Nuchsirikulaphong, N., Dankulchai, P., Thepmongkhol, K., Khachonkham, S., Treechairusame, T., Jaikuna, T., and Sathitwatthanawirot, C.
Head and neck cancer (HNC) patients treated with radiotherapy confront the changing of air cavity during the treatment course. Changing air cavities may influence dose distribution due to the electron return effect (ERE) under a magnetic field in magnetic resonance linear accelerator (MRL) and affect plan quality. This study aims to investigate the impact of air cavity change in the inter-fraction of HNC patients on the adaptive intensity-modulated radiotherapy (IMRT) plan quality. In this study, the five computed tomography (CT) and fifteen surrogate cone-beam CT (CBCT) images at the beginning and middle of the treatment course from five oropharyngeal cancer patients were arbitrarily selected for this retrospective study by considering the dimension of planning target volume in the superior-inferior direction less than 20 cm. Tumor, organs at risk, and air cavity within planning target volume (PTV) were segmented on the CT image by an experienced radiation oncologist and transformed to the CBCT images using rigid registration by the Eclipse treatment planning system. Nine-field IMRT was generated on the CT image using the Monaco Unity treatment planning system and used as an initial plan. This study investigated two scenarios of the air cavity affect plan quality 1) a treatment plan without adaptive radiotherapy (ART) and 2) a treatment plan with ART using the adapt to shape (ATS), adapt to shape lite (ATS-lite), and replan. The initial plan was transferred and recalculated on each surrogate CBCT using CBCT to electron density (ED) correlation curve to investigate the effect of air cavity change without ART (recal technique). For the ATS plan, the full re-optimization was performed based on segmented contours of the actual anatomical border and assigned the Hounsfield unit (HU) of each contour on CBCT using mean ED from the CT image (Bulk density-assigned). The re-optimization plan based on a transferred contour from CT to CBCT using either rigid or deformable registration was observed in the ATS-lite. In addition, the replanning process (replan technique) was performed based on the HU in each voxel on CBCT and used as a reference plan. The recal, ATS, and ATS-lite plans were compared with replan technique by considering the plan quality index, including dose difference at D95%, D98%, and D2% extracted from the dose-volume histogram of PTV and three-dimensional (3D) dose distribution difference via 3D global gamma analysis (3%3mm criteria with 10% threshold). Additionally, the volume and location of 105% of the prescription dose was observed. This study considered the statistically significant at p<0.05 by ANOVA test using Stata version 17.0. The air cavity volume within PTV extracted from each CBCT was changed on average 0.99 ± 2.71 cm3 compared to the CT image. The most considerable dose difference was observed in the high dose region at D2%, around 1.25 ± 0.64 Gy for the recal technique compared to the initial plan, while the difference at D95%, D98% is less than 1 Gy. This study found statistical differences among ART techniques (p<0.05). The most significant difference between the ART and replan techniques was found in ATS-lite with a rigid contour approach by increasing the dose at D2% 0.42 ± 0.77 Gy. In contrast, ATS-lite with a deformed approach and ATS were 0.29 ± 0.67 Gy, and 0.09 ± 0.61 Gy, respectively. However, the coverage doses at D95% and D98% were decreased compared to the replan technique in all ART approaches. The lowest dose difference of D95% and D98% was found when using the ATS technique (0.10 ± 0.42 Gy and 0.27 ± 0.49 Gy for D95% and D98%). The high passing rate of gamma analysis was found in ATS, ATS-lite techniques with a rigid- and deformed-contour, about 86.76 ± 0.07 %, 84.07 ± 0.08 % and 83.37 ± 0.05 % consecutively, compared to replan technique. As expected, the recal technique yielded the lowest gamma passing rate (80.32 ± 0.05 %). The mean difference of the volume received 105% prescription dose was about 20.91 ± 18.97 cm3, 14.99 ± 18.21 cm3, 13.65 ± 13.33 cm3, and 13.02 ± 16.81 cm3 for the recal, ATS, ATS-lite with rigid-, and ATS-lite with deformed-contour compared to the replan which is located within PTV and close to the pharynx, glottis supra, mandibles, oral cavities, and parotids. Accordingly, the air cavity in the HNC patient's PTV changes throughout the treatment course and affects the quality of the IMRT plan. ART in MRL has a significantly improving treatment plan quality, but minimal impact in clinic. The correlation of time series, the volume of the air cavity, and dose will be investigated in the future study. [ABSTRACT FROM AUTHOR]