3D-ultrasound guided radiation therapy in the post-prostatectomy setting

Daily transabdominal ultrasound-directed localization has proven valuable in correcting for setup error and organ motion in the treatment of prostate cancer with three-dimensional conformal radiation therapy (3DCRT). The present study sought to determine whether this transabdominal ultrasound technology could also be reliably applied in the post-operative adjuvant or salvage setting to improve the reproducibility of coverage of the intended volumes and to enhance conformal avoidance of adjacent normal structures. Sixteen consecutive patients who received external beam radiotherapy underwent daily localization using an optically guided 3D-ultrasound target localization system (SonArray, Zmed, Inc., Ashland, MA). Six of the above patients were treated in a post-prostatectomy setting, either adjuvantly or for salvage, while the remaining 10 with intact prostates were treated definitively. Because the bladder neck generally approximates the postoperative prostatic fossa, it was used during ultrasound localization as the primary reference structure for the post-prostatectomy patients. For patients treated definitively, the prostate was the primary reference structure. Daily shifts were recorded and port films were taken weekly immediately after ultrasound-based repositioning. By comparing port films taken after ultrasound localization, which evaluates for both set-up error and internal shift, with the original digitally reconstructed radiographs (DRRs), which represents a zero clinical set-up error situation, the degree of variability in organ position was determined. The average absolute, ultrasound-based shifts from the clinical set-up position in the anterior/posterior, lateral, and cranial/caudal directions for the post-prostatectomy patients were 5 +/- 4 mm SD, 3 +/- 3 mm SD, and 3 +/- 4 mm SD over the entire course of treatment, respectively. The average vector length shift was 8 +/- 4 mm SD. For patients treated with an intact prostate, the analogous average absolute shifts in the anterior/posterior, lateral, and cranial/caudal directions were 4 +/- 3 mm SD, 4 +/- 3 mm SD, and 4 +/- 3 mm SD over the entire course of treatment. The average vector length shift was 7 +/- 4 mm SD. Vector length shifts representing interfraction internal motion were estimated by comparing post-ultrasound port films with DRRs. These were 5 +/- 3 mm SD and 4 +/- 4mm SD for post-prostatectomy and intact prostate patients, respectively. These ultrasound-based displacements were not statistically different in patients with an intact prostate versus patients post-prostatectomy (p > 0.1). In conclusion, daily transabdominal 3D-ultrasound localization proved to be a clinically feasible method of correcting for set-up and internal motion displacements. The bladder neck, which serves as an adequate localization reference structure for the prostatic fossa, could be readily ultrasound imaged and repositioned as necessary. Daily internal motion errors that would have occurred if only pre-treatment port films were used were similar in magnitude to those observed for the patients with intact prostates and were of sufficient magnitude to support the use of daily pre-treatment ultrasound localization in the post-prostatectomy setting.