基于PET-CT的SUV值与基于4DCT的EE时相勾画胸段食管癌GTV相关性分析

目的 探讨基于PET-CT图像SUV阈值≥2.0及20%SUV_max与基于4DCT的EE时相图像勾画胸段食管癌原发肿瘤GTV相关性因素。方法 22例胸段食管癌患者序贯完成3DCT、4DCT、FDG PET-CT 胸部定位扫描。基于4DCT的EE时相图像勾画GTV_50%。基于SUV≥2.0、20%SUV_max分别在PET图像上勾画IGTV_PET并分别命名为IGTV_PET2.0、IGTV_PET20%。获得GTV_50%最大横径、GTV_50%大小、上下方向位移、三维运动矢量和SUV_max。结果 IGTV_PET2.0、IGTV_PET20%与GTV_50%间体积比与GTV_50%最大横径、GTV_50%大小、上下方向位移、三维运动矢量均无相关性(P=0.055~0.932);IGTV_PET2.0、IGTV_PET20%与GTV_50%间CI与GTV_50%最大横径、GTV_50%大小、上下方向位移、三维运动矢量均有相关性(P=0.005~0.033);IGTV_PET20%与GTV_50%间体积比、CI与SUV_max均有相关性(P=0.001、0.016)。结论 基于PET-CT图像构建的IGTV并不能客观真实反映肿瘤空间位置变化及运动信息,而且单一数值的SUV阈值选取也是不可靠的。构建食管癌原发肿瘤靶区时应依据4DCT所构建IGTV纠正PET-CT所构建IGTV边界及其位置。

Analysis of correlation between GTVs delineated based on SUVs of PET-CT and EE phase of 4DCT in thoracic esophageal cancer

Objective To investigate the correlation factors in delineating the gross tumor volume (GTV) of primary tumor in thoracic esophageal cancer based on the positron emission tomography-computed tomography (PET-CT)(thresholds of standardized uptake value or SUV ≥2.0 and ≥20% of maximum SUV or SUV_max) and the end-expiration (EE) phase of four-dimensional computed tomography (4DCT). Methods Twenty-two patients with thoracic esophageal cancer sequentially underwent 3DCT, 4DCT, and F-fluorodeoxyglucose PET-CT for thoracic localization. GTV_50% was delineated by the EE phase of 4DCT. The internal GTVs based on PET-CT images (IGTV_PET) were delineated by two thresholds:SUV ≥2.0 and ≥20% of SUV_max, and designated as IGTV_PET2.0 and IGTV_PET20%, respectively. The maximum transverse diameter of GTV_50%, volume of GTV_50%, the displacement in cranial-caudal (CC) direction, 3D motion vector, and SUV_max were obtained. Results The maximum transverse diameter of GTV_50%, volume of GTV_50%, the displacement in CC direction, and 3D motion vector showed no significant correlations with the volume ratio (VR) of IGTV_PET20% or IGTV_PET2.0 to GTV_50%(P=0.055- 0.932), but showed significant correlations with the conformity index (CI) between IGTV_PET20% or IGTV_PET2.0 and GTV_50%(P=0.005- 0.033). Significant correlations were observed between the VR and CI of IGTV_PET20% to GTV_50% and SUV_max(P=0.001 and 0.016, respectively). Conclusions IGTV based on PET-CT images cannot objectively and accurately reflect the changes in spatial position and motion of the tumor, and the selection of a single SUV threshold is unreliable. To determine the primary tumor target in esophageal cancer, the IGTV based on 4DCT should be used to correct the boundary and position of IGTV constructed based on PET-CT.