A correlation study between clinical dose distribution and gamma passing rates in pre-treatment Tomotherapy quality assurance

Introduction: we focused on Tomotherapy patient-specific Quality Assurance (QA) in prostate cancer cases. Conventionally it is possible to accomplish QA with the calculation of in-phantom gamma passing rate (GPR). However, GPR, although widely used, is not a good predictor of the agreement between calculated and delivered dose to the patient and recent studies suggest the use of clinical metrics. To our best knowledge, there are no similar studies in Tomotherapy. In this work, we used 3DVH software and ArcCheck matrix (SunNuclear Corporation, Melbourne, FL) and we looked for the correlation between clinical metrics results and phantom GPRs. 3DVH applies perturbations to the in-patient TPS planned dose to account for discrepancies in dose delivery. Materials and methods: we retrospectively selected patients between those treated in our centre for prostatic cancer with Tomotherapy between April 2017 and August 2018. We submitted all plans to a specific QA; for data acquisition, we used the ArcCheck matrix and SNC patient software. We compared delivered and calculated volumetric dose distributions via the GPR and evaluated the clinical metrics using the 3DVH software. Unlike what previous authors have done, we did not investigate DVH clinical endpoints (e.g. rectum v70) agreement, but the complete dose distributions agreement applying gamma metric to the comparison of in-patient calculated and delivered dose distribution. For statistical analysis, we made both two tails Pearson's correlation coefficient calculation and a linear regression evaluation. Results: from our results do not emerge any significant correlation between GPR and 3DVH results for every DTA and %D values used in gamma analysis. Conclusions: from this study, the need for employing clinical patient metrics instead of the simple gamma analysis, already suggested for conventional Linacs treatments, is extended to Tomotherapy treatments. Further studies are necessary to extend this result to Tomotherapy pre-treatment QA to other anatomical districts.