Volumetric growth rates of meningioma and its correlation with histological diagnosis and clinical outcome: a systematic review

Lesion delineation framework for vestibular schwannoma, meningioma and brain metastasis for gamma knife radiosurgery using stereotactic magnetic resonance images

BACKGROUND AND OBJECTIVE

GKRS is an effective treatment for smaller intracranial tumors with a high control rate and low risk of complications. Target delineation in medical MR images is essential in the planning of GKRS and follow-up. A deep learning-based algorithm can effectively segment the targets from medical images and has been widely explored. However, state-of-the-art deep learning-based target delineation uses fixed sizes, and the isotropic voxel size may not be suitable for stereotactic MR images which use different anisotropic voxel sizes and numbers of slices according to the lesion size and location for clinical GKRS planning. This study developed an automatic deep learning-based segmentation scheme for stereotactic MR images.

METHODS

We retrospectively collected stereotactic MR images from 506 patients with VS, 1,069 patients with meningioma and 574 patients with BM who had been treated using GKRS; the lesion contours and individual T1W+C and T2W MR images were extracted from the GammaPlan system. The three-dimensional patching-based training strategy and dual-pathway architecture were used to manage inconsistent FOVs and anisotropic voxel size. Furthermore, we used two-parametric MR image as training input to segment the regions with different image characteristics (e.g., cystic lesions) effectively.

RESULTS

Our results for VS and BM demonstrated that the model trained using two-parametric MR images significantly outperformed the model trained using single-parametric images with median Dice coefficients (0.91, 0.05 versus 0.90, 0.06, and 0.82, 0.23 versus 0.78, 0.34, respectively), whereas predicted delineations in meningiomas using the dual-pathway model were dominated by single-parametric images (median Dice coefficients 0.83, 0.17 versus 0.84, 0.22). Finally, we combined three data sets to train the models, achieving the comparable or even higher testing median Dice (VS: 0.91, 0.07; meningioma: 0.83, 0.22; BM: 0.84, 0.23) in three diseases while using two-parametric as input.

CONCLUSIONS

Our proposed deep learning-based tumor segmentation scheme was successfully applied to multiple types of intracranial tumor (VS, meningioma and BM) undergoing GKRS and for segmenting the tumor effectively from stereotactic MR image volumes for use in GKRS planning.

Volumetric measurement of intracranial meningiomas: a comparison between linear, planimetric, and machine learning with multiparametric voxel-based morphometry methods

PURPOSE

To compare the accuracy of three volumetric methods in the radiological assessment of meningiomas: linear (ABC/2), planimetric, and multiparametric machine learning-based semiautomated voxel-based morphometry (VBM), and to investigate the relevance of tumor shape in volumetric error.

METHODS

Retrospective imaging database analysis at the authors' institutions. We included patients with a confirmed diagnosis of meningioma and preoperative cranial magnetic resonance imaging eligible for volumetric analyses. After tumor segmentation, images underwent automated computation of shape properties such as sphericity, roundness, flatness, and elongation.

RESULTS

Sixty-nine patients (85 tumors) were included. Tumor volumes were significantly different using linear (13.82 cm³ [range 0.13-163.74 cm³]), planimetric (11.66 cm³ [range 0.17-196.2 cm³]) and VBM methods (10.24 cm³ [range 0.17-190.32 cm³]) (p < 0.001). Median volume and percentage errors between the planimetric and linear methods and the VBM method were 1.08 cm³ and 11.61%, and 0.23 cm³ and 5.5%, respectively. Planimetry and linear methods overestimated the actual volume in 79% and 63% of the patients, respectively. Correlation studies showed excellent reliability and volumetric agreement between manual- and computer-based methods. Larger and flatter tumors had greater accuracy on planimetry, whereas less rounded tumors contributed negatively to the accuracy of the linear method.

CONCLUSION

Semiautomated VBM volumetry for meningiomas is not influenced by tumor shape properties, whereas planimetry and linear methods tend to overestimate tumor volume. Furthermore, it is necessary to consider tumor roundness prior to linear measurement so as to choose the most appropriate method for each patient on an individual basis.

Clinical Management of Supratentorial Non-Skull Base Meningiomas

Supratentorial non-skull base meningiomas are the most common primary central nervous system tumor subtype. An understanding of their pathophysiology, imaging characteristics, and clinical management options will prove of substantial value to the multi-disciplinary team which may be involved in their care. Extensive review of the broad literature on the topic is conducted. Narrowing the scope to meningiomas located in the supratentorial non-skull base anatomic location highlights nuances specific to this tumor subtype. Advances in our understanding of the natural history of the disease and how findings from both molecular pathology and neuroimaging have impacted our understanding are discussed. Clinical management and the rationale underlying specific approaches including observation, surgery, radiation, and investigational systemic therapies is covered in detail. Future directions for probable advances in the near and intermediate term are reviewed.

MAC-spinal meningioma score: A proposal for a quick-to-use scoring sheet of the MIB-1 index in sporadic spinal meningiomas

Objective

MIB-1 index is an important predictor of meningioma progression. However, MIB-1 index is not available in the preoperative tailored medical decision-making process. A preoperative scoring sheet independently estimating MIB-1 indices in spinal meningioma (SM) patients has not been investigated so far.

Methods

Between 2000 and 2020, 128 patients with clinical data, tumor imaging data, inflammatory laboratory (plasma fibrinogen, serum C-reactive protein) data, and neuropathological reports (MIB-1, mitotic count, CD68 staining) underwent surgery for spinal WHO grade 1 and 2 meningioma.

Results

An optimal MIB-1 index cut-off value (≥5/<5) predicting recurrence was calculated by ROC curve analysis (AUC: 0.83; 95%CI: 0.71-0.96). An increased MIB-1 index (≥5%) was observed in 55 patients (43.0%) and multivariable analysis revealed significant associations with baseline Modified McCormick Scale ≥2, age ≥65, and absence of calcification. A four-point scoring sheet (MAC-Spinal Meningioma) based on Modified McCormick, Age, and Calcification facilitates prediction of the MIB-1 index (sensitivity 71.1%, specificity 60.0%). Among those patients with a preoperative MAC-Meningioma Score ≥3, the probability of a MIB-1 index ≥5% was 81.3%.

Conclusion

This novel score (MAC-Spinal Meningioma) supports the preoperative estimation of an increased MIB-1 index, which might support preoperative patient-surgeon consultation, surgical decision making and enable a tailored follow-up schedule or an individual watch-and-wait strategy.

Preoperative Prediction of Intracranial Meningioma Grade Using Conventional CT and MRI

Objective

Preoperative diagnosis of tumor grade can assist in treatment-related decision-making for patients with intracranial meningioma. This study aimed to distinguish between high-grade and low-grade meningiomas using conventional CT and MRI.

Methodology

We retrospectively analyzed 173 consecutive patients with intracranial meningioma (149 low-grade and 24 high-grade tumors) who were treated surgically at the National Hospital Organization Kyushu Medical Center from 2008 to 2020. Clinical and radiological features, including tumor doubling time (Td) and relative growth rate (RGR), were compared between low-grade and high-grade meningiomas.

Results

Multivariate logistic regression analysis showed that symptomatic tumor (p=0.001), non-skull base location (p=0.006), irregular tumor shape (p=0.043), tumor heterogeneity (p=0.025), and peritumoral brain edema (p=0.003) were independent predictors of high-grade meningioma. In 53 patients who underwent surgery because of tumor progression, progression to symptoms (p=0.027), intratumoral heterogeneity (p<0.001), peritumoral brain edema (p=0.001), larger tumor volume (p=0.005), shorter Td (p<0.001), and higher RGR (P<0.001) were significantly associated with high-grade meningioma. Receiver operating characteristics (ROC) curve analysis showed that the optimal Td and annual RGR cut-off values to distinguish high-grade from low-grade meningioma were 460.5 days and 73.2%, respectively (100% sensitivity and 78.6% specificity).

Conclusion

Based on our findings, conventional CT and MRI are useful methods to predict meningioma grades before surgery. High-grade lesions are associated with non-skull base location, irregular tumor shape, intratumoral heterogeneity, and peritumoral brain edema. High-grade meningioma should be suspected in tumors that exhibit Td <460.5 days or annual RGR >73.2% or those that develop intratumoral heterogeneity or surrounding brain edema on surveillance imaging.

Biology and Treatment of Meningiomas: A Reappraisal

Meningiomas are largely indolent tumors with a benign clinical course, but a minority exhibit aggressive behavior characterized by rapid growth, neurologic deficits, and increased mortality. Identifying high-risk patients requiring intervention is challenging, but recent insights into meningioma biology provide a useful guide for decision making. Standard of care for recurrent or biologically aggressive tumors consists of surgery and radiation therapy. Systemic therapies targeting vascular endothelial growth factor signaling and somatostatin analogues are potential options for those with refractory disease but display only modest activity. New paradigms in meningioma clinical trial design provide hope for improved options in the future.

Consensus core clinical data elements for meningiomas

BACKGROUND

With increasing molecular analyses of meningiomas, there is a need to harmonize language used to capture clinical data across centers to ensure that molecular alterations are appropriately linked to clinical variables of interest. Here the International Consortium on Meningiomas presents a set of core and supplemental meningioma-specific Common Data Elements (CDEs) to facilitate comparative and pooled analyses.

METHODS

The generation of CDEs followed the four-phase process similar to other National Institute of Neurological Disorders and Stroke (NINDS) CDE projects: discovery, internal validation, external validation, and distribution.

RESULTS

The CDEs were organized into patient- and tumor-level modules. In total, 17 core CDEs (10 patient-level and 7-tumour-level) as well as 14 supplemental CDEs (7 patient-level and 7 tumour-level) were defined and described. These CDEs are now made publicly available for dissemination and adoption.

CONCLUSIONS

CDEs provide a framework for discussion in the neuro-oncology community that will facilitate data sharing for collaborative research projects and aid in developing a common language for comparative and pooled analyses. The meningioma-specific CDEs presented here are intended to be dynamic parameters that evolve with time and The Consortium welcomes international feedback for further refinement and implementation of these CDEs.

Memory Modulation Factors in Hippocampus Exposed to Radiation

Although the key scene of the hippocampus in memory processes is obvious, the specificity of its participation in information processing is far from being established. Current advanced neuroimaging enables to operate with precise morphometric parameters. The aim of the study was to reveal fine memory rearrangements under mechanical impact on the hippocampus by a neoplasm and radiation exposure in the course of therapy.

Materials and Methods

We used a homogeneous sample of 28 patients with parasellar meningiomas adjacent to hippocampus. In 10 patients (5 with left-sided and 5 with right-sided meningiomas), the tumor was located near the hippocampus but exhibited no mechanical effect on it. In 18 patients (10 with left-sided and 8 with right-sided tumors), the neoplasm compressed the adjacent hippocampus. The control group consisted of 39 healthy subjects. All three groups were comparable in age, education, and gender characteristics. In order to control tumor growth, the patients underwent radiotherapy when the hippocampus involuntary was exposed to a dose comparable to that in the tumor (30 sessions with a single focal dose of 1.8 Gy, total dose - 54.0 Gy).Based on the literature data on hippocampus involved in mnestic processes, a special methodology to investigate memory was developed. Incorrect responses the subjects made when identifying previously memorized images were classified as neutralizing the novelty factor of an identified stimulus or as wrongly emphasizing its novelty.

Results

At the first observation point (before radiation therapy) all groups underwent a complete standardized neuropsychological examination and performed a battery of cognitive tests. The overall results of the tests assessing attention, memory, thinking processes, and neurodynamic indicators corresponded to standard values. A mild brain compression by the tumor without brain tissue destruction was not accompanied by focal neuropsychological symptoms and deficit manifestations in the cognitive sphere. However, as early as in the first observation point, the number of "pattern separation" errors in the clinical group was significantly higher than that in healthy subjects.The second observation point (immediately after radiotherapy) and the third observation point - 6 months after the treatment - showed that, in general, the patients' cognitive sphere condition was not deteriorating, and in a number of parameters was characterized by positive dynamics, apparently associated with some tumor reduction due to the therapy provided. However, the distribution of errors in the original method significantly changed. When previously memorized stimuli were recognized, the errors neutralizing the novelty factor of the evaluated stimulus increased, while the number of errors with overestimating the stimuli novelty decreased.All tendencies hypothetically (according to the published data) associated with the changes in functional activity of the hippocampus were more pronounced in the subgroup of patients with mechanical impact of the tumor on hippocampus.

Conclusion

The continuous flow of impressions any person has at any moment of his activity is most likely marked by the hippocampus in a continuum "old-similar-new". The present study has shown that mechanical impact on the hippocampus combined with radiation exposure changes the range of assessments towards the prevailing labeling "old, previously seen, already known".

Clinical Outcomes Following Re-Operations for Intracranial Meningioma

Simple Summary

This study investigated patients who underwent more than one operation for a meningioma, a type of brain tumor. Currently, there is little evidence available for this specific patient group. The purpose of this study was to determine if patients had an improvement or deterioration following a second operation for a recurrent meningioma, and to identify any factors that may influence this change. The results demonstrated that following a second operation for meningioma, patients have poorer outcomes. The findings of this study provide supporting information for surgeons and patients, thereby informing decisions related to patient care and re-operation.

Abstract

The outcomes following re-operation for meningioma are poorly described. The aim of this study was to identify risk factors for a performance status outcome following a second operation for a recurrent meningioma. A retrospective, comparative cohort study was conducted. The primary outcome measure was World Health Organization performance. Secondary outcomes were complications, and overall and progression free survival (OS and PFS respectively). Baseline clinical characteristics, tumor details, and operation details were collected. Multivariable binary logistic regression was used to identify risk factors for performance status outcome following a second operation. Between 1988 and 2018, 712 patients had surgery for intracranial meningiomas, 56 (7.9%) of which underwent a second operation for recurrence. Fifteen patients (26.8%) had worsened performance status after the second operation compared to three (5.4%) after the primary procedure (p = 0.002). An increased number of post-operative complications following the second operation was associated with a poorer performance status following that procedure (odds ratio 2.2 [95% CI 1.1–4.6]). The second operation complication rates were higher than after the first surgery (46.4%, n = 26 versus 32.1%, n = 18, p = 0.069). The median OS was 312.0 months (95% CI 257.8–366.2). The median PFS following the first operation was 35.0 months (95% CI 28.9–41.1). Following the second operation, the median PFS was 68.0 months (95% CI 49.1–86.9). The patients undergoing a second operation for meningioma had higher rates of post-operative complications, which is associated with poorer clinical outcomes. The decisions surrounding second operations must be balanced against the surgical risks and should take patient goals into consideration.

Automatic Meningioma Segmentation and Grading Prediction: A Hybrid Deep-Learning Method

The purpose of this study was to determine whether a deep-learning-based assessment system could facilitate preoperative grading of meningioma. This was a retrospective study conducted at two institutions covering 643 patients. The system, designed with a cascade network structure, was developed using deep-learning technology for automatic tumor detection, visual assessment, and grading prediction. Specifically, a modified U-Net convolutional neural network was first established to segment tumor images. Subsequently, the segmentations were introduced into rendering algorithms for spatial reconstruction and another DenseNet convolutional neural network for grading prediction. The trained models were integrated as a system, and the robustness was tested based on its performance on an external dataset from the second institution involving different magnetic resonance imaging platforms. The results showed that the segment model represented a noteworthy performance with dice coefficients of 0.920 ± 0.009 in the validation group. With accurate segmented tumor images, the rendering model delicately reconstructed the tumor body and clearly displayed the important intracranial vessels. The DenseNet model also achieved high accuracy with an area under the curve of 0.918 ± 0.006 and accuracy of 0.901 ± 0.039 when classifying tumors into low-grade and high-grade meningiomas. Moreover, the system exhibited good performance on the external validation dataset.

Volumetric growth of residual meningioma - A systematic review

Surgical resection of meningioma leaves residual solid tumour in over 25% of patients. Selection for further treatment and follow-up strategy may benefit from knowledge of volumetric growth and factors associated with re-growth. The aim of this review was to evaluate volumetric growth and variables associated with growth in patients that underwent incomplete resection of a meningioma without the use of adjuvant radiotherapy. A systematic review was conducted in accordance with the PRISMA statement and registered a priori with PROSPERO (registration number: CRD42020177052). Six databases were searched up to May 2020. Full text articles analysing volumetric growth rates in at least 10 patients who had residual meningioma after surgery were assessed. Four single-centre, retrospective studies totalling 238 patients were included, of which 99% of meningioma were WHO grade 1. The absolute tumour growth rate ranged from 0.09 to 4.94 cm³ per year. The relative growth rate ranged from 5.11 to 14.18% per year. Varying methods of volumetric assessment and definitions of growth impeded pooled analysis. Pre-operative and residual tumour volume, and hyperintensity on T2 weighted MRI were identified as variables associated with residual meningioma growth, however this was inconsistent across studies. Risk of bias was high in all studies. Radiological regrowth occurred in 42-67% of cases. Our review identified that volumetric growth of residual meningioma is scarcely reported. Sufficiently powered studies are required to delineate volumetric growth and prognostic factors to stratify management.

Volumetric growth rate of incidental asymptomatic meningiomas: a single-center prospective cohort study

BACKGROUND

Decision about treatment of incidentally found intracranial meningiomas is controversial and conditioned by the growth potential of these tumors. We aimed to evaluate the growth rate of a cohort of incidentally found asymptomatic meningiomas and to analyze their natural course and the need for eventual treatment.

METHODS

A total of 193 patients harboring intracranial meningiomas (85 with 109 incidental and 108 with 112 symptomatic) were included between 2015 and 2019. In the prospective cohort of incidental meningiomas, we measured size at diagnosis, volumetric growth rate (by segmentation software), appearance of symptoms, and need for surgery or radiotherapy. Progression-free survival and risk factors for growth were assessed with Kaplan-Meier survival and Cox regression analyses.

RESULTS

Among incidental meningiomas, 94/109 (86.2%) remained untreated during a median follow-up of 49.3 months. Tumor growth was observed in 91 (83.5%) and > 15% growth in 40 (36.7%). Neurological symptoms developed in 1 patient (1.2%). Volume increased an average of 0.51 cm³/year (95% CI, 0.20-0.82). Nine patients were operated (9.2%) and 4 underwent radiotherapy (4.7%). Treatment-related complication rates of incidental and symptomatic meningiomas were 0% and 35.4%, respectively. Persistent neurological defects occurred in 46 (40.7%) of symptomatic versus 2 (2.3%) of incidental meningiomas. Among covariates, only brain edema resulted in an increased risk of significant tumor growth in the female subgroup (Cox regression HR 2.96, 95% CI 1.02-8.61, p = 0.046). Size at diagnosis was significantly greater in the symptomatic meningioma group (37.33 cm³ versus 4.74 cm³, p < 0.001).

CONCLUSIONS

Overall, 86% of incidentally found meningiomas remained untreated over the first 4 years of follow-up. The majority grew within the 20% range, yet very few developed symptoms. Treatment-related morbidity was absent in the incidental meningioma group.

The growth rate and clinical outcomes of radiation induced meningioma undergoing treatment or active monitoring

INTRODUCTION

Radiation induced meningioma (RIM) incidence is increasing in line with improved childhood cancer survival. No optimal management strategy consensus exists. This study aimed to delineate meningioma growth rates from tumor discovery and correlate with clinical outcomes.

METHODS

Retrospective study of patients with a RIM, managed at a specialist tertiary neuroscience center (2007-2019). Tumor volume was measured from diagnosis and at subsequent interval scans. Meningioma growth rate was determined using a linear mixed-effects model. Clinical outcomes were correlated with growth rates accounting for imaging and clinical prognostic factors.

RESULTS

Fifty-four patients (110 meningiomas) were included. Median duration of follow-up was 74 months (interquartile range [IQR], 41-102 months). Mean radiation dose was 41 Gy (standard deviation [SD] = 14.9) with a latency period of 34.4 years (SD = 13.7). Median absolute growth rate was 0.62 cm³/year and the median relative growth rate was 72%/year. Forty meningiomas (between 27 patients) underwent surgical intervention after a median follow-up duration of 4 months (IQR 2-35). Operated RIMs were clinically aggressive, likely to be WHO grade 2 at first resection (43.6%) and to progress after surgery (41%). Median time to progression was 28 months (IQR 13-60.5). A larger meningioma at discovery was associated with growth (HR 1.2 [95% CI 1.0-1.5], P = 0.039) but not progression after surgery (HR 2.2 [95% CI 0.7-6.6], P = 0.181). Twenty-seven (50%) patients had multiple meningiomas by the end of the study.

CONCLUSION

RIMs exhibit high absolute and relative growth rates after discovery. Surgery is recommended for symptomatic or rapidly growing meningiomas only. Recurrence risk after surgery is high.

Demographic and socioeconomic disparities of benign cerebral meningiomas in the United States

Epidemiology provides an avenue for deciphering disease pathogenesis. By determining incidence across socioeconomic and demographic variables in the context of benign cerebral meningiomas (BCM), epidemiologic data may aid in elucidating and addressing healthcare inequalities. To investigate BCM incidence (per 100,000) with respect to sex, age, income, residence, and race/ethnicity, we queried the largest United States (US) administrative dataset (1997-2016), the National (Nationwide) Inpatient Sample (NIS), which surveys 20% of US discharges. Annual national BCM incidence was 5.01. Females had an incidence of 6.78, higher (p = 0.0000038) than males at 3.14. Amongst age groups incidence varied (p = 1.65 × 10^-11) and was highest amongst those 65-84 (16.71) and 85+ (18.32). Individuals with middle/high income had an incidence of 5.27, higher (p = 0.024) than the 4.91 of low income patients. Depending on whether patients lived in urban, suburban, or rural communities, incidence varied (χ² = 8.22, p = 0.016) as follows, respectively: 5.23; 4.96; 5.51. Amongst race/ethnicity (p = 8.15 × 10^-14), incidence for Whites, Blacks, Asian/Pacific Islanders, Hispanics, and Native Americans were as follows, respectively: 5.05; 4.59; 4.22; 2.99; 0.55. In the US, BCM annual incidence exhibited disparities amongst socioeconomic and demographic subsets. Disproportionately, incidence was greatest for patients who were White, Black, female, 65 and older, and middle/high income.

Fast meningioma segmentation in T1-weighted magnetic resonance imaging volumes using a lightweight 3D deep learning architecture

Purpose: Automatic and consistent meningioma segmentation in T1-weighted magnetic resonance (MR) imaging volumes and corresponding volumetric assessment is of use for diagnosis, treatment planning, and tumor growth evaluation. We optimized the segmentation and processing speed performances using a large number of both surgically treated meningiomas and untreated meningiomas followed at the outpatient clinic. Approach: We studied two different three-dimensional (3D) neural network architectures: (i) a simple encoder-decoder similar to a 3D U-Net, and (ii) a lightweight multi-scale architecture [Pulmonary Lobe Segmentation Network (PLS-Net)]. In addition, we studied the impact of different training schemes. For the validation studies, we used 698 T1-weighted MR volumes from St. Olav University Hospital, Trondheim, Norway. The models were evaluated in terms of detection accuracy, segmentation accuracy, and training/inference speed. Results: While both architectures reached a similar Dice score of 70% on average, the PLS-Net was more accurate with an F 1 -score of up to 88%. The highest accuracy was achieved for the largest meningiomas. Speed-wise, the PLS-Net architecture tended to converge in about 50 h while 130 h were necessary for U-Net. Inference with PLS-Net takes less than a second on GPU and about 15 s on CPU. Conclusions: Overall, with the use of mixed precision training, it was possible to train competitive segmentation models in a relatively short amount of time using the lightweight PLS-Net architecture. In the future, the focus should be brought toward the segmentation of small meningiomas ( < 2    ml ) to improve clinical relevance for automatic and early diagnosis and speed of growth estimates.

Automated Detection and Segmentation of Brain Metastases in Malignant Melanoma: Evaluation of a Dedicated Deep Learning Model

BACKGROUND AND PURPOSE

Malignant melanoma is an aggressive skin cancer in which brain metastases are common. Our aim was to establish and evaluate a deep learning model for fully automated detection and segmentation of brain metastases in patients with malignant melanoma using clinical routine MR imaging.

MATERIALS AND METHODS

Sixty-nine patients with melanoma with a total of 135 brain metastases at initial diagnosis and available multiparametric MR imaging datasets (T1-/T2-weighted, T1-weighted gadolinium contrast-enhanced, FLAIR) were included. A previously established deep learning model architecture (3D convolutional neural network; DeepMedic) simultaneously operating on the aforementioned MR images was trained on a cohort of 55 patients with 103 metastases using 5-fold cross-validation. The efficacy of the deep learning model was evaluated using an independent test set consisting of 14 patients with 32 metastases. Manual segmentations of metastases in a voxelwise manner (T1-weighted gadolinium contrast-enhanced imaging) performed by 2 radiologists in consensus served as the ground truth.

RESULTS

After training, the deep learning model detected 28 of 32 brain metastases (mean volume, 1.0 [SD, 2.4] cm³) in the test cohort correctly (sensitivity of 88%), while false-positive findings of 0.71 per scan were observed. Compared with the ground truth, automated segmentations achieved a median Dice similarity coefficient of 0.75.

CONCLUSIONS

Deep learning-based automated detection and segmentation of brain metastases in malignant melanoma yields high detection and segmentation accuracy with false-positive findings of <1 per scan.

Demographic and socioeconomic disparities of benign and malignant spinal meningiomas in the United States

INTRODUCTION

Spinal meningiomas constitute the majority of primary spinal neoplasms, yet their pathogenesis remains elusive. By investigating the distribution of these tumors across sociodemographic variables can provide direction in etiology elucidation and healthcare disparity identification.

METHODS

To investigate benign and malignant spinal meningioma incidences (per 100,000) with respect to sex, age, income, residence, and race/ethnicity, we queried the largest American administrative dataset (1997-2016), the National (Nationwide) Inpatient Sample (NIS), which surveys 20% of United States (US) discharges.

RESULTS

Annual national incidence was 0.62 for benign tumors and 0.056 for malignant. For benign meningiomas, females had an incidence of 0.81, larger (P=0.000004) than males at 0.40; yet for malignant meningiomas, males had a larger (P=0.006) incidence at 0.062 than females at 0.053. Amongst age groups, peak incidence was largest for those 65-84 years old (2.03) in the benign group, but 45-64 years old (0.083) for the malignant group. For benign and malignant meningiomas respectively, individuals with middle/high income had an incidence of 0.67 and 0.060, larger (P=0.000008; P=0.04) than the 0.48 and 0.046 of low income patients. Incidences were statistically similar (P=0.2) across patient residence communities. Examining race/ethnicity (P=0.000003) for benign meningiomas, incidences for Whites, Asian/Pacific Islanders, Hispanics, and Blacks were as follows, respectively: 0.83, 0.42, 0.28, 0.15.

CONCLUSIONS

Across sociodemographic strata, healthcare inequalities were identified with regards to spinal meningiomas. For benign spinal meningiomas, incidence was greatest for patients who were female, 65-84 years old, middle/high income, living in rural communities, White, and Asian/Pacific Islander. Meanwhile, for malignant spinal meningiomas incidence was greatest for males, those 45-65 years old, and middle/high income.

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.

Validation of a scoring system to evaluate the risk of rapid growth of intracranial meningiomas in neurofibromatosis type 2 patients

OBJECTIVE

Intracranial meningiomas occur in about half of neurofibromatosis type 2 (NF2) patients and are very frequently multiple. Thus, estimating individual meningiomas' growth rates is of great interest to tailor therapeutic interventions. The Asan Intracranial Meningioma Scoring System (AIMSS) has recently been published to estimate the risk of tumor growth in sporadic meningiomas. The current study aimed to determine predictors of rapid meningioma growth in NF2 patients and to evaluate the AIMSS score in a specific NF2 cohort.

METHODS

The authors performed a retrospective analysis of 92 NF2 patients with 358 measured intracranial meningiomas that had been observed prospectively between 2012 and 2018. Tumor volumes were measured at diagnosis and at each follow-up visit. The growth rates were determined and evaluated with respect to the clinicoradiological parameters. Predictors of rapid tumor growth (defined as growth ≥ 2 cm3/yr) were analyzed using univariate followed by multivariate logistic regression to build a dedicated predicting model. Receiver operating characteristic (ROC) curves to predict the risk of rapid tumor growth with the AIMSS versus the authors' multivariate model were compared.

RESULTS

Sixty tumors (16.76%) showed rapid growth. After multivariate analysis, a larger tumor volume at diagnosis (p < 0.0001), presence of peritumoral edema (p = 0.022), absence of calcifications (p < 0.0001), and hyperintense or isointense signal on T2-weighted MRI (p < 0.005) were statistically significantly associated with rapid tumor growth. It is particularly notable that the genetic severity score did not seem to influence the growth rate of NF2 meningiomas. In comparison with the AIMSS, the authors' multivariate model's prediction did not show a statistically significant difference (area under the curve [AUC] 0.82 [95% CI 0.76-0.88] for the AIMSS vs AUC 0.86 [95% CI 0.81-0.91] for the authors' model, p = 0.1).

CONCLUSIONS

The AIMSS score is valid in the authors' cohort of NF2-related meningiomas. It adequately predicted risk of rapid meningioma growth and could aid in decision-making in NF2 patients.

Histological Grade of Meningioma: Prediction by Intravoxel Incoherent Motion Histogram Parameters

RATIONALE AND OBJECTIVES

To evaluate the usefulness of intravoxel incoherent motion (IVIM) histogram analysis for differentiating low-grade meningiomas (LGMs) and high-grade meningiomas (HGMs).

MATERIALS AND METHODS

Fifty-nine patients with pathologically confirmed meningiomas (45 LGMs and 14 HGMs) underwent IVIM MR imaging. Maps of IVIM parameters (perfusion fraction, f; true diffusion coefficient, D; and pseudo diffusion coefficient, D*), as well as of the apparent diffusion coefficient (ADC), were generated. Histogram analysis was performed using parametric values from all voxels in regions-of-interest manually drawn to encompass the whole tumor. The histogram results of ADC and IVIM parameters were compared using the Mann-Whitney U test. Area under the receiver operating characteristic curve (AUC) values were generated to evaluate how well each parameter could differentiate LGMs from HGMs. Spearman's rank correlation coefficients were used to evaluate correlations between histogram parameters and Ki-67 expression.

RESULTS

Compared to LGM, HGM showed significantly higher standard deviation (SD), variance, and coefficient of variation (CV) of ADC (p< 0.006-0.028; AUC, 0.693-0.748), D (p< 0.004-0.032; AUC, 0.670-0.752), and significantly higher CV of f (p< 0.005-0.024; AUC = 0.737). Means and percentiles of ADC and IVIM parameters did not differ significantly between LGM and HGM. Significant positive correlations were identified between Ki-67 and histogram parameters of ADC (SD, variance, kurtosis, skewness, and CV) and D (SD, variance, kurtosis, and CV), whereas no significant correlation with Ki-67 was shown for mean or percentiles of ADC and IVIM parameters.

CONCLUSION

Heterogeneity histogram parameters of ADC, D, and f may be useful for differentiating LGMs from HGMs.

Automated Meningioma Segmentation in Multiparametric MRI : Comparable Effectiveness of a Deep Learning Model and Manual Segmentation

PURPOSE

Volumetric assessment of meningiomas represents a valuable tool for treatment planning and evaluation of tumor growth as it enables a more precise assessment of tumor size than conventional diameter methods. This study established a dedicated meningioma deep learning model based on routine magnetic resonance imaging (MRI) data and evaluated its performance for automated tumor segmentation.

METHODS

The MRI datasets included T1-weighted/T2-weighted, T1-weighted contrast-enhanced (T1CE) and FLAIR of 126 patients with intracranial meningiomas (grade I: 97, grade II: 29). For automated segmentation, an established deep learning model architecture (3D deep convolutional neural network, DeepMedic, BioMedIA) operating on all four MR sequences was used. Segmentation included the following two components: (i) contrast-enhancing tumor volume in T1CE and (ii) total lesion volume (union of lesion volume in T1CE and FLAIR, including solid tumor parts and surrounding edema). Preprocessing of imaging data included registration, skull stripping, resampling, and normalization. After training of the deep learning model using manual segmentations by 2 independent readers from 70 patients (training group), the algorithm was evaluated on 56 patients (validation group) by comparing automated to ground truth manual segmentations, which were performed by 2 experienced readers in consensus.

RESULTS

Of the 56 meningiomas in the validation group 55 were detected by the deep learning model. In these patients the comparison of the deep learning model and manual segmentations revealed average dice coefficients of 0.91 ± 0.08 for contrast-enhancing tumor volume and 0.82 ± 0.12 for total lesion volume. In the training group, interreader variabilities of the 2 manual readers were 0.92 ± 0.07 for contrast-enhancing tumor and 0.88 ± 0.05 for total lesion volume.

CONCLUSION

Deep learning-based automated segmentation yielded high segmentation accuracy, comparable to manual interreader variability.

Predictors of recurrence and high growth rate of residual meningiomas after subtotal resection

OBJECTIVE

The extent of resection has been shown to improve outcomes in patients with meningiomas. However, resection can be complicated by constraining local anatomy, leading to subtotal resections. An understanding of the natural history of residual tumors is necessary to better guide postsurgical management and minimize recurrence. This study seeks to identify predictors of recurrence and high growth rate following subtotal resection of intracranial meningiomas.

METHODS

Adult patients who underwent primary surgical resection of a WHO grade I meningioma at a tertiary care institution from 2007-2017 were retrospectively reviewed. Volumetric tumor measurements were made on patients with subtotal resections. Stepwise multivariate proportional hazards regression analyses were performed to identify factors associated with time to recurrence, as well as stepwise multivariate regression analyses to assess for factors associated with high postoperative growth rate.

RESULTS

Of the 141 patients (18%) who underwent radiographic subtotal resection of an intracranial meningioma during the reviewed period, 74 (52%) suffered a recurrence, in which the median (interquartile range, IQR) time to recurrence was 14 (IQR 6-34) months. Among those tumors subtotally resected, the median pre- and postoperative tumor volumes were 17.19 cm3 (IQR 7.47-38.43 cm3) and 2.31 cm3 (IQR 0.98-5.16 cm3), which corresponded to a percentage resection of 82% (IQR 68%-93%). Postoperatively, the median growth rate was 0.09 cm3/year (IQR 0-1.39 cm3/year). Factors associated with recurrence in multivariate analysis included preoperative tumor volume (hazard ratio [HR] 1.008,95% confidence interval [CI] 1.002-1.013, p = 0.008), falcine location (HR 2.215, 95% CI 1.179-4.161, p = 0.021), tentorial location (HR 2.410, 95% CI 1.203-4.829, p = 0.024), and African American race (HR 1.811, 95% CI 1.042-3.146, p = 0.044). Residual volume (RV) was associated with high absolute annual growth rate (odds ratio [OR] 1.175, 95% CI 1.078-1.280, p < 0.0001), with the maximum RV benefit at < 5 cm3 (OR 4.056, 95% CI 1.675-9.822, p = 0.002).

CONCLUSIONS

By identifying predictors of recurrence and growth rate, this study helps identify potential patients with a high chance of recurrence following subtotal resection, which are those with large preoperative tumor volume, falcine location, tentorial location, and African American race. Higher RVs were associated with tumors with higher postoperative growth rates. Recurrences typically occurred 14 months after surgery.

Theragnostic Use of Radiolabelled Dota-Peptides in Meningioma: From Clinical Demand to Future Applications

Meningiomas account for approximately 30% of all new diagnoses of intracranial masses. The 2016 World Health Organization's (WHO) classification currently represents the clinical standard for meningioma's grading and prognostic stratification. However, watchful waiting is frequently the chosen treatment option, although this means the absence of a certain histological diagnosis. Consequently, MRI (or less frequently CT) brain imaging currently represents the unique available tool to define diagnosis, grading, and treatment planning in many cases. Nonetheless, these neuroimaging modalities show some limitations, particularly in the evaluation of skull base lesions. The emerging evidence supporting the use of radiolabelled somatostatin receptor analogues (such as dota-peptides) to provide molecular imaging of meningiomas might at least partially overcome these limitations. Moreover, their potential therapeutic usage might enrich the current clinical offering for these patients. Starting from the strengths and weaknesses of structural and functional neuroimaging in meningiomas, in the present article we systematically reviewed the published studies regarding the use of radiolabelled dota-peptides in surgery and radiotherapy planning, in the restaging of treated patients, as well as in peptide-receptor radionuclide therapy of meningioma.

Accuracy of Radiomics-Based Feature Analysis on Multiparametric Magnetic Resonance Images for Noninvasive Meningioma Grading

OBJECTIVE

Meningioma grading is relevant to therapy decisions in complete or partial resection, observation, and radiotherapy because higher grades are associated with tumor growth and recurrence. The differentiation of low and intermediate grades is particularly challenging. This study attempts to apply radiomics-based shape and texture analysis on routine multiparametric magnetic resonance imaging (MRI) from different scanners and institutions for grading.

METHODS

We used MRI data (T1-weighted/T2-weighted, T1-weighted-contrast-enhanced [T1CE], fluid-attenuated inversion recovery [FLAIR], diffusion-weighted imaging [DWI], apparent diffusion coefficient [ADC]) of grade I (n = 46) and grade II (n = 25) nontreated meningiomas with histologic workup. Two experienced radiologists performed manual tumor segmentations on FLAIR, T1CE, and ADC images in consensus. The MRI data were preprocessed through T1CE and T1-subtraction, coregistration, resampling, and normalization. A PyRadiomics package was used to generate 990 shape/texture features. Stepwise dimension reduction and robust radiomics feature selection were performed. Biopsy results were used as standard of reference.

RESULTS

Four statistically independent radiomics features were identified as showing the strongest predictive values for higher tumor grades: roundness-of-FLAIR-shape (area under curve [AUC], 0.80), cluster-shades-of-FLAIR/T1CE-gray-level (AUC, 0.80), DWI/ADC-gray-level-variability (AUC, 0.72), and FLAIR/T1CE-gray-level-energy (AUC, 0.76). In a multivariate logistic regression model, the combination of the features led to an AUC of 0.91 for the differentiation of grade I and grade II meningiomas.

CONCLUSIONS

Our results indicate that radiomics-based feature analysis applied on routine MRI is viable for meningioma grading, and a multivariate logistic regression model yielded strong classification performances. More advanced tumor stages are identifiable through certain shape parameters of the lesion, textural patterns in morphologic MRI sequences, and DWI/ADC variability.

A prospective study of the natural history of incidental meningioma-Hold your horses!

Background

The number of incidental meningiomas has increased because of the increased availability of neuroimaging. Lack of prospective data on the natural history makes the optimal management unclear. We conducted a 5-year prospective study of incidental meningiomas to identify risk factors for tumor growth.

Methods

Sixty-four of 70 consecutive patients with incidental meningioma were included. Clinical and radiological status was obtained at 0, 0.5, 1, 1.5, 2, 3, 4, and 5 years. GammaPlan and mixed linear regression modeling were utilized for volumetric analysis with primary endpoint tumor growth.

Results

None of the patients developed tumor-related symptoms during the study period, although 48 (75%) tumors increased (>15%), 13 (20.3%) remained unchanged, and 3 (4.7%) decreased (>15%) in volume. Mean time to growth was 2.2 years (range, 0.5-5.0 years).The growth pattern was quasi-exponential in 26%, linear in 17%, sigmoidal in 35%, parabolic in 17%, and continuous reduction in 5%. There was significant correlation among growth rate, larger baseline tumor volume (P < .001), and age in years (<55 y: 0.10 cm³/y, 55-75 y: 0.24 cm³/y, and >75 y: 0.85 cm³/y).

Conclusion

The majority of meningiomas will eventually grow. However, more than 60% display a self-limiting growth pattern. Our study provides level-2 evidence that asymptomatic tumors can be safely managed utilizing serial imaging until persistent radiological and/or symptomatic growth.

Incidental intracranial meningiomas: a systematic review and meta-analysis of prognostic factors and outcomes

Background

Incidental discovery accounts for 30% of newly-diagnosed intracranial meningiomas. There is no consensus on their optimal management. This review aimed to evaluate the outcomes of different management strategies for these tumors.

Methods

Using established systematic review methods, six databases were scanned up to September 2017. Pooled event proportions were estimated using a random effects model. Meta-regression of prognostic factors was performed using individual patient data.

Results

Twenty studies (2130 patients) were included. Initial management strategies at diagnosis were: surgery (27.3%), stereotactic radiosurgery (22.0%) and active monitoring (50.7%) with a weighted mean follow-up of 49.5 months (SD = 29.3). The definition of meningioma growth and monitoring regimens varied widely impeding relevant meta-analysis. The pooled risk of symptom development in patients actively monitored was 8.1% (95% CI 2.7–16.1). Associated factors were peritumoral edema (OR 8.72 [95% CI 0.35–14.90]) and meningioma diameter ≥ 3 cm (OR 34.90 [95% CI 5.17–160.40]). The pooled proportion of intervention after a duration of active monitoring was 24.8% (95% CI 7.5–48.0). Weighted mean time-to-intervention was 24.8 months (SD = 18.2). The pooled risks of morbidity following surgery and radiosurgery, accounting for cross-over, were 11.8% (95% CI 3.7–23.5) and 32.0% (95% CI 10.6–70.5) respectively. The pooled proportion of operated meningioma being WHO grade I was 94.0% (95% CI 88.2–97.9).

Conclusion

The management of incidental meningioma varies widely. Most patients who clinically or radiologically progressed did so within 5 years of diagnosis. Intervention at diagnosis may lead to unnecessary overtreatment. Prospective data is needed to develop a risk calculator to better inform management strategies.

Electronic supplementary material

The online version of this article (10.1007/s11060-019-03104-3) contains supplementary material, which is available to authorized users.

An overview of meningiomas

Meningiomas are the most common primary intracranial tumor. Important advances are occurring in meningioma research. These are expected to accelerate, potentially leading to impactful changes on the management of meningiomas in the near and medium term. This review will cover the histo- and molecular pathology of meningiomas, including recent 2016 updates to the WHO classification of CNS tumors. We will discuss clinical and radiographic presentation and therapeutic management. Surgery and radiotherapy, the two longstanding primary therapeutic modalities, will be discussed at length. In addition, data from prior and ongoing investigations of other treatment modalities, including systemic and targeted therapies, will be covered. This review will quickly update the reader on the contemporary management and future directions in meningiomas. [Formula: see text].

Volumetric tumor growth rates of meningiomas involving the intracranial venous sinuses

OBJECT

There is currently no consensus as to whether meningiomas located inside the venous sinuses should be aggressively or conservatively treated. The goals of this study were to identify how sinus-invading meningiomas grow, report and compare growth rates of tumor components inside and outside the different venous sinuses, identify risk factors associated with increased tumor growth, and determine the effects of the extent of tumor resection on recurrence for meningiomas that invade the dural venous sinuses.

METHODS

Adult patients who underwent primary, non-biopsy resection of a WHO grade 1 meningioma invading the dural venous sinuses at a tertiary care institution between 2007 and 2015 were retrospectively reviewed. Rates of tumor growth were fit to several growth models to evaluate the most accurate model. Cohen's d analysis was used to identify associations with increased growth of tumor in the venous sinuses. Logistic regression was used to compare extent of resection with recurrence.

RESULTS

Of the 68 patients included in the study, 34 patients had postoperative residual tumors in the venous sinuses that were measured over time. The growth model that best fit the growth of intrasinus meningiomas was the Gompertzian growth model (r² = 0.93). The annual growth rate of meningiomas inside the sinuses was 7.3%, compared to extrasinus tumors with 13.6% growth per year. The only factor significantly associated with increased tumor growth in sinuses was preoperative embolization (effect sizes (ES) [95% CI], 1.874 [7.633-46.735], p = 0.008).

CONCLUSIONS

This study shows that meningiomas involving the venous sinuses have a Gompertzian-type growth with early exponential growth followed by a slower growth rate that plateaus when they reach a certain size. Overall, the growth rate of the intrasinus portion is low (7.3%), which is half of the reported growth rates for other studies involving primarily extrasinus tumors.

Fully automated detection and segmentation of meningiomas using deep learning on routine multiparametric MRI

Objectives

Magnetic resonance imaging (MRI) is the method of choice for imaging meningiomas. Volumetric assessment of meningiomas is highly relevant for therapy planning and monitoring. We used a multiparametric deep-learning model (DLM) on routine MRI data including images from diverse referring institutions to investigate DLM performance in automated detection and segmentation of meningiomas in comparison to manual segmentations.

Methods

We included 56 of 136 consecutive preoperative MRI datasets [T1/T2-weighted, T1-weighted contrast-enhanced (T1CE), FLAIR] of meningiomas that were treated surgically at the University Hospital Cologne and graded histologically as tumour grade I (n = 38) or grade II (n = 18). The DLM was trained on an independent dataset of 249 glioma cases and segmented different tumour classes as defined in the brain tumour image segmentation benchmark (BRATS benchmark). The DLM was based on the DeepMedic architecture. Results were compared to manual segmentations by two radiologists in a consensus reading in FLAIR and T1CE.

Results

The DLM detected meningiomas in 55 of 56 cases. Further, automated segmentations correlated strongly with manual segmentations: average Dice coefficients were 0.81 ± 0.10 (range, 0.46-0.93) for the total tumour volume (union of tumour volume in FLAIR and T1CE) and 0.78 ± 0.19 (range, 0.27-0.95) for contrast-enhancing tumour volume in T1CE.

Conclusions

The DLM yielded accurate automated detection and segmentation of meningioma tissue despite diverse scanner data and thereby may improve and facilitate therapy planning as well as monitoring of this highly frequent tumour entity.

Key Points

• Deep learning allows for accurate meningioma detection and segmentation

• Deep learning helps clinicians to assess patients with meningiomas

• Meningioma monitoring and treatment planning can be improved

Electronic supplementary material

The online version of this article (10.1007/s00330-018-5595-8) contains supplementary material, which is available to authorized users.

Correlation of volumetric growth and histological grade in 50 meningiomas

INTRODUCTION

Advances in radiological imaging techniques have enabled volumetric measurements of meningiomas to be easily monitored using serial imaging scans. There is limited literature on the relationship between tumour growth rates and the WHO classification of meningiomas despite tumour growth being a major determinant of type and timing of intervention. Volumetric growth has been successfully used to assess growth of low-grade glioma; however, there is limited information on the volumetric growth rate (VGR) of meningiomas. This study aimed to determine the reliability of VGR measurement in patients with meningioma, assess the relationship between VGR and 2016 WHO grading as well as clinical applicability of VGR in monitoring meningioma growth.

METHODS

All histologically proven intracranial meningiomas that underwent resection in a single centre between April 2009 and April 2014 were reviewed and classified according to the 2016 edition of the Classification of the Tumours of the CNS. Only patients who had two pre-operative scans that were at least 3 months apart were included in the study. Two authors performed the volumetric measurements using the Slicer 3D software independently and the inter-rater reliability was assessed. Multiple regression analyses of factors affecting the VGR and VDE of meningiomas were performed using the R statistical software with p < 0.05 considered to be statistically significant.

RESULTS

Of 548 patients who underwent resection of their meningiomas, 66 met the inclusion criteria. Sixteen cases met the exclusion criteria (NF2, spinal location, previous surgical or radiation treatment, significant intra-osseous component and poor quality imaging). Forty-two grade I and 8 grade II meningiomas were included in the analysis. The VGR was significantly higher for grade II meningiomas. Using receiver-operator characteristic (ROC) curve analysis, the optimal threshold that distinguishes between grade I and II meningiomas is 3 cm³/year. Higher histological grade, high initial tumour volume, MRI T2-signal hyperintensity and presence of oedema were found to be significant predictors of higher VGR.

CONCLUSION

Reliable tools now exist to evaluate and monitor volumetric growth of meningiomas. Grade II meningiomas have significantly higher VGR compared with grade I meningiomas and growth of more than 3 cm³/year is strongly suggestive of a higher grade meningioma. A larger, multi-centre prospective study to investigate the applicability of velocity of growth to predict the outcome of patients with meningioma is warranted.