Dosimetric and Anatomical Evaluation of Normal Tissue Changes During IMRT Radiotherapy for Head and Neck Cancer
M Hunt*, N Tyagi, M Tam, P Zhang, Y Hu, S Rao, N Lee, Memorial Sloan Kettering Cancer Center, New York, NYTU-E-141-5 Tuesday 2:00PM - 3:50PM Room: 141
To measure changes to parotid, submandibular glands and oral cavity using weekly CBCT and to determine factors predictive of weekly volume and mean dose changes.
Weekly CBCT images were reviewed for seven patients undergoing standard IMRT. Parotid, submandibular glands (SMG) and oral cavity were segmented both manually and through free-form deformable registration of the CBCT to the planning CT. Structure volume, center-of-mass (COM) and manual/deformed similarity on each CBCT was recorded. All structures were then mapped back to the planning CT and weekly delivered mean doses were calculated. Temporal changes in volume, COM and dose were correlated to pre- and on-treatment anatomic and dosimetric parameters.
On average, 6.5 scans per patient were available for review. Mean decrease in parotid, submandibular gland and oral cavity volume by the end of treatment was 18.9% (σ=10.0), 21.9% (σ=13.1) and 3.1% (σ=7.8), respectively. Volume changes translated into an average 2.9 Gy (σ=2.8) increase in parotid gland mean dose. No significant change in mean dose to SMG or oral cavity was observed. Deformable registration commonly failed in areas of dental fillings requiring subsequent manual editing of parotid and oral cavity volumes. Average Dice Similarity Coefficients of 0.841 (σ=0.051), 0.752 (σ=0.089) and 0.892 (σ=0.037) for the parotids, submandibular glands and oral cavity were obtained. On multivariate analysis, factors correlated with volume and dose change during treatment included the planned dose-to-date, patient weight and changes in structure COMs.
Dosimetric consequences of on-treatment volume changes of normal tissues during head and neck IMRT treatment were limited to an increase in mean dose to the parotid glands. Planned dose-to-date, patient weight, and center-of-mass changes were all predictive of volume changes during treatment. Improvements in CBCT image quality are essential before these images can be used for treatment decisions or adaptive therapy.