Pre-operative versus post-operative radiosurgery for brain metastasis: Effects on treatment volume and inter-observer variability.

PURPOSE/OBJECTIVES: Compared to post-operative whole brain radiotherapy, resection cavity radiosurgery reduces impact on neuro-cognitive function and improves quality-of-life. However, coverage of the operative tract, in addition to tumour bed, may lead to large treatment volumes and inter-observer variability. We hypothesized that pre-operative radiosurgery reduces target volume size and inter-observer variability compared to post-operative radiosurgery. MATERIALS/METHODS: We identified 10 consecutive patients, with solitary brain metastasis, treated with post-operative cavity radiosurgery.Pre- and post-operative axial T1 contrast MRI were co-registered with the planning CT scans. Three radiation oncologists independently contoured the target volumes on the pre- and post-operative imaging. A 2mm-PTV margin was utilized for both strategies and radiosurgery treatment plans were generated. The following parameters were evaluated in the 2 plans: Mean target volume (cc), 50% isodose volume (cc), Inter-observer variability (Jaccard Index JI) and Conformity Index (CI). RESULTS: There was no significant difference in the mean target volume, nor 50% isodose volume, between pre- and post-operative strategies. (17.6 (95% CI 9.98 - 25.22) versus 19.4 (95% CI 10.11 - 28.69) cc, P=0.80; 61.7 (95% CI 38.4 - 85.0) vs 77.7 (95% CI 34.94 - 120.46) cc, P=0.65). There was significantly less inter-observer variability and improved conformity in the pre-operative group (Mean JI 0.84(95% CI 0.82 - 0.86) versus 0.70 (95% CI 0.62 - 0.78), P = 0.005; Mean CI 1.32 (95% CI 1.26 - 1.38) vs 1.45 (95% CI 1.36 - 1.54), P= 0.01). Planned subgroup analysis did not reveal any significant difference (between pre- vs post-op) in the mean volume of cystic versus non-cystic metastasis. Deep lesions (>2.5cm from dura) had a larger post-operative target volume (25.8 (95% CI 15.1 - 36.5) vs 12.3 (95% CI 6.54 - 18.06) cc, P=0.06) compared to superficial lesions. CONCLUSION: Pre-operative radiosurgery has less inter-observer variability and improved plan conformity. However, there was no difference in mean target volume between the pre- versus post-operative radiation. Contouring guidelines, and peer review, may help to reduce inter-observer variability for cavity radiosurgery.

Small composite field correction factors for the CyberKnife radiosurgery system: clinical and PCSR plans.

A formalism has been proposed for small and non-standard photon fields in which [Formula: see text] correction factors are used to correct dosimeter response in small fields (indiviual or composite) relative to that in a larger machine-specific reference (MSR) field. For clinical plans consisting of several fields, a plan-class specific reference (PCSR) plan can also be defined, serving as an intermediate calibration field between the MSR and clinical plans within a certain plan-class. In this work, the formalism was applied in the calculation of [Formula: see text] for 21 clinical plans delivered by the [Formula: see text] radiosurgery system, each plan employing one or two of the smallest diameter collimators: 5 mm, 7.5 mm, and 10 mm. Three detectors were considered: the Exradin A16 and A26 micro chambers, and the W1 plastic scintillator. The clinical plans were grouped into 7 plan-classes according to commonly shared characteristics. The suitability of using a PCSR plan to represent the detector response of each plan within the plan-class was investigated. Total and intermediate correction factors were calculated using the [Formula: see text] Monte Carlo user code. The corrections for the micro chambers were large, primarily due to the presence of the low-density air cavity and the volume averaging effect. The correction for the scintillator was found to be close to unity for most plans, indicating that this detector may be used to measure small clinical plan correction factors in any plan except for those using the 5 mm collimator. The PCSR plan was shown to be applicable to plan-classes comprising isocentric plans only, with plan-classes divided according to collimator size. For non-isocentric plans, the variation of [Formula: see text] as a function of the point of measurement within a single plan, as well as the high inter-plan-class variability of the correction factor, precludes the use of a PCSR plan.

Use of novel fibre-coupled radioluminescence and RADPOS dosimetry systems for total scatter factor measurements in small fields.

A dosimetry system based on Al2O3:C radioluminescence (RL), and RADPOS, a novel 4D dosimetry system using microMOSFETs, were used to measure total scatter factors, (S(c,p))(f(clin))(det), for the CyberKnife robotic radiosugery system. New Monte Carlo calculated correction factors are presented and applied for the RL detector response for the 5, 7.5 and 10 mm collimators in order to correct for the detector geometry and increased photoelectric cross section of Al2O3:C relative to water. For comparison, measurements were also carried out using a micro MOSFET, PTW60012 diode and GAFCHROMIC(®) film (EBT and EBT2). Uncorrected (S(c,p))(f(clin))(det) were obtained by taking the ratio of the detector response for each collimator to that for the 60 mm diameter reference field. Published Monte Carlo calculated correction factors were applied to the RADPOS, microMOSFET and diode detector measurements to yield corrected field factors, Ω(f(clin),f(msr))(Q(clin),Q(msr)), following the terminology of a recent formalism introduced for small and composite field relative dosimetry. With corrections, the RL measured Ω(f(clin),f(msr))(Q(clin),Q(msr)) were 0.656 ± 0.002, 0.815 ± 0.002 and 0.865 ± 0.003 for the 5, 7.5 and 10 mm collimators, respectively. This was in good agreement with RADPOS corrected field factors of 0.650 ± 0.010, 0.816 ± 0.024 and 0.867 ± 0.010 for the 5, 7.5 and 10 mm collimators, respectively. Both RL and RADPOS total scatter factors agreed within approximately two standard deviations of the GAFCHROMIC film values (average of EBT and EBT2) of 0.640 ± 0.006, 0.806 ± 0.007 and 0.859 ± 0.09. Corrected total scatter factors for all dosimetry systems agreed within one standard deviation for collimator sizes 10-60 mm. Our study suggests that the microMOSFET/RADPOS and optical fibre-coupled RL dosimetry system are well suited for total scatter factor measurements over the entire range of field sizes, provided that appropriate correction factors are applied for the collimator diameters smaller than 10 mm.

Sci-Thur AM: Planning - 10: Improved dosimetric accuracy for patient specific quality assurance using a dual-detector measurement method for cyberknife output factors.

The measurement of output factors for small fields is challenging and can lead to large dose errors in patient treatments if corrections for detector size and scatter from high-Z material are not applied. Due to its high spatial resolution and near tissue equivalence, GAFCHROMIC® film potentially provides a correction free measure of output factors but it can be challenging to obtain high quality dosimetric results using this film. We propose minimizing errors in the clinical determination of small field output factors by employing diode measurements with Monte-Carlo generated corrections for small fields ≤10 mm diameter and using small volume ion chambers for apertures >10 mm diameter with independent validation using radiochromic film. We performed patient specific quality assurance (QA) measurements for 9 patients using GAFCHROMIC® film and an A16 small volume ion chamber in a head-shaped phantom, employing this hybrid dual detector method for relative output factor measurements within the Multiplan treatment planning system. Our results suggest that consistent output factors can be determined using this method with experimental verification using GAFCHROMIC® film dosimetry. For the patient specific QA using film, we achieve good dosimetric agreement (<2σ) of the measured and calculated average dose for pixels within the 80% isodose line. For patient specific QA using the micro-ion chamber, we get good agreement (<3%) for cone sizes greater than 5 mm. The differences observed for the 5 mm cone plans are consistent with a 1 mm radial setup uncertainty for patient positioning using the Cyberknife system.

Sci-Fri PM: Delivery - 07: Cyberknife relative output factor measurements using fiber-coupled luminescence, MOSFETS and RADPOS dosimetry system.

Novel dosimetry systems based on Al2 O3 :C radioluminescence (RL) and a 4D dosimetry system (RADPOS) from Best Medical Canada were used to measure the relative output factor (ROF) on Cyberknife. Measurements were performed in a solid water phantom at the depth of 1.5 cm and SSD = 78.5 cm for cones from 5 to 60 mm. ROFs were also measured using a mobileMOSFET system (Best Medical Canada) and EBT1 and EBT2 GAFCHROMIC® (ISP, Ashland) radiochromic films. For cone sizes 12.5-60 mm all detector results were in agreement within the measurement uncertainty. The microMOSFET/RADPOS measurements (published corrections applied) yielded ROFs of 0.650 ± 1.9%, 0.811 ± 0.9% and 0.843 ± 1.7% for the 5, 7.5 and 10 mm cones, respectively, and were in excellent agreement with radiochromic film values (averaged for EBT1 and EBT2) of 0.645 ± 1.4%, 0.806 ± 1.1% and 0.859 ± 1.1%. Monte-Carlo calculated correction factors were applied to the RL readings to correct for excessive scatter due to the relatively high effective atomic number of Al2 O3 (Z=10.2) compared to water for the 5, 7.5 and 10 mm cones. When these corrections are applied to our RL detector measurements, we obtain ROFs of 0.656 ± 0.3% and 0.815 ± 0.3% and 0.865 ± 0.3% for 5, 7.5 and 10 mm cones. Our study shows that the microMOSFET/RADPOS and optical fiber-coupled RL dosimetry system are well suited for Cyberknife cone output factors measurements over the entire range of field sizes, provided that appropriate correction factors are applied for the smallest cone sizes (5, 7.5 and 10 mm).