Jooho Chung, MD, PhD

Robert Manguso, PhD

Macrophages, much like Alice of “Alice in Wonderland,” recognize and consume tumor cells that display “eat me” surface markers. However, tumor cells can evade detection by macrophages if they successfully present “don’t eat me” signals.

To counter this, researchers have developed drugs aimed at turning off these “don’t eat me” signals. However, such treatments haven’t been as effective as anticipated in patients with acute myeloid leukemia (AML) and other blood cancers.

To better understand why, a team of researchers from Mass General Brigham, Dana-Farber Cancer Institute and the Broad Institute of MIT and Harvard looked at samples from patients treated for AML at Mass General Brigham. The team was led by Jooho Chung, MD, PhD, Mounica Vallurupalli, MD, Todd Golub, MD, and Robert Manguso, PhD.

The team conducted a genome-scale loss of function screen in AML cell lines, systematically turning off individual genes and cataloging those that affected detection by macrophages. Surprisingly, the classic CD47 “don’t eat me” signal had only a weak effect.

Instead, the researchers found that another signal—CD43—had a much stronger influence on macrophage detection. Their findings suggest that therapies targeting CD43 could be promising for treating patients with AML and, potentially, a broader range of cancers.

Published in Science on April 9, 2026 | Read the paper: “Sialylated CD43 forms a glyco-immune barrier that restrains anti-leukemic immunity.”

Summary reviewed by: Jooho Chung, MD, PhD, co-lead author; Mounica Vallurupalli, MD, co-lead author; Robert Manguso, PhD, co-senior author