The Clinical Effect of Changing from a Dose to Water to a Dose to Medium-Based Methodology to Calculate Monitor Units for Electron Beams

Abstract

In 2016, The Mid North Coast Cancer Institute (MNCCI) implemented Elekta’s Monaco® electron Monte Carlo (eMC) dose calculation algorithm for electron beams. The new algorithm resulted in a change in clinical practice whereby monitor units (MU) delivered to patients would be calculated in terms of dose to medium (𝐷𝑚), rather than using measured, dose to water (𝐷𝑤)-based MU. Delivering high energy ionising radiation during radiotherapy has inherent risks. Therefore, it is crucial to patient safety that dose calculations performed by the treatment planning system (TPS) are accurate. Any changes in clinical practice should also be understood in terms of their effect on patient outcomes (both tumour control and normal tissue toxicity). This research aims to validate the accuracy of the 𝐷𝑚-based MU calculated by the Elekta Monaco® eMC algorithm and to assess patient outcomes in terms of acute skin toxicity due to this change in clinical practice. To validate the 𝐷𝑚-based MU, the dose measured with an ionisation chamber was compared with the dose calculated by the Elekta Monaco® eMC algorithm in a range of clinically relevant phantoms. The assessment of acute toxicity involved a cohort study that compared acute skin toxicity grades of skin cancer patients treated before the change in clinical practice (cohort 1) and after the change (cohort 2). Various predictors of radiation-induced acute skin toxicity were also investigated. The comparison between measured and calculated dose found that the Elekta Monaco® eMC 𝐷𝑚 calculation is accurate in most clinical scenarios. The level of agreement between the measured and calculated 𝐷𝑚 data is mostly within ± 3.5% for a wide range of tissue types. However, for tissues with densities significantly different from water (i.e. < 0.5 g/cm3 and > 1.5 g/cm3), the method used to determine 𝐷𝑚 from measurements of ionisation resulted in unacceptable levels of uncertainty. For these tissues, a more accurate validation method, such as full Monte Carlo modelling, is required. Two hundred and ninety-four patients were recruited into the cohort study, with 141 patients in cohort 1 and 153 patients in cohort 2. Statistical analysis of patient acute toxicity data was performed using the Cochran-Armitage test for linear trend and binary logistic regression modelling. The results of the Cochran-Armitage test for linear trend found no statistically significant increase in acute skin toxicity for patients in cohort 2 compared with patients in cohort 1. Therefore, the change in clinical practice from using the measured 𝐷𝑤-based MU to using the 𝐷𝑚-based MU as calculated in the Elekta Monaco® eMC algorithm does not increase acute skin toxicity for skin cancer patients treated with electron beams. Binary logistic regression modelling found a statistically significant correlation between baseline toxicity grade and acute toxicity grade, suggesting that baseline grade is often a predictor of acute toxicity grade. This modelling also found that patients treated with the 𝐷𝑚-based MU as calculated in the Elekta Monaco® eMC algorithm experience statistically significant lower levels of pain, while those patients with treatment sites involving bone experience statistically significant lower levels of pain but increased alopecia.