BETHESDA, Maryland—Members of the Oncologic Drugs Advisory
Committee (ODAC) have recommended a change in the endpoint the Food and Drug
Administration (FDA) uses to evaluate hormonal agents as first-line therapy for
metastatic breast cancer. The panel considered the issue of revising the design
of hormonal therapy clinical trials at the agency’s request.

A key ODAC recommendation, made by a 10-to-3 vote, was that in making
drug-approval decisions, FDA should replace response rate, its long-time
endpoint for anticancer hormonal agents in the first-line setting, with time to
tumor progression.

FDA has accepted response rate as an endpoint since 1976, when it approved
the first hormonal anticancer agent, megestrol acetate (Megace), as a
palliative treatment for advanced breast cancer. One year later, FDA cleared
tamoxifen (Nolvadex) for marketing, again relying on response rate findings.

For more than 2 decades, FDA has evaluated breast cancer hormonal agents on
whether they were noninferior (no worse than) or superior to an approved drug,
usually tamoxifen, used as a comparison.

Generally, FDA has interpreted non-inferiority to mean ruling out that a new
drug has a response rate that is 10% lower than the comparison agent. Since
1995, FDA has approved toremifene (Fareston), anastrozole (Arimidex), and
letrozole (Femara) using this standard for evaluation. The agency has required
sponsors to submit data on survival and time to progression as secondary
endpoints.

During a half day of deliberations, ODAC members also:

• Voted 12 to 1 that non-inferiority to tamoxifen no longer be
  considered an acceptable basis for approval. If tamoxifen is the comparator,
  demonstration of superiority should be required.
• Voted 10 to 3 against designation of a single approved drug as the
  only acceptable comparator in future clinical trials. FDA noted that choice of
  a comparator will determine whether superiority or non-inferiority is an
  appropriate trial design and analysis.
• Advised FDA 9 to 4 that data do not suggest that one class of
  hormonal agents is clearly superior to another.

In urging a switch from response rate to a time-to-progression endpoint in
trials of hormonal agents, ODAC members noted that large numbers of patients
would be needed to adequately measure noninferiority using the new standard,
but that patients with bone-only disease could be included. Also, the endpoint
of time to progression could reflect the potential benefit of prolonged stable
disease, the panel said.