文献速递神经创伤神经重症

创伤性轴索损伤:常规MRI与DTI的预后信息是互补还是补充?

Traumatic axonal injury: is the prognostic information produced by conventional MRI and DTI complementary or supplementary?
PubMed 文献 · J Neurosurg · 2021-07-02

本研究纳入185例中重度TBI患者,在亚急性期行常规MRI与DTI检查,评估两者病灶共定位及预后价值,并在92例队列中内部验证。结果显示宏观非出血性TAI病灶与FA降低共定位有限,DTI与常规MRI提供互补预后信息,联合模型性能优于IMPACT模型。

要点

  • 常规MRI非出血性TAI病灶与FA降低共定位仅见于内/外囊、放射冠、下额枕束、胼胝体和脑干。
  • DTI与常规MRI提供互补预后信息,联合可提升传统预后模型性能。
  • 年龄、运动评分、缺氧、胼胝体膝部FA、胼胝体TAI病灶及丘脑/基底节病灶为独立预后因素。

核心结论

  • 宏观非出血性TAI病灶与FA降低共定位非常有限,微出血无共定位。
  • DTI与常规MRI提供互补预后信息,联合可改善传统预后模型性能。

研究设计与样本

  • 研究纳入185例中重度TBI患者,亚急性期行常规MRI与DTI检查。
  • 在至少四种常规MRI序列上确定半球皮质下白质、胼胝体、基底节、丘脑和脑干病灶数目与体积。
  • 采用感兴趣区法测量28个白质束的FA。
  • 多变量逻辑回归评估各神经影像模态独立预后价值,并在92例当代队列中内部验证。

主要发现

  • 不同严重程度和结局患者间病灶负荷存在差异。
  • 宏观非出血性TAI病灶与较低FA共定位仅见于内囊、外囊、放射冠、下额枕束、胼胝体和脑干。
  • 部分半球白质束及胼胝体压部FA较低与脑干高信号数目和体积较高相关。
  • 回归分析显示年龄、运动评分、缺氧、胼胝体膝部FA、胼胝体TAI病灶特征及丘脑/基底节病灶为独立预后因素。
  • 验证队列中提出模型性能高于IMPACT模型。

临床怎么用

  • 亚急性期中重度TBI患者可联合常规MRI与DTI评估TAI预后。
  • 关注胼胝体膝部FA、胼胝体TAI病灶及丘脑/基底节病灶等独立预后因素。

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

OBJECTIVE: A traumatic axonal injury (TAI) diagnosis has traditionally been based on conventional MRI, especially on those sequences with a higher sensitivity to edema and blood degradation products. A more recent technique, diffusion tensor imaging (DTI), can infer the microstructure of white matter (WM) due to the restricted diffusion of water in organized tissues. However, there is little information regarding the correlation of the findings obtained by both methods and their use for outcome prognosis. The main objectives of this study were threefold: 1) study the correlation between DTI metrics and conventional MRI findings; 2) evaluate whether the prognostic information provided by the two techniques is supplementary or complementary; and 3) determine the incremental value of the addition of these variables compared to a traditional prognostic model. METHODS: The authors studied 185 patients with moderate to severe traumatic brain injury (TBI) who underwent MRI with DTI study during the subacute stage. The number and volume of lesions in hemispheric subcortical WM, corpus callosum (CC), basal ganglia, thalamus, and brainstem in at least four conventional MRI sequences (T1-weighted, T2-weighted, FLAIR, T2* gradient recalled echo, susceptibility-weighted imaging, and diffusion-weighted imaging) were determined. Fractional anisotropy (FA) was measured in 28 WM bundles using the region of interest method. Nonparametric tests were used to evaluate the colocalization of macroscopic lesions and FA. A multivariate logistic regression analysis was performed to assess the independent prognostic value of each neuroimaging modality after adjustment for relevant clinical covariates, and the internal validation of the model was evaluated in a contemporary cohort of 92 patients. RESULTS: Differences in the lesion load between patients according to their severity and outcome were found. Colocalization of macroscopic nonhemorrhagic TAI lesions (not microbleeds) and lower FA was limited to the internal and external capsule, corona radiata, inferior frontooccipital fasciculus, CC, and brainstem. However, a significant association between the FA value and the identification of macroscopic lesions in distant brain regions was also detected. Specifically, lower values of FA of some hemispheric WM bundles and the splenium of the CC were related to a higher number and volume of hyperintensities in the brainstem. The regression analysis revealed that age, motor score, hypoxia, FA of the genu of the CC, characterization of TAI lesions in the CC, and the presence of thalamic/basal ganglia lesions were independent prognostic factors. The performance of the proposed model was higher than that of the IMPACT (International Mission on Prognosis and Analysis of Clinical Trials in TBI) model in the validation cohort. CONCLUSIONS: Very limited colocalization of hyperintensities (none for microbleeds) with FA values was discovered. DTI and conventional MRI provide complementary prognostic information, and their combination can improve the performance of traditional prognostic models.

来源:J Neurosurg,查看原文。版权归原作者所有。

引用

  1. Castaño-Leon AM, Cicuendez M, Navarro-Main B, et al. Traumatic axonal injury: is the prognostic information produced by conventional MRI and DTI complementary or supplementary? J Neurosurg. 2021. PMID: 34214979. doi:10.3171/2020.11.JNS203124

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。