abstract
- BACKGROUND AND PURPOSE: It is necessary to include the entire prostate in the high dose treatment volume when planning radical radiation for patients with prostate cancer. We prospectively compared magnetic resonance imaging (MRI) to computed tomography (CT) and urethrography as means of localizing the prostatic apex. MATERIALS AND METHODS: Thirty patients with clinically localized prostate cancer had a sagittal T2-weighted MRI scan and a conventional axial CT scan performed in the treatment position prior to the start of radiotherapy. Twenty of these patients had a static retrograde urethrogram performed at simulation. The position of the MRI and CT apices were localized independently by two radiation oncologists. In addition, the MRI apex was localized independently by a diagnostic radiologist. The urethrogram apex, defined as the tip of the urethral contrast cone, was easily identified and was therefore localized by only one observer. RESULTS: There was good interobserver agreement in the position of the MRI apex. Interobserver agreement was significantly better with MRI than with CT. There were no systematic differences in the position of the MRI and CT apices. However, the MRI apex was located significantly above and behind the urethrogram apex. There was poor correlation between MRI and CT and between MRI and urethrogram in the height of the apex above the ischial tuberosities. There was 83% agreement between MRI and CT and 80% agreement between MRI and urethrogram in the identification of patients with a low-lying apex. The apex, as determined by MRI, was <2 cm above the ischial tuberosities and therefore potentially under-treated in 17% of the patients. CONCLUSIONS: MRI is superior to CT and urethrography for localization of the prostatic apex. All patients undergoing radiotherapy for prostate cancer should have localization of the apex using MRI or a technique of equal precision to assure adequate dose delivery to the entire prostate and to minimize the unnecessary irradiation of normal tissues.