Room: ePoster Forums
Purpose: Cardiac risk is linearly associated (7.4%/Gy, Darby et al, NEJM, 2013) with radiation dose but it is uncertain due to calculation algorithm and dose calculation grid size. Due to volume averaging, smaller organs are particularly vulnerable to incorrect dose estimation. This study evaluates impact of calculation grid size on accurate dosimetry for the lung, heart and the left anterior descending artery (LAD) for radiation treatment of left breast cancer.
Methods: Fifty left breast cancer patients were evaluated (10 treated supine and 40 treated prone). Dose calculations were performed using advanced calculation algorithm (AAA) for inhomogeneity correction by changing calculation grid size incrementally from a minimum of 1x1 mm2 to a maximum of 5x5 mm2, which was the traditionally used grid size. Maximum and mean OAR doses were analyzed for LAD, heart, lung and breast.
Results: With normalization to 5x5 mmÂ² grid, the maximum and mean doses to the heart, lung and LAD decrease with decreasing grid size and nearly identical in prone and supine position with respect to grid size. The correlation between calculation grid size and OAR doses are linearly correlated both in maximum and mean dose for both algorithms. For the mean dose to the LAD, heart, and lung, the linear fit model was found to be optimum with RÂ² values of 0.95, 0.94, and 0.91, respectively suggesting a quantifiable and linear impact of calculation grid size on OAR dose estimation.
Conclusion: Small calculation grid size provides more accurate dosimetry that should be used for smaller structures or segments of structures (e.g. LAD) where dosimetry may be critical, particularly at areas with steep dose gradient, such as the chest wall/breast interface. Achieving more accurate dosimetric estimations will ensure that estimated future risk of cardiopulmonary toxicity is accurate; enabling optimal treatment decision making.