Application of dose compensation in image-guided radiotherapy of prostate cancer.

In image-guided radiation therapy (IGRT), volumetric information on patient anatomy at treatment conditions is made available with in-room imaging devices capable of cone-beam CT. Setup error and inter-fraction rigid motion can be corrected online. The planning margin can therefore be reduced significantly. However, to compensate for uncertainties including organ deformation and intra-fraction motion, offline evaluation and replanning are necessary. The purpose of this study is to investigate the use of an offline dose compensation technique to further reduce the margin safely. In IGRT, online CT scan, rigid image registration and setup correction are performed at each fraction. Later the regions of interest are registered offline between treatment and planning CTs using a finite element method to account for non-rigid organ motion. Cumulative dose distribution is calculated and compared with the prescription dose. The discrepancy, if found significant, is repaired using the dose compensation technique, in which the cumulative dose distribution is incorporated in adaptive IMRT planning for future fractions. Two compensation schedules were tested in this study: single compensation at the end of the treatment course and compensation performed weekly. One patient with one planning CT and 16 treatment CTs were used in this simulation study. Due to the aggressive smaller planning margin used, severe underdose was observed in the clinical target volume. The size and magnitude of the underdose were reduced substantially with online guidance but were still significant. Both dose compensation strategies were able to reduce the dose deficit to an acceptable level without additional planning margin. Weekly compensation is more biologically beneficial and can spread the execution error into multiple fractions. The offline dose compensation technique allows further margin reduction and can complement the online guidance by compensating for uncertainties that cannot be reduced online, thereby increasing the confidence in IGRT delivery.

Duke Scholars

Author Qiuwen Wu Radiation Oncology