- Email: [email protected]
4D Tracking Radiotherapy Delivery Evaluation for Five Patients with Mobile Lung Tumor
I. J. Radiation Oncology d Biology d Physics
S614
Volume 72, Number 1, Supplement, 2008
(used as a reference SUV). For the periodic motion patterns, as the motion amplitude increased, the ungated PET SUVs decreased linearly with increasing amplitude of motion, for 1 cm and 2 cm amplitude motion the SUVs were 54% and 38% of the reference SUV. The spatial-gated and shifted PET images reconstructed the target shape and SUVs more accurately. Conclusions: These phantom studies showed the potential of using a spatially-based binning method for improving the quality and quantitation of free-breathing lung-cancer PET/CT images. Author Disclosure: P.K. Yalavarthy, None; Z. Wei, Philips Medical Systems, A. Employment; J. Wen, None; P. Parikh, None; T. Zhao, None; J. Bradely, None; D. Low, None.
2991
4D Tracking Radiotherapy Delivery Evaluation for Five Patients with Mobile Lung Tumor
Y. Liu, B. Lin, N. Papanikolaou University of Texas Health Science Center, San Antonio, TX Purpose/Objective(s): The aim of this study is to provide a pre-clinical evaluation of a four-dimensional tracking radiotherapy (4D-TRT) delivery to lung tumor using a prototype tracking system. Five patients who were treated with conventional three-dimensional conformal radiotherapy (3D-CRT) were selected for the evaluation studies. The evaluation was carried on using films dosimetric analysis based-on QA dynamic phantom and using DVH analysis based on a 3D-CRT and 4D-TRT planning respectively. Materials/Methods: The concept for four-dimensional tracking radiation therapy (4D-TRT) can be shortly explained as to control the beam following the tumor motion. Currently in clinic the challenging is how to move the MLC leaves to form a tracking beam according to the motion of target. The key component of the evaluated 4D-TRT system was TrackBeam. It consists of image processing tools and first-of-its-kind dual-layer micro MLC. DmMLC has two layers of orthogonal leaves providing advantages in speed and conformality when forming beam aperture for tracking. The TrackBeam was mounted to a Varian Linac and connected to a workstation which process the online MV fluence and controls each leaf’s motion. A Quasar dynamic phantom was used for radiographic film irradiation with 4DTRT and also 3D-CRT. The phantom has a Gafchromic film insert and a gold marker in the insert. The patient respiratory motion data was recorded during the 4D-CT scanning and loaded to the dynamic phantom for QA propose. A 3D-CRT and 4D-TRT planning were developed with 180 cGy at 33 fraction based on 3D-CT and dose volume histogram was compared. Results: The respiratory motion cycles of the five patients were averaged at 3.31, 5.54, 2.67, 3.74 and 6.23 seconds per breathing. The synchronization of marker motion and the DmMLC leaf motion was achieved within less than 0.05 seconds. To evaluate the effect of real-time tumor tracking, a static tumor without motion was used as reference. The films analysis indicated that total 32.15% over the tolerance of 5% for 3D-CRT and 7.12% of over the tolerance of 5% for 4D-TRT compared to the static film respectively. The DVH comparisons indicate 4D-TRT reduces significant dose to the Ring (expanding 2.5 cm from GTV) from 97.5% volume to 61.5% volume at V20 and reduces dose from 33.5 Gy to 11.5 Gy at 80% volume. 4D-TRT also reduces considerable amount dose to the total lung from 28% volume to 18% volume at V20 and reduces dose from 35.2 Gy to 15.0 Gy at 20% volume. Conclusions: We evaluated a 4D-TRT DmMLC-based delivery and provided dosimetric analysis of radiographic films based on dynamic phantoms and DVH analysis based on patients 3D-CRT and 4D-TRT planning. Lung tumors are susceptible to motion due to respiration. The 4D-TRT provided conformal coverage to mobile tumor and meanwhile limited the significant dose to the surrounding normal tissue. Author Disclosure: Y. Liu, None; B. Lin, None; N. Papanikolaou, None.
2992
Evaluation of a 4DCT-based IMRT Plan for Lung using the Free-form Image Deformation Dose Tracking Method
H. Alasti1,2, J. Chow1,2, D. Markel1, C. Lochovsky1, D. Payne1,2 1
Princess Margaret Hospital, Toronto, ON, Canada, 2University of Toronto, Toronto, ON, Canada
Purpose/Objective(s): 4D-CT provides patient specific tumor motion in RT planning of lung cancer patients. For the lung IMRT plan using the 4D-CT dataset, target volumes of phase 0% (end-inhale) and 50% (end-exhale) of patient’s breathing cycle, are fused to form an internal target volume (ITV). Treatment planning using ITV accounts for the target motion due to breathing, but unavoidably results in more exposed normal lung tissue. This study investigated the dose-volume relationship of the ITV-based plan and compared it with a plan considering the GTV and lung motion and deformation. The Free-Form Image Deformation (FFID) model was used to track/register the dose distributions of ten phases of 4D-CT image sets, taking into account the organ motion and deformation. Materials/Methods: Seven-beam IMRT plan was created on the ITV-based plan of lung cancer patient using the Pinnacle3 TPS with prescription dose of 45 Gy/25 fractions. The fused PTV was defined as ITV plus a 6 mm margin and the image set of phase 30% was used for dose calculations. To determine the actual dose deposited at each voxel in which the GTV moves and deforms during breathing, the ITV-based IMRT plan was copied on to the image sets of each phase. Ten dose distributions were calculated for the ten phases as per the same IMRT plan. The ten dose distributions, representing the target and critical organ motions, were input to an in-house developed voxel-matching program based on the FFID algorithm. The program tracked the dose distributions of the ten phases by deforming all voxels in different phases to phase 30%, which is the image set for the ITV-based plans. Results: The dose-volume histograms (DVHs) for the GTV and lung of the rigid ITV-based plan (A) are compared to those of the voxel-tracking non-rigid deformed plan (B). The DVH of the GTV of plan (B) shows inadequate dosimetric coverage compared to that of plan (A), implying that a higher TCP value of 39% is achieved in plan (A) compared to 35% in plan (B) based on the 45 Gy prescription. This uncertainty arises from neglecting the respiratory motion and lung deformation in plan (A). The DVH for the lung shows a small variation among the ten phases, and the 20 Gy covered volume is similar between plans (A) and (B). However the high dose (30-35 Gy) covered volumes of the ipsi lung are 20-30% greater for plan (B) than (A).
S614
Volume 72, Number 1, Supplement, 2008
(used as a reference SUV). For the periodic motion patterns, as the motion amplitude increased, the ungated PET SUVs decreased linearly with increasing amplitude of motion, for 1 cm and 2 cm amplitude motion the SUVs were 54% and 38% of the reference SUV. The spatial-gated and shifted PET images reconstructed the target shape and SUVs more accurately. Conclusions: These phantom studies showed the potential of using a spatially-based binning method for improving the quality and quantitation of free-breathing lung-cancer PET/CT images. Author Disclosure: P.K. Yalavarthy, None; Z. Wei, Philips Medical Systems, A. Employment; J. Wen, None; P. Parikh, None; T. Zhao, None; J. Bradely, None; D. Low, None.
2991
4D Tracking Radiotherapy Delivery Evaluation for Five Patients with Mobile Lung Tumor
Y. Liu, B. Lin, N. Papanikolaou University of Texas Health Science Center, San Antonio, TX Purpose/Objective(s): The aim of this study is to provide a pre-clinical evaluation of a four-dimensional tracking radiotherapy (4D-TRT) delivery to lung tumor using a prototype tracking system. Five patients who were treated with conventional three-dimensional conformal radiotherapy (3D-CRT) were selected for the evaluation studies. The evaluation was carried on using films dosimetric analysis based-on QA dynamic phantom and using DVH analysis based on a 3D-CRT and 4D-TRT planning respectively. Materials/Methods: The concept for four-dimensional tracking radiation therapy (4D-TRT) can be shortly explained as to control the beam following the tumor motion. Currently in clinic the challenging is how to move the MLC leaves to form a tracking beam according to the motion of target. The key component of the evaluated 4D-TRT system was TrackBeam. It consists of image processing tools and first-of-its-kind dual-layer micro MLC. DmMLC has two layers of orthogonal leaves providing advantages in speed and conformality when forming beam aperture for tracking. The TrackBeam was mounted to a Varian Linac and connected to a workstation which process the online MV fluence and controls each leaf’s motion. A Quasar dynamic phantom was used for radiographic film irradiation with 4DTRT and also 3D-CRT. The phantom has a Gafchromic film insert and a gold marker in the insert. The patient respiratory motion data was recorded during the 4D-CT scanning and loaded to the dynamic phantom for QA propose. A 3D-CRT and 4D-TRT planning were developed with 180 cGy at 33 fraction based on 3D-CT and dose volume histogram was compared. Results: The respiratory motion cycles of the five patients were averaged at 3.31, 5.54, 2.67, 3.74 and 6.23 seconds per breathing. The synchronization of marker motion and the DmMLC leaf motion was achieved within less than 0.05 seconds. To evaluate the effect of real-time tumor tracking, a static tumor without motion was used as reference. The films analysis indicated that total 32.15% over the tolerance of 5% for 3D-CRT and 7.12% of over the tolerance of 5% for 4D-TRT compared to the static film respectively. The DVH comparisons indicate 4D-TRT reduces significant dose to the Ring (expanding 2.5 cm from GTV) from 97.5% volume to 61.5% volume at V20 and reduces dose from 33.5 Gy to 11.5 Gy at 80% volume. 4D-TRT also reduces considerable amount dose to the total lung from 28% volume to 18% volume at V20 and reduces dose from 35.2 Gy to 15.0 Gy at 20% volume. Conclusions: We evaluated a 4D-TRT DmMLC-based delivery and provided dosimetric analysis of radiographic films based on dynamic phantoms and DVH analysis based on patients 3D-CRT and 4D-TRT planning. Lung tumors are susceptible to motion due to respiration. The 4D-TRT provided conformal coverage to mobile tumor and meanwhile limited the significant dose to the surrounding normal tissue. Author Disclosure: Y. Liu, None; B. Lin, None; N. Papanikolaou, None.
2992
Evaluation of a 4DCT-based IMRT Plan for Lung using the Free-form Image Deformation Dose Tracking Method
H. Alasti1,2, J. Chow1,2, D. Markel1, C. Lochovsky1, D. Payne1,2 1
Princess Margaret Hospital, Toronto, ON, Canada, 2University of Toronto, Toronto, ON, Canada
Purpose/Objective(s): 4D-CT provides patient specific tumor motion in RT planning of lung cancer patients. For the lung IMRT plan using the 4D-CT dataset, target volumes of phase 0% (end-inhale) and 50% (end-exhale) of patient’s breathing cycle, are fused to form an internal target volume (ITV). Treatment planning using ITV accounts for the target motion due to breathing, but unavoidably results in more exposed normal lung tissue. This study investigated the dose-volume relationship of the ITV-based plan and compared it with a plan considering the GTV and lung motion and deformation. The Free-Form Image Deformation (FFID) model was used to track/register the dose distributions of ten phases of 4D-CT image sets, taking into account the organ motion and deformation. Materials/Methods: Seven-beam IMRT plan was created on the ITV-based plan of lung cancer patient using the Pinnacle3 TPS with prescription dose of 45 Gy/25 fractions. The fused PTV was defined as ITV plus a 6 mm margin and the image set of phase 30% was used for dose calculations. To determine the actual dose deposited at each voxel in which the GTV moves and deforms during breathing, the ITV-based IMRT plan was copied on to the image sets of each phase. Ten dose distributions were calculated for the ten phases as per the same IMRT plan. The ten dose distributions, representing the target and critical organ motions, were input to an in-house developed voxel-matching program based on the FFID algorithm. The program tracked the dose distributions of the ten phases by deforming all voxels in different phases to phase 30%, which is the image set for the ITV-based plans. Results: The dose-volume histograms (DVHs) for the GTV and lung of the rigid ITV-based plan (A) are compared to those of the voxel-tracking non-rigid deformed plan (B). The DVH of the GTV of plan (B) shows inadequate dosimetric coverage compared to that of plan (A), implying that a higher TCP value of 39% is achieved in plan (A) compared to 35% in plan (B) based on the 45 Gy prescription. This uncertainty arises from neglecting the respiratory motion and lung deformation in plan (A). The DVH for the lung shows a small variation among the ten phases, and the 20 Gy covered volume is similar between plans (A) and (B). However the high dose (30-35 Gy) covered volumes of the ipsi lung are 20-30% greater for plan (B) than (A).