In order to mount an immune response against his own disease, the patient with leukemia must have leukemic cells containing antigens not present on normal cells (tumor-associated antigens). This has been demonstrated beyond doubt in various forms of acute leukemia,1-3 though the situation is far from clear in chronic granulocytic and chronic lymphocytic leukemia. Fortunately, there is also evidence that at least some of the tumor-associated antigens are the same for any one type of leukemia (a situation analogous to the virally induced tumors in animals) so that it is possible to use leukemic cells from other patients as a method of giving the antigen(s).
It is equally clear that by the time the patient manifests overt leukemia, whatever immunological responses may be occurring, they are inadequate to control the disease. Nevertheless, animal experiments suggest that under certain circumstances manipulation of immune mechanisms can influence the course of malignant