Registry Based Framework for Plan Quality Improvement: A Head-And-Neck Radiotherapy Application Improving Simultaneous Integrated Boost Treatments
S Qi*, S Lee, P Kupelian, A Pham, D Low, J DeMarco, UCLA School of Medicine, Los Angeles, CASU-E-T-696 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To systemically evaluate the dosimetric plan quality and consistency for multiple IMRT delivery techniques to further improve guidelines for head-and-neck (H&N) irradiation.
Method: One hundred consecutive H&N patients from our patient registry were selected and analyzed retrospectively. Simultaneous integrated boost (SIB) IMRT treatment were generated with the same prescription dose of 69.3/62.7/56.1Gy on TomoTherapy(Tomo, Accuray), TrueBeam(TB, Varian) using fixed-field IMRT(TB_IMRT) or RAPIDARC(TB_RAPIDARC) or Siemens Oncor (Siemens_IMRT, Siemens). The contouring was performed by the same physicians delineating targets and normal structures. The standard in house dose-volume objectives were used as planning guideline for all cases for all techniques. A Matlab based dose-volume extraction and analysis tool was developed to export dose-volume endpoints for each patient. A more specific guideline was proposed for future planning purpose.
Results: While all techniques yielded comparable target coverage, the doses to the critical structures differed. In average, the percent volumes receiving greater than 26Gy (V26 and standard deviation) for the total parotid were 37.8±0.1, 46.1±0.2, 44.3±0.2 and 41.8±12.9%, the V65 for mandible were 4.7±5.6, 6.3±11.1, 3.0±5.3 and 8.0±7.3 % with the mean mandible doses of 42.5±6.1, 41.4±8.2, 37.4±10.7 and 46.6±8.7 Gy for Tomo, TB_IMRT, TB_RAPIDARC and Siemens_IMRT plans respectively. ANOVA analyses showed significant difference for the maximum dose of cord but no significant difference was found for other selected structures among four IMRT techniques. Based on current dataset, the projected constraints (at 95% confidence level) for future H&N IMRT plans for cord maximum doses were established at 34.0±2.9, 42.7±2.5, 43.3±2.3, and 45.4±2.4Gy for the evaluated techniques, respectively. The refined dosimetric constraints may improve plan quality and consistency.
Conclusion: Patient registry based processes enable easy access of systematic dosimetric monitoring of the plan quality and consistency. The self-evolving real-time dosimetric evaluation and ensuing guidelines will help to continuously improve plan quality, resulting in better clinical outcome.