Mathematical models to evaluate the optimal treatment of HIV-1 infection and the response of drug-resistant variants

Recent advances in the chemotherapy of HIV infection have been very successful in delaying or even stopping the progression of disease in many patients and are responsible for the decline in HIV-related deaths. However, there are still many patients who fail to maintain suppressed viral loads, and HIV mortality does persist. Therefore, the goal of many researchers is to develop new treatments and treatment strategies to extend the utility of the currently available drugs. Means to attain this goal include developing improved ways to assess disease status and therapeutic efficacy and the development of resistance.; The models presented here explore various mathematical formulations to simulate infection of an infant and an adult. These models were evaluated to determine their relative outcomes in assigning therapeutic efficacy, to establish any differences among the models in assessing viral and T cell response to therapy, and to assess the optimal trade off between drug efficacy and its adverse effects. Results show that for models including more complexity associated with HIV infection within an individual, a higher treatment efficacy is needed for optimization. These results indicate that less detailed mathematical models may not give the most complete picture of disease course, and that increasing complexity should be explored when advances in clinical and experimental data permit.; A mathematical model to incorporate the presence of resistance mutations is also presented to explore the correlation between phenotypic resistance and duration of viral response to therapy and the possibility of a preferred sequencing of these drugs. It is found that under certain conditions the resistance phenotypes of the strains with the greatest number of mutations are the primary determinants of the total time of successful treatment. On a population basis only one model studied allows for any statistically significant differences in the total treatment times between the forward and reverse administration of two therapies. No differences are noted with variations of specific host parameters with 200 observations.