Modelling CD4 T cell reconstitution in HIV-infected children starting antiretroviral therapy
Rollo L Hoare (1,2), Robin Callard (1,2), Joseph F Standing (1,2,3)
(1) Institute of Child Health, University College London, 30 Guilford St, London, UK; (2) Centre for Mathematics and Physics in the Life Sciences and Experimental Biology, University College London, Gower St, London, UK; (3) Great Ormond Street Hospital NHS Foundation Trust, Great Ormond St, London, UK
Objectives: Antiretroviral therapy (ART) is the standard treatment for adults and children infected with human immunodeficiency virus (HIV). HIV mainly infects CD4 T cells, causing a decline in CD4 T cell concentration. This decline leaves patients immunocompromised and hence vulnerable to opportunistic infections. ART suppresses HIV replication, reducing viral load, allowing CD4 T cells to reconstitute. This reconstitution is slow, taking between one a two years. Studying immune reconstitution in children is challenging because the rapidly developing immune system results in a three-fold decrease in expected CD4 T cell concentrations for age [1].
Methods: This work combines an adapted version of a mechanistic model previously constructed for paediatric HSCT reconstitution [2] with a model for virus dynamics [3]. The resulting model has two compartments: a CD4 concentration compartment and a viral load compartment. The thymus produces new CD4 cells, which once they enter the periphery can proliferate or die. The underlying biology of the system is then taken into account by using mathematical functions on these rates to model age dependence due to the effects of development on thymic output, loss and proliferation, and concentration dependence due to competition for resources such as cytokines and sp-MHC. Virus is produced in CD4 cells, so production depends on their concentration. Furthermore, the HIV-induced increase in loss of CD4 cells is included, proportional to viral load. We apply this model to paired longitudinal data for both CD4 concentration and viral load collected in two different studies.
Results: The final model has good descriptive and simulation properties, giving sensible biological parameters for the reconstitution. Mean CD4 concentrations are found to be much below those expected of a healthy child.
Conclusions: A mechanistic model has been successfully applied to data for the reconstitution of CD4 T cells in HIV-infected children starting ART. The model has the potential to give insight into the effects of a range of covariates, such as socio-economic factors, the ART drugs used, or the age of the patient at the start of ART.
References:
[1] S Huenecke et al., Eur J Haematol, 80(6): 532-39, 2008.
[2] Hoare et al., PAGE 22, Abstract 2676, 2013.
[3] Bonhoeffer et al., PNAS 94: 6971-6, 1997.
