Intended for healthcare professionals


FDA committee votes to withdraw bevacizumab for breast cancer

BMJ 2011; 343 doi: (Published 05 July 2011) Cite this as: BMJ 2011;343:d4244
  1. Jeanne Lenzer
  1. 1New York

An advisory committee to the US Food and Drug Administration voted unanimously to withdraw approval of a top selling drug for the treatment of metastatic breast cancer.

The six member panel concluded on 29 June that bevacizumab (marketed by Genentech as Avastin) had not been shown to be either effective or safe for the treatment of breast cancer. The FDA is inviting public comment on the decision for one month before issuing a final ruling.

This unusual action comes after several surprising developments in the approval process of bevacizumab.

In February 2008 Genentech was granted “accelerated approval” to use bevacizumab to treat breast cancer. The approval relied on a key study, known as E2100, that tested paclitaxel (marketed as Taxol) alone versus bevacizumab together with paclitaxel. The decision was seen as highly unusual, because the study used a controversial primary end point of “progression-free survival.”

Ralph D’Agostino, professor of mathematics and statistics, biostatistics, and epidemiology at Boston University, served on the FDA committee that approved bevacizumab. Writing in a commentary published in the New England Journal of Medicine last month (doi:10.1056/NEJMp1106984), Professor D’Agostino said that the FDA’s acceptance of progression-free survival was a “new precedent” that could have “serious consequences” and could mean that “data on overall survival might never be obtained, since no studies of sufficient duration would be conducted.”

Genentech defended its use of progression-free survival as an end point last December. At that time, it also pointed out that the drug was still approved by the European Medicines Agency to treat breast cancer and argued that women in the United States should have the same treatment option as European women.

Dr D’Agostino, in his commentary, said that the “European community accepts this end point as clinically meaningful, but many methodologic issues remain unresolved.”

Jerome Hoffman, an expert in epidemiology and professor of medicine and emergency medicine at the University of Southern California, told the BMJ that progression-free survival is a poor surrogate marker for overall survival. He said, “It is possible for more people to die—and to suffer significant morbidity—from a ‘successful’ treatment if progression-free survival is used as the primary end point, because it only tells you whether tumour sizes have increased and it tells you nothing about the quality of the patient’s life or whether he or she survived any longer.”

A spokesman for Genentech admitted that after the trials that used progression-free survival as the end point there were no further data showing that overall survival improved. In a statement issued on 29 June, after the committee’s vote, the company said, “There are no data available showing that Avastin improves disease-related symptoms or survival in HER2-negative metastatic breast cancer.”

The company also acknowledged that women in the bevacizumab arm of study E2100 experienced a 21% increase in “severe to life-threatening and fatal side effects, such as gastrointestinal perforation and severe bleeding, over women not receiving the drug.”

Curt Furberg, a former FDA adviser and professor of biostatistics at Wake Forest University in Winston-Salem, North Carolina, told the BMJ, “The diagnosis of progression-free survival is susceptible to detection bias.”

The tools used to measure progression are “fairly crude,” he said. “For example, a negative x ray [picture] of the lungs may not preclude the presence of lung metastases. Ideally I would prefer to use overall survival, since it is a straightforward measure. It’s a count: you’re dead or you’re alive.”

The advisory panel’s findings underscored the concerns of critics. Richard Pazdur, director of the Office of Oncology Drug Products at the FDA’s Center for Drug Evaluation and Research, cited many of these concerns. During public hearings he said, “Although progression-free survival contains the word ‘survival,’ it does not mean that [a] patient’s life will be extended.”

He went on to note that although E2100 showed a five month improvement in progression-free survival, another study showed no such benefit, and none of the five randomised trials of bevacizumab showed a benefit in overall survival.

Dr Hoffman emphasised that when investigators are allowed to decide whether the death of a participant in a clinical trial was, or was not, caused by the treatment drug (as was the case in study E2100), distortions of determinations of cause of death are likely, “especially if the investigators have a financial interest in the drug being studied,” said Dr Hoffman. “For this reason as well, it’s essential that we track all cause mortality and not just progression-free survival.”


Cite this as: BMJ 2011;343:d4244

View Abstract