- Email: [email protected]
Dosimetric evaluation of MLC-based dynamic tumor tracking radiotherapy using digital phantom: Desired setup margin for tracking radiotherapy
ARTICLE IN PRESS Medical Dosimetry ■■ (2017) ■■–■■
Medical Dosimetry j o u r n a l h o m e p a g e : w w w. m e d d o s . o r g
Dosimetry Contribution:
Dosimetric evaluation of MLC-based dynamic tumor tracking radiotherapy using digital phantom: Desired setup margin for tracking radiotherapy Noriyuki Kadoya, Ph.D.,* Kei Ichiji, Ph.D.,† Tomoya Uchida, M.S.,‡ Yujiro Nakajima, M.S.,*,§ Ryutaro Ikeda, B.S.,* Yosuke Uozumi, Ph.D.,‡ Xiaoyong Zhang, Ph.D.,¶ Ivo Bukovsky, Ph.D.,‡,** Takaya Yamamoto, M.D., Ph.D.,* Ken Takeda, M.D., Ph.D.,† Yoshihiro Takai, M.D., Ph.D.,†† Keiichi Jingu, M.D., Ph.D.,* and Noriyasu Homma, Ph.D.‡ *Department of Radiation Oncology, Tohoku University Graduate School of Medicine, Sendai, Japan; †Department of Therapeutic Radiology, Tohoku University Graduate School of Medicine, Sendai, Japan; ‡Department of Radiological Imaging and Informatics, Tohoku University Graduate School of Medicine, Sendai, Japan; §Department of Radiotherapy, Tokyo Metropolitan Cancer and Infectious Diseases Center Komagome Hospital, Tokyo, Japan; ¶ Department of Electrical Engineering, Graduate School of Engineering, Tohoku University, Sendai, Japan; **Department of Instrumentation and Control Engineering, Faculty of Mechanical Engineering, Czech Technical University in Prague, Prague, Czech Republic; and ††Department of Radiation Oncology, Southern Tohoku BNCT Research Center, Koriyama, Japan
A R T I C L E
I N F O
Article history: Received 5 January 2017
Received in revised form 12 July 2017 Accepted 22 August 2017 Keywords:
Radiotherapy Dynamic tumor tracking radiotherapy Setup margin Four-dimensional dose calculation Stereotactic body radiotherapy
A B S T R A C T
The purpose of this study is to evaluate the dosimetric impact of the margin on the multileaf collimator-based dynamic tumor tracking plan. Furthermore, an equivalent setup margin (EM) of the tracking plan was determined according to the gated plan. A 4-dimensional extended cardiac-torso was used to create 9 digital phantom datasets of different tumor diameters (TDs) of 1, 3, and 5 cm and motion ranges (MRs) of 1, 2, and 3 cm. For each dataset, respiratory gating (30% to 70% phase) and tumor tracking treatment plans were prepared using 8-field 3-dimensional conformal radiation therapy by 4-dimensional dose calculation. The total lung V20 was calculated to evaluate the dosimetric impact for each case and to estimate the EM with the same impact on lung V20 obtained with the gating plan with a setup margin of 5 mm. The EMs for {TD = 1 cm, MR = 1 cm}, {TD = 1 cm, MR = 2 cm}, and {TD = 1 cm, MR = 3 cm} were estimated as 5.00, 4.16, and 4.24 mm, respectively. The EMs for {TD = 5 cm, MR = 1 cm}, {TD = 5 cm, MR = 2 cm}, and {TD = 5 cm, MR = 3 cm} were estimated as 4.24 mm, 6.35 mm, and 7.49 mm, respectively. This result showed that with a larger MR, the EM was found to be increased. In addition, with a larger TD, the EM became smaller. Our result showing the EMs provided the desired accuracy for multileaf collimator-based dynamic tumor tracking radiotherapy. © 2017 American Association of Medical Dosimetrists.
Introduction
Reprint requests to Noriyuki Kadoya, Ph.D., Department of Radiation Oncology, Tohoku University Graduate School of Medicine, 1-1 Seiryomachi, Aoba-ku, Sendai 980-8574, Japan. E-mail: [email protected]
Lung cancer is the leading cause of cancer-related mortality in Japan.1 At present, surgical resection remains the standard curative treatment of early stage non–small cell lung cancer (NSCLC).2 However, some patients are contraindicated
http://dx.doi.org/10.1016/j.meddos.2017.08.005 0958-3947/Copyright © 2017 American Association of Medical Dosimetrists
ARTICLE IN PRESS 2
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
for surgery because of advanced age or the presence of comorbidities, such as emphysema or heart disease. Thus, aggressive stereotactic body radiotherapy (SBRT) at a high hypofractionated dose plays a key role in the treatment of earlystage NSCLC. SBRT for the treatment of lung cancer results in high local control rates by focusing a high radiation dose on the target while sparing normal surrounding tissues.3-5 Lung toxicity continues to be a severe side effect of SBRT for early stage NSCLC. Fakiris et al. reported grades 3 to 5 toxicity in 17.1% of patients, and Timmerman et al. reported a rate of grade 3 toxicity of 12.7%.6,7 To reduce the extent of lung toxicity, several research groups have developed multileaf collimator (MLC)-based dynamic tumor tracking radiotherapy techniques.8-10 Our group has also been developing the MLC-based dynamic tumor tracking radiotherapy technique for the treatment of lung cancer.11,12 It is well known that dynamic tumor tracking radiation therapy can potentially reduce the internal margin without prolonging irradiation time.13-15 Saito et al. demonstrated that the gated delivery for 50% and 30% of the respiratory cycle required 20 seconds and 33.3 seconds for 100 monitor units (MU) using dose rate of 600 MU/min, compared with the nongated delivery with 10 seconds for 100 MU.15 In addition, Takao et al. showed that the incidence of baseline shift/ drift exceeding 3 mm was 14.0% for the craniocaudal (C-C) direction within 10 minutes of the start of treatment.16 This result suggested that gated radiotherapy required more treatment time due to intrafractional baseline shift or drift of lung tumor motion during gated radiotherapy in clinical care. Linear accelerator (Linac) has begun to use high dose rate of a flatting filter free mode (e.g., 2400 MU/min) for lung cancer patients in clinical practice.15,17,18 This technique can reduce the treatment time, compared with Linac with normal dose rate of a flatting filter mode (e.g., 600 MU/min). Thus, the high dose rate of the flatting filter free mode has the potential for solving the issue of prolonging irradiation time caused by the gated delivery. However, we used Linac beam with flatting filter in this study because at present, normal dose rate of the flatting filter mode was still commonly used in clinical practice in Japan. The MLC-based dynamic tumor tracking technique requires an extra margin because of the uncertainty of the tumor location and for the prediction of respiratory motion and beam repositioning. The dosimetric impact of MLCbased dynamic tumor tracking is dependent on the setup margin of the tracking plan. Thus, the main focus of this study was to estimate a setup margin equivalent to that for respiratory-gated radiotherapy using different setup margins with tumors of different tumor diameters (TDs) and motion ranges (MRs). An equivalent or smaller setup margin provides the desired accuracy for MLC-based dynamic tumortracking radiotherapy because it provides the dosimetric impact better than or comparable with the gated plan, with
a shorter irradiation time. An equivalent setup margin is expected to be helpful in further improving dynamic tumor tracking radiotherapy. For irradiation technique, in early-stage lung cancer for SBRT, 3-dimensional conformal radiotherapy (3D-CRT) has often been used in Japan.19,20 Furthermore, SBRT with 3DCRT has been focused on in light of the small effect of patient movement resulting from lack of modulated beam.21 Thus, we used 3D-CRT for MLC-based dynamic tracking planning in this study. The aim of the present study was to evaluate an equivalent setup margin of respiratory-gated radiotherapy for MLCbased dynamic tumor tracking radiotherapy using a digital phantom with different TDs and MRs using 3D-CRT.
Material and Methods Digital phantom Nine digital phantom datasets of different TDs of 1, 3, and 5 cm and MRs of 1, 2, and 3 cm were created using 4-dimensional extended cardiac-torso phantom (Duke University, NC).22 The voxel dimension was 1 mm (approximate) × 1 mm (approximate) × 2.5 mm. In this phantom, the tumor was set in the left lower lobe, as shown in Fig. 1. The 4D computed tomography (4D-CT) dataset consisted of 10 discrete respiratory phases (0% to 90% phase). In evaluating the equivalent setup margins for various TDs and MRs, digital phantom is a suitable tool because the TD and MR can be changed arbitrarily. If we use patient data for this study, TD and MR cannot be changed arbitrarily. Thus, we used digital phantom for this study.
Treatment planning To create a treatment plan for MLC-based dynamic tumor tracking radiotherapy and gated radiotherapy, 4D dose calculation with deformable image registration (DIR) was used to calculate the accurate irradiated dose to total lung. As the MLC-based dynamic tumor tracking plan had different 10 dose distributions on each phase (0% to 90%) with different MLC apertures, DIR is essential for summation of all 10 dose distributions, resulting in 1 dose distribution (i.e., 4D dose) on reference image. In the same manner, as gated plan had different 5 dose distributions on each phase (30% to 70%) with same MLC apertures, DIR is essential for summation of all 5 dose distributions. RayStation version 4.5.1 (RaySearch Laboratories, Stockholm, Sweden) was used as the treatment planning system combined with the collapsed cone dose calculation algorithm and a hybrid intensity and structure-based DIR algorithm (the Anatomically constrained deformation algorithm (ANACONDA) algorithm).23
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
3
Fig. 1. A representative digital phantom image. (Color version of figure is available online.)
RayStation can be used to focus on the structure or to control the region of interest (ROI) in the DIR algorithm. A previous study found that the DIR with a focus ROI had higher DIR accuracy than that by controlling the ROI.24 The lung structure was used as the ROI in this study. For the MLC-based dynamic tumor tracking plan, the concept was basically based on method reported by Keall et al.25 To perform MLC-based dynamic tumor tracking simulation to obtain the tracking MLC position, we used 10 images representing 10 phases of respiration. This tumor tracking plan using the 4D-CT (i.e., 10 images) is assumed to be the actual tracking plan with continuous MLC adaption of patient’s anatomy. The gross tumor volume (GTV) was created on the reference CT peak-exhale images (50% phase). The GTV was expanded by 5 to 10 mm in all directions to create the planning target volume (PTV) to compensate for the uncertainty of the tumor location and prediction of respiratory motion. We used 8-field 3D-CRT. Beam directions and weights were automatically selected to achieve optimal PTV coverage and minimal dosages to organs at risk using inverse-optimized 3D-CRT planning technique implanted in RayStation. The dose constraint was based on RTOG0618.26 The prescription dose was set to 20 Gy/fraction for a total of 3 fractions. All plans were normalized to 60 Gy prescribed to the PTV receiving 95% of the prescription dose. In the same manner, a treatment plan for each respiratory phase was created. After treatment planning was performed for each respiratory phase, the dose distribution for each CT phase was transformed back to the reference image
using DIR for evaluation of the entire treatment plan accounting for all respiratory phases (Fig. 2). For the gated plan, an internal target volume (ITV) was created using 5 phases of approximately 50% (i.e., ITV = GTV30% + GTV40% + GTV50% + GTV60% + GTV70%). The gating window level was used in our clinical protocol. Based on the previous paper reported by Saito et al., we determined the gating window width of 50% of the respiratory cycle as the clinical setting for gated radiotherapy, taking the trade-off between lower lung dose and increased treatment time into consideration.15 The ITV was expanded by 5 mm in all directions to create the PTV for setup uncertainty. Afterward, the gated plan on reference images was transferred to the other 4 phases (except for the 50% phase). Finally, the dose distribution for each CT phase was transformed back to the reference image using DIR, as in the tracking plan. In addition, an experienced medical physicist replaced the anatomic landmarks in peak-inhale and peak-exhale images, and the spatial DIR accuracy was evaluated by correspondence to the anatomic landmark.24,27
Estimation of equivalent setup margin for tracking The total lung volume receiving at least 20 Gy (V20) was calculated to evaluate the dosimetric impact for each case and to estimate an equivalent setup margin with the same impact on lung V20 obtained by the gated plan (Fig. 3). The lung V20 of the tracking plan as the function of the setup
ARTICLE IN PRESS 4
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
Table 1 summarizes the V20 of total lung for MLCbased tumor tracking plan and gated plan. This figure showed that the V20 of total lung increased with increasing the setup margin for MLC-based tumor tracking (e.g., TD = 5 cm, MR = 3 cm: 15.65% for 5 mm of setup margin vs 19.22% for 10 mm of setup margin). All cases with various TDs and MRs were similar results. An equivalent setup margin for different TDs and MRs are shown in Table 2 and Fig. 6. The equivalent setup margins for {TD = 1 cm, MR = 1 cm}, {TD = 1 cm, MR = 2 cm}, and {TD = 1 cm, MR = 3 cm} were estimated as 5.00, 4.16, and 4.24 mm, respectively. The equivalent setup margins for {TD = 5 cm, MR = 1 cm}, {TD = 5 cm, MR = 2 cm}, and {TD = 5 cm, MR = 3 cm} were estimated as 4.24 mm, 6.35 mm, and 7.49 mm, respectively. This result showed that with a larger MR, the equivalent setup margin was found to be increased. In addition, with a larger TD, the equivalent setup margin became smaller (e.g., for MR = 3 cm, 9.33 mm [TD = 1 cm] vs 7.49 mm [TD = 5 cm]). Figure 5 shows representative dose distributions for a gated plan and tracking plan with small, equivalent, and large setup margins. This figure shows that the tracking plan with a large setup margin had a higher irradiation area than the gated plan, and the tracking plan with equivalent setup margins had almost the same dose distribution as the gated plan. Fig. 2. Creation of MLC-based tracking plan using 4D dose calculation. (Color version of figure is available online.)
margin is shown in Fig. 4. In this figure, the red line shows the lung V20 of the gated plan. The equivalent setup margin was estimated by a quadratic function interpolation.
Results For DIR accuracy, the anatomic landmark-based target registration error (TRE) for DIR was calculated from peakinhale and peak-exhale in 1 extreme case with a TD of 5 cm and MR of 3 cm, which resulted in a TRE of 2.2 ± 2.1 mm (the number of landmarks = 100).
Discussion The dynamic tumor tracking radiotherapy has the potential to reduce the radiation dose to the total lung compared with that used for conventional radiotherapy. Although the dosimetric impact of a dynamic tracking plan is dependent on the setup margin for the tracking plan, the data showed that the setup margin was not equivalent to that of respiratory-gated radiotherapy, suggesting an equivalent setup margin for the tracking plan for different TDs and MRs, as compared with the gated plan with a gating width of 50% and setup margin of 5 mm. The equivalent setup margin ranged from 4.16 mm (TD = 3 cm, MR = 1 cm) to 9.33 mm (TD = 1 cm, MR = 3 cm).
Fig. 3. Concept of a setup margin for a tracking plan with the same impact on lung V20 obtained by the gated plan (30% to 70% phase) with a setup margin of 5 mm. (Color version of figure is available online.)
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
5
Fig. 4. Estimation of an equivalent setup margin for the tracking plan. The quadratic function was used to calculate the correlation between the setup margin and lung V20 of the tracking plan. Then, the lung V20 value of the gated plan was substituted for y in this equation to obtain an equivalent setup margin for the tracking plan. (Color version of figure is available online.)
Table 1 Summary of V20 of total lung for MLC-based tumor tracking plan with various setup margins and gated plan for various tumor diameters and motion ranges Tumor diameter
Motion range
MLC-based tumor tracking plan Setup margin 5 mm
Setup margin 6 mm
Setup margin 7 mm
Setup margin 8 mm
Setup margin 9 mm
Setup margin 10 mm
1 cm
1 cm 2 cm 3 cm 1 cm 2 cm 3 cm 1 cm 2 cm 3 cm
1.39 1.21 1.27 6.68 6.64 6.34 13.94 15.41 15.65
1.42 1.28 1.32 6.88 6.85 6.59 14.18 15.78 16.01
1.50 1.41 1.44 7.35 7.01 7.03 14.68 16.27 16.73
1.67 1.59 1.65 7.93 7.54 7.63 15.36 17.17 17.46
1.97 1.87 1.87 8.59 8.29 8.37 16.20 18.01 18.33
2.31 2.18 2.20 9.37 9.07 9.32 17.01 19.03 19.22
3 cm
5 cm
As shown in Fig. 6, as the MR increased, the equivalent setup margin also increased. In addition, as the TD increased, the equivalent setup margin decreased. As the equivalent setup margin increased, the effectiveness of dynamic tumor tracking radiotherapy compared with respiratory-gated radiotherapy also increased. That is, dynamic tumor tracking radiotherapy is suitable for tumors with a large MR and small TD. There were several papers focused on comparison between gated and nongated plan or between tracking and nongated plan for lung cancer patients.15,25,28-30 However, to our knowledge, there were no published data focused on
Table 2 Summary of equivalent setup margin for different tumor diameters and motion ranges Motion range
1 cm 2 cm 3 cm
Tumor diameter 1 cm
3 cm
5 cm
5.00 mm 7.14 mm 9.33 mm
4.16 mm 7.04 mm 8.00 mm
4.24 mm 6.35 mm 7.49 mm
Gated plan
1.32 1.43 1.98 6.51 7.09 7.62 13.75 15.97 17.06
comparison between tracking and gated plan for lung cancer. For liver cancer, Yoon et al.31 evaluated the dose difference between gated and tracking plan using CyberKnife for hepatocellular carcinoma. Gated plan used the ITV margin around the tumor determined by measuring its motion over 305% to 70% of respiratory phases using 4D-CT, followed by a 5-mm isotropic margin for PTV, and tracking plan excluded the ITV from PTV. They showed that CyberKnife tracking plan could significantly decrease the volume of normal liver tissue receiving > 15 Gy by 1.61-fold, compared with that in gated plan for 29 hepatocellular carcinoma patients with 14.5 ± 5.7 mm of average respiratory tumor motion and 10 to 20 mm of TD (65.5%). Our result suggested that tracking plan had better dose distribution than gated plan in these patients because the equivalent setup margin was more than 5 mm, which was used for tracking plan in their study. Thus, our result was similar to their study, although the treatment sites were different between our study and their study. The method to calculate the setup margin for the tracking plan is different from that for conventional radiotherapy. In conventional radiotherapy using an image-guided patient setup, the patient’s position is
ARTICLE IN PRESS 6
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
Fig. 5. Representative dose distributions for the gated plan and tracking plan with small, equivalent, and large setup margins (tumor diameter = 5 cm, motion range = 3 cm). (Color version of figure is available online.)
determined by kV or MV images before irradiation. Thus, the intrafractional motion error during dose delivering should be included. On the other hand, dynamic tumor tracking radiotherapy acquires images with the tumor position in real time, so there is no need to include the intrafractional motion error in the setup margin. However, dynamic tumor tracking radiotherapy should include the error related to tumor detection in real time and prediction for respiratory motion, instead of the intrafractional motion error. The equivalent setup margin estimated in this study can also clarify a tolerance of the tumor detection and prediction errors inherent in the dynamic tumor tracking techniques. In other words, by the techniques with such accuracy less than the equivalent setup margin, the dynamic tracking plan achieves a dosimetric impact better than the gated plan. In this sense, the results of this study can be used for defining a performance index to design and develop such techniques for the dynamic tumor tracking radiotherapy better than the gated one. In this study, we only focused on the lung V20 because lung V20 is one of the most commonly used predictors for lung toxicity. However, other DVH parameters for lung also
import predictors for lung toxicity, such as V5, V10, and mean dose. Thus, further investigation study will be needed to clarify the dosimetric and clinical impact of tracking plan using other DVH parameters. There were some limitations to this study. First, only 1 situation was evaluated, in which the tumor was switched to the right lower lobe. With this tumor location, it is assumed that the tumor has the highest respiratory MR. Thus, only this situation was analyzed. However, if the tumor is located in a different lobe, this result may change. Second, in this study, only 1 gating window was employed for the gated plan (i.e., 30% to 70% phase). This gating window is commonly used in clinical practice.32 However, several studies investigated smaller gating windows for gated plans,33,34 and changes to the gating window of the internal margin for the gated plan resulted in a different equivalent setup margin for the tracking plan.35 Thus, further studies are needed to further clarify a relation between gating windows for the gated plan and equivalent setup margins for the tracking plan. Next, our study focused on only 1 direction for tumor motion (i.e., C-C direction). In clinical case, tumor motion is more complex.
Fig. 6. Correlations between motion range and equivalent setup margin (A) and between tumor diameter and equivalent setup margin (B). The equivalent setup margin increases proportionally with the motion range. In addition, as the diameter increases, the equivalent setup margin decreases. (Color version of figure is available online.)
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
However, almost all tumors had the largest respiratory movement in C-C direction.36 Thus, although the setting used in our study could not simulate the clinical situation perfectly, our result may be helpful for further improving MLCbased dynamic tumor tracking radiotherapy. Then, we used 4D dose calculation with DIR. Thus, our result includes the residual DIR error. The residual DIR error (i.e., TRE within 2 mm) was smaller than the image voxel size in thoracic regions (i.e., 2 mm). In addition, in our study, V20 of total lung was used for calculation of equivalent setup margin. The volume metrics, such as V20, has less influence from the residual DIR error, compared with point metrics, such as maximum and minimum dose for PTV. Thus, the influence was expected to be small (or negligible). Finally, to perform MLC-based dynamic tracking simulation to obtain the tracking MLC position, we used only 10 images representing 10 phases of respiration. As a result, the tracking MLC position could not continuously adapt the patient’s anatomy (i.e., shape of tumor), which may yield the slightly different result for MLC-based tumor tracking plan. Conclusions The results of this study showed that an equivalent setup margin changes in response to TD and MR. The tracking plan with a setup margin smaller than the equivalent setup margin achieves a better dosimetric impact than the gated plan. Conflict of interest There is no conflict of interest with regard to this manuscript. Acknowledgements This work was partially supported by Varian Medical Systems (Palo Alto, CA, USA) and JSPS KAKENHI Grant No. 25293258, No. 15J05402 and No. 15K19765. References 1. National Cancer Center. Cancer statistics in Japan-2014. Available at: http://ganjoho.jp/reg_stat/statistics/stat/summary.html. Accessed August 1, 2016. 2. McCann, J.; Artinian, V.; Duhaime, L.; et al. Evaluation of the causes for racial disparity in surgical treatment of early stage lung cancer. Chest 128:3440–6; 2005. 3. Onishi, H.; Shirato, H.; Nagata, Y.; et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for stage I non-small cell lung cancer: updated results of 257 patients in a Japanese multi-institutional study. J. Thorac. Oncol. 2:S94–100; 2007. 4. Koto, M.; Takai, Y.; Ogawa, Y.; et al. A phase II study on stereotactic body radiotherapy for stage I non-small cell lung cancer. Radiother. Oncol. 85:429–34; 2007. 5. Bush, D.A.; Slater, J.D.; Shin, B.B.; et al. Hypofractionated proton beam radiotherapy for stage I lung cancer. Chest 126:1198–203; 2004.
7
6. Fakiris, A.J.; McGarry, R.C.; Yiannoutsos, C.T.; et al. Stereotactic body radiation therapy for early-stage non-small-cell lung carcinoma: four-year results of a prospective phase II study. Int. J. Radiat. Oncol. Biol. Phys. 75:677–82; 2009. 7. Timmerman, R.; Paulus, R.; Galvin, J.; et al. Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA 303:1070–6; 2010. 8. Keall, P.J.; Todor, A.D.; Vedam, S.S.; et al. On the use of EPID-based implanted marker tracking for 4D radiotherapy. Med. Phys. 31:3492– 9; 2004. 9. Poulsen, P.R.; Cho, B.; Sawant, A.; et al. Implementation of a new method for dynamic multileaf collimator tracking of prostate motion in arc radiotherapy using a single kV imager. Int. J. Radiat. Oncol. Biol. Phys. 76:914–23; 2010. 10. Moore, D.; Ruan, D.; Sawant, A. Fast leaf-fitting with generalized underdose/overdose constraints for real-time MLC tracking. Med. Phys. 43:465; 2016. 11. Zhang, X.; Homma, N.; Ichiji, K.; et al. A kernel-based method for markerless tumor tracking in kV fluoroscopic images. Phys. Med. Biol. 59:4897–911; 2014. 12. Zhang, X.; Homma, N.; Ichiji, K.; et al. Tracking tumor boundary in MV-EPID images without implanted markers: a feasibility study. Med. Phys. 42:2510; 2015. 13. Keall, P. 4-Dimensional computed tomography imaging and treatment planning. Semin. Radiat. Oncol. 14:81–90; 2004. 14. Colvill, E.; Booth, J.T.; O’Brien, R.T.; et al. Multileaf collimator tracking improves dose delivery for prostate cancer radiation therapy: results of the first clinical trial. Int. J. Radiat. Oncol. Biol. Phys. 92:1141–7; 2015. 15. Saito, T.; Sakamoto, T.; Oya, N. Effect of gating window width on lung dosimetric parameters and on treatment time at gated radiotherapy for lung cancer. Acta Oncol. 50:1246–9; 2011. 16. Takao, S.; Miyamoto, N.; Matsuura, T.; et al. Intrafractional baseline shift or drift of lung tumor motion during gated radiation therapy with a real-time tumor-tracking system. Int. J. Radiat. Oncol. Biol. Phys. 94:172–80; 2016. 17. Lang, S.; Shrestha, B.; Graydon, S.; et al. Clinical application of flattening filter free beams for extracranial stereotactic radiotherapy. Radiother. Oncol. 106:255–9; 2013. 18. Navarria, P.; Ascolese, A.M.; Mancosu, P.; et al. Volumetric modulated arc therapy with flattening filter free (FFF) beams for stereotactic body radiation therapy (SBRT) in patients with medically inoperable early stage non small cell lung cancer (NSCLC). Radiother. Oncol. 107:414–8; 2013. 19. Onishi, H.; Araki, T.; Shirato, H.; et al. Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanese multiinstitutional study. Cancer 101:1623–31; 2004. 20. Onishi, H.; Shirato, H.; Nagata, Y.; et al. Stereotactic body radiotherapy (SBRT) for operable stage I non-small-cell lung cancer: can SBRT be comparable to surgery? Int. J. Radiat. Oncol. Biol. Phys. 81:1352–8; 2011. 21. Cai, J.; Malhotra, H.K.; Orton, C.G. A 3D-conformal technique is better than IMRT or VMAT for lung SBRT. Med. Phys. 41:040601; 2014. 22. Segars, W.P.; Sturgeon, G.; Mendonca, S.; et al. 4D XCAT phantom for multimodality imaging research. Med. Phys. 37:4902–15; 2010. 23. Weistrand, O.; Svensson, S. The ANACONDA algorithm for deformable image registration in radiotherapy. Med. Phys. 42:40–53; 2015. 24. Kadoya, N.; Nakajima, Y.; Saito, M.; et al. Multi-institutional validation study of commercially available deformable image registration software for thoracic images. Int. J. Radiat. Oncol. Biol. Phys. 96(2): 422–31; 2016. 25. Keall, P.J.; Joshi, S.; Vedam, S.S.; et al. Four-dimensional radiotherapy planning for DMLC-based respiratory motion tracking. Med. Phys. 32:942; 2005.
ARTICLE IN PRESS 8
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
26. Timmerman, R.; Galvin, J.; Michalski, J.; et al. Accreditation and quality assurance for Radiation Therapy Oncology Group: multicenter clinical trials using stereotactic body radiation therapy in lung cancer. Acta Oncol. 45:779–86; 2006. 27. Kadoya, N.; Fujita, Y.; Katsuta, Y.; et al. Evaluation of various deformable image registration algorithms for thoracic images. J Radiat Res 55:175–82; 2014. 28. Saito, T.; Sakamoto, T.; Oya, N. Comparison of gating around endexpiration and end-inspiration in radiotherapy for lung cancer. Radiother. Oncol. 93:430–5; 2009. 29. Underberg, R.W.; van Sornsen de Koste, J.R.; Lagerwaard, F.J.; et al. A dosimetric analysis of respiration-gated radiotherapy in patients with stage III lung cancer. Radiat. Oncol. 1:8; 2006. 30. Matsuo, Y.; Ueki, N.; Takayama, K.; et al. Evaluation of dynamic tumour tracking radiotherapy with real-time monitoring for lung tumours using a gimbal mounted linac. Radiother. Oncol. 112:360–4; 2014. 31. Yoon, K.; Kwak, J.; Cho, B.; et al. Gated volumetric-modulated arc therapy vs. tumor-tracking cyberKnife radiotherapy as stereotactic
32.
33.
34.
35.
36.
body radiotherapy for hepatocellular carcinoma: A dosimetric comparison study focused on the impact of respiratory motion managements. PLoS ONE 11:e0166927; 2016. Jang, S.S.; Huh, G.J.; Park, S.Y.; et al. The impact of respiratory gating on lung dosimetry in stereotactic body radiotherapy for lung cancer. Phys. Med. 30:682–9; 2014. Chin, E.; Loewen, S.K.; Nichol, A.; et al. 4D VMAT, gated VMAT, and 3D VMAT for stereotactic body radiation therapy in lung. Phys. Med. Biol. 58:749–70; 2013. Wolthaus, J.W.; Sonke, J.J.; van Herk, M.; et al. Comparison of different strategies to use four-dimensional computed tomography in treatment planning for lung cancer patients. Int. J. Radiat. Oncol. Biol. Phys. 70:1229–38; 2008. Cai, J.; McLawhorn, R.; Read, P.W.; et al. Effects of breathing variation on gating window internal target volume in respiratory gated radiation therapy. Med. Phys. 37:3927–34; 2010. Liu, H.H.; Balter, P.; Tutt, T.; et al. Assessing respiration-induced tumor motion and internal target volume using four-dimensional computed tomography for radiotherapy of lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 68:531–40; 2007.
Medical Dosimetry j o u r n a l h o m e p a g e : w w w. m e d d o s . o r g
Dosimetry Contribution:
Dosimetric evaluation of MLC-based dynamic tumor tracking radiotherapy using digital phantom: Desired setup margin for tracking radiotherapy Noriyuki Kadoya, Ph.D.,* Kei Ichiji, Ph.D.,† Tomoya Uchida, M.S.,‡ Yujiro Nakajima, M.S.,*,§ Ryutaro Ikeda, B.S.,* Yosuke Uozumi, Ph.D.,‡ Xiaoyong Zhang, Ph.D.,¶ Ivo Bukovsky, Ph.D.,‡,** Takaya Yamamoto, M.D., Ph.D.,* Ken Takeda, M.D., Ph.D.,† Yoshihiro Takai, M.D., Ph.D.,†† Keiichi Jingu, M.D., Ph.D.,* and Noriyasu Homma, Ph.D.‡ *Department of Radiation Oncology, Tohoku University Graduate School of Medicine, Sendai, Japan; †Department of Therapeutic Radiology, Tohoku University Graduate School of Medicine, Sendai, Japan; ‡Department of Radiological Imaging and Informatics, Tohoku University Graduate School of Medicine, Sendai, Japan; §Department of Radiotherapy, Tokyo Metropolitan Cancer and Infectious Diseases Center Komagome Hospital, Tokyo, Japan; ¶ Department of Electrical Engineering, Graduate School of Engineering, Tohoku University, Sendai, Japan; **Department of Instrumentation and Control Engineering, Faculty of Mechanical Engineering, Czech Technical University in Prague, Prague, Czech Republic; and ††Department of Radiation Oncology, Southern Tohoku BNCT Research Center, Koriyama, Japan
A R T I C L E
I N F O
Article history: Received 5 January 2017
Received in revised form 12 July 2017 Accepted 22 August 2017 Keywords:
Radiotherapy Dynamic tumor tracking radiotherapy Setup margin Four-dimensional dose calculation Stereotactic body radiotherapy
A B S T R A C T
The purpose of this study is to evaluate the dosimetric impact of the margin on the multileaf collimator-based dynamic tumor tracking plan. Furthermore, an equivalent setup margin (EM) of the tracking plan was determined according to the gated plan. A 4-dimensional extended cardiac-torso was used to create 9 digital phantom datasets of different tumor diameters (TDs) of 1, 3, and 5 cm and motion ranges (MRs) of 1, 2, and 3 cm. For each dataset, respiratory gating (30% to 70% phase) and tumor tracking treatment plans were prepared using 8-field 3-dimensional conformal radiation therapy by 4-dimensional dose calculation. The total lung V20 was calculated to evaluate the dosimetric impact for each case and to estimate the EM with the same impact on lung V20 obtained with the gating plan with a setup margin of 5 mm. The EMs for {TD = 1 cm, MR = 1 cm}, {TD = 1 cm, MR = 2 cm}, and {TD = 1 cm, MR = 3 cm} were estimated as 5.00, 4.16, and 4.24 mm, respectively. The EMs for {TD = 5 cm, MR = 1 cm}, {TD = 5 cm, MR = 2 cm}, and {TD = 5 cm, MR = 3 cm} were estimated as 4.24 mm, 6.35 mm, and 7.49 mm, respectively. This result showed that with a larger MR, the EM was found to be increased. In addition, with a larger TD, the EM became smaller. Our result showing the EMs provided the desired accuracy for multileaf collimator-based dynamic tumor tracking radiotherapy. © 2017 American Association of Medical Dosimetrists.
Introduction
Reprint requests to Noriyuki Kadoya, Ph.D., Department of Radiation Oncology, Tohoku University Graduate School of Medicine, 1-1 Seiryomachi, Aoba-ku, Sendai 980-8574, Japan. E-mail: [email protected]
Lung cancer is the leading cause of cancer-related mortality in Japan.1 At present, surgical resection remains the standard curative treatment of early stage non–small cell lung cancer (NSCLC).2 However, some patients are contraindicated
http://dx.doi.org/10.1016/j.meddos.2017.08.005 0958-3947/Copyright © 2017 American Association of Medical Dosimetrists
ARTICLE IN PRESS 2
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
for surgery because of advanced age or the presence of comorbidities, such as emphysema or heart disease. Thus, aggressive stereotactic body radiotherapy (SBRT) at a high hypofractionated dose plays a key role in the treatment of earlystage NSCLC. SBRT for the treatment of lung cancer results in high local control rates by focusing a high radiation dose on the target while sparing normal surrounding tissues.3-5 Lung toxicity continues to be a severe side effect of SBRT for early stage NSCLC. Fakiris et al. reported grades 3 to 5 toxicity in 17.1% of patients, and Timmerman et al. reported a rate of grade 3 toxicity of 12.7%.6,7 To reduce the extent of lung toxicity, several research groups have developed multileaf collimator (MLC)-based dynamic tumor tracking radiotherapy techniques.8-10 Our group has also been developing the MLC-based dynamic tumor tracking radiotherapy technique for the treatment of lung cancer.11,12 It is well known that dynamic tumor tracking radiation therapy can potentially reduce the internal margin without prolonging irradiation time.13-15 Saito et al. demonstrated that the gated delivery for 50% and 30% of the respiratory cycle required 20 seconds and 33.3 seconds for 100 monitor units (MU) using dose rate of 600 MU/min, compared with the nongated delivery with 10 seconds for 100 MU.15 In addition, Takao et al. showed that the incidence of baseline shift/ drift exceeding 3 mm was 14.0% for the craniocaudal (C-C) direction within 10 minutes of the start of treatment.16 This result suggested that gated radiotherapy required more treatment time due to intrafractional baseline shift or drift of lung tumor motion during gated radiotherapy in clinical care. Linear accelerator (Linac) has begun to use high dose rate of a flatting filter free mode (e.g., 2400 MU/min) for lung cancer patients in clinical practice.15,17,18 This technique can reduce the treatment time, compared with Linac with normal dose rate of a flatting filter mode (e.g., 600 MU/min). Thus, the high dose rate of the flatting filter free mode has the potential for solving the issue of prolonging irradiation time caused by the gated delivery. However, we used Linac beam with flatting filter in this study because at present, normal dose rate of the flatting filter mode was still commonly used in clinical practice in Japan. The MLC-based dynamic tumor tracking technique requires an extra margin because of the uncertainty of the tumor location and for the prediction of respiratory motion and beam repositioning. The dosimetric impact of MLCbased dynamic tumor tracking is dependent on the setup margin of the tracking plan. Thus, the main focus of this study was to estimate a setup margin equivalent to that for respiratory-gated radiotherapy using different setup margins with tumors of different tumor diameters (TDs) and motion ranges (MRs). An equivalent or smaller setup margin provides the desired accuracy for MLC-based dynamic tumortracking radiotherapy because it provides the dosimetric impact better than or comparable with the gated plan, with
a shorter irradiation time. An equivalent setup margin is expected to be helpful in further improving dynamic tumor tracking radiotherapy. For irradiation technique, in early-stage lung cancer for SBRT, 3-dimensional conformal radiotherapy (3D-CRT) has often been used in Japan.19,20 Furthermore, SBRT with 3DCRT has been focused on in light of the small effect of patient movement resulting from lack of modulated beam.21 Thus, we used 3D-CRT for MLC-based dynamic tracking planning in this study. The aim of the present study was to evaluate an equivalent setup margin of respiratory-gated radiotherapy for MLCbased dynamic tumor tracking radiotherapy using a digital phantom with different TDs and MRs using 3D-CRT.
Material and Methods Digital phantom Nine digital phantom datasets of different TDs of 1, 3, and 5 cm and MRs of 1, 2, and 3 cm were created using 4-dimensional extended cardiac-torso phantom (Duke University, NC).22 The voxel dimension was 1 mm (approximate) × 1 mm (approximate) × 2.5 mm. In this phantom, the tumor was set in the left lower lobe, as shown in Fig. 1. The 4D computed tomography (4D-CT) dataset consisted of 10 discrete respiratory phases (0% to 90% phase). In evaluating the equivalent setup margins for various TDs and MRs, digital phantom is a suitable tool because the TD and MR can be changed arbitrarily. If we use patient data for this study, TD and MR cannot be changed arbitrarily. Thus, we used digital phantom for this study.
Treatment planning To create a treatment plan for MLC-based dynamic tumor tracking radiotherapy and gated radiotherapy, 4D dose calculation with deformable image registration (DIR) was used to calculate the accurate irradiated dose to total lung. As the MLC-based dynamic tumor tracking plan had different 10 dose distributions on each phase (0% to 90%) with different MLC apertures, DIR is essential for summation of all 10 dose distributions, resulting in 1 dose distribution (i.e., 4D dose) on reference image. In the same manner, as gated plan had different 5 dose distributions on each phase (30% to 70%) with same MLC apertures, DIR is essential for summation of all 5 dose distributions. RayStation version 4.5.1 (RaySearch Laboratories, Stockholm, Sweden) was used as the treatment planning system combined with the collapsed cone dose calculation algorithm and a hybrid intensity and structure-based DIR algorithm (the Anatomically constrained deformation algorithm (ANACONDA) algorithm).23
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
3
Fig. 1. A representative digital phantom image. (Color version of figure is available online.)
RayStation can be used to focus on the structure or to control the region of interest (ROI) in the DIR algorithm. A previous study found that the DIR with a focus ROI had higher DIR accuracy than that by controlling the ROI.24 The lung structure was used as the ROI in this study. For the MLC-based dynamic tumor tracking plan, the concept was basically based on method reported by Keall et al.25 To perform MLC-based dynamic tumor tracking simulation to obtain the tracking MLC position, we used 10 images representing 10 phases of respiration. This tumor tracking plan using the 4D-CT (i.e., 10 images) is assumed to be the actual tracking plan with continuous MLC adaption of patient’s anatomy. The gross tumor volume (GTV) was created on the reference CT peak-exhale images (50% phase). The GTV was expanded by 5 to 10 mm in all directions to create the planning target volume (PTV) to compensate for the uncertainty of the tumor location and prediction of respiratory motion. We used 8-field 3D-CRT. Beam directions and weights were automatically selected to achieve optimal PTV coverage and minimal dosages to organs at risk using inverse-optimized 3D-CRT planning technique implanted in RayStation. The dose constraint was based on RTOG0618.26 The prescription dose was set to 20 Gy/fraction for a total of 3 fractions. All plans were normalized to 60 Gy prescribed to the PTV receiving 95% of the prescription dose. In the same manner, a treatment plan for each respiratory phase was created. After treatment planning was performed for each respiratory phase, the dose distribution for each CT phase was transformed back to the reference image
using DIR for evaluation of the entire treatment plan accounting for all respiratory phases (Fig. 2). For the gated plan, an internal target volume (ITV) was created using 5 phases of approximately 50% (i.e., ITV = GTV30% + GTV40% + GTV50% + GTV60% + GTV70%). The gating window level was used in our clinical protocol. Based on the previous paper reported by Saito et al., we determined the gating window width of 50% of the respiratory cycle as the clinical setting for gated radiotherapy, taking the trade-off between lower lung dose and increased treatment time into consideration.15 The ITV was expanded by 5 mm in all directions to create the PTV for setup uncertainty. Afterward, the gated plan on reference images was transferred to the other 4 phases (except for the 50% phase). Finally, the dose distribution for each CT phase was transformed back to the reference image using DIR, as in the tracking plan. In addition, an experienced medical physicist replaced the anatomic landmarks in peak-inhale and peak-exhale images, and the spatial DIR accuracy was evaluated by correspondence to the anatomic landmark.24,27
Estimation of equivalent setup margin for tracking The total lung volume receiving at least 20 Gy (V20) was calculated to evaluate the dosimetric impact for each case and to estimate an equivalent setup margin with the same impact on lung V20 obtained by the gated plan (Fig. 3). The lung V20 of the tracking plan as the function of the setup
ARTICLE IN PRESS 4
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
Table 1 summarizes the V20 of total lung for MLCbased tumor tracking plan and gated plan. This figure showed that the V20 of total lung increased with increasing the setup margin for MLC-based tumor tracking (e.g., TD = 5 cm, MR = 3 cm: 15.65% for 5 mm of setup margin vs 19.22% for 10 mm of setup margin). All cases with various TDs and MRs were similar results. An equivalent setup margin for different TDs and MRs are shown in Table 2 and Fig. 6. The equivalent setup margins for {TD = 1 cm, MR = 1 cm}, {TD = 1 cm, MR = 2 cm}, and {TD = 1 cm, MR = 3 cm} were estimated as 5.00, 4.16, and 4.24 mm, respectively. The equivalent setup margins for {TD = 5 cm, MR = 1 cm}, {TD = 5 cm, MR = 2 cm}, and {TD = 5 cm, MR = 3 cm} were estimated as 4.24 mm, 6.35 mm, and 7.49 mm, respectively. This result showed that with a larger MR, the equivalent setup margin was found to be increased. In addition, with a larger TD, the equivalent setup margin became smaller (e.g., for MR = 3 cm, 9.33 mm [TD = 1 cm] vs 7.49 mm [TD = 5 cm]). Figure 5 shows representative dose distributions for a gated plan and tracking plan with small, equivalent, and large setup margins. This figure shows that the tracking plan with a large setup margin had a higher irradiation area than the gated plan, and the tracking plan with equivalent setup margins had almost the same dose distribution as the gated plan. Fig. 2. Creation of MLC-based tracking plan using 4D dose calculation. (Color version of figure is available online.)
margin is shown in Fig. 4. In this figure, the red line shows the lung V20 of the gated plan. The equivalent setup margin was estimated by a quadratic function interpolation.
Results For DIR accuracy, the anatomic landmark-based target registration error (TRE) for DIR was calculated from peakinhale and peak-exhale in 1 extreme case with a TD of 5 cm and MR of 3 cm, which resulted in a TRE of 2.2 ± 2.1 mm (the number of landmarks = 100).
Discussion The dynamic tumor tracking radiotherapy has the potential to reduce the radiation dose to the total lung compared with that used for conventional radiotherapy. Although the dosimetric impact of a dynamic tracking plan is dependent on the setup margin for the tracking plan, the data showed that the setup margin was not equivalent to that of respiratory-gated radiotherapy, suggesting an equivalent setup margin for the tracking plan for different TDs and MRs, as compared with the gated plan with a gating width of 50% and setup margin of 5 mm. The equivalent setup margin ranged from 4.16 mm (TD = 3 cm, MR = 1 cm) to 9.33 mm (TD = 1 cm, MR = 3 cm).
Fig. 3. Concept of a setup margin for a tracking plan with the same impact on lung V20 obtained by the gated plan (30% to 70% phase) with a setup margin of 5 mm. (Color version of figure is available online.)
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
5
Fig. 4. Estimation of an equivalent setup margin for the tracking plan. The quadratic function was used to calculate the correlation between the setup margin and lung V20 of the tracking plan. Then, the lung V20 value of the gated plan was substituted for y in this equation to obtain an equivalent setup margin for the tracking plan. (Color version of figure is available online.)
Table 1 Summary of V20 of total lung for MLC-based tumor tracking plan with various setup margins and gated plan for various tumor diameters and motion ranges Tumor diameter
Motion range
MLC-based tumor tracking plan Setup margin 5 mm
Setup margin 6 mm
Setup margin 7 mm
Setup margin 8 mm
Setup margin 9 mm
Setup margin 10 mm
1 cm
1 cm 2 cm 3 cm 1 cm 2 cm 3 cm 1 cm 2 cm 3 cm
1.39 1.21 1.27 6.68 6.64 6.34 13.94 15.41 15.65
1.42 1.28 1.32 6.88 6.85 6.59 14.18 15.78 16.01
1.50 1.41 1.44 7.35 7.01 7.03 14.68 16.27 16.73
1.67 1.59 1.65 7.93 7.54 7.63 15.36 17.17 17.46
1.97 1.87 1.87 8.59 8.29 8.37 16.20 18.01 18.33
2.31 2.18 2.20 9.37 9.07 9.32 17.01 19.03 19.22
3 cm
5 cm
As shown in Fig. 6, as the MR increased, the equivalent setup margin also increased. In addition, as the TD increased, the equivalent setup margin decreased. As the equivalent setup margin increased, the effectiveness of dynamic tumor tracking radiotherapy compared with respiratory-gated radiotherapy also increased. That is, dynamic tumor tracking radiotherapy is suitable for tumors with a large MR and small TD. There were several papers focused on comparison between gated and nongated plan or between tracking and nongated plan for lung cancer patients.15,25,28-30 However, to our knowledge, there were no published data focused on
Table 2 Summary of equivalent setup margin for different tumor diameters and motion ranges Motion range
1 cm 2 cm 3 cm
Tumor diameter 1 cm
3 cm
5 cm
5.00 mm 7.14 mm 9.33 mm
4.16 mm 7.04 mm 8.00 mm
4.24 mm 6.35 mm 7.49 mm
Gated plan
1.32 1.43 1.98 6.51 7.09 7.62 13.75 15.97 17.06
comparison between tracking and gated plan for lung cancer. For liver cancer, Yoon et al.31 evaluated the dose difference between gated and tracking plan using CyberKnife for hepatocellular carcinoma. Gated plan used the ITV margin around the tumor determined by measuring its motion over 305% to 70% of respiratory phases using 4D-CT, followed by a 5-mm isotropic margin for PTV, and tracking plan excluded the ITV from PTV. They showed that CyberKnife tracking plan could significantly decrease the volume of normal liver tissue receiving > 15 Gy by 1.61-fold, compared with that in gated plan for 29 hepatocellular carcinoma patients with 14.5 ± 5.7 mm of average respiratory tumor motion and 10 to 20 mm of TD (65.5%). Our result suggested that tracking plan had better dose distribution than gated plan in these patients because the equivalent setup margin was more than 5 mm, which was used for tracking plan in their study. Thus, our result was similar to their study, although the treatment sites were different between our study and their study. The method to calculate the setup margin for the tracking plan is different from that for conventional radiotherapy. In conventional radiotherapy using an image-guided patient setup, the patient’s position is
ARTICLE IN PRESS 6
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
Fig. 5. Representative dose distributions for the gated plan and tracking plan with small, equivalent, and large setup margins (tumor diameter = 5 cm, motion range = 3 cm). (Color version of figure is available online.)
determined by kV or MV images before irradiation. Thus, the intrafractional motion error during dose delivering should be included. On the other hand, dynamic tumor tracking radiotherapy acquires images with the tumor position in real time, so there is no need to include the intrafractional motion error in the setup margin. However, dynamic tumor tracking radiotherapy should include the error related to tumor detection in real time and prediction for respiratory motion, instead of the intrafractional motion error. The equivalent setup margin estimated in this study can also clarify a tolerance of the tumor detection and prediction errors inherent in the dynamic tumor tracking techniques. In other words, by the techniques with such accuracy less than the equivalent setup margin, the dynamic tracking plan achieves a dosimetric impact better than the gated plan. In this sense, the results of this study can be used for defining a performance index to design and develop such techniques for the dynamic tumor tracking radiotherapy better than the gated one. In this study, we only focused on the lung V20 because lung V20 is one of the most commonly used predictors for lung toxicity. However, other DVH parameters for lung also
import predictors for lung toxicity, such as V5, V10, and mean dose. Thus, further investigation study will be needed to clarify the dosimetric and clinical impact of tracking plan using other DVH parameters. There were some limitations to this study. First, only 1 situation was evaluated, in which the tumor was switched to the right lower lobe. With this tumor location, it is assumed that the tumor has the highest respiratory MR. Thus, only this situation was analyzed. However, if the tumor is located in a different lobe, this result may change. Second, in this study, only 1 gating window was employed for the gated plan (i.e., 30% to 70% phase). This gating window is commonly used in clinical practice.32 However, several studies investigated smaller gating windows for gated plans,33,34 and changes to the gating window of the internal margin for the gated plan resulted in a different equivalent setup margin for the tracking plan.35 Thus, further studies are needed to further clarify a relation between gating windows for the gated plan and equivalent setup margins for the tracking plan. Next, our study focused on only 1 direction for tumor motion (i.e., C-C direction). In clinical case, tumor motion is more complex.
Fig. 6. Correlations between motion range and equivalent setup margin (A) and between tumor diameter and equivalent setup margin (B). The equivalent setup margin increases proportionally with the motion range. In addition, as the diameter increases, the equivalent setup margin decreases. (Color version of figure is available online.)
ARTICLE IN PRESS N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
However, almost all tumors had the largest respiratory movement in C-C direction.36 Thus, although the setting used in our study could not simulate the clinical situation perfectly, our result may be helpful for further improving MLCbased dynamic tumor tracking radiotherapy. Then, we used 4D dose calculation with DIR. Thus, our result includes the residual DIR error. The residual DIR error (i.e., TRE within 2 mm) was smaller than the image voxel size in thoracic regions (i.e., 2 mm). In addition, in our study, V20 of total lung was used for calculation of equivalent setup margin. The volume metrics, such as V20, has less influence from the residual DIR error, compared with point metrics, such as maximum and minimum dose for PTV. Thus, the influence was expected to be small (or negligible). Finally, to perform MLC-based dynamic tracking simulation to obtain the tracking MLC position, we used only 10 images representing 10 phases of respiration. As a result, the tracking MLC position could not continuously adapt the patient’s anatomy (i.e., shape of tumor), which may yield the slightly different result for MLC-based tumor tracking plan. Conclusions The results of this study showed that an equivalent setup margin changes in response to TD and MR. The tracking plan with a setup margin smaller than the equivalent setup margin achieves a better dosimetric impact than the gated plan. Conflict of interest There is no conflict of interest with regard to this manuscript. Acknowledgements This work was partially supported by Varian Medical Systems (Palo Alto, CA, USA) and JSPS KAKENHI Grant No. 25293258, No. 15J05402 and No. 15K19765. References 1. National Cancer Center. Cancer statistics in Japan-2014. Available at: http://ganjoho.jp/reg_stat/statistics/stat/summary.html. Accessed August 1, 2016. 2. McCann, J.; Artinian, V.; Duhaime, L.; et al. Evaluation of the causes for racial disparity in surgical treatment of early stage lung cancer. Chest 128:3440–6; 2005. 3. Onishi, H.; Shirato, H.; Nagata, Y.; et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for stage I non-small cell lung cancer: updated results of 257 patients in a Japanese multi-institutional study. J. Thorac. Oncol. 2:S94–100; 2007. 4. Koto, M.; Takai, Y.; Ogawa, Y.; et al. A phase II study on stereotactic body radiotherapy for stage I non-small cell lung cancer. Radiother. Oncol. 85:429–34; 2007. 5. Bush, D.A.; Slater, J.D.; Shin, B.B.; et al. Hypofractionated proton beam radiotherapy for stage I lung cancer. Chest 126:1198–203; 2004.
7
6. Fakiris, A.J.; McGarry, R.C.; Yiannoutsos, C.T.; et al. Stereotactic body radiation therapy for early-stage non-small-cell lung carcinoma: four-year results of a prospective phase II study. Int. J. Radiat. Oncol. Biol. Phys. 75:677–82; 2009. 7. Timmerman, R.; Paulus, R.; Galvin, J.; et al. Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA 303:1070–6; 2010. 8. Keall, P.J.; Todor, A.D.; Vedam, S.S.; et al. On the use of EPID-based implanted marker tracking for 4D radiotherapy. Med. Phys. 31:3492– 9; 2004. 9. Poulsen, P.R.; Cho, B.; Sawant, A.; et al. Implementation of a new method for dynamic multileaf collimator tracking of prostate motion in arc radiotherapy using a single kV imager. Int. J. Radiat. Oncol. Biol. Phys. 76:914–23; 2010. 10. Moore, D.; Ruan, D.; Sawant, A. Fast leaf-fitting with generalized underdose/overdose constraints for real-time MLC tracking. Med. Phys. 43:465; 2016. 11. Zhang, X.; Homma, N.; Ichiji, K.; et al. A kernel-based method for markerless tumor tracking in kV fluoroscopic images. Phys. Med. Biol. 59:4897–911; 2014. 12. Zhang, X.; Homma, N.; Ichiji, K.; et al. Tracking tumor boundary in MV-EPID images without implanted markers: a feasibility study. Med. Phys. 42:2510; 2015. 13. Keall, P. 4-Dimensional computed tomography imaging and treatment planning. Semin. Radiat. Oncol. 14:81–90; 2004. 14. Colvill, E.; Booth, J.T.; O’Brien, R.T.; et al. Multileaf collimator tracking improves dose delivery for prostate cancer radiation therapy: results of the first clinical trial. Int. J. Radiat. Oncol. Biol. Phys. 92:1141–7; 2015. 15. Saito, T.; Sakamoto, T.; Oya, N. Effect of gating window width on lung dosimetric parameters and on treatment time at gated radiotherapy for lung cancer. Acta Oncol. 50:1246–9; 2011. 16. Takao, S.; Miyamoto, N.; Matsuura, T.; et al. Intrafractional baseline shift or drift of lung tumor motion during gated radiation therapy with a real-time tumor-tracking system. Int. J. Radiat. Oncol. Biol. Phys. 94:172–80; 2016. 17. Lang, S.; Shrestha, B.; Graydon, S.; et al. Clinical application of flattening filter free beams for extracranial stereotactic radiotherapy. Radiother. Oncol. 106:255–9; 2013. 18. Navarria, P.; Ascolese, A.M.; Mancosu, P.; et al. Volumetric modulated arc therapy with flattening filter free (FFF) beams for stereotactic body radiation therapy (SBRT) in patients with medically inoperable early stage non small cell lung cancer (NSCLC). Radiother. Oncol. 107:414–8; 2013. 19. Onishi, H.; Araki, T.; Shirato, H.; et al. Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanese multiinstitutional study. Cancer 101:1623–31; 2004. 20. Onishi, H.; Shirato, H.; Nagata, Y.; et al. Stereotactic body radiotherapy (SBRT) for operable stage I non-small-cell lung cancer: can SBRT be comparable to surgery? Int. J. Radiat. Oncol. Biol. Phys. 81:1352–8; 2011. 21. Cai, J.; Malhotra, H.K.; Orton, C.G. A 3D-conformal technique is better than IMRT or VMAT for lung SBRT. Med. Phys. 41:040601; 2014. 22. Segars, W.P.; Sturgeon, G.; Mendonca, S.; et al. 4D XCAT phantom for multimodality imaging research. Med. Phys. 37:4902–15; 2010. 23. Weistrand, O.; Svensson, S. The ANACONDA algorithm for deformable image registration in radiotherapy. Med. Phys. 42:40–53; 2015. 24. Kadoya, N.; Nakajima, Y.; Saito, M.; et al. Multi-institutional validation study of commercially available deformable image registration software for thoracic images. Int. J. Radiat. Oncol. Biol. Phys. 96(2): 422–31; 2016. 25. Keall, P.J.; Joshi, S.; Vedam, S.S.; et al. Four-dimensional radiotherapy planning for DMLC-based respiratory motion tracking. Med. Phys. 32:942; 2005.
ARTICLE IN PRESS 8
N. Kadoya et al. / Medical Dosimetry ■■ (2017) ■■–■■
26. Timmerman, R.; Galvin, J.; Michalski, J.; et al. Accreditation and quality assurance for Radiation Therapy Oncology Group: multicenter clinical trials using stereotactic body radiation therapy in lung cancer. Acta Oncol. 45:779–86; 2006. 27. Kadoya, N.; Fujita, Y.; Katsuta, Y.; et al. Evaluation of various deformable image registration algorithms for thoracic images. J Radiat Res 55:175–82; 2014. 28. Saito, T.; Sakamoto, T.; Oya, N. Comparison of gating around endexpiration and end-inspiration in radiotherapy for lung cancer. Radiother. Oncol. 93:430–5; 2009. 29. Underberg, R.W.; van Sornsen de Koste, J.R.; Lagerwaard, F.J.; et al. A dosimetric analysis of respiration-gated radiotherapy in patients with stage III lung cancer. Radiat. Oncol. 1:8; 2006. 30. Matsuo, Y.; Ueki, N.; Takayama, K.; et al. Evaluation of dynamic tumour tracking radiotherapy with real-time monitoring for lung tumours using a gimbal mounted linac. Radiother. Oncol. 112:360–4; 2014. 31. Yoon, K.; Kwak, J.; Cho, B.; et al. Gated volumetric-modulated arc therapy vs. tumor-tracking cyberKnife radiotherapy as stereotactic
32.
33.
34.
35.
36.
body radiotherapy for hepatocellular carcinoma: A dosimetric comparison study focused on the impact of respiratory motion managements. PLoS ONE 11:e0166927; 2016. Jang, S.S.; Huh, G.J.; Park, S.Y.; et al. The impact of respiratory gating on lung dosimetry in stereotactic body radiotherapy for lung cancer. Phys. Med. 30:682–9; 2014. Chin, E.; Loewen, S.K.; Nichol, A.; et al. 4D VMAT, gated VMAT, and 3D VMAT for stereotactic body radiation therapy in lung. Phys. Med. Biol. 58:749–70; 2013. Wolthaus, J.W.; Sonke, J.J.; van Herk, M.; et al. Comparison of different strategies to use four-dimensional computed tomography in treatment planning for lung cancer patients. Int. J. Radiat. Oncol. Biol. Phys. 70:1229–38; 2008. Cai, J.; McLawhorn, R.; Read, P.W.; et al. Effects of breathing variation on gating window internal target volume in respiratory gated radiation therapy. Med. Phys. 37:3927–34; 2010. Liu, H.H.; Balter, P.; Tutt, T.; et al. Assessing respiration-induced tumor motion and internal target volume using four-dimensional computed tomography for radiotherapy of lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 68:531–40; 2007.