Impact of different treatment techniques for pediatric Ewing sarcoma of the chest wall: IMRT, 3DCPT, and IMPT with/without beam aperture

Abstract Purpose To evaluate the dosimetric differences between photon intensity‐modulated radiation therapy (IMRT) plans, 3D conformal proton therapy (3DCPT), and intensity‐modulated proton therapy (IMPT) plans and to investigate the dosimetric impact of different beam spot size and beam apertures in IMPT for pediatric Ewing sarcoma of the chest wall. Methods and Materials Six proton pediatric patients with Ewing sarcoma in the upper, middle, and lower thoracic spine regions as well as upper lumbar spine region were treated with 3DCPT and retrospectively planned with photon IMRT and IMPT nozzles of different beam spot sizes with/without beam apertures. The plan dose distributions were compared both on target conformity and homogeneity, and on organs‐at‐risk (OARs) sparing using QUANTEC metrics of the lung, heart, liver, and kidney. The total integral doses of healthy tissue of all plans were also evaluated. Results Target conformity and homogeneity indices are generally better for the IMPT plans with beam aperture. Doses to the lung, heart, and liver for all patients are substantially lower with the 3DPT and IMPT plans than those of IMRT plans. In the IMPT plans with large spot without beam aperture, some OAR doses are higher than those of 3DCPT plans. The integral dose of each photon IMRT plan ranged from 2 to 4.3 times of proton plans. Conclusion Compared to IMRT, proton therapy delivers significant lower dose to almost all OARs and much lower healthy tissue integral dose. Compared to 3DCPT, IMPT with small beam spot size or using beam aperture has better dose conformity to the target.

about one third of all bone tumors in children. 2 Ewing sarcoma of the axial skeleton, for example base of skull, chest wall, and pelvis, is frequently treated with radiotherapy that serves as preoperative, definitive, or adjuvant therapy. [3][4][5][6][7] Due to the proximity of critical organs, radiotherapy is commonly associated with side effects and complications. Ewing sarcoma of the chest wall presents management challenges in the pediatric population affected, the need for aggressive adriamycin-based chemotherapy, and the presence of critical organs adjacent to the tumor. Treatment with conventional photon-based radiotherapy is associated with a 26% rate of Grade 3 + toxicity. 8 Proton therapybecause of the physical properties of protonshas the advantage of sparing healthy tissues and organs-at-risk (OARs) and is being used in the management of Ewing sarcoma. 9,10 Within the last couple of decades, proton therapy availability has increased substantially while the therapy itself has evolved from 3D conformal proton therapy (3DCPT) 11 using double scattering technique or uniform scanning technique to pencil beam scanning (PBS) using intensity-modulated proton therapy (IMPT) technique. 12 In this study, dosimetric comparisons were made between photon intensity-modulated radiotherapy (IMRT), 3DCPT, and IMPT plans for six pediatric patients with posterior chest wall Ewing sarcomas. Total integrated dose was also calculated for all the plans as a potential indication of secondary cancer possibilities.
Within the IMPT plans, proton techniques with and without beam apertures were employed with large and small beam spot sizes to evaluate the dosimetric impact of beam spot size and aperture within the IMPT cohort. This study was a clinical dosimetric study that compared the treatment techniques across widely used radiation modalities and technology representations to provide a realistic assessment of Ewing sarcoma treatment options.

2.A | Patient information
Six pediatric Ewing sarcoma patients treated with 3DCPT were selected for this retrospective dosimetric study comparing IMRT, 3DCPT, and IMPT with different beam spot size and with/without beam aperture. Five tumors were unresectable and one was treated postoperatively. These patients were grouped as upper thoracic spine (one patient), middle/lower thoracic spine (three patients), and upper lumbar spine (two patients). Detailed patient information is listed in Table 1. One of the patients who has target volumes at the region of T5 was presented in Fig. 1.

2.B | Treatment planning
In this study, clinical target volume 1 (CTV1) received 45 Gy (RBE) in 1.8 Gy (RBE) per fraction for 25 fractions and CTV2 received a sequential boost of 5.4Gy (RBE) in 1.8 Gy (RBE) per fraction for three fractions resulting in a total dose of 50.4 Gy (RBE). The primary criterion of the comparative plan development was equivalent target coverage and the secondary goal was to reduce dose to OAR, including the lung, heart, kidneys, and liver. Integral dose was also tracked to assess second malignancy risk. Standard institutional guidelines on target design, plan optimization, and critical organ tolerance were applied.
The photon IMRT plans were generated using Pinnacle 8.0 m with a step-and-shoot technique. The double scattering 3DCPT plans were generated using Eclipse Proton planning system version 11. All the IMPT plans were generated using RayStation 8A.
For 3DCPT planning, the number of beams used for each prescription depended on the target geometry. In general, two or three beams were used. When there were patch fields, the total number of fields increased. For each treatment field, beam distal and proximal margins were customized to take into consideration of the range uncertainties based on CTV. The beam aperture margins were based on planning target volume (PTV) and the beam ranges. The treatment nozzle with beam aperture was extended to about 4 cm from patient or table surface to minimize beam penumbra. Beam compensator smearing margin was customized based on setup uncertainties and potential target motion. Distal blocking was used to reduce dose to the OARs for some of the treatment fields. These double scattering plans were the original clinical plans; the CTV1 coverage was 45 Gy (RBE) covering 99% of volume; and the CTV2 coverage was 5.4 Gy (RBE) covering 99% of volume or 95% of the volume given that 99% was not achievable due to spinal cord dose constraints.
For IMRT planning, five to seven coplanar beams were used for initial target volume and boost target volume for each of the six patients. The total number of segments of each plan was 10 segments multiplying the number of beams. PTV was used for target coverage optimization. The CTV to PTV margins were 5 mm, which were based on interfraction setup uncertainties from x-ray-based image guidance, residual interfraction target motion uncertainties, and intrafraction target motion uncertainties. The planning goals were prescription doses covering 95% of the PTV volumes and 99% of the CTV volumes for each of the targets. The OAR dose-volume objectives of the plan optimization were either directly extracted or derived from the QUANTEC metrics of the lung, heart, liver and kidney. When these initial objectives resulted in significant degradation

2.C | Proton machines nozzles and apertures
For the IBA proton therapy systems, there are two types of treatment nozzles designed for IMPT treatment deliveries. The universal nozzle has limited beam vacuum space inside the nozzle that leads to a large beam spot size (in air) at the isocenter (1 sigma from 5 mm to 12 mm for energy from 100 MeV to 227 MeV); the dedicated nozzle has beam vacuum space that extends to the end of the nozzle. 18 The beam spot size (in air) at the isocenter is smaller than CT axial, sagittal, and coronal views of a patient with chest wall Ewing sarcoma at the region of T5. The green-colored contour is CTV1 and the blue-colored contour is CTV2.
those of the universal nozzle (1 sigma from 3 mm to 7 mm for energy from 70 MeV to 227 MeV). Smaller beam spot size leads to smaller penumbra at the edge of treatment field, therefore, proton plans with dedicated nozzle can potentially reduce dose to the OARs surrounding the target. 19 For treating the shallower portion of the targets, even though the proton scatter inside the patient is not dominant, the penumbra increases due to a lower proton energy compared to those of deeper targets. Thus, to sharpen the beam penumbra to reduce dose to the adjacent OARs, beam aperture was proposed for IMPT treatment. [20][21][22][23] In this study, 6-cm thick brass aperture was used for the nozzle with aperture IMPT plans. For each of these fields, aperture opening was custom fit to that beam-eyeview PTV with 0.7 cm treatment lateral margin.
where TV is the target volume, TV100 is the target volume covered by the prescription dose, and V100 is the total patient volume covered by the prescription dose. Note that the defined conformity index value decreases as the treatment plan becomes more conformal. The lowest conformity index is 1, which represents a plan with perfect conformity.
where, D5% is the minimum dose received by 5% volume of the target, D95% is the minimum dose received by 95% volume of the target, and Dp is the prescription dose.

3.B | OAR Doses
The dose-volume results of all the plans were presented in Table 3.  Liver dose in all of the plans was below the threshold value of 31 Gy. However, the proton plans' dose to the liver was substantially lower than those of IMRT plans. For IMRT plans, mean liver dose ranged from 11.5 Gy to 20.9 Gy. However, mean liver dose of the proton plans had a minimum value of 0.1 Gy and a maximum value of 4.6 Gy.

| DISCUSSION
Radiation therapy is an essential part of management of Ewing sarcoma in the chest wall, either as a definitive or an adjuvant therapy.
In the early years of radiotherapy, three-dimensional conformal photon therapy (3DCRT) 3-7 was the main radiation treatment technique.
But in the last two decades, IMRT has largely replaced 3DCRT as the main radiation therapy planning and delivery technique. 28,29 Due to the proximity of the lung and heart to the chest wall tumors as well as the physical properties of high-energy photon beams, multiple beams have to be used and each beam deposits radiation dose T A B L E 3 Lung, heart, and kidneys dose volume parameters of IMRT, 3DCPT, and IMPT plans. 3DCPT, three-dimensional conformal proton therapy; IMPT, intensity-modulated proton therapy; UN, universal nozzle; DN, dedicated nozzle; UN_A, universal nozzle with aperture; DN_A, dedicated nozzle with aperture.
in normal tissues before and after the tumors. Toxicities related to radiation therapy include cardiotoxicities, constrictive pericarditis, scoliosis, and pneumonitis. In our prior institutional retrospective study, we showed that even though the modern systemic therapy improved patient cause-specific survival rates substantially, 26% of the patients experienced Grade 3 + toxicity. 8,30 Furthermore, secondary malignancy is also a concern of radiotherapy, especially for the pediatric patient population. Several early studies 31