No abstract

Animal models have been developed (1-4) in which established and disseminated tumors can be eradicated by the adoptive transfer of syngeneic T cells specifically immune to tumor-associated antigens. These models have served as prototypes for understanding how the immune system of a tumor-bearing host might be manipulated to promote in vivo lysis of advanced tumors. Many factors necessary for successful immunotherapy have now been elucidated, including the need for immune effector T cells capable of persisting in the host (5, 6), the long time period required for elimination of all tumor cells by immune cells in vivo (5), and the importance of ablating suppressor cells, induced in the host by the growing tumor, which interfere with the expression of transferred immunity (3, 7). Although specific T cell immunity is necessary for tumor eradication, it has been difficult to determine which immunologic effector mechanism(s) must be operative in vivo for these transferred T cells to be effective.Cytolytic T cells (CTL) ~ can directly and immediately kill tumor cells in vitro. Adoptively transferred CTL have been shown to prevent the outgrowth of recently transplanted tumors, and to contribute to tumor elimination in some therapy settings (4,8,9); the rejection of a murine sarcoma following transfer of immune T cells has been correlated with the in vivo generation of CTL (10). Thus, CTL may be important for the elimination of established disseminated tumors. However, studies using purified functional subsets of immune T cells have suggested that transfer of the noncytolytic helper T cell subset, which presumably can induce a tumoricidal delayed-type hypersensitivity (DTH) response in vivo, is necessary and sufficient for tumor eradication (11,12). This conclusion was supported by studies (13) suggesting that a similar DTH phenom-

Studies in animal tumor models have shown that localized and disseminated tumors can be eradicated by the adoptive transfer oftumor-specific syngeneic T cells (1-3). Analyses ofthe effector T cell subsets participating in tumor rejection have revealed disparate results in different tumor models . For example, in vivo generation ofclass I-restricted tumor-specific CTL has been shown to correlate with rejection of some murine tumors (4-8), while studies utilizing purified populations ofclass II-restricted tumor-specific Th, capable of inducing delayed-type hypersensitivity (DTH)' responses in vivo, have suggested that noncytolytic class II-restricted effectors are necessary and sufficient for tumor rejection (9-11) . Moreover, in some tumor models both tumor-specific CTL and Th are required for tumor eradication (3) . The disparities observed in the T cell responses required for tumor eradication in these different tumor models could reflect unique susceptibility by the tumor to a particular effector mechanism. However, most tumors are susceptible to lysis both by CD8 + CTL and macrophages activated by lymphokines released by CD4+ Th during DTH responses (12, 13). Therefore, an alternative explanation for these apparent differences in effector requirements for mediating tumor elimination in some models is that particular tumors may preferentially induce and/or activate either class I-or class II-restricted T cell responses .Presentation of antigen for activation ofMHC-restricted T cells requires that the antigen be processed to a peptide that can bind to the appropriate MHC molecule to form an immunogenic complex. The intracellular processing pathways that generate complexes of peptides and MHC antigens are separate for class I and class II molecules (14). Presentation with class II molecules appears to require that antigen enter via an exogenous chloroquine-sensitive pathway in which antigen is endocytosed and degraded in lysozomes (15, 16), whereas presentation with class I molecules appears to require antigen enter via an endogenous pathway in which proteins produced

No abstract