25 Practical evaluation of treatment planning of Cyberknike patients in Institut Canceroloige of Lorraine

Introduction We conducted a practice analysis as part of a quality assurance process related to the commissioning of a Cyberknife M6. In this context, several physicists have been specifically trained. The purpose of this study was to assess whether interoperator differences existed in the implementation of treatment plans and whether these differences could impact treatment. Methods We retrospectively analyzed treatments planned between September 2017 (first Cyberknife M6 treatments) and March 2018 using Precision 1.1 (Accuray) TPS. The data collection and formatting was automated under Excel macro from the export of the “Plan Overview” file generated by the TPS containing the specific technical and dosimetric data (DVH). The analysis focused on dosimetric indexes (PTV coverage, compliance index (reference isodose vol/ PTV flight), gradient index (isodose volt 50%/ reference isodose)] and efficiency indices of treatment delivery (number of beams, MU per session and estimated duration of the session). Numerical parameters were compared by variance analysis or by Kruskal–Wallis test according to the normality of the distribution (studied by a Kolmogorv-Smirnov test). The qualitative parameters were compared by a Chi-square or Fisher Exact test. The level of significance was set at 5%. Results 6 physicists (medical physicists only) with experience ranging from less than 1 year to 11 years completed 226 treatment plans. 40% of the treatment plans were intracranial. The proportion of intracranial/extra-cranial plans, the volume of the PTVs and the multiplicity of volumes were comparable between the 6 physicists. No significant differences between physicists were found on PTV coverage (p = 0.108) or Gradient Index (p = 0.219). Significant differences appear on the number of MUs (p = 0.008), the number of beams (p = 0.001) and the treatment times (p = 0.020). These 3 parameters are related to beam, time or UM reduction operations, during the last stage of plan optimization. No criterion on these 3 parameters or on the reductions was defined a priori within the team; they remain at the discretion of the physicist. Conclusions The statistical evaluation of the treatment plans carried out during the first six months following the initiation of Cyberknife M6 treatments at our institution made it possible to objectify the homogeneity of the dosimetric results obtained by the different physicists, despite different optimization practices. The analysis also showed that the recently integrated medical physicist obtained comparable results to other physicists on all the evaluated indicators. This validates her training by companionship.