Chemotherapy-Based HLA Haploidentical Transplantation with Treg/Tcon Immunotherapy in Unfit/Elderly Leukemia Patients: Powerful Gvl Effect and Insights from Animal Models

In patients with acute leukemia (AL) at high risk of relapse, because of unfavorable cytogenetics, molecular markers and disease status, the most powerful therapy is hematopoietic stem cell transplantation. In HLA-haploidentical stem cell transplantation we showed adoptive immunotherapy with naturally occurring T regulatory cells (nTregs) followed by conventional T cells (Tcons) prevented graft-versus-host-disease (GvHD) (Di Ianni et al., Blood 2011) and was associated with low transplant-related mortality (TRM) and relapse rate (Martelli et al., Blood 2014) in patients with AL conditioned with single-dose TBI.

However, patients that are heavily pretreated (e.g., because of multiple induction chemotherapy cycles, previous autologous or allogeneic transplant, previous irradiation, infections, etc.) and elderly patients are not expected to withstand a high-intensity TBI-based conditioning regimen. For these patients with an indication to transplant we, therefore, designed a chemotherapy only-based conditioning regimen. We enrolled 24 patients with high risk AL (19 AML and 5 ALL, median age: 46, range 21-64). Twenty-two patients were transplanted in CR and 2 in relapse. For 5 patients this was the second transplant. Patients received a conditioning regimen that included thiotepa (10 mg/Kg on days -10 and -9), fludarabine (200 mg/sqm from day -10 to -7) and melphalan (80-120 mg/sqm on day -5). Fiftheen patients also received ATG (6 mg/kg on days -24 and -23). Seven patients also received cyclophosphamide (20 mg/Kg on days -8 and -7) (and no ATG). Two patients received only Treosulfan (36 gr/sqm from day -8 to -6), fludarabine and thiotepa. On day -4 patients were received an infusion of 2x10e6/kg nTregs (93% ± 8 SD FoxP3+ cells). On day 0 they received CD34+ cells (mean 9.5x10e6/kg ±3.1 SD) and Tcons (1x10e6/kg). No post-transplant pharmacologic immune suppressive GvHD prophylaxis was given. Twenty-two of the 24 patients engrafted. CD4+ and CD8+ cell counts reached 200/µL on days 56 (range 30-100) and 35 (range 20-120) respectively with a wide T-cell repertoire as analyzed by Spectratyping. Eight of the 22 engrafted patients (36%) developed ≥ grade II acute GvHD. Three of the 8 patients died, 3 are alive with cGvHD and 2 are currently under immunosuppressive therapy. The incidence of TRM was 32% (7/22). Only one patient (in relapse at the time of transplant) relapsed. At a median follow-up of 24 months (range 5-44), 62% of the patients are alive and leukemia-free. The mechanisms underlying Treg suppression of GVHD with no loss of GVL activity were clarified by our murine transplant models. NSG mice received human myeloid or lymphoblastic leukemia and HLA mismatched Tregs/Tcons. Mice that received leukemia and Tcons (without Tregs) cleared leukemia but died of GvHD. Human T cells harvested from their bone marrow, liver and gut displayed a CD8+ phenotype and mediated alloreactivity against human leukemia. Mice treated with Tregs alone died of leukemia. No human T cells were found in their bone morrow. Human CD4+ T cells were instead recovered in the liver and gut. They had retained their regulatory function as they inhibited mixed lymphocyte reaction. Mice that received human leukemia and Tcons plus Tregs were rescued from leukemia and survived without GvHD. Human T cells harvested from liver and gut were composed of CD4+/FOXP3- T cells (60%) and CD8+ T cells (40%). Purified CD4+ T cells had retained their regulatory function as they inhibited mixed lymphocyte reaction. Purified CD8+ T cells displayed no alloreactivity against human leukemia. In contrast, human T cells harvested from bone marrow were still predominantly CD8+ and still displayed potent alloreactivity against human leukemia. In conclusion, peripheral blood Tregs used for adoptive immunotherapy were predominantly CD45RO+ (and negative for the bone marrow homing receptor CXCR4) and, when infused in mice, migrated to GvHD target organs (i.e., liver and gut) where they blocked Tcons and prevented GvHD. However, they were unable to migrate to the bone marrow, thereby allowing Tcons to exert unopposed alloantigen recognition and leukemia killing.

These results demonstrate haploidentical transplantation with a chemotherapy-based conditioning regimen and Treg/Tcon immunotherapy is not only feasible in fragile patients but also exerts a powerful anti-leukemic effect. Apparently, the Treg/Tcon immunotherapy plays a key role in the GvL effect.