SBRT in the Post-Prostatectomy Setting: Analysis of Interfraction Variability

Abstract

Our institution has a phase 1 clinical trial evaluating acute toxicity in the setting of SBRT to the prostate fossa following prostatectomy. Patients enrolled on this trial are simulated and treated with fiducial markers in the prostate fossa, a rectal balloon and a daily ultrasound (US) of the bladder for immobilization and target visualization. This study reports the interfraction motion analysis based on the described immobilization technique. At this point, the study has enrolled 12 patients onto dose levels 1 and 2 (3.6Gy/fraction x 15 fractions and 4.7Gy/fraction x 10 fractions, respectively). All patients had a rectal balloon for consistency in rectal volume and immobilization of the prostate bed. Patients were treated with a full bladder that was within 20% of the bladder volume on the day of CT simulation as verified by daily US. All patients were implanted with three gold fiducial markers into the prostate bed. KV and CBCT were performed on day 1 to verify fiducial placement and bladder and rectal volume. Daily kV and weekly CBCT were used for daily alignment based on fiducial markers. For 10 out of 12 patients, the patients’ shift data were gathered and a shift vector was calculated. The mean shifts for this cohort of the patients was -0.2 mm, 0.4 mm, -1.1 mm in anterior and posterior, superior and inferior, and left and right, respectively. The standard deviation was 4.0 mm, 3.7 mm, -3.4 mm in anterior and posterior, superior and inferior and left and right, respectively. It was demonstrated that the current bladder filling and IGRT protocol resulted in minimal systematic shift (small mean shifts). Most shifts were random which were reflected in the standard deviation. The mean and standard deviation of the shift vector was 5.3 mm and 3.9 mm, respectively. The calculated percentage bladder volume compared to bladder volume during the CT simulation had a mean of 3.6% and standard deviation of 8.9%. No correlation was found between magnitudes of patient shifts and bladder filling for this cohort of patients. With the current rectal/bladder immobilization and IGRT protocol, this cohort of the patients had a mean shift of no more than 1.1 mm but a relatively large standard deviation of 4.0 mm in each direction. Setup variation did not follow a specific pattern based on bladder or rectal filling. For the subset of the patients who have reasonable CBCT volumetric information, dosimetric analysis will be performed by deforming planning CT and dose to CBCT using VelocityTM.