- Email: [email protected]
1189 poster ENDOBILIARY STENT: A SURROGATE MARKER FOR PATIENT ALIGNMENT IN IGRT IN CARCINOMA GALL BLADDER AND PANCREAS
IGRT: A DAPTIVE TREATMENT SCHEMES AND CORRECTION STRATEGIES
normal tissue dose constraints. Analysis of target coverage and dose to surrounding normal tissues in each phase is ongoing. Recruitment to and followup within the study continues. 1188 poster DOSIMETRIC CONSEQUENCES OF ANATOMIC VARIATIONS DURING HEAD-AND-NECK RADIOTHERAPY TREATMENT M. Marguet1 , S. Warren1 , M. C. Bordage2 , R. Garcia1 1 2
I NSTITUT S AINTE C ATHERINE, Avignon - Cedex 02, France L APLACE, Toulouse, France
Purpose: Anatomic variations during head-and-neck radiotherapy treatment may compromise the delivery of the planned dose distribution, particularly in the case of IMRT treatments. The aim of this study was to establish dosimetric indicators to identify patients whose delivered dose deviates from the planned dose to allow an eventual re-optimisation of the patient’s dosimetry, if necessary, during the course of their radiotherapy treatment. Materials: Basic indicators such as the position of the parotids with respect to high dose gradient regions in the planned dose distribution at the start of radiotherapy may help to identify patients "at risk" of significant dosimetric changes during the course of their treatment. The study was carried out without the need of any additional contouring ; the patient dose distribution analysis and comparison was performed from the initial CT and repeated CT or CBCT using a modified gamma index technique that we named gammaLSC3D. This improved gamma index technique quantified and identified the location of changes in the dose distribution in relation to the planned dose, in a stack of 2D images, pixel by pixel, with particular reference to the target volume (PTV) or organs at risk (parotids) location. Moreover, the superimposition of gammaLSC3D indicator maps with CT or CBCT maps was implemented in order to visualise the dose deviation information directly on the organs. You can see an example for a patient in Figure 1.
S 443
Purpose: To study the position of the stent in pre-treatment MVCT images with respect to the stent on the planning KVCT and analyse the possibility of using it as a surrogate for tumour during matching for daily registration in cases of carcinoma gall bladder and pancreas treated on Tomotherapy Hi ART system. Materials: Patients of carcinoma pancreas and gall bladder with endobiliary stents were included in this study. Computed tomography (CT) was carried out as a part of treatment planning in all patients. Patients were treated on Tomotherapy Hi ART system with Mega Voltage Computed tomography (MVCT) images acquired prior to every fraction delivered. Both the planning CT and MVCT of the first seven fractions were transferred to the focal sim (Vs) contouring station. The planning CT of each patient was independently fused with each of the seven MVCT images of that particular patient using alignment with the vertebral body, liver and the kidneys. The stent was then contoured on the planning CT and the seven MVCT images for each patient. During contouring on the MVCT images, the contours on the planning CT were turned off. The centre of mass (CM) of the stents both on the planning as well as the seven MVCT images was found for all patients. The difference between the three co-ordinates of the CM of planning and MVCT was found. The mean of these differences across all patients was calculated. Results: Results: Three patients of gall bladder cancer and two patients of carcinoma pancreas were included in the study. A total of five CMs of the stents on the planning CT images and thirty five CMs of the stents on the MVCT images were obtained. In total thirty five readings for difference in the planning and MVCT CMs were thus obtained. The mean of difference in CMs in the right -left, supero-inferior and antero-posterior directions was found to be 0.16 cm (±0.7), 0.22 cm (±2.07) and 0.36 cm (±0.7) respectively. Conclusions: The study validates the use of endobiliary stents in gall bladder and pancreas cancers as stable tumour localization surrogates. In this region of body in which there is an absence of any fixed bony point for daily pretreatment matching, stent can be used as a reliable surrogate for tumour and can be used as a valuable method to correct for interfraction target motion and to improve precision in the delivery of radiotherapy in patients of carcinoma gall bladder and pancreas. 1190 poster EVALUATION OF DEFORMABLE IMAGE CO-REGISTRATION IN ADAPTIVE IMRT FOR HEAD AND NECK CANCER A. M. L. Olteanu1 , I. Madani1 , W. De Gersem1 , T. Vercauteren1 , W. De Neve1 1 U NIVERSITAIR Z IEKENHUIS, Department of Radiotherapy, Gent, Belgium
Results: The changes observed in the dose distribution for the PTV or parotids were analysed and presented in the form of gammaLSC3D-volume histograms in order to facilitate the follow-up of dosimetric changes during the radiotherapy treatment. We suggest a replanification for patients whose more than 5% of the selected pixels do no match anymore with planned dose. Conclusions: This analysis method has been automated in order to make it applicable in clinical routine to have the follow-up of dose variations during head and neck radiotherapy treatment. 1189 poster ENDOBILIARY STENT: A SURROGATE MARKER FOR PATIENT ALIGNMENT IN IGRT IN CARCINOMA GALL BLADDER AND PANCREAS R. Engineer1 , P. Shukla1 , S. Chopra1 , Z. Master2 , S. K. Shrivastava1 1
TATA M EMORIAL H OSPITAL, Department of Radiation Oncology, Mumbai, India TATA M EMORIAL H OSPITAL, Department of Medical Physics and Radiation Safety, Mumbai, India 2
Purpose: to assess accuracy of contour deformation applying deformable image co-registration in adaptive IMRT for head and neck cancer. Materials: Data of 12 patients with non-metastatic head and neck cancer treated at the highest dose level of the phase I trial on adaptive 18 F-FDGPET-voxel intensity-based intensity-modulated radiotherapy (IMRT), i.e. dose painting by numbers (DPBN) were used in this study [1]. Each patient had two DPBN plans followed by conventional (homogenous dose) treatment. The initial DPBN plan was based on a pre-treatment PET/CT scan, while the second, adapted DPBN plan (fractions 11-20) was based on a PET/CT scan acquired after the 8th fraction. The median prescription dose to the dose-painted volume was 30 Gy for both DPBN plans. The median dose prescription for the conventional plans (fraction 21-32) was 26 Gy. A commercial available software package MIMvista, version 4.2 (MIM Software Inc., Cleveland, OH) applying intensity-based free-form deformable registration algorithm was used for deformation of contours of pretreatment CT to per-treatment CT. Jaccard (JI) and overlap indices (OI) were computed for deformed regions of interest (ROIdef) as well as regions of interest deformed and adjusted at discretion of an experienced head-and-neck radiation oncologist (ROIdef_ad) in comparison to the manually redrawn ones (ROIm). Results: The radiation oncologist chose to adjust almost all deformed ROIs in all patients. Variable results were obtained for the GTV depending on tumor response: the largest adaptations were made in patients with substantially regressing tumors (ROIdef=11.8 ± 10.9 cm3 vs. ROIdef_ad=5.9±7.8 cm3 vs. ROIm=7.7 ± 7.2 cm3 [p=0.57]). The swallowing structures (superior, middle, inferior pharyngeal constrictor muscles [SPC, MPC, IPC, respectively]; upper esophageal sphincter [UES], esophagus and supraglottic larynx) represented the most frequently adjusted ROIs. The mandible was well propagated and needed little or no adaptations. JI and OI confirm these observations (Table) with the highest for the mandible (JI=0.8 and OI=0.9) and lowest for the UES (JI=0.3 for ROIdef and 0.4 for ROIdef_ad; OI=0.5 for both ROIs). Manual adjustment of deformed contours addressed uncertainty and artefacts. As a result, indices of similarity (JI) and overlap (OI) were better between deformed and deformed-adjusted ROIs, than between deformed and manually redrawn ROIs, especially for the swallowing structures.
normal tissue dose constraints. Analysis of target coverage and dose to surrounding normal tissues in each phase is ongoing. Recruitment to and followup within the study continues. 1188 poster DOSIMETRIC CONSEQUENCES OF ANATOMIC VARIATIONS DURING HEAD-AND-NECK RADIOTHERAPY TREATMENT M. Marguet1 , S. Warren1 , M. C. Bordage2 , R. Garcia1 1 2
I NSTITUT S AINTE C ATHERINE, Avignon - Cedex 02, France L APLACE, Toulouse, France
Purpose: Anatomic variations during head-and-neck radiotherapy treatment may compromise the delivery of the planned dose distribution, particularly in the case of IMRT treatments. The aim of this study was to establish dosimetric indicators to identify patients whose delivered dose deviates from the planned dose to allow an eventual re-optimisation of the patient’s dosimetry, if necessary, during the course of their radiotherapy treatment. Materials: Basic indicators such as the position of the parotids with respect to high dose gradient regions in the planned dose distribution at the start of radiotherapy may help to identify patients "at risk" of significant dosimetric changes during the course of their treatment. The study was carried out without the need of any additional contouring ; the patient dose distribution analysis and comparison was performed from the initial CT and repeated CT or CBCT using a modified gamma index technique that we named gammaLSC3D. This improved gamma index technique quantified and identified the location of changes in the dose distribution in relation to the planned dose, in a stack of 2D images, pixel by pixel, with particular reference to the target volume (PTV) or organs at risk (parotids) location. Moreover, the superimposition of gammaLSC3D indicator maps with CT or CBCT maps was implemented in order to visualise the dose deviation information directly on the organs. You can see an example for a patient in Figure 1.
S 443
Purpose: To study the position of the stent in pre-treatment MVCT images with respect to the stent on the planning KVCT and analyse the possibility of using it as a surrogate for tumour during matching for daily registration in cases of carcinoma gall bladder and pancreas treated on Tomotherapy Hi ART system. Materials: Patients of carcinoma pancreas and gall bladder with endobiliary stents were included in this study. Computed tomography (CT) was carried out as a part of treatment planning in all patients. Patients were treated on Tomotherapy Hi ART system with Mega Voltage Computed tomography (MVCT) images acquired prior to every fraction delivered. Both the planning CT and MVCT of the first seven fractions were transferred to the focal sim (Vs) contouring station. The planning CT of each patient was independently fused with each of the seven MVCT images of that particular patient using alignment with the vertebral body, liver and the kidneys. The stent was then contoured on the planning CT and the seven MVCT images for each patient. During contouring on the MVCT images, the contours on the planning CT were turned off. The centre of mass (CM) of the stents both on the planning as well as the seven MVCT images was found for all patients. The difference between the three co-ordinates of the CM of planning and MVCT was found. The mean of these differences across all patients was calculated. Results: Results: Three patients of gall bladder cancer and two patients of carcinoma pancreas were included in the study. A total of five CMs of the stents on the planning CT images and thirty five CMs of the stents on the MVCT images were obtained. In total thirty five readings for difference in the planning and MVCT CMs were thus obtained. The mean of difference in CMs in the right -left, supero-inferior and antero-posterior directions was found to be 0.16 cm (±0.7), 0.22 cm (±2.07) and 0.36 cm (±0.7) respectively. Conclusions: The study validates the use of endobiliary stents in gall bladder and pancreas cancers as stable tumour localization surrogates. In this region of body in which there is an absence of any fixed bony point for daily pretreatment matching, stent can be used as a reliable surrogate for tumour and can be used as a valuable method to correct for interfraction target motion and to improve precision in the delivery of radiotherapy in patients of carcinoma gall bladder and pancreas. 1190 poster EVALUATION OF DEFORMABLE IMAGE CO-REGISTRATION IN ADAPTIVE IMRT FOR HEAD AND NECK CANCER A. M. L. Olteanu1 , I. Madani1 , W. De Gersem1 , T. Vercauteren1 , W. De Neve1 1 U NIVERSITAIR Z IEKENHUIS, Department of Radiotherapy, Gent, Belgium
Results: The changes observed in the dose distribution for the PTV or parotids were analysed and presented in the form of gammaLSC3D-volume histograms in order to facilitate the follow-up of dosimetric changes during the radiotherapy treatment. We suggest a replanification for patients whose more than 5% of the selected pixels do no match anymore with planned dose. Conclusions: This analysis method has been automated in order to make it applicable in clinical routine to have the follow-up of dose variations during head and neck radiotherapy treatment. 1189 poster ENDOBILIARY STENT: A SURROGATE MARKER FOR PATIENT ALIGNMENT IN IGRT IN CARCINOMA GALL BLADDER AND PANCREAS R. Engineer1 , P. Shukla1 , S. Chopra1 , Z. Master2 , S. K. Shrivastava1 1
TATA M EMORIAL H OSPITAL, Department of Radiation Oncology, Mumbai, India TATA M EMORIAL H OSPITAL, Department of Medical Physics and Radiation Safety, Mumbai, India 2
Purpose: to assess accuracy of contour deformation applying deformable image co-registration in adaptive IMRT for head and neck cancer. Materials: Data of 12 patients with non-metastatic head and neck cancer treated at the highest dose level of the phase I trial on adaptive 18 F-FDGPET-voxel intensity-based intensity-modulated radiotherapy (IMRT), i.e. dose painting by numbers (DPBN) were used in this study [1]. Each patient had two DPBN plans followed by conventional (homogenous dose) treatment. The initial DPBN plan was based on a pre-treatment PET/CT scan, while the second, adapted DPBN plan (fractions 11-20) was based on a PET/CT scan acquired after the 8th fraction. The median prescription dose to the dose-painted volume was 30 Gy for both DPBN plans. The median dose prescription for the conventional plans (fraction 21-32) was 26 Gy. A commercial available software package MIMvista, version 4.2 (MIM Software Inc., Cleveland, OH) applying intensity-based free-form deformable registration algorithm was used for deformation of contours of pretreatment CT to per-treatment CT. Jaccard (JI) and overlap indices (OI) were computed for deformed regions of interest (ROIdef) as well as regions of interest deformed and adjusted at discretion of an experienced head-and-neck radiation oncologist (ROIdef_ad) in comparison to the manually redrawn ones (ROIm). Results: The radiation oncologist chose to adjust almost all deformed ROIs in all patients. Variable results were obtained for the GTV depending on tumor response: the largest adaptations were made in patients with substantially regressing tumors (ROIdef=11.8 ± 10.9 cm3 vs. ROIdef_ad=5.9±7.8 cm3 vs. ROIm=7.7 ± 7.2 cm3 [p=0.57]). The swallowing structures (superior, middle, inferior pharyngeal constrictor muscles [SPC, MPC, IPC, respectively]; upper esophageal sphincter [UES], esophagus and supraglottic larynx) represented the most frequently adjusted ROIs. The mandible was well propagated and needed little or no adaptations. JI and OI confirm these observations (Table) with the highest for the mandible (JI=0.8 and OI=0.9) and lowest for the UES (JI=0.3 for ROIdef and 0.4 for ROIdef_ad; OI=0.5 for both ROIs). Manual adjustment of deformed contours addressed uncertainty and artefacts. As a result, indices of similarity (JI) and overlap (OI) were better between deformed and deformed-adjusted ROIs, than between deformed and manually redrawn ROIs, especially for the swallowing structures.