Background: measuring the extent to which the breast setup error contributed to the total delivered dose discrepancies in all patients treated with Volumetric modulated arc therapy (VMAT) both in deep inspiration breath-hold technique (DIBH) and free breathing (FB) and to compare the dosimetric stability of each technique. Methods: the dose distribution parameters of the mean of 90 cone-beam computer tomographies (CBCTs) reshaped by deformable image registration (DIR) compared to the mean of their reference plan (RP) in left (Lt) sided breast cancer (BC) patients treated with adjuvant loco-regional radiation. 50 CBCTs in the first group treated in DIBHT and 40 CBCTs in the second group treated in FB compared also to the mean of their RP. Deviations in the delivered from the planned dose distributions have been analyzed. The average of this difference has been calculated for all cases and in each group of patients.
Results: 1% reduction in PTV-total D95% caused by -4% decrease in the PTV-supraclavicular (supra). 13% and 3% increase in Dmean of heart and Lt lung, respectively. 8% and 6% increase in spinal cord Dmax and right (Rt) breast D2%, respectively. The difference between both techniques was significant in heart with its sub-volume, Rt breast and spinal cord.
Conclusions: The dosimetric impacts of breast setup errors during Lt BC loco-regional radiation by VMAT technique using DIR was significant for the heart with its sub-volume in the FB group only.
Keywords CBCT, VMAT, Breast setup error, deformable image registration
