ESTRO 2021 Abstract Book

S1461

ESTRO 2021

different rotational errors. Again, rotational errors were measured by means of a CBCT and counteracted by adjusting the rotatable head fixation accordingly. The result was verified by checking the entrance portals and by performing a further CBCT.

Results With the device it was feasible to correct rotational errors. The residual errors were below 0.2°.

Conclusion We designed and built a prototype rotatable headrest. We successfully tested its usability. The device will enable stereotactic treatments with both, C-arm and ring gantry clinical linacs using their standard treatment tables. PO-1738 Reducing Dose Hot Spots for Hypofractionated Gamma Knife Radiosurgery via Hundreds of Isocenters N.W. Cho 1 , D. Raleigh 1 , B. Ziemer 1 , T. Nano 1 , P. Theodosopoulos 1 , P. Sneed 1 , L. Boreta 1 , S. Braunstein 1 , L. MA 1,1 1 University of California San Francisco, Radiation Oncology, San Francisco, USA Purpose or Objective To report our initial clinical implementation of a hypofractionated Gamma Knife brain radiosurgery (GKSRS) approach via packing hundreds of isocenters for treatment of a large target with a substantial GTV expansion in order to reduce dose hot spots in eloquent brain. Materials and Methods A high-density isocenter packing (HDIP) technique similar to “onion peeling” was developed for image-guided hypofractionated GKSRS, where a large target volume of 14.1 cc was sequentially shrunk or “peeled” by a margin of 2 to 4-mm depending on its size. The isocenters were densely packed inside each layer or each “peel” of the target, especially within the GTV-to-PTV margin. Once packed, the isocenter location, shapes and weights of the beams were inversely optimized and manually adjusted to achieve an optimal dose distribution. This technique was clinically implemented on a frameless SRS unit for the purpose of expediting the workflow in treating brain lesions with GTV-to-PTV expansion of at least 5 mm. Results HDIP enabled the target dose to be prescribed to the 80-90% isodose line (normalized to the dose maximum), thus reducing dose hot spots greater than 120% of prescription dose inside the GTV expansion volume comprised mostly of normal brain tissue. Compared to the conventional approach of using only 1/3 to 1/4 as many isocenters, HDIP achieved identical dose coverage (>98%), dose conformality (<= 2.8) and Paddick gradient indices (<= 3.5) for the PTV. An important caveat in implementing HDIP was that when globally increasing the total number of isocenters, the dose contributions from some isocenters decreased rapidly, rendering their beam-on time to fall below the machine limit of 0.01 min. To overcome such a problem, manual blocking of beamlets associated with the affected isocenters was implemented to effectively lower the dose rate, propelling the treatment time at these isocenters > 0.01 min to enable clinical implementation. This strategy negligibly impacted the total treatment time of 20-30 min of a typical HDIP treatment with a dose rate of 2.0 Gy/min of our GKSRS unit. Conclusion HDIP is technically feasible and successfully implemented for the first time at our institution, enabling us to expedite the hypofractionated GKSRS workflow to care for patients with targets in eloquent brain requiring substantial GTV-to-CTV/PTV margin expansion PO-1739 Optimal threshold of model parameters for the respiratory tracking system with helical tomotherapy W. Okada 1 , M. Tanooka 1 , H. Doi 1,2 , K. Sano 1 , M. Shibata 1 , K. Nakamura 1,2 , Y. Sakai 1 , M. Tanaka 1 1 Takarazuka city hospital, Department of Radiotherapy, Takarazuka, Japan; 2 Kindai University Faculty of