|dc.description.abstract||Combination antiretroviral therapy (cART) ensures that millions of people with HIV lead normal lives. However, cART is not a cure and if stopped, even after decades, HIV hidden in the latent reservoir can activate and lead to viral rebound. Given the drawbacks of cART particularly cost and difficulties of adherence to chronic treatment, HIV cures could significantly reduce the burden on patients while reducing the healthcare cost.
In 2008, the “Berlin Patient” was treated with myeloablative irradiation and hematopoietic stem cell transplant (HSCT) from a donor with a CCR5Δ32 mutation conferring resistance to HIV. Since then, the recipient has been “functionally cured”, i.e. has shown no signs of active HIV-1 replication in the absence of cART. This success renewed hope that replacing HIV-susceptible cells with more resistant cells by inserting genes or gene networks into patients’ or matched donors’ stem cells before transplantation could provide HIV-resistance to progeny target cells and lead to cure. This approach was recently shown in macaques to reduce viral load and return T cell counts to normal levels. Key questions remain: (a) given that donor chimerism occurs, what percentage of the cells must be HIV-resistant in order to block HIV; (b) what is the minimal level of anti-HIV activity needed in these cells; (c) which anti-HIV genes will work best, and for which patients; and (d) will combinations of anti-HIV genes synergize?
As few patients have undergone transplants, we built novel molecular-detailed mechanistic models of HIV infection to answer these questions. The models are validated against independent in vitro and in vivo experimental data. Using the models, we study the complex pathogenesis of HIV, design gene-augmented stem cell therapies, and calculate the probability of cure for each therapy. We focus on HSCTs that include knocking out CCR5 and/or inserting anti-HIV genes or gene circuits such as the APOBEC3 family, SAMHD1, and on-demand apoptosis-inducing circuits. Instead of studying a single average course of HIV infection in a typical patient, we apply our models, parameterized using real patient data, to simulate a population of HIV-infected patients. Using this population of models, we run virtual clinical trials of different treatments. We validate the model by predicting recent clinical data from CCR5-modified T-cell therapy. The model has the ability to help design stem cell-based therapies and predict the results in clinical studies.||