• 5/21/2008
  • Robert Langreth
  • Forbes.com

ImClone Systems’ breakthrough colon cancer drug Erbitux has become a billion-dollar best seller as a one-size-fits-all drug for patients who have failed treatment with chemotherapy.

But emerging new gene findings indicate that the high-priced drug may be useless in nearly half of colon cancer patients–and that a simple gene test could allow doctors to predict in advance which patients are likely to benefit. The drug, co-marketed by ImClone and Bristol-Myers Squibb, costs $38,000 for a four-month course of treatment.

The result is one of the first examples of personalized medicine in colon cancer. It helps explain why tumors shrink dramatically in a minority of patients who get the drug while most others get little benefit. The new gene test could allow doctors to give the drug only to patients who are likely to benefit, and avoid those who are likely to just get side effects. Side effects of the drug include skin rashes and infections in the fingers and toes.

A series of studies published over the last several months indicate that about 35% to as much as 45% of colon patients probably should not be taking Erbitux or a similar drug from Amgen called Vectibix. These patients have tumors with a mutation in a growth-promoting gene called kras that may render the drugs ineffective.

Erbitux and Vectibix are antibodies that bind to receptors on the cell surface called EGF and blocks signals that turn kras on. But the mutations apparently cause kras to become permanently stuck in the “on” position. Although the results are still preliminary, this appears to eliminate the possibility of benefiting from the drug.

“It is the first ‘thou shalt not’ gene marker in colon cancer,” says Memorial- Sloan-Kettering colon cancer specialist Leonard Saltz, a pioneer in testing Erbitux. He calls the data “unbelievably airtight.” … “Thou shalt not give Erbitux and Vectibix to patients with mutated kras; all it will do is give toxicity and run up the bill.” Memorial started testing colon cancer patients for kras early this year.

“Largely what personalized medicine will be is telling people which drugs are out there that don’t work so don’t bother. This is the first step in that direction,” Saltz says. That won’t be good news for some patients who find out they have one less treatment option, he admits. “But it is reality, and you can look at it or stick your head in the sand.”

One French study published in the Jan. 20 Journal of Clinical Oncology found that none of the 24 patients with kras mutation responded to Erbitux, while tumors shrank in 40% of those with normal kras. In another study by researchers in Belgium published in March, Erbitux worked in none of 42 patients with the mutation, but 27 of 66 patients without it. Amgen has published a big study showing that its Vectibix only works in patients with normal kras. In Europe, Vectibix is only approved for patients who have normal kras.

Larger trials are needed to confirm the Erbitux results. But UCLA’s Jonsson Cancer Center is also in the process of gearing up to perform the kras gene test, says J. Randolph Hecht, who directs the gastrointestinal oncology program there. “We don’t want to expose people to drugs that won’t benefit them.”

The emerging results could throw the colon cancer drug market in turmoil. Bristol-Myers and ImClone could lose a significant chunk of sales as doctors stop using it in patients with the gene mutation. But Erbitux may also gain sales if doctors start using Erbitux earlier on in the course of the disease in those with normal version of the gene. This could slice into sales of Genentech’s Avastin drug, which is currently a mainstay of initial treatment for metastatic colon cancer.

How this will wash out over the long run isn’t clear at the moment.

M.D. Anderson Cancer Center’s Scott Kopetz says that about one-third of his colon patients, ones who lack the mutation, are now getting Erbitux as an initial treatment instead of Avastin. “The holy grail is trying to identify which patients will benefit,” he says. If there is any benefit of the drugs in those with kras mutations, “it is exceedingly small,” he adds.

“It is potentially a very big breakthrough,” says Bristol-Myers Vice President Martin Birkhofer. He predicts that the finding could expand the use of Erbitux in patients who do not have a kras mutation. But he says the company will follow the science even if it means many patients who would have gotten the drug in the past no longer get it.

Back at Memorial Sloan Kettering, Saltz believes it is premature to start giving Erbitux instead of Avastin to patients. There is no proof that using Erbitux up front can keep people alive longer than using Avastin, even in those with normal kras. Meanwhile, Erbitux is not currently FDA approved for initial treatment of colon cancer, and insurers could balk at paying for earlier Erbitux use.

Much more data on kras will be presented at a big meeting of cancer doctors in Chicago at the beginning of June that could clarify the matter once and for all. In particular, researchers will re-analyze a previous big trial of Erbitux as initial colon cancer treatment to see if the drug combo worked better in the subset of patients who don’t have kras mutations.