IV-063 Alberto Vegas

Modelling & simulation of biomarkers during anti-HIV monotherapy treatment

Vegas Rodríguez, Alberto1; Troconiz, Iñaki1,2,3; Vélez de Mendizábal, Nieves4; Feigelman, Justin4

1: Department of Pharmaceutical Sciences, School of Pharmacy and Nutrition, University of Navarra, Pamplona, Spain. 2: Navarra Institute for Health Research (IdiSNA), Pamplona, Spain. 3: Institute of Data Science and Artificial Intelligence (DATAI), University of Navarra, Pamplona, Spain. 4: Gilead Sciences, Foster City, CA, United States.

Introduction: Since its first discovery, the human immunodeficiency virus (HIV) has infected more than 70M people worldwide and 35M have died. Currently, the number of deaths per year has decreased considerably due to the fact that 75% of people with HIV are currently on therapy[1]. Although currently no cure is available, existing treatments prolong life expectation and quality and reduce transmission[2] by keeping the viral load below the limit of detection (<50 copies/mL)[3]. HIV has received considerable attention from the computational modelling area, and models with different degrees of complexity have been developed, however, as of now there is no single model capable of describing and predicting the entire dynamics of the disease[4].

Objectives: The aim of this work is to compare the dynamics of the systemic levels of virus (HIV-RNA) and CD4+T cell counts predicted using different key models 

Methods: Five models considered relevant were selected. The simplest model[5] (Model 1) comprises just three entities: virus (V), target cells (T) and infected cells (I), and served as a starting point, and was extended in stages by first including a population of long-lived infected cells [chronic cells (C)][6] (model 2), then a latent state (L) for the virus which hampers its elimination from the patient[7] (model 3)[7]. Since the immune system plays an important role as it interacts with the infection process, model 4 additionally includes humoral and cell mediated immunity[8]. Lastly, a model considering the heterogenicity in drug efficacy due to the presence of drug sanctuaries was also included since it is believed that it´s a major obstacle for eradication[9. Simulated anti-retroviral therapy affected viral infectivity or viral replication. Selected models were implemented using mrgsolve in R[10]. During simulations, systemic drug concentrations were maintained constant at levels above the reported IC50 concentration, and therapeutic efficacies of 0% (no treatment) 70%, 80%, 90%, 95% and 100% were considered.

Results: Typical patient profiles were simulated, excluding interindividual variability. Once therapy is administered, the observed decrease from HIV-RNA baseline in the case of models 1 & 2 was of 100%, concluding that eradication would be achieved when efficacies above 80% are achieved. However, this predicted benefit is in excess of real clinical datasets. In contrast, model 3 does not predict extinction, even with a therapy of 100% of effectiveness, aligning with what is observed in clinical practice. In the simulations, a persistent low-level viremia of 50 copies/mL was obtained, followed by a rebound of HIV-RNA to at least 50,000 copies/mL when therapy was removed. Regarding the T cell count, models 1, 2 & 3 reflected correctly the recovery from 200 cells/µL to stable levels between 700-1,000 cells/µL during treatment, matching clinical scenarios. Model 4 produced similar results to models 1 & 2 for viremia levels, while also including representation for elements of the immune system for a more biologically realistic representation. Regarding the model considering drug sanctuaries extinction is not conquer due to the reduced efficacy in the sanctuary site, different reductions on efficacy can lead to very different outcomes, proving also an explanation for the transient viral re-emergence (i.e. “blips”).

Conclusions: This simulation exercise helps to understand how the virus interacts with host immune cells and the impact of different model topologies and assumptions on the simulated dynamics. Of the models tested, model 3 provided the most realistic outcomes. On the drug sanctuary model even though blips were described for a period of time it failed to capture them for longer periods, decreasing their size and presence over time.

 

References:

  1. Global HIV Programme [Internet]. [cited 2023 Dec 15]. Available from: https://www.who.int/teams/global-hiv-hepatitis-and-stis-programmes/hiv/strategic-information/hiv-data-and-statistics
  2. HIV Treatment: The Basics | NIH [Internet]. [cited 2023 Dec 15]. Available from: https://hivinfo.nih.gov/understanding-hiv/fact-sheets/hiv-treatment-basics
  3. WHO. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection. Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach [Internet]. 2016 [cited 2024 Dec 15];(2016):XXXii. Available from: https://www.ncbi.nlm.nih.gov/books/NBK374294/  
  4. Gonçalves A, Mentré F, Lemenuel-Diot A, Guedj J. Model Averaging in Viral Dynamic Models. AAPS J. 2020 Mar 13;22(2):48.
  5. Nowak MA, May RM. Virus Dynamics: Mathematical Principles of Immunology and Virology. Oxford: Oxford University Press; 2000. 16–43 p.
  6. Perelson AS, Essunger P, Cao Y, Vesanen M, Hurley A, Saksela K, et al. Decay characteristics of HIV-1-infected compartments during combination therapy. Nature. 1997 May;387(6629):188–91.
  7. Hill AL. Mathematical Models of HIV Latency. In 2017. p. 131–56.
  8. Lin J, Xu R, Tian X. Threshold dynamics of an HIV-1 model with both viral and cellular infections, cell-mediated and humoral immune responses. Mathematical Biosciences and Engineering. 2019;16(1):292–319.
  9. Kepler, T. B., & Perelson, A. S. (1998). Drug concentration heterogeneity facilitates the evolution of drug resistance. Proceedings of the National Academy of Sciences of the United States of America, 95(20), 11514-11519. https://doi.org/10.1073/pnas.95.20.11514
  10. Baron K (2023). mrgsolve: Simulate from ODE-Based Models. R package version 1.3.0, https://github.com/metrumresearchgroup/mrgsolve.

Reference: PAGE 32 (2024) Abstr 10982 [www.page-meeting.org/?abstract=10982]

Poster: Drug/Disease Modelling - Infection