One Year Radiographic Response Following Prostrate SBRT: An Exploratory Analysis of a Phase III Randomized Trial

Purpose/objective(s)Radiographic MRI response following prostate radiotherapy, particularly stereotactic body radiotherapy (SBRT), remains poorly understood. Our objective was to describe radiographic changes to the prostate gland and prostate tumor following SBRT of men treated on a prospective, randomized trial.Materials/methodsMIRAGE (NCT04384770) is a single center, randomized phase III trial of patients receiving either CT or MRI guided SBRT for localized prostate cancer. Patients underwent pre-treatment and annual post-treatment MRIs, in addition to routine PSA surveillance. Outcomes reported include percent gland shrinkage, percent PSA response at one year, and presence of residual tumor based on radiographic interpretation. Patient characteristics were compared via two-sample t-test or Fischer's exact test. Both univariate and multivariable logistical analysis were employed to identify potential clinical predictors of residual tumor on 1-year follow up MRI.ResultsThis study cohort included 94 eligible patients with baseline characteristics in Table 1. Residual lesions were seen in 13 patients (14%), 5/27 (18.5%) treated without ADT and 8/67 (12%) with ADT. PSA ablation was deep, with a 79% median decrease without ADT and 98% median decrease with ADT. Patients receiving ADT showed more gland shrinkage (17% vs. 34% shrinkage, p = 0.0001), while radiographic non-responders and responders experienced similar gland shrinkage (median 21% vs 29% shrinkage, p > 0.05). No significant clinical predictors of residual tumor were identified on univariate and multivariate analysis. No patient had any clinical or biochemical evidence of failure.ConclusionA total of 14% of patients were found to have residual tumor detected on MRI one year after SBRT. These data highlight the protracted nature of radiographic tumor response to radiation therapy, even with ablative radiation techniques. The analysis is limited by the lack of biopsy data to quantify whether visualized residual tumor harbor active cancer.