Cancer cells appear adept at dodging some therapies

On paper, the new targeted therapies for cancer are a guaranteed success: They take lethal aim at molecules and biochemical pathways that allow cancer cells to proliferate, spread and avoid self-destructing. Yet this is not so as a report by Sharon Begley in the Wall Street Journal shows.

Even the most vaunted targeted therapies being described at the annual meeting of the American Association for Cancer Research tave disappointingly low success rates. ImClone's Erbitux, for instance, shrinks only 10% of colon tumors.

That highlights a property of cancer cells that scientists are only now appreciating. Call it Murphy's Law in reverse: For a cancer cell, whatever can go right, will go right.

Targeted cancer therapies were born from discoveries about the molecular mechanisms that transform normal cells into cancer cells. Biologists have identified some 15 tumor-suppressor genes and more than 100 growth-promoting oncogenes. They also have worked out the pathways in cells by which these genes unleash cancer - by overproducing receptors for growth-signaling molecules, for instance, or overriding the 'suicide' program of cells gone bad.

What could be more obvious in the first case than to fill the receptors with a cancer drug? The signaling molecule that would otherwise dock with the receptor, triggering the cell to divide and multiply, could then no more squeeze into it than a car could get into an occupied parking space.

Or so it would seem. In real life, says molecular biologist Robert Weinberg, of the Whitehead Institute in Cambridge, MA, US: 'cancer cells are much better able to resort to evasive maneuvers than we realized. If you throw one attack at them, they activate a pathway to elude it.'

One evasive manouvre relies on genetic instability. Early in the development of cancer, cells rack up thousands of random mutations, says biologist Lawrence Loeb of the University of Washington, Seattle. Some of those mutations mangle genes whose job is to keep a cell's DNA intact. With these DNA-fidelity genes out of commission, the result is a 'cascading number of mutations,' says Prof. Loeb. This 'mutator phenotype,' as he calls it, lets cancer cells keep generating new and diverse progeny, 'some of which are resistant to standard chemotherapy, and maybe to targeted therapies as well.'

Novartis's Gleevec, for example, is used against a form of leukemia and a rare stomach cancer. It targets a growth-promoting protein called a kinase. But sooner or later, most of the tumors Gleevec targets develop a mutant kinase, whose shape is a tiny bit different from the original.

This mutant still makes the cancer cell grow, explains molecular biologist Scott Lowe of Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. The drug can't fit it to disable it. About three-quarters of patients on Gleevec for more than two years develop resistance.

Shape-changing mutations are only the leading edge of cancer cells' evasive talents. 'If you block one pathway [to growth], other pathways can take over,' says biologist Stanley Riddell of the Fred Hutchinson Cancer Research Center in Seattle.

That may limit the power of both Erbitux and AstraZeneca's Iressa (for non-small-cell lung cancer). Each targets a receptor called epidermal growth factor receptor (EGFr). True to its name, EGFr acts as the docking port for a growth-promoting molecule. Block the receptor, goes the thinking, and the growth promoter can't stimulate the cell to divide and multiply.

'But for a therapy that targets a specific pathway to be effective, that pathway has to be one of the main ones driving that abnormal cell,' says oncologist Roy Herbst of M.D. Anderson Cancer Center in Houston. If the tumor isn't relying on the EGF pathway for growth, the drugs will have no effect. The tumor will choose another path and carry on just fine.

This multiplicity of redundant pathways may explain why antiangiogenesis drugs, once touted as the cure for cancer, fall short. Angiogenesis is the development of blood vessels to nourish tumor cells that journey away from the initial site. They can no more survive without blood vessels than towns in the Old West could survive without stagecoach lines.

But despite the hope and hype (about 40 antiangiogenesis drugs are in clinical trials), these drugs also are subject to cancer cells' quick-change artistry. They might not work, says Dr. Herbst, if the cancer is fed by many angiogenesis factors and the drug blocks only one.

Genentech's Avastin, for example, targets a gene (VEGF) that promotes angiogenesis. By turning off VEGF, Avastin should starve tumors. 'But even when treated with a drug that targets VEGF, a cancer cell can still grow because there are more than 20 factors that drive angiogenesis,' says Dr. Herbst.

Clearly, cancer cells have no single Achilles' heel. Instead, they can resort to any number of pathways to accomplish the same ends. That bolsters the case for combination therapies, using several drugs that knock out the cells' preferred grow-and-divide pathway and, simultaneously, its fallbacks. 'This is a solvable problem,' says Dr. Lowe. But it's not an easy one