It's not as simple as lethality. Generally a pathogen needs to cause some sort of symptom in the host to spread well, but the more severe the symptoms the more

It's not as simple as lethality. Generally a pathogen needs to cause some sort of symptom in the host to spread well, but the more severe the symptoms the more aggressively the person would isolate. Insofar as the vaccines mostly worked upon their initial release, you can imagine a new strain that is much more virulent (replicates significantly faster, etc); such a strain could cause a vaccinated individual who would normally be completely asymptomatic and therefore not infectious [hopefully we all accept by now that there is pre-symptomatic spread but not asymptomatic spread, if not you're behind a year on the research], to become symptomatic and therefore capable of transmitting. If that same strain were to infected an unvaccinated individual who had never been exposed to the real virus, it might cause much more severe symptoms in them. --- There's also the separate phenomenom of immune escape. The current major vaccines (mRNA and adenovirus-vector) all just make the body's own cells express (generate) the spike protein (S2 subunit), and nothing else. So specifically point mutations in the S2 subunit could allow some degree of artificial immunity evasion (it wouldn't work against those naturally infected because they were infected with whole, live-replicating virus and thus have much more diverse epitopes for their immune system to "learn" and possibly a stronger immune stimulus in the first place)