Session Item

To study intra-fraction residual errors occurring in brain SRS treatments with the Cyberknife system depending on the characteristics of patient.

Data from intra-fraction displacements of 70 brain SRS patients (150 treatment fractions) treated in our department with the Cyberknife system were analyzed in this work. Geometrical uncertainties for four different situations were obtained: no intra-treatment tracking, real treatment and treatment if only half and one third of the tracking images were employed. Margins needed to account for intra-fraction uncertainties were assessed by means of the Van Herk´s formula assuming a 70% isodose level and a beam penumbra of σ_p=2mm. Margins due to rotations were calculated by assuming a mean distance between target and imaging center of 5 cm and calculating the effect of rotations on displacements of the target. As tumors can be isotropically distributed around the imaging center, symmetric uncertainties were applied.

All analysis were performed for the whole patient population and for a division of it in three groups according to the mean standard deviation of the corrections observed during treatment.

Treatment margins for the whole patient population are 0.5mm isotropically if no tracking is applied and 0.1 mm if residual errors of the real treatments are considered. Margins in AP and SI directions were increased to 0.2mm when skipping the 50% and 66% of the images, showing little dependency on imaging frequency. Results when dividing the population according to their in-treatment motion (based on the standard deviation of displacements) are shown in Table 1. It can be observed that margins are much larger for the group with the most unstable patients (group 3) than for the other 2 groups. Margins for group 3 in case no tracking is applied are close to 1mm and almost the double than in the other 2 groups, showing the importance of identifying unstable patients when no tracking capabilities are available. Also, margins of this last group are the most affected when images are skipped.

Intra-fraction tracking is able to reduce uncertainties of patient motion to a minimum. In case tracking is not available, most of the patients could receive an appropriate treatment. Nevertheless, attention should be paid to unstable patients for which tracking might be mandatory. Also care should be taken when choosing imaging frequency in this group of patients.