SU-E-T-249: Theoretical Analysis of the Effects Uncertainties Have On Treatment Outcomes.

PURPOSE Mathematical models are used in Industrial & Systems Engineering to analyze complex integrated operational systems. Adapting this approach to radiation therapy can help quantify the precision and accuracy necessary to achieve optimal outcome of radiation treatment. The purpose of this work is to develop such a model using clinical data and assess the effect uncertainties have on treatment outcomes. METHODS The Taguchi Loss Function (TLF) is adapted to radiation therapy using conventional radiobiological models for tumor control probabilities (TCP) and normal tissues complication probabilities (NTCP) based on the equivalent uniform dose. The TCP and NTCP curves are combined to create a failure probability function for a given treatment plan. The composite effects of all uncertainties involved in treating a patient are modeled by a normal distribution. The standard deviation and mean of the normal distribution represent the precision and accuracy of a treatment. The failure probability function is convolved with the normal distribution to arrive at an expected failure probability. Precision was varied from 0.5% to 25% while accuracy ranged from ±5% to investigate uncertainties effects on complication-free local tumor control. 3D 4-field box plans where compared to IMRT plans for 18 prostate patients using this method. RESULTS The average expected failure probability at the prescription dose for the 3D 4-field box plans was 30.02% and 18.13% for the IMRT plans at zero uncertainty. At 25% uncertainty the expected failure probabilities were 76.85% and 64.36%, respectively. On average the IMRT plans failure probability was 14.84% less than the 3D 4-field box plans for all uncertainty levels. CONCLUSION This study demonstrates that uncertainty in radiotherapy procedures has a quantifiable effect on treatment outcome. To further improve complication-free local tumor control we must both improve treatment technologies and improve quality to minimize the uncertainties in radiation therapy.