FFF‐VMAT for SBRT of lung lesions: Improves dose coverage at tumor‐lung interface compared to flattened beams

Abstract Purpose To quantify the differences in dosimetry as a function of ipsilateral lung density and treatment delivery parameters for stereotactic, single dose of volumetric modulated arc therapy (VMAT) lung stereotactic body radiation therapy (SBRT) delivered with 6X flattening filter free (6X‐FFF) beams compared to traditional flattened 6X (6X‐FF) beams. Materials/methods Thirteen consecutive early stage I–II non‐small‐cell‐lung cancer (NSCLC) patients were treated with highly conformal noncoplanar VMAT SBRT plans (3–6 partial arcs) using 6X‐FFF beam and advanced Acuros‐based dose calculations to a prescription dose of 30 Gy in one fraction to the tumor margin. These clinical cases included relatively smaller tumor (island tumors) sizes (2.0–4.2 cm diameters) and varying average ipsilateral lung densities between 0.14 g/cc and 0.34 g/cc. Treatment plans were reoptimized with 6X‐FF beams for identical beam/arc geometries and planning objectives. For same target coverage, the organs‐at‐risk (OAR) dose metrics as a function of ipsilateral lung density were compared between 6X‐FFF and 6X‐FF plans. Moreover, monitor units (MU), beam modulation factor (MF) and beam‐on time (BOT) were evaluated. Results Both plans met the RTOG‐0915 protocol compliance. The ipsilateral lung density and the tumor location heavily influenced the treatment plans with 6X‐FFF and 6X‐FF beams, showing differences up to 12% for the gradient indices. For similar target coverage, 6X‐FFF beams showed better target conformity, lower intermediate dose‐spillage, and lower dose to the OAR. Additionally, BOT was reduced by a factor of 2.3 with 6X‐FFF beams compared to 6X‐FF beams. Conclusion While prescribing dose to the tumor periphery, 6X‐FFF VMAT plans for stereotactic single‐dose lung SBRT provided similar target coverage with better dose conformity, superior intermediate dose‐spillage (improved dose coverage at tumor interface), and improved OAR sparing compared to traditional 6X‐FF beams and significantly reduced treatment time. The ipsilateral lung density and tumor location considerably affected dose distributions requiring special attention for clinical SBRT plan optimization on a per‐patient basis. Clinical follow up of these patients for tumor local‐control rate and treatment‐related toxicities is in progress.


| INTRODUCTION
Due to the recent technological advances in lung stereotactic body radiation therapy (SBRT) treatments and reported comparable tumor local-control rates, 1-7 single-dose lung SBRT has become a viable treatment option for peripherally located lung lesions for medically inoperable early-stage nonsmall-cell lung cancer (NSCLC) patients. [8][9][10] Additionally, there has been growing interest in the clinical use of flattening filter-free (FFF) beams to deliver lung SBRT treatment. [11][12][13][14][15] FFF-beams have much higher dose rates compared to flattened beams and consequently reduce beam on time significantly. This results in better patient comfort (less time on the table), reduced dose delivery uncertainty due to intrafraction motion, and reduced out-of-field dose due to head scatter and electron contamination. 15 Combining FFF-beams with volumetric modulated arc therapy (VMAT) 16,17 results in even greater treatment efficiency for complex lung SBRT plans compared to historically used 8-15 noncoplanar fixed fields or several noncoplanar dynamic conformal arc (DCA) plans. Linac-based intensity modulated radiation therapy (IMRT), helical TomoTherapy or optimized robotic CyberKnife treatments significantly prolong SBRT treatment time, comparatively. [18][19][20][21] However, FFF-beams have different beam characteristics compared to flattened beams as mentioned earlier. 15 These include a nonuniform beam profile, reduced mean energy, and differing penumbra at depth. Given these distinct physical characteristics of FFF-beams, a few previous researchers have studied the dosimetric advantages of FFF-vs FF-beams in SBRT lung treatments. [11][12][13][14] The majority of the previous studies showed similar target coverage and clinically insignificant dose differences to the organs-at-risk (OAR) for both beam types, however, FFF-beams resulted in much faster treatment times.
While using flattened beams, rings of underdosing around the tumor and at lung tissue interfaces have been previously reported by Chetty et al. 22,23 In the most recent study by Vassiliev et al., 24 it was demonstrated that 6X-FFF beams can mitigate dose loss at the tumor-lung periphery due to low energy secondary electron dose buildup at the interface. Their experiment was conducted in a phantom comprised of a chest wall, lung tissue, and spherical tumors of 1, 3, and 5 cm diameters. Three lung densities of 0.1, 0.2, and 0.26 g/cc were considered. Treatment plans were generated using 7-coplanar static-fields with 6X-FFF and 6X-FF beams for a total dose of 50 Gy in five fractions prescribed to the tumor center.

2.A | Patient characteristics
After obtaining an institutional review board (IRB) approval from our institution, 13 peripherally located tumors in early stage I-II NSCLC patients who received a single dose of 30 Gy via lung SBRT were evaluated. Tumor characteristics are summarized in Table 1   guidelines. 8 The relevant critical structures such as bilateral lungs excluding the ITV (normal lung), spinal cord, ribs, heart, big vessels, esophagus, and skin were delineated on the 3D-CT images for dose reporting.
The average ipsilateral lung density was calculated using the following equation: ρ lung = ρ water (1.0 + avg. HU/1000). Average HU of the ipsilateral lung (excluding ITV) was obtained from the Eclipse TPS using ipsilateral lung contour for each patient.

2.C | Clinical 6X-FFF plans and treatment delivery
For each patient, highly conformal, clinically optimal VMAT SBRT plans were generated in Eclipse TPS using 3-6 (median, 4) partial The patient-specific quality assurance (QA) of these plans was performed by delivering VMAT SBRT plans on an Octavius phantom (PTW, Freiburg, Germany). All VMAT QA plans were delivered on Truebeam before the patient start date. The measured cumulative 2D dose plan was compared with the computed dose distributions calculated on the Octavius QA phantom plan in Eclipse TPS. Upon completion of dose delivery, data were analyzed with Octavius MEPHYSTO Navigator (VeriSoft Patient Plan Verification, Version 6.3, PTW) using the clinical gamma passing rate criteria of 3%/3mm maximum dose difference and distance-to-agreement (DTA) with a 10% threshold. The average VMAT QA pass rate was 97.6 ± 2.7%.
All patients were treated with CBCT-guided imaging. Patient setup prior to single-dose lung SBRT was performed using an in-house • Gradient index, GI: ratio of 50% prescription isodose volume to the PTV. GI has to be smaller than 3-6, depending on the PTV.
• Maximum dose at any point 2 cm away from the PTV margin in any direction, D 2cm : D 2cm has to be smaller than 50-70%, depending on the PTV size.
• Percentage of normal lung receiving dose equal to 20 Gy or more, V 20Gy : V 20Gy should be less than 10% per protocol, V 20Gy <15% is acceptable with minor deviations. V 20Gy is for total lungs minus the ITV.
• Heterogeneity index, HI: Dmax/prescribed dose was used to evaluate the dose heterogeneity within the PTV.
• Gradient distance, GD: GD is the average distance from 100% prescribed dose to 50% prescribed dose which indicates how sharp the dose falls off. The gradient distance (GD) is used to evaluate dose sparing to normal lung volume.  Table 2) compared to traditional 6X-FF plans. Statistically significant P-values are in bold (see Table 2). In general, mean dose to the ITV was similar between the plans, however, the maximum and minimum doses to the ITV were significantly higher (see P values in  12 In addition, the imbalance of the rates of secondary electron production in lung parenchyma and the tumor itself caused dose buildup differences at the tumor center. These appeared higher in lower lung densities (lung heterogeneities effect) with the 6X-FF beams (see Fig. 3).
For the same planning objectives, optimization parameters, and similar target coverage, very small, yet statistically significant differences were observed for all normal lung parameters including V 20Gy .
F I G . 1. Scatter plot: Ratio between 6X-FF and 6X-FFF plans for GI, D2cm, GD and HI as a function of average ipsilateral lung density for all 13 patients. Up to 12% higher values were observed for GI and GD with lower ipsilateral lung density with 6X-FF plans. In general, as the ipsilateral lung density decreases, the deviation in the gradient indices between the plans increases.
Comparison of dose distributions in the axial view for patient #6 (lowest average ipsilateral lung density of 0.14 gm/cc) with both 6X-FFF (right panel) and 6X-FF (left panel) beams. The red contour represents the ITV, orange represents the PTV (10.7 cc), green represents the 95% isodose line that encompasses the PTV and the blue line represents the 50% isodose-spillage (much tighter with 6X-FFF beam). The tumor was located in the middle of the right lung. The viewing plane intersection shows the isocenter location. The light blue ring was generated to calculate D 2cm around the target volume. ITV, Internal target volume; PTV, planning target volume.
These values were uniformly lower with clinical 6X-FFF beams compared to 6X-FF beams (see P-values in Table 3).
A comparison of other OAR dosimetric parameters for 6X-FFF and 6X-FF plans for all 13 lung SBRT patients is presented in Table 4.
Critical organs such as spinal cord (D max , and D 0.35cc ), heart (D max and D 15cc ), esophagus (D max and D 5cc ), ribs (D max and D 1cc ), and skin (D max and D 10cc ) were evaluated per SBRT protocol guidelines. It was observed that the volumetric dose differences to the heart, esophagus, and ribs were statistically significant (see P-values in Table 4) between the two plans. Overall, the doses with 6X-FF plans were higher by 1-15% for most of the critical organs, suggesting that the average values of absolute dose differences could be higher ITV, internal target volume; n. s., not significant; OAR, organs-at-risk; SBRT, stereotactic body radiation therapy. The statistical significance at P < 0.05 is bold.
with 6X-FF plans (of the order of 1.0 Gy) compared to clinical 6X-FFF plans. However, we predict the difference will not be clinically significant.

3.B | Treatment delivery parameters
For the given lung SBRT plan, the total number of MU did not change significantly while using 6X-FF vs 6X-FFF beams for plan optimization. This suggests that both plans had similar plan complexity providing similar MF.
However, the average BOT was 6.5 ± 1.5 min (range, 4.6-  Table 5). The BOT for 6X-FF vs 6X-FFF plans on a per-patient basis is also shown in Fig. 4.  Tables 2-5). The tighter dose distributions with 6X-FFF beams was due to its unique beam profile, softer energy spectrum, and smaller out-of-field scatter and leakage characteristics compared to traditional 6X-FF beams. This effect was much more prominent for the island tumors and increased as a function of lower ipsilateral lung density as shown in Fig. 2.

| DISCUSSION
However, insignificant differences between 6X-FFF and 6X-FF dose distributions (see Fig. 5 decreased the variation of intrafraction motion error due to coughing or pain making a geographic miss less likely and improving the patient stability as well as the clinic workflow.
In summary, each 6X-FFF and 6X-FF plan was rigorously evaluated using the dosimetric parameters listed in the Tables 2-5. All parameters were deemed acceptable per SBRT protocol suggesting that 6X-FFF plans are dosimetrically superior to 6X-FF plans. Furthermore, 6X-FFF plans would also deliver much faster lung SBRT treatments which would potentially improve patient compliance and clinic efficiency.
While evaluating the target coverage and OAR doses as a function of average ipsilateral lung density, it was observed that the island tumors with surrounding ipsilateral low lung density showed much higher variation between the modalities (6X-FFF vs 6X-FF plans) compared to lesions located near the chest wall with a higher ipsilateral lung density; suggesting that FFF-beam improves dose coverage at tumor-lung interface. Dose-limiting toxicity after hypofractionated dose-escalated radiotherapy in NSCLC patients is still an issue in lung SBRT treatment. 40 Utilizing 6X-FFF beams for VMAT SBRT lung planning may potentially reduce dose to OAR, help enhance dose to tumor peripheries, and deliver much faster treatments. However, while optimizing VMAT SBRT lung plans, planners are advised to pay special attention to the ipsilateral lung density and the tumor location on a per-patient basis as a function of beam modality.

| CONCLUSION
For our 30 Gy single-dose lung SBRT treatments, 6X-FFF plans showed dosimetrically superior isodose distributions, lower OAR doses, and much faster treatment deliveries compared to traditional 6X-FF plans.
Additionally, the isodose distributions were significantly affected by the ipsilateral lung density and tumor location as a function of beam modality. The dose enhancement at the tumor periphery was achieved by prescribing dose at the tumor margin (rather than prescribing dose at the tumor center) in addition to using 6X-FFF beams. Given that FFF-beam are already widely available in clinics, we hope this study will help in implementation of 6X-FFF beams for fast, effective, and safe treatment of NSCLC patients treated with SBRT. Clinical follow-up results of these single-dose SBRT lung patients are underway.

CONF LICT OF I NTEREST
None.