Ptcy + ATG Vs Ptcy Alone As Gvhd Prophylaxis for Peripheral Blood Stem Cells Haplo-Transplants: Comparison of NK and T Cell Effector Reconstitution

Backgound: A higher incidence of graft-versus-host disease (GVHD) has been observed after haplo-transplant of hematopoietic stem cells with post-transplant cyclophosphamide (PTCY) using peripheral blood stem cells (PBSC) instead of bone marrow as source of graft. Combining PTCY with anti-thymoglobulin (ATG) may help to reduce GVHD incidence. Here, early immune reconstitution, especially of NK and T cell compartments, was compared after both types of transplant (PTCY vs PTCY+ATG) investigating their influence on patients (pts) outcomes.

Methods: This retrospective study included 58 adults who received a reduced-intensity conditioning (RIC) PBSC haplo-transplant with ciclosporine and mycophenolate mofetyl + PTCY (n=32, period: 08/2014-05-2017) or PTCY+ATG (N=26, period: 05/017-03/2019) as GVHD prophylaxis. The PTCY group received a Baltimore-based RIC regimen with fludarabine (n=10) or clofarabine (n=22). The PTCY+ATG group received a CloB2A1 regimen consisting of clofarabine 30 mg/m²/day (d), d-6 to -2, busulfan 3,4 mg/kg d-3 and d-2, ATG 2,5 mg/kg d-1. Both groups shared similar characteristics except for the median number of CD3⁺ T cells infused, significantly higher for CloB2A1 pts (27.5x10⁷/kg vs 22.3, p=0.01). Samples from all pts were collected three times a week from d0 until d30, then at d60 and d90/100 in order to evaluate NK and T cells reconstitution by flow cytometry (FCM). All pts provided informed consent for data and samples collection before graft.

Results: The median follow-up was 32 and 10,5 months for alive pts in PTCY and PTCY+ATG groups, respectively. Primary graft failure (PGF) was observed in 5 pts in the PTCY+ATG group (19%) vs none in the PTCY group (p=0.03). The PTCY and PTCY+ATG groups shared similar survivals at 1-year with PTCY OS of 71.5% vs PTCY+ATG 60,2% (p=0.73), DFS of 62,5% vs 58.6% (p=0,8) and GRFS of 46.8% vs 5.7% (p=0.47). Incidences of relapse and death were also comparable. Conversely, considering engrafted patients, the incidence of grade 2-4 acute GVHD was significantly lower in the CloB2A1 sub-group (23.8% vs 59%, p=0.02) despite higher numbers of CD3+ T cells infused in this group. The incidence of grade 3-4 acute GVHD was also lower in the CloB2A1 sub-group (4.7% vs 18.7%, p=0.29). Immune reconstitution of the PTCY group has been already reported (Willem, J Immunol 2019).

Considering the PTCY+ATG group, T cell reconstitution did not influence the occurrence of acute post-transplant GVHD. Yet, in the group of pts with myeloid diseases (n=20), a higher frequency of CD56- T cells at d20 was associated with a higher relapse incidence (39% vs 18%, p=0.04). The NK cells frequency was lower at d+30 for patients with PGF (52.5% vs 19.5%, p=0.01). Pts who did not relapse were documented with higher d30 frequency of NKG2A+ NK cells (94.4% vs 72.1% of all lymphocytes, p=0.04). Specifically, these pts showed higher NKG2A⁺ KIR2DL2/3/S2^- NK cell levels at d30 (74.1% vs 56.3%, p=0.05) and d60 (74.1% vs 55.5%, p=0.03). Inhibitory KIR/HLA incompatibilities, documented in 4 donor/recipient pairs, were associated with lower NK cell frequency at d60 (15% vs 69.2% of all lymphocytes, p= 0.01) and d100 (9% vs 61.5%, p=0.01)

Comparing the PTCY vs the PTCY+ATG groups, the frequency of CD56^- T cells was found to be significantly lower during the first month when using PTCy+ATG: d5 (34.3% vs 75.7% of all lymphocytes, p<0.0001), d20 (26.4% vs 54.8%, p<0.0001), d25 (35.9% vs 55.7%, p=0.01) and d30 (29.1% vs 48.1%, p=0.008). However, higher d5 (p=0.0001) and d30 (p=0.003) KIR2DL2/3/S2+ T cell levels were observed in this group. However, this proportion strongly decreased in comparison to the PTCY group at d100 (p>0.0001). NK cell levels were higher at d25 (34.9% vs 18.4% of all lymphocytes, p=0.03) and d30 (42.7% vs 27.9%, p=0.04) in the PTCY+ATG group. Specifically, faster reconstitutions at d20 of mature NKp46+ 2B4- (p=0.002), NKG2A+ (p=0.01) and NKG2A+ KIR2DL2/3/S2- (p=0.0004) cell subsets were found in this group.

Conclusion: These results show that PTCy+ATG vs PTCy alone as GVHD prophylaxis limits acute grade 2-4 GVHD occurrence after RIC PBSC haplo-transplant, perhaps because of the combined effect of T and NK cell effector reconstitution. Indeed, the slower T cell reconstitution with PTCy+ATG may limit GvHD occurrence. In parallel, the quicker reconstitution of some NK cells sub-types may limit relapse occurrence. Larger prospective studies are needed to confirm these preliminary results.