I don't do modeling.  Geodynamist can answer your question.  One way to model this is to assume earth is consisted of layers with different viscosity.  If you hydrate the lithosphere throughout subduction processe, then the bottom part of the lithosphere will become less viscous and boundary layer instability will develop.  Small scale convection can gradually remove the lithosphere. 

Another way is to thicken the crust (60-80 km) then the base of the magic crust will change phase into eclogite.  The higher density will start a gravitational instability and remove a large piece of lithosphere. Some data in the Andes seems suggest such process can occur, but Not in the Himalyas nor Tibet.