Abstract
Acute myeloid leukemia (AML) is the most common acute leukemia amongst adults with a 5-year overall survival lower than 30%. Emerging evidence suggest that immune alterations favor leukemogenesis and/or AML relapse thereby negatively impacting disease outcome. Over the last years myeloid derived suppressor cells (MDSCs) have been gaining momentum in the field of cancer research. MDSCs are a heterogeneous cell population morphologically resembling either monocytes or granulocytes and sharing some key features including myeloid origin, aberrant (immature) phenotype, and immunosuppressive activity. Increasing evidence suggests that accumulating MDSCs are involved in hampering anti-tumor immune responses and immune-based therapies. Here, we demonstrate increased frequencies of CD14(+) monocytic MDSCs in newly diagnosed AML that co-express CD33 but lack HLA-DR (HLA-DR(lo)). AML-blasts induce HLA-DR(lo) cells from healthy donor-derived monocytes in vitro that suppress T-cells and express indoleamine-2,3-dioxygenase (IDO). We investigated whether a CD33/CD3-bispecific BiTE® antibody construct (AMG 330) with pre-clinical activity against AML-blasts by redirection of T-cells can eradicate CD33(+) MDSCs. In fact, T-cells eliminate IDO(+)CD33(+) MDSCs in the presence of AMG 330. Depletion of total CD14(+) cells (including MDSCs) in peripheral blood mononuclear cells from AML patients did not enhance AMG 330-triggered T-cell activation and expansion, but boosted AML-blast lysis. This finding was corroborated in experiments showing that adding MDSCs into co-cultures of T- and AML-cells reduced AML-blast killing, while IDO inhibition promotes AMG 330-mediated clearance of AML-blasts. Taken together, our results suggest that AMG 330 may achieve anti-leukemic efficacy not only through T-cell-mediated cytotoxicity against AML-blasts but also against CD33(+) MDSCs, suggesting that it is worth exploring the predictive role of MDSCs for responsiveness towards an AMG 330-based therapy.