Thomas Bénéteau 1, Nathanaël Hozé 1, Jérémie Guedj 1
1 Infections, Antimicrobials, Modelling, Evolution (IAME) Research Centre, Unité mixte de recherche (UMR) 1137, Université Paris Cité, Institut national de la santé et de la recherche médicale (INSERM), Paris, France. (, )
Introduction: Before 2022, mpox (formerly known as `monkeypox’) was primarily circulating in endemic areas (West and Central Africa) with sporadic spillover outside these regions [1]. In May 2022, the World Health Organization (WHO) issued a first Public Health Emergency of International Concerns for the worldwide outbreak of mpox [2], and a second one in August 2024 for the upsurge of mpox in Central Africa [3]. Tecovirimat, an antiviral specific to Orthopoxviruses, notably developed preventively against the use of smallpox as a biological weapon [4], was put forward as a candidate drug for the treatment of mpox [5].
The approval of tecovirimat in 2018 by the FDA was based on human safety data and efficacy data derived from animal models under the `Animal Rule’ [6] as the antiviral demonstrated protective efficacy against Orthopoxviruses in various lethal models, including against mpox in non-human primates (NHP).
Following the 2022 outbreak, tecovirimat was used under an expanded-access protocol [5] and three randomised clinical trials (RCT) were quickly launched to investigate the efficacy of the treatment on both the time to lesion resolution and on the time to viral clearance in human infections. So far, the three randomized clinical trials concluded that the treatment was safe but failed to demonstrate significant differences on the two outcomes [7, 8].
Objectives: We re-analysed data from the mpox model in NHP [9] to investigate the link between blood viral load, tecovirimat’s blood concentration, and the formation of new lesion in order to evaluate more accurately the impact of the initiation date of the treatment on two main clinical endpoints: the time to lesion resolution and the time to viral clearance. We then used the data from the UNITY cohort to translate the results from the NHP model to human infections.
Methods: For all 24 animals, the virus was inoculated by intravenous injection with 5×10^7 PFU/mL of clade Ia mpox (Zaire 79 strain) and were treated for 4 days after inoculation for 14 days [9]. In UNITY, 446 participants were included in the positive intention-to-treat and were followed weekly for up to 28 days and were treated with 600mg of tecovirimat twice a day for 14 days.
We relied on a mechanistic framework to model the virus lifecycle and the action of the drug on the egress of viral particles with a system of ODE equations. We then used this model to build a multivariate framework that link the formation of new lesion with the infectious blood viral particles. We then simulated various scenarios in which we delayed the treatment initiation by several days to account for the extra time between symptom onset and inclusion in RCTs at which the treatment is initiated in human.
We opted for a non-linear mixed effect framework to account for individual variation. The estimation of the parameters was performed using the SAEM algorithm [10] implemented in Monolix and the simulation were conducted with Simulx (version2024).
Results: Our approach captures the rapid clearance of the virus in blood following the initiation of the treatment – 14 days reduction in the median time to viral clearance with 14 doses at 10 mg/kg – and the rebounds in case of suboptimal dosage in the monkey model. Our approach also distinguishes between the fast decline of viruses in blood (median time to viral clearance equals 28 days) and the prolonged presence of viral materials in replicative areas (resp. 44 days). We found that the formation of lesion was almost negligible when the levels of infectious viral particles in blood fall below 5 logs and exploded rapidly as the levels exceeded 7 logs. Delaying the treatment by more than a week after symptom onset provided little to no effect on both the time to viral clearance and the time to lesion resolution compared to the placebo.
Conclusions: Assuming the relationship observe in monkeys also holds in human, our approach suggests that as most participants entered the clinical trials – and received treatment – with undetectable blood viral load, their infection was already under control and thus treating with tecovirimat had little effect on the time to lesion resolution or the time to viral clearance in blood or other compartment.
References:
[1] Gessain A et al., Monkeypox, NEJM (2022)
[2] https://www.who.int/europe/news/item/23-07-2022-who-director-general-declares-the-ongoing-monkeypox-outbreak-a-public-health-event-of-international-concern
[3] https://www.who.int/news/item/14-08-2024-who-director-general-declares-mpox-outbreak-a-public-health-emergency-of-international-concern
[4] Merchlinsky M et al., The development and approval of tecoviromat (TPOXX®), the first antiviral against smallpox, Antivir. Res. (2019)
[5] Sherwat A et al., Tecovirimat and the Treatment of Monkeypox — Past, Present, and Future Considerations, NEJM (2022)
[6] Grosenbach D et al., Oral Tecovirimat for the Treatment of Smallpox, NEJM (2018)
[7] PALM007 writing group, Tecovirimat wfor Clade I MPXV Infection in the Democratic Republic of Congo, NEJM (2025)
[8] Zucker J et al., Tecovirimat for the Treatment of Mpox, NEJM (2026)
[9] Nguyen, B. T. et al. Early administration of tecovirimat shortens the time to mpox clearance in a model of human infection. PLoS Biology(2023).
[10] Kuhn E & Lavielle M, Coupling a stochastic approximation version of EM with an MCMC procedure, ESAIM probab. Stat (2004)
Reference: PAGE 34 (2026) Abstr 12214 [www.page-meeting.org/?abstract=12214]
Poster: Drug/Disease Modelling - Infection