Of course plasma will degrade its just easier to separate out the products. You could feed the plasma back into the reactor and use cyclotron resonance. Alpha and Beta decay being one of the big problems with reactor design as the walls degrade over time. So designing for that with an active system seems to me to be a way a viable solution to minimize maintenance.
Unlikely as resonance in this case is referring the minuscule difference in mass between isotopes which results in preferential orbits within a tokamak which can be exploited for separation. I guess depending on what you're separating and whether or not someone is paying attention I guess conceivably a runaway event could occur but I think the masses and densities involved are too small to be of any concern.
