SAN ANTONIOIn a retrospective study of prostate cancer patients with a rising PSA after external beam radiation therapy, pretreatment factors appeared to have little independent influence on the development of metastatic disease. However, two post-treatment factors, PSA doubling time and time to onset of a rising PSA, were found to be major determinants of metastatic relapse, Lewis G. Smith III, MD, reported at the 41st Annual Scientific Meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO).
Of note, no patient developed distant metastasis without a rising PSA level, said Dr. Smith, assistant professor of radiation oncology, M.D. Anderson Cancer Center.
Patients in the cohort represented an average spectrum of patients who received radiation therapy during the 1980s and 1990s for prostate cancer, Dr. Smith said. He added that even those who developed a rising PSA still had fairly good outcomes. Their overall survival was the same as predicted survival, and I think that is a key significant point, he said.
The parent population consisted of 1,119 patients, all treated between 1987 and 1997. All patients had T1 to T3 disease, and their pretreatment PSA levels ranged from 0.3 to 150 ng/mL (median, 8.4 ng/mL).
The external beam radiation dose varied from 60 to 78 Gy (median, 68 Gy). None of the patients had received androgen ablation at the time of initial therapy. Follow-up ranged from 12 to 146 months (median, 59 months).
Of this parent population, there were 357 patients who developed a rising PSA level (the test cohort). The time to onset of the rise in PSA was from 1 to 100 months (median, 20 months). Follow-up after the onset of the rise of PSA ranged from 3 to 132 months (median, 53 months).
Treatment for the rising PSA was at the discretion of the investigating physician: 91 patients received no additional treatment; 216 patients were given androgen ablation (181 for a rising PSA and 35 who were diagnosed with distant metastasis at the time of the onset of the rise in PSA); 42 patients underwent cryosurgery; and 8 patients underwent prostatectomy.
Local control was a main issue for the test patients, Dr. Smith said. At 5 years, 50% had failed locally, and at 8 years, 59% had failed locally.
Freedom from distant metastasis was 87% at 5 years, and 69% at 8 years. And this is from the point of the rise in PSA, not from the time of treatment, he said.
On first analysis of the information, it appeared that stage, Gleason score, and pretreatment PSA were going to be of importance. However, when we re-analyzed data with 8 years of follow-up and added a few more patients and looked more thoroughly, we found that we really couldnt find any pretreatment factors that, under multivariate analysis, would indicate a predictor for distant metastasis, Dr. Smith said.
On multivariate analysis for distant metastasis, only two factors were found to be significant: PSA doubling time and onset of a rise in PSA. PSA doubling time of greater than 12 months was favorable for this group of patients, and the onset of a rise in PSA beyond 12 months was also favorable.
The researchers analyzed data for the patients untreated for rising PSA as well as for the patients with distant metastases at the time of rising PSA. Basically, the worst case and the best case scenarios, Dr. Smith said. The untreated patients had a median PSA doubling time of 23 months vs 6 months for the metastatic patients. And the median onset of a rise in PSA following treatment was 20 months for the untreated patients vs 10 months for those patients who were diagnosed with metastatic disease.
For the patients who received early androgen ablation for a rising PSA, two factors were predictive of distant metastasis: Gleason score and time to rise in PSA. Additionally, the PSA doubling time was bordering on significance, he said.
On multivariate analysis, only the time to onset of the rise in PSA was significant in these patients. Those who had a rise in PSA 12 months or longer after treatment had roughly an 82% freedom from distant metastasis at 8 years, while those patients whose PSA level began to rise less than a year from the time of initial treatment had only a 27% probability of being free from distant metastasis at 8 years.