ESTRO 38 Abstract book

S277 ESTRO 38

reconstructed by an iterative algorithm which provides 20 breathing phases per measurement. The deformation vector fields between the breathing phases were used to warp a 3D planning CT of the corresponding patient to obtain multiple synthetic 4DCTs per patient. These synthetic 4DCTs were used to perform single field uniform dose 4D dose calculations with 1.8 Gy (RBE) per fraction and two oblique proton beams. The interplay effect was quantified by means of the homogeneity index d5/d95 and CTV coverage v95 for gradually increasing numbers of treatment fractions. This was realized by randomly accumulating single fraction dose distributions with both variable initial breathing phases and underlying synthetic 4DCTs. Moreover, correlations between the interplay effect and pancreatic motion amplitudes were analyzed. Results Variable initial breathing phases and day-to-day organ motion variations lead to an incremental mitigation of the interplay effect. For single fractions, pronounced underdosage down to v95=70% with heterogeneous dose distributions were observed. On average, after 7 fractions, sufficient CTV coverage of 95%-107% was obtained for the patients. However, for patients with large underlying motion amplitudes, more fractions were needed to mitigate the interplay effect sufficiently. Despite of the pronounced interplay impact on the CTV, no significant differences were found for the organs at risk, comparing static and 4D dose distributions. Significant correlations were found between the CTV motion amplitudes and the resulting interplay effect.

Purpose or Objective To determine the accuracy of a surface imaging system (AlignRT) for positioning of breast cancer patients in breath hold (BH) by comparison with conebeam computed tomography (CBCT) data. Second, to evaluate with AlignRT the intrafraction variability and stability of the breast surface position during BHs guided with the active breathing coordinator (ABC) system. Material and Methods eighteen consecutive left-sided breast cancer patients treated with deep inspiration breath hold radiotherapy (DIBH-RT) were included. During CBCT acquisition and treatment, AlignRT monitored the breast surface. CBCT registrations were performed both on the target volume and on the patient’s surface. The setup error differences between the CBCT and AlignRT were analyzed in terms of the group mean ( M ), the systematic error ( Σ ), the random error ( σ ) and the 95% limits of agreement (LOA). A linear regression analysis was performed to determine the correlation. Furthermore, the intrafraction variability and stability of the ABC guided BHs were evaluated with AlignRT. The variability was determined as the average over the maximum differences between different BH levels within a treatment fraction for a patient. The stability was determined as the difference between the start and end position of a BH. Results In total 143 treatment fractions were included with an average of 7.9 ± 3.6 CBCTs per patient. Figure 1 shows the Bland-Altman plots for the setup errors of AlignRT and the CBCT registrations. The LOA values are on average tighter when the CBCT is registered to the patient’s surface. The LOA values (mean ± 2SD) were 0.1 ± 3.0, 0.6 ± 4.1 and 0.4 ± 3.4 mm in left-right, craniocaudal and anterior-posterior directions respectively for the CBCTs registered to the patient’s surface. Systematic and random errors of the setup error differences were ≤2 mm in all directions. AlignRT data showed higher correlations with CBCT data for the patient’s surface then for the target volume (0.61 vs 0.44 respectively). This was as expected as AlignRT monitors the same surface as was registered during the CBCT surface registration. For the variability and stability a total of 1705 BHs were analyzed. The results are presented in Table 1. The average intrafraction variability between BHs was 2.2, 2.8 and 2.3 mm whereas the average stability was -1.0, 2.1 and 1.5 mm in left-right, craniocaudal and anterior-posterior directions respectively. The largest error in intrafraction variability was 12.4 mm (craniocaudal direction) and in stability it was 11.7 mm (anterior-posterior direction).

Conclusion 4D dose evaluations, based on repeated 4D-MRI images, are a promising tool to investigate interplay effects and their dependency on the number of fractions. Without exposing the patients to additionally ionizing radiation, repeated 4D-MRI are useful to include motion variability into 4D dose analyses. Generally, for hypofractionated treatments, the interplay effect showed to be more pronounced than for standard fractionation schemes for pancreatic cancer treatments.

Proffered Papers: RTT 5: Improving accuracy in patient positioning

OC-0527 Evaluation of AlignRT for deep inspiration breath hold positioning and intrafraction monitoring V. Hamming 1 , C. Visser 1 , D. Busz 1 , E. Batin 1 , J.A. Langendijk 1 , S. Both 1 , N.M. Sijtsema 1 1 UMCG University Medical Center Groningen, Radiation Oncology, Groningen, The Netherlands