Changes in prostate-specific antigen midway through salvage radiotherapy may be associated with long-term outcomes

Abstract

To examine the relationship between changes in prostate-specific antigen (PSA) by midpoint of salvage radiotherapy (SRT) for biochemically recurrent prostate cancer and long-term clinical outcomes. We conducted a retrospective study of men treated with SRT for biochemically recurrent prostate cancer at a single practice from 2004 to 2016. Patients were grouped based on PSA response at treatment midpoint: group 0, no change; group 1, decrease; group 2, increase. The primary endpoint was clinical failure measured from time of SRT completion and defined as serum PSA ≥ 0.2 ng/mL above post-radiotherapy nadir, initiation of androgen deprivation therapy, development of bone metastasis, or death from prostate cancer. The Kaplan-Meier method was used to estimate freedom from clinical failure for each group, and differences between groups were examined using pairwise multiple comparison. Median follow-up was 51.6 months (range, 3.3–138.0). Of 76 eligible men, 13.1% experienced no change in PSA at midpoint of SRT (group 0), 68.4% experienced a decrease (group 1), and 18.4% experienced an increase (group 2). Four-year freedom from clinical failure rates were as follows: group 0, 60.0%; group 1, 58.3%; and group 2, 41.7%. Median time to clinical failure was 71.7 months (95% confidence interval, 46.9–96.5) for group 1; 26.8 months (95% confidence interval, 0.0–55.9) for group 2; and was not reached for group 0. There was a significant difference in four-year freedom from clinical failure between groups 1 and 2 (p = 0.036). PSA changes by the midpoint of SRT predict long-term prostate cancer control with increases in PSA associated with decreased freedom from disease progression.