Purpose: To investigate an alternative approach to VMAT optimization for hypofractionation lung treatment which increases average aperture opening and results in lower total Monitor Units (MU) without significantly sacrificing plan quality. Methods: Benchmark Volumetric Modulated Arc Therapy (bVMAT) plans were generated for 10 lung Stereotactic Body radiotherapy (SBRT) cases using Eclipse Version 11.0.42 (Varian Medical Systems) without a maximum MU constraint. Prescriptions ranged from 40 to 54Gy in 3 to 5 fractions. AAA dose calculation and PRO fluence based optimization was utilized. Two comparison VMAT plans were generated for each case, one that forced an initial “open” mlc aperture conformal to the tumor as a starting condition (oVMAT) with similar optimization parameters and arc geometries, and one that repeated the bVMAT optimization but added a maximum MU constraint (muVMAT). All plans used two arcs with lengths between 168 to 230 degrees. PTV D 95% and Dmean, lung V20 Gy, chest wall V30 Gy, average aperture opening and MU's were compared. Statistical significance was evaluated using Wilcoxon signed rank test. Results: Average PTV D(95), PTV mean and lung V20Gy over all plans was 99.2 ± 1.7%, 103.3 ± 0.6% and 7.8 ± 2.4% respectively. The average chest wall V30Gy was 61 ± 61 cc and ranged between 0 to 166 cc. There were no significant differences between the three techniques for the dosimetric quantities. MUs were reduced by 11 ±11% (p<0.01) and 25 ± 5% (p<0.01) and the average aperture size was increased by 13.7 ± 14% (p=0.02) and 35.8 ± 10% (p<0.01) with muVMAT and oVMAT, respectively, compared to bVMAT. Conclusion: oVMAT and muVMAT techniques were both able to increase average aperture size and reduce total MU compared to the benchmark VMAT plan, but the magnitude of the changes observed for oVMAT was larger.

Purpose: We have developed a method of tracking irregularly shaped implanted markers using KV projection images acquired in rotational mode and assess its potential for detecting intra‐fractional target motion. This is a feasibility study directed toward long‐range goals of acquiring such images during rotational treatment and using them for motion correction. Methods: KV projection images were acquired (Varian TrueBeam) during seven cone beam scans of two gastroesophageal and two pancreas cancer patients (IRB‐approved protocol). Each had at least one irregularly shaped radiopaque marker (Visicoil) implanted in or near the tumor. Specialized digitally reconstructed radiographs (DRRs) used for template based tracking were created from a breath‐hold planning CT at end expiration, in which the ray tracing was confined to a small volume of interest surrounding each marker. Sobel filter preprocessing of KV images served to enhance marker visibility and suppress background features. DRRs were matched with processed KV images both manually (ground truth) and automatically (normalized cross‐correlation with simplex minimization). Anthropomorphic phantom studies were also done to evaluate measurement uncertainty. Results: The mean (over patient scans) and standard deviation of the differences (Auto‐manual) were −0.04 ± 0.68 mm and 0.08 ± 0.89 mm in transverse and superior‐inferior (SI) directions respectively. The percentages of matches with difference exceeding 2 mm were 1.8% transverse and 5.0% SI. Intra‐observer consistency of manual registration was checked by repeating the manual registration for all 657 projections in one patient; the standard deviation of the difference was 0.4 mm. Phantom studies showed the measurement uncertainty of automatic registration to be approximately 0.15 mm. Conclusions: The proposed method can track arbitrary marker shapes using templates generated from a breath‐hold CT or alternatively, respiration‐correlated CT scan at one phase. Preliminary results indicate accuracy and robustness are adequate for clinical application but confirmation in larger numbers of patients is required. Research grant from Varian Medical Systems

Purpose: To investigate the sensitivity of trajectory log analysis for detecting small VMAT delivery errors with an automated program. Methods: For each treatment, Varian TrueBeam™ generates a set of trajectory log files that record accumulated MU, MLC positions, and gantry angle every 10 ms during delivery. An automated computer program analyzes discrepancies between planned and actual leaf positions. Any leaf that exceeds 0.5 mm discrepancy for >15% of the total beam‐on time is flagged and an alert sent to a physicist for further investigation. To validate the method, 6 different “induced‐error” VMAT plans were generated by modifying a lung VMAT plan with an intended error: gantry angle error of 0.5 and 1.0 deg, position error of one leaf of 0.5, 1.0, 1.5, and 2.0 mm. All plans were delivered and trajectory logs were collected. The dosimetric effect of the induced errors was evaluated by reconstructing dose distributions from the trajectory logs and comparing to the original error‐free plan. Results: The proposed method is sensitive and can detect gantry error down to 0.5 deg and MLC leaf error of 0.5 mm without any false positives (i.e. no error detected for other leaves). No dose difference was observed for gantry angle errors of 0.5–1.0 deg, while the maximum dose discrepancy from a MLC position error of 0.5 mm was 1% and increased to 3% for a leaf error of 2 mm, with a passing rate of 99.6% and 97.7% for gamma analysis (1%/0 mm), respectively. Conclusion: Analysis of trajectory log files is highly sensitive and provides an efficient and accurate method of detecting treatment delivery errors down to sub‐millimeters. The automated program can be run daily for treatment validation and can be used also as a pre‐treatment QA tool. Research grant from Varian Medical Systems.

Purpose: To investigate the feasibility of real‐time mis‐alignment correction in Rapid Arc treatment and design a corresponding tomosynthesis acquisition protocol. Method and Materials: A CT image set of an anthropomorphic pelvic phantom was used in the study. Simulated projection images were produced to resemble a simultaneous kV fluoro in Rapid Arc treatment. A modified Feldkamp algorithm was used to reconstruct the tomosynthesis images. Various combinations of imaging reconstruction parameters including scan angle, angular interval, and slice thickness (mm) were tested: 1) 60°, 6°, 2.4; 2) 60°, 6°, 0.8; 3) 60°, 3°, 2.4; and 4) 30°, 3°, 2.4. A predefined 5 mm displacement in all three orthogonal directions modeled patient motion during treatment. After each successive tomosynthesis acquisition, registrations were performed between current reconstructions and reference images. The phantom position was corrected accordingly by shifting the treatment couch. Residual errors and their root mean square (RMS) values were recorded for evaluation. Results: The residual errors (L‐R, A‐P and S‐I directions in mm) for the 4 schemes after the first tomosynthesis acquisition were (1.2, 2.5, −0.2), (−1.2, 1.1, 0.0), (1.1, 1.9, 0.0) and (−1.2, 3.1, 0.0), and the corresponding RMSs were 1.6, 0.9, 1.3 and 1.9 respectively. The RMSs after a full arc delivery were 0.7, 0.5, 0.5 and 0.7. All schemes tested accurately corrected displacement in the SI direction after first acquisition. Scheme 2 performed better than scheme 1 at the expense of more computation time. By doubling projection numbers in scheme 1, scheme 3 improved correction ability in the L‐R direction. With a smaller 30° scan angle, scheme 4 was acceptable and will be improved after several acquisitions. Conclusions: Tomosynthesis scans can be used for real‐time mis‐alignment correction in Arc therapy after 30° gantry rotation.

Purpose: Volumetric modulated arc therapy (VMAT) (Otto K, Medical Physics, 2008) is an emerging treatment paradigm which modulates MLC aperture and dose rate during gantry rotation. The purpose of this study is to implement and evaluate VMAT relative to the standard IMRT approach. Method and Materials: A gantry rotation up to 360° is modeled as 360 evenly divided beams. Beam apertures and dose rate are optimized with respect to gantry angles under a DVH constraints based objective. Differences between our VMAT implementation and previous VMAT are: using gradient search for dose rate optimization rather than random search, and sampling multiple MLC leaf positions within the allowed leaf speed constraints rather than a single one in each iteration. A planning study including 5 prostate patients with a prescription dose of 86.4Gy was performed to evaluate VMAT verse the standard 5‐field IMRT approach. VMAT treatment plans are normalized such that certain critical organ dose limits are met, and are comparable to IMRT plans. V95, D95 and mean dose of PTV are used to evaluate plans, while monitor unit (MU) and delivery time are used to assess delivery efficiency. Results: The PTV V95, D95 and mean dose in VMAT plans are 97.0±0.8%, 96.4±0.5%, and 101.7±0.4%, respectively, vs 97.1±0.8%, 97.5±1.0% and 103.0±0.7% in IMRT plans. VMAT and IMRT plans are indistinguishable measured by these dose indices. The advantage that VMAT presents is it reduces MU by 49.8±7.4%. Secondary scatter dose to patient is reduced accordingly. A typical prostate treatment is shortened from about 5 minutes for IMRT to 2.6±0.2 minutes for VMAT. The better delivery efficiency of VMAT is accomplished by having a larger time averaged beam aperture: 49.1±5.2cm2 vs 19.5±1.5 cm2 in IMRT. Conclusion: VMAT technique can reduce treatment time by up to 50% while maintaining comparable dosimetric quality to standard IMRT approach.