Impact of short course hormonal therapy on overall and cancer specific survival after permanent prostate brachytherapy.

PURPOSE

To review the impact of prior hormonal therapy on 10-year overall and prostate cancer specific survival after primary brachytherapy.

METHODS AND MATERIALS

A retrospective review was performed on the Arizona Oncology Services tumor registry for 2,378 consecutive permanent prostate brachytherapy cases from 1988 through 2001. Hormonal therapy was administered before the implant in 464 patients for downsizing of the prostate or at the discretion of the referring physician. All deceased patients with known clinical recurrence were considered to have died of prostate cancer, irrespective of the immediate cause of death. Risk groups were defined, with 1,135 favorable (prostate-specific antigen [PSA] < 10, Gleason < 7, Stage T1-T2a), 787 intermediate (single adverse feature), and 456 unfavorable (two or more adverse features) patients. Kaplan-Meier actuarial survival curves were generated for both overall and cause-specific survival from the time of treatment. Multivariate analysis was performed to assess the impact of hormonal intervention in comparison with known risk factors of grade, PSA, and age.

RESULTS

With follow-up ranging up to 12.6 years and a median of 4.1 year, a total of 474 patients died, with 67 recorded as due to prostate cancer. Overall and cause-specific 10-year survival rates are 43% and 88%, respectively. Overall survival is 44% for the hormone naive patients, compared with 20% for the hormone-treated cohort (p = 0.02). The cancer-specific survival is 89% vs. 81% for the same groups (p = 0.133). Multivariate analysis confirms the significance of age > 70 years (p = 0.0013), Gleason score >/= 7 (p = 0.0005), and prior hormone use (p = 0.0065) on overall survival.

CONCLUSIONS

At 10 years, in prostate cancer patients receiving brachytherapy, overall survival is worse in men receiving neoadjuvant hormonal therapy, compared with hormone naive patients. This does not appear to be due to other known risk factors for survival (i.e., stage, grade, PSA, age) on multivariate analysis. The leading causes of death were cardiovascular, prostate cancer, and other cancers with no obvious discrepancy between the two groups. This finding is unexpected and requires confirmation from other centers.