In March, the FDA's Cellular, Tissue and Gene Therapies Advisory Committee endorsed approval of Dendreon's sipuleucel-T vaccine (Provenge) for treating men with metastatic prostate cancer. The committee voted 17-0 that the agent is safe, and 13-4 that it is effective.
However, neither of the two small studies presented by Dendreon in support of approval met the primary endpoint of progression-free survival, and overall survival was significant only in post hoc and combined analyses.
At least one dissenter said that approval should be delayed until Dendreon completes its phase III IMPACT study of Provenge in men with hormone-refractory prostate cancer, to answer lingering questions about the drug's ability to improve overall survival.
Two months later, counter to its advisory committee's recommendation, FDA denied approval of the vaccine, citing a need for more data on overall survival. FDA has agreed with Dendreon to accept either a positive interim or final analysis of survival from the IMPACT study to complete the licensure of the agent.
The denial of approval for Provenge set off strident protests from advocates who believe the drug should have been approved based on the available data. These advocates' voices were joined by others, including the Abigail Alliance for Better Access to Developmental Drugs, who think that terminally ill patients should have access to investigational drugs after phase I trials have shown safety. In September, the advocacy groups held a small, staged protest outside of FDA's Rockville, Maryland, offices (see photographs of the protest below, courtesy of Leslie Mulkey).
The mainstream press got into the fray, with pieces such as the Wall Street Journal Op-Ed "The FDA's deadly track record" [WSJ.com 08/14/07], in which representatives from the Abigail Alliance excoriated the FDA, claiming that Provenge was "kept away from prostate cancer sufferers because of a small but aggressive club of FDA advisors hand-picked by the director of the agency's Office of Oncology Drugs who thinks the statistics are not yet perfect enough."