An extracellular stochastic model of early HIV infection and the formulation of optimal treatment policy

The problem of scheduling optimal treatment strategies for patients at the early stage of human immunodeficiency virus (HIV) infection is investigated. Unlike patients with an established HIV infection, complete eradication of the infection is still possible at this stage and treatment can further increase the probability of eradication. However, high dosages of drugs should be avoided, if possible, because of toxic side effects. Stochastic simulation is capable of determining the infection establishment probability at the early infection stage. Consequently, to obtain acceptable treatment strategies, an optimization problem was formulated, employing a stochastic model to predict the response of an average patient to treatment. Optimal treatment strategies for prompt and also a few days latency in treatment initiation were computed. These strategies were compared with constant treatment strategies and were shown to be more beneficial in silico, i.e., they either decreased the infection establishment probability or the dosage of the drugs.