Guillaume Lingas
Université de Paris, IAME, INSERM, F-75018 Paris, France
Objectives: Lassa fever is a potentially lethal/disabling hemorrhagic fever, endemic in West Africa, caused by Lassa virus (LASV), that afflicts from 100000 to 300000 people each year according to the CDC, resulting in an estimated 5,000 deaths due to hemorrhage and multi-organ failure [1]. With no vaccine currently available, only one treatment, ribavirin (RBV), is currently recommended by the WHO [2]. However, its efficacy is not proven in humans. Furthermore, this treatment is associated with potential hematologic side effects [3]. In non-human primates (NHP), polymerase inhibitor favipiravir (FPV), an anti-Influenza A agent, at high doses prevents death in 100% cases [4], and therefore constitutes a very promising drug for the treatment of LASV.
Taking profit of our previous experience regarding favipiravir, especially its pharmacokinetics [5] in NHP and use in other viral diseases such as Ebola [6], our goal was to build a within-host model of Lassa viral infection in order to unravel the host-pathogen-drug interactions. We aimed at determining RBV and FPV most likely modes of action (MOA) against LASV as well as EC50s and average efficacies of both drugs on LASV. Finally, using this model, we predicted the impact of FPV and RBV on LASV when used at human concentrations.
Methods: Here, we develop the first mathematical model recapitulating Lassa viral dynamics in NHP, using virological data from 24 cynolmogus macaques, treated by either FPV, RBV or left untreated. Animals treated with FPV received a loading dose of 300 mg/kg intravenously the first day of treatment, followed by subcutaneous injection of 50 mg/kg TID (N = 4) or 300 mg/kg QD (N = 4). Animals treated with RBV received a loading dose of 30 mg/kg the first day of treatment, followed by subcutaneous injections of 10 mg/kg TID (N = 4) or 30 mg/kg QD (N = 4). Animals were frequently sampled for RNA viral load and TCID50, the latter being considered as the subgroup effectively infectious among all viruses produced. We built a viral dynamic model describing the links between different types of virions, cells, immune system and treatments effects.
Favipiravir and ribavirin are generally considered to block viral production by disrupting the RNA polymerase. However recent findings in Marburg and Ebola viruses show that FPV could cause lethal mutagenesis [7]. Therefore, we considered both drugs to either act as viral production blockers or as mutagenesis agents i.e. decreasing infectivity, under the form of an Emax model, adding a Hill coefficient for FPV. We compared both MOA in their ability to reproduce our data in order to discriminate the one most likely to occur, and considered uncertainty on FPV Hill coefficient for both MOA using a model averaging (MA) approach. Finally, we simulated viral kinetic profiles using human concentrations of FPV found in the JIKI trial that assessed its efficacy in Ebola and human concentrations of RBV found in LASV infected patients.
Results: With this biological model of LASV infection, we were able to properly reproduce our data and enlightened a specific MOA of both drugs against LASV, causing mutagenesis and decreasing infectivity. Indeed, models considering this MOA consistently provided a better fit than a model assuming that the drugs blocked viral production, irrespective of the Hill coefficient.
We estimated average efficacies throughout treatment course in reducing infectivity of RBV at 40% at 30 mg/kg/day and FPV at 60% and 91% at doses 150 mg/kg/day and 300 mg/kg/day respectively. High dose favipiravir efficacy was about twice the one found in Ebola with the same daily dose [6]. Conducting MA on Hill coefficient showed a consistency in EC50s estimations, and simulations taking into account model uncertainty on this parameter showed a high level of antiviral efficacy for high dose FPV regardless of MOA.
Finally, at human plasmatic concentrations, simulations showed that FPV would be able to suppress infectivity within 2 days of treatment, even when started late in the course of the disease.
Conclusions: This model describes for the first time LASV pathogenesis and clarifies RBV and FPV modes of action. However, estimating whether those two MOA could act together could not be brought to light. FPV was shown to be much more efficient than RBV and simulations using human concentrations showed a very high antiviral efficacy of FPV (>99%). Taken together, these results provide a milestone towards an eventual use in humans.
References:
[1] CDC. Estimated number of Lassa fever cases in West Africa each year Death rate from Ebola Death rate from Lassa fever. 2015
[2] World Health Organization, Lassa Fever; https://www.who.int/news-room/fact-sheets/detail/lassa-fever.
[3] Lee DU, Je SH, Yoo SJ, Kwon T, Shin JY, Byun JJ, et al. Hematological adverse effects and pharmacokinetics of ribavirin in pigs following intramuscular administration. Journal of Veterinary Pharmacology and Therapeutics. 2017;40(5):561{568. doi:10.1111/jvp.12394.
[4] Rosenke K, Feldmann H, Westover JB, Hanley PW, Martellaro C, Feldmann F, et al. Use of favipiravir to treat lassa virus infection in Macaques. Emerging Infectious Diseases. 2018;24(9):1696{1699. doi:10.3201/eid2409.180233
[5] Madelain V, Guedj J, Mentré F, Nguyen THT, Jacquot F, Oestereich L, et al. Favipiravir Pharmacokinetics in Nonhuman Primates and Insights for Future Efficacy Studies of Hemorrhagic Fever Viruses. Antimicrobial Agents and Chemotherapy. 2017;61(1). doi:10.1128/AAC.01305-16.
[6] Madelain V, Baize S, Jacquot F, Reynard S, Fizet A, Barron S, et al. Ebola viral dynamics in nonhuman primates provides insights into virus immuno-pathogenesis and antiviral strategies. Nature Communications. 2018;9(1):1{11. doi:10.1038/s41467-018-06215-z.
[7] Espy N, Nagle E, Pfeffer B, Garcia K, Chitty AJ, Wiley M, et al. T-705 induces lethal mutagenesis in Ebola and Marburg populations in macaques. Antiviral Research. 2019;170(May):104529. doi:10.1016/j.antiviral.2019.06.001.
Reference: PAGE 29 (2021) Abstr 9608 [www.page-meeting.org/?abstract=9608]
Poster: Drug/Disease Modelling - Infection