Accelerating therapeutics development during a pandemic: population pharmacokinetics of the long-acting antibody combination AZD7442 (tixagevimab/cilgavimab) in the prophylaxis and treatment of COVID-19

Abstract

ABSTRACT AZD7442 is a combination of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-neutralizing antibodies, tixagevimab and cilgavimab, developed for pre-exposure prophylaxis (PrEP) and treatment of coronavirus disease 2019 (COVID-19). Using data from eight clinical trials, we describe a population pharmacokinetic (popPK) model of AZD7442 and show how modeling of “interim” data accelerated decision-making during the COVID-19 pandemic. The final model was a two-compartmental distribution model with first-order absorption and elimination, including standard allometric exponents for the effect of body weight on clearance and volume. Other covariates included were as follows: sex, age >65 years, body mass index ≥30 kg/m 2 , and diabetes on absorption rate; diabetes on clearance; Black race on central volume; and intramuscular (IM) injection site on bioavailability. Simulations indicated that IM injection site and body weight had > 20% effects on AZD7442 exposure, but no covariates were considered to have a clinically relevant impact requiring dose adjustment. The pharmacokinetics of AZD7442, cilgavimab, and tixagevimab were comparable and followed linear kinetics with extended half-lives (median 78.6 days for AZD7442), affording prolonged protection against susceptible SARS-CoV-2 variants. Comparison of popPK simulations based on “interim data” with a target concentration based on 80% viral inhibition and assuming 1.81% partitioning into the nasal lining fluid supported a decision to double the PrEP dosage from 300 mg to 600 mg to prolong protection against Omicron variants. Serum AZD7442 concentrations in adolescents weighing 40–95 kg were predicted to be only marginally different from those observed in adults, supporting authorization for use in adolescents before clinical data were available. In these cases, popPK modeling enabled accelerated clinical decision-making.