Accuracy and sources of error of out-of field dose calculations by a commercial treatment planning system for intensity-modulated radiation therapy treatments

Although treatment planning systems are generally thought to have poor accuracy for out-of-field dose calculations, little work has been done to quantify this dose calculation inaccuracy for modern treatment techniques, such as intensity-modulated radiation therapy (IMRT), or to understand the sources of this inaccuracy. The aim of this work is to evaluate the accuracy of out-of-field dose calculations by a commercial treatment planning system (TPS), Pinnacle(3) v. 9.0, for IMRT treatment plans. Three IMRT plans were delivered to anthropomorphic phantoms, and out-of-field doses were measured using thermoluminescent detectors (TLDs). The TLD-measured dose was then compared to the TPS-calculated dose to quantify the accuracy of TPS calculations at various distances from the field edge and out-of-field anatomical locations of interest (i.e., radiosensitive organs). The individual components of out-of-field dose (patient scatter, collimator scatter, and head leakage) were also calculated in Pinnacle and compared to Monte Carlo simulations for a 10 x 10 cm(2) field. Our results show that the treatment planning system generally underestimated the out-of-field dose and that this underestimation worsened (accuracy decreased) for increasing distances from the field edge. For the three IMRT treatment plans investigated, the TPS underestimated the dose by an average of 50%. Our results also showed that collimator scatter was underestimated by the TPS near the treatment field, while all components of out-of-field dose were severely underestimated at greater distances from the field edge. This study highlights the limitations of commercial treatment planning systems in calculating out-of-field dose and provides data about the level of accuracy, or rather inaccuracy, that can be expected for modern IMRT treatments. Based on our results, use of the TPS-reported dose could lead to an underestimation of secondary cancer induction risk, as well as poor clinical decision-making for pregnant patients or patients with implantable cardiac pacemakers and defibrillators.