Post-treatment PSA Nadir Predicts Prostate Cancer Outcome

SAN FRANCISCO—Patients with lower PSA values after radiation therapy are more likely to be alive and free from distant metastasis 10 years later than those with higher values, according to study results presented at the 43rd Annual Meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO abstract 203).

PSA doubling time was also shown to be predictive of later death from prostate cancer, said Gerald Hanks, MD, who recently retired as director of the Department of Radiation Oncology, Fox Chase Cancer Center. Dr. Hanks, a coauthor of this study, reported the results on behalf of study statistician Alexandra Hanlon, PhD.

"This is the first report that demonstrates a very strong predictive power of postradiation PSA nadir for determining distant failure and death due to prostate cancer. These results shed light on clinical progression as a function of PSA nadir and PSA doubling time," Dr. Hanks said.

The researchers followed 615 men who were treated at Fox Chase with three-dimensional conformal radiation therapy for prostate cancer between April 1989 and December 1995. The median dose was 73 Gy (range, 64 to 82 Gy).

After a median follow-up period of 64 months, 186 of the patients had experienced biochemical relapse (using the ASTRO definition of three increases in PSA), 40 had developed distant metastasis, and 18 had died of their prostate cancer. Of the 186 biochemical failure patients, 48 received androgen-deprivation therapy.

Multivariate analysis revealed that several factors were extremely predictive of improved freedom from distant metastasis: lower post-treatment PSA nadir (P < .0001), longer interval to nadir from start of treatment (P = .0002), favorable tumor grade (P = .005), androgen-deprivation therapy after PSA failure (P = .001), and stage T1/T2A tumors (P < .01). Pretreatment PSA level, however, was not predictive of freedom from distant metastasis.

The 10-year incidence of freedom from distant metastasis was 96% for patients with post-treatment PSA nadirs of less than 1 ng/mL vs 89% and 49% for patients with values of 1.1 to 2 ng/mL and greater than 2 ng/mL, respectively (

P

< .0001).

Only post-treatment PSA nadir and time to achieve nadir from start of radiotherapy were predictive of death due to prostate cancer. At 10 years, cause-specific survival rates were 96% for patients with post-treatment PSA nadirs of less than 1 and between 1.1 and 2 ng/mL vs 78% for those with PSA nadirs greater than 2 ng/mL (P < .0001).

There was sufficient information to calculate PSA doubling times for 136 of the 186 patients who experienced biochemical failure. Doubling times were significantly predictive of freedom from distant metastasis (P = .0029), but there were insufficient data to determine if doubling time rates were related to the incidence of cause-specific survival.

Dr. Hanks acknowledged that these results may not have immediate application for patients who do not have access to clinical trials. He did, however, advocate that these results be used to direct high-risk patients to phase III trials. "The results may be used for the early identification of patients at high risk who may benefit from clinical trials directed at the metastatic disease that a certain proportion are fated to develop," he said.