Monte Carlo Simulation of Flattening Filter Free Linacs With Emphasis on Photoneutron Production.

The aim of this work is to study the effect of removing the flattening filter (FF) from the head of a linear accelerator (linac) on photon beam dosimetric properties and treatment outcomes with emphasis on photoneutron production. An Intensive Monte Carlo (MC) study was carried out for this purpose using two MC codes: FLUKA and BEAMnrx/DOSXYZnrc. These codes were used to model the head of Varian Clinac and/or TrueBeam linacs with and without a FF. For the Varian Clinac, several aspects of the FF removal were addressed. These include: first, study of the dosimetric characteristics of the 15 MV unflattened beam such as dose rate, percentage depth dose on the central axis, beam profile, out-of-field dose, surface dose and neutron contamination. Second, the neutron fluence in treatment rooms housing flattening filter free (FFF) linacs at different energies. Third, the quality of megavoltage cone beam computed tomography (MVCBCT) images by using the 6FFF photon beam with a copper target. Fourth, the feasibility of using the 15FFF beam in intensity modulated radiotherapy (IMRT) for prostate cancer with regard to planning target volume (PTV) coverage, organs at risk sparing (OAR) and neutron dose to the patient was studied. Finally, the dosimetric effect of combining two FFF beams and investigating its practicality was studied. For the Varian TrueBeam, the dosimetric properties of 6 and 10 MV photon beams with and without a FF were calculated using the second version of Varian TrueBeam phase-space files and compared with measurements. In addition, four volumetric modulated radiotherapy (VMAT) plans of a prostate cancer case for these photon beams were modelled using MC techniques for comparison. Results obtained showed that the 15FFF photon beam of the Varian Clinac has a dose rate 4.86±0.09% times higher than the flattened one and the average out-of-field dose from the edge of the field to the edge of the phantom was reduced by 44.0±0.5%. The neutron fluence and neutron dose equivalent at the isocentre were also reduced by 77±3%. However, the photon surface dose of the unflattened beam increased by 13.0=1=0.3%. Moreover, a significant drop-off in neutron fluence at different locations inside the treatment room was found when the FF was removed. For example: the neutron fluence at the isocentre decreased by 54.0±4.2%, 76.0=1=1.4% and 75.0=1=0.8% for 10, 15 and 18 MV, respectively. This would decrease the neutron dose to patient and medical staff as well as reduce the shielding cost of the treatment room. A reduction was observed in the mean energy of the photon spectrum and an increase is obtained in the low energy photon ratio when the 6FFF beam with the copper target was used leading to an improvement in the quality of MVCBCT images. The local contrast in the MVCBCT images when two cylindrical bones with a 2 cm diameter were placed inside a. water phantom with a 5 cm thickness was improved by 31.0±1.3% when the copper target was used. The 15FFF did not provide sufficient coverage to the PTV because the modelled 15FFF plan was not optimised as the hospital treatment planning system did not have the 15FFF data required for the optimising process. However, the calculated neutron dose from a full step and shoot IMRT plan on a water phantom decreased significantly with the use of the 15FFF. Combining the 6 and 15FFF beams provided better sparing of the rectum and bladder and less surface dose. This technique could improve the current IMRT treatments. The calculated dosimetric properties using new TrueBeam phase-space files agreed with measured data. Results of out-of-field dose were found to decrease by 30.4±0.6% and 44.8±0.8% for the unflattened 6 and 10 MV beams, respectively, compared to the flattened beams. The homogeneity index (HI) within the PTV for 6 and 10 photon beams with and without FF plans were found to be 0.08, 0.07, 0.09, and 0.11, respectively. The volume receiving 70 Gy within the bladder were 11.9%, 11.3%, 11.0% and 4.7%, respectively. The VMAT plan treated with the conventional 10 MV beam provided the lowest HI, while the VMAT plan treated with the 10FFF beam provided the highest OAR sparing. All VMAT plans were found to be acceptable clinically. From all results obtained, it can be concluded that the unflattened beam can lead to better treatment outcomes and could provide a step forward towards optimum photon radiotherapy