ESTRO 2024 - Abstract Book

S5351

Radiobiology - Tumour biology

ESTRO 2024

1 Kaiser Permanente - South San Francisco, Radiation Oncology, South San Francisco, USA. 2 Cromwell Hospital, GenesisCare Radiotherapy, London, United Kingdom

Purpose/Objective:

Treatment time duration in stereotactic radiosurgery (SRS) is known to affect the biological effectiveness of a given physical dose 1,2 and this has now been demonstrated in several clinical studies for solid, vascular, and functional targets 3,4,5,6 The mechanism is attributed to sub-lethal damage repair (SLDR) that contains a fast and slow component 7 that is best described with biphasic repair kinetics 8 . The basic linear-quadratic (LQ) model does not account for SLDR because it assumes the entire dose is delivered instantaneously, significantly overestimating the BED delivered in a clinical setting. When SRS treatments are delivered quickly, it is expected that the resultant BED will increase for the same physical dose compared with an equivalent treatment given over a longer exposure time. A simplification 9 of Millar's BED model is applied here to demonstrate the impact of shortened treatment times on the BED in the treatment of vestibular schwannoma (VS).

Material/Methods:

Treatment and prescription data were reviewed for 814 VS patients treated with Gamma Knife (GK) (Elekta, Sweden), Cyberknife (CK) (Accuray, USA) or Truebeam (Varian Medical Systems, USA), from 2001 to 2023. 334 of the 814 patients reviewed were included in the study and received either 12 Gy in one fraction (238) or 18 Gy in three fractions (96). The 12 Gy/1# cohort were treated using GK Model B (76), GK Perfexion (56), CK VSI (80) and Truebeam (26). The 18 Gy/3# cohort patients were treated with CK (40) or Truebeam (56). GK plans were created using TMR10 algorithm in GammaPlan, employing 2 - 36 isocenters (mean of 4.9 on Model B; 12.9 on Perfexion) and a 60Co dose rate of 3.5 - 1.5 Gy/min. CK plans used Ray Tracing algorithm in Multiplan and had 56 - 219 beams with 1 - 3 fixed cones and a dose rate of 1000 MU/min using 6MV. Truebeam plans used the Pencil Beam algorithm in the Cranial SRS Element (Brainlab, Germany) with 3 – 5 arcs and a dose rate of up to 1400 MU/min using 6FFF. A subset of Truebeam patients had an added GTV-to-PTV margin of 0.5 mm (42) or 1 mm (2). Treatment data collected for analysis included physical prescription dose and fractionation, number of treatment isocenters/beams/arcs, total treatment time from first beam-on to last beam-off, total beam-on time, average time between each treatment isocenter/beam/arc and dose received by 95% of the GTV. Using an α/β = 2.47 Gy, repair half-times of 11.4 and 129.6 minutes, a partition coefficient, x, of 0.987, and each patient’s extracted treatment data, the BED 2.47 was calculated and compared. Due to the use of margins for some Truebeam patients the GTV D95% was used for BED 2.47 to permit comparison to GK and CK treatments where no margins were used.

Results:

BED 2.47 for 12Gy/1# and 18Gy/3# calculated from the basic LQ model were 70.3 Gy 2.47 and 61.7 Gy 2.47 , respectively, regardless of treatment time. Treatment times varied between platforms with 17 - 127 min and 18 - 84 min for Model B and Perfexion, respectively, 17 - 79 min for CK and 4 - 19 minutes for Truebeam. Accounting for fast and slow repair components, using Equation A9 9 , resulted in varying BED 2.47 values, even for the same prescription dose. Table 1 and figure 1 illustrate the range of BED 2.47 for each treatment modality for 12 Gy/1# or 18 Gy/3#, respectively, and include the BED 2.47 as predicted by the LQ model for each fractionation scheme for the 12 Gy example shown in Figure 1.