Extended SSD VMAT treatment for total body irradiation

Abstract In this work, we develop a total body irradiation technique that utilizes arc delivery, a buildup spoiler, and inverse optimized multileaf collimator (MLC) motion to shield organs at risk. The current treatment beam model is verified to confirm its applicability at extended source‐to‐surface distance (SSD). The delivery involves 7–8 volumetric modulated arc therapy arcs delivered to the patient in the supine and prone positions. The patient is positioned at a 90° couch angle on a custom bed with a 1 cm acrylic spoiler to increase surface dose. Single‐step optimization using a patient CT scan provides enhanced dose homogeneity and limits organ at risk dose. Dosimetric data of 109 TBI patients treated with this technique is presented along with the clinical workflow. Treatment planning system (TPS) verification measurements were performed at an extended SSD of 175 cm. Measurements included: a 4‐point absolute depth‐dose curve, profiles at 1.5, 5, and 10 cm depth, absolute point‐dose measurements of an treatment field, 2D Gafchromic® films at four locations, and measurements of surface dose at multiple locations of a Alderson phantom. The results of the patient DVH parameters were: Body‐5 mm D98 95.3 ± 1.5%, Body‐5 mm D2 114.0 ± 3.6%, MLD 102.8 ± 2.1%. Differences between measured and calculated absolute depth‐dose values were all <2%. Profiles at extended SSD had a maximum point difference of 1.3%. Gamma pass rates of 2D films were greater than 90% at 5%/1 mm. Surface dose measurements with film confirmed surface dose values of >90% of the prescription dose. In conclusion, the inverse optimized delivery method presented in the paper has been used to deliver homogenous dose to over 100 patients. The method provides superior patient comfort utilizing a commercial TPS. In addition, the ability to easily shield organs at risk is available through the use of MLCs.


| INTRODUCTION
Total body irradiation (TBI) is an essential part of bone marrow transplant conditioning as it has been shown to eliminate residual chemotherapy-resistant cancer cells and it provides additional immunosuppression to enhance engraftment. 1 TBI is combined with chemotherapy to enhance dose intensity of the preparative regimen while avoiding overlapping toxicity that may occur with high-dose multi-drug regimens without radiation. While the goal of TBI is to deliver a homogeneous dose of radiation to the body 1,2 there is currently no consensus on the optimal technique or fractionation to deliver the prescription dose most safely and effectively. [3][4][5][6] Conditioning regimens that use low dose TBI generally do not result in significant side effects; however lung shielding is used to maintain an even dose throughout the body.
A variety of TBI delivery techniques have been developed. Most of these techniques involve opposing beams that are either lateral or anterior-posterior (AP/PA) and are performed with the patient at an extended distance to limit the need for junctioning fields. 7,8 An example of a lateral technique is lateral parallel opposed pair (POP) beams with the patient under full bolus, which provides good dose homogeneity and utilizes simple dose calculation algorithms. 9,10 Drawbacks to this approach include reduced patient comfort and the limited ability to shield organs at risk without compromising dose to bones. Examples of AP/PA techniques include translating bed and multiple field techniques, such as the "mick" technique that uses multiple fields and boosts to achieve the required dose distribution. [11][12][13][14][15][16] AP/PA techniques provide the ability to shield organs at risk (usually through poured blocks) but other challenges exist. For instance, translating bed techniques require external beds custom designed to be under computer control with custom dosimetry calculations. In addition, multiple field techniques require accurate matching of field junctions in order to limit hot and cold regions during delivery to the patient.
Recently, there have been attempts to incorporate aspects of volumetrically modulated arc treatment (VMAT) into TBI treatments.
Kirby et al. 17 developed an inversely modulated pseudo-arc technique. Multiple static beams are delivered in an arc configuration using inverse planning to develop the monitor units (MU) for each.
The patient is treated in an AP/PA orientation with cerrobend blocks used to shield the lungs, and a hanging spoiler was used to increase the surface dose. Springer et al. 18 19 demonstrated an ultra-efficient modulated arc delivery method consisting of multiple consecutive modulated 5°s ubarcs in order to produce an homogenous dose. Jahnke et al. 20 recently proposed single modulated sweeping arc version of this technique where the patient is treated AP/PA at extended source-tosurface distance (SSD) and a sweeping arc covers their entire body.
The arc speed is varied to account for inverse square law (ISL) effects and to provide a homogenous dose. Organs at risk were shielded with poured blocks, and surface dose was increased with a beam spoiler. Jahnke et al. reported dose homogeneity of ±10% for a block phantom, however homogeneity was not considered for actual patients. In this work we build on this technique by modifying Jahnke et al.'s standard arcs to accommodate for SSD variation by using patient CT data. We also employ an inverse planned, single setup optimization from a commercial treatment planning system (TPS) to provide a VMAT solution (Eclipse ® ; Varian Medical Systems, Palo Alto, CA, USA) with the AAA algorithm that has been shown to perform well at extended SSD. 21,22 The optimization modifies the multi-leaf collimator (MLC) positions to shield organs at risk and to provide dose homogeneity throughout the body, producing a personalized, deliverable treatment plan that interfaces with the record and verify system. We report on our patient experience and the measurements made to verify the accuracy of our TPS's beam model at extended SSD.   Using the Eclipse ® TPS (Varian Medical Systems), an extended CT scan, equal to the patient's height, was created by replicating a slice from midthigh to the correct length ( Fig. 1). In addition to the Body contour, additional structures were created: Body retracted by 5 mm from skin (Body-5 mm), a Flash structure consisting of a 3 cm rind around the Body (Flash), both lungs in a single contour (Lungs), and a calculation volume including all of the structures (Calc Volume).

Standard plans
Custom VMAT arcs, ranging from 310°to 60°were created with a static 10 × 40 cm 2 opening (Fig. 2a). Plan meterset weights were altered using custom Python code to deliver more MUs at the periphery of the arc to account for the ISL effects. 20   Patient-specific QA is performed for one supine and one prone arc from all patient plans. This is done through delivery of an EPID on the treatment unit used for treatment. Plans are analyzed for 95% pass rates at 3%/3 mm.

2.B.3 | Treatment delivery
Treatment was delivered with the patient setup on a modified massage table capable of lowering the patient to a position of 175 cm SSD. The bed is setup perpendicular to the conventional couch and modified to have a 1 cm thick acrylic spoiler that traverses the entire length of the patient (Fig. 3). Both the spoiler and the bed can be raised and lowered for ease of patient transfer.
Typical treatment workflow is as follows: 1. The bed is setup at 90°and then centered and aligned with the in-room lasers. (Fig. 4) 2. The patient lies on the bed. With the gantry at 0°, the crosshair is aligned to the umbilicus, the patient is raised/lowered to the correct SSD, and the spoiler is lowered.

5.
The plan is delivered and diode readings are confirmed to match the dose predicted from the TPS for each arc.
6. The patient is then setup prone and the procedure is repeated.
The entire process takes approximately 45-60 minutes for the first fraction and less for subsequent fractions.

2.C | Shift analysis
To assess the plan robustness to setup uncertainties, the dosimetric effect of setup uncertainties was tested by systematically shifting the CT scan in crainial-caudal (CC), lateral (Lat) and anterior-posterior (AP) directions and recalculating the dose. The shifts were applied to both the prone and supine plans in the same direction to simulate the maximum error. Simulated shifts were +1 and +2 cm in the Lat and CC directions and ±1 and 2 cm in the AP direction (simulating SSD mismatch). Simulated setup uncertainties were chosen to be multiple times larger than uncertainties generally seen in patients that are setup with minimal immobilization. 17 Plans were compared to unshifted plans for dosimetric consequences. This was tested for 15 patients and the results of D98% coverage and D2% hot spots to the Body-5 mm was reported, as well as the change to MLD.

2.D | Treatment beam verification measurements
A series of measurements at extended SSD were performed to ensure that the TPS correctly models the machine output and beam profiles at these distances. These measurements are not typically performed at commissioning time and are required for extended SSD treatments.
The first measurement was made to verify the ability of the TPS to model the absolute output (cGy) as a function of depth at extended SSD for a fixed field. A 10 × 10 cm 2 field size was used to isolate the output measurement from field size effects. The full plan was delivered as per the protocol outline in the treatment delivery section.

3.A | Patient dosimetric results
Multiple slices from a patient CT scan with dose in color wash are shown in Fig. 5. Two dose profiles are shown, one down the midline and one across the lungs (Fig. 6). These profiles illustrate a homogenous dose within ±10% of the prescription coverage of the Body- Lungs contour is shown in Fig. 7. Results are presented as a percentage of the prescription dose in Fig. 8. The largest percentage of patients receives 400 cGy in two fractions (86/109), and their coverage was consistently over 95%. MLD was limited to approximately 100% of the prescription dose. The D2% hot spot was greater than our target of 110%, however the difference between the D98% and the D2% was consistently <20%.
Results of OSLD measurements from five patients for the standard eight OSLD sites (head [temple], belly button, back, foot and four leg positions) are shown in Table 2 with a comparison to a single fraction dose of 200 cGy. Our tolerance for these measurements is ±10% for the morning fractions. If this is exceeded, we would consider a change for the afternoon. To date, this has not occurred. An example of the patient specific QA is shown in Fig. 9. The pass rate at 3%/3 mm is 99%, which is typical for these plans. Of note is how flat the dose profile is, due to the small amount of modulation present.  Table 3.  2D film analysis of the representative AP patient arc results in gamma pass rates, at 5%/1 mm, with respect to corresponding superior-inferior locations, of 95.3% (+55 cm; head), 97.6% (+35 cm; lungs), 99.5% (0 cm; abdomen), and 91.5% (−55 cm; legs). A more in-depth examination of the TPS predicted dose and film dose from the section occurring in the phantom's "lungs" is shown in Fig. 12. Presented with the planar images of the predicted dose and measured film are profiles taken in the two directions of the film. The profile through the lateral direction (Fig. 12 c) shows excellent correspondence between the two scans. The crainial-Caudal profile (Fig. 12 d) displays the limitation of F I G . 6. (a) Dose profile across the chest as indicated in Fig. 5a; (b) dose profile from head to legs as indicated in Fig. 5(a). Both dose profiles exhibit the homogenous dose to the patient at a prescription of 400 cGy.

3.C | Treatment beam verification measurements
F I G . 7. Dose volume histogram for a representative patient. Both the target (Body-5 mm) and the Lungs are shown, and obtain the prescription dose (400 cGy) with a sharp falloff.
the TPS to calculate VMAT arcs at extended SSD-specifically the sinuous appearance created by calculating the dose only at controls points, not between control points. This pattern did not appear to greatly influence the pass rate of the film; however, it is likely contributing to the smaller pass rate seen in the leg film.
Results of the eight films placed on the surface of the Alderson phantom are shown in Table 5. Multiple ROI were taken to assess different regions. All doses were above 175 cGy (87.5% of prescription dose), confirming the SCF and the use of the spoiler to promote skin dose.

| DISCUSSION
In this work, we present a new TBI delivery technique that uses MLC shaping to accomplish dose homogeneity. The use of a modulated gantry speed addresses ISL changes for arc delivery, similar to the technique presented by Jahnke et al. 20 The optimizer from Eclipse ® is used to shape the MLC and adjust the MUs of each arc.
Using only the final step of the PRO we maintain the set meterset weights for each control point while altering the total MU per arc and the MLC positions. The standard beam model using the AAA 11.0.31 calculation algorithm is used to calculate the dose.
Employing a commercial beam optimizer ensures a plan's deliverability and permits the use of the record and verify system, which is an integral part of radiation therapy treatment delivery and not always available for TBI treatments.  The low dose TBI regimen used at our center does not necessitate significant shielding to organs at risk, like the lung or kidneys.
However, the use of higher doses often seen in TBI (eg 12 Gy in 6 fractions) would require significant shielding. We have tested our technique for a common TBI prescription of 12 Gy in six fractions, and have been able to achieve a lung dose as low as 6 Gy. The modified technique relied on manual manipulation of the MLCs and a slightly smaller field size (10 × 40 cm 2 ), but the results were promising.

| CONCLUSIONS
In this work, we presented a TBI technique that delivers at extended SSD and provides organ at risk shielding with minimal MLC modulation. The workflow is completed entirely using commercial treatment planning software, ensuring deliverability and consistency. The method has been implemented for over 100 patients at our center.
Dosimetric verification measurements were performed prior to technique implementation and showed that separate beam model data was not required. Data measured at extended SSD match predicted data from the TPS. This technique has been shown to be robust and patient sensitive, while provided a safe treatment that utilizes both the record and verify system and the commercial TPS.

ACKNOWLEDGMENTS
We acknowledge the entire radiation therapy team at the Tom Baker Cancer Centre for making this project a success.

CONF LICTS OF INTEREST
The authors have no relevant conflicts of interest to disclose.