April 19, 2006 -- A new study from Columbia University Medical Center researchers who are members of the Southwest Oncology Group (SWOG), suggests that certain changes in prostate-specific antigen (PSA) levels may serve as surrogate endpoints for prostate cancer survival.
Doctors, drug companies and cancer patients all look to speed up the process of clinical trials. Researchers have suggested that for prostate cancer, changes in PSA level could be used as a biomarker to measure how well a treatment works overall.
Currently, the U.S. Food and Drug Administration accepts only survival as an endpoint of measure. Survival as a primary endpoint was used in phase III studies of novel chemotherapeutic drugs for men with androgen-independent prostate.
For this new study of surrogate endpoints Daniel P. Petrylak, M.D., an associate professor of medicine who directs the genitourinary oncology program at New York-Presbyterian/Columbia Hospital, led a research team in analyzing results of 551 men with prostate cancer treated in the Southwest Oncology Group Protocol S99-16. This was a Phase III clinical trial that randomized men with hormone-refractory metastatic prostate cancer to take either Taxotere and Estramustine or Mitoxantrone and Prednisone. The trial found that Taxotere and Estramustine increased survival.
When Petrylak's team looked at the trial results, they saw patterns in changes in the patients' PSA levels. These changes matched the patients' outcomes during the trial in way which, the researchers concluded, could possibly serve as surrogate endpoints for survival.
The risk of death, in men whose serum PSA levels declined by at least 30 percent in the first three months of treatment was reduced more than 50 percent.
In an editorial in the Journal of the National Cancer Institute, Stuart Baker says that Petrylak's team's findings must be viewed with caution and taken as tentative since they analyzed only the one clinical trial. "The growing view," Baker says, "is that measures to validate surrogate endpoints should be based on multiple trials involving surrogate and true endpoints." Petrylak's group, Baker says, "correctly emphasized" that their findings may apply "to only one class of drugs." Only with additional data," Baker says, "will it be possible to determine if surrogate endpoints are closer to wishful thinking or reality."
"The findings show that PSA levels can be a reliable endpoint measure of prostate cancer treatment efficacy," said Dr. Petrylak. "However, this and other candidate surrogate endpoints must be validated in independent clinical trials of men with prostate cancer."
