Centre for Molecular Imaging, Peter MacCallum Cancer Centre, East Melbourne, Victoria, Australia.
Int J Radiat Oncol Biol Phys. 2013 Jul 15;86(4):749-54. doi: 10.1016/j.ijrobp.2013.02.030. Epub 2013 Apr 16.
The delineation of internal target volumes (ITVs) in radiation therapy of lung tumors is currently performed by use of either free-breathing (FB) (18)F-fluorodeoxyglucose-positron emission tomography-computed tomography (FDG-PET/CT) or 4-dimensional (4D)-CT maximum intensity projection (MIP). In this report we validate the use of 4D-PET-MIP for the delineation of target volumes in both a phantom and in patients.
A phantom with 3 hollow spheres was prepared surrounded by air then water. The spheres and water background were filled with a mixture of (18)F and radiographic contrast medium. A 4D-PET/CT scan was performed of the phantom while moving in 4 different breathing patterns using a programmable motion device. Nine patients with an FDG-avid lung tumor who underwent FB and 4D-PET/CT and >5 mm of tumor motion were included for analysis. The 3 spheres and patient lesions were contoured by 2 contouring methods (40% of maximum and PET edge) on the FB-PET, FB-CT, 4D-PET, 4D-PET-MIP, and 4D-CT-MIP. The concordance between the different contoured volumes was calculated using a Dice coefficient (DC). The difference in lung tumor volumes between FB-PET and 4D-PET volumes was also measured.
The average DC in the phantom using 40% and PET edge, respectively, was lowest for FB-PET/CT (DCAir = 0.72/0.67, DCBackground 0.63/0.62) and highest for 4D-PET/CT-MIP (DCAir = 0.84/0.83, DCBackground = 0.78/0.73). The average DC in the 9 patients using 40% and PET edge, respectively, was also lowest for FB-PET/CT (DC = 0.45/0.44) and highest for 4D-PET/CT-MIP (DC = 0.72/0.73). In the 9 lesions, the target volumes of the FB-PET using 40% and PET edge, respectively, were on average 40% and 45% smaller than the 4D-PET-MIP.
A 4D-PET-MIP produces volumes with the highest concordance with 4D-CT-MIP across multiple breathing patterns and lesion sizes in both a phantom and among patients. Freebreathing PET/CT consistently underestimates ITV when compared with 4D PET/CT for a lesion affected by respiration.
目前，在肺癌的放射治疗中，内部靶区（ITV）的描绘是通过使用自由呼吸（FB）（18）F-氟脱氧葡萄糖正电子发射断层扫描-计算机断层扫描（FDG-PET/CT）或四维（4D）-CT 最大强度投影（MIP）来实现的。在本报告中，我们验证了 4D-PET-MIP 在体模和患者中用于靶区描绘的适用性。
准备了一个带有 3 个空心球体的体模，周围是空气和水。球体和水背景用（18）F 和放射性对比剂的混合物填充。使用可编程运动装置，对体模进行了 4 种不同呼吸模式的 4D-PET/CT 扫描。纳入了 9 名接受 FB 和 4D-PET/CT 检查且肿瘤运动超过 5mm 的 FDG 阳性肺癌患者进行分析。在 FB-PET、FB-CT、4D-PET、4D-PET-MIP 和 4D-CT-MIP 上，使用 2 种勾画方法（最大 40%和 PET 边缘）对 3 个球体和患者病变进行勾画。使用 Dice 系数（DC）计算不同勾画体积之间的一致性。还测量了 FB-PET 和 4D-PET 体积之间的肺肿瘤体积差异。
体模中使用 40%和 PET 边缘的平均 DC 分别为 FB-PET/CT 最低（DCAir = 0.72/0.67，DCBackground = 0.63/0.62），4D-PET/CT-MIP 最高（DCAir = 0.84/0.83，DCBackground = 0.78/0.73）。9 名患者中使用 40%和 PET 边缘的平均 DC 也分别为 FB-PET/CT 最低（DC = 0.45/0.44），4D-PET/CT-MIP 最高（DC = 0.72/0.73）。在 9 个病变中，FB-PET 使用 40%和 PET 边缘的靶区体积平均分别比 4D-PET-MIP 小 40%和 45%。
4D-PET-MIP 在多个呼吸模式和病变大小的体模和患者中，与 4D-CT-MIP 具有最高的一致性。与 4D-PET/CT 相比，自由呼吸 PET/CT 对受呼吸影响的病变始终低估 ITV。
