Functional immune reconstitution after allogeneic hematopoietic stem cell transplantation in myeloablative and non-myeloablative conditioned patients

Abstract

Allogeneic hematopoietic stem cell transplantation (allo-HCT) is a treatment modality for several hematological and immune-driven diseases. A conditioning regimen precedes transplantation. These comprise myeloablative (MA) conditioning consisting of chemotherapeutics often in combination with high-dose total body irradiation (TBI), while non-myeloablative (NMA) conditioning regimens use reduced dosage of chemotherapy and TBI allowing allo-HCT to patients who would otherwise not be eligible for treatment. While cellular immune reconstitution in allo-HCT patients has been well studied, differences between MA and NMA conditioned patients including the functional status of the immune system post-transplantation remains unclear. Seventy-seven patients undergoing allo-HCT were included in the main study, only including patients receiving peripheral blood stem cell grafts, no ATG treatment and no other GVHD prophylaxis than tacrolimus + methotrexate or cyclosporine + MMF + sirolimus (median age 59; IQR: 48-65). The cohort consisted of 34 patients receiving MA conditioning and 43 NMA conditioned patients. As a proxy for post-transplantation immune function, we analyzed stimulated cytokine release patterns in whole-blood samples from MA and NMA patients before and after transplantation alongside major immune cell phenotypes and T cell chimerism on day 28 after transplantation. Notably, among patients receiving grafts from peripheral blood apheresis, MA patients exhibited higher T cell counts, and elevated CD3/CD28-stimulated cytokine release compared to NMA patients. Assessment of associations between cytokine release and immune cell concentration associations indicated that variation in T cell or other immune cell concentrations between the cohorts could not account for differences in CD3/CD28-stimulated cytokine release. Meanwhile, LPS- and R848-stimulated cytokine release was associated with innate immune cell subtypes. A secondary study amongst MA conditioned patients further revealed that those who received fludarabine and treosulfan (n = 35) had higher T cell concentration and stimulated immune function compared to patients receiving more intense MA regimens (n = 15). These findings highlight the complex impact conditioning has on immune function after allo-HCT. Further analyses of T cell compartments and myeloid/lymphoid innate cells are needed to further understand the functional differences observed between conditioning groups and the potential impact on clinical outcomes.

Keywords: Bone marrow transplantation; Conditioning; Hematopoietic stem cells; Immune function; Immune stimulation; TruCulture.

Fig. 1

Study design overview. (A) Study subgroups. (B) Applied immunophenotyping assays. The primary study analyzed differences in myeloablative (MA) versus NMA conditioned patients (total n = 77) where only patients receiving peripheral blood stem cell (PBSC) grafts and either combined cyclosporin (CsA), mycophenolate mofetil (MMF), and sirolimus or combined tacrolimus (Tac) and methotrexate (MTX) as GVHD-prophylaxis, were included, while patients receiving anti-thymocyte globulin (ATG) were excluded. The secondary study featured analyses of patients receiving MA conditioning with FluTreo (fludarabine + treosulfan) versus other MA conditioning regimens (cyclophosphamide + total body irradiation, etopophos + total body irradiation or fludarabine + busulfan. Samples were taken before conditioning and around day 28 post-HCT. Created in BioRender. Ostrowski, S. (2025) https://BioRender.com/28qxbcs.

Fig. 2

Stimulated cytokine release in myeloablative (MA, n = 34) and non-myeloablative (NMA, n = 43) conditioned patients on day 28 post-HCT. Concentrations were log transformed. (A) Unstimulated (B) CD3/CD28 (T cell stimulation) (C) LPS (TLR4 stimulation) (D) R848 (TLR7/8 stimulation) (E) Poly I:C (TLR3 stimulation). Significance levels: *** p < 0.001, ** p < 0.01, * p < 0.05. Blue shaded area represents normal concentration reference ranges for stimulated cytokine release.

Fig. 3

Distribution of immune cell subset concentrations day 28 after allo-HCT between myeloablative (MA)(n = 34) and non-myeloablative (NMA) (n = 43) conditioned patients. (A) Stacked barplots of cell percentages between MA and NMA conditioning groups. (B) Boxplots of log-transformed cell concentrations. Significance levels: *** p < 0.001, ** p < 0.01, * p < 0.05. Grey rectangles represent normal reference ranges.

Fig. 4

Heatmaps of correlation analyses. Day 28 post-HCT stimulated cytokine release and immune cell types from all patients (n = 77), as well as subset cohorts of patients undergoing myeloablative (MA) (n = 34) or non-myeloablative conditioning (NMA) (n = 43). Only significant results after Bonferroni adjustment for multiple tests are shown with asterisks. Red boxes indicate stimulated cytokine release and immune cell correlation analysis results. Inter/intra correlation of cytokines and immune cells are located outside red boxes. Colored grading of intensity presented in bottom bar of each plot shows R-values where blue coloring shows negative correlation and red shows positive correlation between variables. (A) CD3/CD28 stimulation. (B) LPS stimulation. (C) R848 stimulation. Significance levels: *** p < 0.001, ** p < 0.01, * p < 0.05.

Fig. 5

Volcano plots showing -log10 p-values versus linear regression model coefficient estimates from univariate (A & C) and multivariate (B & D) modelling of associations between post-HCT stimulated cytokine release as dependent variables and immune cell subtype concentration variables as covariates. Results from both the main study cohort (n = 77, green) and patients undergoing myeloablative (MA, n = 34, red) and non-myeloablative conditioning (NMA, n = 43, blue) are presented. Multiple immune cell subtypes (CD4, CD8, NK-cells, monocytes and neutrophils) were adjusted for in the multivariate model, that also included adjustment for donor age over 30, and conditioning in linear models with all patients (n = 77). Only significant results (Bonferroni adjusted p-values correcting for 9 tests per each stimulus) and coefficient estimates >/< ±0.3) are reported with labels. Only patients transplanted with a peripheral blood stem cell graft were included and patients receiving ATG or TAC + MMF/CsA + MMF GVHD prophylaxis were excluded. Dashed red line is p = 0.05, while solid red line represents Bonferroni corrected p-value threshold (p = 0.055). (A) Univariate linear regression model results of CD3/CD28, LPS, and R848 for all patients. (B) Multivariate analyses of CD3/CD28, LPS, and R848 linear regression model results for all patients. (C) Univariate analyses of CD3/CD28, LPS, and R848 linear models results for MA and NMA patients. (D) Multivariate analyses of CD3/CD28, LPS, and R848 linear models results for MA and NMA patients.

Fig. 6

Distribution of CD4, CD8 donor/recipient chimerism percentages at day 28 after allo-HCT. (A) Myeloablative (MA) conditioned patients (n = 30). (B) Non-myeloablative (NMA) conditioned patients (n = 35).

Fig. 7

Secondary study results. Day 28 post-HCT comparison of patients undergoing the myeloablative (MA) conditionings; Combined fludarabine + treosulfan (n = 35) compared to patients receiving “other” MA regimens (n = 15): cyclophosphamide + total body irradiation (TBI), etoposide + TBI or fludarabine + busulfan. (A) Distribution of CD3/CD28-stimulated between fludarabine + treosulfan and “other” MA regimens. (B) Stacked barplots of cell percentages. (C) Immune cell concentrations between groups. Boxplots of log-transformed cell concentrations. Significance levels: *** p < 0.001, ** p < 0.01, * p < 0.05. Grey rectangles represent normal reference ranges.