PURPOSE/OBJECTIVES:(A) To examine the alignment accuracy of CBCT guidance for brain metastases with off centered isocenters, (B) to test dose delivery and targeting accuracy for single isocenter treatments with multiple brain metastases. We report the results of the end-to-end test for Truebeam stereotactic radiosurgery (SRS).MATERIALS/METHODS:An anthropomorphic CT head phantom was drilled with five MOSFET inserts and two PTW Pinpoint chamber inserts. The phantom was simulated, planned, and delivered. For the purpose of comparing the accuracy of alignment, CBCTs were acquired with the isocenter centered and offset superiorly 8 cm, inferiorly 8 cm, anteriorly 7 cm, posteriorly 7 cm, and right 5 cm. There were six degrees of freedom corrections applied to the plans, as well as intentional rotational and translational errors for dose comparisons. Dose accuracy checks were performed with MOSFET and PTW Pinpoint chamber, and targeting accuracy was assessed with GafChromic films.RESULT:(A) Compared to centered CBCT, off-centered CBCT scan showed some alignment errors, with a maximum difference of 0.6-degree pitch and 0.9 mm translation when the phantom was placed 8 cm inferior off center. (B) For the single isocenter plan, measured doses of the five MOSFET were 95%-100% of the planned dose, whereas the multiple isocenter plans were 96%-100%. With intentional setup errors of 1-degree pitch, doses were 97.1%-100.4% compared to the perfect setup. The same was found for the two pinpoint chamber readings with 1-degree rotation and 1 mm translation. (C) Targeting accuracy for targets at the isocenter is 0.67 mm, within the machine specification of 0.75 mm. Targeting accuracy for isocenters 6-12 cm away from the target is in the range 0.67-1.18 mm.CONCLUSION:(A) Single isocenter HyperArc treatments for multiple brain metastases are feasible and targeting accuracy is clinically acceptable. (B) The vertex in a cranial scan is very important for proper alignment.

OBJECTIVE:Preoperative stereotactic radiosurgery (SRS) is emerging as a viable alternative to standard postoperative SRS. Studies have suggested that preoperative SRS provides comparable tumor control and overall survival (OS) and may reduce the incidence of leptomeningeal disease (LMD) and adverse radiation effects (AREs). It is unknown, however, if preoperative SRS remains effective in cohorts including large brain metastases (> 14 cm3) or if preoperative SRS affects steroid taper/immunotherapy. Here, the authors report the results of a phase 2 single-arm trial assessing a prospectively acquired series of 26 patients who underwent preoperative SRS, without a volumetric cutoff, compared with a propensity score-matched concurrent cohort of 30 patients who underwent postoperative SRS to address these salient questions. METHODS:Demographics, oncological history, surgical details, and outcomes were collected from the medical records. Coprimary endpoints were local tumor control (LTC) and a composite outcome of LTC, ARE, and LMD. Additional outcomes were OS, steroid taper details, and immunotherapy resumption. For survival analyses, cohorts were propensity score matched. RESULTS:Preoperative and postoperative SRS patients were comparable in terms of age, sex, Karnofsky Performance Status score, oncological history, and operative details. Gross tumor volume (GTV) was significantly higher in the preoperative group (median 12.2 vs 5.3 cm3, p < 0.001). One-year LTC (preoperative SRS: 77.2% vs postoperative SRS: 82.5%, p = 0.61) and composite outcome (68.3% vs 72.7%, p = 0.38) were not significantly different between the groups. In multivariable analysis, preoperative SRS did not have a significant effect on LTC (HR 1.57 [95% CI 0.38-6.49], p = 0.536) or the composite outcome (HR 1.18 [95% CI 0.38-3.72], p = 0.771), although the confidence intervals were large. The median OS (preoperative SRS: 17.0 vs postoperative SRS: 14.0 months, p = 0.61) was not significantly different. Rates of LMD were nonsignificantly lower in the preoperative SRS group (3.8% vs 16.7%, p = 0.200). Greater GTV volume was associated with prolonged (> 10 days) steroid taper (OR 1.24 [95% CI 1.04-1.55], p = 0.032). However, in multivariable analysis, preoperative SRS markedly reduced the steroid taper length (OR 0.13 [95% CI 0.02-0.61], p = 0.016). Time to immunotherapy was shorter in the preoperative SRS group (36 [IQR 26, 76] vs OR 228 [IQR 129, 436] days, p = 0.02). CONCLUSIONS:Compared with postoperative SRS, preoperative SRS is a safe and effective strategy in the management of cerebral metastases of all sizes and provides comparable tumor control without increased adverse effects. Notably, preoperative SRS enabled rapid steroid taper, even in larger tumors. Future studies should specifically examine the interaction of preoperative SRS with steroid usage and resumption of systemic therapies and the subsequent effects on systemic progression and OS.

The use of small fields in radiation therapy techniques has increased substantially in particular in stereotactic radiosurgery (SRS) and stereotactic body radiation therapy (SBRT). However, as field size reduces further still, the response of the detector changes more rapidly with field size, and the effects of measurement uncertainties become increasingly significant due to the lack of lateral charged particle equilibrium, spectral changes as a function of field size, detector choice, and subsequent perturbations of the charged particle fluence. This work presents a novel 3D dose volume-to-point correction method to predict the readings of a 0.015 cc PinPoint chamber (PTW 31014) for both small static-fields and composite-field dosimetry formed by fixed cones on the CyberKnife® M6™ machine. A 3D correction matrix is introduced to link the 3D dose distribution to the response of the PinPoint chamber in water. The parameters of the correction matrix are determined by modeling its 3D dose response in circular fields created using the 12 fixed cones (5 mm-60 mm) on a CyberKnife® M6™ machine. A penalized least-square optimization problem is defined by fitting the calculated detector reading to the experimental measurement data to generate the optimal correction matrix; the simulated annealing algorithm is used to solve the inverse optimization problem. All the experimental measurements are acquired for every 2 mm chamber shift in the horizontal planes for each field size. The 3D dose distributions for the measurements are calculated using the Monte Carlo calculation with the MultiPlan® treatment planning system (Accuray Inc., Sunnyvale, CA, USA). The performance evaluation of the 3D conversion matrix is carried out by comparing the predictions of the output factors (OFs), off-axis ratios (OARs) and percentage depth dose (PDD) data to the experimental measurement data. The discrepancy of the measurement and the prediction data for composite fields is also performed for clinical SRS plans. The optimization algorithm used for generating the optimal correction factors is stable, and the resulting correction factors were smooth in the spatial domain. The measurement and prediction of OFs agree closely with percentage differences of less than 1.9% for all the 12 cones. The discrepancies between the prediction and the measurement PDD readings at 50 mm and 80 mm depth are 1.7% and 1.9%, respectively. The percentage differences of OARs between measurement and prediction data are less than 2% in the low dose gradient region, and 2%/1 mm discrepancies are observed within the high dose gradient regions. The differences between the measurement and prediction data for all the CyberKnife based SRS plans are less than 1%. These results demonstrate the existence and efficiency of the novel 3D correction method for small field dosimetry. The 3D correction matrix links the 3D dose distribution and the reading of the PinPoint chamber. The comparison between the predicted reading and the measurement data for static small fields (OFs, OARs and PDDs) yield discrepancies within 2% for low dose gradient regions and 2%/1 mm for high dose gradient regions; the discrepancies between the predicted and the measurement data are less than 1% for all the SRS plans. The 3D correction method provides an access to evaluate the clinical measurement data and can be applied to non-standard composite fields intensity modulated radiation therapy point dose verification.

BACKGROUND:Historically, survival for even highly select cohorts of brain metastasis patients selected for SRS alone is <2 yr; thus, limited literature on risks of recurrence exists beyond 2 yr. OBJECTIVE:To investigate the possibility that for subsets of patients the risk of intracranial failure beyond 2 yr is less than the commonly quoted 50% to 60%, wherein less frequent screening may be appropriate. METHODS:As a part of our institutional radiosurgery database, we identified 132 patients treated initially with stereotactic radiosurgery (SRS) alone (± pre-SRS surgical resection) with at least 2 yr of survival and follow-up from SRS. Primary study endpoints were rates of actuarial intracranial progression beyond 2 yr, calculated using the Kaplan-Meier and Cox regression methods. RESULTS:The median follow-up from the first course of SRS was 3.5 yr. Significant predictors of intracranial failure beyond 2 yr included intracranial failure before 2 yr (52% vs 25%, P < .01) and total SRS tumor volume ≥5 cc (51% vs 25%, P < .01). On parsimonious multivariate analysis, failure before 2 yr (HR = 2.2, 95% CI: 1.2-4.3, P = .01) and total SRS tumor volume ≥5 cc (HR = 2.3, 95% CI: 1.2-4.3, P = .01) remained significant predictors of intracranial relapse beyond 2 yr. CONCLUSION:Relapse rates beyond 2 yr following SRS alone for brain metastases are low in patients who do not suffer intracranial relapse within the first 2 yr and with low-volume brain metastases, supporting a practice of less frequent screening beyond 2 yr. For remaining patients, frequent (every 3-4 mo) screening remains prudent, as the risk of intracranial failure after 2 yr remains high.

We performed an evaluation of the CyberKnife InCise MLC by comparing plan qualities for single and multiple brain lesions generated using the first version of InCise MLC, fixed cone, and Iris collimators. We also investigated differences in delivery efficiency among the three collimators. Twenty‐four patients with single or multiple brain mets treated previously in our clinic on a CyberKnife M6 using cone/Iris collimators were selected for this study. Treatment plans were generated for all lesions using the InCise MLC. Number of monitor units, delivery time, target coverage, conformity index, and dose falloff were compared between MLC‐ and clinical cone/Iris‐based plans. Statistical analysis was performed using the nonparametric Wilcoxon‐Mann‐Whitney signed‐rank test. The planning accuracy of the MLC‐based plans was validated using chamber and film measurements. The InCise MLC‐based plans achieved mean dose and target coverage comparable to the cone/Iris‐based plans. Although the conformity indices of the MLC‐based plans were slightly higher than those of the cone/Iris‐based plans, beam delivery time for the MLC‐based plans was shorter by . For smaller targets or cases with OARs located close to or abutting target volumes, MLC‐based plans provided inferior dose conformity compared to cone/Iris‐based plans. The QA results of MLC‐based plans were within 5% absolute dose difference with over 90% gamma passing rate using gamma criteria. The first version of InCise MLC could be a useful delivery modality, especially for clinical situations for which delivery time is a limiting factor or for multitarget cases.PACS number(s): 87.53.Ly, 87.55.D‐

Purpose:The InCise™ Multileaf Collimator (MLC) of CyberKnife M6™ System has been released recently. The purpose of this study was to explore the dosimetric characteristics of the new MLC. In particular, the penumbra characteristics of MLC fields at varying locations are evaluated.Methods:EBT3‐based film measurements were performed with varying MLC fields ranging from 7.5 mm to 27.5 mm. Seventeen regions of interests (ROIs) were identified for irradiation. These are regions located at the central area (denoted as reference field), at the left/right edge areas of reference open field, at an intermediate location between central and edge area. Single beam treatment plans were designed by using the MultiPlan and was delivered using the Blue Phantom. Gafchromic films were irradiated at 1.5 cm depth in the Blue Phantom and analyzed using the Film Pro software. Variation of maximum dose, penumbra of MLC‐defined fields, and symmetry/flatness were calculated as a function of locations of MLC fields.Results:The InCise™ MLC System showed relatively consistent dose distribution and penumbra size with varying locations of MLC fields. The measured maximum dose varied within 5 % at different locations compared to that at the central location and agreed with the calculated data well within 2%. The measured penumbrae were in the range of 2.9 mm and 3.7 mm and were relatively consistent regardless of locations. However, dose profiles in the out‐of‐field and in‐field regions varied with locations and field sizes. Strong variation was seen for all fields located at 55 mm away from the central field. The MLC leakage map showed that the leakage is dependent on position.Conclusion:The size of penumbra and normalized maximum dose for MLC‐defined fields were consistent in different regions of MLC. However, dose profiles in the out‐field region varied with locations and field sizes.

Purpose:Accuray has recently released a new collimator, the InCise™ Multileaf Collimator (MLC), for clinical use with the CyberKnife M6™ System. This work reports the results of measurements of output factors (OF) for fields shaped by the InCise™ Multileaf collimator, Iris collimator and fixed cones.Methods:The MLC consists of 41 pairs of 2.5 mm wide leaves projecting a minimum and maximum field size of 7.6 mm × 7.5 mm and 110 mm × 97.5 mm at 800 mm SAD. OF measurements were made using 6 different detectors: PTW stereotactic and electron diodes, PTW microDiamond detector, Sun Nuclear Edge diode, IBA SFD diode and PFD 3G photon diode. Measurements were made for 14 MLC field sizes and 12 cone/Iris field sizes (5mm to 60 mm diameter field sizes). All measurements were made in a Wellhofer water tank at the depth of maximum dose. Correction factors from Francescon et al (PMB 59, 2014, N11‐N17; 17, 2012 3741–3748) were used to correct the diode responses for the fixed cone and Iris measurements.Results:For the MLC fields no correction factors are currently available. For the cones and Iris collimator, the average of corrected OF for all diodes ranged from 0.651 ± 0.014 (5mm cone) to 0.994 ± 0.001 (50 mm cone); for the Iris collimator the corresponding values are 0.510 ±0.004 and 0.997 ±0.000 respectively. The OF for the MLC field sizes ranged from 0.806 ±0.009 for the 7.6 mm × 7.5 mm field to 1.023± 0.003 for the 110 mm × 97.5 mm field size.Conclusion:The present results are the first set of output factor data reported for the InCise™ MLC, Iris collimator and the fixed cones for the CyberKnife M6™ system. Good agreement is obtained among all the corrected data for the cones and Iris data.

Accuray recently released a new collimator, the InCise Multileaf Collimator (MLC), for clinical use with the CyberKnife M6 System. We performed an evaluation of the MLC by comparing plan qualities for single and multiple brain lesions generated using the InCise MLC, cone, and Iris collimators. We also investigated difference in delivery efficiency among the three collimators. Finally, we evaluated the capability of the MLC for treating multiple targets, larger and/or complicated irregular shaped targets with or without an organ-at-risk (OAR) adjacent to the lesions. Twenty-four patients with single or multiple brain mets treated previously in our clinic on a CyberKnife M6 System using cone/Iris collimators were selected for this study. Treatment plans were generated for all lesions using the InCise MLC. Monitor unit (MU), delivery time, target dose/coverage including dose-to-OAR, conformity index (CI), and dose to low-dose region were compared between MLC and clinical cone/Iris plans. Statistical analysis was performed using the nonparametric Wilcoxon-Mann-Whitney signed rank test. The planning accuracy of the MLC plans was validated using chamber and film measurements. Mean dose to target and target coverage was found to be comparable between the InCise MLC and cone/Iris plans. For minimum dose to target, the MLC plans had more cold spots than the cone/Iris plans. Median CI for MLC and cone/Iris plans was 1.37 and 1.28, respectively. CI for MLC plans was higher than that of cone/Iris plans regardless of size of target, number of targets, and complexity of the target shape. For irregularly shaped targets, there was a greater variation in CI and the difference was not reduced with increasing number of beams in the MLC plans. Statistically, the difference in CI between MLC and cone/Iris plans was significant for all cases. Median MU for the MLC and cone/Iris plans was 9,091 and 17,041, respectively. Median delivery time was 58 minutes for the cone/Iris plans and 36 minutes for the MLC plans, which resulted from the combination of reduction in MU and number of beams (median number = 166 in cone/Iris plans versus 58 in MLC plans). Statistically, the differences in MU, number of beams, and delivery time between MLC and cone/Iris plans were significant for all cases regardless of single versus multiple lesions, size of target, and irregularity of target. The QA results demonstrated good agreement between calculated and measured doses. Circular collimators produce higher quality plans than MLC; however the InCise MLC can produce clinically acceptable plans with approximately 40% decrease in delivery time, which may be useful in cases where delivery time is a limiting factor including multi-target cases.

Accuray recently released a new collimator, the InCise Multileaf Collimator (MLC), for clinical use with the CyberKnife M6 System. The new collimator will allow robotic radiosurgery to be delivered to larger targets, with substantially shorter treatment times compared to fixed circular or Iris collimators. This work reports the results an evaluation of the performance and reliability characteristics of this collimator performed from August 2014 to February 2015. Initial MLC tests included the following: acceptance and commissioning of the collimator, output factors for small fields using different types of detectors, alignment tests (to confirm pointing accuracy of the 6-DOF robot with the heavier MLC), MLC leaf positioning tests, end-to-end tests, beam data spot checks, leakage measurements, patient specific QA for SRS treatment plans, and multiple deliveries of simulated patient plans. The number and types of errors seen during the testing period were logged. The new MLC proved to be very reliable, with no mechanical malfunctions reported during 6 months of testing. Two forms of laser alignment test were done: a therapist performed a single alignment test daily, while physicists performed a more extensive, 4-component alignment test on 25 days. 92% of the therapists tests passed (47/51, all failures occurring on the first day of testing), while 72% of the physics tests passed all components (a single component failed 7/25 days, all offset values were less than 0.4 mm). MLC leaf positioning tests were performed to confirm leaf positioning accuracy to within 0.5 mm. Relative leaf positioning tests (‘picket fence’) passed 59/59 times, while absolute positioning tests (‘garden fence’) passed 10/32 times, due to manual errors in the setup of the test. All end-to-end test results were less than 0.3 mm. Spot check data for beam parameters agreed with acceptance results and the measured maximum and average leakage was 0.37% and 0.3% respectively. A total of 24 recoverable interruptions were logged during this extensive evaluation period. 10 patients originally planned for circular SRS collimators were re-planned for the MLC. Plan-specific QA measurements with chamber were all within 5% absolute dose agreement, and film measurements all passed 2%/2mm gamma evaluation for more than 95% of measurement points. In addition, 51 simulated patient plan deliveries were conducted over the evaluation period to confirm the reliability of the MLC during heavy use. The InCise MLC collimator passed all the tests satisfactorily, and demonstrated good reliability during the evaluation period. The MLC is approved for clinical treatments.

Purpose: Accuray recently released a new collimator, the InCise™ Multileaf Collimator (MLC), for clinical use with the CyberKnife M6™ System. This work reports the results of acceptance testing and commissioning measurements for this collimator. Methods: The MLC consists of 41 pairs of 2.5 mm wide leaves projecting a clinical maximum field size of 110 mm x 97.5 mm at 800 mm SAD. The leaves are made of tungsten, 90 mm in height and tilted by 0.5 degree. The manufacturer stated leaf positioning accuracy and reproducibility are 0.5 mm and 0.4 mm respectively at 800 mm SAD. The leaf over-travel is 100% with full interdigitation capability. Acceptance testing included, but are not limited to, the verification of the specifications of various parameters described above, leakage measurements and end-to-end tests. Dosimetric measurements included, but not limited to, measurements of output factors, open beam profiles, tissue-phantom ratios, beam flatness and symmetry, and patient specific QA. Results: All measurements were well within the manufacturer specifications. The values of output factors ranged from 0.804 (smallest field size of 7.6 mm x 7.5 mm) to 1.018 (largest field size of 110.0 mm x 97.5 mm). End-to-end test results for the various tracking modes are: Skull (0.27mm), fiducial (0.16mm), Xsight Spine (0.4mm), Xsight Lung (0.93 mm) and Synchrony (0.43mm). Measured maximum and average leakage was 0.37% and 0.3%, respectively. Patient-specific QA measurements with chamber were all within 5% absolute dose agreement, and film measurements all passed 2%/2mm gamma evaluation for more than 95% of measurement points. Conclusion: The presented results are the first set of data reported on the InCise™ MLC. The MLC proved to be very reliable and is currently in clinical use.

BACKGROUND:A significant number of patients have recurrent or persistent lung cancer despite complete resection or treatment with definitive chemoradiation. Stereotactic radiosurgery (SRS)/stereotactic body radiation therapy is emerging as an important modality for the treatment of early-stage lung neoplasm; SRS may also offer an alternative treatment option for patients with recurrent lung disease. We evaluated outcomes after treatment with SRS for recurrent lung neoplasm in a large series of patients. METHODS:Selected patients with limited recurrent, persistent, or progressive disease after one or more prior treatments for lung cancer were offered SRS. Thoracic surgeons evaluated all patients, placed fiducials when needed, and planned treatment in close collaboration with radiation oncologists and medical physicists. In our early experience, a single fraction of 20 Gy radiation was prescribed and was subsequently increased to 45 to 60 Gy in three to five fractions. The primary endpoint evaluated was overall survival. RESULTS:We treated 100 patients with recurrent lung cancer (median age 72 years) with SRS. The postprocedure 30-day mortality rate was 0%; median follow-up was 51 months (range, 5 to 123). The median overall survival for the entire group was 23 months (95% confidence interval: 19 to 41). The probability of 2-year and 5-year overall survival was 49% (95% confidence interval: 40% to 60%) and 31% (95% confidence interval: 23% to 43%), respectively. CONCLUSIONS:Our experience indicates that SRS is safe, and offers an alternative modality for selected patients with recurrent oligometastatic or persistent lung cancer. Thoracic surgeons should actively participate in SRS and continue to evaluate the efficacy of this treatment strategy.

BACKGROUND:Postoperative stereotactic radiosurgery for brain metastases potentially offers similar local control rates and fewer long-term neurocognitive sequelae compared to whole brain radiation therapy, although patients remain at risk for distant brain failure (DBF). OBJECTIVE:To describe clinical outcomes of adjuvant stereotactic radiosurgery for large brain metastases and identify predictors of intracranial failure and their implications on optimal patient selection criteria. METHODS:We performed a retrospective review on 100 large (>3 cm) brain metastases in 99 patients managed by resection followed by postoperative stereotactic radiosurgery to a median dose of 22 Gy (range, 10-28) in 1 to 5 fractions (median, 3). Primary histology was nonsmall cell lung in 40%, breast cancer in 18%, and melanoma in 17%. Forty (40%) patients had uncontrolled systemic disease. RESULTS:With a median follow-up of 12.2 months (range, 0.6-87.4), the 1-year Kaplan-Meier local control was 72%, DBF 64%, and overall survival 55%. Nine patients (9%) developed evidence of radiation injury, and 6 (6%) developed leptomeningeal disease. Uncontrolled systemic disease (P=.03), melanoma histology (P=.04), and increasing number of brain metastases (P<.001) were significant predictors of DBF on Cox multivariate analysis. Patients with <4 metastases, controlled systemic disease, and nonmelanoma primary (n=47) had a 1-year DBF of 48.6% vs 80.1% for all others (P=.01). CONCLUSION:Postoperative stereotactic radiosurgery to the resection cavity safely and effectively augments local control of large brain metastases. Patients with <4 metastases and controlled systemic disease have significantly lower rates of DBF and are ideal treatment candidates.

PURPOSE With the sequential optimization algorithm in MultiPlan system, clinical objectives (homogeneity, PTV coverage, conformity, normal tissue protection) can be optimized in sequence. However, the prescription isodose line (RxIDL) varies widely among institutions, which can influence the optimized dose distribution. The aim of this study is to investigate the impact of different prescription isodose lines on plan quality for the treatment of brain metastases using CyberKnife Multiplan system. METHODS Ten patients with multiple metastases were selected for this study. Four plans were generated for each patient such that 100% of the target volume receives the prescribed dose of 18 Gy, which was 50%, 60%, 70%, and 80% prescription Isodose line, separately. The prescription isodose was calculated as the ratio of the prescription dose and the maximum dose in target volume. The dosimetric parameters, including PTV coverage, conformity index (CI), gradient index (GI) and the volume covered by 12 Gy (V12Gy) were analyzed. The plan Monitor Units (MU) and treatment time were also compared. RESULTS All plans can provide the same target coverage (100%) and similar conformity index (1.26, 1.30, 1.32, and 1.29 on average for 80%, 70%, 60%, and 50% RxIDL plans, separately); there was no difference in critical structure dose. The 50% RxIDL plans have much lower GI (4.21±1.79 for 50% and 5.56±2.92 for 80% RxIDL plans) and V12Gy (13.36±10.31cc for 50% and 15.87±11.85cc for 80% RxIDL plans). The variation in estimated treatment delivery time was insignificant. CONCLUSION The dose falloff is much faster for the lower RxIDL plans in terms of GI and V12Gy. For 50% RxIDL plans, the average V12Gy decreases by 16% compared to 80% RxIDL plans, which indicates that the normal tissue can be better protected by decreasing the prescription Isodose line.

Purpose/Objective(s)For lung robotic radiosurgery, the Monte Carlo (MC) dose calculation is desired due to its advantage in accounting for the heterogeneous effect. It has been reported that MC calculations are sensitive to the mass density change. For example, the coverage of a plan optimized for lung density of 0.2 g/cc may drop more than 20% if the lung density is changed to 0.1 g/cc. Since the MC algorithm uses a CT curve that extrapolates data not directly measured for the low density range (e.g. 0∼ 0.2g/cc), and the corresponding electron density is always set to zero, the discrepancy may affect the accuracy of the MC dose calculation. The purpose of this study is to evaluate the performance of MC dose calculation for low density materials.Materials/MethodsA lung phantom was built with four different lung insert materials, of which the mass densities were 0.05 g/cc, 0.1 g/cc, 0.15 g/cc, and 0.27g/cc, respectively. TLDs and ion chamber were placed near the chest wall inside a simulated tumor which has a density of 1.0 g/cc. The volume of PTV was approximately 30cc. For each insert, a MC optimized plan was generated using the robotic radiosurgery treatment planning system, where the prescribed isodose line was 60-75% that covered at least 95% of PTV. The uncertainty was set 0.5% for all the MC calculations. The plans were delivered with a newly commissioned robotic radiosurgery system, and the measured doses were compared with the MC calculation.ResultsFor all measurements, high agreement was found in the center of the tumor (i.e. within 80% isodose line), where the largest difference between measured dose and MC calculated dose was only 1.8%. At the peripheral region (i.e. near the prescribed isodose line), the discrepancy became larger and the measured doses were off by 2.1%, 1.6%, 2.6% and 6.8%, for lung density of 0.27g/cc, 0.15g/cc, 0.1g/cc and 0.05g/cc, respectively.ConclusionsOur study shows that for typical patient lung density, ranging from 0.1g/cc to 0.35g/cc, the robotic radiosurgery MC-calculated dose is in good agreement with the measured dose. Purpose/Objective(s)For lung robotic radiosurgery, the Monte Carlo (MC) dose calculation is desired due to its advantage in accounting for the heterogeneous effect. It has been reported that MC calculations are sensitive to the mass density change. For example, the coverage of a plan optimized for lung density of 0.2 g/cc may drop more than 20% if the lung density is changed to 0.1 g/cc. Since the MC algorithm uses a CT curve that extrapolates data not directly measured for the low density range (e.g. 0∼ 0.2g/cc), and the corresponding electron density is always set to zero, the discrepancy may affect the accuracy of the MC dose calculation. The purpose of this study is to evaluate the performance of MC dose calculation for low density materials. For lung robotic radiosurgery, the Monte Carlo (MC) dose calculation is desired due to its advantage in accounting for the heterogeneous effect. It has been reported that MC calculations are sensitive to the mass density change. For example, the coverage of a plan optimized for lung density of 0.2 g/cc may drop more than 20% if the lung density is changed to 0.1 g/cc. Since the MC algorithm uses a CT curve that extrapolates data not directly measured for the low density range (e.g. 0∼ 0.2g/cc), and the corresponding electron density is always set to zero, the discrepancy may affect the accuracy of the MC dose calculation. The purpose of this study is to evaluate the performance of MC dose calculation for low density materials. Materials/MethodsA lung phantom was built with four different lung insert materials, of which the mass densities were 0.05 g/cc, 0.1 g/cc, 0.15 g/cc, and 0.27g/cc, respectively. TLDs and ion chamber were placed near the chest wall inside a simulated tumor which has a density of 1.0 g/cc. The volume of PTV was approximately 30cc. For each insert, a MC optimized plan was generated using the robotic radiosurgery treatment planning system, where the prescribed isodose line was 60-75% that covered at least 95% of PTV. The uncertainty was set 0.5% for all the MC calculations. The plans were delivered with a newly commissioned robotic radiosurgery system, and the measured doses were compared with the MC calculation. A lung phantom was built with four different lung insert materials, of which the mass densities were 0.05 g/cc, 0.1 g/cc, 0.15 g/cc, and 0.27g/cc, respectively. TLDs and ion chamber were placed near the chest wall inside a simulated tumor which has a density of 1.0 g/cc. The volume of PTV was approximately 30cc. For each insert, a MC optimized plan was generated using the robotic radiosurgery treatment planning system, where the prescribed isodose line was 60-75% that covered at least 95% of PTV. The uncertainty was set 0.5% for all the MC calculations. The plans were delivered with a newly commissioned robotic radiosurgery system, and the measured doses were compared with the MC calculation. ResultsFor all measurements, high agreement was found in the center of the tumor (i.e. within 80% isodose line), where the largest difference between measured dose and MC calculated dose was only 1.8%. At the peripheral region (i.e. near the prescribed isodose line), the discrepancy became larger and the measured doses were off by 2.1%, 1.6%, 2.6% and 6.8%, for lung density of 0.27g/cc, 0.15g/cc, 0.1g/cc and 0.05g/cc, respectively. For all measurements, high agreement was found in the center of the tumor (i.e. within 80% isodose line), where the largest difference between measured dose and MC calculated dose was only 1.8%. At the peripheral region (i.e. near the prescribed isodose line), the discrepancy became larger and the measured doses were off by 2.1%, 1.6%, 2.6% and 6.8%, for lung density of 0.27g/cc, 0.15g/cc, 0.1g/cc and 0.05g/cc, respectively. ConclusionsOur study shows that for typical patient lung density, ranging from 0.1g/cc to 0.35g/cc, the robotic radiosurgery MC-calculated dose is in good agreement with the measured dose. Our study shows that for typical patient lung density, ranging from 0.1g/cc to 0.35g/cc, the robotic radiosurgery MC-calculated dose is in good agreement with the measured dose.

Objective To evaluate tumor control, hearing, tinnitus, and balance outcomes of patients treated with CyberKnife (CK) radiosurgery for vestibular schwannoma (VS). Study Design Retrospective series review. Setting Tertiary referral center. Patients All patients treated with CK radiosurgery for vestibular schwannoma by a multidisciplinary radiosurgical team from August 2005 to November 2011. The median age was 59 years, and mean follow-up was 40 months. Seventy-three patients were treated (63 primary radiosurgery and 10 postsurgical). Interventions CK radiosurgery, serial MRI imaging, comprehensive audiometry, Tinnitus Handicap Inventory (THI) scores, and Activities-Specific Balance Confidence Scale (ABC). Main Outcome Measures Tumor control defined as 2 mm linear growth or lower or less than 20% increase in tumor volume (TV), measured in cubic centimeter, after a minimum of 12 months of monitoring, audiogram profiles, THI, and ABC surveys. Results Of those treated with CK as primary modality, 83% had 0- to 2-mm growth (tumor control or stable) and 17% grew greater than 2 mm. Of the tumors that were stable, 29% shrank 2 mm or greater. Volumetric analysis found that 74% of tumors had less than 20% TV growth, whereas 26% exhibited 20% or greater increase in TV. Of those deemed stable, 65% shrank 20% or greater TV; 95% of patients did not need additional surgical intervention, 3 required salvage surgery and 1 underwent additional radiosurgery. The majority of patients started with Class D hearing, but of those with Class A or B hearing before treatment, 53.5% maintained serviceable hearing at 3 years of follow-up. The pretreatment and posttreatment median THI Grades were both 1. The pretreatment and posttreatment ABC scores were unchanged at 81%. Conclusion The LINAC-based CK (18 Gy over 3 fractions at 80% isodose line) provides tumor control rates comparable to other forms of radiosurgery. Analysis for tumor growth was positive for 17% using maximum linear diameters and 26% with a volumetric workstation. This discrepancy is consistent with previous reports where volumetric models were found to be more sensitive in establishing growth. Serviceable hearing was comparable to previous SRS and SRT reports with an overall hearing preservation of 53.5%. This number was 77% in those with pre-Class A hearing. SRS did not affect pretreatment tinnitus or vestibular function.