A Study to Quantify Dosimetric Consequences of Pancreatic Tumor Motion on the Robustness of IMRT Plans using a Motion Kernel Convolution Method

Abstract

Purpose/Objective(s)Pancreatic tumors show significant, respiration-induced motion whose magnitudes vary from patient to patient. Since IMRT treatments are highly conformal, the understanding of this target motion and how it affects the dose distribution is essential to the accurate delivery of IMRT. Using a method involving convolution with a 3D motion kernel, we studied the dosimetric consequences of pancreatic tumor motion on IMRT plans.Materials/MethodsIMRT plans of 5 patients with pancreatic cancer were examined in this study. PTvs. were obtained by expanding the CTvs. by 1 cm and 50.4 Gy was then prescribed to the PTV. The plans were generated in the Eclipse TPS (Varian) using 5-7 beams with a sliding window delivery technique (Varian 2300IX). Phantom verification plans (PVP) were also created from the treatment plans in the Eclipse TPS. The 3D convolution method, which used a motion kernel to simulate 3D pancreatic tumor motion, was applied for all plans to take into account tumor motion. Using motion-corrected PVP, 2D IMRT QA dose maps were measured and simulated. The IMRT QA passing rates were obtained by comparing the simulated measurements with 2D verification plans for a criterion of 90% at 3%, 3 mm distance to agreement, threshold = 10%. For the motion-corrected plan, the DVH parameters of CTV, PTV and OAR were analyzed. PTV's receiving >95% of prescription dose (V95%) and CTvs. receiving >98% of prescription dose (V98%) were obtained, respectively.ResultsPTV coverage and IMRT QA passing rate were reduced significantly with an increase in the magnitude of pancreatic tumor motion. The V95% of PTV was reduced by 12% when 1.0 cm tumor motion was applied. By contrast, the CTV coverage was maintained when the magnitude of tumor motion was 1.0 cm. The V98% of CTV was reduced by 16% when the largest tumor motion (2.0 cm) was applied. The IMRT QA passing rate was ˜80% when 1.5 cm motion was applied, which corresponds to a 23% and 6% reduction of the PTV and CTV coverage, respectively. Dose to OARs (small bowel, liver and kidneys) were not affected significantly with tumor motion. The general tendency in dose to the OAR was an increase in the low dose region and a decrease in the high dose region as the magnitude of motion increased.ConclusionsThis study suggests that CTV coverage of pancreatic tumor can be maintained with IMRT delivery as long as the pancreatic tumor motion is within the PTV. The CTV coverage, however, decreases if the actual tumor motion moves beyond PTV. Therefore, 4D-CT or MRI should be employed to evaluate the extent of pancreatic tumor motion. Instead of a universal margin, the ITV should be used to take into account the pancreatic tumor motion. Purpose/Objective(s)Pancreatic tumors show significant, respiration-induced motion whose magnitudes vary from patient to patient. Since IMRT treatments are highly conformal, the understanding of this target motion and how it affects the dose distribution is essential to the accurate delivery of IMRT. Using a method involving convolution with a 3D motion kernel, we studied the dosimetric consequences of pancreatic tumor motion on IMRT plans. Pancreatic tumors show significant, respiration-induced motion whose magnitudes vary from patient to patient. Since IMRT treatments are highly conformal, the understanding of this target motion and how it affects the dose distribution is essential to the accurate delivery of IMRT. Using a method involving convolution with a 3D motion kernel, we studied the dosimetric consequences of pancreatic tumor motion on IMRT plans. Materials/MethodsIMRT plans of 5 patients with pancreatic cancer were examined in this study. PTvs. were obtained by expanding the CTvs. by 1 cm and 50.4 Gy was then prescribed to the PTV. The plans were generated in the Eclipse TPS (Varian) using 5-7 beams with a sliding window delivery technique (Varian 2300IX). Phantom verification plans (PVP) were also created from the treatment plans in the Eclipse TPS. The 3D convolution method, which used a motion kernel to simulate 3D pancreatic tumor motion, was applied for all plans to take into account tumor motion. Using motion-corrected PVP, 2D IMRT QA dose maps were measured and simulated. The IMRT QA passing rates were obtained by comparing the simulated measurements with 2D verification plans for a criterion of 90% at 3%, 3 mm distance to agreement, threshold = 10%. For the motion-corrected plan, the DVH parameters of CTV, PTV and OAR were analyzed. PTV's receiving >95% of prescription dose (V95%) and CTvs. receiving >98% of prescription dose (V98%) were obtained, respectively. IMRT plans of 5 patients with pancreatic cancer were examined in this study. PTvs. were obtained by expanding the CTvs. by 1 cm and 50.4 Gy was then prescribed to the PTV. The plans were generated in the Eclipse TPS (Varian) using 5-7 beams with a sliding window delivery technique (Varian 2300IX). Phantom verification plans (PVP) were also created from the treatment plans in the Eclipse TPS. The 3D convolution method, which used a motion kernel to simulate 3D pancreatic tumor motion, was applied for all plans to take into account tumor motion. Using motion-corrected PVP, 2D IMRT QA dose maps were measured and simulated. The IMRT QA passing rates were obtained by comparing the simulated measurements with 2D verification plans for a criterion of 90% at 3%, 3 mm distance to agreement, threshold = 10%. For the motion-corrected plan, the DVH parameters of CTV, PTV and OAR were analyzed. PTV's receiving >95% of prescription dose (V95%) and CTvs. receiving >98% of prescription dose (V98%) were obtained, respectively. ResultsPTV coverage and IMRT QA passing rate were reduced significantly with an increase in the magnitude of pancreatic tumor motion. The V95% of PTV was reduced by 12% when 1.0 cm tumor motion was applied. By contrast, the CTV coverage was maintained when the magnitude of tumor motion was 1.0 cm. The V98% of CTV was reduced by 16% when the largest tumor motion (2.0 cm) was applied. The IMRT QA passing rate was ˜80% when 1.5 cm motion was applied, which corresponds to a 23% and 6% reduction of the PTV and CTV coverage, respectively. Dose to OARs (small bowel, liver and kidneys) were not affected significantly with tumor motion. The general tendency in dose to the OAR was an increase in the low dose region and a decrease in the high dose region as the magnitude of motion increased. PTV coverage and IMRT QA passing rate were reduced significantly with an increase in the magnitude of pancreatic tumor motion. The V95% of PTV was reduced by 12% when 1.0 cm tumor motion was applied. By contrast, the CTV coverage was maintained when the magnitude of tumor motion was 1.0 cm. The V98% of CTV was reduced by 16% when the largest tumor motion (2.0 cm) was applied. The IMRT QA passing rate was ˜80% when 1.5 cm motion was applied, which corresponds to a 23% and 6% reduction of the PTV and CTV coverage, respectively. Dose to OARs (small bowel, liver and kidneys) were not affected significantly with tumor motion. The general tendency in dose to the OAR was an increase in the low dose region and a decrease in the high dose region as the magnitude of motion increased. ConclusionsThis study suggests that CTV coverage of pancreatic tumor can be maintained with IMRT delivery as long as the pancreatic tumor motion is within the PTV. The CTV coverage, however, decreases if the actual tumor motion moves beyond PTV. Therefore, 4D-CT or MRI should be employed to evaluate the extent of pancreatic tumor motion. Instead of a universal margin, the ITV should be used to take into account the pancreatic tumor motion. This study suggests that CTV coverage of pancreatic tumor can be maintained with IMRT delivery as long as the pancreatic tumor motion is within the PTV. The CTV coverage, however, decreases if the actual tumor motion moves beyond PTV. Therefore, 4D-CT or MRI should be employed to evaluate the extent of pancreatic tumor motion. Instead of a universal margin, the ITV should be used to take into account the pancreatic tumor motion.