【作者】 张艳；

【导师】 伍兵；

【作者基本信息】 四川大学 ， 影像医学与核医学（放射）（专业学位）， 2022， 硕士

【摘要】 目的:按照2019年最新版的WHO（5th）胃肠/肝胆胰神经内分泌肿瘤的分类分级标准,结合临床分析直肠神经内分泌肿瘤（rectal neuroendocrine neoplasms,R-NENs）的磁共振成像（magnetic resonance imaging,MRI）特征,并初步探讨MRI特征与病理分级的相关性。材料与方法:回顾性分析四川大学华西医院2010年4月-2021年10月经术后样本病理或内镜活检病理证实的R-NENs患者的临床、病理以及高分辨直肠MRI图像资料,经过纳入以及排除标准,最终纳入患者41例。根据2019年最新版的WHO（5th）胃肠/肝胆胰神经内分泌肿瘤的分类分级标准,本研究将41例患者分为两组,分别为高分化的直肠神经内分泌肿瘤（rectal neuroendocrine tumours,R-NETs）组以及低分化的直肠神经内分泌癌（rectal neuroendocrine carcinomas,R-NECs）组。高分化的R-NETs组包括R-NETs G1级、R-NETs G2级以及R-NETs G3级。在高分辨直肠MRI图像上评估肿瘤位置、大小、固有肌层受累情况、边界、形态、均质性、T1WI、T2WI、ADC值以及盆腔淋巴结短轴直径等情况。使用单因素分析来分析两组间是否有统计学差异。使用受试者工作特性曲线（receiver operating characteristic curve,ROC）来寻找ADC（apparent dispersion coefficient,表观弥散系数）值、肿瘤大小以及盆腔淋巴结短轴直径对于鉴别R-NETs与R-NECs的最佳截断值,以及相应的敏感度和特异度。结果:本研究共纳入41例R-NENs患者,高分化的R-NETs组患者19例（R-NETs G1级患者11例、R-NETs G2级患者7例、R-NETs G3级患者1例）,低分化的R-NECs组患者22例。41例R-NENs的患者中,男性26例,女性15例;平均年龄为52.7±13.9岁。R-NENs的平均肿瘤大小为2.7（2.5）cm,平均盆腔淋巴结短轴直径为0.6（0.4）cm,肿瘤平均ADC值为0.97±0.23x10^-3mm²/s。ADC值等于0.959x10^-3mm²/s时,在鉴别高分化的R-NETs与低分化的R-NECs有较高的敏感度（91%）和特异度（84%）,AUC=0.935。肿瘤大小等于2.85cm时,在鉴别高分化的R-NETs与低分化的R-NECs有较高的特异度（95%）,但敏感度一般（77%）,AUC=0.921。盆腔淋巴结短轴直径等于0.45cm时,在鉴别高分化的R-NETs与低分化的R-NECs有较高的敏感度（100%）,但特异度差（58%）,AUC=0.837。两组间的直肠壁固有肌层是否受累差异无统计学意义（P=0.082）。与高分化的R-NETs相比,低分化的R-NECs一般边界不清楚、形态不规则、信号不均匀（坏死囊变）、T1WI呈稍低信号、T2WI呈稍高信号。本研究22例患者有MRI动态增强图像,包括7例R-NETs G1级患者、3例R-NETs G2级患者以及12例R-NECs患者。分化越好的肿瘤其强化程度越明显,分化较差的肿瘤一般呈轻中度强化。肝转移可发生在任何级别的R-NENs患者中,R-NETs G1级患者也可出现肝转移（6例患者发生肝转移,包括1例R-NETs G1级患者、3例R-NETs G2级患者以及2例R-NECs患者）。R-NENs患者多数有便血的临床症状,且更容易出现在低分化的R-NECs患者中。R-NETs G1级患者多数为临床偶然发现,一般无临床症状。结论:ADC值相较于肿瘤大小以及盆腔淋巴结短轴直径而言,更有助于鉴别高分化的R-NETs和低分化的R-NECs。此外,肿瘤边界、形态、均质性、T1WI、T2WI等对于鉴别高分化的R-NETs和低分化的R-NECs也有一定的临床意义。肝转移可发生在任何级别的R-NENs患者中,R-NETs G1级的患者也可能出现肝转移。更多还原

【Abstract】 Objective:With the latest WHO（5th）classification and grading criteria for neuroendocrine neoplasms（NENs）of the gastrointestinal tract and hepatopancreatobiliary organs in2019,to analyze the magnetic resonance imaging（MRI）features of R-NENs,and preliminarily discuss the correlation with pathological grading.Materials and Methods:The clinical,pathological and rectal MRI data of patients with pathologically confirmed R-NENs in West China Hospital of Sichuan University from April 2010to October 2021 were retrospectively analyzed.After inclusion and exclusion criteria,41 patients were finally included.According to the latest WHO（5th）classification and grading criteria for NENs of the gastrointestinal tract and hepatopancreatobiliary organs in 2019,in this study,41 patients were divided into two groups:the well-differentiated rectal neuroendocrine tumours（R-NETs）group and the poorly-differentiated neuroendocrine carcinomas（R-NECs）group.The well-differentiated R-NETs group includes R-NETs G1,R-NETs G2 and R-NETs G3.MRI assessment of tumor location,size,tumor infiltration depth,boundary,morphology,homogeneity,T1WI,T2WI,ADC value,short axis diameter of pelvic lymph nodes and other conditions were considered.Univariate analysis was used to analyze whether there was significant difference between the two groups.Receiver operating characteristic（ROC）curve was used to find the best cut-off value of ADC value,tumor size and short axis diameter of pelvic lymph nodes for distinguishing R-NETs from R-NECs,as well as the corresponding sensitivity and specificity.Results:A total of 41 patients with R-NENs were included in this study.There were 19patients in the well-differentiated R-NETs group（11 patients with R-NETs G1,7patients with R-NETs G2 and 1 patient with R-NETs G3）and 22 patients in the poorly-differentiated R-NECs group.26 males and 15 females,with an average age of 52.7±13.9 years.The average tumor size was 2.7（2.5）cm,the average short axis diameter of pelvic lymph nodes was 0.6（0.4）cm,and the average ADC value was0.97±0.23x10^-3mm²/s.ROC curve analysis results showed that when using ADC value of 0.959x10^-3mm²/s as a cut-off value,it showed a sensitivity of 91%,specificity of 84%,AUC=0.935;when using tumor size of 2.85cm as a cut-off value,it showed a sensitivity of 95%,specificity of 77%,AUC=0.921;when using When the short axis diameter of pelvic lymph nodes of 0.45cm as a cut-off value,it showed a sensitivity of 100%,specificity of 58%,AUC=0.837.There was no significant difference between the two groups in whether the muscularis propria of rectal wall was involved（P=0.082）.Compared with the well-differentiated R-NETs,R-NECs generally have unclear boundary,irregular shape,the signal intensity of MRI was not homogeneous（necrosis and cystic transformation）,slightly lower signal on T1WI and slightly higher signal on T2WI.Enhanced images can be observed in 22 patients,including 7 patients with R-NETs G1,3 patients with R-NETs G2 and 12 patients with R-NECs.The better the differentiation of the tumors,the more obvious the enhancement degree,and the poorly differentiated tumors generally show mild to moderate enhancement.Liver metastasis can occur in R-NENs of any grade.Liver metastasis can also occur in patients with R-NETs G1（liver metastasis occurred in 6patients,including 1 patient with R-NETs G1,3 patients with R-NETs G2 and 2patients with R-NECs）.Most patients with R-NENs have hematochezia,and it is more likely to occur in patients with R-NECs.Most patients with R-NETs G1 are mostly clinically discovered accidentally,and generally have no clinical symptoms.Conclusion:Compared with tumor size and short axis diameter of pelvic lymph nodes,ADC value is helpful to distinguish the well-differentiated R-NETs from the poorly-differentiated R-NECs.In addition,tumor boundary,morphology,homogeneity,T1WI and T2WI also have certain clinical significance for distinguishing the well-differentiated R-NETs from the poorly-differentiated R-NECs.更多还原