文献速递神经肿瘤

酰胺质子转移加权MRI在IDH野生型胶质母细胞瘤假性进展单时间点诊断中的常规临床应用

Amide proton transfer-weighted (APTw) CEST MRI in clinical routine for single time point diagnosis of pseudoprogression in IDH-wildtype glioblastoma.
PubMed 文献 · Neuro Oncol · 2026-01-01

本研究将APTw MRI整合入脑肿瘤常规影像流程,筛查626例患者共870次扫描,最终分析143例胶质母细胞瘤患者的256次MRI扫描,其中65次显示PD(42次)或TRE(23次)。PD组APTw信号中位数高于TRE组(2.23% vs 1.76%,P=.001),ROC曲线下面积0.82;放疗后6个月内早期PD或TRE的AUC升至0.93;抗血管生成治疗降低APTw信号(P<.01);APTw联合DWI和PWI将AUC提升至0.90。

要点

  • APTw MRI可无创区分胶质母细胞瘤PD与TRE,总体AUC为0.82。
  • 放疗后6个月内早期鉴别诊断效能更高,AUC达0.93。
  • 抗血管生成治疗降低APTw信号,联合DWI和PWI可提高诊断准确度。

核心结论

  • APTw MRI是一种适用于临床常规的无创成像工具,有助于区分胶质母细胞瘤的早期进展与假性进展。
  • 抗血管生成治疗应用后及随访较晚时间点,其诊断准确度下降。
  • APTw MRI联合PWI和DWI的多模态方案诊断准确度最高。

研究设计与样本

  • 将APTw MRI整合入脑肿瘤常规影像工作流程后,筛查626例患者共870次扫描。
  • 最终分析143例胶质母细胞瘤患者的256次MRI扫描,其中65次显示PD(n=42)或TRE(n=23)。
  • 采用基于深度学习的3D肿瘤分割在钆增强T1w和FLAIR感兴趣区自动量化基于体素的APTw信号。
  • PD与TRE比较采用非配对t检验,诊断准确度通过ROC和逻辑回归分析评估。

主要发现

  • PD组APTw信号中位数高于TRE组(2.23% vs 1.76%;P=.001)。
  • ROC分析显示AUC为0.82。
  • 放疗后6个月内早期PD或TRE患者中AUC升至0.93。
  • 抗血管生成治疗降低APTw信号(P<.01)。
  • APTw MRI联合DWI和PWI提高诊断准确度(AUC=0.90)。

临床怎么用

  • 可将APTw MRI作为常规脑肿瘤影像流程的一部分,用于单时间点鉴别胶质母细胞瘤PD与TRE。
  • 放疗后6个月内早期鉴别诊断效能最高,适合在此时间窗应用。
  • 联合DWI和PWI的多模态评估可进一步提高诊断准确度。
  • 接受抗血管生成治疗或随访较晚的患者,需注意APTw诊断准确度下降。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Differentiating progressive disease (PD) from treatment-related effects (TRE) in glioblastoma remains challenging, particularly at single time point evaluations. TRE can occur at any disease stage, and its underlying biology is poorly understood. This study evaluates the clinical feasibility and diagnostic performance of amide proton transfer-weighted (APTw) MRI in this challenge. METHODS: Following the integration of APTw MRI into the routine clinical workflow for brain tumor imaging, we screened a total of 870 scans from 626 patients. APTw signal (voxel-based measurement) was automatically quantified in gadolinium-enhanced T1w and FLAIR regions of interest using a deep learning-based approach for 3D tumor segmentations. PD and TRE were compared using unpaired t-tests, and diagnostic accuracy was assessed via ROC- and logistic regression analysis. RESULTS: Among 256 MRI scans of 143 patients with glioblastoma, 65 scans showed PD (n = 42) or TRE (n = 23). The median APTw signal was higher in PD (2.23%) vs TRE (1.76%; P = .001). ROC analysis showed an area under the curve (AUC) of 0.82. In patients with early PD or TRE (<6 months post-radiotherapy), the AUC increased to 0.93. Anti-angiogenic therapy decreased APTw signal (P < .01). Combining APTw MRI with DWI and PWI improved diagnostic accuracy (AUC = 0.90). CONCLUSIONS: APTw MRI is a non-invasive imaging tool that is feasible for clinical routine and aids in differentiation of early progression from pseudoprogression in glioblastoma. Its diagnostic accuracy decreases with application of anti-angiogenic treatment and at later follow-up time points. Highest diagnostic accuracy was found in a multimodal approach combining APTw MRI, PWI and DWI.

来源:Neuro Oncol,查看原文。版权归原作者所有。

引用

  1. Zeyen T, Krause I, Decker A, et al. Amide proton transfer-weighted (APTw) CEST MRI in clinical routine for single time point diagnosis of pseudoprogression in IDH-wildtype glioblastoma. Neuro Oncol. 2026. PMID: 41233976. doi:10.1093/neuonc/noaf261

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。