WORK GROUP ON IMRT REPORT Written By WGIMRT

Article Watch 2016

Planning

Ahanj et al. created a planning technique that boosts the superior/inferior edges of a lung tumor to account for respiratory-induced target motion. The edge-enhancing boost was found to reduce lung dose compared to the ITV treatment of moving lung targets. However, their technique was more sensitive to changes in tumor size and fluctuations in the patient's respiratory motion.

Ziegenhein et al. presented a new planning model called interactive dose shaping (IDS) that uses user-requested dose modifications instead of a traditional optimization cost function. The algorithm works in two phases: 1) first, it adapts the plan to conform to the requested modification, 2) next it attempts to recover undesired disturbances of the dose pattern caused by the modification. The IDS optimization has been combined with an "ultra-fast" dose update calculation method to form a new planning system Dynaplan. The fast dose calculation and modification nature of optimization are ideal for adaptive planning for changing patient anatomies.

Irene Hazell et al. reported a significant clinical study result stating that Philips Pinnacle TPS's Auto-Planning module can automatically generate clinical acceptable IMRT plans for virtually all cases in their study. They included a total of 26 head and neck IMRT cases in their comparison — to compare automatically generated plans from Pinnacle auto-planning module, and dosimetrist-optimized plans on the same patients. A blind plan quality clinical review is conducted as well as direct DVH comparisons. Each plan is given a score between 1 and 6. They reported that for 94% of cases, auto-planning was able to generate IMRT plans having at least as high as a score as the manually optimized IMRT plans from dosimetry. This is certainly promising news for one direction in IMRT treatment planning – that is automation and class solutions.

Kamran et al. studied the potential benefit of using multi-criteria optimization (MCO) for planning non-small cell lung IMRT cases. RayStation's MCO algorithm was used, which allows the planner to vary OAR/target criteria interactively, to see how improving one criterion worsens another. 5-field IMRT plans were created in RayStation using MCO and non-MCO planning techniques. Planning time was capped at 4 hours. MCO plans required less planning time and resulted in improved OAR sparing, while maintaining target coverage.

Dosimetry

Bailey et al. reported on the use of a measurement uncertainty correction function when calculating gamma pass rates for IMRT QA using MapCHECK. The use of the correction function—enabled by default—was found to have a noticeable effect upon QA results. The magnitude of the effect was both plan type and technique dependent with a maximum change in pass rate of 8.7% for a single field when 3%/3mm criteria were used. The authors conclude that the use of this uncertainty function should be specified in reports of MapCHECK-based QA results in the literature.

Chuter et al. demonstrated the ability to use EPID-based in vivo portal dosimetry in FFF radiotherapy. It was noted that while the Elekta iView EPID saturated at a dose rate of 800 MU/min — below the maximum FFF dose rate for the linac — FFF beams limited to this dose rate were successfully modelled using the procedures developed for flattened beams despite the differing spectra of their FFF counterparts. It was concluded that in vivo portal dosimetry in its current iteration was viable for FFF treatments limited to lower dose rates.

Kim et al. reported the results of a multi-institutional IMRT QA study. The goal was to evaluate confidence limits for both point dose and planar QA in the style of TG-119, and to determine if the confidence limit concept was appropriate for the data. Across all plans they report confidence limits of 2.7% for point doses in high dose regions, 6.2% for point doses in low dose regions, and 93.7% pass rate on for single field planar gamma analysis (3%/3mm). However they conclude that the concept of confidence limits was appropriate only for point dose measurements due to the lack of normality in the distribution of planar QA results.

Steers and Fraass presented a study of the sensitivity of gamma analysis in detecting delivery errors as a function of DTA, % dose error, and threshold %. Intentional errors are introduced to IMRT plans and several combinations of gamma criterion are evaluated. The authors promote the concept of an "error curve" which is stated to represent the sensitivity of each gamma criterion to a particular induced error. From the evaluation of these error curves the authors determine that the typical 3%/3mm/10%TH gamma criterion could allow clinically meaningful errors to go unnoticed at a 90% pass rate cutoff. Instead it is suggested that 3%/2mm/50%TH or 2%/3mm/50% with 90% to 95% pass rate cutoff would be more clinically valuable.

Ricketts et al. reported on their experience using EPID-based transit dosimetry and their efforts to create dose-deviation action limits. The authors evaluated reconstructed point doses in 58 patients using this transit dosimetry system, including breast, prostate and head and neck cases. The authors conclude that, for their particular system, the reconstructed point doses for IMRT treatments were consistently lower than the planned values, and that asymmetric action levels were required as a result of this. Action levels were set at -6% ± 7% for their system.

Pérez et al. attempted to mitigate lateral artifacts in EBT3 dosimetry by using an additional thin linear polarized film during the initial EBT3 scanning process. A thorough description of the typical optical scanning system is presented, and the effects on the response of the system of the inherent polarization of the light transmitted through an EBT3 film are analyzed. These polarization effects are shown to be counteracted by using a polarized light source for transmission scanning. This can be done by placing a thin, linearly polarized sheet of film on top of the EBT3 during scanning.

Knill et al. presented an investigation of ion recombination effects in a liquid-filled ion chamber array designed for SRS and SBRT dosimetry. The relationship between collection efficiency and dose per pulse, pulse frequency, and energy are studied for VMAT SBRT plans using both 6MV and 10FFF beams. It is concluded that ion recombination effects are negligible for 6MV, but the increased dose per pulse for the unflattened beam could result in up to a 4.8% difference in measured 2%/2mm gamma pass rates with and without corrections for the ion recombination effects.

Palma et al. performed a planning study comparing "conventional" VMAT versus theoretical very high-energy electron (VHEE) treatments. BEAMnrc/EGSnrc was used to create the initial phase space files for VHEE that were imported into RayStation. Optimization was done using RayStation's proton pencil beam scanning optimization algorithm. VHEE treatments consisted of sixteen to thirty–two 100MeV electron beams delivered using pencil beam scanning in a fixed-gantry coplanar technique. VHEE treatments produced comparable coverage and better OAR sparing, especially for centrally located tumors >4cm. The authors hope the advantages of VHEE will motivate the design of a VHEE treatment machine.

MacFarlane et al. compared unified intensity-modulated arc therapy (UIMAT), a technique involving both static-gantry IMRT and VMAT, to IMRT and VMAT alone. UIMAT begins with the optimization of multiple static beams distributed along an arc range. Beams with fewer segments are converted into VMAT phases while beams with more segments are converted into IMRT phases. The IMRT/VMAT phases are optimized simultaneously and finally merged into a single plan. UIMAT was tested for 30 head and neck cases and was found to produce better target coverage with increased OAR sparing.

Back to Article List Next Previous