Application of cone-beam CT to study the positional errors of radiation therapy for neuroblastoma in children Neuroblastoma is one of the common malignant tumors in children, ranking fourth in incidence after leukemia, central nervous system tumors and lymphoma, and accounting for about 8% to 10% of pediatric malignant solid tumors. Children usually require a combination of treatment, including surgery, chemotherapy and radiotherapy to the tumor bed. Post-surgical children are usually treated with radiation doses of 21 Gy to 26 Gy, but some treatment protocols now suggest higher doses, and it has been documented that increasing the dose to the tumor target may increase the local control rate to 90%. Increasing the target dose places greater demands on the precision of radiation therapy. Positioning error is an unavoidable problem in radiation therapy, and it is common to place a certain boundary outside the clinical target area to form a planned target area to compensate for the dosimetric impact of the error. The outline of CTV is mainly based on clinical experiments and has been basically unified, while the external radiation boundary is influenced by the experience of radiotherapy institutions and radiotherapy equipment, and the size of the external radiation boundary varies among institutions, too large external radiation boundary will lead to an increase in the irradiated dose to normal tissues, and insufficient external radiation boundary will result in insufficient dose to the target area. The most common sites treated for abdominal neuroblastoma in children include the adrenal and paravertebral regions, with the kidney usually being closer to the tumor bed. Reducing the boundary between the CTV of the clinical target area to the PTV of the planned target area is important to protect normal tissues, especially the kidney. The positional error and the internal body dynamics are two main factors that constitute the external radiation boundary, among which the positional error includes interfraction and intrafraction error. There are several literature reports on the positional error of adult head and neck, pelvis or abdomen in China and abroad. 1. Materials and methods Ten children with retroperitoneal neuroblastoma, with a mean age of 4 years (2.5 years ~7 years), were selected from October 2012 to May 2013. Each child was treated in the supine position with hands raised on either side of the head and received a radiation dose of 25.2 Gy for 14 sessions at a single dose of 1.8 Gy. Acquisition parameters of CBCT images: rack The counterclockwise scanning angle was 200° (100° to 260°) rotation. The voltage was 100 kV, S20 collimator, filter plate of F1, medium resolution reconstruction, and 414 frames of images. In this paper, weekly CBCT scans were used instead of weekly verification films, and the CBCT scans were aligned with the localized CT images, and the alignment range was the vertebral area including the PTV, using grayscale alignment. The technologist will decide whether to continue manual adjustment according to the alignment situation, and record the error values of the center point between the CBCT image and the positioning CT before treatment in three directions: left and right, head and foot, and thoracic and dorsal directions, and if the error in one direction exceeds 3 mm, the patient’s position will be adjusted, and the error of the rotation angle in the three directions should be less than 3° (>3° requires repositioning). If the error in one direction is more than 3 mm, the patient’s position will be adjusted and the error in the three directions of rotation must be less than 3° (more than 3° requires repositioning). If the scanning error is greater than 3 mm, the scan data will be adjusted before each treatment; the second method is the daily CBCT scan verification, where the CBCT scan is performed before treatment and again after treatment, and the post-treatment CBCT images are obtained and the error values in each of the three directions are recorded. Each CBCT image is compared with the localization CT image, so the error of the CBCT scan before treatment can be regarded as interfractional error, and the error of the CBCT scan after that treatment can be regarded as intrafractional error. According to Van Herk et al.’s formula: 2.5Σ+0.7σ (Σ is the systematic error and σ is the random error), the pose error is: For weekly CBCT scan validation, the pose error needs to consider the contribution of both intra-fraction and inter-fraction errors, and for daily CBCT scan, the inter-fraction error can be considered as 0, and only the contribution of intra-fraction error is considered. 2. Results The mean volume of CTV was 155.2 cc (range 83.2 cc to 260.1 cc) for ten patients with 10 pre-treatment CBCT scans and 10 post-treatment CBCT scans, respectively. Table 1 below shows the inter- and intra-positioning errors of the ten patients. The inter-positioning errors were 4.9 mm in the left-right direction, 5.3 mm in the head and foot direction, and 4.8 mm in the chest and back direction. The intra-positioning errors were 2.1 mm in the left-right direction, 1.7 mm in the head and foot direction, and 1.4 mm in the chest and back direction. If we follow the daily CBCT image guidance, the inter-fraction error can be considered as 0, and only the intra-fraction error is considered, in which case the positional error is only 2.1 mm in the left-right direction, 1.7 mm in the head-foot direction, and 1.4 mm in the chest-dorsal direction. The outward radiation boundary from CTV to PTV can also be reduced accordingly. CBCT image scanning can be aligned according to the image grayscale to obtain the positional error during the treatment. The pose error of childhood neuroblastoma in radiation therapy is about 5 mm-6 mm if the image is verified once a week, but it can be reduced to about 2 mm with daily image verification. In this paper, the average volume of CTV of neuroblastoma is about 155 cc, and assuming that it is spherical, the volume of target area increases to 235 cc if the pose error is 5 mm, and to 184 cc if the pose error is 2 mm. Reducing the pose error directly reduces the illumination range of approximately normal tissue. In pediatric patients, additional radiation dose needs to be carefully considered when using image guidance, and the radiation dose with CBCT is comparable to that of weekly orthogonal field verification films . Due to respiratory motion, there is motion in both the tumor target area and the kidney during treatment, and the error caused by this motion needs to be evaluated by internal boundaries. several papers have reported the evaluation of the size of the internal boundaries by 4DCT, and the size of the internal boundaries was not considered during the study in this paper. hua C et al. reported that the kidney motion in children is smaller than that in adults, about 1 mm in the thoracodorsal direction and about 2 mm-3 mm in the head and foot direction. this is similar to the daily This is comparable to the size of the positional error of the daily CBCT scan. The use of daily CBCT image guidance can effectively reduce the positional errors in pediatric neuroblastoma radiotherapy, narrow the CTV to PTV exenteration boundary, and reduce normal tissue damage. Therefore, the use of CBCT daily image guidance is recommended for radiation therapy, especially for intensity-modulated radiation therapy.