ESTRO 2021 Abstract Book

S1539

ESTRO 2021

PO-1810 Potential clinical benefits of FLASH compared with fractionation in scanned proton therapy M. Krieger 1,2 , S. Van de Water 2 , M.M. Folkerts 3 , A. Mazal 4 , S. Fabiano 2,5,6 , N. Bizzocchi 2 , D.C. Weber 2,6,7 , S. Safai 2 , A.J. Lomax 2,5 1 Varian Medical Systems International AG, FLASH, Steinhausen, Switzerland; 2 Paul Scherrer Institute, Center for Proton Therapy, Villigen PSI, Switzerland; 3 Varian Medical Systems Inc., FLASH, Palo Alto, USA; 4 Quironsalud, Centro de Protonterapia, Madrid, Spain; 5 ETH Zurich, Department of Physics, Zurich, Switzerland; 6 University Hospital Zurich, Department of Radiation Oncology, Zurich, Switzerland; 7 University of Bern, Department of Radiation Oncology, Bern, Switzerland Purpose or Objective To explore the potential FLASH effect in pencil beam scanned (PBS) proton therapy for a variety of patient cases and planning approaches as well as to compare the clinical benefits of FLASH with fractionation effects. Materials and Methods FLASH effects were estimated using a quantitative FLASH effectiveness model. In this model, each dose contribution to any voxel triggers FLASH if the dose is >=5Gy and the dose rate is >=40Gy/s. FLASH is considered to persist for 200ms after the trigger has ended, regardless of the dose and dose rate. All dose contributing to the FLASH trigger as well as any dose delivered within the persistence time is multiplied by a FLASH effectiveness factor of 0.67 (i.e., 33% sparing) to obtain the FLASH dose distribution in healthy tissues. This model was used to estimate the relative integral dose (ID) reduction through FLASH for a variety of anatomical sites (brain, lung, nasal cavity, pancreas, prostate) and for 3 PBS plans each, assuming a Varian ProBeam machine: a) multi-field, upstream degrader as a clinical reference plan (Ref), b) single-field, downstream range-shifter to exploit the increased beamline transmission efficiency at high energies, c) single-field transmission. Clinical benefit of FLASH was quantified as the ID reduction when comparing any plan’s FLASH dose with the Ref’s physical dose. FLASH simulations were performed for a single fraction of 22.3Gy. To calculate fractionation effects, prescription doses for a given number of fractions were chosen to be equivalent to 30x2Gy using a representative tumour alpha/beta ratio of 10Gy. The fractionation effects were quantified as the relative reduction in integral biological effects of any fractionation scheme compared with 1x22.3Gy and assuming a representative healthy tissue alpha/beta ratio of 4Gy, without considering FLASH effects. The number of fractions for which FLASH outweighed fractionation was then obtained. Results Figure 1a shows the FLASH-induced ID reduction with mean(standard deviation) values for all patients being 0.7(0.4)%, 9.6(2.9)% and 30.8(1.2)% for Ref, downstream and transmission plans, respectively. The numbers of fractions for which FLASH reductions were greater than fractionation effects are shown in Figure 1b and average 1.0(0.0), 1.1(0.3) and 2.9(0.9), respectively.

Representative dose and FLASH distributions for one pancreas case are shown in Figure 2a. Clinical benefits, as shown in Figure 2b, varied largely among patients and were 0.7(0.4)%, -1.6(11.1)% and -15.1(18.0)% for Ref, downstream and transmission plans, respectively.