基于临床和实验室指征的脑膜瘤/胶质瘤Nomogram鉴别诊断模型建立

Establishment of a Nomogram differential diagnosis model for meningioma/glioma based on clinical and laboratory indications

目的建立本地区脑膜瘤和胶质瘤鉴别诊断的Nomogram图。方法2011年1月—2021年4月在本院住院手术的脑膜瘤患者98例,胶质瘤患者35例,其中男性56例,女性77例,年龄为20岁~86岁,平均年龄为(58.67±13.13)岁。收集临床指征,采集血浆检测血浆凝血酶原时间(PT)、部分活化凝血酶时间(APTT)等。免疫组化检测病理组织的孕激素受体(PR)、上皮膜抗原(EMA)、中枢神经特异蛋白(S-100)、细胞角蛋白(CK),神经胶质纤维酸性蛋白(GFAP)、增殖指数(Ki67)、波形蛋白(VIM)等。Logistic回归模型筛选独立风险因子,通过Nomogram图进行可视化输出。结果多因素Logistic回归结果显示,风险因子分别是性别(OR=15.645,P<0.05)、APTT(OR=0.731,P<0.05)、PR(OR=0.049,P<0.05)、EMA(OR=0.117,P<0.05)、GFAP(OR=74.886,P<0.01)、S-1000(OR=73.166,P<0.01)。结论建立得到一种简单又可操作性强的预测胶质瘤的临床积分系统,有助于胶质瘤和脑膜瘤的早期鉴别诊断。

Objective To establish a Nomogram spectrum for differential diagnosis of meningioma and glioma in this area.Methods From January 2011 and April 2021,98 patients with meningioma and 35 patients with glioma were collected.There were 56 males and 77 females enrolled with an average age of 58.67±13.13.Clinical indications were collected and plasma was collected to detect plasma prothrombin time(PT)and partially activated thrombin time(APTT).Immunohistochemistry was used to detect progesterone receptor(PR),epithelial membrane antigen(EMA),central nervous specific protein(S-100),cytokeratin(CK),glial fibrillary acidic protein(GFAP),proliferation index(Ki67),vimentin(VIM)in pathological tissues.Independent risk factors were screened by Logistic regression model and visualized by Nomogram.Results Multivariate Logistic regression showed that the risk factors were gender(OR=15.645,P<0.05),APTT(OR=0.731,P<0.05),PR(OR=0.049,P<0.05),EMA(OR=0.117,P<0.05),GFAP(OR=74.886,P<0.01),S-1000(OR=73.166,P<0.01).Conclusion The establishment of a simple and operable clinical score system for glioma prediction is helpful for the early differential diagnosis of glioma and meningioma.