Human Epstein-Barr virus (EBV)-specific cytotoxic T lymphocytes home preferentially to and induce selective regressions of autologous EBV-induced B cell lymphoproliferations in xenografted C.B-17 scid/scid mice

C.B-17 scid/scid (severe combined immunodeficiency [SCID]) mice inoculated with peripheral blood lymphocytes from Epstein-Barr virus (EBV)-seropositive donors, or with EBV-transformed lymphoblastoid B cell lines (EBV-LCL), develop lethal human EBV(+) B cell lymphoproliferative disorders (EBV-LPD) with characteristics similar to those arising in immunodeficient patients. Using this model, we examined the capacity of human effector cells to control human EBV-LPD. SCID mice received rabbit anti-asialo GM1 antiserum to abrogate endogenous natural killer-cell function. Preliminary experiments showed that adoptive transfer of peripheral blood mononuclear cells (PBMC), purified T cells, interleukin (IL) 2-activated PBMC or anti-CD3-activated T cells derived from EBV-seropositive donors did not result in improved survival of treated mice (in vivo effector/target ratio 2:1 to 1:1). In contrast, EBV-specific cytotoxic T lymphocytes (CTL), derived from EBV-seropositive donors and expanded in vitro, exhibited strong EBV-specific and HLA-restricted activity both in vitro and in vivo. SCID mice inoculated intraperitoneally with autologous but not with HLA-mismatched EBV-LCL. had significantly improved survival relative to untreated mice after inoculation of EBV-specific CTL either intraperitoneally (P <0.001) or intravenously (P <0.001) (in vivo effector/target ratio 1:1). SCID mice bearing large subcutaneous EBV(+) tumors and treated intravenously with 10(7) EBV-specific CTL achieved complete tumor regression. Both CTL- and CTL-plus-IL-2-treated mice survived significantly longer than untreated animals or animals treated with IL-2 alone (P = 0.004 and P <0.02, respectively). SCID mice bearing two subcutaneous EBV(+) tumors, one autologous and the other HLA mismatched to the EBV-specific CTL donor, had regression of only the autologous tumor after intravenous infusion of 10(7) EBV-specific CTL. Moreover, we could demonstrate preferential homing of PKH26-labeled EBV-specific CTL to autologous but not to HLA-mismatched EBV(+) tumors as early as 24 h after intravenous adoptive transfer. Immnunophenotypic analyses also demonstrated preferential infiltration of T cells into the autologous EBV(+) tumor in SCID mice bearing both the autologous and either fully HLA-mismatched or genotypically related haplotype-sharing EBV(+) tumors. The human T cells infiltrating EBV(+) tumors were CD3(+) and, predominantly, CD8(+)CD4(-). Our results indicate that EBV-specific CTL preferentially localize to and infiltrate EBV(+) tumors bearing the appropriate HLA antigens and thereafter induce targeted regressions of disease.