ESTRO 2021 Abstract Book

S1595

ESTRO 2021

quality assurance (PSQA) with gamma index evaluation was performed for every case to verify the reliability and robustness of rIMRT treatment delivery. Results Semi-automated rIMRT inverse planning allowed fast treatment planning: mean of 5.5 ± 0.3 minutes duration script. rIMRT showed a significantly better homogeneity in dose distribution than 3DCRT. In addition, rIMRT substantially limited the hotspots in the CTV breast while ensuring a coverage equivalent to 3DCRT or VMAT (Table I). rIMRT significantly reduced low doses spread compared to VMAT with lungs volume receiving 5 Gy of 13.4 ± 4.8 % versus 26.3 ± 10.7 % (p < 0.001) for rIMRT and VMAT respectively.

The mean dose to lungs, heart and contralateral breast were within dose constraints, however higher with rIMRT than 3DCRT, and lower with rIMRT than VMAT (Table II). Gamma index analyses of PSQAs confirmed excellent reliability of rIMRT treatment delivery with a 3%, 3 mm local gamma index average of 98.7 ± 0.7 %.

Conclusion rIMRT plans were a good compromise between 3DCRT and VMAT: with a better homogeneity than 3DCRT and lower mean dose to OAR than VMAT. Semi-automation ensured rapidity, standardization and strict framing of treatment planning. These results and the robust reliability of treatment delivery allowed us to implement whole breast rIMRT in daily practice within our department, without systematic PSQA. PO-1873 HyperArc and CyberKnife dose distribution comparison for 1, 2, and 3 brain metastases radiosurgery A. Skrobala 1 , W. Kijeska 1 , M. Kruszyna-Mochalska 1 1 Greater Poland Cancer Centre, Medical Physics Department, Poznan, Poland Purpose or Objective The purpose of this study was to compare the dose distribution in terms of plan quality and analysis of accuracy in dose delivery of HyperArc-based (HA) and CyberKnife (CK) for one, two, and three intracranial metastases treated, respectively. Materials and Methods In total, 15 patients with multiple intracranial metastases (3 equal groups with respectively 1 to 3 tumors) treated with CK were examined. These patients were replaned using HA. The prescribed dose of 21 Gy in single-fraction was prescribed to the 80% isodose line covering 98% of the PTV for both systems. PTV had a mean dimension of 4.2cm 3 (range 1.0-11.5cm 3 ). Plans had to respect the constraints on maximum dose to organ at risk reported by Timmerman et al. (Semin Radiat Oncol 2008). HA and CK plans were compared in terms of dose-volume metrics by Paddick and RTOG conformity index (CI), homogeneity index, and gradient index (GI). V 12 (the brain's volume receiving no less than 12 Gy) and mean dose to the brain-minus-PTV (BmPTV) were evaluated for the healthy brain. In terms of treatment time, the number of monitor units (MU) and overall treatment time (OTT) per fraction were examined. Following indicators (3%,3mm), (3%,2mm), (2%,2mm) and (L3%,3mm), (L3%,2mm), (L2%,2mm) were used in the passing-rate (PR) γ-analysis of accuracy in dose delivery, realized in-phantom (SRS 1000 and Octavius II) dosimetric verifications. The accepted level for PR was ≥95%, with a threshold of 20%. Results Both HA and CK plans were judged clinically acceptable, but significant differences were noticed in the plan quality parameters with the increasing number of treated lesions. A significant difference in both CI and GI favoring CK plans was observed for 1,2, and 3 metastasis (p<0.01). CK and HA plans were computed for doses to a healthy brain and were not equivalent, CK achieved lower values for V 12 7.93 vs. 10.06 Gy and BmPTV 1.15 vs. 1.33 Gy. Finally, both MU and OTT were significantly reduced by HA plans, especially 2 and 3 brain lesions. PR for CK and HA filed assumed criteria, and for HA plans the global approach (3%,3mm) the average PR valueswere 99.32±1.07, for (3%,2mm) 97.92±2.75, and for (2%,2mm) 96.25±3.72, while for the local approach these values were for (L3%,3mm) 98.35 ± 1.86, for (L3%,2mm) 95.25±4.19, and for (L2%,2mm) 95.15±4.39. Conclusion Comparing the dose distributions for the plans made for CK and HA, it was shown that the overall treatment time for patients with one, two, and three intracranial lesions is shorter with HA. CI and other plan quality parameters were higher or slightly higher for HA plans. The homogeneity index values are comparable but have an advantage over HA. Parameters determining brain sparing are more favorable for CK. Doses for organs at risk are comparable for both techniques but with a slight advantage over CK. The dosimetric consistency between planned and delivered was satisfactory from CK and HA independent of a number of treated lesions.