胶质母细胞瘤术后早期复发与假性进展的鉴别诊断模型

发布时间：2018-06-12 23:12

本文选题：胶质母细胞瘤 + 假性进展　；参考：《中南大学》2014年博士论文

【摘要】：目的：胶质母细胞瘤是最常见和病死率最高的成人颅内肿瘤,其治疗方法主要采用手术和放化疗等多种治疗方式相结合的综合治疗。即使是接受综合治疗的患者,其预后也未达到令人满意的效果。在接受综合治疗的胶质母细胞瘤患者中,部分患者在同步放化疗结束后12周内复查颅脑磁共振时发现影像学早期进展。影像学早期进展的性质可能为早期复发,亦可能为假性进展。而早期复发和假性进展在治疗方案和预后方面存在显著差异。目前,仅通过即时影像学方法无法有效鉴别两者；而再次开颅手术或立体定向活检病理诊断由于其有创性和局限性并未在临床中广泛开展；在临床实践中,比较可行的鉴别诊断方法是影像学随访,但其需要2个月以上的随访时间,明显影响了患者个体化治疗方案的及时制定。本研究旨在建立及时、准确、便捷的判别模型,鉴别诊断胶质母细胞瘤术后早期复发与假性进展。以期有助于胶质母细胞瘤患者个体化治疗方案的及时制定和预后的提高。方法：本研究为病例对照分析,回顾性收集了47例发生影像学早期进展的胶质母细胞瘤患者的临床资料(包括：神经功能状态变化、手术切除程度、性别、年龄、磁共振检查、类固醇使用剂量、放化疗方案等)；利用首次开颅手术或立体定向活检获取的病理组织,检测相关分子标记(免疫组化方法检测P53表达情况,甲基特异性聚合酶链式反应检测MGMT启动子甲基化状态,限制性片段长度多态性聚合酶链式反应检测IDH1类型)。对于获得的数据,首先运用Logistic回归分析筛选早期复发与假性进展之间存在显著性差异的因素,验证各因素之间的交互作用,并比较各因素效应的相对大小,然后建立基于多因素的Fisher判别模型,以判别发生影像学早期进展的胶质母细胞瘤患者属于早期复发或假性进展。并对判别模型进行系统性评价。结果：胶质母细胞瘤早期复发与假性进展之间存在显著性差异的因素包括：神经功能状态变化(P=0.015)、MGMT启动子甲基化状态(P=0.005)、IDH1类型(P=0.019)。未发现各因素之间存在显著的交互作用。各因素对于判别结果的影响程度由大到小排列为：MGMT启动子甲基化状态(X4)、IDH1类型(X5)、神经功能改变(X1)。Fisher判别函数夕判别界值Z。为-0.151。当判别函数值Z-0.151时,判为假性进展；当判别函数值Z-0.151时,判为早期复发。将研究对象回代入判别模型,有9例误判,误判率为0.19(9/47)。结论：发生影像学早期进展的胶质母细胞瘤患者中,神经功能状态加重的患者更有可能为早期复发；MGMT启动子甲基化的患者更有可能为假性进展；IDH1野生型患者更有可能为早期复发。神经功能状态未加重、MGMT启动子甲基化胶质母细胞瘤患者发生的影像学早期进展极有可能为假性进展；神经功能状态加重、MGMT启动子未甲基化、IDH1野生型胶质母细胞瘤患者发生的影像学早期进展极有可能为肿瘤复发。
[Abstract]:Objective: glioblastoma is the most common and highly fatrate adult intracranial tumor. Its treatment is mainly combined with a combination of surgery, radiotherapy and chemotherapy. The prognosis is not satisfactory even in patients receiving comprehensive treatment. Patients receiving comprehensive treatment of glioblastoma have been treated with a comprehensive treatment. In some patients, the early progress of imaging was found during the reexamination of craniocerebral magnetic resonance (MRI) within 12 weeks after the end of concurrent chemo radiotherapy. The nature of early imaging progress may be an early recurrence and may be a false progress. The early recurrence and false progress have significant differences in the treatment and prognosis. In clinical practice, the more feasible differential diagnosis method is imaging follow-up, but it needs more than 2 months of follow-up, which obviously affects individual treatment program. The purpose of this study is to establish a timely, accurate and convenient discriminant model for the differential diagnosis of the early recurrence and false progression of glioblastoma, in order to contribute to the timely formulation of the individualized treatment scheme for glioblastoma patients and the improvement of the prognosis.
Methods: This study was a case-control analysis. The clinical data of 47 patients with glioblastoma in the early stage of imaging (including nerve function change, surgical excision degree, sex, age, magnetic resonance, steroid use dose, radiotherapy and chemotherapy) were collected, and the first craniotomy or stereotaxic orientation was used. Pathological tissue obtained by biopsy, detection of related molecular markers (immunohistochemical method for detection of P53 expression, methyl specific polymerase chain reaction detection of MGMT promoter methylation status, restrictive fragment length polymorphism polymerase chain reaction detection of IDH1 type). For data obtained, first use of Logistic regression analysis to screen early stage There is a significant difference between the recurrence and the false progression, verifies the interaction between the factors, and compares the relative size of the factors, and then establishes a Fisher discriminant model based on multiple factors to identify the early progression of glioblastoma in the early progression of imaging and to the discriminant model. Systematic evaluation.
Results: there were significant differences between the early recurrence and the false progression of glioblastoma, including the changes of neural function (P=0.015), the methylation status of the MGMT promoter (P=0.005), and the IDH1 type (P=0.019). The arrangement is: MGMT promoter methylation state (X4), IDH1 type (X5), neural function change (X1).Fisher discriminant function Z. is -0.151. when the discriminant function is Z-0.151, and when discriminant function value Z-0.151 is false progress; when discriminant function value Z-0.151, it is found to be early recurrence. The study object is replaced by the discriminant model, there are 9 cases misjudged, misjudgment rate is 0.19 (9/47).
Conclusions: in patients with glioblastoma with early imaging progress, patients with increased neurologic status are more likely to relapse early; patients with MGMT promoter methylation are more likely to be pseudoprogressive; IDH1 wild-type patients are more likely to relapse early. The neurologic state is not aggravated, and the MGMT promoter methylation of glioblastoma is more likely. Early progress in imaging of the patients with cytomatoma is likely to be false progress; neural function is aggravated, MGMT promoter is not methylation, and early progress in imaging of patients with IDH1 wild glioblastoma is likely to be a tumor recurrence.
【学位授予单位】：中南大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R447;R739.41

【相似文献】