SP-0399: FFF and new dosimetric challenges
Radiotherapy and Oncology
Current normal tissue complication probability (NTCP) models rely heavily on dosimetric parameters that are derived from dose-volume histograms (DVHs) which lack the spatial information. These parameters assume a homogenous response to radiation and tend to ignore any inter or intra-organ variation in radiosensitivity. To address these issues in our analysis, we developed an image-based approach which utilizes the full 3D dose distribution and does not assume any prior knowledge of the OARs.
... ng image registration, we deformed CT scans, anatomic structures, and 3D dose distributions from a patient cohort onto the same frame of reference "standard patient". Voxel-based correlations between dose and the complication endpoint (trismus) were computed over the entire voxel space to construct a 3D correlation map. We applied this methodology to a cohort of 110 oropharyngeal cancer (OPC) patients treated with IMRT and Chemotherapy at MSKCC between 2004 and 2009. The results showed a good agreement between the image-based approach and the standard NTCP models. Both approaches showed a statistically significant correlation between the dose to mastication muscles and trismus. Furthermore, the voxelwise 3D correlation map provided additional information and revealed regions of high correlation outside the mastication muscle region which might suggest the involvement of other organs or tissues at risk. However, the clinical implication of these new findings needs further investigation in an independent data set and might be related to beam arrangement and configuration. In addition, we also developed a new metric to estimate the geometric uncertainty of the underlying deformable registration which we denote as distance discordance (DD). More technical details and a few examples will be presented to illustrate the concept of DD and its impact on the uncertainty of dose mapping. In summary, this study demonstrates the validity of this approach and emphasizes the need to consider the full 3D dose distribution in conjunction with DVH analysis. Many new treatment units are already on the market that operate without a flattening filter (FF) or can be operated in a dedicated clinical flattening filter free (FFF) mode, i.e. Tomotherapy, Cyberknife, Varian's True-Beam or ELEKTA's Agility linacs, respectively. The commonly accepted dosimetric benefits of FFF high energy photon beams, which range from increased dose rate and dose per pulse, to favorable output ratio in-air variation with field size, reduced energy variation across the beam, reduced leakage and outof-field dose, respectively, imply new dosimetric challenges. These are related to beam quality specification, beam calibration, small field dosimetry, acceptance testing/commissioning, regular quality assurance procedures, and finally treatment plan verification. As far as the acquisition of basic relative beam parameters is concerned, the commissioning of a FFF treatment delivery unit is not very different from a standard linac. However, the increased dose rate and the high dose per pulse needs to be considered,since some point detectors show a pronounced dose rate dependency (e.g. diamond detector, liquid filled ionization chamber), possible saturation or reduced collection efficiency. The forward peaked dose distribution of FFF beams is fundamentally different and makes standard parameters as used for quality control of linacs, e.g. homogeneity, field flatness and beam penumbra, not always viable. Recently suggestions have been made to overcome this issue in clinical dosimetry. Small field dosimetry is of special importance since all relevant clinical applications of FFF beams are closely related to high precision radiotherapy techniques, i.e. stereotactic RT or IMRT, where small fields/segments build the basis for treatment delivery. Today's challenge of small beam dosimetry and its link to absorbed dose determination in reference geometry is enhanced by unflattened beams and dedicated high precision radiotherapy delivery units that do not allow to setup the reference conditions, as defined by the IAEA TRS 398 or AAPM TG 51 codes of practice. Incorrectly measured beam output in small beams can result in systematic uncertainties when implementing basic beam data in treatment planning systems(TPS), misleading results in acceptance testing/ commissioning of a TPS, and finally erroneous patient treatments leading to radiation incidents or accidents. Therefore correction factors for a variety of commercially available small ion chambers and other radiation detectors are currently being investigated in order to extend their use towards small beam dosimetry in FFF photon beams. The accuracy of dose calculation in TPS represents another dosimetric aspect. Several studies demonstrated that inverse planning tools for IMRT are able to handle unflattened beams and that the dose calculation accuracy of commercial TPS is not impaired when removing the FF. The advent of the Flattening Filter Free beams (FFF) available on common linacs, is facing the community to the ability of treatment planning system to accurately model such beams, and with the understanding of which are the best sites and treatments which could take advantages from using FFF beams. An overview of the published data on dose calculation algorithms accuracy for FFF fields is given. The characteristics of the FFF beams are here analyzed to evaluate the clinical sites where such beams could take advantage. The main characteristics relate to the out-of-field dose (peripheral dose) that could lower the organs at risk dose; to the high dose rate, allowing shorter treatment time for stereotactical treatments; to the profile shape that could be better suitable for small targets or simultaneous integrated boost treatments. A review of published planning studies presents plan comparisons in different sites. Prostate, lung, brain, head and neck plans were evaluated for IMRT treatment with different beams (6 and 18 MV, flattened and unflattened). Also whole breast plans (with different techniques from tangents to IMRT) and esophageal treatment (conformal, IMRT, VMAT) where analysed by different groups. Results of such studies presenting rather large volumes showed no significant difference in plan quality between standard flattened and FFF beams. A substantial reduction of MUs and consequently of radiation outside the target was evidentiated for most of sites, although the opposite could be found for very large (or very elongated) targets, where the lower off-axis dose would play an unbeneficial role. The lower radiation dose outside the target was also studied on anthropomorphic phantom to simulate pediatric IMRT treatments with FFF beams in the intracranial regions. The measured dose reduction in the periphery (from thyroid to ovaries and testes) was lower of 23 to 70% relative to the standard flattened beams. Particular attention is paid to stereotactical treatments, where the high dose rate is the clear and natural benefit in treating with FFF beams. Studied sites were early stage lung treatments, liver metastases and pelvic/abdominal metastases. Examples of lung and brain stereotactical comparisons with and without flattening filter are given. Conclusions: there are benefits in using FFF beams for some treatment site and lesion sizes. The primary benefit is, for high dose per fraction treatments, the high dose rate that reduces the beam-on time. Other proven benefits are the lower out-of-field and peripheral dose, that decrease the critical structures dose and the risk of secondary cancer induction. Flattening filter free (FFF) options are now commercially available and treatments utilising the higher dose-rates obtained are being introduced into clinical use. Removal of the flattening filter results in an increase, by a factor of 2-6 times, in the instantaneous dose-rate of the x-ray pulse compared to the standard flattened beams. The increased dose per pulse, result in an equivalent increase in the average dose-rate and a potentially similar reduction in overall treatment time. The potential radiobiological factors that result from changing the instantaneous dose-rate include potential changes to sub-lethal damage repair mechanisms, potential synergistic effects and the potential radiobiological effects resulting from changes to the overall treatment time. A number of in-vitro studies have been undertaken using x-ray beams from linear accelerators designed specifically to study the implications of changes in dose-rate of the treatment beams resulting from treating with flattening filter free beams. These studies have used a variety of experimental methodologies to achieve the required instantaneous dose-rates and differing ways to ensure dose uniformity over the in-vitro flask. These differences in the experimental methodologies have not led to