关键词:
Brain metastasis
Dynamic PET
Glioma
Kinetic modeling
PET-MRI
摘要:
目的: 探讨l8F-氟代脱氧葡萄糖(18F-FDG)动态PET-MRI在鉴别治疗后胶质瘤的肿瘤进展(TP)及治疗后改变(PTRC)的作用。 方法: 横断面研究。纳入2022年7月至2024年9月在中山大学肿瘤防治中心可疑复发/进展的治疗后胶质瘤患者,共15例,其术后放疗后进行18F-FDG PET-MRI检查。从静态的PET图像提取所有病灶的标准化摄取值(SUV)和靶病灶背景比值(TBR),包括最大SUV(SUVmax)、SUV均值(SUVmean)、血糖水平校正最大SUV(SUVgluc max)、血糖水平校正平均SUV(SUVgluc mean)、靶病灶灰质背景比值(TBRgm)及靶病灶白质背景比值(TBRwm)。对所有病灶时间-活动曲线(TAC)进行双组织隔室模型(2TCM)拟合和Patlak图形分析,以获取动力学参数包括速率常数K1、k2、k3、Ki以及氟代脱氧葡萄糖代谢速率(MRFDG)。分析不同代谢参数在TP和PTRC组间分布差异,以及通过绘制受试者工作特征(ROC)曲线评估不同参数区分TP和PTRC的诊断性能。 结果: 15例患者年龄[M(Q1,Q3)]为48(41,54)岁,其中男5例(33.3%),在增强MRI图像上共检出18个病灶,最终共13个病灶确定进展。18F-FDG PET-MRI多参数分析显示TP的2TCM-Ki[0.018(0.015,0.022)比0.013(0.012,0.014),P=0.016]、2TCM-MRFDG[8.334(7.041,9.836)比6.281(5.713,7.469),P=0.046]、Patlak-Ki[0.017(0.014,0.020)比0.010(0.007,0.011),P=0.004]及Patlak-MRFDG[7.742(6.904,9.084)比4.892(3.833,5.640),P=0.003]高于PTRC。在检测肿瘤方面Patlak-Ki[曲线下面积(AUC)为0.954(95%CI:0.856~1.000)],灵敏度为92.3%,特异度为100.0%。 结论:18F-FDG PET-MRI提供的代谢动力学参数Patlak-Ki作为一种无创生物标志物,在胶质瘤治疗后鉴别TP和PTRC方面具有良好的性能。.;Objective: To investigate dynamic positron emission tomography/magnetic resonance (dPET-MRI) using 18F-deoxyglucose (FDG) in distinguishing tumor progression (TP) from post treatment related change (PTRC) in patients with glioma. Methods: This study is a cross-sectional study. From July 2022 to September 2024, a total of 15 suspected recurrent/progressing post-treatment glioma patients were recruited at the Cancer Prevention and Treatment Center of Sun Yat sen University. They underwent 18F-FDG PET-MRI examination after postoperative radiotherapy in terms of static parameters, the SUVmax, SUVmean of all lesions and SUVmean of normal brain (gray matter and white matter) were measured. Additionally, target-to-background ratios (TBRgm and TBRwm) and SUV corrected for blood glucose (SUVgluc max and SUVgluc mean)were calculated. The two-tissue compartment model (2TCM) and Patlak plot were used to derive kinetic metrics, including K1, k2, k3, 2TCM-Ki, 2TCM-MRFDG, Patlak-Ki and Patlak-MRFDG. Receiver operating characteristic (ROC)