Experience of ultrasound-based daily prostate localization

The NOMOS (Sewickley, PA) B-mode Acquisition and Targeting System (BAT) ultrasound system provides a rapid means of correcting for interfraction prostate positional variation. In this investigation, we report our experience on the clinical issues relevant to the daily use of the BAT system and the analysis of combined setup error and organ motion for 3509 BAT alignment procedures in 147 consecutive patients treated with IMRT for prostate cancer.

After setup to external skin marks, therapists performed the BAT ultrasound alignment procedure before each IMRT treatment. In this study, a single physician (A.C.) reviewed all BAT images and classified image quality and accuracy of image alignment by the therapist. On a scale of 1-3, near-perfect image quality or alignment was given a 1, fair image quality or misalignment > or = 5 mm (likely within the PTV) was given a 2, and unacceptable image quality or misalignment >5 mm (potential to violate the PTV) was given a value of 3. The distribution of shifts made was analyzed in each dimension and for all patients. The time required to perform the BAT alignment was also assessed in 17 patients.

Among the 3509 attempted BAT procedures, the image quality was judged to be poor or unacceptable in 5.1% (181). Of the remaining 3328 BAT images, with quality scores of 1-2, alignments were unacceptable (>5 mm misalignment as judged by the reviewing physician) in 3% (100). The mean shift in each direction, averaged over all patients, was 0.5-0.7 mm. Interfraction standard deviation (1 SD) of prostate position based on combined setup error and internal organ motion is 4.9 mm, 4.4 mm, and 2.8 mm in the anteroposterior (AP), superior-inferior (SI), and lateral (RL) dimensions, respectively. The distribution of the shifts was a near-random Gaussian-type in all three major axes, with greater variations in AP and SI directions. The percent of BAT procedures in which the shift was >5 mm was 28.6% in AP, 23% in SI, and 9% in RL directions. The average BAT procedure took extra 5 min out of a 20-min time slot in a typical eight-field IMRT treatment.

The quality of the daily ultrasound images was deemed acceptable in 95%. Major alignment errors by therapists were only 3%. The BAT system is clinically effective and feasible in a matter of 5 min. Although the accuracy of the BAT was not addressed in this investigation, we found a significant percentage of large shifts being made from the initial alignment position.