Immune Escape after Allogeneic Hematopoietic Cell Transplantation in Pediatric Patients with Acute Myeloid Leukemia

Introduction: Although allogeneic hematopoietic cell transplantation (allo-HCT) is curative for high-risk pediatric acute myeloid leukemia (AML), post-transplantation relapse occurs in 20-50% of patients and comprises the primary cause of treatment failure and mortality. To explore the effects of immune-mediated selective pressures imposed by allo-HCT on tumor evolution in relapsed pediatric AML, we performed exploratory analysis at single-cell resolution using a multimodal proteogenomic approach.

Methods: We evaluated bone marrow aspirate (BMA) samples obtained at diagnosis and at post-transplant relapse for four pediatric AML patients at Memorial Sloan Kettering Cancer Center (MSKCC). Using single-cell RNA-seq, T cell receptor (TCR)-seq, and cellular indexing of transcriptomes and epitopes (CITE)-seq, we identified several epigenetic and phenotypic alterations that may permit pediatric AML cells to escape the graft-versus-leukemia effect.

Results: A total of 79,334 single-cell transcriptomes were profiled from nine BMA samples obtained from four pediatric AML patients at diagnosis (n = 45,345) and post-transplant relapse (n = 15,412) and from one healthy 20-year-old female donor (n = 18,577). We identified several significantly downregulated immune-related processes in AML blasts at disease relapse following allo-HCT compared to initial disease diagnosis, including antigen processing and presentation, cytokine-mediated signaling, and IFN-γ mediated signaling. We observed a global downregulation of MHC class II (MHC-II) genes HLA-DR, -DP, -DQ, -DM, -DO and MHC-II regulators CIITA and CD74 when comparing mean expression at diagnosis with post-transplant relapse.

Given the significant intratumoral heterogeneity present in pediatric AML, we further evaluated the downregulation of MHC-II at relapse within subpopulations of AML blasts. AML blast cells were stratified into six subtypes along the myeloid differentiation axis: hematopoietic stem cell (HSC)-like, progenitor-like, granulocyte-monocyte progenitor (GMP)-like, promonocyte-like, monocyte-like, or cDC-like. Downregulation of MHC-II expression observed at post-transplant relapse was most profound in progenitor-like blasts and was accompanied by correlative changes in upstream MHC-II transcriptional regulation.

To elucidate the immune pressures leading to the MHC-II changes in AML blasts, detailed profiling of the leukemia-associated immune compartment was performed, including single-cell CITE-seq and TCR-seq. When compared to BMA samples obtained at diagnosis, post-transplant relapse samples exhibited clonal expansion of exhausted CD8+ T cells, suggesting recruitment of cytotoxic T cells to the bone marrow environment and lower effector functionality at the time of post-transplant relapse. In comparison to diagnosis samples, post-transplant relapse samples revealed an over-representation of regulatory T cells with a relative decrease in naïve CD4+ T cells, indicative of an immunosuppressive tumor microenvironment. Pathway enrichment validated an overall dysfunctional pattern in T cells and NK cells at post-transplant relapse with lower observed response to IFN-γ, TNF-α signaling via NF-κB, and IL-2/STAT5 signaling.

Conclusions: We observed an overall downregulation of MHC-II genes at post-transplant relapse not previously reported in pediatric AML. A dampened CD4+ T cell response, suggested by the reduced conventional CD4+ T cell population and diminished IFN-γ, TNF-α, and IL-2 response signature in NK and CD8+ T cells, could be a possible mechanism for post-transplant relapse driven by downregulation of AML MHC-II expression. We further propose that progenitor-like AML blasts may be critical to this phenomenon based on the relative expansion of progenitor-like AML clones at post-transplant relapse and further suppression of MHC-II relative to other subpopulations. The novel computational methods utilized here serve as a strong foundation to guide future single-cell studies to define the functions and mechanisms of immunomodulatory AML populations, to evaluate their relationship to the bone marrow microenvironment, and to modify their activities for therapeutic benefit.