Interference with HIV infection of the first cell is essential for viral clearance in a pre-exposure prophylaxis model

Pre-exposure prophylaxis (PrEP) uses relatively weak HIV inhibition to reduce transmission between individuals. Why this approach is successful is unclear. Here we derive and experimentally validate a mathematical model for predicting infection clearance with PrEP based on the measured effect of a drug on the HIV replication ratio and number of initial infected cells. We tested the model by inhibiting low dose HIV infection with tenofovir, which reduces infection frequency per cell, and atazanavir, which reduces the cellular burst size of viable virions. Both drugs were at concentrations which allowed similar HIV replication. Reducing infection frequency dramatically increased infection clearance, while reducing burst size did not. This indicates that initial infection is vulnerable to inhibition before it infects the first cell, but not thereafter. Our model explains why PrEP is potent at drug concentrations which are ineffective against established infection, and provides a framework to test drug effectiveness for PrEP.