Avastin Fight Highlights Limits Of Big, Random Trials
An FDA advisory panel on Wednesday unanimously rejected arguments from Genentech that its anti-cancer drug should continue as an approved therapy for... View Enlarged Image
A Food and Drug Administration special advisory panel voted 6-0 on Wednesday to withdraw approval of the drug Avastin for treating late-stage breast cancer, sticking to its usual scientific standards despite some patients' pleas to disregard them.
The recommendation was based on the best scientific evi dence currently available, but the controversy highlights the limits of the science the FDA uses to determine drug effectiveness.
"Large, randomized-controlled trials are the 20th century," said Peter Pitts, president of the conservative Center for Medicine in the Public Interest. "The 21st century is the age of personalized medicine, where we need better molecular diagnostics to determine which subpopulation reacts to which treatments."
Avastin blocks the growth of new blood vessels to tumors, and is already FDA-approved for late-stage colon and lung cancer.
In 2008 the FDA granted accelerated approval of Avastin for use in late-stage breast cancer based on one randomized trial that found it slowed the progression of late-stage breast cancer.
In late 2010, the FDA withdrew approval based on two trials that found Avastin did little to slow progression, did not improve overall survival from late-stage breast cancer, and that it posed serious health risks, including death.
Randomized-controlled trials often involve large samples of patients. Such studies determine if a drug has a widespread benefit. But this is also a limitation. A treatment may work for a subpopulation in the sample, but since it doesn't work for most people in the sample, the average shows that the treatment is ineffective.
"The concept that you're looking for one magic bullet that cures a ton of patients at once is the wrong idea," said Terry Kalley, who founded Freedom of Access to Medicines in response to the FDA's 2010 decision on Avastin. "Cancer is likely to be cured by several therapies that target different subpopulations."
Kalley says his wife Arlene has been taking Avastin, along with chemotherapy, for 27 months.
Roche asked the FDA to extend its approval of Avastin until it conducts a study that "would include a biomarker component to identify patients who may be more likely to derive a more substantial benefit from Avastin."
But that is in the future.
"Unfortunately, the data we have right now (don't) allow us to know if there is such a subpopulation," said Karuna Jaggar, executive director of Breast Cancer Action.
Jaggar is concerned that Roche is using this as a stalling tactic to keep Avastin on the market. The Swiss drug giant could lose nearly $1 billion in sales, analysts say.
"We must insist that breast-cancer patients are receiving drugs that have been demonstrated to work," Jaggar added.
Others say regulators should keep Avastin approved for now.
"The FDA should say the evidence isn't there right now, but we will work with Genentech and patient groups to develop the appropriate diagnostics," said Pitts. "But the FDA needs to leave the indication on the label so people can still use it and be reimbursed for that."
If the FDA rescinds Avastin for late-stage breast cancer, at least until subpopulations are studied, then Genentech/Roche cannot market the drug for that purpose.
Doctors could still prescribe Avastin for the off-label use. But it's unlikely that government programs like Medicare and Medicaid or private insurance will cover Avastin — which costs $80,000 or more per year — for breast cancer.
An estimated 17,500 women suffer from late-stage breast cancer.
Many people like Terry testified at the hearings about Avastin. But without better data, such stories are little more than anecdotal.
The flip side is that if it turns out that Avastin does help some subpopulations, such women today will suffer as long as the FDA does not approve the drug for late-stage breast-cancer treatment.
