Immunologic Predictors for Clinical Responses in Patients with Myelodysplastic Syndromes Treated with Immune Checkpoint Blockade

Backgound Myelodysplastic Syndrome (MDS) is a disorder in hematopoiesis where mutations may alter the epigenetic landscape facilitating leukemogenesis. Such epigenetic alterations may contribute to the immune-dysregulation leading to ineffective anti-tumor immunity required for the immune-surveillance during cancer immune-editing process. Hypomethylating agents (HMAs) such as 5-azacytidine and 5-aza-2' deoxycytidine are standard treatment for patients with high risk MDS. Resistance to HMAs ultimately occurs in most MDS patients leading to treatment failure. The expression of immune checkpoint proteins on leukemia blasts and immune cells from MDS patients during HMA treatment was reportedly increased, and is thought to be one of acquired immune-mechanism of resistance to HMA. Therefore, the combination therapy with HMA and immune checkpoint blockade have been tried to provide additional immune-pressure to shift this equilibrium in tumor microenvironment towards tumor elimination, through the promotion of inflammation, putative neo-antigen presentation, and anti-tumor immune responses in MDS. However, there are limited data on biomarkers for response to facilitate the patient selection and beneficial combinations.

Aims This study investigates whether the immunotherapy (IMT) using immune checkpoint blockade in combination with hypomethylating agent (HMA) can restore anti-tumor T cell responses eradicating leukemic clones, and attempt to determine the immune-related biomarkers to predict effective anti-Tumor immunity in MDS.

Methods Peripheral blood samples from 55 MDS patients who receive IMT (aCTLA and/or aPD1) +/- HMA, and HMA alone were collected at pre- and post-treatments. Comprehensive immunophenotypic analysis by multiparameter flow cytometry, next generation sequencing (NGS)-based TCR sequencing, immune-related gene expression profiling by NanoString Technologies, and NGS-based targeted sequencing were performed.

Results Thirty-seven patients who received IMT+/-HMA, and 18 patients who received with HMA alone were included in the analysis. We further divided 37 patients receiving IMT+/-HMA into responder (n=21) and non-responder (n=16), and 18 patients receiving HMA into responder (n=12) and non-responder (n=6). The patients who received aCTLA-4 blockade +/- HMA (n=28) showed significant increases in activated T cells, central memory T cells, and Treg after C#1 treatment, compared to those who received aPD-1 blockade +/- HMA (n=9) or HMA alone (n=18). When we evaluated the changes of immune cell subsets post treatment in relation to clinical response, a significant increase in central memory CD8+ T cells was associated with clinical responses in patients with IMT+/-HMA but not with HMA alone. Additionally, we found that the increase in clonality after 1 cycle of treatment was shown for responders (p=0.03), in contrast there was no marked change in clonality for non-responders. Next, immune signature in tumor immune-microenvironment (TIME) associated with response during IMT were evaluated. Among the responders, the expression of genes related to processes such as the adaptive immune system, cell adhesion, host-pathogen interaction, lymphocyte activation, and T cell receptor signalling significantly increased after IMT +/- HMA treatment. Lastly, we evaluated whether non-synonymous mutations detected at pre-treatment correlated to the clinical response. Responders among patients with IMT+/- HMA (p=0.0043) but not HMA alone showed a trend toward higher burden of individual mutations than those of non-responders, suggesting the presence of potential neoantigens may drive antigen specific expansion of central memory CD8+ T cells in responders during immunotherapy.

Conclusions The results demonstrated that aCTLA4 but not aPD-1 blockade elicited significant increases in activated T cell, central memory T cell, and Treg frequencies. Significant expansion of central memory CD8+ T cells and clonal expansion of T cells were associated with clinical response in MDS patients who received IMT +/- HMA. In addition, distinct immune microenvironment signature and higher individual mutation burden were observed in responders among MDS patients who received IMT+/-HMA. Further investigation is warranted to determine the immunological and genetic factors critical for effective anti-Tumor immunity in MDS patients during immunotherapy.