Purpose: To determine if, for prostate bed treatment, generating synthetic CTs (sCT) from one daily cone-beam CT (CBCT) per week represents the average dose delivered to the patient over the full course of treatment.
Methods: The planning CT plus 34 daily CBCT scans of a post-operative prostate bed patient were anonymized and sCTs were generated. The planning and organ at risk (OAR) volumes were deformed to the new sCT and checked for consistency. The treatment plan was recalculated on each of sCT. Single fraction Dose Volume Histogram (DVH) data was recorded for AnoRectum, Bladder, Body, Bowel, Femoral Heads, and Planning Target Volumes including prostate bed and pelvic lymph nodes. DVH data for all 34 fractions sCTs were averaged to get the treatment average daily dose to targets and OARs. Average DVHs were calculated for subsets of the sCT DVH data including for the first 7 days, the last 7 days, and every fifth day starting at day 1, 2, 3, and 4. ANOVA tests used to determine if the subgroup averages differed significantly from the average over all 34 fractions.
Results: According to the ANOVA tests, which compared the subgroup averages to the total daily averages, the subset are not statistically different than the total average. For each structure, the p value was greater than 0.64.
Conclusion: The recalculated DVH data for any of the checked 7 day subsets of the 34 sCTs were not significantly different than the average of the entire 34 fractions, so a once weekly subset appears to be an acceptable alternative to using the whole data set.