FDA Expected to Approve Surrogate End Point for Neoadjuvant Breast Cancer Trials

Caroline Helwick

June 2013, Vol 4, No 5 - ASCO Annual Meeting


Chicago, IL—For years, the cancer research community has pushed for the use of surrogate end points in clinical trials as a means of hastening the drug approval process. These efforts will soon bear fruit, with the anticipated release by the US Food and Drug Administration (FDA) of its final guidance to drug manufacturers for accelerated drug approval for neoadjuvant breast cancer therapies. At the 2013 American Society of Clinical Oncology (ASCO) annual meeting, the speakers discussed the potential implications for researchers, providers, and patients.

Under the new guidance for the pharmaceutical industry, accelerated approval could be granted based on a surrogate end point that predicts for clinical benefit; such an approval would require subsequent confirmation in a clinical trial.

For neoadjuvant breast cancer, this end point would likely be patholog­­ic complete response (pCR) rate. Although achieving a pCR has been associated with superior outcomes in multiple studies, the correlation needs further validation, said Tatiana M. Prowell, MD, Breast Cancer Scientific Lead, Breast Oncology Group, FDA, and Assistant Professor of Oncology, Johns Hopkins, Baltimore, MD.

It is expected that pCR will be the surrogate end point in this setting, but other outcomes have been studied in patients with breast cancer, including tumor response on imaging, measurement of the tumor proliferation marker Ki-67, preoperative endocrine prognostic index, and residual cancer burden after treatment.

Faster Drug Approval to Most Benefit High-Risk Patients
The traditional drug approval process averages 10 to 15 years. The goal of accelerated approval is to provide early access to effective drugs and to provide incentives for developing drugs for particular breast cancer subtypes with unmet needs.

Under the new approval process, a single trial or a short-duration trial would be appropriate for accelerated approval, and a simultaneous longer trial would be conducted to support “regular” approval, according to speakers at a session at the ASCO meeting entitled “Pushing the Limits of Upfront Care and Drug Devel­opment: Neoadjuvant Opportuni­ties in Breast Cancer.”

The use of pCR to support accelerated approval is appropriate “for highly promising drugs and high-risk patient populations,” Dr Prowell maintained.

Session Chair Angela DeMichele, MD, MSCE, Co-Program Leader, Breast Cancer Program, Abramson Cancer Center, University of Penn­sylvania, Philadelphia, noted that there is a “fertile environment” for new drug development in the neoadjuvant breast cancer treatment setting, because the many benefits of neoadjuvant therapy are now established.

Moving clinical trials from adjuvant to the neoadjuvant setting can yield important information about drugs that are under evaluation, such as the ability of pharmacodynamic markers to provide in vivo evidence of biologic effects. But most important, this should identify active agents faster, which serves patients, Dr DeMichele said.

She emphasized that surrogate markers must be robust (ie, reproducible and standardized), and their correlation with the true end point, such as overall survival, must be more than a simple “correlation.” In this case, pCR must be strictly defined as the elimination of all invasive and noninvasive disease in the breast and lymph nodes, rather than a “looser” definition. Its relationship with the final end point can only be assured through long-term follow-up of patients, Dr DeMichele added.

She also proposed that pCR may not be a valid surrogate end point for every breast cancer. Although it may be appropriate for HER2-positive, triple-negative, and highly proliferative estrogen receptor (ER)-positive disease, it may not be a good surrogate for low-proliferation ER-positive tumors. At present, Dr DeMichele believes that Ki-67 is the best surrogate for the latter, although it is not perfect. Better surrogates are needed for this subset, she noted.