2268: MRI And MRSI Following Prostate Brachytherapy: Is it of Value?

Abstract

The use of Magnetic Resonance Imaging (MRI) and Magnetic Resonance Spectroscopy Imaging (MRSI) following radiation treatment have been reported previously. We attempted to reproduce the findings of these previous reports. Eligibility criteria for this study included patients treated with prostate brachytherapy alone at least five years ago. Institutional Research Board approval and informed patient consent were obtained prior to imaging. A 1.5 Tesla MRI with standard endorectal coil was used to acquire axial, coronal, and transverse T1-weighted and T-2 weighted images and MRSI. The volume for MRSI was selected to maximize coverage of the prostate while minimizing coverage of periprostatic fat and rectal air. Automated and manually shimming was used to optimize homogeneity of the magnetic field. 1024 voxels were obtained per patient. Each voxel contained one signal. A single radiologist interpreted MRI images and a single physicist and radiologist interpreted MRSI data independently. Each reviewer was blinded to patient information. Fifteen men were accrued. Median follow-up from the time of prostate brachytherapy is 81 months. The median pretreatment PSA was 8.1 (range 5.4–13.9). Twelve patients were stage T1c, and all but two patients had a Gleason score of 6. All men were treated using I-125 permanent interstial seeds alone to a dose of 145 Gy. The median post-treatment PSA is 0.04 ng/ml. (range 0.01–1.19). Three patients have biochemical recurrence by ASTRO consensus definition with no known metastasis and continue to be observed. No man developed complications following MRI/MRSI. For all men, the peripheral zone of the treated prostate on T2-weighted images was relatively thin and only slightly higher in signal intensity than the central gland. On T1-weighted imaging, the treated prostate is of uniform signal intensity with no differentiation between the zones. One patient with biochemical recurrence had a vague mass like area of low signal intensity suspicious for tumor. The MRI results of the remaining men were interpreted as normal. Fifteen MRSI data sets were collected but ten were found to be uninterpretable including the patient with an abnormal MRI and biochemical failure. No abnormal spectra suspicious for tumor activity are appreciated in the 5 men with interpretable MRSI. MRI and MRSI with endorectal coil following LDRPB alone is feasible in our hands but a readable MRSI was achieved in a minority of patients on this study. MRI, specifically axial T2 weighted imaging of the treated prostate, was not useful. Outside of a study, we do not recommend MRI and MRSI to evaluate for local recurrence.