Assessing prostate, bladder and rectal doses during image guided radiation therapy--need for plan adaptation?

The primary application of Image Guided Radiotherapy (IGRT) in the treatment of localized prostate cancer has been to assist precise dose delivery to the tumor. With the ability to use in-room Computed Tomography (CT) imaging modalities, the prostate, bladder and rectum can be imaged before each treatment and the actual doses delivered to these organs can be tracked using anatomy of the day. This study evaluates the dosimetric uncertainties caused by inter-fraction organ variation during IGRT for 10 patients using kilovoltage cone beam CT (kvCBCT) on the Elekta Synergy system and Megavoltage CT (MVCT) on the Tomotherapy Hi-ART system. The actual delivered doses to the prostate, bladder and rectum were based on dose recomputation using CT anatomy of the day. The feasibility of dose calculation accuracy in kvCBCT images from the Elekta Synergy system was investigated using the ComTom phantom. Additionally, low contrast resolution, image uniformity and spatial resolution between the three imaging modalities of kilovoltage CT (kvCT), kvCBCT and MVCT images were quantitatively evaluated using the Catphan 600 phantom. The Planned Adaptive software was used on the Tomotherapy Hi-ART system to construct a cumulative Dose Volume Histogram (DVH) incorporating anatomical information provided by the daily MVCT scans. The cumulative DVH was examined to identify large deviation (10 % or greater) between the planned and delivered mean doses. The study proposes a framework that applies the cumulative DVH to evaluate and adapt plans which are based on actual delivered doses. Due to the large deviation in CT number ( 300 HU) between the kvCBCT images and the kvCT, a direct dose recomputation on the kvCBCT images from the Elekta Synergy system was found to be inaccurate. The maximum deviation to the prostate was only 2.7% in our kvCBCT study when compared to the daily prescribed dose. However, there was a large daily variation in rectum and bladder doses based on the anatomy of the day. The maximum variation in rectum and bladder volumes receiving the percentage of prescribed dose was 12% and 40% respectively. We have shown that by using Planned Adaptive software on the Tomotherapy Hi-ART system, plans can be adapted based on the image feedback from daily MVCT scans to allow the actual delivered doses to closely track the original planned doses.