A Computational Framework for the Optimisation of Antivenom Pharmacokinetics and Pharmacodynamics

Abstract

Snakebite envenomation causes over 100,000 deaths and 400,000 cases of disability each year, most of which are preventable through timely access to safe and effective antivenoms. Antivenoms are currently derived from the sera of hyper-immune animals, which incurs considerable limitations on their efficacy, cost, and safety. Owing to this, there has been significant interest in the recombinant production of antivenoms using cellular culture, which would make treatments more targeted and less likely to trigger adverse reactions, and could enable the use of alternative scaffolds with novel biophysical properties. The application of modern antibody engineering techniques may also allow further control of binder properties. While the use of alternative engineered scaffolds has been suggested to offer pharmacokinetic benefits in the treatment of different snakebites, their performance relative to conventional antivenom formats is currently unknown. This thesis explores the impact of antivenom design on treatment outcome and establishes a computational framework for the optimisation of antivenom pharmacokinetics and pharmacodynamics. A compartmental model of envenomation and treatment was defined to allow simulation of a rescue experiment. The model was parameterised using existing experimental data from rabbits, and enables user-control of various envenomation and treatment parameters alongside the prediction of antivenom pharmacokinetics from molecular size. A series of case studies indicate that the model can recapitulate the dynamics of envenomation-treatment systems, and illustrate the large dose reductions that recombinant antivenoms confer. The model was subsequently applied to define guidelines for optimised antivenom design. The treatment of an elapid and viper envenomation with a set of theoretical antivenoms was simulated at a range of treatment time delays, allowing an exploration of the effects of multiple sources of variability. Using this dataset, optimised antivenoms were identified with an area under the curve metric and a global sensitivity analysis was performed to ascertain the influence of different antivenom parameters on treatment outcome. These simulations show k_on to primarily mediate treatment efficacy. While molecular weight has a negligible direct impact on treatment outcome, small scaffolds can be more easily designed for high neutralising efficacy, particularly when treatment is delayed. Similar parametric trends are apparent in the effective treatment of both envenomation cases, however the optimised bounds for viper treatment are more constrained.

Date of Award	4 Mar 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Sabine Hauert (Supervisor), Johanna A Blee (Supervisor) & Ian R Collinson (Supervisor)