Guillaume Lingas
Université de Paris, IAME, INSERM, F-75018 Paris, France
Context: In the last two decades, modelling has largely contributed to unravel the determinants of host/pathogen interaction of viral infections, identify the most vulnerable populations and optimize treatment strategies [1–3]. During a pandemic, rapidly identifying the key parameters of viral evolution can provide elements for understanding the natural history of the disease, and answer critical questions such as the pertinence of certain treatment strategies or provide a target efficacy of antiviral treatments required to interrupt disease progression.
As both a medical and PhD student in a group whose expertise include both pharmacometrics and clinical research, I used mixed-effects viral dynamics modelling to unravel the host-pathogen-treatment interactions in hospitalized patients. Our group led two major projects that provided virological data that allowed for a description the natural viral evolution of the disease, a quantification of the impact of viral dynamics on disease evolution and provided a clearer picture of the impact of antiviral treatments in hospitalized patients.
To obtain an accurate description of populational viral dynamics and evaluate viral burden on disease severity, we used nasopharyngeal swabs from the first hospitalized patients at the beginning of the pandemic, included in the multicentric French Covid Cohort (N=655), among which 284 patients had repeated samples throughout hospitalization.
In parallel, the European Discovery trial evaluated during 2020 a number of repurposed drugs according to the initial WHO recommendations, that were subsect to controversy. In particular, the antivirals treatments utility was criticized considering that upon hospital admission, the dysregulated immune response or cytokine storm was considered as the main driver of clinical degradation [4]. Among them, remdesivir showed some contradictory results across studies regarding clinical efficacy in hospitalized patients [5–7] but did show a clinical efficacy in outpatients treated within 4 days after symptom onset [8]. Moreover, several studies did not find any impact of remdesivir on viral load levels [5,9,10]. However, the statistical power to identify this virological efficacy, being often a perquisite for clinical efficacy, was greatly impaired by a limited number of patients or by omitting several sources of variabilities in the time course, such as incubation period or delay between symptom onset and treatment initiation. Using the knowledge acquired previously on the natural viral evolution, we evaluated the impact of remdesivir on viral dynamics, by modelling centralized and normalized viral load data collected sequentially from hospitalized patients treated with either Standard of Care (N=329) or Standard of Care + Remdesivir (N=336).
Objectives:
1. Describe the viral course in hospitalized patients from infection until viral clearance
2. Explore the link between viral dynamics and mortality and anticipate the impact of potent antiviral treatments reducing viral load on mortality rates in the most at-risk patients.
3. To evaluate the antiviral effect of remdesivir in hospitalized patients
Methods:
We built several within-host models of the infection to reconstruct the viral course starting from the infection. Each model incorporated a different immune response impacting viral dynamics, and we selected the one best describing our data. Clinical covariates were then tested on the viral dynamic parameters of this selected model.
We used a non-linear joint modelling approach to provide unbiased estimations of our parameters and estimated the association between viral dynamics and death. Relying on this model, we could then explore the impact of lowering viral burden on mortality rates, by performing population simulations of patients treated upon hospital admission with antiviral treatments that would block viral production at several levels of efficacy close the pharmacodynamic target [11].
We incorporated remdesivir effect as a parameter inhibiting viral production. Due to the latency required to reach its active form, we further added a pharmacological delay, ranging from 0 to 5 days after treatment initiation, and performed model averaging on those models. We stratified our population according to viral load at randomization, in order to identify the category of individuals most likely to respond to treatment.
Results: The model best describing the data incorporated an immune response proportional to the quantity of infected cells, that exerted a cytotoxic effect on the cells producing viruses. We used this model to reconstruct the whole viral course from the infection until viral clearance. We identified older age (being 65 or older) to be associated with a diminution of the immune response, leading to an extended time of viral clearance, going from 13 days in patients < 65 to 16 days in patients ≥ 65. We estimated the incubation period between infection and onset of symptoms to be 5 days, with viral load peak occurring on average 1 day before symptom onset.
Using joint modelling and adjusting on the clinical risk factors of mortality, we identified viral dynamics as an independent predictor of disease progression, with a strong relation with survival (HR=1.31, P<10-4). Through simulations, we predicted a relative 41% reduction of mortality rate in the population most at risk (being 65 or older and having at least an additional risk factor), when treated with a treatment blocking viral production at 99% [12].
Finally, using data from both treated and untreated hospitalized patients from the Discovery trial, we estimated a weak remdesivir-induced reduction of viral production by 52% (%95CI:35-69). Simulations with this treatment effect predicted a diminution of only 0.7 (0-1.3) day in terms of time to viral clearance. We found that this effect rose to 80% (%95CI:65-96) in patients whose viral load at admission was higher than the threshold of infectiosity (3.5 log10 copies/104 cells), leading to a time of viral clearance shortened by 2.4 (IQR: 0.9-4.5) days in comparison to untreated patients [13].
Conclusions:
We first developed a semi-mechanistic within-host model of SARS-CoV-2 infection that allowed us to identify patients 65 or older to have a longer viral shedding. Symptom onset was estimated to occur 5 days after infection and prdicted peak viral load to coincide with symptom onset, a key factor for asymptomatic transmission [14]. We then showed that viral load is an independent predictor of mortality, and that a potent treatment whose effect reaches the pharmacodynamic target could eventually reduce mortality rates, especially in the most-at-risk populations. Finally, we estimated that remdesivir had a significant but modest antiviral effect, that was increased in patients with a high viral load at admission.
References:
[1] Li C-C, Wang L, Eng H-L, et al. Correlation of Pandemic (H1N1) 2009 Viral Load with Disease Severity and Prolonged Viral Shedding in Children. Emerg Infect Dis [Internet]. Centers for Disease Control and Prevention; 2010 [cited 2021 Oct 15]; 16(8):1265. Available from: /labs/pmc/articles/PMC3298297/
[2] Perelson AS, Neumann AU, Markowitz M, Leonard JM, Ho DD. HIV-1 Dynamics in Vivo: Virion Clearance Rate, Infected Cell Life-Span, and Viral Generation Time. Science (80- ) [Internet]. American Association for the Advancement of Science ; 1996 [cited 2021 Oct 15]; 271(5255):1582–1586. Available from: https://www.science.org/doi/abs/10.1126/science.271.5255.1582
[3] Neumann AU, Lam NP, Dahari H, et al. Hepatitis C Viral Dynamics in Vivo and the Antiviral Efficacy of Interferon-α Therapy. Science (80- ) [Internet]. American Association for the Advancement of Science; 1998 [cited 2021 Oct 15]; 282(5386):103–107. Available from: https://www.science.org/doi/abs/10.1126/science.282.5386.103
[4] Valle DM Del, Kim-Schulze S, Huang HH, et al. An inflammatory cytokine signature predicts COVID-19 severity and survival. Nat Med [Internet]. Springer US; 2020; 26(10):1636–1643. Available from: http://dx.doi.org/10.1038/s41591-020-1051-9
[5] Wang Y, Zhang D, Du G, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. Lancet. Lancet Publishing Group; 2020; 395(10236):1569–1578.
[6] Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for the Treatment of Covid-19 — Final Report. N Engl J Med. Massachusetts Medical Society; 2020; 383(19):1813–1826.
[7] WHO Solidarity Trial Consortium. Repurposed Antiviral Drugs for Covid-19 — Interim WHO Solidarity Trial Results. N Engl J Med. Massachusetts Medical Society; 2021; 384(6):497–511.
[8] Gottlieb RL, Vaca CE, Paredes R, et al. Early Remdesivir to Prevent Progression to Severe Covid-19 in Outpatients. N Engl J Med [Internet]. 2021; :1–11. Available from: http://www.ncbi.nlm.nih.gov/pubmed/34937145
[9] Barratt-Due A, Olsen IC, Nezvalova-Henriksen K, et al. Evaluation of the Effects of Remdesivir and Hydroxychloroquine on Viral Clearance in COVID-19?: A Randomized Trial. Ann Intern Med [Internet]. 2021 [cited 2021 Oct 6]; 174(9):1261–1269. Available from: http://www.ncbi.nlm.nih.gov/pubmed/34251903
[10] Ader F, Bouscambert-Duchamp M, Hites M, et al. Remdesivir plus standard of care versus standard of care alone for the treatment of patients admitted to hospital with COVID-19 (DisCoVeRy): a phase 3, randomised, controlled, open-label trial. Lancet Infect Dis [Internet]. Elsevier; 2021; 0(0). Available from: http://www.thelancet.com/article/S1473309921004850/fulltext
[11] Gonçalves A, Bertrand J, Ke R, et al. Timing of Antiviral Treatment Initiation is Critical to Reduce SARS-CoV-2 Viral Load. CPT Pharmacometrics Syst Pharmacol. American Society for Clinical Pharmacology and Therapeutics; 2020; 9(9):509–514.
[12] Néant N, Lingas G, Hingrat Q Le, et al. Modeling SARS-CoV-2 viral kinetics and association with mortality in hospitalized patients from the French COVID cohort. Proc Natl Acad Sci [Internet]. 2021; 118(8):e2017962118. Available from: http://www.pnas.org/content/118/8/e2017962118.abstract
[13] Lingas G, Néant N, Gaymard A, et al. Effect of remdesivir on viral dynamics in COVID-19 hospitalized patients: a modelling analysis of the randomized, controlled, open-label DisCoVeRy trial. J Antimicrob Chemother. England; 2022; .
[14] Killingley B, Mann AJ, Kalinova M, et al. Safety, tolerability and viral kinetics during SARS-CoV-2 human challenge in young adults. Nat Med [Internet]. 2022; . Available from: https://doi.org/10.1038/s41591-022-01780-9
Reference: PAGE 30 (2022) Abstr 10200 [www.page-meeting.org/?abstract=10200]
Poster: Drug/Disease Modelling - Infection