By Alexandra Goho
An Imperfect Substitute
The increased use of surrogate endpoints in cancer drug trials is raising both hopes and fears
By Alexandra Goho
Illustration by Nicolle Rager Fuller
In 2004, the U.S. Food and Drug Administration (FDA) approved a new drug for metastatic colorectal cancer that was unlike any on the market. Called Avastin (bevacizumab), the angiogenesis inhibitor, which targets a tumor’s nourishing blood vessels, represented a new approach to treating cancer. Two years later, the FDA went on to approve the drug for the treatment of lung cancer.
But no sooner had Avastin become a blockbuster than it came under fire, in February 2008, when the FDA approved it for the treatment of patients with advanced breast cancer. The agency’s decision was at odds with its own advisory committee, which voted 5–4 against recommending the drug for approval. Avastin, the majority pointed out, did not extend women’s lives in clinical trials. Nonetheless, the FDA gave the drug the nod because it temporarily stopped the disease from worsening.
An improvement in patient survival would certainly be the most ideal demonstration of Avastin’s—or any new cancer therapy’s—effectiveness. “There’s no question in anyone’s mind that we would prefer overall survival as a gold standard,” says Indiana University oncologist George Sledge, who presided over the first phase II trial of Avastin for the treatment of breast cancer at the university’s Melvin and Bren Simon Cancer Center, in Indianapolis. “However, these tests can cost hundreds of millions of dollars and take several years.” What’s more, enormous numbers—sometimes thousands—of patients frequently need to be recruited. The process and its requirements, many drug companies and patient advocates worry, could delay the delivery of potentially effective and life-saving therapies to patients.
To hasten the approval of new drugs, researchers are increasingly exploring the use of alternate indicators of a drug’s effectiveness. These so-called surrogate endpoints, such as tumor shrinkage or the concentration of a particular molecule in the blood, can be evaluated more quickly than patient survival and require fewer study participants. And they allow researchers to rapidly draw conclusions about a drug’s effectiveness without waiting to directly measure a benefit to the patient, such as improved survival or quality of life.
Rather than using patient survival rates to evaluate the effectiveness of a new drug, many studies use yardsticks such as tumor-response rate, disease-free progression, or the amount of a specific molecule in the blood. Above, a tumor (lower left corner) grows and spreads cells (purple) into a blood vessel, while a tumor biomarker—prostate-specific antigen, or PSA (blue)—travels through the blood stream.
When the FDA was considering Avastin’s suitability as a treatment for advanced breast cancer, the agency looked at a surrogate endpoint known as progression-free survival—the length of time during and after treatment when a patient’s disease does not worsen. Women taking Avastin demonstrated longer progression-free survival than women who didn’t take the drug. But considering Avastin’s significant side effects, which include heart failure and gastrointestinal perforation in some patients, many people question whether such a surrogate is truly an adequate yardstick for approving a new drug. In fact, researchers sometimes discover that a surrogate endpoint doesn’t correlate at all with a clinically useful benefit like overall survival.
“Everyone is interested in shortening the time to getting drugs on the market,” says Richard Pazdur, the director of the Office of Oncology Drug Products within the FDA’s Center for Drug Evaluation and Research. “But the FDA and other regulatory agencies really have to have confidence that there is a real relationship between these surrogates and a direct clinical benefit.”