A mathematical framework for estimating pathogen transmission fitness and inoculum size using data from a competitive mixtures animal model.

Journal article

We present a method to measure the relative transmissibility ("transmission fitness") of one strain of a pathogen compared to another. The model is applied to data from "competitive mixtures" experiments in which animals are co-infected with a mixture of two strains. We observe the mixture in each animal over time and over multiple generations of transmission. We use data from influenza experiments in ferrets to demonstrate the approach. Assessment of the relative transmissibility between two strains of influenza is important in at least three contexts: 1) Within the human population antigenically novel strains of influenza arise and compete for susceptible hosts. 2) During a pandemic event, a novel sub-type of influenza competes with the existing seasonal strain(s). The unfolding epidemiological dynamics are dependent upon both the population's susceptibility profile and the inherent transmissibility of the novel strain compared to the existing strain(s). 3) Neuraminidase inhibitors (NAIs), while providing significant potential to reduce transmission of influenza, exert selective pressure on the virus and so promote the emergence of drug-resistant strains. Any adverse outcome due to selection and subsequent spread of an NAI-resistant strain is exquisitely dependent upon the transmission fitness of that strain. Measurement of the transmission fitness of two competing strains of influenza is thus of critical importance in determining the likely time-course and epidemiology of an influenza outbreak, or the potential impact of an intervention measure such as NAI distribution. The mathematical framework introduced here also provides an estimate for the size of the transmitted inoculum. We demonstrate the framework's behaviour using data from ferret transmission studies, and through simulation suggest how to optimise experimental design for assessment of transmissibility. The method introduced here for assessment of mixed transmission events has applicability beyond influenza, to other viral and bacterial pathogens.

Authors: McCaw, J; Arinaminpathy, N; Hurt, A; McVernon, J; Mclean, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Public Library of Science
• Journal: PLoS computational biology
• Volume: 7
• Issue: 4
• Article number: e1002026
• Publication date: 2011-04-01
• DOI: 10.1371/journal.pcbi.1002026
• EISSN: 1553-7358
• ISSN: 1553-734X
• Language: English
• Keywords: Algorithms; Models, Biological; Humans; Drug Resistance, Viral; Genetic Fitness; Computational Biology; Influenza, Human; Computer Simulation; Orthomyxoviridae