Potential reduction of lung dose via VMAT with jaw tracking in the treatment of single‐isocenter/two‐lesion lung SBRT

Abstract Purpose/objectives Due to higher radiosensitivity, non‐target normal tissue dose is a major concern in stereotactic body radiation therapy (SBRT) treatment. The aim of this report was to estimate the dosimetric impact, specifically the reduction of normal lung dose in the treatment of single‐isocenter/two‐lesion lung SBRT via volumetric modulated arc therapy with jaw tracking (JT‐VMAT). Materials/methods Twelve patients with two peripherally located early‐stage non‐small‐cell‐lung cancer (NSCLC) lung lesions underwent single‐isocenter highly conformal non‐coplanar JT‐VMAT SBRT treatment in our institution. The mean isocenter to tumors distance was 5.6 ± 1.9 (range 4.3–9.5) cm. The mean combined planning target volume (PTV) was 38.7 ± 22.7 (range 5.0–80.9) cc. A single isocenter was placed between the two lesions. Doses were 54 and 50 Gy in three and five fractions, respectively. Plans were optimized in Eclipse with AcurosXB algorithm utilizing jaw tracking options for the Truebeam with a 6 MV‐FFF beam and standard 120 leaf millennium multi‐leaf collimators. For comparison, the JT‐VMAT plans were retrospectively re‐computed utilizing identical beam geometry, objectives, and planning parameters, but without jaw tracking (no JT‐VMAT). Both plans were normalized to receive the same target coverage. The conformity and heterogeneity indices, intermediate‐dose spillage [D2cm, R50, Gradient Index (GI), Gradient Distance (GD)], organs at risks (OAR) doses including normal lung as well as modulation factor (MF) were compared for both plans. Results For similar target coverage, GI, R50, GD, as well as the normal lung V5, V10, V20, mean lung dose (MLD), and maximum dose received by 1000 cc of lungs were statistically significant. Normal lung doses were reduced by 8%–11% with JT‐VMAT. Normal lung dose increased as a function of tumor distance from isocenter. For the other OAR, up to 1%–16% reduction of non‐target doses were observed with JT‐VMAT. The MF and beam‐on time were similar for both plans, however, MF increased as a function of tumors distance, consequently, delivering higher dose to normal lungs. Conclusion Utilizing jaw tracking options during optimization for single‐isocenter/two‐lesion lung SBRT VMAT plans reduced doses to the normal lung and other OAR, reduced intermediate‐dose spillage and provided superior/similar target coverage. Application of jaw tracking did not affect delivery efficiency and provided excellent plan quality with similar MF and beam‐on time. Jaw tracking is recommended for future clinical SBRT plan optimization.

100 Gy while minimizing the dose to the adjacent organs at risk (OAR). [1][2][3] Several studies have shown that safely delivering a higher BED to the lung lesions improved therapeutic ratio and local control rates. [4][5][6][7][8][9][10] In addition, utilizing volumetric modulated arc therapy (VMAT) planning with a flattening filter free (FFF) beam in lung SBRT treatment reduced the total number of monitor units (MUs) 11,12 and the treatment time compared to intensity modulated radiotherapy, Tomotherapy, or CyberKnife. [13][14][15][16] Reduction in MUs provides faster treatment delivery that can improve patient comfort, decrease potential setup/motion related errors and promote efficient clinical workflow. Owing to those advantages, VMAT SBRT planning using single isocenter for multiple targets has been gaining popularity in clinics for treating multiple intracranial tumors 17,18 as well as extracranial oligometastases lesions. [19][20][21][22][23] Conversely, VMAT averages the dose delivery over more angles and produces slightly higher non-target low dose distribution compared to intensity modulated radiation therapy (IMRT). Generally, the treatment fields are designed with the jaw apparatus and tertiary multi-leaf collimators (MLCs) shaping the target volume. The jaw apparatus is fixed on the maximum field size of MLCs during treatment delivery, and thus leakage and transmission of radiation through the MLCs is present in the optimized IMRT/VMAT plan. This effect is noticeable while utilizing single-isocenter/multitarget VMAT plan. When the isocenter to tumor distance is large (on the order of 4-10 cm), the MLCs have to travel a longer distance to provide the target coverage to each lesion, potentially delivering higher non-trivial low-dose spillage to the non-target tissues such as normal lungs.
Due to the higher radiosensitivity, non-target normal tissue dose is one of the major concerns for SBRT treatments. 24 In this report we retrospectively evaluated 12 single-isocenter/twolesion early stage NSCLC patient's plans who underwent SBRT treatment in our clinic using JT-VMAT. For those patients, the nontarget low dose was minimized by using jaw tracking options for the Truebeam Linac with a 6 MV-FFF beam (in Eclipse treatment (ITV) were delineated on the 3D CT images with reference to the MIP images. Planning target volumes (PTV) were generated by adding non-uniform 5-10 mm margins to the ITV to accommodate the patient setup uncertainties based on tumor size, location, and synchronous tumor motion. The critical structures, such as bilateral lungs excluding the ITV (normal lung), spinal cord, ribs, heart, great vessels, esophagus, and skin were delineated on the 3D CT images.
The tumor characteristics for the single-isocenter/two-lesion lung SBRT patients are summarized in Table 1, including isocenter to tumors distance, normal lung volume, and tumor location. The combined PTV was defined as PTV1 plus PTV2. Both lesions were treated synchronously with a total dose of 54 Gy or 50 Gy in three and five fractions, respectively. Normal lung volume ranged from 1893 to 6543 cc, mean 3881 cc. The average value of isocenter to tumors distance was 5.6 cm (range 3.4 to 9.5 cm). Additionally, the jaw tracking (JT) option was chosen during plan optimization to further minimize the non-target dose. A dose of 54 or 50 Gy in three and five fractions was prescribed to the PTV of which D95% received at least 100% of the prescription. All hot spots were within each ITV (i.e., the center of each ITV was 20% hotter). All clinical treatment plans were calculated using the Eclipse TPS with Acuros XB (version 13.6.0, Varian Medical Systems, Palo Alto, CA) algorithm on the 3D CT images with heterogeneity corrections using a 2.0 × 2.0 × 2.0 mm 3 dose calculation grid-size. Dose to medium was reported. All clinical plans were inversely optimized using variation of gantry rotation speed, dose rate, and MLC positions. In addition to optimization ring structures, the generalized normal tissue objective (NTO) parameters were used to control the gradients for each target. Planning objectives were per RTOG 0915 guidelines. These patients were treated every other day per lung SBRT protocol.

2.B.2 | Quality assurance and treatment delivery
For each plan, a verification plan was generated in the Eclipse TPS using an Octavius phantom (PTW, Freiburg, Germany). Doses re-calculated on the phantom's 2D ionization chamber array were exported and compared to a measured dose distribution. Using the γ-evaluation method of VeriSoft (Version 6.3, PTW) the two distributions were compared using the standard clinical gamma passing rate criteria of 3%/3 mm maximum dose difference and distance-toagreement with 10% threshold as well as maximum point dose. The Octavius QA pass rates for the single-isocenter/two-lesion lung SBRT plan were 98.8 ± 2.5%, on average, for 3%/3 mm clinical gamma pass rate criteria and the maximum point dose measurement was 1.0 ± 0.7%, on average, suggesting that lung SBRT plans using JT can be accurately delivered. The beam-on time was estimated by using dose rates of 1400 MU/min for these plans. The dose-rate was confirmed by reviewing each VMAT arc for all patients under the MLC properties in Eclipse. Additionally, maximum dose rate of 1400 MU/min was visually observed during VMAT QA delivery at Truebeam for all single-isocenter/two-lesion lung SBRT plans.
Before delivering each JT-VMAT SBRT treatment, a daily quality assurance check on kilovoltage to megavoltage imaging isocenter coincidence was performed, including IsoCalc measurement for precise and accurate target localization. Our IsoCalc localization accuracy for Truebeam was <0.5 mm. All the quality assurance procedures were in compliance for SBRT treatment delivery.
The patients received daily cone beam CT per image-guidance procedures established in our clinic.

2.C | Plan evaluation
The dose volume histograms (DVHs) and isodose curves of JT-VMAT vs no JT-VMAT plans were compared. The Conformity index (CI), heterogeneity index (HI), gradient index (GI), gradient distance (GD), T A B L E 1 Characteristics of single-isocenter/two-lesion lung stereotactic body radiation therapy (SBRT) patients treated with volumetric modulated arc therapy with jaw tracking (JT-VMAT) plan included in this study.

Parameters
Mean ± SD (range or no. of patients)

3.A | Targets coverage
Both plans were normalized to receive the same target coverage  Table 2) compared to no JT-VMAT plan.

3.B | Dose to lungs
The absolute differences between single-isocenter JT-VMAT and no JT-VMAT SBRT plans for normal lung V20, V10, V5, MLD, and the maximum dose received by 1000 cc of lungs are listed in Table 3.
All patients had V20 < 10%-15% for JT-VMAT treatment plans per protocol. The absolute differences of V20, V10, and V5 were up to 2%, 3%, and 4% higher, respectively with no JT-VMAT plans. Doses to all lung parameters increase uniformly with no JT-VMAT plan T A B L E 2 Plan quality evaluation for single-isocenter/two-lesion lung stereotactic body radiation therapy (SBRT) volumetric modulated arc therapy with jaw tracking (JT-VMAT; clinical) and no JT-VMAT (replanned) plans for all 12 patients. Combined planning target volume (PTV) = PTV1 plus PTV2. CI = conformity index, total volume covered by the 100% isodose line divided by the volume of the combined PTV. HI, heterogeneity index = D10%/ D95%, where D10% is the dose to the hottest 10% of the combined PTV and D95% is the dose to the 95% of the combined PTV coverage. R50 (%) = ratio of 50% prescription isodose volume to the combined PTV. D2 cm (%) = maximum dose (in % of dose prescribed) 2 cm away from PTV in any direction. GI = R50%/R100%, R50% is the ratio of 50% prescription isodose volume to the combined PTV and R100% is the ratio of 100% prescription isodose volume to the combined PTV. GD (cm) = is the average distance from 100% prescription dose to 50% of the prescription dose. Statistically significant P-values are in bold, n.s. = not significant.

3.C | Dose to other OAR
A comparison of other OAR dosimetric parameters for single-isocenter/two-lesion JT-VMAT and no JT-VMAT plans for all 12 lung SBRT patients is presented in Table 4. Critical organs such as spinal cord

3.D | Modulation factor and beam-on time
The MF for no JT-VMAT vs JT-VMAT and the MF as a function of the isocenter to tumor distance is shown in Fig. 4. For the given lung SBRT plan, the total number of MUs did not change However, MF increases as a function of isocenter to tumor distance (see right panel in Fig. 4), suggesting that farther apart the tumors, the more MUs are required to deliver the target coverage and consequently more low-dose spillage to the non-target tissues.
F I G . 3. Scatter plot: Ratios of normal lungs V5, V10, V20, MLD, and maximum dose to 1000 cc of lungs calculated by volumetric modulated arc therapy with no jaw tracking (no JT-VMAT) and JT-VMAT plans as a function of isocenter to tumor distance. For the identical plan parameters and objectives, the no JT-VMAT plans gave higher V5, V10, V20, MLD and maximum dose to 1000 cc of lungs by 6%, 8%, 8%, 8%, and 11%, on average, respectively, compared to JT-VMAT plans.
T A B L E 4 Average values and ranges of absolute dose differences between volumetric modulated arc therapy with no jaw tracking (no JT-VMAT) vs JT-VMAT plans for the major dose distribution parameters of the other OAR for all 12 lung stereotactic body radiation therapy (SBRT) patients.    In summary, the potential benefit of applying jaw tracking approach in Truebeam (with 6MV-FFF beam) for single-isocenter/

CONF LICT OF I NTEREST
None.