| Objective(s):To retrospectively analyze the general clinical data and preoperative magnetic resonance imaging(MRI)features of patients with atypical meningioma(AM)confirmed by pathology,and to analyze the correlation between imaging features and brain invasion(BI).To construct a predictive model for brain invasion of atypical meningiomas and analyze its predictive efficiency,so as to deepen the understanding of invasive meningiomas and explore the correlation between preoperative MRI features and brain invasion of atypical meningiomas.Methods:The clinical data and preoperative cranial magnetic resonance imaging(MRI)of 62 patients with pathologically confirmed atypical meningioma treated by craniotomy at the First Affiliated Hospital of Kunming Medical University from January 2018 to October 2022 were retrospectively included.The tumour morphology,tumour enhancement pattern,smoothness of the tumour-brain junction,tumour base,ADC signal,longest tumour diameter,longest oedema diameter and tumour volume were statistically analysed to investigate the factors associated with the occurrence of invasion in atypical meningioma and to establish a prediction model for the occurrence of brain invasion,with a view to providing guidance in clinical practice.Results:Univariate analysis showed that tumour morphology(P<0.001),tumour enhancement pattern(P=0.002),smoothness of the tumour-brain junction(P=0.043),tumour base(P=0.011),tumour volume(P=0.008),longest tumour diameter(P=0.001)and longest oedema diameter(P=0.002)were significantly correlation(P<0.05).The gender,age,tumor location and ADC signal of the patients were not statistically significant.Spearman’s correlation analysis showed that the longest tumour diameter was highly correlated with tumour volume(r=0.800,P=0.000<0.001),while the other two variables were weakly or not correlated.Multi-factor binary logistic regression analysis revealed that: 1.When the two-dimensional model was included in the binary logistic regression analysis,tumour junction smoothness(P=0.015,OR: 12.647,95%CI: 1.648 to 97.042),tumour base(P=0.009,OR: 13.548,95% CI: 1.896 to 96.824),edema longest diameter(P=0.015,OR: 1.536,95% CI: 1.087 to 2.17),and tumour longest diameter(P=0.033,OR: 1.863,95% CI: 1.051 to 3.303)were significantly correlated with the occurrence of brain invasion(P<0.05);2.When the three-dimensional model was included in the binary logistic regression analysis,the tumour junction smoothness(P=0.021,OR: 6.724,95% CI: 1.332 to 33.930),tumour base(P=0.004,OR: 17.167,95% CI: 2.478 to 118.935),and tumour volume(P=0.006,OR: 1.041,95% CI: 1.011 to 1.072)were significantly associated with brain invasion was significantly correlated with the occurrence of invasion(P<0.05).The results of the subject working characteristic curves showed good predictive efficacy for the longest diameter of tumour,longest diameter of oedema,and tumour volume models among the single indicators(all AUC values > 0.7),with AUCs of 0.769(sensitivity =76.7%,specificity = 75.0%),0.742(sensitivity 83.3%,specificity = 62.5%),0.714(sensitivity = 56.7%,specificity=78.1%),with the longest tumour diameter having the highest AUC value(AUC=0.769,sensitivity=76.7%,specificity=75.0%)and the best predictive efficacy.However,the combined index-two-dimensional model had the highest AUC of 0.904 and also the highest sensitivity and specificity of 90.0% and84.4%,respectively,which were greater than the results of either single index.Conclusion(s): 1.Based on preoperative MRI features,it was found that the smoothness of tumor-brain interface,tumor base,tumor volume,longest diameter of tumor and longest diameter of edema were closely related to the occurrence of brain invasion of atypical meningioma;2.Among the single indicators,the longest diameter of tumor had the highest predictive efficiency.In the combined analysis of multiple indicators,the "comprehensive index-two-dimensional model" has the best prediction efficiency and can better predict brain invasion.3.Preoperative imaging can reflect the biological behavior of brain invasion,which is of great significance for the preoperative detection of atypical meningiomas with brain invasion,adjustment of surgical plan and prognosis evaluation... |
PDF Full Text Request