Abstract
Purpose/Objective: To examine whether the impact of respiratory motion on the delivered dose distributions for a range of tumour sizes in SABR, differs between 3DCRT,sliding window IMRT and RapidArc. Materials and Methods: Three patient data sets with different lung tumour sizes were selected: T1=6cc, T2=31cc and T3=60cc. Each was planned in Eclipse for SABR using 3DCRT, sliding window IMRT and VMAT, creating 9 treatment plans which were then delivered to a dynamic thorax phantom. The phantom was programmed to move at a typical patient breathing amplitude of 15mm with a period of 5 seconds and Varian linacs were used for the delivery. EBT3 Gafchromic film was used in coronal and sagittal planes for measuring dose distributions. Static phantom measurements were compared with the TPS calculated plans to establish agreement between expected and measured dose distributions without motion, using the software OmniPro I'mRT. Comparisons of static and dynamic phantom measurements followed. Global gamma analysis was used to carry out a relative comparison between the three delivery techniques. Five regions of the gamma index map (Middle, Proximal Left, Proximal Right, Distal Left, Distal Right) were analysed to quantify the differences along the axis of motion. The criteria used for the gamma analysis were 3%/3mm, with a threshold of 20%. Results: The setup and delivery accuracy was confirmed by the agreement between planned and static delivered dose distributions. The average percentage of pixels passing was 100% (T1),100% (T2) and 98.46% (T3). The comparison of films with and without motion gave lower percentages of pixels passing, ranging between 33.68 - 59.94% (T1), 47.86 - 61.77% (T2) and 43.44 - 64.32% (T3). Comparison of the delivery technique, showed passing rates of 33.63 - 52.25% (3DCRT), 43.44 - 64.52% (IMRT) and 46.58- 56.08% (VMAT). Analysis of the five regions for all delivery techniques gave averages of 93.76% (Middle), 58.7% (Proximal) and 12.8% (Distal). For 3DCRT results were 87.08% (Middle), 46.52% (Proximal) and 12.54% (Distal), for IMRT were 96.45%, 69.20%, 14.14% and for VMAT 97.75%, 60.39% and 11.71%, respectively. Conclusions: The results are indicative of the intra-fractional respiratory motion-induced dosimetric inaccuracies caused in three SABR delivery techniques.On average, the impact is greatest in the distal regions, significant in the proximal regions, whereas the middle region is less susceptible to these effects. No noticeable difference was observedbetween coronal and sagittal planes. The results also suggest that the effect of motion is greater in the proximal regions for 3DCRT in relation to the other techniques, particularly with smaller tumour sizes.