Treatment margins for prostate brachytherapy

The purpose of this article was to determine what planned treatment margin (TM) would allow for implant-related prostate volume changes and still achieve an adequate periprostatic cancercidal dose. Twenty consecutive, unselected patients who underwent (125)I implantation (144 Gy prescription dose) were studied. The treated volume (TV) was calculated as the volume encompassed by the 144 Gy isodose distribution. A post-implant computed tomography scan was obtained the following day, using 5-mm images at every 5 mm. The distances between the prostate margin (GTV) and TV were determined by measuring the distance between the ultrasound-defined prostatic margin and the prescription isodose, perpendicular to the prostatic margin. The lateral, anterior, and posterior TM margins were determined at the base, mid-level, and apex of the prostate. The pre-implant TV was nearly twice as large as the GTV, ranging from 36 to 199 mL (median, 73 mL). The anterior, lateral, and posterior planned TMs varied substantially between patients, due to lack of a consistent policy the magnitude of the CTV and the acceptable CTV-to-TV distance. For all measurement points, the median planned treatment margin was 3 mm (range, -16 mm to 14 mm). Overall, there was only a loose correlation between pre- and post-implant treatment margins primarily due to variable, implant-related prostatic dimensional changes. Patients with a greater implant-related volume increase tended to have smaller post-implant treatment margins. The post-implant TMs were negatively correlated with dimensional changes, and the negative correlation was most marked for the anterior and posterior TMs due to predominant anterior-posterior dimensional increase. As expected, the post-implant target coverage was higher when larger planning TMs were used, but the correlation was loose due to the unpredictable, highly variable degree of implant-related volume increase. We currently are using 5-mm TMs around the GTV, as identified on pre-implant transrectal ultrasonography or computed tomography. However, the poor correlation between planned and actual post-implant TMs call into question any attempt to make a rational recommendation regarding optimal TMs.