Whole prostate D90 and V100: A dose–response analysis of 2000 consecutive 125I monotherapy patients

Abstract

To examine the relationship between whole prostate dose metrics and disease-free survival (DFS) after (125)I low-dose-rate prostate brachytherapy (LDR-PB). Data for the first 2000 LDR-PB monotherapy implants were extracted from a database containing patient, tumor, dosimetric, and outcomes information. By National Comprehensive Cancer Network criteria, half (n = 1006) had low-risk disease and half (n = 990) had intermediate-risk disease (four had high-risk disease). Most patients (58.4%) and 75.3% of intermediate-risk patients received 3 months neoadjuvant and 3 months concomitant androgen deprivation therapy (ADT). Univariate and multivariate analyses were conducted using recognized prognostic factors and the whole prostate dose metrics D90 (the minimum dose received by 90% of the postimplant CT-based prostate volume) and V100 (the percent of the postimplant CT-based prostate volume that received at least 100% of the prescription dose). The median followup is 5 years (maximum, 12.5 years); the 5-, 7-, and 10-year actuarial DFS estimates are 96.0%, 94.4%, and 93.0%, respectively. Of the recognized prognostic factors, only pretreatment prostate-specific antigen (p = 0.012) and Gleason sum (p = 0.010) were predictive of DFS. When analyzed as continuous variables, dose metrics were not predictive of DFS. However, most nonsignificant trends favored higher doses, and D90 values <130 Gy were predictive of an increased risk of recurrence in the non-ADT subset (N = 833; log rank, p = 0.018). Although D90 values of <130 Gy were predictive of an increased risk of recurrence in the non-ADT subset, neither D90 nor V100, when used as continuous variables, was predictive of DFS when applied to the entire cohort or in the subset analysis. This observation informs us that dose metrics are not equivalent to oncologic end points and must be calibrated against DFS for each physician and each institution offering LDR-PB.