Leukemia talks to itself, or how AML cells can signal their own growth

We put lots of hurdles up to slow or stop cancer growth, but cancer cells often get over them. The more we know about the how they do it, though, the more we can do about it. (Phil Roeder/Wikimedia Commons)

Leukemia and other cancer cells are really good at hurdling over the obstacles we throw at them. It’s the whole basis of drug resistance: we attack a mutation with one drug, and another mutation arises to cancel out the drug’s effect. Or the cell ramps up other pathways to compensate for the one blocked by the drug.

But the more we learn about the backup systems cancer cells use to get around our treatment strategies, the better we can get at controlling or eliminating cancer cells. Alex Kentsis, MD, PhD, and Thomas Look, MD, of Dana-Farber/Children’s Hospital Cancer Center, came across one such backup while trying to find which genes a blood cancer called acute myeloid leukemia (AML) relies on to survive. In a nutshell: the cells talk to themselves—chemically, that is.

Knocking down gene after gene individually using a technique called RNA interference (RNAi), the pair had become particularly interested in the gene for a growth factor called HGF (hepatocyte growth factor). HGF acts as an ignition switch for cancer cell growth in nearly half of all AML patients, switching on a receptor called MET, a portal for incoming cell signals. When the researchers muzzled the gene for HGF or genes activated by MET, the cells in nearly half their AML samples either slowed their growth or died.

Knowing cancer cells’ habit of rebounding after treatment, Kentsis and Look weren’t shocked when the cells began growing again. What surprised them was the speed of the rebound: Within days of being treated, the cells were growing as if nothing had happened—much too quickly to have resulted from mutation.

The reason, as Look and Kentsis explain in a recent paper in Nature Medicine, was that the shutdown of the MET-associated genes prompted the cells to crank up their own production of HGF—in effect, telling themselves to grow by stimulating MET more. The HGF produced by the AML cells surged to 30 times its previous level, overpowering the growth-slowing effect of molecules the researchers were using to block MET.

“A built-in feedback loop enables the cells to sense when the MET receptor has been shut down and to adapt by increasing their supply of HGF,” Kentsis explains. “This kind of self-induced growth has been known to occur in solid tumors; this is the first time it has been found in AML.”

When Kentsis and Look’s team added a second inhibitor that blocked the feedback loop, it produced a sustained die-off of AML cells that lasted many weeks. Their findings suggest that therapies directed at both the initial growth trigger and the backup system may have the best chance for success in treating AML; a similar strategy could work for other cancers, as well, giving tumor cells a one-two knockout punch.

Look notes that, “Our study demonstrates that in developing personalized cancer medicines, we will also need to block the intrinsic escape mechanisms that are launched by the cancer cells if patients are to realize a sustained benefit.”

[Ed. note: The original version of this story, written by Robert Levy, ran on page one of Dana-Farber Cancer Institute’s Inside the Institute on June 19, 2012.]