Department of Radiology, Lanzhou University Second Hospital, Lanzhou, 730030, China.
Second Clinical School, Lanzhou University, Lanzhou, 730030, China.
J Cancer Res Clin Oncol. 2023 Dec;149(19):17427-17436. doi: 10.1007/s00432-023-05463-x. Epub 2023 Oct 25.
To investigate the predictive value of a model combining conventional MRI features and apparent diffusion coefficient (ADC) histogram parameters for meningioma recurrence.
Seventy-two meningioma patients confirmed by surgical and pathological findings in our hospital (January 2017-June 2020) were retrospectively and divided into the recurrence and non-recurrence group. MaZda software was used to delineate the region of interest at the largest tumor level and generate histogram parameters. Univariate and multivariate logistic regression analysis were used to construct the nomogram for predicting recurrence. The predictive efficacy and diagnostic of this model were assessed by calibration and decision curve analysis, and receiver operating characteristic curve, respectively.
Maximum diameter, necrosis, enhancement uniformity, age, Simpson, tumor shape, and ADC first percentile (ADCp1) were significantly different between the two groups (p < 0.05), with the latter four being independent risk factors for recurrence. The model constructed combining the four factors had the best predictive efficacy, and the area under the curve, accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 0.965(0.892-0.994), 90.3%, 92.6%, 88.9%, 83.3%, and 95.2%, respectively. The calibration curve showed good agreement between the model-predicted and actual probabilities of recurrence. The decision curve analysis indicated good clinical availability of the model.
This model based on conventional MRI features and ADC histogram parameters can directly and reliably predict meningioma recurrence, providing a guiding basis for selecting treatment options and individualized treatment.
探讨联合常规 MRI 特征和表观扩散系数（ADC）直方图参数的模型对脑膜瘤复发的预测价值。
回顾性分析我院 2017 年 1 月至 2020 年 6 月间经手术和病理证实的 72 例脑膜瘤患者的临床资料，根据术后是否复发分为复发组和未复发组。使用 MaZda 软件在最大肿瘤层面勾画感兴趣区并生成直方图参数。采用单因素和多因素逻辑回归分析构建预测脑膜瘤复发的列线图。通过校准和决策曲线分析评估该模型的预测效能和诊断价值，采用受试者工作特征曲线评估模型的诊断效能。
复发组和未复发组患者的最大直径、坏死、强化均匀度、年龄、Simpson 分级、肿瘤形状、ADC 最小值（ADCp1）差异均有统计学意义（p<0.05），其中后 4 项是复发的独立危险因素。结合这 4 个因素构建的模型预测效能最佳，曲线下面积、准确率、敏感度、特异度、阳性预测值和阴性预测值分别为 0.965（0.892-0.994）、90.3%、92.6%、88.9%、83.3%和 95.2%。校准曲线显示模型预测的复发概率与实际复发概率具有良好的一致性。决策曲线分析表明该模型具有良好的临床应用价值。
该模型基于常规 MRI 特征和 ADC 直方图参数，能直接、可靠地预测脑膜瘤的复发，为选择治疗方案和个体化治疗提供指导依据。
