背景:放疗后高级别神经胶质瘤中经常发生辐射诱导变化的室管膜下增强(SE)的特定模式经常被忽视。灌注MRI可以提供诊断线索。
目的:区分成人高级别弥漫性神经胶质瘤标准治疗后放射诱发的SE和进展。
方法:回顾性。
方法:94例连续高级别弥漫性神经胶质瘤患者(平均年龄,55±14岁;54[57.4%]男性)在完成手术加放化疗后的随访MRI中发现新的SE:进展(N=74)与回归(N=20)。
■3T,梯度回波动态磁化率对比增强MRI,3D梯度回波对比增强T1加权成像。
结果:为了区分SE评估中的辐射变化和进展,使用SE出现间隔中的显著变量进行多变量逻辑回归,IDH突变,形态特征,和rCBV。进行Cox回归预测肿瘤进展。对于rCBV的附加值,在有和没有rCBV的多变量逻辑回归模型之间进行了对数秩检验.
方法:Logistic回归,Cox回归,接收机工作特性分析,对数秩检验。
结果:38.3%(36/94)的患者出现首次特异性SE(术后9.2±9.5个月),在放射后MRI首次出现后5.8±5.8个月后,21.3%(20/94)消失。IDH突变,细长的,rCBV较低的小病变倾向于消退:IDH突变,伸长率,直径,和rCBV_p95;赔率比,分别为0.32、1.92、1.70和2.47。对形状的定性评估表明,薄且曲线状的SE趋于退化,表明与定量形状特征显著相关(r=0.31)。在Cox回归中,rCBV和病变形状显着(风险比分别为1.09和0.54)。对于亚厘米病变,rCBV在预测结果方面显示出附加值(曲线下面积,0.873vs.0.836;对数秩检验)。
结论:较小,rCBV和IDH突变较低的细长病变在区分放疗后SE的高级别神经胶质瘤的放疗变化和进展时与消退相关.
方法:3技术效果:阶段2。
BACKGROUND: The specific patterns of subependymal enhancement (SE) that frequently occur as radiation-induced changes in high-grade gliomas following radiotherapy are often overlooked. Perfusion MRI may offer a diagnostic clue.
OBJECTIVE: To distinguish between radiation-induced SE and progression in adult high-grade diffuse gliomas after standard treatment.
METHODS: Retrospective.
METHODS: Ninety-four consecutive high-grade diffuse glioma patients (mean age, 55 ± 14 years; 54 [57.4%] males) with new SE identified in follow-up MRI after completion of surgery plus chemoradiation: progression (N = 74) vs. regression (N = 20).
UNASSIGNED: 3 T, gradient-echo dynamic susceptibility contrast-enhanced MRI, 3D gradient-echo contrast-enhanced T1-weighted imaging.
RESULTS: To differentiate between radiation changes and progression in SE evaluation, multivariable logistic regression was performed using significant variables among SE appearance interval, IDH mutation, morphological features, and rCBV. Cox regression was performed to predict the tumor progression. For the added value of the rCBV, a log-rank test was conducted between the multivariable logistic regression models with and without the rCBV.
METHODS: Logistic regression, Cox regression, receiver operating characteristic analysis, log-rank test.
RESULTS: 38.3% (36/94) patients had first specific SE (9.2 ± 9.5 months after surgery), which disappeared in 21.3% (20/94) after 5.8 ± 5.8 months after initial appearance on post-radiation MRI. IDH mutation, elongated, small lesions with lower rCBV tended to regress: IDH mutation, elongation, diameter, and rCBV_p95; odds ratio, 0.32, 1.92, 1.70, and 2.47, respectively. Qualitative evaluation of shape revealed that thin and curvilinear-shaped SE tended to regress, indicating a significant correlation with quantitative shape features (r = 0.31). In Cox regression, rCBV and lesion shape were significant (hazard ratio = 1.09 and 0.54, respectively). For sub-centimeter lesions, the rCBV showed added value in predicting outcomes (area under the curve, 0.873 vs. 0.836; log-rank test).
CONCLUSIONS: Smaller, elongated lesions with lower rCBV and IDH mutation are associated with regression when differentiating radiation changes from progression in high-grade glioma with post-radiotherapy SE.
METHODS: 3 TECHNICAL EFFICACY: Stage 2.