Anti-HTLV-I Tax-Protein-Specific CTL Is Not Necessarily Resposible for Graft-v.s.-ATL Effect in Allogeneic SCT for Patients with Adult T-Cell Leukemia/Lymphoma.

[Background] Although the anti-HTLV-I Tax protein-specific CTL (anti-Tax-CTL) have been reported to play a crucial role for the Graft-v.s.-ATL effect (GvATL) in allo-SCT for ATLL, whether or not anti-Tax-CTL is the real weapon of the GvATL still remains undetermined, because large part of primary ATLL cells do not express Tax protein via several gene alternations, and chronic GVHD seems highly correlated to the longer survival.

[Study design] To explore how significantly anti-Tax CTL contribute to GvATL, we investigated the correlation between anti-Tax-CTL and other immunological parameters after RIST without T-cell depletion (TCD) from HLA-mismatched relative donors for patients with refractory ATLL and who lack suitable donors. All 4 cases did not have KIR incompatibility in the direction of GVHD. Three cases of 4 received PBSC from HTLV-I negative donors, and survived longer than 7 months. About these long survivors, we longitudinally monitored immunological parameters including Tax-protein derived peptides-specific CD8+ T-cells by tetramer assay in accordance with HTLV-I proviral load in the periphery determined by quantitative-PCR as a surrogate marker of residual ATLL.

[Case report] A 54-year-old woman with ATLL received non-TCD RIST from HTLV-I negative and HLA-haploidentical her son (recipient;HLA-A;*0201,-, B;*54,*40*11, Cw;*0102,-,DRB1;*04,*11, donor; HLA-A;*11, *2402,B;*54,-,Cw;*0102,-,DRB1;*04,-). She achieved complete donor chimerism on day 15. By day 30 HTLV-I proviral load became undetectable (less than 0.5 copies/1x10e3 PBMC), and she achieved CR. 6 months later, her ATLL relapsed with several skin nodules. At relapse, donor-derived HLA-A*2402 restricted HTLV-I Tax _301-309-specific CTL were documented (0.11% of CD8) for the first time. Three weeks after 1x10e6/kg of CD3 cells of DLI, she suffered from GVHD (skin; stage 3), and those anti-Tax CTL further increased upto 0.38% of CD8 in number. These CTL bore cytoplasmic granzyme B, and produced IFN-γ in response to identical peptide in vitro. Interestingly those donor-derived anti-Tax CTL were unable to kill recipient ATLL cells, because recipient cells were negative for HLA-A*2402. Eventually her relapsed ATLL was well controlled mainly by single course of DLI. Additionally the other 2 of 3 long survivors (who had either HLA-A*0201 or A*2402 molecule) did not show any-anti-Tax-CTL throughout their clinical courses. One patient surviving longer than 20 months who had chronic GVHD (extended) and shared HLA-A*2402 molecule with his donor, that theoretically suggests the more feasible condition for the induction of anti-Tax CTL.

[Discussion] In the setting of HLA-mismatched SCT, our observation clearly demonstrated that anti-Tax-specific CTL unable to kill recipient ATLL cells could be induced. Alloantigens and undetermined ATLL-related tumor antigens other than HTLV-I Tax protein should be equally considered for the better understanding of GvATL and improve the outcome of allo-SCT against ATLL.