Immune reconstitution dynamics after unrelated allogeneic transplantation with post-transplant cyclophosphamide compared to classical immunosuppression with anti-thymocyte globulin: a prospective cohort study

Abstract

Post-transplant cyclophosphamide (PTCy) has contributed to the success of haploidentical hematopoietic stem cell transplantation (HSCT) and is also used in transplantation from matched donors. However, limited data on the immune reconstitution after this type of immunosuppression is available. We aimed to evaluate immune reconstitution after HSCT from unrelated donors, comparing anti-thymocyte globulin (ATG) and PTCy. Consecutive patients undergoing HSCT from unrelated donors and receiving either ATG or PTCy were prospectively included. Immune reconstitution analyses were performed by flow cytometry pre-transplant and on days 30, 60, 90, and 180 post-transplant. We included 36 patients, 20 in the ATG group and 16 in the PTCy group. In the early post-transplant period (day [d]+30), the ATG group showed a higher number of total lymphocytes, T, B, and natural killer (NK) cells compared to the PTCy group. However, at d+180, the PTCy group exhibited a higher number of B cells. On d+60 and d+90, the ATG group displayed higher number of NK cells CD56dim compared to the PTCy group, while on d+180, the PTCy group showed higher number of CD56-, CD16+, and, NKG2D+ NK cells. Naive CD4+, transition CD4+, and naive CD8+ T cells on d+60 were identified as risk factors for acute graft-versus-host disease grade 2-4, and a higher count of CD4+ memory cells on d+180 was identified as a risk factor for chronic graft-versus-host disease. In the context of unrelated allogeneic transplantation, immunosuppression with PTCy was associated with later B-, T- and NK-cell reconstitution compared to ATG.

Introduction

Different immunosuppression protocols are employed in allogeneic hematopoietic stem cell transplantation (HSCT) to mitigate the acute graft-versus-host disease (aGVHD) and chronic GVHD (cGVHD) risk. These protocols are especially used in the scenario of transplants from unrelated donors, in which classical immunosuppression with calcineurin inhibitor and methotrexate are associated with high GVHD risk.¹ Some studies have shown that the association of anti-thymocyte globulin (ATG) in transplant protocols results in reduction in grade 2-4 aGVHD and moderate to severe cGVHD.^2-5

The use of post-transplant cyclophosphamide (PTCy) has revolutionized haploidentical HSCT, with impressive outcomes related do GVHD, while it is relatively simple, cost-effective, easily reproducible, and eliminates the need for expensive T-cell depletion techniques. The PTCy strategy, when compared with haploidentical transplantation using depleted in vitro T cells, yields better outcomes related to infectious complications.^6-8 Given the encouraging outcomes observed in haploidentical HSCT, especially concerning GVHD, and the inherent assumption that PTCy could mitigate GVHD following a matched transplant, some trials are being conducted to investigate the value of PTCy in HLA-matched transplants,^9-14 with encouraging results in the context of myeloablative conditioning regimens,^15,16 or reduced-intensity conditioning.¹⁷ Poor immune reconstitution post-allogeneic HSCT is directly related to significant post-transplant morbidity and mortality, including viral, bacterial, and fungal infections, as well as relapse of the underlying disease. T-specific immunity is affected by a combination of factors, including the conditioning regimen used, thymic involution, patient age, graft type, type of T-cell depletion (in vivo or in vitro), HLA mismatches between donor and patient, immunosuppression used, and GVHD.¹⁸ In general, innate immunity recovers early after transplant, initially with the recovery of monocytes, followed by neutrophils. Natural killer (NK) are the first lymphoid cells to reconstitute and remain the predominant lymphocytes in circulation in the first 3 months post-transplant¹⁹ and its early reconstitution may be associated with transplant outcomes.²⁰ The recovery of adaptive immunity, especially T lymphocytes and B lymphocytes, is slower, potentially longer than 1 year.²¹ The reconstitution of T lymphocytes occurs through two main mechanisms: i) survival and peripheral expansion of donor-infused memory T lymphocytes, or ii) generation of donor T lymphocytes via the thymus from donor hematopoietic precursors.

However, immune reconstitution in the context of unrelated transplant, using PTCy, has still been little studied, with limited data comparing immune reconstitution with this immunosuppression strategy to the classical approach with methotrexate, calcineurin inhibitor, and ATG. The assessment of immunological profiles may allow preventive measures after transplant. Thus, our objectives were to evaluate the kinetics of immune reconstitution post-allogeneic HSCT from unrelated donors with two types of immunosuppression, ATG or PTCy. In addition, we aimed to correlate the impact of the immune reconstitution with complications in HSCT.

Methods

Study design, setting, and ethics

In this prospective cohort study, we included patients undergoing allogeneic HSCT from unrelated donors using immunosuppression with ATG or PTCy in the years 2019 to 2021 in a single private tertiary hospital (Hospital Israelita Albert Einstein). The study protocol was approved by the local ethics committee (CAAE 00527018.6.0000.0071). Informed consent for inclusion in this study was obtained from all research participants.

Participants’ eligibility and recruitment for this study

We included patients of any age undergoing their first allogeneic HSCT for hematological diseases with an unrelated donor, using an immunosuppression protocol with ATG or PTCy. The choice for one or another approach was based on the allocation in a concurrent study at the same institution, a clinical trial in which patients were randomized between ATG and PTCy (clinicaltrials gov. Identifier: NCT03818334) or the clinical choice of the bone marrow transplant team. We excluded cases in which both patient and donor had negative pre-transplant cytomegalovirus (CMV) serology. Clinical evaluations and treatments may be found in the Online Supplementary Appendix.

Immune reconstitution monitoring

Patients’ peripheral blood samples were collected pre-transplant and on day (d)+30, d+60, d+100 and d+180 post-transplant. Flow cytometric immunophenotyping for immune recovery was performed prospectively. For the analysis of mononuclear cells, an initial gate on CD45 was used. For T cells, CD3 and subsets CD3CD4 and CD3CD8 were used. For NK cells, CD3^-, CD14^-, CD11b^-, CD19^-, CD56⁺ or CD56^-, CD16⁺ and NKG2D⁺ were used, and for B cells, CD3^-, CD14^-, CD11b^-, CD19⁺. Monocytes included CD3^-, CD14⁺ and CD11b⁺. A subset analysis of CD4⁺ and CD8⁺ naive (CD45RA⁺), memory (CD45RO⁺) and transitional T cells (CD45RA^+/-CD45RO^+/-) was also performed. We also analyzed some patients data using the Cytobank platform to support and illustrate the findings of flow cytometry.²²

Outcomes

The primary endpoint in this study was B-, T-, and NK-cell immune recovery in patients undergoing unrelated allogeneic transplantation using PTCy or ATG. The secondary endpoints studied included overall survival, cumulative incidence of treatment-related mortality (TRM), 100-day mortality, aGVHD, CMV reactivation, and disease.

Statistical analysis

Categorical variables were summarized using frequencies and percentages, and continuous variables using means and standard deviations, or medians and ranges where appropriate, as well as the number of observations.

To compare categorical variables, the χ² or the Fisher’s exact tests were used, as appropriate. To compare continuous variables, the parametric Student’s t test was used, if applicable, or the non-parametric Mann-Whitney test. Overall survival was defined from the date of HSCT to the patient’s death, and was estimated using the Kaplan-Meier method, with patients alive at the time of data collection being censored. The log rank test was used for comparison of survival curves. Cumulative incidence curves (non-relapse mortality and relapse) were built with the Gray method. For relapse, non-relapse mortality was the competing event, and vice versa. For uni and multivariable analyses for survival outcomes, we used the Cox model.

Statistically significant values were defined as P≤0.05. Statistical tests were performed using the R (Vienna, Austria) statistical software, version 4.3.2.

Results

Clinical outcomes

The demographic characteristics and clinical outcomes of the included population and their donors are reported in Table 1. At 180 days post-transplant, aGVHD grade 2-4 occurred in 40% of the ATG group and 25% of the PTCy group (P=0.26). cGVHD at 3 years was 31% for ATG and 19% for PTCy (P=0.19) (Online Supplementary Figures S1, S2). In a follow-up of 3.25 years, overall survival was 59% for ATG and 74% for PTCy (P=0.4) (Online Supplementary Figure S3). CMV reactivation rates were 70% in the ATG group and 75% in the PTCy group within 180 days (P=0.98) (Online Supplementary Figure S4), with similar median viral loads (median 959 Ui/mL in ATG and 932 Ui/mL in PTCy group), P=0.81. Second CMV reactivation occurred in 20% of ATG patients and 6% of PTCy patients, with no CMV-related disease in the ATG group compared to three cases in the PTCy group. Other viral infections are described in the Online Supplementary Appendix.

Table 1.Patient characteristics and clinical outcomes.

Table 2.Lymphomononuclear profile.

Immune reconstitution

There was no difference between the ATG and PTCy groups in terms of pre-transplant lymphomononuclear cell profile (represented in Table 2; Figures 1, 2; Online Supplementary Figure S5). On d+30, in the ATG group, there was a significantly higher number of total lymphocytes (924 cells/µL vs. 473 cells/µL; P=0.0004), T lymphocytes (238 cells/µL vs. 82 cells/µL; P=0.022), B lymphocytes (17 cells/µL vs. 2 cells/µL; P=0.004), and NK cells (385 cells/µL vs. 96 cells/ uL; P<0.001) compared with the PTCy group (2 illustrative cases of patients from the ATG and PTCy groups, using the Cytobank platform, are represented in Figure 3 and the Online Supplementary Appendix). On d+180, there was a higher count of B cells in the PTCy group (54 cells/µL vs. 282 cells/uL; P=0.003).

The subtypes of T cells are represented in Table 3. A predominance of CD8 T cells was observed compared to CD4 T cells from the early stages of transplantation, persisting until d+180 in both analyzed groups. We observed a significantly higher population of double-positive CD4⁺ and CD8⁺ T cells in the PTCy group compared to the ATG group on d+180 (21 cells/µL vs. 98 cells/µL; P=0.003) (Online Supplementary Figure S6).

Among the subsets of T cells, we found no difference between the groups regarding the profile of CD4⁺ and CD8⁺ T cells. Pre-transplant, there was a predominance of CD4⁺ memory T lymphocytes and naive CD8⁺ T lymphocytes in both groups. On d+30, in both groups, there was a predominance of CD4⁺ and CD8⁺ memory T lymphocytes. On d+60 and d+90 in both groups, we observed a persistence of predominance of CD4⁺ memory T cells, but with an increase in naive and transitional CD8⁺ T cells, while on d+180, naive CD8⁺ T cells were predominant in both groups (Online Supplementary Figures S7, S8; Online Supplementary Table S1).

Figure 1.Total lymphocytes, T-, B-, and natural killer-lymphocyte reconstitution. (A) Earlier lymphocyte reconstitution was observed in the anti-thymocyte globulin (ATG) group; however, there was a greater increase in lymphocytes in the post-transplant cyclophosphamide (PTCy) group after day (d)+100. (B) There was a significant early reduction in T, B, and natural killer (NK) lymphocytes in the PTCy group on d+30, while B-lymphocyte recovery was observed on d+180. (C) In NK-cell subsets, CD56^dim, CD56 ^high and CD56^-CD16⁺ was higher in the ATG group and there was higher, and higher counts of CD56^dim on d+60 and d+90 in ATG group compared with PTCy. (D) There was an inversion of the CD4/CD8 ratio in both groups. (E) CD4⁺ T cells were predominantly represented by memory CD4⁺ cells in both groups. (F) Compared do CD4⁺ cells, there was earlier recovery of transitonal and naive CD8⁺ cells in both groups.

An analysis of NK-cell subsets was conducted based on the expression of CD56^high or CD56^dim and the expression of CD16 (Table 4). On d+30, there was a significantly higher number of all NK-cell subtypes in the ATG group. On d+60 and d+90, there was a higher number of CD56^dim cells in the ATG group. On d+180, there was a significantly higher number of CD56^- NK cells (CD3^-, CD11b^-, CD14^-, CD19^-, CD56^-, CD16⁺, NKG2D⁺) in the PTCy group (Figure 4).

Figure 2.Distribution of T, B and natural killer lymphocytes by type of prophylatic immunosupression used in patients undergoing hematopoietic stem cell transplantation: anti-thymocyte globulin or post-transplant cyclophosphamide for pre-transplant and 30, 60, 90 and 180 days. (A) There was a higher number of T, (B) natural killer (NK) lymphocytes and (C) B cells on day (d)+30 in the anti-thymocyte globulin (ATG) group compared to the post-transplant cyclophosphamide (PTCy) group. There were no statistically significant differences between the groups on d+60 and d+90; however, on d+180, there was a higher number of B lymphocytes in the PTCy group compared to the ATG group (C).

Figure 3.Cell clusters from the Cytobank platform showing differences between two patients, one who received anti-thymocyte globulin and another who received post-transplant cyclophosphamide regarding markers of B, T, natural killer cells and monocytes on the day +30. In the patient receiving anti-thymocyte globulin (ATG), there is a greater aggregate of CD56⁺ natural killer (NK) cells, a higher aggregate of CD3⁺ T cells, mainly represented by CD8⁺ cells, while in the patient who received post-transplant cyclophosphamide (PTCy), a higher population of CD14⁺ and CD11b⁺ monocytes can be visualized.

Impact of immune reconstitution on clinical outcomes

Performance status of 90 or 100 in the Karnofsky scale was a protective factor for overall survival (hazard ratio [HR]=0.3, 95% confidence interval [CI]: 0.1-0.96; P=0.042), and a higher absolute number of transitional CD4⁺ cells was associated with worse survival (HR=1.01, 95% CI: 1.01-1.04; P=0.0083).

Higher naive CD4⁺ cells (HR=1.04, 95%: CI 1-1.09; P=0.047), CD4⁺ cells transitional stage (HR=1.04, 95% CI: 1.01-1.07; P=0.012), and naive CD8⁺ cells (HR=1, 95% CI: 1-1.01; P= 0.026) on d+60 were risk factors associated with aGVHD grade 2-4. On d+180 post-transplant, a higher quantity of memory CD4⁺ cells (HR=1.01, 95% CI: 1-1.02; P=0.014) was a risk factor for cGVHD. We also analyzed the impact of aGVHD grade 2-4 in immune reconstitution and found that patients who had GVHD had lower lymphocytes count on d+60 (939 cells/µL vs. 476 cells/µL; P=0.009), lower B cells on d+90 (78.8 cells/uL vs. 19.8 cells/µL; P=0.042) and lower CD4⁺ memory T cells on d+90 (76.5 cells/µL vs. 36.8 cells/ uL; P=0.025) (Online Supplementary Table S2).

Regarding graft source, peripheral blood, compared to bone marrow, was associated with higher counts of naïve CD4⁺ T cells on d+30 (2.24 cells/µL vs. 0.44 cells/µL; P=0.03), higher counts of T cells on d+60 (523 cells/µL vs. 188 cells/ µL; P=0.03), higher CD4⁺ T cells on d+60 (39 cells/µL vs. 22 cells/µL; P=0.04), higher CD8⁺ T cells on d+60 (89 cells/µL vs. 30 cells/µL; P=0.01) and higher counts of NK cells CD56^dim

on d+60 (105 cells/µL vs. 47 cells/µL; P=0.04). We found no strong or median correlation with CD34 product counts and immune reconstitution on Spearman correlation test. We also analyzed the impact of conditioning regimen on immune reconstitution and found that reduced intensity conditioning, compared with myeloablative, was associated with higher counts of CD4⁺ naïve T cells on d+30 (3.86 cells/µL vs. 0.53 cells/µL; P=0.011), higher NK CD56^high on d+30 (287.8 cells/µL vs. 79.4 cells/µL; P=0.01) and NK CD-56^dim on d+60 (137 cells/µL vs. 52.9 cells/µL; P=0.03) while myeloablative conditioning was associated with higher NK cells CD56^-, CD16⁺ on d+90 (20.7 cells/µL vs. 9.21 cells/µL; P=0,03) and higher double-positive CD4 and CD8 T cells on d+90 (34 cells/µL vs. 7.4, cells/µL; P=0.005).

Discussion

In the present study, we prospectively analyzed patients with onco-hematological diagnoses who underwent allogeneic transplantation with unrelated donors and compared two different post-transplant immunosuppression approaches (classical immunosuppression with ATG or PTCy) according to post-transplant immune reconstitution. Early after the transplant, at d+30, the ATG group exhibited a higher number of total lymphocytes including T, B, and NK lymphocytes, as well as CD56^high, CD56^dim and CD56^-, CD16⁺,

Table 3.T-cell subtypes.

Table 4.Subsets of natural killer cells.

NKG2D⁺ NK cells, compared with the PTCy group. However, we observed a greater increase in the total lymphocyte count in the PTCy group compared to the ATG group after d+90, mainly represented by CD8⁺ T lymphocytes. Additionally, there was a statistically significant higher number of B lymphocytes and CD56^-, CD16⁺, NKG2D⁺ NK lymphocytes in the PTCy group at d+180.

Rambaldi et al.²³ also assessed immune reconstitution in patients undergoing haploidentical transplantation with PTCy or matched transplantation with classic prophylaxis. Their study described a delayed reconstitution of T cells, with a significant reduction in naive T lymphocytes, and an early decrease in NK cells following haploidentical HSCT with PTCy. Furthermore, the authors observed a preferential expansion of immature CD56^brightCD16 NK cells in the haploidentical group, compared to classic immunosuppression prophylaxis. The later reconstitution of total lymphocytes, B, T, and NK cells in the PTCy group may explain the higher grade 2 infection rate observed in the PTCy group in the phase III study comparing PTCy and classical immunosuppression in matched and mismatched transplants with a reduced-intensity conditioning regimen¹⁷ or the higher incidence of CMV reactivation in haploidentical and matched sibling transplants with PTCy compared to classical immunosuppression in sibling transplants in the retrospective CIBMTR analysis.²⁴

In the present study, at d+60 and d+90, the ATG group showed a higher number of mature NK cells (CD56^dim) compared to the PTCy group. Early-reconstituted NK cells exhibit an immature CD56^bright phenotype, representing 40-50% of NK cells in the initial months, acquiring a predominantly CD56^dim phenotype after several months, whereas, in healthy adults, CD56^bright phenotype constitute approximately 5-10%.²⁵ It is believed that NK-cell reconstitution primarily derives from the differentiation and maturation of progenitor cells rather than the expansion of mature NK cells within the graft. This idea is supported by the finding that NK-cell reconstitution is mainly independent of graft type and its NK-cell content.²⁶ The difference found in the NK maturation profile between the groups may be explained by the selective purging of dividing cells by PTCy, which has already been demonstrated by Russo et al. and may be supported by the fact that proliferating NK cells after HSCT infusion do not express aldehyde dehydrogenase and may be killed by cyclophosphamide.²⁷ Previous studies have already shown an immature NK phenotype in haploidentical transplantation in more intensive immunossupression with positive CD34 selection.^25,28

Figure 4.Distribution of natural killer cells and subtypes of CD56^high, CD56^dim and CD56^-CD16⁺ natural killer cells by type of prophylatic immunosupression used in patients undergoing hematopoietic stem cell transplantation: anti-thymocyte globulin or post-transplant cyclophosphamide for pre-transplant and 30, 60, 90 or 180 days. (A) Natural killer (Nk) cells, (B) CD56^high, (C) CD56^dim, (D) CD56^-CD16⁺ NK cells. There is a higher number of total NK cells and CD56^dim, CD56^high, and CD56^-CD16⁺ NK cells in the anti-thymocyte globulin (ATG) group on day (d)+30, a higher number of CD56^dim cells in the ATG group on d+60 and d+90, and a higher number of CD56^-CD16⁺ cells in the post-transplant cyclophosphamide (PTCy) group on d+180.

In the present study, we identified a population of NK cells characterized as CD45^pos, CD3^neg, CD11b^neg, CD14^neg, CD19^neg, CD56^neg, CD16^pos, NKG2D^pos that was higher in PTCy group on d+180. Human NK cells are generally defined as CD3^-CD56⁺ lymphocytes. However, a subset of lymphocytes, CD56^-CD16⁺ (CD56^-), displaying associations with NK markers, expands during chronic viral infections such as HIV-1 and HCV, and herpesvirus.²⁹ Many questions persist regarding the origin, development, phenotype, and function of the CD56^- NK-cell population. Gaddy J et al. have shown a subset of NK cells CD56^- CD16^- in cord blood transplants that has low lytic activity and are possible precursors of mature NK cells.³⁰ A meta-analysis evaluating 757 individuals from a total of 28 studies found that CD56^- cells constitute 5.67% of NK cells in healthy peripheral blood, while HIV-1 infection increases this population to 10.69% with no evidence of increased exhaustion or decreased proliferation within the CD56^- subset.³¹ In another study with healthy elderly donors, co-infection with CMV and EBV was associated with expansion of CD56^- NK cells with reduced cytotoxic capacity, what correlated with end-stage-differentiated T cells and reduced CD4/CD8 ratio, reflecting immune risk profile.³² The immune dysfunction after transplant and high prevalence of CMV and EBV serology in the population of the study may explain the identification of this NK cell in the present study.

The current study observed an increase in CD8⁺ T lymphocytes, leading to inversion of the CD4/CD8 ratio. CD4⁺ T lymphocytes at all evaluated time points and CD8⁺ T lymphocytes at d+30 were phenotypically represented mostly by memory lymphocytes, as previously described in patients who received PTCy immunosuppression.³³ We observed an elevation in CD8⁺ naive and transitional CD4⁺ T cells in both groups from d+90, with a predominance of CD8⁺ naive T cells at d+180 in both groups. This increase was accompanied by low counts of naive and transitional CD4⁺ T lymphocytes compared to CD4⁺ memory T cells, indicating slow reconstitution of CD4⁺ T cells.^19,34 Kanakry et al. also described that the reconstitution of CD4⁺ and CD8⁺ T lymphocytes primarily involved effector memory cells (CD45RA^-CCR7^-) and terminally differentiated cells (CD45RA⁺CCR7^-) in patients undergoing transplantation with PTCy.³³ It is known that the first “wave” of T cells post-transplant comprises donor T cells undergoing proliferation induced by lymphopenia and alloactivation, resulting in polyclonal T cells with a restricted T-cell receptor (TCR) repertoire, limited antigenic specificity or alloreactivity activity, causing GVHD.

Complete T-cell reconstitution, with TCR diversity, requires thymic regeneration and endogenous T-cell development, making this process very slow.³⁵

Besides that, in the PTCy group, we found a higher number of double-positive T lymphocytes (CD4⁺CD8⁺) at d+180. The recovery of de novo T lymphocytes is dependent on a functional thymus. The infused hematopoietic stem cells (CD34), are able to recover the bone marrow function and generate de novo lymphoid progenitors that migrate to the thymus. These progenitors, initially double-negative (without CD4 and CD8 expression), undergo a maturation process, becoming double-positive. At this stage, these cells undergo negative selection, where T cells that recognize self-antigens undergo apoptosis, an essential step for GVHD control.³⁶ If double-positive T cells do not pass through negative selection, escaping thymic selection, it may be related to the development of GVHD. Hussen et al.³⁷ observed the development of double-positive T cells during xenogeneic GVHD and Hess et al.³⁸ identified, in an observational study with 35 patients submitted to PTCy, that CD4/CD8 double-positive cells is a predictive biomarker of GVHD. In a xenogeneic model, the authors identified this population as transcriptionally, metabolically, and functionally distinct from single-positive CD4 and CD8 T cells, demonstrating that these are chronically activated CD8 T cells, highly inflammatory, secreting IFN-γ, TNF-α, GM-CSF, IL-17A, IL-22, granzyme, and perforin. These cells acquire effector functions of the CD4 lineage and are directly implicated in the pathogenesis of GVHD. As all patients in this study received PTCy, it is uncertain whether the predictive value of GVHD would also apply to the classic immunosuppression protocol. However, in our study, these lymphocytes were more frequent in the PTCy population. In our study of the correlations of clinical outcomes with immune reconstitution, at d+30, a higher absolute number of CD4⁺ transitional cells were associated with worse survival. CD4 transitional cells represent an intermediate stage between naive and memory T cells. The transition from CD45RA⁺ naive T cells to CD45RO⁺ memory T cells, initiated by TCR-mediated signaling in the G1 phase of the cell cycle, suggests that these transitional cells are in an active state of differentiation and proliferation.³⁹ This hyperactivation can lead to an increased risk of GVHD, what may be related to increased mortality risk. Additionally, the immune reconstitution involving these highly reactive transitional T cells may not provide effective protection against infections due to their immature functional state, further contributing to increased mortality. Therefore, the presence of a higher number of transitional CD4⁺ T cells may indicate an imbalance in immune homeostasis, resulting in adverse clinical outcomes post-transplantation. Besides that, higher quantity of CD4⁺ naive, transitional CD4⁺, and CD8⁺ naive cells at d+60 were risk factors for aGVHD. Additionally, a higher quantity of CD4⁺ memory cells at d+180 was a risk factor for cGVHD. In aGVHD, after activation, alloreactive memory T cells cause damage to various tissues, such as the gastrointestinal tract, liver, skin, and thymus, through direct cytotoxicity or cytokine-mediated mechanisms. However, uncertainties persist regarding whether these effector memory cells present in the patient’s blood with aGVHD are derived from the graft, proliferating upon encountering the antigen, or if they differentiate from naive cells in the graft.⁴⁰ Some studies suggest that naive T cells are even more potent inducers of aGVHD and may be associated with more severe forms, possibly due to being the most potent inducers of alloreactive responses in vitro. A possible explanation could be that naive T cells have a broader TCR repertoire, and a higher proliferative capacity, resulting in a greater potential for clonal expansion and for recognizing antigens presented by antigen-presenting cells (APC). A study evaluating the association of donor-infused CD4⁺ T cells with aGVHD found that a higher quantity of infused CD4⁺ naive T cells was associated with a higher risk of grade 2-4 aGVHD and cGVHD.⁴¹ The increased risk of GVHD associated with naive CD4⁺ T cells may explain the worse overall survival found in the present study with more transitional CD4⁺ T cells at d+60.

We found higher counts of naïve CD4⁺ T cells on d+30 and higher counts of T cells, CD4⁺ T cells, and CD8⁺ T cells on d+60 in peripheral blood compared to the bone marrow graft source. Other authors have found higher CD4⁺ T cell counts with peripheral blood, which may be explained by the higher number of lymphocytes infused, approximately one log higher with peripheral blood compared to bone marrow.^42,43 Reduced-intensity conditioning was associated with higher counts of CD4⁺ naive T cells on d+30 and higher NK CD56^high cells on d+30 and NK CD56^dim on d+60. Petersen et al.⁴³ also compared non-myeloablative and myeloablative conditioning regimens in HLA-identical sibling donor transplants and found earlier normalization of NK cells, CD8⁺ T cells, and CD4⁺ T cells but delayed B-cell recovery with less intensive conditioning.

The present study has some limitations, primarily related to the number of patients included in the analyses. However, this prospective study comprehensively analyzed the immune reconstitution of patients using ATG or PTCy, providing valuable information for clinical practice regarding this innovative form of post-transplant immunosuppression. In conclusion, in the context of unrelated allogeneic transplantation, the use of PTCy compared to classical immunosuppression with ATG is associated with a lower number of total lymphocytes, T cells, NK cells, and B cells on d+30, and a lower number of CD56^dim NK cells on d+60 and d+90, but with a higher proportion of double-positive T cells, B cells, and a higher proportion of CD56^-CD16⁺NKG2D⁺ NK cells on d+180. These results may contribute to a deeper understanding of the differences in immune reconstitution and their impact on clinical practice, which is crucial for refining transplant strategies, such as vaccinations, minimizing complications, and improving patient survival. Collaborative research efforts hold the potential to further advance our comprehension of reconstitution biology and translate these findings into customized therapeutic interventions for HSCT recipients.

Footnotes

• Received May 21, 2024
• Accepted September 3, 2024

Correspondence

^#LCM and NH contributed equally as senior authors

M. N. Kerbauy mariana.kerbauy@einstein.br

Disclosures

PB, LNK and The University of Texas MD Anderson Cancer Center have institutional financial conflict of interest with Takeda Pharmaceutical for the licensing of the technology related to CAR-NK-cell research. MD Anderson has implemented an Institutional Conflict of Interest Management and Monitoring Plan to manage and monitor the conflict of interest with respect to MDACC’s conduct of any other ongoing or future research related to this relationship. PB participates on the scientific advisory board for Dialectica.

Contributions

MNK contributed with study design and methodology, data collection and analysis, results interpretation, writing of the manuscript draft and subsequent revisions. FAR and PSC contributed with data collection and contribution to the execution of laboratory tests in flow cytometry. LJA contributed with statistical analyses of data and results interpretation. LNK, CMM and AAFR contributed with study conceptions, results interpretation and critical review of the manuscript drafts. PB contributed with Cytobank analysis. LCM and NH contributed with study conception and methodology, general supervision, results interpretation and critical review of the manuscript drafts. All authors reviewed and approved the final version submitted.

Funding

This study was funded by Amigo H.

Acknowledgments

The authors thank Patricia Logullo, PhD, CMPP (Palavra Impressa) for manuscript editing services.