ESTRO 36 Abstract Book

S828 ESTRO 36 _______________________________________________________________________________________________

Electronic Poster: Physics track: Treatment planning: applications

EP-1540 Optimal fractionation schemes for radiosurgery of large brain metastases J. Unkelbach 1 , H.A. Shih 2 1 University Hospital, Radiation Oncology, Zürich, Switzerland 2 Massachusetts General Hospital, Radiation Oncology, Boston, USA Purpose or Objective Stereotactic radiosurgery (SRS) is an established treatment option for patients with brain metastases. While small metastases are successfully treated with single-fraction SRS, the optimal fractionation scheme for large lesions is unclear and involves a trade-off between the number of fractions, tumor dose, and dose to normal brain. In this work, we demonstrate that spatiotemporal fractionation schemes, ie. delivering distinct dose distributions in different fractions, may improve the therapeutic ratio for these patients. Material and Methods Fractionation effects are described using the biologically effective dose (BED) model, assuming alpha-beta ratios of 10 in the tumor and 2 in normal brain. Treatment planning is performed based on objective functions evaluated for BED instead of physical dose. Constrained optimization techniques are used to ensure that all treatment plans have identical target coverage and conformity. Plans are compared regarding integral BED to normal brain, and BED to normal brain adjacent to the tumor. Results Traditional fractionation schemes deliver the same dose distribution in all fractions. Increasing the number of fractions reduces the BED to normal brain in the high dose region adjacent to the tumor for a fixed tumor BED. However, the integral BED to normal brain typically remains approximately the same. Spatiotemporal fractionation can lower the integral BED via the following mechanism: Distinct treatment plans for different fractions are designed such that each dose distribution creates a similar dose bath in the normal brain surrounding the tumor, ie. exploit the fractionation effect. However, each fraction delivers a nonuniform target dose such that a high single-fraction dose is delivered to alternating parts of the tumor (Figure 1). Thereby, partial hypofractionation in the tumor can be achieved along with relatively uniform fractionation in normal brain, leading approximately to a 10% reduction of BED in normal brain.

Figure 1: Illustration of spatiotemporal fractionation using 4 fractions for a large brain metastasis (26cc). Treatment planning was performed for rotation therapy (VMAT or Tomotherapy) such that a cumulative BED equivalent to 20 Gy single-fraction dose is delivered to the target, while minimizing normal brain BED. It is assumed that 1x20Gy in the tumor corresponds to 2x13Gy, 3x10Gy, and 4x8Gy. Hence, a uniformly fractionated 4- fraction treatment increases the total physical dose from 20 Gy to 32 Gy. Spatiotemporal fractionation delivers approximately 20 Gy in a single fraction to parts of the tumor. Thereby, the same tumor BED is achieved with a lower physical dose, which translates into a net BED reduction in the normal brain where the dose is relatively uniformly fractionated. Conclusion Delivering distinct dose distributions in different fractions may improve the therapeutic ratio. For patients with brain metastasis this may lower integral dose to normal brain and reduce cognitive decline. EP-1541 4D dose reconstruction using a standard TPS in combination with a respiratory motion model M. Ziegler 1 , J. Woelfelschneider 1 , H. Prasetio 1 , C. Bert 1 1 University Hospital Erlangen, Radiation Oncology, Erlangen, Germany Purpose or Objective Dynamic tracking ( DT ) is one approach to treat intra- fractionally moving tumors due to a conformal irradiation while sparing of healthy tissue. Clinical tracking systems rely on correlation models to predict the internal tumor position based on external surrogates. However, assessment of the actually delivered dose is still challenging as many treatment planning systems ( TPS ) do not have the ability to calculate dose on time-resolved ( 4D ) computed tomography images. The aim of this study was to determine the possibility for 4D dose reconstruction of DT patients using a common TPS and a respiratory motion model that is based on external surrogates. Material and Methods The University Hospital in Erlangen is equipped with a Vero system (Brainlab, Feldkirchen, Germany) that is used to treat patients with intra-fractionally moving tumors by DT. This system further provides the extraction of the patients’ surface as a surrogate by external infrared