Adrien Mitard de Girardier 1,2, Mathilde Galhaut 3, Benoit Delache 3, Vanessa Contreras 3, Roger Le Grand 3, Mélanie Prague 2,4, Jérémie Guedj 1
1 Université Paris Cité, IAME, INSERM F-75018 (Paris, France), 2 Inria Bordeaux Sud-Ouest, Inserm, Bordeaux Population Health Research Center, SISTM Team, UMR 1219, University of Bordeaux (Bordeaux, France), 3 Université Paris-Saclay, Inserm, CEA, Center for Immunology of Viral, Auto-immune, Hematological and Bacterial diseases (IMVA-HB/IDMIT) (Fontenay-aux-Roses & Le Kremlin-Bicêtre, France), 4 Vaccine Research Institute (Créteil, France)
Introduction:
Pre-clinical modeling of viral dynamics has provided key insights during the covid-19 pandemic (specificities of each variant, treatment efficacy against viral replication, vaccine protection…) [1, 2, 3, 4]. However, one remaining question, common to all viral infections, is the relationship between viral load and symptom dynamics. As non-human primates only developed asymptomatic forms of SARS-CoV-2, it could not be established. Fortunately, it is not the case with influenza. We decided to focus on fever as it is one of the most common symptoms and temperature is more reliable than commonly used symptom scores [5]. Moreover, temperature quasi-continuous measurements could help address the identifiability issues inherent to viral load modeling that greatly limit mechanistic modelling. In fact, if no NHP Influenza viral load model has been published, those fitted to human data were shown to be missing at least one aspect of the immune response [5, 6, 7, 8,]. As a consequence, the characteristic double peak of influenza could not be captured [6].
In addition, fever is not only a symptom but an integral part of the immune response [9]. For example, the use of antipyretic drugs to diminish fever leads to an increase in mortality in humans ongoing an influenza infection [10]. Is this protection linked to a lower viral replication at high temperatures?
Objectives:
• Model jointly fever and viral load
• Investigate the possible mitigating effect of fever on viral replication.
• With a simulation work, identify viral trajectories likely to cause severe symptomatic diseases. Is severity associated with a high viral peak or rather to a difficulty to clear the virus?
Methods:
Our study includes 42 NHP challenged with the H1N1 California 09 pandemic strain which is substantial in a pre-clinical context. Inoculum vary from 104 to 108 PFU. Most animals are naïve but six have been previously challenged with H3N2. Viral load is measured daily and temperature is given every hour through an implanted chip. In addition, lung lesions were also monitored for some primates.
A latent interferon compartment, known to impact both viral replication and temperature was used to link the two biomarkers. Temperature was modelled with a sinusoidal function to capture its circadian rhythm. Fever was modelled as a rise from baseline temperature thanks to an Emax function depending on interferon level (produced by infected cells).
Different structural and statistical models were compared on the BICc criterion, using Monolix. Starting from a target cell limited model, we tested different mechanisms of action on viral replication for the innate and adaptative immune responses.
Results:
Despite biological causal knowledge and visual adequation of viral and temperature trajectories among some animals, correlations between viral and thermal endpoints were found not significant. Double fever peaks were observed in 22% of the animals when only 7% developed virtually no fever (less than a 0.5°C rise for less than a day).
The final model included a reduction of the viral production due to the interferons (innate response) and a quicker elimination of infected cells when effector cell count increases (adaptative response).
Interferon and effector cell compartments were produced proportionally to the infected cells with respective delays τ_F = 0.3 day (95% CI: [0.2 – 0.4]) and τ_E = 2.6 day (95% CI: [1.9 – 3.5]). Individual parameters suggest that higher inoculum induces a quicker innate response (p=0.008) but delays the adaptative one (p=0.02).
Conclusion:
Through the inclusion of innate and adaptative immune response, we have developed the first model describing longitudinally viral load and temperature during a viral infection. As it is routinely followed in most pre-clinical studies, we recommend to include temperature in models to increase identifiability in addition to gaining insights on physiopathology. In fact, its inclusion allowed us to estimate a viral load model with more degrees of freedom than the ones published using only viral data.
References:
1: Goyal et al., “Modeling explains prolonged SARS-CoV-2 nasal shedding relative to lung shedding in remdesivir-treated rhesus macaques”, eLife, 2021.
2: Alexandre et al., “Modelling the response to vaccine in non-human primates to define SARS-CoV-2 mechanistic correlates of protection”, eLife, 2022.
3: Marc et al., “Impact of variants of concern on SARS-CoV-2 viral dynamics in non-human primates”, PLOS Comp Biology, 2023.
4: Mitard de Girardier et al., “Exposure history shapes SARS-CoV-2 viral dynamics in Non-Human Primates and provides insights into correlates of protection against infection and transmission”, Phil Tans Roy Soc B, 2025
5: Canini et Carrat, “Population modeling of influenza A/H1N1 virus kinetics and symptom dynamics”, J. Virology, 2011.
6: Baccam et al., « Kinetics of influenza A virus infection in humans », J. Virology, 2006.
7: Handel et al., “Towards a quantitative understanding of the within-host dynamics of influenza A infections”, J. R. Soc. Interface, 2010.
8: Dobrovolny et al., “Assessing Mathematical Models of Influenza Infections Using Features of the Immune Response”, PLOS One, 2013
9: Evans et al., “Fever and the thermal regulation of immunity: the immune system feels the heat”, Nat Rv Immunol., 2016.
10: Schulman et al., “The effect of antipyretic therapy upon outcomes in critically ill patients: a randomized, prospective study”, Surg Infect, 2005
Reference: PAGE 34 (2026) Abstr 12049 [www.page-meeting.org/?abstract=12049]
Poster: Oral: Drug/Disease Modelling - Other Topics