Effects of GvHD Prophylaxis with Post-Transplant Cyclophosphamide on Immune Reconstitution and TCR Repertoire after Haploidentical and HLA-Matched Transplantation Compared to Prophylaxis with Calcineurin Inhibitor Plus Methotrexate in HLA-Matched Transplantation

Background: Post-transplant cyclophosphamide (PTCy) significantly reduces the risk of graft versus host disease and is increasingly employed in both HLA-haploidentical HSCT (haplo-PTCy) and HLA-matched transplantation (matched-PTCy). We previously reported that early T and NK cell recovery is delayed after haplo-PTCy compared with HLA-matched HCT receiving conventional prophylaxis with calcineurin inhibitor and methotrexate (matched-CNI) (Rambaldi et al, Blood Adv 2021). In the current study we examined whether delays in immune reconstitution after haplo-PTCy were also observed after matched-PTCy and whether PTCy affects recovery of a diverse T cell receptor (TCR) repertoire.

Methods: Immune reconstitution after HCT was evaluated in 210 patients: n=80 haplo-PTCy; n=60 matched-PTCy; and n=70 matched-CNI, that were transplanted between 2017-2023 at the Dana-Farber Cancer Institute. An extensive panel of markers was used to characterize immune cells in peripheral blood by flow cytometry at specific timepoints after HCT: 1 month (M), 2M, 3M, 6M and 12M. CD4 T regulatory cells (Treg), CD4 conventional T cells (Tcon) and CD8 T cells were also purified from cryopreserved peripheral blood mononuclear cells by cell sorting from 9 healthy donors (HDs), 18 haplo-PTCy patients, 20 matched-PTCy patients and 17 matched-CNI patients at 3M, 6M and 12M after HCT. TCRβ sequencing was performed using the ImmunoSEQ Assay (Adaptive Biotechnologies). Key TCR repertoire metrics, including clonality and richness, were calculated and compared between groups and time points.

Results: Median age at HCT was 62 years (range, 19-76) for haplo-HCT, 57 years (range, 21-75) for matched-PTCy and 65 years (range, 22-78) for matched-CNI. The three cohorts were similar in terms of patient/donor sex, CMV patient/donor serostatus, hematologic disease and comorbidity index. 162 patients (77%) received reduced intensity conditioning regimens and 185 patients (88%) received peripheral blood stem cells. Absolute lymphocyte counts were significantly lower after haplo-PTCy and matched PTCy compared to matched-CNI at 1 month but were similar at subsequent timepoints. Patients that received PTCy had lower CD3 T-cell numbers at 1M and 3M (median values: 71 vs 291 cells/uL at 1M, p<.00001; 236 vs 432 cells/uL at 3M, p=.0004). This was primarily due to a significant reduction in CD4Tcon and CD8 T-cells while CD4Treg were only decreased 1M post HSCT. Naïve (CR45RA+CCR7+) CD4Tcon recovery was significantly reduced for 1 year in both PTCy groups, while naïve CD4Treg and CD8 T-cells recovered sooner and were comparable with matched-CNI by 6M after HCT. TCRgd T cells and CD3+CD56+ NK-T cells were significantly decreased at all time points in the first year in patients receiving PTCy. Recovery of CD56^dimCD16+ NK cells was also delayed in both PTCy groups for 3 months (median values: 17 vs 99 cells/uL at 1M, p<.00001; 65 vs 112 cells/uL at 3M, p=.002) compared to matched-CNI.

Analysis of TCR repertoire in CD4Tcon, CD4Treg and CD8 T cells at 3, 6 and 12M revealed that TCR repertoire was consistently less diverse (Simpson clonality) in CD8 T cells compared to CD4Tcon and CD4Treg in all groups at all timepoints. In the matched-CNI group, TCR repertoire diversity and richness (Daley-smith estimator) recovered by 3 months in all T cell populations and became similar to HD values. Comparing CD8 repertoire diversity and richness in the 3 groups, Simpson clonality was significantly higher (i.e. lower TCR diversity) and Daley-smith richness was significantly lower after haplo-PTCy compared with matched-CNI at all timepoints. In contrast, CD8 TCR diversity and richness after matched-PTCy recovered and became similar to matched-CNI after 6M. Similarly, CD4Treg and CD4Tcon repertoire after haplo-PTCy was less diverse than matched-CNI, but the level of TCR diversity after matched-PTCy was similar to matched-CNI at 6 and 12M.

Conclusions: Recovery of TCRgd T cells, NK cells and NKT cells is delayed in patients receiving PTCy for GVHD prophylaxis in both haplo-mismatched and HLA-matched settings. Although numerical recovery of CD4Tcon, CD4Treg and CD8 T-cells was delayed in both haplo-PTCy and matched-PTCy groups, recovery of TCR repertoire diversity was most significantly delayed after haplo-PTCy. Recovery of TCR repertoire diversity and richness in each T cell population after matched-PTCy was faster than haplo-PTCy and more similar to matched-CNI.