Motion and Deformation of the Target Volumes During IMRT for Cervical Cancer: What Margins Do We Need?

Abstract

Purpose/Objective(s)For cervical cancer patients the clinical target volume (CTV) consists of an assembly of structures such as the uterus, the cervix, the upper part of the vagina, parametria and the gross tumor volume (GTV). As these structures can move relative to each other, and change in volume, the changes in the encompassing CTV can be complex. The purpose of this study is to use weekly MR imaging of these patients to define margins for the PTV that accommodate these changes.Materials/MethodsTwenty patients with cervical cancer underwent an MRI exam before and weekly during IMRT. T2-weighted sagittal and axial scans were made on a 1.5 Tesla MRI scanner. This resulted in five sets of images per patient. The CTV, GTV and surrounding organs were delineated on all sets. In order to derive PTV margins to accommodate the complex changes in the GTV and CTV, we focused on boundaries of the GTV and CTV in the superior, inferior, right lateral, left lateral, posterior and inferior direction. The shift of these boundaries relative to the pre-treatment MRI exam was determined and correlated with changes in the volumes of the GTV, CTV and the volume of the bladder, rectum, sigmoid and bowel.ResultsThe largest shifts of both the GTV and CTV were in the anterior and posterior direction. To accommodate 95% of all changes during IMRT, an inhomogeneous margin around the GTV of 12, 14, 12, 11, 4 and 8 mm to the anterior, posterior, right lateral, left lateral, superior and inferior direction would have been needed. Around the CTV even larger margins of 24, 17, 12, 16, 11 and 8 mm apply. The posterior and inferior shift of the GTV and CTV were significantly correlated to the rectal volume before treatment (p < 0.001). No significant correlations were found with bladder, sigmoid and bowel volumes or the changes in the volumes of the GTV and CTV.ConclusionsWe used weekly MRI scans during IMRT to derive inhomogeneous PTV margins that accommodate changes in GTV and CTV. The rectal volume before treatment correlates with shifts in the posterior and inferior direction. This could be taken into consideration in defining the margins. The absence of correlation of bladder filling with shifts seems to indicate that measures to control bladder filling are not likely to be effective. Purpose/Objective(s)For cervical cancer patients the clinical target volume (CTV) consists of an assembly of structures such as the uterus, the cervix, the upper part of the vagina, parametria and the gross tumor volume (GTV). As these structures can move relative to each other, and change in volume, the changes in the encompassing CTV can be complex. The purpose of this study is to use weekly MR imaging of these patients to define margins for the PTV that accommodate these changes. For cervical cancer patients the clinical target volume (CTV) consists of an assembly of structures such as the uterus, the cervix, the upper part of the vagina, parametria and the gross tumor volume (GTV). As these structures can move relative to each other, and change in volume, the changes in the encompassing CTV can be complex. The purpose of this study is to use weekly MR imaging of these patients to define margins for the PTV that accommodate these changes. Materials/MethodsTwenty patients with cervical cancer underwent an MRI exam before and weekly during IMRT. T2-weighted sagittal and axial scans were made on a 1.5 Tesla MRI scanner. This resulted in five sets of images per patient. The CTV, GTV and surrounding organs were delineated on all sets. In order to derive PTV margins to accommodate the complex changes in the GTV and CTV, we focused on boundaries of the GTV and CTV in the superior, inferior, right lateral, left lateral, posterior and inferior direction. The shift of these boundaries relative to the pre-treatment MRI exam was determined and correlated with changes in the volumes of the GTV, CTV and the volume of the bladder, rectum, sigmoid and bowel. Twenty patients with cervical cancer underwent an MRI exam before and weekly during IMRT. T2-weighted sagittal and axial scans were made on a 1.5 Tesla MRI scanner. This resulted in five sets of images per patient. The CTV, GTV and surrounding organs were delineated on all sets. In order to derive PTV margins to accommodate the complex changes in the GTV and CTV, we focused on boundaries of the GTV and CTV in the superior, inferior, right lateral, left lateral, posterior and inferior direction. The shift of these boundaries relative to the pre-treatment MRI exam was determined and correlated with changes in the volumes of the GTV, CTV and the volume of the bladder, rectum, sigmoid and bowel. ResultsThe largest shifts of both the GTV and CTV were in the anterior and posterior direction. To accommodate 95% of all changes during IMRT, an inhomogeneous margin around the GTV of 12, 14, 12, 11, 4 and 8 mm to the anterior, posterior, right lateral, left lateral, superior and inferior direction would have been needed. Around the CTV even larger margins of 24, 17, 12, 16, 11 and 8 mm apply. The posterior and inferior shift of the GTV and CTV were significantly correlated to the rectal volume before treatment (p < 0.001). No significant correlations were found with bladder, sigmoid and bowel volumes or the changes in the volumes of the GTV and CTV. The largest shifts of both the GTV and CTV were in the anterior and posterior direction. To accommodate 95% of all changes during IMRT, an inhomogeneous margin around the GTV of 12, 14, 12, 11, 4 and 8 mm to the anterior, posterior, right lateral, left lateral, superior and inferior direction would have been needed. Around the CTV even larger margins of 24, 17, 12, 16, 11 and 8 mm apply. The posterior and inferior shift of the GTV and CTV were significantly correlated to the rectal volume before treatment (p < 0.001). No significant correlations were found with bladder, sigmoid and bowel volumes or the changes in the volumes of the GTV and CTV. ConclusionsWe used weekly MRI scans during IMRT to derive inhomogeneous PTV margins that accommodate changes in GTV and CTV. The rectal volume before treatment correlates with shifts in the posterior and inferior direction. This could be taken into consideration in defining the margins. The absence of correlation of bladder filling with shifts seems to indicate that measures to control bladder filling are not likely to be effective. We used weekly MRI scans during IMRT to derive inhomogeneous PTV margins that accommodate changes in GTV and CTV. The rectal volume before treatment correlates with shifts in the posterior and inferior direction. This could be taken into consideration in defining the margins. The absence of correlation of bladder filling with shifts seems to indicate that measures to control bladder filling are not likely to be effective.