Department of Nuclear Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands.
Int J Radiat Oncol Biol Phys. 2012 Apr 1;82(5):e725-31. doi: 10.1016/j.ijrobp.2011.11.039. Epub 2012 Feb 11.
[(18)F]-fluorodeoxyglucose-positron emission tomography (FDG-PET) images are usually quantitatively analyzed in "whole-tumor" volumes of interest. Also parameters determined with dynamic PET acquisitions, such as the Patlak glucose metabolic rate (MR(glc)) and pharmacokinetic rate constants of two-tissue compartment modeling, are most often derived per lesion. We propose segmentation of tumors to determine tumor heterogeneity, potentially useful for dose-painting in radiotherapy and elucidating mechanisms of FDG uptake.
In 41 patients with 104 lesions, dynamic FDG-PET was performed. On MR(glc) images, tumors were segmented in quartiles of background subtracted maximum MR(glc) (0%-25%, 25%-50%, 50%-75%, and 75%-100%). Pharmacokinetic analysis was performed using an irreversible two-tissue compartment model in the three segments with highest MR(glc) to determine the rate constants of FDG metabolism.
From the highest to the lowest quartile, significant decreases of uptake (K(1)), washout (k(2)), and phosphorylation (k(3)) rate constants were seen with significant increases in tissue blood volume fraction (V(b)).
Tumor regions with highest MR(glc) are characterized by high cellular uptake and phosphorylation rate constants with relatively low blood volume fractions. In regions with less metabolic activity, the blood volume fraction increases and cellular uptake, washout, and phosphorylation rate constants decrease. These results support the hypothesis that regional tumor glucose phosphorylation rate is not dependent on the transport of nutrients (i.e., FDG) to the tumor.
[(18)F]-氟脱氧葡萄糖正电子发射断层扫描（FDG-PET）图像通常在“全肿瘤”感兴趣区域进行定量分析。此外，通过动态 PET 采集确定的参数，如 Patlak 葡萄糖代谢率（MR(glc))和双组织室模型的药代动力学速率常数，通常是每个病变确定的。我们提出对肿瘤进行分割，以确定肿瘤异质性，这对于放射治疗的剂量描绘和阐明 FDG 摄取机制可能很有用。
在 41 名患者的 104 个病变中，进行了动态 FDG-PET 检查。在 MR(glc)图像上，肿瘤按背景减去最大 MR(glc)的四分位数（0%-25%、25%-50%、50%-75%和 75%-100%）进行分割。使用最高 MR(glc)的三个段中的不可逆双组织室模型进行药代动力学分析，以确定 FDG 代谢的速率常数。
从最高到最低四分位数，摄取（K(1)）、洗脱（k(2)）和磷酸化（k(3)）速率常数显著降低，组织血容量分数（V(b)）显著增加。
MR(glc)最高的肿瘤区域具有较高的细胞摄取和磷酸化速率常数，相对较低的血容量分数。在代谢活性较低的区域，血容量分数增加，细胞摄取、洗脱和磷酸化速率常数降低。这些结果支持了这样一种假设，即区域肿瘤葡萄糖磷酸化速率不依赖于营养物质（即 FDG）向肿瘤的运输。
