High-Dose Rate Brachytherapy Alone for Treatment of Unfavorable Intermediate Risk Prostate Cancer: A Propensity-Score Matched Analysis.

Abstract

To demonstrate the feasibility of high-dose rate brachytherapy (HDR BT) as monotherapy for unfavorable intermediate risk (UIR) prostate cancer by comparing survival outcomes of HDR BT alone against external beam radiation therapy (EBRT) + HDR BT boost, +/- androgen deprivation therapy (ADT) using propensity-score matched (PSM) data. This retrospective study queried two data registries collecting patient data from 1991 to present. 633 patients with UIR prostate cancer treated with HDR BT alone, HDR BT+EBRT or HDR+EBRT+ADT were included. HDR BT patients received 42-45Gy/6 fractions (fx) or 27 Gy/2 fx. For HDR BT+EBRT, the HDR dose was 20-24 Gy/2 fx, 24 Gy/4 fx, or 15 Gy/1 fx. EBRT patients received 45 Gy/25 fx to the prostate +/- pelvic nodes. GU/GI toxicities were graded according to the Common Terminology Criteria for Adverse Events (CTCAE) v5.0. Time-to-event analyses were carried out to evaluate the relationship between treatments and five primary endpoints of interest: freedom from biochemical recurrence (FFBC), freedom from distant metastasis (FFDM), freedom from local failure (FFLF), cancer specific survival (CSS), and overall survival (OS) at 5 years. PSM was performed with one-to-n matching. Logistic regression was used to estimate the respective propensity scores. The five potential confounders identified were T-stage, Gleason score, pre-treatment PSA, age, and percent positive cores. Balance was checked using the standardized mean difference of covariates. Univariate and multivariate analyses were conducted on the matched data. Toxicity analysis was performed via association between a change in pre- and post-treatment GU/GI toxicity status and the treatment group, as well as incidence of post-treatment severe GI/GU toxicity (grade 3 or higher) and the treatment group. Univariate analysis with Kaplan-Meier method and log rank test comparison between the three cohorts demonstrated no significant difference in all survival outcomes FFBC, FFDM, FFLF, CSS, OS (p = 0.15, 0.19, 0.29, 0.57, 0.28, respectively). Multivariate analysis with Cox proportional hazard regression showed no differences in HR for FFBC and OS (p = 0.95, 0.11) with addition of EBRT, or with EBRT+ADT (p = 0.17, 0.24); no fit was obtainable for FFDM, CSS, FFLF. Toxicities between the three cohorts were not significantly different when comparing post-treatment and baseline GI/GU symptoms (p = 0.53/1). No Grade 2 or 3 GI toxicities were identified, while 8%/1% HDR patients, 10%/1% HDR+EBRT patients, and 12%/2% HDR+EBRT+ADT patients experienced Grade 2/3 GU toxicities. The incidence of grade 3 or higher GU toxicities between the three groups was not significantly different (p = 0.91). This propensity-score matched study demonstrates the feasibility of HDR BT alone for effective treatment of UIR prostate cancer when compared to HDR+EBRT or HDR+EBRT+ADT, while potentially minimizing the added toxicities of EBRT and the undesirable side effect profile of ADT.