Cellular features of graft vs host disease prophylaxis impact on CD4 T cell reconstitution post allo-HCT

Abstract Introduction: Immune reconstitution after allogeneic hematopoietic cell transplantation (allo-HCT) is required for protection from infection and response to vaccination. The phase 3 BMT-CTN 1703 trial comparing Graft versus Host Disease (GVHD) prophylaxis regimens suggested that post-transplant cyclophosphamide (PTCy) significantly improved GVHD-free relapsed-free survival without a significant impact on overall survival. Its companion study (BMT CTN 1801) showed that PTCy resulted in an early reduction of T-Cell Receptor (TCR) repertoire diversity that persisted for at least 2 years. However, this important finding was observed amongst total CD3+ T cells. We therefore asked whether the restricted clonality and clonal behavior of T cells over time was driven by reduced naive cell frequency or whether clonality defects were intrinsic to the transplant protocol and persisted into CD4 T cells after antigen exposure. Method: To capture features of the emerging immune systems between conventional GVHD prophylaxis regimens (Tac/MTX or Tac/MMF) used after matched unrelated donor (MUD) transplant versus tacrolimus with post-transplant cyclophosphamide (Tac/PTCy) used as GVHD prophylaxis after haploidentical HCT, we performed deep immunophenotyping on a cohort of 10 adults who received allo-HCT from peripheral blood stem cells (PBSCs). Of the 10 individuals, 6 received Tac/MTX or Tac/MMF and 4 received Tac/PTCy. Acute and chronic GVHD were diagnosed in 6 of 10 individuals, distributed across both conventional and PTCy prophylaxis groups. Peripheral blood mononuclear cells (PBMCs) were drawn at 1, 2, 3, 6, and 12 months post-transplant. PBMCs from 8 age-approximated healthy individuals were included as normal immune controls. PBMCs were analyzed with a high-parameter panel on a Cytek spectral flow cytometer, and viable lymphocytes were sorted for downstream 10x 5' scRNA-seq and immunoreceptor sequencing. Result: Broad measures of immune reconstitution in our analysis recapitulated prior studies, suggesting that PTCy is associated with early and sustained reduction of naive CD4 T cells through 1yr post-transplant as well as reduced clonal diversity assessed in both single cell and bulk TCR sequencing (TCRseq). Instead of naive cells, individuals who received haplo/PTCy HCT were more likely to expand effector memory CD4 T cells. Due to the single cell nature of our data, we were able to ask if the reduced clonal diversity observed after PTCy might be driven solely by a loss of naive CD4 T cells, which are the least clonally expanded and most clonally diverse T cell population in blood. In the analysis of non-naive CD4 T cells, our scTCRseq data indicate similar TCR diversity in haplo/PTCy and MUD/MTX HCT recipients once cells encounter antigen. We next used the single cell resolution in our data to understand whether clonal plasticity – the ability of a clone to be found across multiple CD4 T cell phenotypes – was observed after transplant and if differences exist between haplo/PTCy vs MUD/conventional prophylaxis. We performed clonal plasticity analysis of the top 50 clones per person across all timepoints. We observed significantly higher clonal plasticity rates in haplo/PTCy recipients. The most plastic clones were observed as a dynamic transition between central or effector memory subsets. Additionally, clonal interconversion was observed among central, effector memory and CD4 T cells re-expressing CD45RA cell subsets. Conversely, MUD/conventional HCT recipients demonstrated limited clonal plasticity, including clonal interconvention over time, consistent with findings in healthy adults. Moreover, the plasticity rate was independent of GVHD development, suggesting that early transplant decisions are the driving force for clonal behavior of reconstituting CD4 T cells over time. Conclusion: These data suggest that the clonal bottleneck observed after PTCy may be due entirely to the loss of naive T cells rather than a fundamental shift in the clonal diversity of effector and memory T cells in circulation. Our data further suggest that the phenotype of post-haplo/PTCy non-naive cells–although similar in clonality to MUD/MTX–is biased towards the effector memory state, with increased rates of clonal conversion between the dominant CD4 T cell memory phenotypes. These findings may underpin infectious risk after allo-HCT with PTCy and suggest that CD4 T cell reconstitution is substantially altered by early transplant decisions.