Impact of varying air cavity on planning dosimetry for rectum patients treated on a 1.5 T hybrid MR‐linac system

Abstract Purpose To investigate the dosimetric impact of magnetic (B) field on varying air cavities in rectum patients treated on the hybrid 1.5 T MR‐linac. Methods Artificial air cavities of varying diameters (0.0, 1.0, 1.5, 2.0, 2.5, 3.0, and 5.0 cm) were created for four rectum patients (two prone and two supine). A total of 56 plans using a 7 MV flattening filter‐free beam were generated with and without B‐field. Reference intensity‐modulated radiation therapy treatment plans without air cavity in the presence and absence of B‐field were generated to a total dose of 45/50 Gy. The reference plans were copied and recalculated for the varying air cavities. D95(PTV45–PTV50), D95(PTV50–aircavity), V50(PTV50–aircavity), Dmax(PTV50–aircavity), and V110%(PTV50–aircavity) were extracted for each patient. Annulus rings of 1‐mm‐diameter step size were generated for one of the air cavity plans (3.0 cm) for all four patients to determine Dmax (%) and V110% (cc) within each annulus. Results In the presence of B‐field, hot spots at the cavity interface start to become visible at ~1 cm air cavity in both supine and prone positioning due to electron return effect (ERE). In the presence of B‐field Dmax and V110% varied from 5523 ± 49 cGy and 0.09 ± 0.16 cc for 0 cm air cavity size to 6050 ± 109 cGy and 11.6 ± 6.7 cc for 5 cm air cavity size. The hot spots were located within 3 mm inside the rectal‐air interface, where Dmax increased from 110.4 ± 0.5% without B‐field to 119.2 ± 0.8 % with B‐field. Conclusions Air cavities inside rectum affects rectum plan dosimetry due ERE. Location and magnitude of hot spots are dependent on the size of the air cavity.

presence of a magnetic field scatter and their trajectories are influenced by the resultant Lorentz force. When the electrons move from a high-density medium to a low-density medium, the electrons loop back due to Lorentz force and re-enter the high-density medium.
This results in a dose enhancement at the interface of the high and low media, often called the electron return effect (ERE). [4][5][6][7] In the presence of 1.5 T magnetic field, the radius of curvature of the electrons looping back can be as much as 1 cm. Various methods have been suggested to compensate for ERE such as the use of opposed beams or including the impact of magnetic fields in the inverse planning optimization which works well for stationary air cavities but not for variable air volume. 8 Electron return effect can pose a dosimetric concern for rectum patients treated on the hybrid MR-linac system using intensity-modulated radiation therapy (IMRT) as the air in the rectum is not consistent from day-to-day and may even change during the daily online plan adaptation process that can take as long as 45-60 min. 9 The goal of this study is to investigate and predict the effect of the magnetic field on rectum planning dosimetry due to the electron return effect. We specifically sought to investigate the impact and location of ERE on plan coverage and plan hot spots for rectum patients with air cavities of varying size, treated in both supine and prone positions.

| MATERIALS AND METHODS
Four patients previously treated at our institution with IMRT on a conventional linac were selected for this study. Patients were treated either in prone position on a belly board (mid and upper rectal) or supine (distal rectal) position in a customized aquaplast immobilization mold. All four of these patients received a total dose of 45 Gy to at-risk lymph nodes and 50 Gy to gross disease in 25 fractions using IMRT and simultaneous integrated boost (SIB) techniques. Figure 1 shows a typical beam arrangement and IMRT dose distribution, for example, rectum patient treated on a conventional linac. Our department constraints for 25 fractions dose regime are listed in Table 1.
To evaluate the effect of magnetic field on changing air volumes in rectum patients, two prone and two supine patients were selected, and artificial air cavities of diameters 0.0, 1.0, 1.5, 2.0, 2.5, 3.0, and 5.0 cm were created for all four patients. Air cavities extended superiorly and inferiorly within the high-dose PTV volume following the length of rectum anatomy, yielding a "tube"-like structure as shown in Fig. 2. For each patient, a reference plan with no air cavity and in the presence of magnetic field was optimized following department guidelines for target coverage/OARs constraints and prescribed to a total dose of 45 Gy to nodes and 50 Gy to gross disease in 25 fractions using IMRT and simultaneous integrated boost planning. Plans were generated using a 7 MV flattening filterfree beams and step and shoot delivery with the following settings: minimum segment area of 4 cm 2 , minimum segment width of 0.5 cm, minimum MU/seg of 5 MUs, and a total of 100 segments per plan. Nine beams spread around the patient, as shown in Fig. 1, avoiding entrance through high-density couch material were used.
Plans were calculated in Monaco version 5.4.0 treatment planning system using Unity beam data and the GPU Monte Carlo calculation algorithm (GPUMCD) with 0.3 cm grid size and 1% statistical uncertainty per calculation. 10,11 For patients with air gas inside bowel/rectum on the initial planning CT scan, both rectum/bowel contours F I G . 1. Typical dose distribution for a two-level PTV (PTV45/PTV50) rectum plan using IMRT and simultaneous integrated boost (SIB) techniques. were assigned density of water (relative ED = 1) during planning due to the fact that the air cavity will not be consistent from day to day.
For one case where small and scattered air pockets (few voxels in volume) were present, no density override was done on initial planning/optimization since its impact on dose calculation was assumed minimal.
To evaluate the effect of varying air cavity during online adaptive workflow in the presence of magnetic field, the initial reference plan was copied and recalculated (using same segments and MUs) for each new artificial air cavity diameter scenario with its relative electron density assigned to 0.18 (−824 HU), as shown in Fig. 2. Consequently, a total of seven plans in the presence of magnetic field were created for each patient: reference plan with no air cavity, 1.0-, 1.5-, 2.0-, 2.5-, 3.0-, and 5.0-cm-diameter air cavity plan. To investigate the effect of the ERE alone due to magnetic field, those seven scenarios were also evaluated without the presence of magnetic field: a different reference plan with no air cavity and with the same IMRT constraints was generated using the Unity research beam model with magnetic field turned off, and then copied and recalculated (using same segments and MUs) for the different air cavity diameters. A total of 56 plans were generated for this study.

2.A | Plan analysis
The following dosimetric parameters were extracted for each patient: D 95 PTV45DVH (PTV45 excluding PTV50), D 95 (PTV50-aircavity), V 50 (PTV50-aircavity), D max (PTV50-aircavity), and V 110% (PTV50-aircavity). Box plots with mean and standard deviation (along with individual data points) were generated for each of the above dosimetric parameters and plotted against air cavity diameter size. A polynomial fit was also performed to predict the magnitude of dose hot spots as a function of air cavity diameter size. To determine the location of hot spots due to ERE, annulus rings of 1-mm-diameter step size were generated for one of the air cavity plans  realistic values within the annulus, the dose calculation for 3 cm air cavity plan was redone using a 1 mm grid size.

3.A | PTV coverage vs air cavity size
The effect of air cavity size on PTV coverage was evaluated using

3.B | PTV hot spot versus air cavity size
PTV50-air cavity D max and V 110% experienced the greatest dosimetric impact due to varying air cavity size as shown in Figs. 7(a)-7(d).
In the presence of magnetic field, D max varied from 5523 ± 49 cGy for 0 cm air cavity size to 6050 ± 109 cGy for 5cm air cavity size.
The corresponding change in hot spot volume, described as V 110 (in cc) was 0.09 ± 0.16 cc for 0 cm air cavity to 11.6 ± 6.7 cc for a 5 cm air cavity size. A polynomial fit to the D max (%) and V 110% (cc) values was performed to predict hot spot values with varying air cavity size (Fig. 8)

Annulus results
The location and volume of ERE hot spots were estimated by gener-

| DISCUSSION
In this study, we investigated the dosimetric effect of varying air cavity size on rectum plans for patients treated on a hybrid 1.5 T Unity MR-linac system. We found that, in the absence of a magnetic field, the impact of varying air cavity size on plan dosimetry was limited to decreased attenuation due to increasing air volume and the variations seen were within statistical uncertainty. In the presence of a magnetic field, in addition to decreased attenuation due to increased air cavity, the curving path of electrons (or the electron return effect) at the tissue-air interface affected the dosimetric parameters. In terms of location, most of the hot spots due to ERE were located at the posterior interfaces of rectum wall. This is because the target is more  shown for the entire course of treatment, they are more appropriate for individual dose fractions since the air cavity may not be present for the entire course. Our study helps put things in perspective for the planner regarding the magnitude and location of hot spots to expect on a given day depending on the volume of the air cavity.
F I G . 7. D max and V 110% coverage of (PTV50-air cavity volume) structure in the presence (a,c) and absence (b,d) of magnetic field.
F I G . 8. D max and V 110% polynomial fit for (PTV50-air cavity volume) structure in the presence of magnetic field.

| CONCLUSIONS
In this study, we investigated the effect of electron return effect on rectum planning dosimetry for MR-guided RT. Our work shows that the location and magnitude of hot spots are dependent on the size of the air cavity. We also showed that tumor coverage degrades as a function of air cavity in relation to tumor volume.
The study has a potential to help physicists and physicians take appropriate intervention for daily adaptation based on the air cavity size for rectum patients treated on the Unity MR-linac machine.

ACKNOWLEDGMENTS
This research was partially supported by the NIH/NCI Cancer Center Support Grant/Core Grant (P30 CA008748). Authors would like to thank Miguel Hinojosa from Elekta for providing Unity beam data with no magnetic field.

CONF LICT OF I NTEREST
None.