Post-Transplantation Cyclophosphamide Induces Alloreactive T-Cell Functional Impairment and Depends on Foxp3⁺ Regulatory T Cells Not Selective Elimination or Intrathymic Clonal Deletion of Alloreactive T Cells

Our understanding of how post-transplantation cyclophosphamide (PTCy) prevents graft-versus-host disease (GVHD) in hematopoietic cell transplantation (HCT) is largely extrapolated from major histocompatibility complex (MHC)-matched murine skin allografting models. This current paradigm of understanding posits that PTCy preferentially eliminates alloreactive T cells and is also dependent on intrathymic clonal deletion of alloreactive T cells. Skin allografting data used to generate this model leveraged strain-specific differential expression of certain T-cell receptor (TCR) Vβs as markers of alloreactive cells.

To test the hypothesis that PTCy works via selective elimination of alloreactive T cells and investigate the biology of PTCy, we developed a murine MHC-haploidentical HCT model that directly parallels human sibling haploidentical HCT. In this model (B6C3F1→B6D2F1), the administration of 40 x 10⁶ splenocytes and 10 x 10⁶ T-cell-depleted bone marrow cells following 10.5 Gy irradiation induced universally fatal GVHD, with the Vβ6 TCR serving as a marker of alloreactive T cells. Administration of PTCy on days +3 and +4 at doses ranging from 10 to 50 mg/kg effectively prevented lethal GVHD.

In assessments of the blood, lymph nodes, spleen, and liver, PTCy 25 mg/kg did reduce proliferation of CD4⁺ T cells but not CD8⁺ T cells at day +7 post-transplant. Vβ6⁺ T cells in both the CD4⁺ and CD8⁺ subsets in all four tissue compartments persisted at or above normal donor frequencies at days +7, +21, and +200 post-transplant. Persistence of Vβ6⁺ T cells in all four tissue compartments also was demonstrated in three other murine HCT models: B6→B6D2F1 (MHC-haploidentical; day +7), C3H→B6D2F1 (MHC-disparate; days +6 and +200), and C3H→AKR (MHC-matched; day +7). Using our B6C3F1→B6D2F1 model, we also created grafts in which 8% of donor CD8⁺ T cells expressed the 2C TCR, which is specific for H2-L^d (a major histocompatibility antigen in our model). 2C TCR⁺ T cells continued to proliferate highly and markedly expand from days +3 to +7 despite administration of PTCy 25 mg/kg on days +3 and +4, disproving selective alloreactive T-cell elimination after PTCy treatment in HCT. Additionally, PTCy was equally effective in our wild-type B6C3F1→B6D2F1 model when using thymectomized or non-thymectomized recipients, negating a necessary role for the thymus in GVHD prevention by PTCy.

Compartmentalization could not explain GVHD amelioration despite persistence of alloreactive T cells as similarly normal Vβ6⁺ frequencies were found infiltrating the liver, a GVHD target organ, as were found in the blood, lymph nodes, or spleen. Indeed, 2C TCR⁺ T cells composed 70-80% of liver-infiltrating CD8⁺ T cells at day +7 in mice treated with or without PTCy. Yet, compared with vehicle-treated mice, liver-infiltrating wild-type T cells obtained at day +21 from PTCy-treated mice had slightly reduced proliferation but dramatically lower cytokine production when restimulated in vitro with DBA/2 (host parental) irradiated splenocytes. Similar results were seen when repeating these experiments using liver-infiltrating T cells that were flow cytometrically depleted of CD4⁺CD25⁺ T cells, suggesting impaired functionality of alloreactive T cells after PTCy.

CD4⁺CD25⁺Foxp3⁺ T cells did preferentially expand in all four tissue compartments by day +21 post-transplant. A one-log reduction of these cells in the donor graft using an AutoMACS had minimal impact on transplant outcomes in our model. However, a two-log reduction of Foxp3⁺ cells immediately post-transplant in vivo using Foxp3-DTR-expressing donor cells resulted in marked worsening of the clinical scores and weights, which began to diverge starting approximately 30 days post-transplant. PTCy-treated mice also were resistant to reinfusion of 40 or 120 x 10⁶ new donor splenocytes at +120-150 days, consistent with suppression being an active mechanism underlying PTCy's efficacy. Overall, our findings disprove the existing paradigm of understanding of how PTCy prevents GVHD, support our previously demonstrated role for regulatory T cells, and may have direct translational relevance towards improving outcomes for patients.