Long-term outcomes of radical dose escalation by high-dose-rate brachytherapy for localized prostate cancer: Comparison of 3 different risk stratifications according D'amico, Zelefsky, and Kiel protocol

Purpose/Objective(s): Radical dose escalation >74 Gy has been shown to be superior over conventional radiation in a series of randomized trials. High-dose-rate brachytherapy (HDR-BT) is an established method to escalate the dose to very high equivalent dose at 2 Gy (EQD2). Long-term results are traditionally reported according risk stratifications. However, there is specifically in this context no “golden standard” in the literature. This study aimed to evaluate the 5- and 8-year outcomes following radical dose escalation using HDR-BT reported in comparison of 3 different risk stratification definitions. Materials/Methods: Pelvis external beam radiation therapy (P-EBRT) was performed in combination with a HDR-BT-boost to total EQD2 >100 Gy (α/β =3) in 459 consecutively treated and analyzed patients with a minimum follow-up (FU) of 24 months in 3 subsequent protocols that were based constructively on each other. The recorded endpoints were ASTRO biochemical control (BC) rates according to D'Amico, Zelefsky, and Kiel Protocol risk stratification definitions (RSD). All endpoints were compared for each RSD across different risk strata (low, intermediate, and high) using Kaplan-Meier and log-rank test. Results: Mean follow-up was 77 months. For D'Amico classification and across the protocols the 5/8-year ASTRO BC survival rates were between 89.4% and 100% for low risk (LR) (P=.776), 69.7% to 84.6% for intermediate risk (P=.349) (IR), and 59.9% to 83.4% for high risk (HR) (P=.037), respectively. For Zelefsky classification the 5/8-year ASTRO BC survival rates were 83.3% to 100% for LR (P=.897), 61.6% to 74.3% for IR (P=.231), and 52.1% to 70.6% for HR (P=.431). Finally, for the Kiel Protocol classification the 5/8-year ASTRO BC survival rates were 76.6% to 96.2% for LR (P=.32), 58% to 81% (P=.159), and 49.1% to 78.9% (P=.053). Conclusion: Comparing these differences across the used risk stratification methods the results according D'Amico, Zelefsky, and Kiel Protocol seem to not differ in low-risk cohorts. However in IR and HR the D'Amico method appears to overestimate the outcomes. In contrast, the Kiel Protocol definition underestimates the long-term results, while the Zelefsky definition takes an intermediate role in the method comparison. However, only the D'Amico method discriminated statistically significant between the protocols in the high-risk group. The Kiel method showed in the HR cohort a marginal significance.