Evaluation of dose distributions in the overlapping regions of the modulated volumetric arcs used in total body irradiation in pediatric patients

Abstract

Background

The application of total body irradiation (TBI), using intensity-modulated techniques with dynamic arcs, creates overlap at the treatment arc boundaries, which can lead to dosimetric variations. Therefore, this research aimed to evaluate the dose distributions in the overlap regions of the arcs used in total body irradiation in pediatric patients.

Material and methods

This dosimetric evaluation was conducted on a CIRS 715-TY0710 pediatric body simulator, where a treatment plan with three isocenters was designed, generating two 4 cm overlap regions along the line connecting the isocenters. The plan was verified using the gamma criteria of 3%/3mm and 5%/5mm with the Octavius 4D software.

Results and Conclusions

The planning results showed 98.2% dose coverage over 95% of the planning target volume (PTV) with adequate tolerance for risk organs. Regarding the gamma analysis, the acceptance rate was greater than 95% at the treatment isocenters with criteria of 3%/3 mm and 5%/5 mm. However, there is a decrease in the approved points in the overlap zone when evaluated with a 3%/3 mm criterion, while maintaining an approval percentage above 95% when evaluated with a 5%/5 mm criterion.

Introduction

Total body irradiation (TBI) is a technique that has become common in radiation therapy services due to its use as a conditioning regimen for bone marrow transplantation, indicated for a variety of neoplasms. This technique is designed to deliver a uniformly absorbed dose throughout the body with considerations for protecting risk organs such as the lungs, liver, and kidneys, among others. Additionally, this technique is classically concomitant with chemotherapy, which is used as a myeloablative conditioning regimen aiming to eliminate both neoplastic cells and stem cells (leading to new cells in the bone marrow) and to prevent rejection of the new donor cells [1, 2].

Traditionally, TBI uses static fields with lung-protecting blocks and requires the patient to stand or lie at a long source-to-surface distance (> 3 m), using lateral fields, placing the patient on a bed adapted with acrylic walls and filled with water bags to homogenize the body volume to be irradiated. In this arrangement, the patient runs the risk of developing certain conditions such as pneumonitis, pulmonary fibrosis, or renal failure [2, 3] due to complete organ irradiation; there are also other complications, such as non-uniform distribution of the absorbed dose and its inaccurate estimation, greater logistics, and extended treatment times, which in turn lead to complications when reproducing the position, as the configuration used is exhausting for an immunocompromised patient [4]. These complications have led to research seeking solutions, such as variation of the configuration or patient position [5], in which a reclined anterior posterior (AP)/posterior anterior (PA) position is proposed to decrease treatment difficulty. Another solution is the use of special implements and the implementation of a rotating bed [6]. Still another study proposes the use of a modified bed that includes an acrylic sheet to increase the surface dose [7].

Recently, there has been a proposal to apply modern technologies in radiation therapy for TBI treatments, such as volumetric modulated arc therapy (VMAT). This technique can significantly modulate the absorbed dose, generating homogeneous dose distributions [8]. In contrast to the traditional TBI modality, the TBI approach with VMAT has demonstrated the ability to improve treatment effectiveness and reduce dose exposure to critical organs [9]. This technological evolution promises to increase treatment efficacy, mitigate the risk to sensitive organs, and increase patient comfort due to a better configuration of the patient’s irradiation position. Despite the multiple benefits offered by this technique, it is crucial to consider relevant aspects such as the limitation in the movement capacity of the bed [6] and the need to use multiple treatment isocenters to cover the entire irradiation volume.

The use of multiple treatment isocenters leads to the overlap of the arcs at the boundaries of the irradiation fields, creating a region with the possibility of high dose gradients, meaning a slight deviation in positioning could cause underdosing or overexposure, with potential adverse effects to the patient. Therefore, there is a critical need for meticulous quality control to avoid these dosimetric risks, hence the objective of the work is to evaluate the dose distributions in the overlap regions of the arcs used in total body irradiation in pediatric patients, using an anthropomorphic pediatric phantom.

Materials and methods

Image acquisition

The CIRS 715T pediatric phantom was used, which replicates the thoracic structure of a 5-year-old child whose soft tissue is designed based on urethane and epoxy compounds. This material is applicable in imaging and dosimetry, covering a photon spectrum from 50 keV to 25 MeV. The tomographic images of the phantom allow for the distinction of soft tissues, bones, and low-density areas corresponding to the lungs. These tomographic images of the entire phantom volume were acquired with 3 mm slices using the PHILIPS Brilliance CT Big Bore machine. Six fiducial markers were placed to generate two isocenters at the height of the sternum and another at the navel, allowing for better positioning of the phantom on the bed. These images are exported to the Monaco treatment planning software (TPS), version 6.1.3.

Contouring and planning

For a dosimetric prescription of 12 Gy in 6 fractions, the Monaco TPS was used to perform the virtual simulation in two phases. First, the radiation oncologist delineated the volumes of interest (tumoral volume and risk organs) on the tomography images of the pediatric phantom. In this phase, the complete volume of the phantom was contoured considering an internal margin of 0.3 cm. In the second phase, the physicist set up the treatment beam configuration (arc) and obtained a dose distribution for an X-ray beam with a nominal voltage of 6 MV provided by the Synergy Full linear accelerator, equipped with an Agility 160-leaf collimator with 5° angulation to all arcs.

The treatment plan configuration was executed using the Monte Carlo calculation algorithm [10] with a calculation grid (grid spacing) of 0.3 cm, a surface margin of 0.3 cm, and a beamlet width of 0.3 cm. It considered 3 isocenters spaced 14.7 cm apart; at each isocenter, 3 complete 360° arcs were generated, and the field overlap option was implemented, which encompassed a length of 4.0 cm on the cranio-caudal axis.

Dosimetric verification

For the dosimetric verification, the Octavius 4D system was used (with an array of 1500 ionization chambers), synchronized with the movement of the gantry. An inclinometer is used to constantly check this synchronization, which allows the collection of dosimetric information at 200 ms intervals throughout the reading process [11]. The VeriSoft software was used to analyze this information, through the evaluation of the gamma index, which is calculated by comparing the dose distributions generated by measurements carried out with the Octavius 4D and the Monaco TPS, carried out for the three planes (axial, sagittal and coronal) and in the entire volume [12, 13].

The comparison of dose distributions was carried out for five regions of the pediatric phantom, where the first three were performed at the treatment isocenters, following the traditional approach; while the next two were in the overlap regions, where additional considerations are required, such as positioning the center of the arc overlap in the center of the equipment (Fig. 2). In this way, the region of interest was aligned with the isocenter of the phantom, and verifications in these overlap regions were carried out by creating two quality control treatment plans, where the first included the first and second arcs, and the second plan included the second and third arcs. The dose distribution measurements were performed with the Octavius 4D equipment, and then the dose distributions generated from the quality control were compared with those calculated by the Monaco TPS through γ index analysis.

Figure 1.

Alignment of fiducial markers placed on the CIRS 715T phantom

Figure 2.

Schematic for the dosimetric verification of the overlap region with the Octavius 4D phantom

An acceptance criterion of 95% was established; that is, it is considered satisfactory when the percentage obtained is equal to or greater than 95%. Additionally, a dose threshold equivalent to 10% of the maximum dose was defined as the minimum value required to consider including a specific voxel in the statistics. The evaluation criteria were based on a 5%/5 mm. In addition to these criteria, a second evaluation was carried out applying a 3%/3 mm criterion for both overlap regions. These parameters were also used in the measurements corresponding to the second overlap.

Results

Treatment planning

A total of 1,288 segments were generated (with a greater proportion in the first arc, which covers a larger treated area due to the modulation performed by the TPS on the lungs). The resulting treatment plan is considered adequate because it encompasses more than 95% of the pediatric phantom volume (excluding the lungs, kidneys, and liver) with 95% of the prescribed dose (also, 100% of the prescribed dose encompasses 90% of the volume), and the risk organs have absorbed dose values below the tolerable limits (Tab. 1) as per the constraints given by Morrison et al. (2017) and Tas et al. (2018) [14, 15].

Table 1.

Reference parameters and parameters obtained in the generated plan

Volume	Reference parameters	Parameters obtained
PTV	V100% ≥ 90%	90%
	V95% ≥ 95%	98.23%
	V107% ≤ 2%	1.24%
Lungs	Average dose right lung (< 8 Gy)	8.1 Gy
	Average dose left lung (< 8 Gy)	8.2 Gy
Kidneys	Average dose right kidney (< 9 Gy)	8.3 Gy
	Average dose left kidney (< 9 Gy)	8.1 Gy
Liver	Average dose (< 9Gy)	8.5 Gy

PTV — planning target volume

During the planning process, it was evident that the (internal) margins assigned to the risk organs significantly contributed to the creation of an appropriate dose gradient in the region of the internal margins (between the PTV and these organs), achieving effective control of hot spots within the PTV. This is reflected in a visual inspection, where no significant changes in the overlap region were observed; only a dose drop in the risk organs could be noted, as expected (Fig. 3).

Figure 3.

Dose distribution in the CIRS 715T phantom following treatment planning

In the dose-volume histogram (DVH), the dose in the risk organs varies from approximately 6 Gy to 11.5 Gy, as shown in Figure 4.

Figure 4.

Dose volume histogram from the treatment planning of the CIRS 715T phantom. PTV — planning target volume

Figure 5A displays the distribution of absorbed dose in the pediatric phantom, CIRS 715T, generated by each treatment arc individually and spaced 5 cm apart from each other, allowing for a clear view of the dose distribution from the individual arcs.

Figure 5.

Analysis of the dose profiles in the overlap of the treatment fields

Figure 6.

Comparison of the planning system data and the dosimetric measurements with the Octavius 4D

In Figure 5B, we can observe a profile generated along the midline of the pediatric phantom, where the dose overlap from the fields is visible. Also noticeable are small fluctuations in absorbed dose around the prescribed dose level, continuing across the entire profile until reaching the overlap region. Here, there is a sharp drop in dose, coinciding with the dose fall-off tail from the profile of the subsequent arc. This compensates for the high and low doses so that the summed values are around the prescribed dose.

Dosimetric verification

This verification reports small differences (in regions of high dose gradients) between the dose distributions generated by the TPS and the dosimetric measurements with the Octavius 4D. A not very pronounced peak is also observed in the center of the overlap region, having a small impact relative to all the voxels analyzed; in the regions of low dose gradient, there was no significant variation.

Additionally, some slices of the peripheral regions of the pediatric phantom show notable variation in the regions of low doses, which have limited dosimetric relevance. It is important to note that the information from the planning system covers the entire extent of the pediatric phantom, while the dosimetric data from the Octavius are limited to the sensitive part of the chamber array. On the other hand, some localized hot spots are observed in the overlap region.

Regarding the γ index criterion, it is important to note that the analysis was conducted using distance to agreement (DTA) and dose percentage values of 3%/3 mm and 5%/5 mm, with a dose threshold of 10% and using the maximum dose from the calculated volume for the analysis. Three-dimensional γ index analyses of the absorbed dose distributions in both field overlap regions were obtained. The low-dose regions that were distant from the center of the phantom showed highly discrepant values; however, this region was not analyzed because the values were below the 10% threshold.

The results of the gamma analysis for the treatment isocenters as well as the overlap regions can be seen in Figure 7. It shows that for the treatment isocenters, the gamma analysis with either 3%/3 mm or 5%/5 mm criteria achieves an approval percentage above 95%. However, when analyzing the overlap zones, the 3%/3 mm criteria result in an approval percentage above 86%, but using the 5%/5 mm criteria, the approval percentage exceeds 95%.

Figure 7.

Results of the gamma analysis with dose percentage and distance to agreement (DTA) of 3%/3 mm and 5%/5 mm

Discussion

Concerning the gamma analysis of TBI with VMAT, Pierce et al. (2019) [7] achieved an acceptance rate of over 90% using a 5%/1 mm γ index criterion, according to a two-dimensional analysis in certain surface areas, but no data were reported on the overlap of the treatment arcs. Tas et al. (2018) [15] conducted an analysis focused on the dose-volume histogram, which is not directly comparable with the methodology used in this research. On the other hand, Chakraborty et al. (2015) [16] coupled all arcs into a single isocenter, reporting approval rates of 93% with a 3%/3 mm γ index criterion. However, their evaluation did not include the overlap areas. In contrast to these findings, our study reveals approval rates over 95% at the treatment isocenter when using the 3%/3 mm criterion.

Morrison et al. (2019) [17] evaluated the gamma index criterion in the overlap region, as did the present study. In their analysis, the authors used the γ index criterion of 3%/3mm and achieved a pass rate of 97.5%, while our study had a pass rate above 86% using the same criterion. However, it is important to note that different software was used. Morrison et al. (2017) [14] used Arccheck, while we used Octavius 4D. It is worth mentioning that we found an approval rate above 95% when using the 5%/5mm analysis criterion.

It is crucial to consider that the positioning at each isocenter is of paramount importance, as small variations in positioning can cause the arcs to converge or diverge in the overlapping zone, potentially leading to underdosing or overdosing. Furthermore, volumetric evaluation of the gamma index demonstrates a greater ability to identify significant errors compared to a 2D analysis.

Conclusion

In this study, a pediatric anthropomorphic phantom was used to obtain the absorbed doses in the overlapping regions of the arcs used in total body irradiation. The dosimetric results obtained provide valuable insights into how radiation doses are being distributed to these critical areas of the body. With this data, it is possible to improve the knowledge of the dose distribution on the overlapping regions. For the treatment isocenters, the γ index with either 3%/3 mm or 5%/5 mm criteria achieves an approval percentage above 95%, which is the same for overlapping zones using the 5%/5 mm criteria.

Given the use of large treatment fields and the complexity involved, we believe it is acceptable to apply the 5%/5 mm criterion with an approval rate above 95% in the overlap region. For this purpose, it is important to consider the area where overlaps occur, as the dose limits in organs at risk must remain below the maximum permissible threshold; otherwise, the assumed tolerances could lead to localized dose increases in the overlap zones.