Comparison of MRI pulse sequences in defining prostate volume after permanent implantation

Abstract

Purpose : To determine the relative value of three MRI pulse sequences in defining the prostate volume after permanent implantation. Methods and Materials : A total of 45 patients who received a permanent 125I implant were studied. Two weeks after implantation, an axial CT scan (2 mm thickness) and T 1-weighted, T 1-weighted fat saturation, and T 2-weighted axial MRI (3-mm) studies were obtained. The prostate volumes were compared with the initial ultrasound planning volumes, and subsequently the CT, T 1-weighted, and T 1-weighted fat saturation MRI volumes were compared with the T 2-weighted volumes. Discrepancies in volume were evaluated by visual inspection of the registered axial images and the registration of axial volumes on the sagittal T 2-weighted volumes. In a limited set of patients, pre- and postimplant CT and T 2-weighted MRI studies were available for comparison to determine whether prostate volume changes after implant were dependent on the imaging modality. Results : T 1-weighted and T 1-weighted fat saturation MRI and CT prostate volumes were consistently larger than the T 2-weighted MRI prostate volumes, with a volume on average 1.33 (SD 0.24) times the T 2-weighted volume. This discrepancy was due to the superiority of T 2-weighted MRI for prostate definition at the following critical interfaces: membranous urethra, apex, and anterior base-bladder and posterior base-seminal vesicle interfaces. The differences in prostate definition in the anterior base region suggest that the commonly reported underdose may be due to overestimation of the prostate in this region by CT. The consistent difference in volumes suggests that the degree of swelling observed after implantation is in part a function of the imaging modality. In patients with pre- and postimplant CT and T 2-weighted MRI images, swelling on the T 2-weighted images was 1.1 times baseline and on CT was 1.3 times baseline, confirming the imaging modality dependence of prostate swelling. Conclusion : Postimplant T 2-weighted MRI images provided superior prostate definition in all critical regions of the prostate compared with CT and the other MRI sequences tested. In addition to defining an optimal technique, these findings call two prior observations into question. Under dosing at the anterior base region may be overestimated because of poor definition of the prostate-bladder muscle interface. The swelling observed after implantation was lower on T 2-weighted images as well, suggesting that a fraction of postimplant swelling is a function of the imaging modality. These findings have implications for preimplant planning and postimplant evaluation. As implant planning techniques become more conformal, and registration methods become more efficient, T 2-weighted MRI after implantation will improve the accuracy of postimplant dosimetry.