font size
Sign inprintPrint
PERSONALIZED MEDICINE

More Testing Needed

FDA plays it safe, remaining unconvinced about retrospective analyses.

MARIE DAGHLIAN

The Burrill Report

“Hagop Youssoufian, a senior vice president at ImClone, characterized the data as 'nothing less than transformational.'”

One of the aims of personalized medicine is to reduce costs by treating only those patients most likely to benefit, but personalizing drugs may become more expensive, especially if personalization comes after the drug is approved. This is the take-away message following a key U.S. Food and Drug Administration Oncology Drugs Advisory Committee meeting in mid December. The meeting looked at what standards need to be applied before gene tests are recommended as companion diagnostics for targeted cancer drugs and whether retrospectively identified biomarkers should be included in regulatory decisions.
 
The advisory panel met for what FDA officials characterized as an educational dialogue, as the agency seeks guidance about what should be done when an already approved drug is later found to work only in a subset of patients with certain gene variants.
 
It’s an issue with which the agency is having to wrestle. Amgen’s Vectibix and ImClone System’s Erbitux, both EGFR-inhibiting monoclonal antibodies approved for the treatment of colorectal cancer, only work in the subset of patients without a mutant K-RAS gene. The drugs are ineffective in the 40 percent of patients who have mutant K-RAS tumors. Both companies want the FDA to approve label changes to tell doctors that patients with mutant K-RAS should not be treated with their drugs. The problem is that these drugs were approved before the mutant-K-RAS gene discovery connection, which occurred this past year.
 
Both companies submitted data to the advisory panel from past studies, claiming the information was sufficient to show that their colorectal cancer drugs best helped patients with a normal or “wild-type” version of the KRAS gene. Metastatic colorectal cancer patients who carry the wild-type version of KRAS are much more likely than patients with the mutated form of the gene to benefit from Erbitux, according to data from a large multinational prospective clinical trial that was presented at this year’s American Society of Clinical Oncology’s annual meeting in June by Eric Van Cutsem, a professor at the University Hospital Gasthuisberg in Leuven, Belgium.
 
Amgen has gone ahead and signed an agreement with the U.K.-based personalized medicine company DxS to provide its TheraScreen KRAS mutation kit as a companion diagnostic for Vectibix in the United States, a move that gives Amgen added incentive to get the drug label changed.
 
Hagop Youssoufian, a senior vice president at ImClone, characterized the data as “nothing less than transformational.” But the advisory panel didn’t agree. They said that drug companies will need to conduct additional studies and gather more data before they can win their case to use genetics to target patients.
 
In briefing documents ahead of the meeting, the FDA made it clear that it discourages retrospective genomic biomarker assessments because these analyses, which look back at data to try to explain events, can introduce biases. The FDA prefers prospective studies that ask a question in advance, try to control for the variables, and then answer the question directly. The FDA prefers that companies conduct simultaneous prospective analysis of the clinical studies to establish the efficacy of the drug and gauge the prognostic and predictive value of the genomic biomarker with a “well-characterized” assay. 
 
The FDA recognizes that in cases where researchers have uncovered a genetic association after a drug has been approved, simultaneous development of Rx/Dx is not possible. But it was clear that Amgen and ImClone will have to run larger studies to gather enough data to get agency to change the labels to include a companion diagnostic. 
 
None of this comes as a surprise. The FDA has been under attack for not being cautious enough and has become increasingly risk adverse. While the European Medicines Agency has gone ahead and recommended that doctors establish patients’ K-RAS status before prescribing Vectibix and Erbitux, the FDA has held back issuing similar recommendations.
 
Genentech claims its colon cancer drug Avastin works in patients with or without the K-RAS mutation. The company, which uses pharmacogenomics information on the label of its breast cancer drug Herceptin requiring it to be prescribed in conjunction with a diagnostic test to determine HER2 status, filed a citizen petition in early December urging the FDA to take on greater oversight of diagnostic tests used to guide therapeutic decisions and to regulate all laboratory developed tests. Genentech expressed concern that many companies are starting to make and sell tests with claims not verified by the FDA.
 
The outcome of the meeting sent a significant message to companies striving for personalized medicines that they will have to conduct more studies and gather more data before they can make their case for Rx/Dx approvals.
 


December 19, 2008
http://www.burrillreport.com/article-more_testing_needed.html

[Please login to post comments]

Other recent stories

Sign Up to recevie the Burrill Weekly Brief


Follow burrillreport on Twitter