Received: 10 July 2020 | Revised: 15 December 2020 | Accepted: 9 January 2021

DOI: 10.1002/acm2.13189

RADIATION ONCOLOGY PHYSICS

Automatic treatment planning for VMAT-based total body

irradiation using Eclipse scripting

Jose R. Teruel | Sameer Taneja | Paulina E. Galavis | K. Sunshine Osterman |

Allison McCarthy | Martha Malin | Naamit K. Gerber | Christine Hitchen |
David L. Barbee

Department of Radiation Oncology, NYU

Langone Health, New York, NY, USA Abstract
The purpose of this work is to establish an automated approach for a multiple isocen-
Author to whom correspondence should be
addressed. Jose R. Teruel ter volumetric arc therapy (VMAT)-based TBI treatment planning approach. Five anon-
E-mail: [email protected] ymized full-body CT imaging sets were used. A script was developed to automate and
Telephone: 212-731-6166
standardize the treatment planning process using the Varian Eclipse v15.6 Scripting
API. The script generates two treatment plans: a head-first VMAT-based plan for
upper body coverage using four isocenters and a total of eight full arcs; and a feet-first
AP/PA plan with three isocenters that covers the lower extremities of the patient.
PTV was the entire body cropped 5 mm from the patient surface and extended 3 mm
into the lungs and kidneys. Two plans were generated for each case: one to a total
dose of 1200 cGy in 8 fractions and a second one to a total dose of 1320 cGy in 8
fractions. Plans were calculated using the AAA algorithm and 6 MV photon energy.
One plan was created and delivered to an anthropomorphic phantom containing 12
OSLDs for in-vivo dose verification. For the plans prescribed to 1200 cGy total dose
the following dosimetric results were achieved: median PTV V100% = 94.5%; median
PTV D98% = 89.9%; median lungs Dmean = 763 cGy; median left kidney Dmean =
1058 cGy; and median right kidney Dmean = 1051 cGy. For the plans prescribed to
1320 cGy total dose the following dosimetric results were achieved: median PTV
V100% = 95.0%; median PTV D98% = 88.7%; median lungs Dmean = 798 cGy;
median left kidney Dmean = 1059 cGy; and median right kidney Dmean = 1064 cGy.
Maximum dose objective was met for all cases. The dose deviation between the treat-
ment planning dose and the dose measured by the OSLDs was within 4%. In sum-
mary, we have demonstrated that scripting can produce high-quality plans based on
predefined dose objectives and can decrease planning time by automatic target and
optimization contours generation, plan creation, field and isocenter placement, and
optimization objectives setup.

KEY WORDS
advanced treatment planning, eclipse scripting, plan automation, special procedures, total body
irradiation (TBI)

----------------------------------------------------------------------------------------------------------------------------------------------------------------------
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2021 The Authors. Journal of Applied Clinical Medical Physics published by Wiley Periodicals LLC on behalf of AmericanAssociation of Physicists in Medicine.

J Appl Clin Med Phys 2021; 22:3:119–130 wileyonlinelibrary.com/journal/jacmp | 119

120 | TERUEL ET AL.

1 | INTRODUCTION of lungs, kidneys, and any other organs with increased risk due to
patient comorbidities or previous radiation history.24–28
Total body irradiation (TBI) is a special radiation therapy (RT) pro- Treatment planning approaches for VMAT-based TBI are not yet
cedure in which radiation is administered to the full body of the standardized, can be hard to develop, and may vary across institu-
patient. In combination with chemotherapy, TBI is one of the tions. Placement of fields, isocenters, and planning technique is
therapeutic components of conditioning regimens used to condi- solely based on each institution’s individual efforts and experience
tion patients with hematological neoplasms for hematopoietic stem due to the novelty of the technique and the lack of standardization.
cell transplantation (HCT), primarily those affected by acute mye- The purpose of this work is to establish a fully automated approach
loid leukemia (AML) and acute lymphoid leukemia (ALL). TBI for isocenter and treatment field placement as well as dose objec-
enhances antineoplastic therapeutic efficacy due to its potential to tives selection for optimization for VMAT-based TBI treatment plan-
reach sanctuary sites, such as the testis and the central nervous ning. This objective serves two purposes: First, to provide a
system (CNS), and provides immunosuppression that prevents consistent and standardized planning technique. Second, to ensure
bone marrow transplant rejection.1–4 Radiation-induced interstitial consistent shifts at treatment in order to prevent errors when treat-
pneumonitis is a major concern for patients undergoing TBI. ing multiple isocenters.
According to the International Lymphoma Radiation Oncology
Group, pneumonitis occurs in about 25% of patients receiving
2 | MATERIAL AND METHODS
fractionated TBI.1
From a physics standpoint, guidelines for administering TBI are
2.A | Subjects
outlined in report no. 17 from the AAPM Task Group 29.5 Histori-
cally, administration of TBI is delivered with the patient at an Five full body anonymized CT scans from adult patients previously
extended distance (extended SSD), such that the radiation field treated at our institution using conventional TBI were used retro-
encompasses the patient’s entire body. This technique continues as spectively. CT images were obtained using a CT SOMATOM scanner
the standard of practice in most cancer centers performing TBI. (Siemens, Erlangen, Germany). Patients were simulated in the supine
Open field treatments used for conventional TBI normally require position. The protocol employed a 1 cm slice thickness, 500 mm
the use of a beam spoiler to ensure coverage at shallow depths acquisition diameter, and extended field of view (FOV) reconstruc-
when using high-energy beams. Compensators are recommended as tion of 650 mm. The use of anonymized retrospective CT scans for
well in an effort to obtain homogenous dose distributions.6,7 How- dosimetry studies was approved by our Internal Review Board and
ever, previous work has demonstrated that conventional hand calcu- consent was waived. Additionally, an anthropomorphic body phan-
lations for TBI without tissue heterogeneity correction result in tom (CIRS, Norfolk, VA) was scanned using the same CT simulator
significant dose underestimation, particularly in the lungs for high- and a 5mm slice thickness. The phantom included all anatomical sec-
energy bilateral treatments.8 tions from mid-thigh to head with no upper or lower extremities.
One of the main drawbacks of at-distance open beam treatments
used for TBI is the inability for selective sparing of organs at risk
2.B | Treatment planning
(OARs). Interstitial pneumonitis is the major side effect from high-
dose TBI radiation therapy and can be fatal in some instances.1,9–11 A script was developed to automate and standardize the treatment
Additionally, high-dose TBI treatments can produce late toxicities, planning process. The script was developed in C# programming lan-
12–15
such as chronic kidney dysfunction and secondary malignancies. guage using the Varian Eclipse v15.6 Scripting API (Varian Medical
In order to mitigate TBI-induced acute and chronic side effects, the Systems, Palo Alto, CA). A summary of the treatment planning tasks
use of shielding blocks, particularly for lungs and in some instances that were automated is presented in Supplemental Figure S1. The
for kidneys, is widely accepted in high-dose TBI treatments.16–19 input of the script is a structure set that must include at a minimum
However, accurate placement of beam modifying devices relative to the following contours: body, lungs (including unilateral contours)
the patient in treatment position presents challenges due to limita- and kidneys (including unilateral contours). These contours are cre-
tions in image verification, intrafraction patient motion, and repro- ated manually by the dosimetrist. Additionally, the user origin loca-
ducible patient setups. tion needs to be entered prior to running the script. For these
More recently, several alternatives to at-distance treatment for retrospective cases, an approximate location at body midline (ante-
TBI have been reported. Helical Tomotherapy (Accuray Inc, Sunny- rior–posterior), lungs midline (craniocaudal), and sternum (left–right)
vale, CA) has been employed in single institution studies for TBI and was employed. For prospective patients to be treated with VMAT-
total marrow irradiation (TMI) treatments to obtain higher coverage TBI the user origin location will be determined at simulation and
for sites at high risk of recurrence while sparing major OARs, such radiopaque ball bearings (bbs) will be placed to allow for user origin
20–23
as the lungs, liver, and kidneys. In addition, there is ongoing placement during treatment planning.
effort to explore the use and feasibility of volumetric arc therapy Script execution prompts the user to select a fractionation option
(VMAT)-based TBI treatments in order to obtain a more homoge- and provide the intercept of the laser with the couch top longitudi-
neous dose distribution, better target coverage, and better sparing nal scale at the user origin location recorded at simulation (Fig. 1).
TERUEL ET AL. | 121

Currently, two options are available for planning: 150 cGy × 8 frac- couch travel limitation presented above and resulting in an isocenter
tions BID to a total dose of 1200 cGy; and 165 cGy × 8 fractions spacing of 22.7 cm. The treatment fields are set with 90° collimator
BID to a total dose of 1320 cGy. rotation and jaws are set asymmetrically to obtain an overlap of
After the planner selects the option according to prescription, 2 cm for fields sharing an isocenter, and an overlap of 5.3 cm for
the script performs the following tasks automatically. The script cre- arcs of adjacent isocenters. The 2 cm overlap for the fields sharing
ates one treatment course that contains two automatically generated the chest isocenter can be adjusted automatically by the script, if
plans, an upper body VMAT plan and a lower body 3D plan. The required, to improve lungs sparing. The script calculates the center
upper body plan contains four isocenters with two full arcs per slice of the lungs contour in the craniocaudal direction and will
isocenter. For the generation of this upper body plan the script uses increase the jaw openings if the field edge of any of the jaws used
prior knowledge of the couch indexing and couch longitudinal travel to create the MLC based island block is more than 2 cm away from
to maximize the area of the patient treated with VMAT. Two couch the center of the lungs. An example is presented in Fig. 3. The over-
tops (Orfit Industries, Wijnegem, Belgium) were selected for prospec- lap is defined at the user origin coronal plane. For the inferior half of
tive VMAT-TBI treatments to provide immobilization required for the patient that cannot be accommodated in the head-first orienta-
VMAT delivery and reduce setup variability. The superior couch top tion, the script creates a lower body, feet-first plan using three
provides a fixed location for the head rest with five point indexing equally spaced isocenters. This lower body plan is a 3D anterior–-
for head and neck thermoplastic immobilization mask and four point posterior/posterior–anterior (AP/PA) plan with field matching at the
indexing for chest and abdomen thermoplastic immobilization cast. isocenter plane. The isocenters for the lower body plan are placed
The inferior couch top is an extremity plate that allows individual 4 cm posteriorly compared to the upper body plan to account for
indexing of both feet separately. The complete arrangement includ- the smaller anterior–posterior width on the legs compared to the
ing couch tops and immobilization masks is presented in Fig. 2. Stan- chest and abdomen. Overlap between the lower body plan and
dardized indexing of the couch tops on top of the linear accelerator upper body plan, as well as flash for the most inferior field is
couch together with the fixed position of the head rest in the couch accounted by the script using the following steps: (a) The script finds
tops, allows for a consistent head isocenter location at 87 cm longi- the most inferior slice of the body contour. (b) The script adds 3 cm
tudinal (1 cm tolerance) and a maximum accepted 155 cm longitu- inferiorly to the longitudinal location detected in the prior step. (c)
dinal (1 cm) for the pelvis isocenter that is built into the script The script calculates the distance from this slice to the most inferior
(maximum longitudinal of 160 cm on Varian 6 degrees of freedom slice included in the upper body VMAT plan. (d) The script adds
couch). The script places equidistant isocenters starting superiorly 5 cm to that distance creating overlap between the lower body and
and includes 2 cm flash from the most superior slice of the body the upper body plan. (e) The script splits that distance into three
contour. The exact process employed by the script for isocenter equally spaced isocenters. (f) The script creates two fields (AP/PA) at
placement for the upper body VMAT plan is the following: (a) The each isocenter location with symmetric jaws and field matching at
script finds the body contour and locates the most superior slice of the user origin plane.
the contour (top of the head). (b) The script adds 2 cm superiorly to The script creates four targets: PTV; PTV_Sup; PTV_Inf; and
the CT image longitudinal coordinate (to allow for flash). (c) The PTV_Sup_Norm. The automatically generated PTVs are defined
script sets the first isocenter (head isocenter) 19 cm inferior from according to the following definitions:
the previously calculated longitudinal coordinate. (d) The script cre-
ates all subsequent isocenters using the information regarding the

F I G . 2 . Orfit couch tops and thermoplastic immobilization for

VMAT-based TBI CT simulation. The presented thermoplastic
devices were prepared on a healthy volunteer as part of establishing
the program. Thermoplastic immobilization for the upper body board
F I G . 1 . Graphic user interface of the treatment planning tool have fixed (one location) indexing. Thermoplastic immobilization for
showing dropdown to select fractionation. Additionally, it requests each individual foot can be placed in several locations in the board
an input value to calculate all the treatment couch longitudinal to allow for comfortable immobilization for patients with different
values based on the laser intercept at simulation. heights as well as to control for leg separation.
122 | TERUEL ET AL.

fractionation selected and will load them automatically for optimiza-

tion. Some additional options, such as aperture shape controller (set
to moderate), air cavity correction (set to on), and jaw tracking (set
to enabled), are set automatically by the script for the photon opti-
mizer PO v.15.6. Finally, the script uses the user origin location, the
couch intercept value entered by the planner in the graphic user
interface, and the couch entered in the CT to provide the planner
with all couch coordinates pertaining to the created isocenter
arrangement. The script presents these values on the screen (Fig. 4),
and creates a “.csv” file to be loaded in an in-house tool employed
to assist with imaging and treatment delivery (out of the scope of
this report). Additionally, the time required to run the script was
compared to the time required to manually perform all the tasks
completed by the script.
After script execution, the weights of the inferior 3D plan are
manually adjusted to maximize coverage of the inferior target
structure (PTV_Inf) while maintaining the maximum dose in that
region below 130%. In the next planning step, a second instance
of the script is executed to automatically run the optimization of
the upper body VMAT plan using the preloaded dose objectives,
and automatically calculate the plan after optimization. This second
part of the script can be run in standalone mode, therefore, open-
F I G . 3 . Initial VMAT field arrangement for two cases right after ing an independent instance of Eclipse, saving the progress and
script execution. In both cases the total craniocaudal length covered closing Eclipse allowing the planner to execute it after hours or
by the fields remain the same (106 cm). Fields that share an
overnight.
isocenter overlap by 2 cm. Fields that not share an isocenter overlap
All treatment plans were created using a Varian TrueBeam
by 5.3 cm. The final jaw shape is defined during optimization using
jaw tracking. Left: Larger patient where maximum coverage of upper machine with Millennium MLC, 2.5 mm optimization grid, 5 mm cal-
body VMAT fields stop mid-thigh. Right: Shorter patient culation grid, air cavity correction, and the analytic anisotropic algo-
demonstrating coverage up to the superior aspect of the knee. The rithm (AAA). The energy employed was 6 MV photons for all fields
superior jaw of the fourth field (chest isocenter, inferior field) in this with a maximum allowed nominal dose rate of 600 MU/min.
case is automatically extended 3 cm to compensate for the isocenter
location inferiorly in the lungs.

• PTV is the body contour with a 5 mm margin inside the body sur-
face and extended 3 mm into the lungs and kidneys.
• PTV_Sup is defined as the subsection of the PTV that is covered
by the superior VMAT fields.
• PTV_Inf is the subsection of the PTV defined as the PTV minus
PTV_Sup.
• PTV_Sup_Norm is defined as the PTV_Sup excluding the region of
overlap with the inferior AP/PA fields.

Additionally, the script creates eight optimization structures.

Optimization structures are defined as the union of each individual
VMAT field with the PTV structure. Due to the divergence of the
fields, we consider the intercept at the coronal plane of the user ori-
gin. These optimization structures are named opt_ptv_X, where X is
the beam number of each individual field from the upper body
VMAT plan. Therefore, the optimization structures are named con- F I G . 4 . Script output presenting treatment couch values to assist
dosimetrist with treatment preparation. The script outputs each
secutively from opt_ptv_1 (union of the PTV and the area covered
isocenter’s longitudinal couch value and vertical couch values
by first field of the head isocenter) to opt_ptv_8 (union of the PTV grouped by plan and field (in case an extended distance is required
and the area covered by the second field of the pelvis isocenter). for the lower body plan) and lateral couch values for each plan.
The script will create all optimization objectives based on the Additionally, these values are exported by the script into a “.csv” file.
TERUEL ET AL. | 123

each individual axis (lateral, vertical, and longitudinal). Second, we

2.C | Dose objectives
evaluated dose profiles at the overlap regions of fields not sharing
Two levels of planning goals were established for this treatment an isocenter.
planning study. Primary goals must be met for every plan and
include: target coverage of PTV V100% >90%, target near minimum
dose of D98% >85%, and target maximum dose of PTV D2cc 3 | RESULTS
<130%. The primary goals for the organs at risk (OARs) include: lung
mean dose <800 cGy for 1200 cGy plans or <900 cGy for The result from the script execution is illustrated in Figs. 4 and 5.
1320 cGy plans, individual kidney mean dose <1100 cGy, and maxi- The total time the initial script takes to run is negligible in terms of
mum dose to any OAR D0.03cc <120%. Secondary goals include the planning time (less than 10 s using a Varian workstation). This
following: target coverage PTV V100% >95%, target near minimum included course creation, creation of superior and inferior plans, cal-
dose of PTV D98% >90%, and a target mean dose <110%. For the culation of optimal location of isocenters, placement of inferior plan
OARs, the secondary goal is lung mean dose <800 cGy for fields, placement of superior plan fields, creation of four target con-
1320 cGy plans. tours and eight optimization contours, and loading the appropriate
optimization template, and calculation of all treatment couch coordi-
nate for each isocenter. The time required to perform all these tasks
2.D | In-vivo dose verification, QA, and plan
manually was variable, but it was estimated to comprise approxi-
uncertainty analysis
mately 2–3 hrs with the calculation of optimal location of isocenters
For in-vivo dose verification one treatment plan was created for the taking about 20–40 min, the setup of fields for both plans with
CT dataset of the anthropomorphic phantom. This plan was created appropriate overlap and jaw setting about 40–60 min, the creation
using the script and following the same rationale as described for of all target and optimization contours about 30–50 min, manual cal-
the patient dataset with the only difference being the absence of a culation of each treatment couch coordinate for 7 isocenters about
lower body plan (phantom does not include extremities). The treat- 1020 min, and the rest of the tasks required prior to optimization
ment plan was delivered to the phantom previously loaded with taking about 20 min. The time required to perform all the tasks man-
optically stimulated luminescence dosimeters (OSLDs). Twelve ually considers that they are all completed successfully. However,
OSLDs were placed in different key locations including lungs (four we found that due to the complexity, even a small error in setting
OSLDs), kidneys (two OSLDs), soft tissue (two OSLDs), and bone up isocenters or manual field overlap may go unnoticed until after
(four OSLDs). OSLDs were placed inside the phantom using drilled optimization which may result in up to a full day of time savings
OLSD holders provided by the vendor. OLSDs were read using a when using the script for these tasks.
microSTARii system (Landuaer, Glenwood, IL). The OSLD reader was The body habitus volumes of the cases under study ranged
calibrated following vendor recommendations, daily QA was per- between 56599 and 90932 cm3 with a height range between 152
formed prior to reading OSLDs, and nonpreviously irradiated or and 181 cm, maximum lateral width between 45.4 and 62.2 cm, and
annealed nanoDots were employed (quoted uncertainty by the ven- lung volume between 1716 and 3260 cm3. All measurements were
dor of 5.5%). The dose difference between the recorded dose by the obtained directly from the planning CT and represent the dimension
OSLD and the treatment planning system (TPS) dose was evaluated of the patient in simulation position. In Fig. 3, two cases with differ-
using a normalized dose difference metric defined as: ent patient heights are presented, illustrating how the script priori-
tizes coverage in the craniocaudal orientation with the upper body
DoseOSLD DoseTPS
Dosedeviationð%Þ ¼ 100 VMAT plan.
DoseRx
All primary dosimetric goals were met for all calculated plans. A
Patient-specific QA was performed for this plan using three dif- summary of the dosimetric results is presented in Table 1 for
ferent methods: (a) A dose verification plan was created for a SNC 1200 cGy total dose plans and in Table 2 for 1320 cGy total dose
ArcCheck (SunNuclear, Melbourne, FL) diode array device and each plans. These results were obtained with only one optimization using
field was evaluated separately; (b) A portal dosimetry verification the generic goals loaded by the script. For the plans prescribed to
plan was delivered to the electronic portal imager device (EPID) and 1200 cGy total dose (150 cGy × 8 fractions) the following dosimet-
analyzed using Varian Portal Dosimetry tool for each field; (c) Portal ric results were achieved: median PTV V100% = 94.5%; median PTV
Dosimetry measurements were analyzed using SNC PerFRACTION D98% = 89.9%; median lungs Dmean = 763 cGy; median left kidney
Fraction 0. Additionally to patient-specific QA verification using the Dmean = 1058 cGy; and median right kidney Dmean = 1051 cGy.
described three methods for each individual field, the area of overlap For the plans prescribed to 1320 cGy total dose (165 cGy × 8 frac-
between fields not sharing an isocenter was measured and analyzed tions) the following dosimetric results were achieved: median PTV
using SNC ArcCheck. V100% = 95.0%; median PTV D98% = 88.7%; median lungs
Finally, we analyzed plan uncertainty for the phantom treatment Dmean = 798 cGy; median left kidney Dmean = 1059 cGy; and
plan in two ways. First, we evaluated plan uncertainty by applying median right kidney Dmean = 1064 cGy. The isodose coverage at
the same rigid shifts to all isocenters for 5 mm and 10 mm shifts for four different relevant levels (lungs, different isocenter field
124 | TERUEL ET AL.

F I G . 5 . Results after execution of the automatic planning script. Left: Structure set. The structures inside the blue box are automatically
generated as described in the Methods section. Middle: Automatic field placement result. Right: Example of some of the optimization
objectives loaded by the script.

matching, kidneys, superior VMAT, and inferior APPA match region) (V100%), our plan uncertainty analysis revealed that the decrease in
are presented in Fig. 6 for both treatment plans (1200 cGy and coverage was below 3% for all scenarios analyzed. After analyzing
1320 cGy total dose) for one case. dose profiles at field junctions we found a dose increase/decrease of
The normalized dose deviation for all OSLD measurements was about 20% per cm when uncertainty shifts were applied to a single
less than 4% for each individual OSLD. Individual values are pre- isocenter in the longitudinal direction. This amount was expected
sented in Table 3. Figure 7 illustrates the location of the OSLDs in considering the use of the auto-feathering tool in Eclipse and the
the anthropomorphic phantom as well as the location taken in the overlap distance of 5.3 cm. However, there is some variability in
treatment planning system to obtain treatment plan dose. Patient- areas mostly located at higher dose heterogeneity locations or
specific QA for this plan obtained passing rate over 95% for each around high/low-dose interfaces. Figure 9 includes an example of
individual field and the areas of field overlap using a gamma criterion auto-feathering at one of the dose junctions.
of 2% at 2 mm. The passing rate was >95% for all three modalities
of patient-specific QA employed. Figure 8 presents the analysis using
ArCheck and Portal Dosimetry for one of the chest isocenter fields. 4 | DISCUSSION
Our plan uncertainty analysis revealed that the increase in the
mean dose to the lungs was always below 3% for all plan uncertain- Our results demonstrate that automating VMAT-based treatment
ties evaluated with the exception of a 10 mm lateral shift that planning for TBI can reduce treatment planning time, increase con-
resulted in a 4.3% mean lung dose increase. Regarding PTV coverage sistency in isocenter placement, and decrease variability on field

T A B L E 1 Dose objectives and results for 1200 cGy total dose plans T A B L E 2 Dose objectives and results for 1320 cGy total dose plans
(150 cGy per fraction) generated using treatment planning script. (165 cGy per fraction) generated using treatment planning script.
Median values with minimum and maximum in parentheses. Median values with minimum and maximum in parentheses.
Objectives defined as a range are primary–secondary goals. Objectives defined as a range are primary–secondary goals.
Structure Dose objective Result Structure Dose Objective Result
PTV V100% >90-95% 94.5% (93.7, 95.2)% PTV V100% > 90-95% 95.0% (93.1, 95.2)%
D98% >85-90% 89.9% (88.7, 91.3)% D98% > 85-90% 88.7% (85.9, 90.3)%
Dmean <110% 108.6% (106.1, 110.0)% Dmean <110% 108.8% (107.1, 110.4)%
Lungs Dmean <800 cGy 763 cGy (729, 783) cGy Lungs Dmean <900-800 cGy 798 cGy (772, 811) cGy
D0.03cc <120% 116.1% (114.4, 118.0)% D0.03cc <120% 111.1% (109.0, 116.4)%
Left kidney Dmean <1100 cGy 1058 cGy (1030, 1066) cGy Left kidney Dmean <1100 cGy 1059cGy (1047, 1067) cGy
D0.03cc <120% 110.2% (108.4, 112.7)% D0.03cc <120% 106.7% (104.7, 110.3)%
Right kidney Dmean <1100 cGy 1051 cGy (1025, 1083) cGy Right kidney Dmean <1100 cGy 1064 cGy (1051, 1081) cGy
D0.03cc <120% 111.3% (105.6, 114.7)% D0.03cc <120% 107.7% (105.6, 108.9)%
Body D2cc <130% 129.9% (126.9, 130.0)% Body D2cc <130% 128.7% (125.7, 130.0)%
TERUEL ET AL. | 125

F I G . 6 . Dosimetric distribution in four axial views. Top to bottom: (a) Area of overlap between two fields with different isocenters (head
isocenter and chest isocenter). (b) Slice showing the lungs. (c) Slice showing the kidneys. (d) Area of overlap between upper body VMAT
treatment plan and lower body AP/PA treatment plan (AP/PA plan used as base plan for VMAT optimization). Left: Treatment plan for
1200 cGy (150 cGy per fraction). Right: Treatment plan to 1320 cGy (165 cGy per fraction).

configuration, field matching and field overlap. All the plans gener- do not produce large dose deviations as long as they are maintained
ated using this approach met our expected primary dosimetry goals. within 10 mm. The largest uncertainties will occur if the planned lon-
Additionally, standardization with regard to isocenter placement, and gitudinal distance between isocenters is not maintained at treatment.
particularly isocenter shifts, can prevent errors and streamline the For this reason, we have developed an in-house software to assist
process for treatment delivery. Furthermore, automation can facili- with our image-guided radiation therapy (IGRT) approach. While a
tate the optimization tasks by providing dosimetrists with automated full description of this tool is out of the scope of this report, in brief,
upload of optimization templates. Our in-vivo end-to-end dosimetry the software will prompt therapists to acquire IGRT at three loca-
delivery to a phantom using multiple QA approaches and OSLDs tions (head, chest, and pelvis isocenters) and will calculate an optimal
inside key locations demonstrated that plans are deliverable and global shift based on the desired shifts at each location. Using this
dosimetry was accurate using this approach, including areas of approach, we guarantee that the distance between all isocenters
heterogeneity, such as the lungs and bone. Our plan uncertainty remains constant. The software calculates the residuals between the
analysis demonstrated that global shifts (same shift for all isocenters) global shift and the individual shifts desired at each location and will
126 | TERUEL ET AL.

T A B L E 3 Normalized dose deviation between treatment plan presented here has the potential to include options for additional
system dose and OLSD measured dose. Differences were normalized OARs sparing or simultaneous integrated boost (SIB) regimens beside
using the prescribed dose (165 cGy per fraction). the two current regimens. There is a clear interest in the radiation
TPS OSLD Normalized oncology community to provide solutions to some of the common
dose dose Absolute dose dose deviation problems of TBI, such as organ sparing, lack of or limited imaging
Location (cGy) (cGy) difference (cGy) (%)
capabilities, dosimetric uncertainties, the need to manufacture com-
Pelvic 185.0 184.6 −0.4 −0.3
pensators/blocks and patient comfort. Gruen et al.22 reported their
bone
initial use of Tomotherapy for VMAT-based TBI treatments on ten
Pelvic 183.3 186.4 3.1 1.9
patients treated to 12 Gy with 2 Gy per fraction. The lungs mean
bone
dose for this series was 9.14 Gy and no grade 3–4 toxicities were
Kidney R 123.4 116.9 −6.5 −3.9
observed. Springer et al.26 reported the use of VMAT-based TBI on
Kidney L 103.3 103.8 0.5 0.3
a linear accelerator on seven patients. None of the patients reported
Lung 87.9 83.1 −4.8 −2.9
severe lung toxicities and the authors were able to decrease the
Lung 66.5 63. 7 −2.8 −1.7
dose to the kidneys for patients with renal comorbidities to 7–8 Gy.
Lung 65.6 62.2 −3.4 −2.1
Tas et al.25 reported a VMAT-TBI technique used to treat 30
Lung 58.7 52.5 −6.2 −3.8 patients. The mean dose to the lungs was 9.7 Gy, and the mean
Vertebral 182.3 186.8 4.5 2.7 dose to the kidneys was 9.6 Gy. Grade 3 toxicity or higher was not
body
observed for any of the treated patients (mean follow-up of
Vertebral 182.6 187.6 5.0 3.0
18 months). Ouyang et al.27 reported no pulmonary toxicity for eight
body
patients treated with VMAT-based TBI on a linac with a mean dose
Brain 179.7 178.4 −1.3 −0.8
to the lungs of 8 Gy. They introduced a novel rotational immobiliza-
Brain 185.1 180.6 −4.5 −2.7
tion system for this type of treatment. Along those same lines, other
institutions are reporting their efforts to develop accurate immobi-
lization for these challenging treatments such as Mancosu et al.29 for
trigger an inspection of setup if established thresholds are not met. VMAT-based TMI treatments and Losert et al.30 for linac-based
Based on our plan uncertainty analysis, thresholds were established VMAT-TBI.
at 5 mm for the chest isocenter and 10 mm everywhere else. Among the different options available to build this automatic
The plans presented in this study were obtained with only one planning tool, we decided on fields collimated at 90° that could keep
optimization and calculation iteration in order to demonstrate that the X jaw field size at 15 cm except from the most superior arc to
an automated approach is feasible if properly constructed. However, allow for flash, and the most inferior arc to allow the optimizer to
it would be possible to manually adjust the optimization objectives properly overlap with the lower body 3D base plan. This allows the
based on experience or to reoptimize to improve the treatment plan. MLCs from each bank to reach the full extent of the field opening
This may actually be necessary if a particular case with a challenging giving the optimizer higher flexibility compared to scenarios where
anatomy is presented or if coverage/sparing in a particular area is MLCs cannot reach the end of the opposite bank due to larger field
desired. In any case, a workflow such as the one presented in this sizes. Two fields per isocenter with one field covering superiorly to
study should support standardization and robustness, and provide a the isocenter and a second one covering inferiorly (with a 2 cm
good starting point. Additionally, an automatic tool like the one overlap) proved useful to create MLC-based island blocks for the

F I G . 7 . Illustration demonstrating
physical location of OSLDs on the CIRS
anthropomorphic phantom and contours
placed on the treatment planning system
to obtain planned dose (purple contours)
for one slice. Four OSLDs located in the
lungs (plug number 71, 72, 76, and 78) and
one OSLD located in the vertebral body
(plug number 82). Treatment plan to
1320 cGy in 8 fractions.
TERUEL ET AL. | 127

F I G . 8 . Patient-specific QA analysis of one chest isocenter field. (a) ArcCheck: 99.3% passing rate at 2%/2mm gamma. (b) Varian Portal
Dosimetry: 97.7% passing rate at 2%/2mm gamma.

lungs and the kidneys. The script is able to adjust that overlap for advance the standardization, reproducibility, and automation of
the chest isocenter fields based on the distance between the chest treatment planning for VMAT-based TBI. Springer et al.25 provide a
isocenter and the user origin (Fig. 3). Additionally, for the purpose of significant advancement in the use of VMAT for TBI. However, the
this work we kept the overlap of fields not sharing an isocenter to technique for planning was variable among the seven patients with
5.3 cm. This should provide overlap of the diverging beams up to the number of isocenters between 9 and 15 and a different number
19 cm away from the isocenter and that sufficed to obtain good of fields, full and partial arcs and both longitudinal and lateral shifts.
coverage for the patient sample used here. However, that overlap Symons et al.24 described a robust technique for the upper body
can be increased if a challenging case requires it with the only draw- VMAT plan, but the combined lower body plan was not reported
back of an increased X jaw field size. We are currently updating the and the process required several sequential optimizations to obtain
script to directly increase the jaw field if required based on patient the final treatment plan. In our manuscript, we report a fully stan-
geometry for prospective cases. dardized and reproducible technique with the same number of
Our approach for setting up the automatic treatment planning isocenters, same shifts between upper body isocenters and consis-
used similar principles as the ones presented by some studies dis- tent overlap of fields and jaw settings to create MLC driven island
cussed earlier.24–26 The main contribution from our manuscript is to blocks. In addition, we have described the use of ESAPI to automate
128 | TERUEL ET AL.

F I G . 9 . Dose profile at an area of field

overlap between chest and abdomen
demonstrating linear and smooth
feathering between the two fields
contributing to the total dose in the area
with no abrupt change in field weight
within the overlap region.

most of the process so only an estimate of 10–15 min of manual procedure adds about 30–40 min for the first side, and another
tasks are required (5 min for OAR contouring plus 5–10 min for field 10 min to rotate the patient and verify setup for the opposed field.
weight adjustment of the lower body plan). Using the proposed VMAT-TBI approach the beam-on time per
Dose rate is a subject of controversy with regard to TBI. We treatment is about 10 min. While VMAT-TBI eliminates several com-
selected an energy of 6 MV photons and a nominal maximum dose rate ponents pertaining only to the conventional approach, the addition
of 600 MU/min for this work. For conventional treatments it is recom- of imaging, need for shifts and patient orientation shift are also
mended to keep the dose rate below 20 cGy/min or even 10 cGy/min time-consuming and have to be executed cautiously. Our initial
to minimize lung toxicities.2,10,31 However, there is not an accurate way experience through dry-runs and test of end-to-end process indi-
to compare the influence of dose rates between modulated and non- cates that a total treatment time between 1 hr and 1 hr and 15 min
modulated treatments. For open field at-distance treatments, the lungs should be expected for VMAT-TBI and that remains within the same
will receive the dose uniformly, while on modulated treatments there time range of the conventional approach.
are regions of the lungs receiving doses as low as 40 cGy per fraction Our study has several limitations. First, the CT datasets
(Fig. 7) with a gradual increase as we reach the external boundary of employed were acquired on a different position compared to the
the lungs. Also, the complete volume of the lungs is never irradiated at one expected for a real VMAT-based TBI program (currently under
the same time, and the mean dose can be decreased significantly. Cur- development at our institution). In the CT datasets used the patient
rently, there is no clear evidence as to how dose rates may affect mod- was positioned with forearms crossed over the chest, arms at the
ulated TBI treatments where the dose to the lungs is significantly lungs level, and lower extremity separation similar to pelvis or shoul-
decreased. Tas et al.25 reported in their study a mean instantaneous der separation. This positioning differs from our defined CT simula-
dose rate of 250 MU/min (range 50–600 MU/min) and did not encoun- tion positioning for VMAT-TBI with legs tighter together, arms at
ter any grade 3 or higher lung toxicity for any patient (n = 30). The dose patient side and thermoplastic immobilization. While this is a limita-
rate reported by Tas et al. is practically identical to the one presented in tion, our scripted treatment planning approach was able to provide
this study. Despite using a 600 MU/min maximum nominal dose rate, high-quality plans even using the less favorable positioning for
during delivery to our phantom the instantaneous dose rate was VMAT planning. Second, while the five cases explored here include
observed to range between 200 and 250 MU/min during most of the patients with diverse body habitus, this is a very limited number and
delivery, with some lows of 50–100 MU/min, and highs of up to patients with even larger anatomic variation might pose a challenge
400–500 MU/min. In spite of this, the automatic treatment planning at presentation. Finally, some information that the script uses for the
workflow presented in this study can be employed with a different field arrangement is specific to our institution. As an example, the
nominal dose rate or beam energy. script will always maximize the body region covered by the upper
After treatment is completed and signed-off the tool prints out a body VMAT plan. The longitudinal (craniocaudal) distance that is
summary report of the treatment. From our experience with conven- allowed is based on the indexing of the immobilization device to be
tional TBI using lateral beams, a total beam-on time of about 20 min employed at our institution and the longitudinal travel limit of the
is common. However, setup of the patient at distance, OSLD place- couch (with a margin to prevent reaching maximum travel). However,
ment, spoiler placement, compensation with solid water for the head, this does not conflict with the applicability of our field arrangement
compensation with rice bags between the legs and checklist to other immobilization or setup scenarios.
TERUEL ET AL. | 129

5 | CONCLUSIONS Paulina E. Galavis was involved in data analysis, interpretation of

data for the work, draft revision, final approval of the version to be
In our study, we have demonstrated that VMAT-based TBI treat- submitted, and responsible of accuracy and integrity of the work.
ment planning can be automated using scripting. Scripting can pro- K. Sunshine Osterman, Allison McCarthy, Martha Malin, and Naa-
duce high-quality plans based on predefined dose objectives and can mit K. Gerber were in involved in interpretation of data for the work,
decrease planning time by automatic target and optimization con- draft revision, final approval of the version to be submitted, and
tours generation, plan creation, field and isocenter generation, and responsible of accuracy and integrity of the work.
optimization objectives setup. Additionally, a robust and standardized Christine Hitchen was involved in conception and design of the
planning approach that accounts for couch longitudinal limits and work, interpretation of data for the work, drafting the work, final
immobilization facilitates treatment delivery ensuring consistent approval of the version to be submitted, and responsible of accuracy
shifts and isocenter placement. and integrity of the work.
David L. Barbee was involved in conception and design of the
work, acquisition of data, interpretation of data for the work, draft-
6 | PREVIOUS PUBLICATION OF ing the work, final approval of the version to be submitted, and
MANUSCRIPT TEXT OR DATA responsible of accuracy and integrity of the work.

Part of the material presented in this manuscript was previously sub- REFERENCES
mitted and presented at the 2019 AAPM annual meeting and the
1. Wong JYC, Filippi AR, Dabaja BS, Yahalom J, Specht L. Total body
2019 ASTRO annual meeting. The submitted abstracts are published
irradiation: guidelines from the international lymphoma radiation
in the conference proceedings: oncology group (ILROG). 2018.
2. ACR-ASTRO Committee. ACR–ASTRO practice parameter for the
1 . Teruel, J. R.; Taneja, S.; McCarthy, A.; Galavis, P.; Malin, M.; performance of total body irradiation (CSC/BOC). Am Coll Radiol.
Osterman, S.; Gerber, N. K.; Barbee, D.; Hitchen, C. Robust 2017;1076:1–11.
VMAT-based Total Body Irradiation (TBI) Treatment Planning 3. Paix A, Antoni D, Waissi W, et al. Total body irradiation in allogeneic
bone marrow transplantation conditioning regimens: a review. Crit
Assisted by Eclipse Scripting. International journal of radiation
Rev Oncol Hematol. 2018;123:138–148.
oncology biology physics. 2019:105(1):E788-E789.
4. Park H, Byun JM, Koh Y, et al. Comparison of different conditioning
2. Teruel, J; Taneja, S; Galavis, P; Osterman, K; Malin, M; Gerber, N; regimens in allogeneic hematopoietic stem-cell transplantation shows
Hitchen, C; Barbee, D. VMAT-based total body irradiation treat- superiority of total body irradiation-based regimen for younger
ment plans with eclipse scripting for field configuration: A dosi- patients with acute leukemia: a nationwide study. Clin Lymphoma,
Myeloma Leuk. 2019;19:e605–e615.
metric evaluation. Medical physics. 2019. https://doi.org/10.
5. Van Dyk J, Galvin JM, Glasgog JP, Podgorsak E. The physical aspects
1002/mp.13589. of total and half body photon irradiation. A report of task group 29
radiation therapy committee American Association of Physicists in
Medicine. American Association of Physicists in Medicine. 1986:
ACKNOWLEDGMENTS Report No. 17.
6. Kassaee A, Xiao Y, Bloch P, Goldwein J, Rosenthal DI, Bjärngard BE.
We thank Pine Cheng CMD, Maxwell Schulz CMD and Bhanumati Doses near the surface during total-body irradiation with 15 MV X-
Patel CMD (Department of Radiation Oncology, NYU Langone rays. Int J Cancer. 2001;96(S1):125.
Health) for their support developing the treatment planning 7. Park S-Y, Kim J-I, Joo YH, Lee JC, Park JM. Total body irradiation
with a compensator fabricated using a 3D optical scanner and a 3D
approach and testing the treatment planning script.
printer. Phys Med Biol. 2017;62:3735–3756.
8. Bailey DW, Wang IZ, Lakeman T, Hales LD, Singh AK, Podgorsak
MB. TBI lung dose comparisons using bilateral and anteroposterior
CONFLICT OF INTEREST
delivery techniques and tissue density corrections. J Appl Clin Med
Phys. 2015;16:291–301.
Honorarium from Varian Medical Systems (Jose Teruel).
9. Girinsky T, Benhamou E, Bourhis J-H, et al. Prospective randomized
comparison of single-dose versus hyperfractionated total-body irradi-
ation in patients with hematologic malignancies. J Clin Oncol.
AUTHOR CONTRIBUTION
2000;18:981–986.
10. Gao RW, Weisdorf DJ, DeFor TE, Ehler E, Dusenbery KE. Influence
Jose R. Teruel was involved in conception and design of the work,
of total body irradiation dose rate on idiopathic pneumonia syn-
acquisition of data, data analysis, interpretation of data for the work,
drome in acute leukemia patients undergoing allogeneic hematopoi-
drafting the work, final approval of the version to be submitted, and etic cell transplantation. Int J Radiat Oncol Biol Phys. 2019;103:
responsible of accuracy and integrity of the work. 180–189.
Sameer Taneja was involved in design of the work, acquisition of 11. Buchali A, Feyer P, Groll J, Massenkeil G, Arnold R, Budach V. Immedi-
ate toxicity during fractionated total body irradiation as conditioning
data, interpretation of data for the work, data analysis, draft revision,
for bone marrow transplantation. Radiother Oncol. 2000;54:157–162.
final approval of the version to be submitted, and responsible of 12. Gerstein J, Meyer A, Sykora K-W, Frühauf J, Karstens JH, Bremer
accuracy and integrity of the work. M. Long-term renal toxicity in children following fractionated total-
130 | TERUEL ET AL.

body irradiation (TBI) before allogeneic stem cell transplantation 23. Wong JYC, Rosenthal J, Liu A, Schultheiss T, Forman S, Somlo G.
(SCT). Strahlentherapie und Onkol. 2009;185:751. Image Guided Total Marrow Irradiation (TMI) Using Helical
13. Tauchmanovà L, Selleri C, Rosa GD, et al. High prevalence of endo- TomoTherapy in Patients with Multiple Myeloma and Acute Leuke-
crine dysfunction in long-term survivors after allogeneic bone mar- mia Undergoing Hematopoietic Cell Transplantation (HCT). Int J
row transplantation for hematologic diseases. Cancer. Radiat Oncol Biol Phys. 2009;73:273.
2002;95:1076–1084. 24. Symons K, Morrison C, Parry J, Woodings S, Zissiadis Y. Volumetric
14. Abboud I, Porcher R, Robin M, et al. Chronic kidney dysfunction in modulated arc therapy for total body irradiation: a feasibility study
patients alive without relapse 2 years after allogeneic hematopoietic using Pinnacle3 treatment planning system and Elekta AgilityTM linac.
stem cell transplantation. Biol Blood Marrow Transplant. J Appl Clin Med Phys. 2018;19:103–110.
2009;15:1251–1257. 25. Tas B, Durmus IF, Okumus A, et al. Total-body irradiation using
15. Benyunes MC, Sullivan KM, Joachim Deeg H, et al. Cataracts after linac-based volumetric modulated arc therapy: its clinical accuracy,
bone marrow transplantation: Long-term follow-up of adults treated feasibility and reliability. Radiother Oncol. 2018;129(3):527–533.
with fractionated total body irradiation. Int J Radiat Oncol. 26. Springer A, Hammer J, Winkler E, et al. Total body irradiation with
1995;32:661–670. volumetric modulated arc therapy: dosimetric data and first clinical
16. Bradley J, Reft C, Goldman S, et al. High-energy total body irradia- experience. Radiat Oncol. 2016;11:46.
tion as preparation for bone marrow transplantation in leukemia 27. Ouyang L, Folkerts M, Zhang Y, et al. Volumetric modulated arc
patients: treatment technique and related complications. Int J Radiat therapy based total body irradiation: workflow and clinical experi-
Oncol Biol Phys. 1998;40:391–396. ence with an indexed rotational immobilization system. Phys Imaging
17. Kal HB, van Kempen-Harteveld ML. Renal dysfunction after total Radiat Oncol. 2017;4:22–25.
body irradiation: Dose-effect relationship. Int J Radiat Oncol Biol 28. Blomain ES, Kovalchuk N, Neilsen BK, Skinner L, Hoppe RT, Hiniker
Phys. 2006;65:1228–1232. SM. A preliminary report of gonadal-sparing TBI using a VMAT tech-
18. Igaki H, Karasawa K, Sakamaki H, et al. Renal dysfunction after nique. Pract Radiat Oncol. 2020;S1879–8500:30199–30205.
total-body irradiation. Strahlentherapie und Onkol. 29. Mancosu P, Navarria P, Muren LP, et al. Development of an immobi-
2005;181:704–708. lization device for total marrow irradiation. Pract Radiat Oncol.
19. Lawton CA, Cohen EP, Murray KJ, et al. Long-term results of selec- 2021;11:e98–e105.
tive renal shielding in patients undergoing total body irradiation in 30. Losert C, Shpani R, Kießling R, et al. Novel rotatable tabletop for
preparation for bone marrow transplantation. Bone Marrow Trans- total-body irradiation using a linac-based VMAT technique. Radiat
plant. 1997;20:1069–1074. Oncol. 2019;14:244.
20. Sarradin V, Simon L, Huynh A, Gilhodes J, Filleron T, Izar F. Total 31. Ozsahin M, Pène F, Touboul E, et al. Total-body irradiation before
body irradiation using Helical Tomotherapy®: treatment technique, bone marrow transplantation. Results of two randomized instanta-
dosimetric results and initial clinical experience. Cancer/Radiotherapie. neous dose rates in 157 patients. Cancer. 1992;69:2853–2865.
2018;22:17–24.
21. Sun R, Cuenca X, Itti R, et al. First french experiences of total body
irradiations using helical TomoTherapy®. Cancer/Radiothérapie.
2017;21:365–372. SUPPORTING INFORMATION
22. Gruen A, Ebell W, Wlodarczyk W, et al. Total Body Irradiation (TBI)
Additional supporting information may be found online in the
using Helical Tomotherapy in children and young adults undergoing
stem cell transplantation. Radiat Oncol. 2013;8:92. Supporting Information section at the end of the article.