Coevolving JAK2V617F+ relapsed AML and donor T cells with PD-1 blockade after stem cell transplantation: an index case

Blood Adv. 2021 Aug 25;bloodadvances.2021004335. doi: 10.1182/bloodadvances.2021004335. Online ahead of print.

Abstract

Relapse of myeloproliferative neoplasms (MPN) after allogeneic hematopoietic stem cell transplantation (HSCT) is associated with poor outcomes, as therapeutic approaches to reinstate effective graft-versus-leukemia (GvL) responses remain suboptimal. Immune escape through overexpression of PD-L1 in JAK2V617F mutated MPN provides a rationale for therapeutic PD-1 blockade, and indeed clinical activity of nivolumab in relapsed MPN post-HSCT has been observed. Elucidation of the features of response following PD-1 blockade in such patients could inform of novel therapeutic concepts for how to enhance GvL. Here, we report an integrated high-dimensional analysis using single cell RNA-, TCR-, CITE- and ATAC-sequencing together with mass cytometry on peripheral blood mononuclear cells collected at 6 timepoints before, during and after transient response to PD-1 blockade from an index case of relapsed MPN following HSCT. Before nivolumab infusion, AML blasts demonstrated high expression of chemokines, and T cells were characterized by expression of interferon-response genes. This baseline inflammatory signature disappeared after nivolumab infusion. Clinical response was characterized by transient expansion of a polyclonal CD4+ T cell population and contraction of an AML subpopulation that exhibited megakaryocytic features and elevated PD-L1 expression. At relapse, the proportion of the AML subpopulation with progenitor-like features progressively increased, suggesting co-evolution of AML blasts and donor-derived T cells. We thus demonstrate how single cell technologies can provide complementary insight into cellular mechanisms underlying response to PD-1 blockade, motivating future longitudinal high-dimensional single cell studies of GvL responses in relapsed myeloid disease.