Objective: Meningiomas presented preferred intracranial distribution, which may reflect potential biological natures. This study aimed to analyze the preferred locations of meningioma according to different biological characteristics. Method: A total of 1,107 patients pathologically diagnosed with meningiomas between January 2012 and December 2016 were retrospectively analyzed. Preoperative MRI were normalized, and lesions were semiautomatically segmented. The stereospecific frequency and p value heatmaps were constructed to compare two biological phenotypes using two-tailed Fisher's exact test. Age, sex, WHO grades, extent of resection (EOR), recurrence, and immunohistochemical markers including p53, Ki67, epithelial membrane antigen (EMA), progesterone receptor (PR), and CD34 were statistically analyzed. Recurrence-free survival (RFS) were analyzed by Kaplan-Meier method. Result: Of 1,107 cases, convexity (20.8%), parasagittal (16.1%), and falx (11.4%) were the most predominant loci of meningiomas. The p-value heatmap suggested lesion predominance in the left frontal and occipital convexity among older patients while in the left sphenoid wing, and right falx, parasellar/cavernous sinus, and middle fossa among younger patients. Lesions located at anterior fossa and frontal structures were more frequently seen in the male while left parietal falx and tentorial regions, and right cerebellopontine angle in the female. Grades II and III lesions presented predominance in the frontal structures compared with grade I ones. Meningiomas at the left parasagittal sinus and falx, tentorium, intraventricular regions, and skull-base structures were significantly to receive subtotal resection. Lesions with p53 positivity were statistically located at the left frontal regions and parasellar/cavernous sinus, higher Ki67 index at the left frontal and bilateral parietal convexity and right parasellar/cavernous sinus, EMA negativity at the right olfactory groove and left middle fossa, and CD34 positivity at the sellar regions and right sphenoid wing. Tumor recurrence rates for grades I, II, and III were 2.8, 7.9, and 53.8%, respectively. Inferior RFS, higher Ki67 index, grades II and III, and a larger preoperative volume were observed in older patients. Recurrent meningiomas were more frequently found at the occipital convexity, tentorium, sellar regions, parasagittal sinus, and left sphenoid wing. Conclusion: The preferred locations of meningioma could be observed according to different biological characteristics, which might be helpful for clinical decisions.
Keywords: lesion location preference; magnetic resonance imaging; meningiomas; recurrence; voxel-wise analysis.
Copyright © 2020 Sun, Dou, Wu, Jiang, Iranmanesh, Yu, Li, Zhou, Zhong, Peng, Zhuang, Yu, Wu, Yan, Xie and Chen.