The recent recommendation by the Oncologic Drugs Advisory Committee (ODAC) to withdraw conditional approval of bevacizumab (Avastin) in combination with chemotherapy for previously untreated metastatic breast cancer has been received with controversy. While many are genuinely concerned about withdrawal of an effective drug, others-such as the author of a recent editorial in The Wall Street Journal (August 18, 2010)-have distorted the science and facts to argue that the decision was political and that ODAC is beholden to the politics of the FDA and Congress.  Unfortunately, these latter arguments serve no purpose but to promote an agenda at the expense of breast cancer patients.

It is important that the breast cancer community rest assured that ODAC is an independent, unbiased panel of experts in oncology whose mission is to provide their best objective scientific and clinical recommendations to the FDA for drug approval. In my experience on this panel, I have never witnessed improper influence from any government agency. In fact, the FDA has taken great steps to maintain the independence and objectivity of ODAC.
The FDA's mandate is to protect the American public by approving drugs that are effective while posing acceptable risks. One need only remember events surrounding the use of thalidomide in the 1950s-the drug caused immeasurable suffering in babies born without limbs, and the FDA blocked approval in the United States. Many more examples exist of the role the FDA serves in protecting the American public from unsafe drugs and/or indications and unscrupulous promotion. Indeed, given that fewer than 1% of drugs ever show significant benefit, it is essential that the American people feel confident that an approved drug will be of help to them and not promoted for economic purposes.

I present this as backdrop to the controversy over bevacizumab, which is a good example of a drug that has benefit but also significant side effects.  When evaluating a drug, one needs to weigh these factors within the context of the disease. For example, in lung and colon cancers, the therapeutic effect of bevacizumab on survival was relatively short and the toxicity significant, but the benefit was nonetheless meaningful because few drugs prolong survival in these devastating diseases. In breast cancer, however, patients experienced only a 1- to 2-month delay in disease progression-without improved survival-and had serious toxicity. The panel recognized that a drug may improve a patient's quality of life without prolonging survival, and that is a very legitimate reason to approve a drug. Unfortunately, the manufacturer (Genentech) did not officially incorporate quality-of-life measurements in their investigations, and the quality-of-life data they did present suggested there was no improvement. In fact, Genentech's response was, "We are not making patients worse." This is not sufficient. We need to make patients better.

It is understandable that some oncologists may want bevacizumab to remain available with a breast cancer indication because it could benefit a subset of yet-to-be-identified patients. Of course, this requires controlled studies employing biomarkers and cannot be presaged through the art of medicine. Indeed, bevacizumab does not target a specific driving oncogenic pathway. Angiogenesis is a complex phenotype, and all tumors show some degree of this process. We must not lose sight of  the big picture, which is that most patients with metastatic breast cancer do not achieve meaningful benefit, and in the absence of benefit, only the potential of harm remains. Indeed, it is important to recognize that although many breast cancer survivors are currently benefiting from a combination of bevacizumab and chemotherapy, the results of the randomized studies indicate that it is the chemotherapy drug and not bevacizumab that is providing the significant benefit. ■

-Wyndham Wilson, MD, PhD, ODAC Chair, Chief Lymphoid Therapeutics Section, Center for Cancer Research, National Cancer Institute

In December 2007, I voted to approve bevacizumab for advanced breast cancer on the basis of the E2100 study. At the time, it was noted that the 22% response rate to paclitaxel was surprisingly low and the apparent difference with the combination might be exaggerated. The higher response rate with bevacizumab suggests drug activity. However, a 1% increased incidence of death with the combination observed in all the clinical trials weighed heavily not only on me, but on breast cancer advocates at the hearing.

E2100 was not meant to be a registration trial. Only grade 3-5 toxicities were reported, and there is no record of patients discontinuing therapy because of toxicity. The FDA review of the data found that 10% of patients lacked CT scans and approximately one-third of patients were not followed until progression or end of study. E2100 is considered the most positive bevacizumab trial, and the missing data are of concern.

The additional data from AVADO and RIBBON-1 supported an increased response rate, but also confirmed the increased toxicity and lack of survival advantage with bevacizumab. Like the majority of panel members, I could not recommend approval. ■

-Joanne Mortimer, MD, Director of the Women's Cancers' Program
and Professor of Medical Oncology and Experimental Therapeutics, City of Hope Comprehensive Cancer Center, Duarte, California

Compiled and reported by Caroline Helwick.

Disclosure of Potential Conflicts of Interest

Wyndham Wilson, MD, PhD: Dr. Wilson has no financial interest or other relationship with the manufacturers of any products discussed in this report.
Joanne Mortimer, MD: Dr. Mortimer has no financial interest or other relationship with the manufacturers of any products discussed in this report.