Xuanlin Liu (1), Kaley C. Hanrahan (2), Sean Avedissian (3), Evelyn J. Franco (2), J. G. Coen van Hasselt (1), Ashley N. Brown (2), Anne-Grete Märtson (1)
(1) Leiden Academic Centre for Drug Research, Leiden University, Leiden, the Netherlands, (2) Department of Pharmaceutics, College of Pharmacy, University of Florida, Orlando, Florida, USA, (3) Department of Pharmacy Practice and Science, University of Nebraska Medical Center, Omaha, NE, USA
Objectives: The coronavirus disease 2019 (COVID-19) pandemic caused by the RNA-virus SARS-CoV-2 has shown the development of new antiviral agents. Nirmatrelvir inhibits the 3C-like protease (3CLpro) enzyme in SARS-CoV-2 and is used in combination with ritonavir to treat COVID-19. Nuceloside/nucleotide analogs, like favipiravir and remdesivir have been used for SARS-CoV-2 with different levels of effectiveness. In vitro investigations are needed to maximize drug efficacy and gain understanding to viral dynamics under drug therapy. The aim of this study is to investigate the viral dynamics under static nirmatrelvir treatment and estimate the effect of therapy delay on viral load decline.
Methods: Experiments: Antiviral assays with static drug concentrations were done in 6-well plates with ACE2-A549 cells: 2 x 106 cells were inoculated with an multiplicity of infection (MOI) of 0.03 for an hour, after the incubation the monolayers were washed and replaced with tissue culture medium. Five drug concentrations were evaluated alongside a non-treatment control. The tested drug concentrations were: 0.004 ug/mL, 0.0156 ug/mL, 0.0625 ug/mL, 0.25 ug/mL, and 1 ug/mL. Viral supernatants were collected for 4 days. A plaque assay was performed on Vero E6 cells to determine the infectious burden in PFU/mL (plaque-forming unit/mL). Model development: The viral dynamic modelling was performed using the R package nlmixr2. A previously established target cell-limited (TCL) model was applied to build the basic infection model using data from the control group [1]. Viral kinetic parameters, i.e., viral infection rate (β), death rate of infect cells (δ), viral production rate (ρ), and viral clearance (c) were estimated for control experiments only. Next, the drug effect was incorporated using an Emax effect model to describe the inhibition of viral production rate (ρ). In this step, δ and c were fixed to their final estimates in the TCL model without drug effect to achieve a better fitting result. An additive residual error model was applied in all the models tested. The estimation was performed using the first-order conditional estimation with interaction (FOCEi) method. The final model was selected based on OFV, goodness of fit (GOF) plots, and visual predictive check (VPC).
Results: The in vitro infection experiments showed the fastest viral load decline when therapy was started on day 0 and nirmatrelvir dose was 0.25 ug/mL or 1 ug/mL. Antiviral effect was limited for all doses when therapy was initiated from day 1 onwards. The in vitro data was best described by a TCL model with Emax drug effect. The final estimates (RSE%) of viral kinetic parameters (β = 4.62×10-8 (1.28%) PFU-1/day, δ = 8 day-1 (fix), c = 3.46 day-1 (fix), ρ = 980 (2.81%) PFU/cell/day) and drug effect parameters (Emax = 0.712 (42.2%), hill coefficient = 1.47 (137%), EC50 = 0.039 (11.8%) ug/mL) fit the experimental data well.
Conclusions: The study showed that the antiviral effect was largely suppressed due to delayed therapy initiation time, thus a higher concentration was needed to reach the expected viral load decline. A viral dynamic model was developed to describe the viral dynamics under nirmatrelvir therapy. A TCL viral infection model with Emax drug effect best described the drug concentration-effect relationship. Further model optimization will be done to quantify the effect of therapy delay and design new dosing regimens.
References:
[1] Zhang S, Agyeman AA, Hadjichrysanthou C, Standing JF. SARS-CoV-2 viral dynamic modeling to inform model selection and timing and efficacy of antiviral therapy. CPT Pharmacometrics Syst Pharmacol. 2023 Oct;12(10):1450-1460. doi: 10.1002/psp4.13022.
Reference: PAGE 32 (2024) Abstr 11227 [www.page-meeting.org/?abstract=11227]
Poster: Drug/Disease Modelling - Infection