动态增强CT在胰腺神经内分泌肿瘤病理分级中的可行性研究

The feasibility study of dynamic contrast-enhanced CT in pathological grading of pancreatic neuroendocrine tumors

目的:探讨双期动态增强CT鉴别不同病理级别胰腺神经内分泌肿瘤(pNET)的可行性。方法:回顾性分析50例经手术病理证实的pNET的CT表现,观察病变的位置、最大径、包膜、钙化、囊变坏死、胰胆管扩张,有无周围淋巴结肿大、血管侵犯和肝转移。分别测量平扫、动脉期、静脉期图像上肿瘤病灶和病灶周围正常胰腺实质的平均CT值,计算肿瘤的平扫相对密度指数、动脉期相对强化指数、静脉期相对强化指数。根据2010年WHO消化道神经内分泌肿瘤分类方法将50例肿瘤分为两组:胰腺神经内分泌瘤(G1、G2级肿瘤,记为G1/2级)和胰腺神经内分泌癌(G3级肿瘤)。比较两组肿瘤的CT影像表现差异。结果:50例pNET均为单发病灶。G1、G2、G3级pNET分别为21、22和7例。胰腺神经内分泌瘤(G1/2级)43例,胰腺神经内分泌癌(G3级)7例。G1/2和G3级肿瘤的病灶位置、囊变坏死和钙化差异均无统计学意义(P均>0.05)。G1/2、G3级肿瘤的平均最大径分别为(27.2±15.9)和(35.6±16.4)mm,随着病理级别的增高,肿瘤的平均最大径有增大趋势,但两者差异无统计学意义(P>0.05)。两组肿瘤的包膜、胰胆管扩张、周围淋巴结肿大、血管侵犯的差异均具有统计学意义(P分别为0.001,0.001,0.002,0.029),肝转移差异无统计学意义(P=0.370)。G1/2级肿瘤的动脉期相对强化指数(1.33±0.27)和静脉期相对强化指数(1.31±0.26)均较G3级肿瘤高(分别为0.96±0.16、1.03±0.14),且差异均具有统计学意义(P均<0.05)。动脉期相对强化指数、静脉期相对强化指数鉴别不同病理级别pNET的ROC曲线分析:两者的曲线下面积分别为0.897、0.817,动脉期相对强化指数较静脉期相对强化指数能更好地鉴别G1/2和G3级pNET。结论:双期动态增强CT有助于G1/2级胰腺神经内分泌瘤和G3级胰腺神经内分泌癌的鉴别诊断。

Objective：To identify the feasibility of dual-phase dynamic contrast-enhanced CT in pathological grading of pancreatic neuroendocrine tumors. Methods：The abdominal plain CT and dual- phase contrast-enhanced CT images of 50 lesions in 50 patients with surgically and pathologically con- firmed pancreatic neuroendocrine tumors （pNET） were studied retrospectively. CT image analysis in- cluded tumor location, maximum size, capsule, calcification, cystic necrosis, pancreaticobiliary duct dila tation, peripheral lymphadenopathy, vascular invasion and hepatic metastasis. The CT values of the tumors and normal pancreatic parenchyma in plain, arterial and venous phase were measured respec tively, then the mean relative density index, arterial relative enhancement index,and portal relative en- hancement index were calculated. According to the WHO 2010 pathological classfication, all tumors were divided into 2 groups：pancreatic neuroendocrine tumor （grade 1 and grade 2,G1/2） and pancre- atic neuroendocrine carcinoma （grade 3,G3）. The CT features of tumors were compared between 2 groups. Results：All 50 cases were single lesions,in which 21 lesions were grade 1,22 lesions were grade 2 and 7 lesions were grade 3. There were 43 cases of pancreatic neuroendocrine tumor （G1/2 grade） in group 1 and 7 cases of pancreatic neuroendocrine carcinoma （G3 grade） in group 2. There was no significant difference in tumor location, cystic change or necrosis and calcification between group 1 and group 2 （P〉0.05）. Although the tumor maximum size was larger in group 2 （35.6±16.4） than group 1 （27.2±15.9mm） ,the difference in maximum size between two groups demonstra- ted no statistical significance. There were statistically significant differences in tumor capsule, pancre- aticobiliary duct dilatation,lymphadenopathy and vascular invasion between group 1 and group 2 （P= 0. 001,0. 001,0. 002,0. 029 respectively）, and no significant difference in hepatic metastasis （P = 0. 370）. The arterial relative enhancement index and portal relative enhancement index were larger in group 1 （1.33±0.27 and 1.31±0.26） than those in group 2 （0.96±0.16 and 1.03±0.14） ,showing a significant difference between the two groups （P〈0.05）. In the ROC curve,the arterial relative en- hancement index was more sensitive than the portal relative enhancement index in differentiating group 2 from group 1, as the area under the ROC curve of the arterial relative enhancement index （0. 897） was larger than that of the portal relative enhancement index （0. 817）. Conclusion. Dual-phase dynamic contrast-enhanced CT is helpful in the differential diagnosis of G1/2 pancreatic neuroendo- crine tumor from G3 pancreatic neuroendocrine carcinoma.