Abstract
CD47 is a surface membrane protein expressed by all normal tissues. It is the so-called "don't eat me signal" because it protects the cells against phagocytosis. The CD47 interacts with the signal regulatory protein alpha (SIRPα) on the surface of macrophages, leading to downstream inhibitory signaling that dampens phagocytic capacity. Since macrophages exert immune surveillance against cancers, cancer cells overexpress CD47 to defend themselves against phagocytosis. Acute myeloid leukemia (AML) is a cancer of hematopoietic stem/progenitor cells (HSPC), and similar to other types of cancers, leukemic blasts show enhanced levels of CD47. In patients with AML, CD47 has been associated with a higher disease burden and poor overall survival. Blockage of CD47-SIRPα signaling leads to improved phagocytosis of AML cells and better overall survival in xenograft models. However, the introduction of a pro-phagocytic signal is needed to induce greater phagocytic capacity. These pro-phagocytic signals can be either Fc receptor stimulants (such as monoclonal antibodies) or natural pro-phagocytic molecules (such as calreticulin). Based on these pre-clinical findings, various clinical trials investigating the blockade of CD47-SIRPα interaction have been designed as monotherapy and in combination with other anti-leukemic agents. In this review, we will discuss CD47 biology, highlight its implications for AML pathophysiology, and explore the potential clinical translation of disrupting CD47-SIRPα to treat patients with AML.