CD47 is a protein found on the surface of nearly every cell in your body. Its primary job is to send a “don’t eat me” signal to your immune system, protecting healthy cells from being destroyed by roaming immune cells called macrophages. This protective function is essential for normal health, but it also plays a central role in how cancers hide from immune detection, making CD47 one of the most actively studied targets in cancer therapy.
How the “Don’t Eat Me” Signal Works
Macrophages are immune cells that patrol your body looking for damaged, infected, or abnormal cells to engulf and destroy, a process called phagocytosis. To avoid being eaten by mistake, healthy cells display CD47 on their surface. When a macrophage encounters a cell, the CD47 protein binds to a receptor on the macrophage called SIRPα (signal regulatory protein alpha). This binding triggers a chain of events inside the macrophage that essentially tells it to stand down.
Specifically, the SIRPα receptor activates internal braking signals that shut off the macrophage’s cell-eating machinery, including the motor proteins it needs to physically engulf a target. The result: the macrophage moves on, and the healthy cell is left alone. This checkpoint system runs constantly across your tissues, keeping your immune system from attacking your own body.
CD47 and Red Blood Cell Turnover
One of the clearest examples of CD47 in action involves red blood cells. Young, healthy red blood cells carry abundant CD47 on their surface, organized in small clusters. This keeps them circulating safely for about 120 days. As red blood cells age, though, two things change. The total number of CD47 molecules on the cell surface drops significantly, and the remaining molecules rearrange into larger, denser clusters.
These bigger clusters attract a binding partner called thrombospondin-1, which flips the signal from “don’t eat me” to something closer to “eat me.” The combination of fewer CD47 molecules and increased thrombospondin-1 binding marks aged red blood cells for removal. Macrophages in the spleen then clear them out, making room for fresh replacements. This is how your body continuously recycles its blood supply without any conscious effort. The same principle applies to dying cells elsewhere in the body: as cells become damaged or reach the end of their life, their CD47 levels drop or change shape, allowing the immune system to clean them up.
How Cancer Exploits CD47
Cancer cells hijack this protective system. By overexpressing CD47 on their surface, tumor cells effectively disguise themselves as healthy tissue, sending a constant “don’t eat me” signal to surrounding macrophages. This allows them to avoid immune detection, proliferate, and spread.
CD47 was first identified as a tumor marker on human ovarian cancer in the 1980s. Since then, elevated CD47 levels have been found across a wide range of cancers: acute myeloid leukemia, chronic myeloid leukemia, acute lymphoblastic leukemia, non-Hodgkin lymphoma, multiple myeloma, bladder cancer, breast cancer, lung cancer, and brain tumors among them. High CD47 expression is a prognostic indicator of poor outcomes, meaning patients whose tumors express more CD47 tend to have worse survival rates.
Cancer stem cells, the subset of tumor cells thought to drive regrowth and metastasis, appear especially reliant on CD47 for protection. This makes CD47 a particularly appealing therapeutic target, since blocking it could potentially eliminate the cells most responsible for cancer recurrence.
CD47 Beyond Immunity
CD47 does more than just regulate immune cell behavior. It also interacts with thrombospondin-1 in blood vessels, where it suppresses the growth of new blood and lymphatic vessels by blocking nitric oxide signaling. Nitric oxide is a molecule your blood vessels need to relax, grow, and function properly. When thrombospondin-1 binds to CD47 on the cells lining lymphatic vessels, it reduces nitric oxide production and increases oxidative stress, which inhibits new vessel growth.
This has implications for cardiovascular disease. Research in vascular biology has shown that CD47 activation in lymphatic vessel cells limits lymphatic drainage and promotes the formation of atherosclerotic plaques. When researchers silenced CD47 in these cells, nitric oxide production increased, oxidative stress dropped, and atherosclerotic lesions shrank. So CD47’s influence extends well beyond cancer into heart disease and blood vessel health.
Blocking CD47 as Cancer Therapy
The therapeutic logic is straightforward: if tumors hide behind CD47, blocking that signal should expose them to immune attack. Anti-CD47 antibodies work by binding to CD47 on tumor cells and preventing it from engaging SIRPα on macrophages. Without the “don’t eat me” signal, macrophages recognize and engulf the cancer cells.
In preclinical studies, this approach has shown striking results. Blocking CD47 enabled macrophages to phagocytose lung cancer cells, including cancer stem cells, and inhibited tumor growth in animal models. The effects were even more dramatic when anti-CD47 antibodies were combined with other cancer drugs. In mouse models of non-Hodgkin lymphoma, combining an anti-CD47 antibody with rituximab (a standard lymphoma drug targeting a different surface protein, CD20) cured up to 89% of treated animals. The combination worked through two complementary mechanisms: the anti-CD47 antibody removed the immune brake, while rituximab actively flagged cancer cells for destruction. Neither drug alone achieved anything close to the same result.
Similar combination strategies have been tested with azacitidine, a drug used in blood cancers like acute myeloid leukemia. The leading anti-CD47 antibody, magrolimab, reached Phase 3 clinical trials in combination with azacitidine for patients with a specific aggressive form of AML.
Why CD47 Therapy Is Complicated
The central challenge is that CD47 isn’t only on cancer cells. It’s on virtually every cell in the body, and red blood cells carry it in especially high amounts. When anti-CD47 antibodies enter the bloodstream, they bind to red blood cells as well as tumor cells, triggering macrophages to destroy healthy red blood cells. This causes anemia, sometimes significant enough to require transfusions.
A second problem is hemagglutination, where anti-CD47 antibodies cause red blood cells to clump together. Because the antibody has two binding arms, it can bridge CD47 molecules on adjacent red blood cells, sticking them to one another. This clumping contributes to the red blood cell toxicity seen in clinical use and has been a major obstacle to getting these drugs through trials safely.
Researchers are working on several solutions: engineering antibodies that bind to a different spot on CD47 to minimize red blood cell cross-linking, using dosing strategies that start low to gradually reduce red blood cell counts before full treatment begins, and developing molecules that preferentially target CD47 on tumor cells rather than healthy tissue. The biology is promising, but safely translating it into treatments patients can tolerate remains the field’s biggest hurdle.