Avastin OK

"In a surprise decision, the Food and Drug Administration played against type and gave approval for Avastin as a treatment for metastatic breast cancer. It was the right option for terminally ill women, who will gain another weapon against a disease that kills about 40,000 every year.

In clinical trials, Genentech's biologic drug was shown to control the growth and spread of tumors, doubling the amount of time before illness worsened. That translates into an improvement in quality of life, and the results were corroborated by further studies.

Avastin shouldn't have been controversial. But an FDA panel ruled that "progression-free survival" was not sufficient, because the agency's usual acid test for anticancer agents is extending life overall. Such an analysis overlooks the real benefits to women in the months they have left. But as late as last week, it looked as though approval would be delayed or rejected outright.

The "accelerated approval" granted to Avastin is contingent on follow-up trials, and Avastin could be pulled from the market if future research fails to demonstrate that treatment prolongs life. In an interview with us on Monday, oncology drugs chief Richard Padzur said that the FDA was "not demanding" a survival advantage but would consider it a factor. In 2005, the lung cancer drug Iressa was approved and then withdrawn under similar conditions, and the danger is that history will repeat itself.

As for the notion that the decision portends a shift in the way the FDA evaluates cancer drugs, Dr. Padzur assured us that it "not a new step in our regulatory decision making." He emphasized that prolonging life was still the FDA's primary criterion, and noted that Avastin was green-lighted because its quality-of-life benefits were "statistically robust." Yet the narrowness with which the FDA balances risk and reward was the reason Avastin was contentious -- especially Dr. Padzur's statistical models.

FDA evaluation methods insist on large average effects and simplistic mortality rates. Only about 10% of patients responded to Iressa, for instance, and could be identified by genetic tests; but these targeted results didn't mesh with the arbitrary FDA approach. Other innovative medicines like Provenge (for prostrate cancer) and Junovan (for pediatric bone cancer) are pointlessly blocked by Dr. Padzur's division, and the risk is that Avastin still could be too, whatever its current reprieve."