Original works studying CTL/memory responses to allogeneic tumor cells in mice used tremendous quantities of immunizing tumor cells (3 x 10(7) per mouse) to achieve optimal primary CTL response and successful formation of memory pool. Does it mean, that preexisting effector memory/resident memory/central memory pools with restricted proliferative potentials are sufficient for full destruction of lower immunizing doses of injected tumor cells? Does anyone know anything about exact cellular source of immediate cytotoxicity besides NK-cells? I think, this question may be highly relevant to overall biology of CD8+ CTL responses.