Prostate-specific Antigen Bounce After Stereotactic Body Radiotherapy for Prostate Cancer: A Pooled Analysis of Four Prospective Trials

Aims: We conducted a pooled analysis of four prospective stereotactic body radiotherapy (SBRT) trials of low- and intermediate-risk prostate cancer to evaluate the incidence of prostate-specific antigen (PSA) bounce and its correlation with the time-dose-fraction schedule. The correlation between bounce with PSA response at 4 years (nadir PSA < 0.4 ng/ml) and biochemical failure-free survival (BFFS) was also explored.

Materials and methods: The study included four treatment groups: 35 Gy/five fractions once per week (QW) (TG-1; n = 84); 40 Gy/five fractions QW (TG-2; n = 100); 40 Gy/five fractions every other day (TG-3; n = 73); and 26 Gy/two fractions QW (TG-4; n = 30). PSA bounce was defined as a rise in PSA by 0.2 ng/ml (nadir + 0.2) or 2 ng/ml (nadir + 2.0) above nadir followed by a decrease back to nadir. Patients with fewer than three follow-up PSA tests were excluded from the pooled analysis.

Results: In total, 287 patients were included, with a median follow-up of 5.0 years. The pooled 5-year cumulative incidence of bounce by nadir + 2.0 was 8%. The 2-year cumulative incidences of PSA bounce by nadir + 0.2 were 28.9, 21, 19.6 and 16.7% (P = 0.12) and by nadir + 2.0 were 7.2, 8, 2.7 and 6.7% (P = 0.32) for TG-1 to TG-4, respectively. Multivariable analysis revealed that for nadir + 2.0, pre-treatment PSA (odds ratio 0.49; 95% confidence interval 0.26-0.97) correlated with PSA bounce. Although PSA bounce by nadir + 0.2 (odds ratio 0.10; 95% confidence interval 0.04-0.24) and nadir + 2.0 (odds ratio 0.29; 95% confidence interval 0.09-0.93) was associated with a lower probability of PSA response at 4 years, there was no association between bounce by nadir + 0.2 (hazard ratio 0.36; 95% confidence interval 0.08-1.74) or nadir + 2 (hazard ratio 1.77; 95% confidence interval 0.28-11.07) with BFFS.

Conclusion: The incidence of PSA bounce was independent of time-dose-fraction schedule for prostate SBRT. One in 13 patients experienced a bounce high enough to be misinterpreted as biochemical failure, and clinicians should avoid early salvage interventions in these patients. There was no association between PSA bounce and BFFS.