POST-TRANSPLANT CYCLOPHOSPHAMIDE (PT/Cy) PRESERVES EFFECTOR (T-eff) AND REGULATORY (T-reg) T CELLS DURING EARLY IMMUNE RECONSTITUTION AFTER ALLOGENEIC BONE MARROW TRANSPLANTATION

Emerging clinical data suggest that high-dose PT/Cy is effective in preventing both acute and chronic GVHD, but the mechanisms accounting for these effects have not been fully characterized. We hypothesized that PT/Cy prevents acute GVHD (aGVHD) by selectively depleting highly proliferative alloreactive T cells stimulated early after transplant, while sparing Tregs. We analyzed the phenotypic reconstitution of CD4+ T cells at days 30 and 60 after transplant in a cohort of 47 consecutive patients (median age 50) that received HLA-matched allografts and PT/Cy as sole GVHD prophylaxis. CD4+Foxp3+ T cells were partitioned into three phenotypic fractions (Fr) with distinct functions: CD45RA+Foxp3+lo (Fr I), CD45RA-Foxp3+hi (Fr II), and CD45RA-Foxp3+lo (Fr III), representing naive Tregs, activated Tregs, and non-suppressive, cytokine-secreting Teff, respectively. In both patients with and without aGHVD, Teff (Fr III) percentages were increased compared with healthy donors (p = 0.004), and were higher in patients with aGVHD compared with patients without aGVHD (p = 0.0003). Interestingly, patients with aGVHD also had higher percentages of activated Tregs (Fr II, p = 0.01) compared with patients without aGVHD. These relative increases in Fr II and III persisted despite steroid-induced lymphopenia. The relative expansion of Tregs in patients with aGVHD was an unexpected finding, since aGVHD is associated with lower Tregs in studies using traditional GVHD prophylaxis incorporating calcineurin inhibitors (CNIs). To assess differential effects of PT/Cy and CNI on T cell reconstitution in alloreactive responses, mixed lymphocyte reactions (MLR) were performed (n = 6). High-dose, activated cyclophosphamide (mafosfamide) given 3 days after the start of MLR caused increases in Fr I-III at 7 days (Fr II, p<0.0001; Fr III, p = 0.0002), while cyclosporine-treated cells had similar to lower levels of these fractions compared with controls. The addition of mafosfamide at day 3 to cyclosporine-cultured cells in MLR caused similar increases in Fr I and III as seen with mafosfamide alone, but the Fr II increase was no longer significant. In summary, T cell reconstitution in patients treated with PT/Cy differs from that seen with CNI-based GVHD prophylaxis. For patients treated with PT/Cy who still develop aGVHD, relative expansion of Tregs may be a critical compensatory mechanism accounting for the remarkably low incidence of chronic GVHD seen in these patients. Emerging clinical data suggest that high-dose PT/Cy is effective in preventing both acute and chronic GVHD, but the mechanisms accounting for these effects have not been fully characterized. We hypothesized that PT/Cy prevents acute GVHD (aGVHD) by selectively depleting highly proliferative alloreactive T cells stimulated early after transplant, while sparing Tregs. We analyzed the phenotypic reconstitution of CD4+ T cells at days 30 and 60 after transplant in a cohort of 47 consecutive patients (median age 50) that received HLA-matched allografts and PT/Cy as sole GVHD prophylaxis. CD4+Foxp3+ T cells were partitioned into three phenotypic fractions (Fr) with distinct functions: CD45RA+Foxp3+lo (Fr I), CD45RA-Foxp3+hi (Fr II), and CD45RA-Foxp3+lo (Fr III), representing naive Tregs, activated Tregs, and non-suppressive, cytokine-secreting Teff, respectively. In both patients with and without aGHVD, Teff (Fr III) percentages were increased compared with healthy donors (p = 0.004), and were higher in patients with aGVHD compared with patients without aGVHD (p = 0.0003). Interestingly, patients with aGVHD also had higher percentages of activated Tregs (Fr II, p = 0.01) compared with patients without aGVHD. These relative increases in Fr II and III persisted despite steroid-induced lymphopenia. The relative expansion of Tregs in patients with aGVHD was an unexpected finding, since aGVHD is associated with lower Tregs in studies using traditional GVHD prophylaxis incorporating calcineurin inhibitors (CNIs). To assess differential effects of PT/Cy and CNI on T cell reconstitution in alloreactive responses, mixed lymphocyte reactions (MLR) were performed (n = 6). High-dose, activated cyclophosphamide (mafosfamide) given 3 days after the start of MLR caused increases in Fr I-III at 7 days (Fr II, p<0.0001; Fr III, p = 0.0002), while cyclosporine-treated cells had similar to lower levels of these fractions compared with controls. The addition of mafosfamide at day 3 to cyclosporine-cultured cells in MLR caused similar increases in Fr I and III as seen with mafosfamide alone, but the Fr II increase was no longer significant. In summary, T cell reconstitution in patients treated with PT/Cy differs from that seen with CNI-based GVHD prophylaxis. For patients treated with PT/Cy who still develop aGVHD, relative expansion of Tregs may be a critical compensatory mechanism accounting for the remarkably low incidence of chronic GVHD seen in these patients.