Multiple Inhibitory Receptors Are Expressed on Central Memory and Memory Stem T Cells Infiltrating the Bone Marrow of AML Patients Relapsing after Allo-HSCT

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

Patients and Methods: To this,we longitudinally analyzed the T-cell dynamics in bone marrow (BM) and peripheral blood (PB) of 32 AML patients receiving HSCT from HLA identical (HLAid, 20 pts) or HLA haploidentical (haplo, 12 pts) donors. Samples were analysed by multi-parametric flow cytometry to investigate the expression of inhibitory receptors (IRs) on CD4 and CD8 T-cell subsets defined by CD45RA, CD62L and CD95 expression, and to assess the proportion of regulatory T cells (Tregs; CD4+CD25+FoxP3+). Results were also analyzed with the BH-SNE algorithm, an unbiased computational method for the analysis of FACS data. To evaluate T-cell effector functions, the CD107a degranulation assay was performed and the production of cytokines (IL-2, IFNg and TNFa) was measured by intracellular staining. BM and PB were collected 60 days after HSCT and at relapse (median 237 days; 16 pts) or, when complete remission was maintained (CR; 16 pts), at 1 year. Samples from 8 healthy donors (HD) were used as controls.

Results:After transplant, BM and PB T cells showed a lower CD4/CD8 ratio (p<0.01) and a preferential late differentiation profile (p<0.05) when compared to HD. A higher proportion of BM Tregs was documented at relapse (p<0.01), independently from the donor source.

We next investigated the expression of several IRs as T-cell exhaustion markers. After haplo-HSCT, PD-1, CTLA-4, 2B4 and Tim-3 were significantly upregulated in BM and PB T cells at all time-points, compared to HD and independently from the clinical outcome. Conversely, after HLAid-HSCT, at the late time-point, 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). We then investigated the profile of each T-cell subset in our cohort. In the BM of HD the IR expression was confined to effector memory and effectors. While a similar IR distribution was observed in CR, at relapse, PD-1, 2B4 and Tim-3 were also upregulated in BM infiltrating central memory (p<0.01) and memory stem T cells (p<0.05). Interestingly, at relapse, leukemia expressed PD-L1 (9/9 cases) and Galectin-9 (6/9). The levels of Tim-3 on BM CD8 cells associates with that of Galectin-9 on autologous blasts (p<0.05), suggesting a preferential role for this immunomodulatory axis after HSCT. Based on phenotype similarities, the BH-SNE algorithm positioned HD samples separately from transplanted pts in bi-dimensional maps. 93 significant clusters were identified. Clusters associated with relapse after HLA-id (5) and after haplo (15) were composed of T cells expressing multiple IRs, while CR-specific clusters were diminished in IR fluorescence.

To verify whether the T-cell exhaustion phenotypic profile at relapse associates with functional impairment, we evaluated T-cell effector functions upon polyclonal stimulation. Strikingly, we observed a lower degranulation ability of CD8 cells at relapse when compared to CR (p<0.05). In two patients, selected based on samples availability, we isolated and expanded by rapid expansion protocol (REP) T cells expressing one or more IRs (IR+) or no IR (IR-). Expansion rates were high and similar in IR+ and IR- T cells (mean fold increase 624 and 781, respectively at day 21). The degranulation ability measured ex-vivo in those patients (mean 4.4% on CD8 cells) was dramatically increased upon REP expansion (95% and 88.9% for IR+ and IR-, respectively). Similarly, the frequency of IFN-g producing CD8 cells increased in IR+ and IR- cells upon REP, indicating that the T-cell dysfunction observed at relapse can be efficiently reversed. We next challenged IR+ and IR- T cells against autologous blasts. Preliminary results suggest that IR+ T cells are enriched in leukemia specificity (elimination index of 66% and 44% in IR+ and IR- cells respectively at an E/T ratio of 100:1).

Conclusions: After HSCT, the molecular signature of exhausted CD8 cells in relapsing pts includes PD-1, CTLA-4, 2B4 and Tim-3. The expression of IRs on early differentiated central memory and memory stem T cells at relapse suggests a wide, though reversible, immunological dysfunction mediated by AML relapsing blasts.