Scoring radiologic characteristics to predict proliferative potential in meningiomas

Deep learning and radiomics-based approach to meningioma grading: exploring the potential value of peritumoral edema regions

Objective.To address the challenge of meningioma grading, this study aims to investigate the potential value of peritumoral edema (PTE) regions and proposes a unique approach that integrates radiomics and deep learning techniques.Approach.The primary focus is on developing a transfer learning-based meningioma feature extraction model (MFEM) that leverages both vision transformer (ViT) and convolutional neural network (CNN) architectures. Additionally, the study explores the significance of the PTE region in enhancing the grading process.Main results.The proposed method demonstrates excellent grading accuracy and robustness on a dataset of 98 meningioma patients. It achieves an accuracy of 92.86%, precision of 93.44%, sensitivity of 95%, and specificity of 89.47%.Significance.This study provides valuable insights into preoperative meningioma grading by introducing an innovative method that combines radiomics and deep learning techniques. The approach not only enhances accuracy but also reduces observer subjectivity, thereby contributing to improved clinical decision-making processes.

Impact of Shape Irregularity in Medial Sphenoid Wing Meningiomas on Postoperative Cranial Nerve Functioning, Proliferation, and Progression-Free Survival

Medial sphenoid wing meningiomas (MSWM) are surgically challenging skull base tumors. Irregular tumor shapes are thought to be linked to histopathology. The present study aims to investigate the impact of tumor shape on postoperative functioning, progression-free survival, and neuropathology. This monocentric study included 74 patients who underwent surgery for primary sporadic MSWM (WHO grades 1 and 2) between 2010 and 2021. Furthermore, a systematic review of the literature regarding meningioma shape and the MIB-1 index was performed. Irregular MSWM shapes were identified in 31 patients (41.9%). Multivariable analysis revealed that irregular shape was associated with postoperative cranial nerve deficits (OR: 5.75, 95% CI: 1.15-28.63, p = 0.033). In multivariable Cox regression analysis, irregular MSWM shape was independently associated with tumor progression (HR:8.0, 95% CI: 1.04-62.10, p = 0.046). Multivariable regression analysis showed that irregular shape is independently associated with an increased MIB-1 index (OR: 7.59, 95% CI: 2.04-28.25, p = 0.003). A systematic review of the literature and pooled data analysis, including the present study, showed that irregularly shaped meningiomas had an increase of 1.98 (95% CI: 1.38-2.59, p < 0.001) in the MIB-1 index. Irregular MSWM shape is independently associated with an increased risk of postoperative cranial nerve deficits and a shortened time to tumor progression. Irregular MSWM shapes might be caused by highly proliferative tumors.

Imaging score for differentiation of meningioma grade

PURPOSE

We sought to establish a comprehensive imaging score indicating the likelihood of higher WHO grade meningiomas pre-operatively.

METHODS

All surgical intracranial meningioma patients at our institution between 2014 and 2018 underwent retrospective chart review. Preoperative MRI sequences were reviewed, and imaging features were included in the score based on statistical and clinical significance. Point values for each significant feature were assigned based on the beta coefficients obtained from multivariate analysis. The imaging score was calculated by adding up the points, for a total score of 0 to 5. The predictive ability of the score to identify higher-grade meningiomas was evaluated.

RESULTS

Ninety patients, 50% of whom had a postoperative diagnosis of WHO grade II meningioma, were included. The mean age for the population was 59.9 years and 70% were female. Tumor volume ≥ 36.0 cc was assigned 2 points, presence of irregular tumor borders was assigned 2 points, and presence of peritumoral edema was assigned 1 point. The probability of having a WHO grade II meningioma was 0% with a score of 0, 25.0% with a score of 1, 38.5% with a score of 2, 65.4% with a score of 3, and 83.3% with a score of 4 or greater. A threshold of ≥ 3 points achieved a recall of 0.80, precision of 0.73, F1-score of 0.77, accuracy of 0.76, and AUC of 0.82.

CONCLUSION

The proposed imaging scoring system had good predictive capability for WHO grade II meningiomas with good discrimination and calibration. External validation is needed.

Principal component analysis of texture features for grading of meningioma: not effective from the peritumoral area but effective from the tumor area

PURPOSE

To investigate whether texture features from tumor and peritumoral areas based on sequence combinations can differentiate between low- and non-low-grade meningiomas.

METHODS

Consecutive patients diagnosed with meningioma by surgery (77 low-grade and 28 non-low-grade meningiomas) underwent preoperative magnetic resonance imaging including T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), and contrast-enhanced T1WI (CE-T1WI). Manual segmentation of the tumor area was performed to extract texture features. Segmentation of the peritumoral area was performed for peritumoral high-signal intensity (PHSI) on T2WI. Principal component analysis was performed to fuse the texture features to principal components (PCs), and PCs of each sequence of the tumor and peritumoral areas were compared between low- and non-low-grade meningiomas. Only PCs with statistical significance were used for the model construction using a support vector machine algorithm. k-fold cross-validation with receiver operating characteristic curve analysis was used to evaluate diagnostic performance.

RESULTS

Two, one, and three PCs of T1WI, apparent diffusion coefficient (ADC), and CE-T1WI, respectively, for the tumor area, were significantly different between low- and non-low-grade meningiomas, while PCs of T2WI for the tumor area and PCs for the peritumoral area were not. No significant differences were observed in PHSI. Among models of sequence combination, the model with PCs of ADC and CE-T1WI for the tumor area showed the highest area under the curve (0.84).

CONCLUSION

The model with PCs of ADC and CE-T1WI for the tumor area showed the highest diagnostic performance for differentiating between low- and non-low-grade meningiomas. Neither PHSI nor PCs in the peritumoral area showed added value.

A radiomics nomogram for predicting the meningioma grade based on enhanced T1WI images

OBJECTIVES

The objective of this study was to develop a radiomics nomogram for predicting the meningioma grade based on enhanced T1 weighted imaging (T1WI) images.

METHODS

188 patients with meningioma were analyzed retrospectively. There were 94 high-grade meningioma to form high-grade group and 94 low-grade meningioma were selected randomly to form low-grade group. Clinical data and MRI features were analyzed and compared. The clinical model was built by using the significant variables. The least absolute shrinkage and selection operator regression was used to select the most valuable radiomics feature. The radiomics signature was built and the Rad-score was calculated. The radiomics nomogram was developed by the significant variables of the clinical factors and Rad-score. The calibration curve and the Hosmer-Lemeshow test were used to evaluate the radiomics nomogram. Different models were compared by Delong test and decision curve analysis curve.

RESULTS

The sex, size and surrounding invasion were used to build clinical model. The area under the receiver operator characteristic curve (AUC) of clinical model was 0.870 (95% CI: 0.782-0.959). Nine features were used to construct the radiomics signature. The AUC of the radiomics signature was 0.885 (95% CI: 0.802-0.968). The AUC of radiomics nomogram was 0.952 (95% CI: 0.904-1). The AUC of radiomics nomogram was higher than that of clinical model and radiomics signature with a significant difference (p＜0.05). The decision curve analysis curve showed that the radiomics nomogram had a larger net benefit than the clinical model and radiomics signature.

CONCLUSION

The radiomics nomogram based on enhanced T1 weighted imaging images for predicting the meningioma grade showed high predictive value and might contribute to the diagnosis and treatment of meningioma.

ADVANCES IN KNOWLEDGE

1. We first constructed a radiomic nomogram to predict the meningioma grade.2. We compared the results of the clinical model, radiomics signature and radiomics nomogram.

Bone Invasive Meningioma: Recent Advances and Therapeutic Perspectives

Meningioma is the most common primary neoplasm of the central nervous system (CNS). Generally, these tumors are benign and have a good prognosis. However, treatment can be challenging in cases with aggressive variants and poor prognoses. Among various prognostic factors that have been clinically investigated, bone invasion remains controversial owing to a limited number of assessments. Recent study reported that bone invasion was not associated with WHO grades, progression, or recurrence. Whereas, patients with longer-recurrence tended to have a higher incidence of bone invasion. Furthermore, bone invasion may be a primary preoperative predictor of the extent of surgical resection. Increasing such evidence highlights the potential of translational studies to understand bone invasion as a prognostic factor of meningiomas. Therefore, this mini-review summarizes recent advances in pathophysiology and diagnostic modalities and discusses future research directions and therapeutic strategies for meningiomas with bone invasion.

The meningioma surface factor: a novel approach to quantify shape irregularity on preoperative imaging and its correlation with WHO grade

OBJECTIVE

Atypical and anaplastic meningiomas account for 20% of all meningiomas. An irregular tumor shape on preoperative MRI has been associated with WHO grade II-III histology. However, this subjective allocation does not allow quantification or comparison. An objective parameter of irregularity could substantially influence resection strategy toward a more aggressive approach. Therefore, the aim of this study was to objectively quantify the level of irregularity on preoperative MRI and predict histology based on WHO grade using this novel approach.

METHODS

A retrospective study on meningiomas resected between January 2010 and December 2018 was conducted at two neurosurgical centers. This novel approach relies on the theory that a regularly shaped tumor has a smaller surface area than an irregularly shaped tumor with the same volume. A factor was generated using the surface area of a corresponding sphere as a reference, because for a given volume a sphere represents the shape with the smallest surface area possible. Consequently, the surface factor (SF) was calculated by dividing the surface area of a sphere with the same volume as the tumor with the surface area of the tumor. The resulting value of the SF ranges from > 0 to 1. Finally, the SF of each meningioma was then correlated with the corresponding histopathological grading.

RESULTS

A total of 126 patients were included in this study; 60.3% had a WHO grade I, 34.9% a WHO grade II, and 4.8% a WHO grade III meningioma. Calculation of the SF demonstrated a significant difference in SFs between WHO grade I (SF 0.851) and WHO grade II-III meningiomas (SF 0.788) (p < 0.001). Multivariate analysis identified SF as an independent prognostic factor for WHO grade (OR 0.000009, 95% CI 0.000-0.159; p = 0.020).

CONCLUSIONS

The SF is a proposed mathematical model for a quantitative and objective measurement of meningioma shape, instead of the present subjective assessment. This study revealed significant differences between the SFs of WHO grade I and WHO grade II-III meningiomas and demonstrated that SF is an independent prognostic factor for WHO grade.

Advanced MRI shape analysis as a predictor of histologically aggressive supratentorial meningioma

BACKGROUND AND PURPOSE

A subset of aggressive meningioma is associated with higher morbidity and requires a different therapeutic management. This subset consists of World Health Organization (WHO) grade II and III meningioma, characterized particularly with microscopic brain invasion. Numerous studies tried to screen aggressive meningioma on pre-operative MRI. The objective of the study was to determine if an advanced shape analysis of supratentorial meningioma outlines could reliably predict WHO II-III grade and histological brain invasion.

MATERIALS AND METHODS

We performed a retrospective analysis for all consecutive patients who underwent surgery for supratentorial histologically-proven meningioma from 2010 to 2018. Pre-operative MRI T1WI contrast enhanced axial, coronal and sagittal slices were collected from 101 patients. Advanced shape analysis including fractal analysis and topological skeleton analysis was performed. Shape analysis parameters were correlated with histopathological WHO grading and brain invasion on surgical pieces.

RESULTS

Shape analysis features such as a low circularity, a low solidity, a high fractal dimension and a high number of skeleton's branches were significantly correlated with both WHO II-III meningioma and histological brain invasion. Cross-validated regression models including these features were predictive of WHO II-III meningioma and brain invasion with respective AUC of 0.71 and 0.72.

CONCLUSIONS

MRI shape analysis provides informative imaging biomarkers to predict high WHO grade and histological brain invasion of supratentorial meningioma. Further prospective studies including the evaluation of a fully-automatized and totally reproducible process are required to confirm the results.

Machine learning-based radiomics analysis in predicting the meningioma grade using multiparametric MRI

PURPOSE

To investigate the prediction performance of radiomic models based on multiparametric MRI in predicting the meningioma grade.

METHOD

In all, 229 low-grade [Grade I] and 87 high-grade [Grade II/III] patients with pathologically diagnosed meningiomas were enrolled. Radiomic features from conventional MRI (cMRI), ADC maps and SWI were extracted based on the volume of entire tumor. Classification performance of different radiomic models (cMRI, ADC, SWI, cMRI + ADC, cMRI + SWI, ADC + SWI, and cMRI + ADC + SWI models) was evaluated by a nested LOOCV approach, combining the LASSO feature selection and RF classifier that was trained (1) without subsampling, and (2) with the synthetic minority over-sampling technique (SMOTE). The prediction performance of radiomic models was assessed using ROC curve and AUC of them was compared using Delong's test.

RESULTS

The cMRI + ADC + SWI model demonstrated the best performance without or with subsampling, which AUCs were 0.84 and 0.81, respectively. Following the cMRI + ADC + SWI model, the AUC range of the other models was 0.75-0.80 without subsampling, and was 0.71-0.79 with subsampling. Although the cMRI + ADC model and cMRI + SWI model showed higher AUCs than the cMRI model without subsampling (0.77 vs 0.80, P = 0.037 and 0.77 vs 0.80, P = 0.009, respectively), there was no significant difference among these models with subsampling (0.78 vs 0.77, P = 0.552 and 0.78 vs 0.79, P = 0.246, respectively).

CONCLUSIONS

Multiparametric radiomic model based on cMRI, ADC map and SWI yielded the best prediction performance in predicting the meningioma grade, which might offer potential guidance in clinical decision-making.

Imaging and diagnostic advances for intracranial meningiomas

The archetypal imaging characteristics of meningiomas are among the most stereotypic of all central nervous system (CNS) tumors. In the era of plain film and ventriculography, imaging was only performed if a mass was suspected, and their results were more suggestive than definitive. Following more than a century of technological development, we can now rely on imaging to non-invasively diagnose meningioma with great confidence and precisely delineate the locations of these tumors relative to their surrounding structures to inform treatment planning. Asymptomatic meningiomas may be identified and their growth monitored over time; moreover, imaging routinely serves as an essential tool to survey tumor burden at various stages during the course of treatment, thereby providing guidance on their effectiveness or the need for further intervention. Modern radiological techniques are expanding the power of imaging from tumor detection and monitoring to include extraction of biologic information from advanced analysis of radiological parameters. These contemporary approaches have led to promising attempts to predict tumor grade and, in turn, contribute prognostic data. In this supplement article, we review important current and future aspects of imaging in the diagnosis and management of meningioma, including conventional and advanced imaging techniques using CT, MRI, and nuclear medicine.

Predicting the risk of postoperative recurrence and high-grade histology in patients with intracranial meningiomas using routine preoperative MRI

Risk factors for prediction of prognosis in meningiomas derivable from routine preoperative magnetic resonance imaging (pMRI) remain elusive. Correlations of tumor and edema volume, disruption of the arachnoid layer, heterogeneity of contrast enhancement, enhancement of the capsule, T2-intensity, tumor shape, and calcifications on pMRI with tumor recurrence and high-grade (WHO grade II/III) histology were analyzed in 565 patients who underwent surgery for WHO grade I (N = 516, 91%) or II/III (high-grade histology, N = 49, 9%) meningioma between 1991 and 2018. Edema volume (OR, 1.00; p = 0.003), heterogeneous contrast enhancement (OR, 3.10; p < 0.001), and an irregular shape (OR, 2.16; p = 0.015) were associated with high-grade histology. Multivariate analyses confirmed edema volume (OR, 1.00; p = 0.037) and heterogeneous contrast enhancement (OR, 2.51; p = 0.014) as risk factors for high-grade histology. Tumor volume (HR, 1.01; p = 0.045), disruption of the arachnoid layer (HR, 2.50; p = 0.003), heterogeneous contrast enhancement (HR, 2.05; p = 0.007), and an irregular tumor shape (HR, 2.57; p = 0.001) were correlated with recurrence. Multivariate analyses confirmed tumor volume (HR, 1.01; p = 0.032) and disruption of the arachnoid layer (HR, 2.44; p = 0.013) as risk factors for recurrence, independent of histology. Subgroup analyses revealed disruption of the arachnoid layer (HR, 9.41; p < 0.001) as a stronger risk factor for recurrence than high-grade histology (HR, 5.15; p = 0.001). Routine pMRI contains relevant information about the risk of recurrence or high-grade histology of meningioma patients. Loss of integrity of the arachnoid layer on MRI had a higher prognostic value than the WHO grading, and underlying histological or molecular alterations remain to be determined.

Differentiating microcystic meningioma from atypical meningioma using diffusion-weighted imaging

PURPOSE

Microcystic meningioma (MCM) appears similar to atypical meningioma(AM) as per conventional diagnostic imaging. However, considering their different recurrence rate and prognosis, accurate differential diagnosis is essential for determine the appropriate treatment strategy. The aim of the study was to differentiate MCM from AM by diffusion-weighted imaging (DWI), in order to provide the basis for accurate preoperative diagnosis.

METHODS

The preoperative clinical data, conventional MRI and DWI data of 15 MCM and 30 AM cases were retrospectively analyzed. The average apparent diffusion coefficient (ADC_mean), minimum ADC (ADC_min) and normalized ADC (nADC) between MCM and AM were compared using two sample t-tests. The value of ADC_mean, ADC_min and nADC in the differential diagnosis of MCM and AM were calculated by the receiver operating curve (ROC) analysis.

RESULTS

The ADC_mean (1.06 ± 0.10 vs 0.80 ± 0.11 × 10^-3 mm²/s; P < 0.001), ADC_min (0.99 ± 0.10 vs 0.74 ± 0.12 × 10^-3 mm²/s; P < 0.001) and nADC (1.45 ± 0.17 vs 1.07 ± 0.17; P < .0001) were significantly higher in MCM compared to AM. ADC_mean of 0.91 × 10^-3 mm²/s showed an optimum area under the ROC curve of 0.967 ± 0.022, and distinguished between MCM and AM with 86.67% sensitivity, 100% specificity and 88.89% accuracy. In addition, its positive and negative predictive values were 96.29% and 77.78% respectively.

CONCLUSIONS

DWI can differentially diagnose MCM and AM, and ADC_mean is a potential quantitative tool that can improve preoperative diagnosis of both tumors.

Prediction of High-Grade Histology and Recurrence in Meningiomas Using Routine Preoperative Magnetic Resonance Imaging: A Systematic Review

OBJECTIVE

Estimating the risk of recurrence after surgery remains crucial during care of patients with meningioma. Numerous studies identified correlations of characteristics on routine preoperative magnetic resonance imaging (MRI) with postoperative recurrence or high-grade histology but showed partially inconclusive results.

METHODS

A systematic review of the literature was performed about findings on preoperative MRI and their correlation with high-grade histology and recurrence. Quality of the included studies was analyzed using standardized Quality Assessment of Diagnostic Accuracy Studies criteria.

RESULTS

Among the 35 studies included, quality of the series according to the Quality Assessment of Diagnostic Accuracy Studies criteria differed widely. Remarkably, MRI variables found to be associated with high-grade histology were commonly not consistently associated with prognosis and vice versa. Correlations of the tumor size, the peritumoral edema size, and contrast-enhancement of the tumor capsule with high-grade histology were controversial. In most studies, non-skull base tumor location, cyst formation, heterogenous contrast-enhancement, an irregular tumor shape, and disruption of the tumor/brain border but not intensity of the lesion on T2-weighted images, calcifications, or bone involvement were associated with grade II/III histology. Although tumor and edema size were usually found to correlate with recurrence, heterogenous contrast enhancement, cyst formation, intensity of the tumor on T2-weighted MRI, and enhancement of the tumor capsule were mostly not related with progression.

CONCLUSIONS

Several mostly consistent but partially inconsistent variables associated with high-grade histology or prognosis were identified. Although standardized studies are needed to provide further clarification, consideration of these findings can help to improve estimation of prognosis and can therefore improve postoperative care in patients with meningioma.

A Logistic Regression Model for Detecting the Presence of Malignant Progression in Atypical Meningiomas

OBJECTIVE

To develop a method to distinguish atypical meningiomas (AMs) with malignant progression (MP) from primary AMs without a clinical history.

METHODS

The clinical, radiologic, and pathologic data of 33 previously Simpson grade I resected (if any) as well as no radiotherapy treated intracranial AMs between January 2008 and December 2015 were reviewed. Immunohistochemical staining for connexin 43 (Cx43) and Ki-67 was performed. Descriptive analysis and univariate and multivariate logistic regression analyses were used to explore independent predictors of MP. A multivariable logistic model was developed to estimate the risk of MP, and its diagnostic value was determined from a receiver operating characteristic curve.

RESULTS

There were 11 AMs (33.3%) with histopathologically confirmed MP from benign meningiomas. The other 22 (66.7%) were initially diagnosed AMs with no histopathologically confirmed MP during a median 60.5 months (range, 42-126 months) of follow-up. Univariate and multivariate logistic analyses showed that irregular tumor shape (P = 0.010) and low Cx43 expression (P = 0.010) were independent predictors of the presence of MP, and the predicted probability was calculated by the following formula: P = 1/[1+exp.{1.218-(3.202×Shape)+(3.814×Cx43)}]. P > 0.5 for an irregularly shaped (score 1) AM with low Cx43 expression (score 0) indicated a high probability of MP. The sensitivity, specificity, positive predictive value, negative predictive value, and overall predictive accuracy were 63.6, 95.6, 87.5, 84.0, and 84.8%, respectively.

CONCLUSIONS

Low Cx43 expression and irregular tumor shape were independent predictors of the presence of MP. The relevant logistic regression model was found to be effective in distinguishing MP-AMs from primary AMs.

Non-operative meningiomas: long-term follow-up of 136 patients

BACKGROUND

Improving access to neuroradiology investigations has led to an increased rate of diagnosis of incidental meningiomas.

METHOD

A cohort of 136 incidental meningioma patients collected by a single neurosurgeon in a single neurosurgical centre is retrospectively analysed between 2002 and 2016. Demographic data, imaging and clinical features are presented. The radiological factors associated with meningiomas progression are also presented.

RESULTS

The mean age at diagnosis was 65 (range, 33-94) years. Univariate analysis showed oedema was most strongly correlated with progression (p = 0.010) followed by hyperintensity in T2-weighted (T2W) MRI (p = 0.029) and in Flair-T2W MRI (p = 0.017). Isointensity in Flair-T2W MRI (0.004) was most strongly correlated with non-progression of the meningioma followed by calcification (p = 0.007), older age (p = 0.087), hypointensity in Flair-T2W MRI (p = 0.014) sequences and in T2W MRI (p = 0.096). In multivariate analysis, the strongest radiological factor predictive of progression was peritumoural oedema (p = 0.016) and that of non-progression was calcification (p = 0.002). At the end of the median follow-up (FU) of 43 (range, 4-150) months, 109 (80%) patients remained clinically stable, 13 (10%) became symptomatic and 14 (10%) showed clinical and radiological progression.

CONCLUSIONS

One hundred and nine (80%) patients remained stable at the end of FU. Peritumoural oedema was predictive of meningiomas progression. Further prospective study is needed to identify the combination of factors which can predict the meningioma progression for an early surgery or early discharge.

Histogram Analysis Parameters Apparent Diffusion Coefficient for Distinguishing High and Low-Grade Meningiomas: A Multicenter Study

Low grade meningiomas have better prognosis than high grade meningiomas. The aim of this study was to measure apparent diffusion coefficient (ADC) histogram analysis parameters in different meningiomas in a large multicenter sample and to analyze the possibility of several parameters for predicting tumor grade and proliferation potential. Overall, 148 meningiomas from 7 institutions were evaluated in this retrospective study. Grade 1 lesions were diagnosed in 101 (68.2%) cases, grade 2 in 41 (27.7%) patients, and grade 3 in 6 (4.1%) patients. All tumors were investigated by MRI (1.5 T scanner) by using diffusion weighted imaging (b values of 0 and 1000 s/mm²). For every lesion, the following parameters were calculated: mean ADC, maximum ADC, minimum ADC, median ADC, mode ADC, ADC percentiles P10, P25, P75, P90, kurtosis, skewness, and entropy. The comparison of ADC values was performed by Mann–Whitney-U test. Correlation between different ADC parameters and KI 67 was calculated by Spearman's rank correlation coefficient. Grade 2/3 meningiomas showed statistically significant lower ADC histogram analysis parameters in comparison to grade 1 tumors, especially ADC median. A threshold value of 0.82 for ADC median to predict tumor grade was estimated (sensitivity = 82.2%, specificity = 63.8%, accuracy = 76.4%, positive and negative predictive values were 83% and 62.5%, respectively).

All ADC parameters except maximum ADC showed weak significant correlations with KI 67, especially ADC P25 (P = −.340, P = .0001).

Comprehensive analysis of lung cancer pathology images to discover tumor shape and boundary features that predict survival outcome

Pathology images capture tumor histomorphological details in high resolution. However, manual detection and characterization of tumor regions in pathology images is labor intensive and subjective. Using a deep convolutional neural network (CNN), we developed an automated tumor region recognition system for lung cancer pathology images. From the identified tumor regions, we extracted 22 well-defined shape and boundary features and found that 15 of them were significantly associated with patient survival outcome in lung adenocarcinoma patients from the National Lung Screening Trial. A tumor region shape-based prognostic model was developed and validated in an independent patient cohort (n = 389). The predicted high-risk group had significantly worse survival than the low-risk group (p value = 0.0029). Predicted risk group serves as an independent prognostic factor (high-risk vs. low-risk, hazard ratio = 2.25, 95% CI 1.34–3.77, p value = 0.0022) after adjusting for age, gender, smoking status, and stage. This study provides new insights into the relationship between tumor shape and patient prognosis.

A novel weighted scoring system for estimating the risk of rapid growth in untreated intracranial meningiomas

OBJECTIVE

Advances in neuroimaging techniques have led to the increased detection of asymptomatic intracranial meningiomas (IMs). Despite several studies on the natural history of IMs, a comprehensive evaluation method for estimating the growth potential of these tumors, based on the relative weight of each risk factor, has not been developed. The aim of this study was to develop a weighted scoring system that estimates the risk of rapid tumor growth to aid treatment decision making.

METHODS

The authors performed a retrospective analysis of 232 patients with presumed IM who had been prospectively followed up in the absence of treatment from 1997 to 2013. Tumor volume was measured by imaging at each follow-up visit, and the growth rate was determined by regression analysis. Predictors of rapid tumor growth (defined as ≥ 2 cm3/year) were identified using a logistic regression model; each factor was awarded a score based on its own coefficient value. The probability (P) of rapid tumor growth was estimated using the following formula:

RESULTS

Fifty-nine tumors (25.4%) showed rapid growth. Tumor size (OR per cm3 1.07, p = 0.000), absence of calcification (OR 3.87, p = 0.004), peritumoral edema (OR 2.74, p = 0.025), and hyperintense or isointense signal on T2-weighted MRI (OR 3.76, p = 0.049) were predictors of tumor growth rate. In the Asan Intracranial Meningioma Scoring System (AIMSS), tumor size was categorized into 3 groups of < 2.5 cm, ≥ 2.5 to < 4.0 cm, and ≥ 4.0 cm in diameter and awarded a score of 0, 3, and 6, respectively; the parameters of calcification and peritumoral edema were categorized into 2 groups based on their presence or absence and given a score of 0 or 2 and 1 or 0, respectively; and the signal on T2-weighted MRI was categorized into 2 groups of hypointense and hyperintense/isointense and given a score of 0 or 2, respectively. The risk of rapid tumor growth was estimated to be < 10% when the total score was 0–2, 10%–50% when the total score was 3–6, and ≥ 50% when the total score was 7–11 (Hosmer-Lemeshow goodness-of-fit test, p = 0.9958). The area under the receiver operating characteristic curve was 0.86.

CONCLUSIONS

The authors suggest a weighted scoring system (AIMSS) that predicts the specific probability of rapid tumor growth for patients with untreated IM. This scoring system will aid treatment decision making in clinical settings by screening out patients at high risk for rapid tumor growth.

The Potential Value of Preoperative MRI Texture and Shape Analysis in Grading Meningiomas: A Preliminary Investigation

OBJECT: Preoperative knowledge of meningioma grade is essential for planning treatment and surgery. The purpose of this study was to investigate the diagnostic value of MRI texture and shape analysis in grading meningiomas. METHODS: A surgical database was reviewed to identify meningioma patients who had undergone tumor resection between January 2015 and December 2016. Preoperative MR images were retrieved and analyzed. Texture and shape analysis was conducted to quantitatively evaluate tumor heterogeneity and morphology. Three machine learning classifiers were trained with these features to build classification models. The performance of the features and classification models was assessed. RESULTS: A total of 131 patients were included in this study: 21 with high-grade meningiomas and 110 with low-grade meningiomas. Three texture features were selected: Horzl_RLNonUni, S(2,2)SumOfSqs, and WavEnHL_s-3; three shape features were selected: GeoFv, GeoW4, and GeoW5b. The Mann–Whitney test indicated that all six features were significantly different between high-grade and low-grade meningiomas. AUC values were generally greater than 0.50 (range, 0.73 to 0.88). Sensitivities and specificities ranged from 47.62% to 90.48% and 69.09% to 96.36%, respectively. Among the nine classification models obtained, the one built by training the SVM classifier with all six features achieved the best performance, with a sensitivity, specificity, diagnostic accuracy, and AUC of 0.86, 0.87, 0.87, and 0.87, respectively. CONCLUSIONS: Texture and shape analysis, especially when combined with a SVM classifier, can provide satisfactory performance in the preoperative determination of meningioma grade and is thus potentially useful for clinical application.

Signal Intensities in Preoperative MRI Do Not Reflect Proliferative Activity in Meningioma

BACKGROUND

Identification of high-grade meningiomas in preoperative magnetic resonance imaging (MRI) is important for optimized surgical strategy and best possible resection. Numerous studies investigated subjectively determined morphological features as predictors of tumor biology in meningiomas. The aim of this study was to identify the predictive value of more reliable, quantitatively measured signal intensities in MRI for differentiation of high- and low-grade meningiomas and identification of meningiomas with high proliferation rates, respectively.

PATIENTS AND METHODS

Sixty-six patients (56 World Health Organization [WHO] grade I, 9 WHO grade II, and 1 WHO grade I) were included in the study. Preoperative MRI signal intensities (fluid-attenuated inversion recovery [FLAIR], T1 precontrast, and T1 postcontrast as genuine and normalized values) were correlated with Ki-67 expression in tissue sections of resected meningiomas. Differences between the groups (analysis of variance) and Spearman rho correlation were computed using SPSS 22.

RESULTS

Mean values of genuine signal intensities of meningiomas in FLAIR, T1 native, and T1 postcontrast were 323.9 ± 59, 332.8 ± 67.9, and 768.5 ± 165.3. Mean values of normalized (to the contralateral white matter) signal intensities of meningiomas in FLAIR, T1 native, and T1 postcontrast were 1.5 ± 0.3, 0.8 ± 0.1, and 1.9 ± 0.4. There was no significant correlation between MRI signal intensities and WHO grade or Ki-67 expression. Signal intensities did not differ significantly between WHO grade I and II/III meningiomas. Ki-67 expression was significantly increased in high-grade meningiomas compared with low-grade meningiomas (P < 0.01). Objectively measured values of MRI signal intensities (FLAIR, T1 precontrast, and T1 postcontrast enhancement) did not distinguish between high-grade and low-grade meningiomas. Furthermore, there was no association between MRI signal intensities and Ki-67 expression representing proliferative activity.

Prediction of pediatric meningioma recurrence by preoperative MRI assessment

Preoperative identification of high-recurrent pediatric meningiomas with MRI features would help clinicians to make optimal treatment strategies; however, the relationships between radiological features and recurrence of meningiomas in pediatric population have not been clearly demonstrated yet. The aim of this study is to identify preoperative MRI features which are significant risk factors for recurrence of pediatric meningiomas. From January 2005 to December 2012, we retrospectively reviewed 52 pediatric meningiomas in terms of preoperative MRI features and their clinical data and followed them up from 22 to 128 months (mean 63 months) after the initial surgery. The relationships between these radiological findings and relapse-free survival (RFS) time were assessed initially with univariate Cox analysis and then corrected by multivariate Cox analysis. According to univariate analysis, irregular shape, narrow-based attachment, and skull base location were significantly correlated with shorter time to recurrences of meningiomas in pediatric patients. When corrected by multivariate analysis, irregular shape (P = 0.05; OR 3.442, 95 % CI 1.001-11.831) and narrow-based attachment (P = 0.004; OR 7.164, 95 % CI 1.894-27.09) were strong independent predictive factors for worse RFS of pediatric meningiomas. In pediatric population, narrow-based attachment and irregular shape were significantly correlated with recurrences of meningiomas. Our results could help clinicians to make optimal therapeutic strategies for pediatric patients with intracranial meningiomas before surgery.

Imaging and extent of surgical resection predict risk of meningioma recurrence better than WHO histopathological grade

BACKGROUND

Risk stratification of meningiomas by histopathological grade alone does not reliably predict which patients will progress/recur after treatment. We sought to determine whether preoperative imaging and clinical characteristics could predict histopathological grade and/or improve prognostication of progression/recurrence (P/R).

METHODS

We retrospectively reviewed preoperative MR and CT imaging features of 144 patients divided into low-grade (2007 WHO grade I; n = 118) and high-grade (2007 WHO grades II/III; n = 26) groups that underwent surgery between 2002 and 2013 (median follow-up of 49 months).

RESULTS

Multivariate analysis demonstrated that the risk factors most strongly associated with high-grade histopathology were male sex, low apparent diffusion coefficient (ADC), absent calcification, and high peritumoral edema. Remarkably, multivariate Cox proportional hazards analysis demonstrated that, in combination with extent of resection, ADC outperformed WHO histopathological grade for predicting which patients will suffer P/R after initial treatment. Stratification of patients into 3 risk groups based on non-Simpson grade I resection and low ADC as risk factors correlated with the likelihood of P/R (P < .001). The high-risk group (2 risk factors; n = 39) had a 45% cumulative incidence of P/R, whereas the low-risk group (0 risk factors; n = 31) had no P/R events at 5 years after treatment. Independent of histopathological grade, high-risk patients who received adjuvant radiotherapy had a lower 5-year crude rate of P/R than those without (17% vs 59%; P = .04).

CONCLUSIONS

Patients with non-Simpson grade I resection and low ADC meningiomas are at significantly increased risk of P/R and may benefit from adjuvant radiotherapy and/or additional surgery.

Use of Diffusion Weighted Imaging in Differentiating Between Maligant and Benign Meningiomas. A Multicenter Analysis

BACKGROUND

Meningioma is the most frequent intracranial tumor and is often an incidental finding on imaging. Some imaging-based scores were suggested for differentiating low- and high-grade meningiomas. The purpose of this work was to compare diffusion-weighted imaging findings of different meningiomas in a large multicenter study by using apparent diffusion coefficient (ADC) values for predicting tumor grade and proliferation potential.

METHODS

Data from 7 radiologic departments were acquired retrospectively. Overall, 389 patients were collected. All meningiomas were investigated by magnetic resonance imaging (1.5-T scanner) by using diffusion-weighted imaging (b values of 0 and 1000 s/mm(2)). The comparison of ADC values was performed by Mann-Whitney U test.

RESULTS

World Health Organization grade I was diagnosed in 271 cases (69.7%), grade II in 103 (26.5%), and grade III in 15 patients (3.9%). Grade I meningiomas showed statistically significant higher ADC values (1.05 ± 0.39 × 10(-3) mm(2)s(-1)) in comparison with grade II (0.77 ± 0.15 × 10(-3) mm(2)s(-1); P = 0.001) and grade III tumors (0.79 ± 0.21 × 10(-3) mm(2)s(-1); P = 0.01). An ADC value of <0.85 × 10(-3) mm(2)s(-1) was determined as the threshold in differentiating between grade I and grade II/III meningiomas (sensitivity, 72.9%; specificity, 73.1%; accuracy, 73.0%). Ki67 was associated with ADC (r = -0.63, P < 0.001). The optimal threshold for the ADC was (less than) 0.85 × 10(-3) mm(2)s(-1) for detecting tumors with high proliferation potential (Ki67 ≥5%).

CONCLUSIONS

The estimated threshold ADC value of 0.85 can differentiate grade I meningioma from grade II and III tumors. The same ADC value is helpful for detecting tumors with high proliferation potential.

Preoperative radiologic classification of convexity meningioma to predict the survival and aggressive meningioma behavior

Background

A subgroup of meningioma demonstrates clinical aggressive behavior. We set out to determine if the radiological parameters can predict histopathological aggressive meningioma, and propose a classification to predict survival and aggressive meningioma behavior.

Methods

A retrospective review of medical records was conducted for patients who underwent surgical resection of their convexity meningioma. WHO-2007 grading was used for histopathological diagnosis. Preoperative radiologic parameters were analyzed, each parameter was scored 0 or 1. Signal intensity on diffusion weighted MRI (DWI) (hyperintensity=1), heterogeneity on T1-weighted gadolinium enhanced MRI (heterogeneity=1), disruption of arachnoid at brain-tumor interface=1and peritumoral edema (PTE) on T2-weighted MRI (presence of PTE=1) and tumor shape (irregular shape=1). Multivariate logistic regression analyses were conducted to determine association of radiological parameters to histopathological grading. Kaplan-Meier and Cox regression models were used to determine the association of scoring system to overall survival and progression free survival (PFS). Reliability of the classification was tested using Kappa co-efficient analysis.

Results

Hyperintensity on DWI, disruption of arachnoid at brain-tumor interface, PTE, heterogenicitiy on T1-weighted enhanced MRI and irregular tumor shape were independent predictors of non-grade I meningioma. Mean follow-up period was 94.6 months (range, 12-117 months). Median survival and PFS in groups-I, II and III was 114.1±1.2 and 115.7± 0.8, 88± 3.3 and 58.5±3.9, 43.2± 5.1 and 18.2±1.7 months respectively. In cox regression analysis model, age (P<0.0001, OR–1.039, CI-1.017-0.062), WHO non-grade-I meningioma (P=0.017, OR–3.014, CI-1.217-7.465), radiological classification groups II (P=0.002, OR–6.194, CI–1.956-19.610) and III (P<0.0001, OR–21.658, CI–5.701-82.273) were independent predictors of unfavorable survival outcomes.

Conclusions

Preoperative radiological classification can be used as a supplement to the histopathological grading. Group-I meningiomas demonstrate benign radiological, histopathological and clinical features; group-III demonstrates aggressive features. Group-II meningiomas demonstrate intermediate features; the need for more aggressive follow-up and/or treatment should be further investigated.

Correlation between magnetic resonance imaging grading and pathological grading in meningioma

OBJECT

This study investigated the specific preoperative MRI features of patients with intracranial meningiomas that correlate with pathological grade and provide appropriate preoperative planning.

METHODS

From 2006 to 2012, 120 patients (36 men and 84 women, age range 20-89 years) with newly diagnosed symptomatic intracranial meningiomas undergoing resection were retrospectively analyzed in terms of radiological features of preoperative MRI. There were 90 WHO Grade I and 30 WHO Grade II or III meningiomas. The relationships between MRI features and WHO histopathological grade were analyzed and scored quantitatively.

RESULTS

According to the results of multivariate logistic regression analysis, age ≥ 75 years, indistinct tumorbrain interface, positive capsular enhancement, and heterogeneous tumor enhancement were identified factors in the prediction of advanced histopathological grade. The prediction model was quantified as a scoring scale: 2 × (age) + 5 × (tumor-brain interface) + 3 × (capsular enhancement) + 2 × (tumor enhancement). The calculated score correlated positively with the probability of high-grade meningioma.

CONCLUSIONS

This scoring approach may be useful for clinicians in determining therapeutic strategy and in surgical planning for patients with intracranial meningiomas.

Prediction of high-grade meningioma by preoperative MRI assessment

High-grade (World Health Organization grades II and III) meningiomas grow aggressively and recur frequently, resulting in a poor prognosis. Assessment of tumor malignancy before treatment initiation is important. We attempted to determine predictive factors for high-grade meningioma on magnetic resonance (MR) imaging before surgery. We reviewed 65 meningiomas (39 cases, benign; 26 cases, high-grade) and assessed four factors: (1) tumor-brain interface (TBI) on T1-weighted imaging (T1WI), (2) capsular enhancement (CapE), i.e., the layer of the tumor-brain interface on gadolinium-enhanced T1WI (T1Gd), (3) heterogeneity on T1Gd, and (4) tumoral margin on T1Gd. All four factors were useful in distinguishing high-grade from benign meningiomas, according to univariate analysis. On multivariate regression analysis, unclear TBI and heterogeneous enhancement were independent predictive factors for high-grade meningioma. In meningiomas with an unclear TBI and heterogeneous enhancement, the probability of high-grade meningioma was 98%. Our data suggest that this combination of factors obtained from conventional sequences on MR imaging may be useful to predict high-grade meningioma.

Radiological predictive factors for regrowth of residual benign meningiomas

The pre- and postoperative radiological predictive factors for the regrowth of residual benign meningiomas were investigated in 80 of 327 patients who underwent first surgery for intracranial meningioma, who met the following conditions: residual tumor observed on postoperative imaging, follow up for more than 5 years or until regrowth of the residual tumor, histological diagnosis of World Health Organization grade I, and no additional therapy performed within 1 month after surgery. These 80 patients were divided into those with no regrowth during the follow-up period (Group A, n = 54) and those with regrowth (Group B, n = 26), and the clinical characteristics and pre- and postoperative imaging findings were compared. Univariate analysis of factors influencing regrowth showed 6 factors were significant: tumor size ≥4 cm (p = 0.043), tumor volume ≥30 cm(3) (p = 0.026), presence of edema (p = 0.036), unclear brain-tumor interface (p < 0.001), presence of a pial-cortical blood supply (p = 0.031), and residual tumor volume ≥3.0 cm(3) (p < 0.001). Multivariate analysis showed only residual tumor volume ≥3.0 cm(3) was significant (p = 0.001). Generally, the significant imaging findings on univariate analysis suggest malignant meningioma. Similar findings may be observed even in grade I cases, and residual tumors may regrow in such cases. The possibility is particularly high if the residual tumor volume exceeds 3.0 cm(3), so early radiotherapy should be performed to prevent regrowth.

Use of preoperative magnetic resonance imaging T1 and T2 sequences to determine intraoperative meningioma consistency

Background:

Meningioma firmness is a critical factor that influences ease of resection and risk, notably when operating on tumors intimate with neurovascular structures such as the mesial sphenoid wing. This study develops a predictive tool using preoperative magnetic resonance imaging (MRI) characteristics to determine meningioma consistency.

Methods:

101 patients with intracranial meningioma (50 soft/51 firm) were included. MRI characteristics of 38 tumors (19 soft/19 firm) were retrospectively reviewed to identify preoperative imaging features that were then correlated with intraoperative description of the tumor as either “soft and/or suckable” or “firm and/or fibrous”. Criteria were developed to predict consistency and then blindly applied to the remaining 63 meningiomas (31 soft/32 firm).

Results:

The overall sensitivities for detecting soft and firm consistency were 90% and 56%, respectively (95% CI = 73–97% and 38–73%; P < 0.001). Compared to gray matter, meningiomas that were T2 hypointense were almost always firm. Soft meningiomas were hyperintense on T2 and hypointense on T1. Soft meningiomas were slightly larger and less likely to be associated with edema. There was a slight preponderance of firm meningiomas in the infratentorial compartment. Grade of meningioma was not predictive. Contrast enhancement, diffusion restriction, changes in overlying bone, intratumoral cysts, and angiographic features were not predictable.

Conclusions:

This tool using T1 and T2 series predicts meningioma consistency. Such knowledge should assist the surgeon in preoperative planning and counseling.

Significance of tumor volume related to peritumoral edema in intracranial meningioma treated with extreme hypofractionated stereotactic radiation therapy in three to five fractions

BACKGROUND

To investigate the treatment results of intracranial meningiomas treated with hypofractionated stereotactic radiation therapy in three to five fractions.

METHODS

Thirty-one patients (32 lesions) with intracranial meningioma were treated with hypofractionated stereotactic radiation therapy in three to five fractions using CyberKnife. Fifteen lesions were diagnosed as Grade I (World Health Organization classification) by surgical resection and 17 lesions were diagnosed as meningioma based on radiological findings. The median follow-up time was 48 months. The median planning target volume was 6.3 cm(3) (range, 1.4-27.1), and the prescribed dose (D90≤) ranged from 21 to 36 Gy (median, 27.8) administrated in three to five fractions.

RESULTS

Five-year overall and progression-free survival rate of all 31 patients with intracranial meningioma was 86 and 83%, respectively. Five-year progression-free rate of all 32 lesions was 87%. Six of the 31 patients (19%) developed marked peritumoral edema, three of whom were asymptomatic and three symptomatic, the latter with late adverse effects of more than or equal to Grade 3. The mean planning target volume of the six lesions with marked peritumoral edema was 15.6 cm(3), and for the remaining 26 lesions without marked peritumoral edema was 7.1 cm(3) (P = 0.004). The threshold diameter of 2.56 cm for meningioma was calculated from the planning target volume (11 cm(3)) and was used as marker of developing peritumoral edema (P = 0.003).

CONCLUSIONS

Tumor volume is a significant indicative factor for peritumoral edema in intracranial meningioma treated with hypofractionated stereotactic radiation therapy in three to five factions.

An unusual growth of an intraventricular meningioma: a case report

Intraventricular meningiomas are rare often histologically benign tumors arising most always from the trigonal region of the lateral ventricle. We report the first described case of a rapidly growing histologically benign intraventricular meningioma in a 68-year-old woman whose magnetic resonance imaging (MRI) executed 1 year before surgical operation was negative for intracranial mass lesion.

¹H-MRS of intracranial meningiomas: what it can add to known clinical and MRI predictors of the histopathological and biological characteristics of the tumor?

OBJECTIVE

The main goal of the present study was evaluation of proton magnetic resonance spectroscopy (¹H-MRS) in diagnosis of histopathologically aggressive intracranial meningiomas.

METHODS

Single-voxel ¹H-MRS of 100 intracranial meningiomas was performed before their surgical resection. Investigated metabolites included mobile lipids, lactate, alanine, N-acetylaspartate (NAA), and choline-containing compounds (Cho). According to criteria of World Health Organization (WHO) 82 meningiomas were assigned histopathological grade I, 11 grade II, and 7 grade III. The MIB-1 index varied from 0% to 27.3% (median, 1.6%). In 43 cases tight adhesion of the tumor to the pia mater or brain tissue was macroscopically identified at surgery. The consistency of 49 meningiomas was characterized as soft, 26 as hard, and 25 as mixed.

RESULTS

No one metabolic parameter had statistically significant association with histopathological grade and subtype, invasive growth, and consistency of meningioma. Univariate statistical analysis revealed greater ¹H-MRS-detected Cho content (P=0.0444) and lower normalized NAA/Cho ratio (P=0.0203) in tumors with MIB-1 index 5% and more. However, both parameters lost their statistical significance during evaluation in the multivariate model along with other clinical and radiological variables. It was revealed that non-benign histopathology of meningioma (WHO grade II/III) is mainly predicted by irregular shape (P=0.0076) and large size (P=0.0316), increased proliferative activity by irregular shape (P=0.0056), and macroscopically invasive growth by prominent peritumoral edema (P=0.0021).

CONCLUSION

While ¹H-MRS may be potentially used for the identification of meningiomas with high proliferative activity, it, seemingly, could not add substantial diagnostic information to other radiological predictors of malignancy in these tumors.

Serial volumetric assessment of the natural history and growth pattern of incidentally discovered meningiomas

OBJECT

Due to advances in neuroimaging and the increasing use of imaging to screen for brain disease ("brain checkups"), meningiomas are now often detected as an incidental finding. The natural history of these asymptomatic meningiomas remains unclear, however. In this study, the authors investigated the natural history and growth pattern of incidentally detected meningiomas using serial volumetric assessment and regression analysis.

METHODS

In 70 patients with incidentally discovered meningiomas who underwent follow-up for longer than 1 year, tumor volumes were calculated volumetrically at each follow-up visit, and tumor growth was determined. In patients with tumor growth, regression analysis was performed to determine the pattern of growth.

RESULTS

Forty-four tumors exhibited growth and 26 did not. In a regression analysis, 16 of the tumors that grew followed an exponential growth pattern and 15 exhibited linear growth patterns. The presence of calcification was the only imaging characteristic that significantly distinguished the group with tumor growth from that without, although no radiological characteristics significantly distinguished the exponential growth group from the linear growth group. Two patients with obvious tumor growth underwent surgical removal and the pathological specimens extracted showed a high proliferative potential.

CONCLUSIONS

The authors found that incidentally discovered meningiomas did not always follow an exponential growth pattern but often exhibited more complex patterns of growth. Serial monitoring of tumor volumes and regression analysis may reveal the growth pattern of incidental meningiomas and provide information useful for determining treatment strategy.