A Tran

The development and verification of a highly accurate collision prediction model for automated noncoplanar plan delivery

PURPOSE Significant dosimetric benefits had been previously demonstrated in highly noncoplanar treatment plans. In this study, the authors developed and verified an individualized collision model for the purpose of delivering highly noncoplanar radiotherapy and tested the feasibility of total delivery automation with Varian TrueBeam developer mode. METHODS A hand-held 3D scanner was used to capture the surfaces of an anthropomorphic phantom and a human subject, which were positioned with a computer-aided design model of a TrueBeam machine to create a detailed virtual geometrical collision model. The collision model included gantry, collimator, and couch motion degrees of freedom. The accuracy of the 3D scanner was validated by scanning a rigid cubical phantom with known dimensions. The collision model was then validated by generating 300 linear accelerator orientations corresponding to 300 gantry-to-couch and gantry-to-phantom distances, and comparing the corresponding distance measurements to their corresponding models. The linear accelerator orientations reflected uniformly sampled noncoplanar beam angles to the head, lung, and prostate. The distance discrepancies between measurements on the physical and virtual systems were used to estimate treatment-site-specific safety buffer distances with 0.1%, 0.01%, and 0.001% probability of collision between the gantry and couch or phantom. Plans containing 20 noncoplanar beams to the brain, lung, and prostate optimized via an in-house noncoplanar radiotherapy platform were converted into XML script for automated delivery and the entire delivery was recorded and timed to demonstrate the feasibility of automated delivery. RESULTS The 3D scanner measured the dimension of the 14 cm cubic phantom within 0.5 mm. The maximal absolute discrepancy between machine and model measurements for gantry-to-couch and gantry-to-phantom was 0.95 and 2.97 cm, respectively. The reduced accuracy of gantry-to-phantom measurements was attributed to phantom setup errors due to the slightly deformable and flexible phantom extremities. The estimated site-specific safety buffer distance with 0.001% probability of collision for (gantry-to-couch, gantry-to-phantom) was (1.23 cm, 3.35 cm), (1.01 cm, 3.99 cm), and (2.19 cm, 5.73 cm) for treatment to the head, lung, and prostate, respectively. Automated delivery to all three treatment sites was completed in 15 min and collision free using a digital Linac. CONCLUSIONS An individualized collision prediction model for the purpose of noncoplanar beam delivery was developed and verified. With the model, the study has demonstrated the feasibility of predicting deliverable beams for an individual patient and then guiding fully automated noncoplanar treatment delivery. This work motivates development of clinical workflows and quality assurance procedures to allow more extensive use and automation of noncoplanar beam geometries.

Viability of Noncoplanar VMAT for liver SBRT compared with coplanar VMAT and beam orientation optimized 4π IMRT

Purpose The 4π static noncoplanar radiation therapy delivery technique has demonstrated better normal tissue sparing and dose conformity than the clinically used volumetric modulated arc therapy (VMAT). It is unclear whether this is a fundamental limitation of VMAT delivery or the coplanar nature of its typical clinical plans. The dosimetry and the limits of normal tissue toxicity constrained dose escalation of coplanar VMAT, noncoplanar VMAT and 4π radiation therapy are quantified in this study. Methods and materials Clinical stereotactic body radiation therapy plans for 20 liver patients receiving 30 to 60 Gy using coplanar VMAT (cVMAT) were replanned using 3 to 4 partial noncoplanar arcs (nVMAT) and 4π with 20 intensity modulated noncoplanar fields. The conformity number, homogeneity index, 50% dose spillage volume, normal liver volume receiving >15 Gy, dose to organs at risk (OARs), and tumor control probability were compared for all 3 treatment plans. The maximum tolerable dose yielding a normal liver normal tissue control probability <1%, 5%, and 10% was calculated with the Lyman-Kutcher-Burman model for each plan as well as the resulting survival fractions at 1, 2, 3, and 4 years. Results Compared with cVMAT, the nVMAT and 4π plans reduced liver volume receiving >15 Gy by an average of 5 cm3 and 80 cm3, respectively. 4π reduced the 50% dose spillage volume by ∼23% compared with both VMAT plans, and either significantly decreased or maintained OAR doses. The 4π maximum tolerable doses and survival fractions were significantly higher than both cVMAT and nVMAT (P < .05) for all normal liver normal tissue control probability limits used in this study. Conclusions The 4π technique provides significantly better OAR sparing than both cVMAT and nVMAT and enables more clinically relevant dose escalation for tumor local control. Therefore, despite the current accessibility of nVMAT, it is not a viable alternative to 4π for liver SBRT.

SU‐F‐T‐186: A Treatment Planning Study of Normal Tissue Sparing with Robustness Optimized IMPT, 4Pi IMRT, and VMAT for Head and Neck Cases

PURPOSE We performed a retrospective dosimetric comparison study between the robustness optimized Intensity Modulated Proton Therapy (RO-IMPT), volumetric-modulated arc therapy (VMAT), and the non-coplanar 4? intensity modulated radiation therapy (IMRT). These methods represent the most advanced radiation treatment methods clinically available. We compare their dosimetric performance for head and neck cancer treatments with special focus on the OAR sparing near the tumor volumes. METHODS A total of 11 head and neck cases, which include 10 recurrent cases and one bilateral case, were selected for the study. Different dose levels were prescribed to tumor target depending on disease and location. Three treatment plans were created on commercial TPS systems for a novel noncoplanar 4π method (20 beams), VMAT, and RO-IMPT technique (maximum 4 fields). The maximum patient positioning error was set to 3 mm and the maximum proton range uncertainty was set to 3% for the robustness optimization. Line dose profiles were investigated for OARs close to tumor volumes. RESULTS All three techniques achieved 98% coverage of the CTV target and most photon plans had less than 110% of the hot spots. The RO-IMPT plans show superior tumor dose homogeneity than 4? and VMAT plans. Although RO-IMPT has greater R50 dose spillage to the surrounding normal tissue than 4π and VMAT, the RO-IMPT plans demonstrate better or comparable OAR (parotid, mandible, carotid, oral cavity, pharynx, and etc.) sparing for structures closely abutting tumor targets. CONCLUSION The RO-IMPTs ability of OAR sparing is benchmarked against the C-arm linac based non-coplanar 4π technique and the standard VMAT method. RO-IMPT consistently shows better or comparable OAR sparing even for tissue structures closely abutting treatment target volume. RO-IMPT further reduces treatment uncertainty associated with proton therapy and delivers robust treatment plans to both unilateral and bilateral head and neck cancer patients with desirable treatment time.

TH-EF-BRB-03: Significant Cord and Esophagus Dose Reduction by 4π Non-Coplanar Spine Stereotactic Body Radiation Therapy and Stereotactic Radiosurgery

PURPOSE To demonstrate significant organ-at-risk (OAR) sparing achievable with 4π non-coplanar radiotherapy on spine SBRT and SRS patients. METHODS Twenty-five stereotactic spine cases previously treated with VMAT (n = 23) or IMRT (n = 2) were included in this study. A computer-aided-design model of a Linac with a 3D-scanned human surface was utilized to determine the feasible beam space throughout the 4π steradian and beam specific source-to-target-distances (STD) required for collision avoidance. 4π radiotherapy plans integrating beam orientation and fluence map optimization were then created using a column-generation algorithm. Twenty optimal beams were selected for each case. To evaluate the tradeoff between dosimetric benefit and treatment complexity, 4π plans including only isocentrically deliverable beams were also created. Beam angles of all standard and isocentric 4π plans were imported into Eclipse to recalculate the dose using the same calculation engine as the clinical plans for unbiased comparison. OAR and PTV dose statistics for the clinical, standard-4π, and isocentric-4π plans were compared. RESULTS Comparing standard-4π to clinical plans, particularly significant average percent reduction in the [mean, maximum] dose of the cord and esophagus of [41%, 21.7%], and [38.7%, 36.4%] was observed, along with global decrease in all other OAR dose statistics. The average cord volume receiving more than 50% prescription dose was substantially decreased by 76%. In addition, improved PTV coverage was demonstrated with a maximum dose reduction of 0.93% and 1.66% increase in homogeneity index (D95/D5). All isocentric-4π plans achieved dosimetric performance equivalent to that of the standard-4π plans with higher delivery complexity. CONCLUSION 4π radiotherapy significantly improves stereotactic spine treatment dosimetry. With the substantial OAR dose sparing, PTV dose escalation is considerably safer. Isocentric-4π is sufficient to achieve the dosimetric gain. The successful implementation of 4π using an FDA approved planning system paves the way for a prospective clinical trial. Varian Medical Systems, NIH R43CA183390 and R01CA188300, NSF graduate research fellowship DGE-1144087.

TH-EF-BRB-01: BEST IN PHYSICS (THERAPY): Dosimetric Comparison of 4π and Clinical IMRT for Cortex-Sparing High-Grade Glioma Treatment

PURPOSE Thinning of the cerebral cortex has been observed in patients treated with fractionated partial brain radiation therapy and may contribute to cognitive decline following treatment. The extent of this thinning is dose-dependent, and was shown comparable to that of neurodegenerative diseases such as Alzheimers disease at one year post-therapy. This study investigates whether 4π radiotherapy can enable better sparing of the cortex and other critical structures when compared to conventional clinical IMRT plans. METHODS Clinical cortex-sparing IMRT plans for 15 high-grade glioma patients were included in this study. 4π radiotherapy plans were created for each patient with 20 intensity-modulated non-coplanar fields selected with a greedy column-generation optimization. All plans were normalized to deliver 100% of the prescribed dose to 95% of the planning target volume (PTV). The mean and maximum dose to the cerebral cortex and other organs at risk (OARs) were compared for the two plan types, as well as the conformity index (CI), homogeneity index (HI), and 50% dose spillage volume (R50). RESULTS The 4π plans significantly reduced the mean cortex dose by an average of 16% (range 6% to 27%) compared to the clinical plans. The mean dose to every other OAR compared was also reduced by 15% to 43%, with statistically significant reductions to the brainstem, chiasm, eyes, optic nerves, subcortical whit, and hippocampus. The average maximum doses were also reduced for 10/12 OARs. The R50 was significantly reduced with the 4π plans (>14%) and the homogeneity index was significantly improved. CONCLUSION 4π enables significant sparing of the cerebral cortex when treating high-grade gliomas with fractionated partial brain radiation therapy, potentially reducing the risk of harmful dose-dependent cortical thinning. NIH R43CA183390, NIH R01CA188300, Varian Medical Systems.

SU-D-BRB-01: A Comparison of Learning Methods for Knowledge Based Dose Prediction for Coplanar and Non-Coplanar Liver Radiotherapy

PURPOSE The predictive power of knowledge based planning (KBP) has considerable potential in the development of automated treatment planning. Here, we examine the predictive capabilities and accuracy of previously reported KBP methods, as well as an artificial neural networks (ANN) method. Furthermore, we compare the predictive accuracy of these methods on coplanar volumetric-modulated arc therapy (VMAT) and non-coplanar 4π radiotherapy. METHODS 30 liver SBRT patients previously treated using coplanar VMAT were selected for this study. The patients were re-planned using 4π radiotherapy, which involves 20 optimally selected non-coplanar IMRT fields. ANNs were used to incorporate enhanced geometric information including liver and PTV size, prescription dose, patient girth, and proximity to beams. The performance of ANN was compared to three methods from statistical voxel dose learning (SVDL), wherein the doses of voxels sharing the same distance to the PTV are approximated by either taking the median of the distribution, non-parametric fitting, or skew-normal fitting. These three methods were shown to be capable of predicting DVH, but only median approximation can predict 3D dose. Prediction methods were tested using leave-one-out cross-validation tests and evaluated using residual sum of squares (RSS) for DVH and 3D dose predictions. RESULTS DVH prediction using non-parametric fitting had the lowest average RSS with 0.1176(4π) and 0.1633(VMAT), compared to 0.4879(4π) and 1.8744(VMAT) RSS for ANN. 3D dose prediction with median approximation had lower RSS with 12.02(4π) and 29.22(VMAT), compared to 27.95(4π) and 130.9(VMAT) for ANN. CONCLUSION Paradoxically, although the ANNs included geometric features in addition to the distances to the PTV, it did not perform better in predicting DVH or 3D dose compared to simpler, faster methods based on the distances alone. The study further confirms that the prediction of 4π non-coplanar plans were more accurate than VMAT. NIH R43CA183390 and R01CA188300.

TU-CD-304-05: 4Ï€ Non-Coplanar Radiotherapy: From Mathematical Modeling to Clinical Implementation

Purpose: To develop and clinically implement 4π radiotherapy, an inverse optimization platform that maximally utilizes non-coplanar intensity modulated radiotherapy (IMRT) beams to significantly improve critical organ sparing. Methods: A 3D scanner was used to digitize the human and phantom subject surfaces, which were positioned in the computer assisted design (CAD) model of a TrueBeam machine to create a virtual geometrical model, based on which, the feasible beam space was calculated for different tumor locations. Beamlets were computed for all feasible beams using convolution/superposition. A column generation algorithm was employed to optimize patient specific beam orientations and fluence maps. Optimal routing through all selected beams were calculated by a level set method. The resultant plans were converted to XML files and delivered to phantoms in the TrueBeam developer mode. Finally, 4π plans were recomputed in Eclipse and manually delivered to recurrent GBM patients. Results: Compared to IMRT utilizing manually selected beams and volumetric modulated arc therapy plans, markedly improved dosimetry was observed using 4π for the brain, head and neck, liver, lung, and prostate patients. The improvements were due to significantly improved conformality and reduced high dose spillage to organs mediolateral to the PTV. The virtual geometrical model was experimentally validated. Safety margins with 99.9% confidence in collision avoidance were included to the model based model accuracy estimates determined via 300 physical machine to phantom distance measurements. Automated delivery in the developer mode was completed in 10 minutes and collision free. Manual 4 π treatment on the GBM cases resulted in significant brainstem sparing and took 35–45 minutes including multiple images, which showed submillimeter cranial intrafractional motion. Conclusion: The mathematical modeling utilized in 4π is accurate to create and guide highly complex non-coplanar IMRT treatments that consistently and significantly outperform human-operator-created plans. Deliverability of such plans is clinically demonstrated. This work is funded by Varian Medical Systems and the NSF Graduate Research Fellowship DGE-1144087.

SU-F-BRB-04: Comparison of Coplanar VMAT, Non-Coplanar VMAT, and 4π Treatment Plans

Purpose: The 4π non-coplanar radiotherapy delivery technique has demonstrated significantly better normal tissue sparing and dose conformality than the clinically used volumetric modulated arc therapy (VMAT). It is unclear whether this is a fundamental limitation of VMAT delivery or the coplanar nature of its typical clinical plans. The non-coplanar basis of 4π is incorporated into VMAT treatment planning to compare its effect on plan quality. Methods: Clinical stereotactic body radiation therapy plans for 9 liver patients treated with 30–60 Gy using coplanar VMAT (cVMAT) were re-planned using non-coplanar VMAT (nVMAT) with 3 arcs and 4 π with 20 intensity-modulated non-coplanar fields. All plans were optimized to deliver 100% of the prescribed dose to 95% of the planning target volume (PTV), and nVMAT and 4π plans were tailored to match the maximum and mean PTV dose from the clinical plan. The conformality index (CI), 50% dose spillage volume (R50), normal liver volume receiving >15 Gy (VL>15), and doses to organs at risk (OARs) were compared for all three treatment plans. Results: Compared to cVMAT, the nVMAT and 4π plans reduced VL>15 by an average of 30.6 cm3 and 96.3 cm3, respectively. The average CI was also reduced from 1.22 (cVMAT) to 1.17 (nVMAT) and 1.14 (4π), indicating higher conformality in the same order. Similarly, R50 was reduced from 3.87 (cVMAT) to 3.58 (nVMAT) and 2.74 (4π). With the exception of the mean right kidney dose, which increased by an average of only 0.6 Gy for nVMAT, the dose differences to OARs were not statistically significant between the two VMAT plans. 4π plans either significantly decreased or maintained OAR doses. Conclusion: While the manual selection of intuitive non-coplanar arcs does show some improvement over coplanar VMAT, the automated beam selection for 4π still results in superior plan quality. This project is supported in part by Varian Medical Systems and NIH R43 CA183390.

SU-E-T-765: Treatment Planning Comparison of SFUD Proton and 4Ï€ Radiotherapy for Prostate Cases

Purpose: Single-Field Uniform Dose (SFUD) proton scanning beams and non-coplanar 4π intensity-modulated radiation therapy (IMRT) represent the most advanced treatment methods based on heavy ion and X-rays, respectively. Here we compare their performance for prostate treatment. Methods: Five prostate patients were planned using 4π radiotherapy and SFUD to an initial dose of 54Gy to a planning target volume (PTV) that encompassed the prostate and seminal vesicles, then a boost prescription dose of 25.2Gy to the prostate for a total dose of 79.2 Gy. 4π plans were created by inversely selecting and optimizing 30 beams from 1162 candidate non-coplanar beams using a greedy column generation algorithm. The SFUD plans utilized two coplanar, parallel-opposing lateral scanning beams. The SFUD plan PTV was modified to account for range uncertainties while keeping an evaluation PTV identical to that of the X-ray plans for comparison. PTV doses, bladder and rectum dose volumes (V40, V45, V60, V70, V75.6, and V80), R50, and PTV homogeneity index (D95/D5) were evaluated. Results: Compared to SFUD, 4π resulted in 6.8% lower high dose spillage as indicated by R50. Bladder and rectum mean doses were 38.3% and 28.2% lower for SFUD, respectively. However, bladder and rectum volumes receiving >70Gy were 13.1% and 12% greater using proton SFUD. Due to the parallel-opposing beam arrangement, SFUD resulted in greater femoral head (87.8%) and penile bulb doses (43.7%). 4π PTV doses were slightly more homogeneous (HI 0.99 vs. 0.98) than the SFUD dose. Conclusion: Proton is physically advantageous to reduce the irradiated normal volume and mean doses to the rectum and bladder but it is also limited in the beam orientations and entrance dose, which resulted in greater doses to the femoral heads and penile bulb, and larger volumes of rectum and bladder exposed to high dose due to the required robust PTV definition. This project is supported by Varian Medical Systems.

SU-F-BRB-10: A Statistical Voxel Based Normal Organ Dose Prediction Model for Coplanar and Non-Coplanar Prostate Radiotherapy

Purpose: Knowledge learned from previous plans can be used to guide future treatment planning. Existing knowledge-based treatment planning methods study the correlation between organ geometry and dose volume histogram (DVH), which is a lossy representation of the complete dose distribution. A statistical voxel dose learning (SVDL) model was developed that includes the complete dose volume information. Its accuracy of predicting volumetric-modulated arc therapy (VMAT) and non-coplanar 4π radiotherapy was quantified. SVDL provided more isotropic dose gradients and may improve knowledge-based planning. Methods: 12 prostate SBRT patients originally treated using two full-arc VMAT techniques were re-planned with 4π using 20 intensity-modulated non-coplanar fields to a prescription dose of 40 Gy. The bladder and rectum voxels were binned based on their distances to the PTV. The dose distribution in each bin was resampled by convolving to a Gaussian kernel, resulting in 1000 data points in each bin that predicted the statistical dose information of a voxel with unknown dose in a new patient without triaging information that may be collectively important to a particular patient. We used this method to predict the DVHs, mean and max doses in a leave-one-out cross validation (LOOCV) test and compared its performance against lossy estimators including mean, median, mode, Poisson and Rayleigh of the voxelized dose distributions. Results: SVDL predicted the bladder and rectum doses more accurately than other estimators, giving mean percentile errors ranging from 13.35–19.46%, 4.81–19.47%, 22.49–28.69%, 23.35–30.5%, 21.05–53.93% for predicting mean, max dose, V20, V35, and V40 respectively, to OARs in both planning techniques. The prediction errors were generally lower for 4π than VMAT. Conclusion: By employing all dose volume information in the SVDL model, the OAR doses were more accurately predicted. 4π plans are better suited for knowledge-based planning than the VMAT plans that are strongly biased in its dose gradient orientation. This project is supported by Varian Medical Systems.