Novel method to evaluate the risk of tumor adhesions and post-operative hemorrhage of meningiomas using 320 row CT-DSA: A clinical research study.

2020 
OBJECTIVE: Meningioma is an extra-axial tumor that forms adhesions toward the brain surface in the course of its growth. Predicting adhesions between the tumor and the brain surface leads to better predictions of surgical results. There are few studies on brain-tumor adhesions or postoperative hemorrhage. This study aimed to assess tumor vascularity of the dura and cerebral surface, and predict surgical outcomes using four-dimensional computed tomography angiography (4D CTA). PATIENTS AND METHODS: Using a dynamic contrast CT, we conducted a retrospective study of 27 patients with convexity (n = 15), falx (n = 6), and parasagittal (n = 6) meningiomas treated in our hospital from January 2016 to September 2018. We set the region of interest on the dural layer and cerebral surface side of meningiomas and calculated the mean CT value in each region. Distribution of blood flow in the tumor was classified into two groups: A, which has a higher CT value of the dural side than that of the brain surface side at every timing, and B, which meets the criteria other than those in group A. Demographic data, preoperative characteristic images, and postoperative complications were compared between the groups. RESULTS: Twelve and 15 patients were classified into groups A and B, respectively. The extent of adhesions against the cerebral cortex in group A was significantly less severe compared with that in group B (p = 0.038). The rate of postoperative hemorrhage occurrence in group B (53%) was significantly higher than that in group A (8%) (p = 0.04). There were no significant differences in the other preoperative characteristic images or perioperative parameters between groups A and B. CONCLUSION: A 320-row dynamic contrast CT scanner can detect meningiomas with a high probability of severe adhesion toward the brain surface and postoperative intraparenchymal hematoma.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    28
    References
    0
    Citations
    NaN
    KQI
    []