| Abstract | While on antiretroviral treatment (ART) for HIV, an infected individual's viral load remains non-zero, though it is very low and undetectable using standard equipment. Occasionally, regular blood tests show viral blips: very short periods of detectable viral load. Interestingly these viral blips have been shown not to be associated with patient or demographic variables, and to be only marginally associated with reported episodes of nonadherence to treatment. Further, there is evidence that the virus is closely related to the pre-treatment virus, suggesting that the low viral load is not due to ongoing viral replication, since HIV replication is a highly error-prone process. We will present a stochastic model that shows that this very low viral load can be explained principally by the activation of cells in the latent reservoir, seeded before the initiation of treatment, and that viral blips represent deviations from the mean. The model is calibrated using patient data. We will discuss the implications that a central role of the latent reservoir would have on the emergence of drug resistance. Our results reinforce the notion that the latent reservoir is an important factor to be considered in the design of therapies for HIV. |
|---|
