CD47 is a protein flag normally expressed on the surfaces of certain cells, such as circulating blood stem cells, to protect them from an organism's immune system. About 10 years ago, Irving L. Weissman (Stanford University School of Medicine, CA) and colleagues showed that certain types of cancer, especially leukemia and lymphoma cells, also expressed CD47, helping them to evade destruction by immune cells. In the past few years, various research groups have successfully used antibodies against CD47 to cure some cases of leukemia and lymphoma in mice.

Now, a team led by Weissman reports that blocking CD47 is also effective in treating a host of human solid cancers transplanted into mice: breast, ovary, colon, bladder, brain, liver and prostate tumors shrank substantially or even disappeared completely when treated with antibodies against CD47 (Proc. Natl. Acad. Sci. USA doi:10.1073/pnas.1121623109; published online 26 March 2012).

Weissman explained the findings in a press release: “What we've shown is that CD47 isn't just important on leukemias and lymphomas. It's on every single human primary tumor that we tested.” Higher expression levels of CD47 were associated with lower rates of progression-free and overall patient survival, suggesting that CD47 expression may be a useful prognostic tool in some solid tumors.

When placed together in vitro, immune cells called macrophages failed to destroy cancer cells. But with the addition of monoclonal antibodies against CD47, the macrophages engulfed and destroyed the cancer cells. Furthermore, when human-derived tumors were transplanted into mice and treated with anti-CD47, the tumors shrank and did not spread throughout the body. In five mice with breast cancers, treatment with anti-CD47 prevented tumor growth, and the mice remained cancer-free for 4 months afterwards. In mice with colon cancers, the tumors shrank to about one-third of their original sizes after treatment with anti-CD47. And in mice with bladder tumors, cancer spread to the lymph nodes in all 10 mice that did not receive anti-CD47 but in only 1 of 10 mice that were given the antibody.

Weissman stated, “Blocking this 'don't-eat-me' signal inhibits the growth in mice of nearly every human cancer we tested, with minimal toxicity. This shows conclusively that this protein, CD47, is a legitimate and promising target for human cancer therapy.” The CD47 antibody wasn't 100% effective, however: mice with tumors transplanted from one person with breast cancer did not respond to the treatment. Nevertheless, Weissman's group is eager to begin testing the treatment in human clinical trials.