I can't point to a citation but I'm sure I've seen it hypothesized that it requires both the infection, then some sort of injury to allow the virus to cross the blood brain barrier.