University College Hospitals NHS Foundation Trust, London, UK.
Clin Oncol (R Coll Radiol). 2011 Jun;23(5):364-71. doi: 10.1016/j.clon.2010.11.001. Epub 2010 Nov 24.
This study investigated an adaptive threshold-based method to delineate the target volume using (18)fluoro-2-deoxyglucose ((18)FDG) positron emission tomography/computed tomography (PET/CT) before and during a course of radical radiotherapy or chemoradiotherapy in locally advanced squamous cell carcinoma of the head and neck.
Ten patients were enrolled between March 2006 and May 2008. (18)FDG PET/CT scans were carried out 72h before the start of radiotherapy and then at three time points during radiotherapy (8-18, 36-50 and 66Gy). Functional volumes were delineated using an adaptive iterative algorithm weighted according to the mean standard uptake value (SUV(mean)) within the region of interest. The background (18)FDG uptake, maximum standard uptake value (SUV(max)) and SUV(mean) within the volumes were assessed.
There was no significant reduction in the primary target volumes defined by the adaptive threshold during radiotherapy. However, the SUV(max) significantly reduced within the primary (P=0.003-0.011) and lymph node (P<0.0001) target volume at 36-50 and 36-66Gy compared with 0Gy. The SUV(mean) was negatively correlated to radiation dose (P<0.0001-0.014). The ratio between the background uptake of (18)FDG and the SUV(mean) significantly reduced for both the lymph node target volume at 36-50Gy and the primary volume at 66Gy. The lack of significant correlation between the defined volume and radiation dose was because the SUV(mean) within the region of interest used to define the edge of the volume was equal to or less than the background (18)FDG uptake and the software was unable to effectively differentiate between tumour and background uptake.
The adaptive threshold method may be of benefit when used to define the target volume before the start of radiotherapy. This method was not beneficial during radiotherapy because the software is not sensitive enough to distinguish tumour from background and define a volume. (18)FDG PET/CT-guided volumes delineated by automatic adaptive thresholding methods should only be used for dose escalation with the pretreatment imaging.
本研究旨在探讨一种基于自适应阈值的方法,用于在根治性放化疗或放化疗过程中勾画局部晚期头颈部鳞状细胞癌的靶区,该方法使用(18)氟-2-脱氧葡萄糖((18)FDG)正电子发射断层扫描/计算机断层扫描(PET/CT)。
2006 年 3 月至 2008 年 5 月期间,共纳入 10 例患者。在放疗开始前 72h 进行(18)FDG PET/CT 扫描,然后在放疗期间的三个时间点(8-18Gy、36-50Gy 和 66Gy)进行扫描。功能体积通过根据感兴趣区域内的平均标准摄取值(SUV(mean))加权的自适应迭代算法进行描绘。评估背景(18)FDG 摄取、最大标准摄取值(SUV(max))和体积内的 SUV(mean)。
在放疗过程中,自适应阈值定义的主要靶区体积没有明显减少。然而,与 0Gy 相比,在 36-50Gy 和 36-66Gy 时,主要靶区(P=0.003-0.011)和淋巴结靶区(P<0.0001)内的 SUV(max)显著降低。SUV(mean)与辐射剂量呈负相关(P<0.0001-0.014)。在 36-50Gy 时,淋巴结靶区的背景摄取与 SUV(mean)的比值以及 66Gy 时的原发肿瘤体积的比值显著降低。由于用于定义体积边缘的感兴趣区域内的 SUV(mean)等于或小于背景(18)FDG 摄取,软件无法有效区分肿瘤和背景摄取,因此定义体积与辐射剂量之间缺乏显著相关性。
在放疗前,自适应阈值方法可能有助于定义靶区。然而,在放疗过程中,该方法并没有带来益处,因为软件对区分肿瘤和背景以及定义体积不够敏感。(18)FDG PET/CT 引导的自动自适应阈值勾画的体积应仅用于预治疗成像时的剂量升级。