合理选择内对照客观评价靶组织摄取能力

Ability of Target Tissue Uptake Can Be Reflected More Objectively by Proper Internal Controls

目的:评价一种更客观反映靶组织摄取能力的内对照方法。方法:在Jaszczak SPECT模型内插入7个灌满[18F]-FDG水溶液(放射性比活度23.7 k Bq/m L)的空心球(热区模型),它们的容积依次为:3.55 m L,5.88 m L,11.74 m L,20.52 m L,21.31 m L,64.56 m L和179.50 m L,同时在Jaszczak SPECT模型内灌满放射性比活度8.1 k Bq/m L的[18F]-FDG水溶液组成一低靶/本底比(2.92)实验模型。在Jaszczak SPECT模型内插入10个灌满[18F]-FDG水溶液(放射性比活度36.33 k Bq/m L)的空心球(热区模型),它们的容积依次为:1.91 m L,3.55 m L,5.67 m L,18.74 m L,18.94 m L,19.61 m L,22.84 m L,59.47 m L,64.56 m L和70.33 m L,同时在Jaszczak SPECT模型内灌满放射性比活度4.02 k Bq/m L的[18F]-FDG水溶液组成一高靶/本底比(9.04)实验模型。通过Discovery ST8 PET/CT(通用电气,美国)依次采集其2D及3D PETCT图像。测量球体及水本底平均标准摄取值(SUVmean),计算靶/本底比、靶/内对照比,与理论值比较。结果:低靶/本底比模型热区模型SUVmean真实值2.68,容积为3.55 m L模型测得的SUVmean为1.84(2D图像)和1.70(3D图像),低于真实值(P<0.01);高靶/本底比模型热区模型SUVmean真实值6.74,容积为1.91 m L,3.55 m L和5.67 m L的模型测得的SUVmean为4.47,4.97,4.78(2D图像)和3.47,4.10,4.29(3D图像),低于真实值(P<0.01)。容积≥5.88 m L球体SUVmean与理论值无统计学差异(P>0.05),靶/本底比低于理论值,有统计学差异(P≤0.02);除高靶/本底比组2D PET图像外,另3组靶/内对照比与理论值无统计学差异(P>0.05)。结论:靶/内对照比更能客观反映靶组织的摄取能力。

Objective To evaluate an internal control for more objective reflection of target tissue uptake ability. Methods Several hollow balls filled with some specific activity [Fluorine-I 8]-fluorodeoxyglucose （[18F]-FDG） solution were fixed on the bottom of a Jaszczak SPECT phantom and used to simulate the human body and tumors for test. A low target-background ratio （2.92） model consisted of seven hollow balls of different shapes and sizes; their volumes were 3.55, 5.88, 11.74, 20.52, 21.31, 64.56 and 179.50 mL. Filled with a specific activity （23.7 kBq/mL） [18F]-FDG, they were fixed on the bottom of the Jaszczak SPECT phantom filled with an unlike specific activity （8.1 kBq/mL） [18F]-FDG. A high target-background ratio （9.04） model was formed with ten hollow balls of different shapes and sizes; their volumes were 1.91, 3.55, 5.67, 18.74, 18.94, 19.61, 22.84, 59.47, 64.56, and 70.33 mL. Filled with a specific activity （36.33 kBq/mL） [IqZ]-FDG, they were fixed on the bottom of the Jaszczak SPECT phantom filled with an unlike specific activity （4.02 kBq/mL） [18F]-FDG. A three-dimensional （3D） PET-CT scan followed by a two-dimensional （2D） PET-CT scan was done for each model using Discovery ST8 PET/CT scanner （GE healthcare, USA）. The average standardized uptake value （SUVmean） of the balls and phantoms were measured. The ball-contrast ratios were computed. The data were analyzed by U-test and Student＇s t-test properly. Results The actual SUVm~ of the low target-background ratio phantom was 2.68. The measured SUV~ of the ball whose volume was 3.55 mL was 1.84 and 1.70 for 2D and 3D, respectively. Both of them were significantly lower than their actual SUVmean （P〈0.01）. Those measured SUVmean, of the remainder six balls in the low target-background ratio phantom were not significantly different from their actual SUVmean （P〉0.05）. The actual SUVmean of balls in the high target-background ratio phantom was 6.74. The measured SUVmean of the balls whose volume were 1.91, 3.55, and 5.67 mL were 4.47, 4.97, 4.78, and 3.47, 4.10, 4.29 for 2D and 3D, respectively. All of them were significantly lower than their actual SUVmean （P〈0.01）. Those measured SUV~, of the remainder seven balls in the high target-background ratio phantom were not significantly different from their actual SUVmean （P〉0.05）. As the volume ≥ 5.88 ml, the ball-phantom ratios were smaller than their actual value （P≤0.02）. Compared with the ball whose volume was 64.75 mL, the ball-contrast ratios of those balls （≥15.88 mL） were not different from their actual value （P〉0.05） except the high target-background ratio model by 2D PET-CT scan （P=-0.04）. Conclusion Target-contrast ratio is better than target-background ratio in reflecting the actual target tissue uptake ability.