Background: Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) is the only cure for high-risk acute myeloid leukemia (AML). Unfortunately, relapse remains the major cause of death after HSCT. We investigated if T-cell dysfunction is associated to post-transplant relapse.

Patients and Methods: We longitudinally analysed T-cell dynamics in bone marrow (BM) and peripheral blood (PB) of 32 AML patients receiving HSCT from HLA-matched (20 pts) or HLA haploidentical (Haplo, 12 pts) donors. BM and PB samples were collected 60 days after HSCT and at relapse (median 237 days; 16 pts) or, when complete remission was maintained, at 1 year (CR; 16 pts). 10 healthy donors (HD) were used as controls. Samples were analysed by polychromatic flow cytometry for the expression of inhibitory receptors (IR) on CD4 and CD8 T-cell subsets defined by CD45RA, CD62L and CD95 expression. Results were also analysed with BH-SNE, an unbiased dimensionality reduction algorithm. To evaluate T-cell function and specificity, CD107a expression, cytokine profiles and killing of autologous blasts were quantified.

Results: After Haplo-HSCT, PD-1, CTLA-4, 2B4 and Tim-3 were expressed in BM and PB T cells at higher percentage when compared to HD and independently from the clinical outcome. Conversely, after HLA-matched HSCT, patients who relapsed displayed a higher frequency of BM-infiltrating T cells expressing PD-1, CTLA-4 and Tim-3 than CR pts (p<0.05) and HD (p<0.01). When investigating each T-cell subset, in both HD and CR pts the IR expression was confined to effector memory and effectors; at relapse, instead, PD-1, 2B4, KLRG1 and Tim-3 were more commonly expressed in BM-infiltrating central memory (TCM) and memory stem T cells (TSCM, p<0.01). Interestingly, at relapse leukemia expressed PD-L1 (9/9 cases) and Galectin-9 (6/9) and the latter linearly correlated with Tim-3+ BM CD8 T-cell percentages, suggesting a preferential role for this immunomodulatory axis after HSCT. Based on phenotype similarities, the BH-SNE algorithm separated HD and HSCT samples in bi-dimensional maps and identified T cell clusters that associated with relapse (32) and were characterized by T cells co-expressing high levels of multiple IRs.

Consistent with the exhausted phenotype, BM CD8 T-cells at relapse displayed lower degranulation ability and IL-2 production compared to CR (p<0.05). The ability to degranulate and to produce pro-inflammatory cytokines of BM-T cells expressing at least 1 inhibitory receptor (IR+) could be recovered in five patients upon polyclonal stimulation in vitro with high doses of IL-2, suggesting that T-cell dysfunction can be efficiently reversed. Upon challenge with autologous leukemic blasts, higher proportion of IR+ lymphocytes upregulated HLA-DR (p<0.05), granzyme A and B (p<0.001) and killed blasts (p<0.05) compared to IR- T cells. In addition, IR+ but not IR- cells were efficiently expanded by stimulation with matched Leukemic Antigen-Presenting Cells (fold increase 146 and 3, respectively).

Interestingly, we observed that, at day 60 after HLA-matched HSCT, BM CD8 TSCM and TCM cells from patients who shall subsequently relapse already expressed PD-1, Tim-3 and LAG3 at higher percentage when compared to patients who shall maintain long-term CR (p<0.05). Co-expression of at least 3 IR on TSCM and TCM T-cells, but not on effectors, or TEM, at this early time-point was associated with significantly decreased disease free survival.

Conclusions: After HSCT, the molecular signature of exhausted CD8 cells in relapsing patients includes PD-1, CTLA-4, 2B4 and Tim-3, expressed on early differentiated T cells. Exhausted BM-T cells are preferentially enriched in specificities directed against leukemic blasts. Taken together, these results highlight a wide, though reversible, immunological dysfunction mediated by AML relapsing blasts. Interestingly, the analysis on early time points suggests that the exhausted phenotype may be predictive of leukemia relapse.

Disclosures

Bondanza: TxCell SA: Research Funding; Menarini Biotech: Consultancy; Molmed Spa: Research Funding. Bonini: MolMed: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution