Cone Beam CT-Based Adaptive Intensity Modulated Proton Therapy Assessment Using Automated Planning for Head-and-Neck Cancer

Abstract

Abstract Background To assess the feasibility of CBCT-based adaptive intensity modulated proton therapy (IMPT) using automated planning for treatment of head and neck (HN) cancers. Methods Twenty HN cancer patients treated with IMPT and had daily CBCTs were included. Initial IMPT plans were created using automated planning software for all patients. Synthetic CTs (sCT) were created by deforming planning CT (pCT) to CBCTs. rCTs were deformed to CBCT obtained on the same day to create deformed rCT (rCTdef) as gold standard. Doses recalculated on sCT and rCTdef with the were compared using gamma analysis. The dosimetric parameters were assessed based on rCTdef, sCT with deformably propagated contours or with physician-corrected contours. Weekly sCTs were then created for all patients. For non-adaptive group, dose was directly recalculated on weekly sCTs. For adaptive group, adaptive IMPT plans were created on weekly sCT using the automated planning software. The non-adapted and adapted dose-volume parameters on weekly sCTs were evaluated. The weekly doses were accumulated back to the pCT for both groups, and dosimetric parameters were compared in between. Results High dosimetric agreement was found between sCT and rCTdef with high gamma passing rate of 97.4%±1.6% using 3mm/3% criteria. With contour correction on sCT by physician, the uncertainty range of using sCT to estimate mean dose for organ at risk (OARs) can be reduced to (-2.37%, 2.19%), while for V95 of primary or secondary CTVs, the uncertainty can be controlled within (-1.09%, 0.29%). For non-adapted treatments, 15.7%, 34% and 55.7% of the fractions didn’t meet the V95 > 98% constraint for primary, secondary and tertiary CTVs respectively while all adaptive fractions met the constraint. In the accumulated dose, adaptive planning significantly reduced mean dose to constrictors (-1.42Gy ± 2.79Gy) and larynx (-2.58Gy ± 3.09Gy) compared to non-adaptive group, resulting in significantly reduction on the normal tissue complication probability (NTCP) of larynx edema by 7.52%± 13.59%. Conclusion sCT can be a powerful tool for accurate proton dose calculation. Physician contour correction is recommended to reduce the uncertainty for dose evaluation. Daily adaptation resulted in better CTV coverage, OAR sparing and lower NTCP for some OARs as compared with non-adaptive IMPT.