epidemic, Republic of Korea. Subsequently, we explored the impact of varying the culling radius on the local transmission of HPAI given the transmission risk estimates. The domestic duck farm density was positively associated with higher transmissibility. Ring culling over a radius of 3 km may be effective for areas with high dense duck holdings, but this approach does not appear to significantly reduce the risk for local transmission in areas with chicken farms. This study provides the first estimation of the local transmission dynamics of HPAI in the Republic of Korea as well as insight into determining an effective ring culling radius.

Keywords: H5N6 subtype; HPAI; avian influenza; controlling strategy; culling; local transmission; phylogenetic; poultry; spatial analyses; transmission kernel.

Figure 1

Distribution of distances between HPAI-infected premises for all virus genotypes and specific genotypes. The frequency of each distance range for each genotype category is stacked within a column. Green indicates all infected farms, and purple, orange, yellow, and blue indicate premises infected with the C2, C3, C4, and C5 HPAI virus genotypes, respectively. HPAI, highly pathogenic avian influenza.

Figure 2

Spatiotemporal clusters of poultry farms involved in HPAI H5N6 outbreaks associated with four virus genotypes during the 2016 to 2017 epidemic in the Republic of Korea. Two significant clusters were identified. Farms in cluster A (red shading) were at higher relative risk (RR) of infection with the C3 genotype (RR 41.28), whereas those in cluster B (yellow shading) were at high risk for C4 infection (RR 3.43). Red dots, blue triangles, brown stars, and green crosses denote premises infected with the C2-genotype virus, C3 virus, C4 virus, and C5 virus, respectively. Gray dots indicate all other poultry holdings. HPAI, highly pathogenic avian influenza.

Figure 3

Daily probability of a susceptible farm in two clusters being infected by infectious premises according to inter-farm distance. The distance-based daily probability of HPAI infection at a poultry holding located at different distances from individual infectious premises was estimated using transmission kernel estimates for each cluster assuming an infectious period of 7 days. Spatial transmission kernel estimates were derived from clustered HPAI outbreaks associated with two specific virus genotypes. HPAI, highly pathogenic avian influenza.

Figure 4

Cumulative number of infected premises (IPs) and preemptive culled premises (PCs) from simulation of four different pre-emptive culling ranges among the poultry holdings in cluster A (top) and cluster B (bottom); 0.5 km (red solid line), 1 km (green dotted line), 2 km (blue dot dash line) and 3 km (purple dashed line) from an IP. Top left and top right plots represented cumulative number of infected premises and pre-emptively culled premises in cluster A over the days. Bottom left and bottom right plots represented cumulative number of infected premises and pre-emptively culled premises in cluster B over the time (unit: days).