Application Of Cone-beam Computed Tomography To Radiotherapy For Head-and-neck Tumor, Thoracic Tumor And Abdominal Tumor

[PurPose]To investigate the optimal alignment methods and planning target volume(PTV) margins for the treatment of head-and-neck tumor,thoracic tumor and abdominal tumor. To analyze changes in parotid glands and submandibular glands dose resulting from anatomic changes throughout a course of image-guided radiotherapy in nasopharyngeal carcinoma( NPC). To evaluate the effect of the setup error on the physics dose distribution of peripheral organs at risk ( OAR) undergoing the intensity modulated radiation therapy for nasopharyngeal carcinoma, such as spinal cord,brain stem, optic nerve and optic chiasm , and to analyze their impact on treatment by using image-guided radiotherapy (IGRT).[Materials and Methods]62 patients were treated with the Elekta Synergy IGRT system,including 22 head and neck tumors , 20 thoracic tumor and 20 abdominal tumor. Head-and-neck cancer patients received on-line kilovoltage cone-beam computed tomography(KV-CBCT) scan during the first 5 fraction , and then in the remaining fractions KV-CBCT was Performed weekly before and after irradiation. thoracic tumor and Abdominal Tumor patients received on-line kv-cbct every fraction . The difference in bone alignment,grey value alignment and manual alignment was compared. The study also analyzed the setup errors and PTV margins. The study population consisited of 13 nasopharyngeal carcinoma patients treated definitively with intensity-modulated radiotherapy. The dynamic changing contouring of parotid glands and submandibular glands were manually delineated on each axial slice to investigate the anatomic changes.The original treatment plans were overlaid onto each registered CT images,and each hybrid plan was evaluated.The weekly and cumulative dose deviation from the original treatment plan were analyzed. Corrections between dosimetric and anatomic changes were investigated. The On- Borad- Imager was used to collect the images of Cone Beam Computerized Tomography ( CBCT) in realtime, which were matched with the positioning CT images before radiation treatment through target center. To calculate the errors in x, y and z direction. Then the treatment planning system was used to study the effects of the setup errors on the dose distribution undergoing the radiation therapy of spinal cord, brain stem, optic nerve and optic chiasm.[Results]Total of 229 sets of CBCT images were analyzed for 22 head-and-neck cancer. The mean±standard deviation in the X,Y and Z axis directions were 0.04±0.16cm,-0.05±0.15 cm,-0.03±0.14 cm ; 0.05±0.15 cm,-0.01±0.14 cm,-0.01±0.15 cm; 0.04±0.16 cm,-0.02±0.14 cm,-0.02±0.15 cm in head and neck cancer with bone alignment,grey value alignment and manual alignment respectively.The mean±standard deviation in the X,Y and Z rotation were 0.13±0.97°,-0.28±0.91°,0.36±0.83°; 0.20±1.87°,0.08±2.04°,0.09±2.06°; 0.13±1.09°,0.06±1.05°,0.09±1.12°in head and neck cancer with bone alignment,grey value alignment and manual alignment respectively.Under the condition of non-correction,the margins required to account for total error are 3.64 mm,3.68 mm,3.68 mm; 3.27 mm,3.28 mm,3.22 mm; 3.17 mm,3.37 mm,3.32 mm in X, Y, and Z axis directions respectively. When the tumor was corrected every fraction, the margins required to account for intrafraction error are 1.44mm,1.46 mm,1.49 mm; 1.50 mm,1.34mm,1.21mm; 1.59 mm,1.69 mm,1.47 mm in X, Y, and Z axis directions respectively. There was no statistical difference between the bone and gray value alignment methods, but there was significant difference between manual alignment and automatic alignment. Total of 384 sets of CBCT images were analyzed for 20 thoracic tumor. The mean±standard deviation in the X,Y and Z axis directions were 0.01±0.29cm,0.04±0.31 cm,-0.02±0.28 cm; 0.11±0.41 cm,0.12±0. 45 cm,0.09±0.41 cm; -0.11±0.23 cm,-0.05±0.22 cm,-0.08±0.23 cm in thoracic tumor with bone alignment , grey value alignment and manual alignment respectively.The mean±standard deviation in the X,Y and Z rotation were 0.61±1.09°,0.34±1.44°,0.66±1.28°; 0.06±0.83°,0.04±1.89°,0.16±1.6°; -0.17±1.45°,-0.19±1.53°,-0.13±1.45°in thoracic tumor with bone alignment,grey value alignment and manual alignment respectively.Under the condition of non-correction,the margins required to account for total error are 6.04mm,6.78mm,5.53mm; 9.89mm,8.78mm,8.93mm ; 5.81mm,6.22mm,5.89mm in X, Y, and Z axis directions respectively. When the tumor was corrected every fraction, the margins required to account for intrafraction error are 2.97 mm,3.57 mm,1.62 mm; 2.9mm,4.14 mm,1.74mm; 2.37 mm,3.53 mm,和1.7 mm in X, Y, and Z axis directions respectively. There was significant difference between three alignments . Total of 564 sets of CBCT images were analyzed for 20 abdominal tumor. The mean±standard deviation in the X,Y and Z axis directions were -0.02±0.27cm,-0.04±0.28cm,-0.03±0.27cm ; 0.08±0.46cm,0.11±0.48cm,0.09±0.48 cm; -0.01±0.28 cm,-0.02±0.28cm,-0.02±0.28cm in abdominal tumor with bone alignment,grey value alignment and manual alignment respectively.Under the condition of non-correction,the margins required to account for total error are 6.04mm,6.78mm,5.53mm; 9.89mm,8.78mm,8.93mm; 5.81mm,6.22mm,5.89mm in X, Y, and Z axis directions respectively. The mean±standard deviation in the X,Y and Z rotation were 0.57±1.78°,0. 73±2.17°,0.67±1.98°; 0.15±1.24°,0.12±1.51°,0.23±1.47°; 0.05±0.92°,0.02±1.02°,0.18±1.04°in abdominal tumor with bone alignment,grey value alignment and manual alignment respectively.When the tumor was corrected every fraction, the margins required to account for intrafraction error are .57 mm,3.86 mm,2.69 mm; 4.79mm,4.65 mm,2.8mm; 3.42 mm,3.61 mm and 3.3mm in X, Y, and Z axis directions respectively. There was no significant difference between bone alignment and grey value alignment,but manual alignment was different from them.Parotid glands decreased in volume to 17.64%～61.6% from theoriginal volume. At the end of the treatment ,the cumulative parotid mean dose of the 13 NPC patients differed from the plan dose by an average of 16.27%, with an average increase rate of 0.29Gy/treatmend day. Compared with original treatment plans ,the weekly volume planned receiving 26Gy at treatment completion increased 11.31%～45.02%. submandibular glands decreased in volume to 6.46%～61.82% from theoriginal volume. At the end of the treatment ,the cumulative parotid mean dose of the 13 npc patients differed from the plan dose by an average of 1.07%, with an average increase rate of 0.05Gy/treatmend day . There was no no statistical difference between with or withoutⅠb lymph nodes region irradiation. The parotid glands and submandibular glands dose difference was correlated with weight loss .Compared with the setup error before adjustment, optic chiasm maximum dose significant decreased by 2.47% after adjustment. For the spinal cord and brainstem,the maximum dose decreased by 1.08% and 1.85% respectively, but no significant dosimetric changes were seen. Left optic nerve maximum dose increased by 0.11%, and right optic nerve maximum dose decreased by 1.41%. There was improved significantly intarget dose coverage after adjustment.[Conclusions]There exists some extent of setup error in 3DCRT or IMRT of head-and-neck tumor,thoracic tumor and abdominal tumor patients. The online CBCT correction can be used to reduce the impact of setup error obviously, the required margins for the PTV was reduced accordingly. Three alignments can not replace each other, which alignment can be choiced based on the diseased region .It is suggested that manul adjustment after automatic alignment is necessary. The use of weekly kilovoltage computed tomography imaging and deformable image registration techniques make it possible to investigate the dynamic changes in parotid glands and submandibular glands , and to calculate the dose deviations resulting from anatomic changes. Significant dose variations were observed as result of interfraction anatomic changes. On-line correet reduce the setup errors and dosimetric effect of setup uncertainty in IMRT.These techniques enable the implementation of dose-adaptive radiotherapy.