Room: Karl Dean Ballroom A2
Purpose: Lung toxicity is the primary limiting factor for TMI, while computationally intensive planning and accurate delivery are the technical challenges. To this end, we have developed a step-and-arc technique, that exploits a larger fluence space via longitudinally translating the couch in 0.4-6.0cm increments while delivering partial arc(24-180Â°) VMAT at each step. Auto-plan parameters that optimize the mean lung dose (MLD) with efficient autoplanning in Pinnacle 9.10 were investigated.
Methods: Twenty-two trials were planned for a thorax phantom with the same target coverage and normal tissue sparing objectives while varying autoplan parameters; 1-99% target-conformity-tuning-balance, 1.0-4.5cm dose-fall-off-margin and 100-140% hot-spot-maximum-goal. The acceptable plans (PTV V100>95% and normal tissue maximum dose NT_Dmax<140%) were ranked and the parameters from the plan with lowest MLD were used for autoplanning 6 patients. In addition to the MLD, mean doses to the brain, oral cavity, liver, kidneys and bowels were also compared to clinical benchmarks from conventional IMRT/VMAT treatments. One patient plan was delivered to a Delta4 phantom on the Truebeam developer platform for validation.
Results: All thorax phantom plans met acceptance criteria, NT_Dmax between 106-130% and 40-50% MLD. Adjusting dose fall-off margin barely impacted dose distributions; however, favoring OAR sparing and allowing hot spots were effective in reducing MLD. Using the same set of parameters (99% tuning balance, 3.0cm dose-fall-off and 120% hot-spot-maximum-goal), patient plans with lower mean dose for all OARs were possible from a single autoplan run. In particular, the MLD was 57.8Â±4.0% (meanÂ±SD) compared to the benchmark 69.4Â±6.1%. The QA measurement had a 3%/3mm Gamma passing rate of 97.6%.
Conclusion: Tuning the Pinnacle 9.10â€™s autoplan parameters can provide efficient and automated Step-and-Arc TMI plans with improved dosimetric outcomes and accurate treatment delivery.