NK Cell Impairment after Haploidentical Allogeneic Hematopoietic Cell Transplant in Patients Receiving Post-Transplant Cyclophosphamide (PTCy)

Abstract

Haploidentical hematopoietic cell transplantation (haplo-HCT) with post-transplant cyclophosphamide (PTCy) as graft versus host disease (GVHD) prophylaxis is frequently used for patients who do not have HLA-identical donors. However, the effect of PTCy on immune reconstitution (IR) has not been studied extensively in large prospective cohorts, despite a high incidence of early viral infections. We quantified IR in 60 patients after haplo-HCT with PTCy, mycophenolate mofetil and tacrolimus (TAC) and compared results to 34 patients with 8/8 HLA matched related or unrelated donors (MD) receiving TAC and methotrexate for GVHD prophylaxis. Both groups received reduced intensity conditioning for hematologic malignancies. Samples were prospectively collected at 1, 2, 3, 6, 9, 12, 18, 24 and 30 months after HCT and analyzed using multi-color flow cytometry to characterize distinct lymphocyte populations. Peripheral blood NK cells 1 month after PTCy were characterized using CyTOF and a panel of 35 metal-labeled monoclonal antibodies and compared to age-matched healthy donor NK cells. Recovery of all T cell subsets (CD4Tcon, CD4Treg, CD8) was significantly reduced in the PTCy cohort compared to MD early after HCT (Figure 1A, B, C). Despite delayed CD4Treg recovery, the Treg:Tcon ratio was significantly higher early after HCT in the PTCY cohort due to relatively deeper depression in CD4Tcon (0.18 vs 0.06, p<0.0001 at 1 month, 0.12 vs 0.06, p=0.0004 at 3 months, Figure 1D). Absolute NK cells were lower at 1 month after PTCy (Figure 1E), entirely due to reduced numbers of CD56^dimCD16⁺ NK cells (Figure 1G). Subsequently, CD56^dim NK cells remained significantly lower in the PTCy group until 3 months after HCT. In contrast, recovery of CD56^brightCD16⁻ NK cells was significantly increased in the PTCy cohort at 2, 3 and 6 months after HCT (p<0.01. Figure 1F). The distribution of NK subsets 1 month post-HCT was also significantly different between two cohorts due to a higher frequency of immature NK cells in the PTCY cohort (Figure 1H). CyTOF analysis performed 1 month after haplo-HCT showed that NK cells expressed higher level of activating receptor NKp46, CD95, granzyme-B and inhibitory receptor NKG2A and lower expression of CD57 compared to HD (Figure 2). Functional assays showed that PTCy NK cells displayed reduced degranulation and release of IFNγ after activation compared to HD. These functions were rescued after priming with low dose IL-15 in vitro. In summary, the effect of PTCy on IR was most pronounced at 1 month with significantly delayed recovery of CD3, CD4Tcon, CD4Treg, CD8 and CD56^dim NK cells. After PTCy, NK cells display an immature phenotype and functional impairment that is at least partly corrected by priming with low dose IL-15. Impairment of NK and T cell recovery early after PTCy-based HCT underscores the need to adopt novel strategies to overcome immune defects induced by this platform.