Amgen’s Bid to Make Cancer Drug “Personalized” Lacks Data, FDA Panel Says

Xconomy Seattle — 

Amgen, the world’s largest biotechnology company, with operations in Seattle and Cambridge, MA, came up short in its bid to make its colorectal cancer drug personalized today in front of a panel of expert advisers to the FDA. The panel said Amgen and Eli Lilly will have to run more clinical trials to gather enough evidence that doctors should screen patients for their genetic status, to see whether they are likely to benefit from the drug or not.

The FDA panel, according to this Reuters report, said the companies need to have a well-organized plan to review their data to support this conclusion. They will also need enough patients and tissue samples to analyze genetic issues. “It’s going to make life more complicated, costly…it’s going to require larger clinical trials,” said panel member Richard Simon of the National Cancer Institute, according to the Reuters story.

Studies by Amgen and ImClone Systems, recently acquired by Lilly, have shown that about 40 percent of colorectal cancer patients with a normal form of a gene called KRAS are likely to benefit from their drugs, while the rest, who have a more aggressive form of disease, are likely to get a whole lot of side effects with little chance of benefit. The statistical flaw in this work, however, is that these are “retrospective” analyses, in which researchers look back at data and try to explain events, which can introduce biases. The FDA usually prefers “prospective” studies that ask a question in advance, try to control for a whole lot of potential variables, and then answer the question directly.

I laid out the background of this issue in a detailed preview story on Monday, so I won’t repeat it all here. Suffice to say, Amgen would have been happy to get a go-ahead to market its drug to normal-KRAS patients, but it didn’t sound heartbroken. It is awaiting data from a pair of clinical trials in 2009, called “181” and “Prime” which are designed to ask prospectively whether KRAS status predicts whether a patient will respond to its drug, panitumumab (Vectibix).

“We are happy to have been provided the opportunity to participate in today’s discussions and look forward to our continued dialogue with FDA on how best to address both KRAS and future biomarker data in our clinical trials,” the company said in a statement after the hearing.

By posting a comment, you agree to our terms and conditions.

3 responses to “Amgen’s Bid to Make Cancer Drug “Personalized” Lacks Data, FDA Panel Says”

  1. Richard says:

    It’s articles like this that display the real difference between Xconomy and MSM. There is a fair assessment of what happened and why, with the background, links, etc. needed for a dedicated reader to gain much more information. Less spin and more perspective.

    I think that this drug will be a great example of ‘personalized’ medicine but we are early in this new realm and the FDA wants to be extra certain.

    If Amgen and Lilly can produce really strong statistical numbers for this drug in people with a specific genetic background, it will be a great step forward towards our understanding of why some drugs work only for some groups of people.

  2. Anon says:

    From same Reuters story linked in piece:

    “Despite ImClone and Amgen’s request, the FDA did not seek advice on Erbitux and Vectibix but cited them as examples.”