Neuroradiological assessment of newly diagnosed glioblastoma

Tumour growth rate predicts overall survival in patients with recurrent WHO grade 4 glioma

PURPOSE

Accurate prognostication may aid in the selection of patients who will benefit from surgery at recurrent WHO grade 4 glioma. This study aimed to evaluate the role of serial tumour volumetric measurements for prognostication at first tumour recurrence.

METHODS

We retrospectively analyzed patients with histologically-diagnosed WHO grade 4 glioma at initial and at first tumour recurrence at a tertiary hospital between May 2000 and September 2018. We performed auto-segmentation using ITK-SNAP software, followed by manual adjustment to measure serial contrast-enhanced T1W (CE-T1W) and T2W lesional volume changes on all MRI images performed between initial resection and repeat surgery.

RESULTS

Thirty patients met inclusion criteria; the median overall survival using Kaplan-Meier analysis from second surgery was 10.5 months. Seventeen (56.7%) patients received treatment post second surgery. Univariate cox regression analysis showed that greater rate of increase in lesional volume on CE-T1W (HR = 2.57; 95% CI [1.18, 5.57]; p = 0.02) in the last 2 MRI scans leading up to the second surgery was associated with a higher mortality likelihood. Patients with higher Karnofsky Performance Score (KPS) (HR = 0.97; 95% CI [0.95, 0.99]; p = 0.01) and who received further treatment following second surgery (HR = 0.43; 95% CI [0.19, 0.98]; p = 0.04) were shown to have a better survival.

CONCLUSION

Higher rate of CE-T1W lesional growth on the last 2 MRI images prior to surgery at recurrence was associated with increase mortality risk. A larger prospective study is required to determine and validate the threshold to distinguish rapidly progressive tumour with poor prognosis.

Differentiation of Glioblastoma and Brain Metastases by MRI-Based Oxygen Metabolomic Radiomics and Deep Learning

Glioblastoma (GB) and brain metastasis (BM) are the most frequent types of brain tumors in adults. Their therapeutic management is quite different and a quick and reliable initial characterization has a significant impact on clinical outcomes. However, the differentiation of GB and BM remains a major challenge in today's clinical neurooncology due to their very similar appearance in conventional magnetic resonance imaging (MRI). Novel metabolic neuroimaging has proven useful for improving diagnostic performance but requires artificial intelligence for implementation in clinical routines. Here; we investigated whether the combination of radiomic features from MR-based oxygen metabolism ("oxygen metabolic radiomics") and deep convolutional neural networks (CNNs) can support reliably pre-therapeutic differentiation of GB and BM in a clinical setting. A self-developed one-dimensional CNN combined with radiomic features from the cerebral metabolic rate of oxygen (CMRO₂) was clearly superior to human reading in all parameters for classification performance. The radiomic features for tissue oxygen saturation (mitoPO₂; i.e., tissue hypoxia) also showed better diagnostic performance compared to the radiologists. Interestingly, both the mean and median values for quantitative CMRO₂ and mitoPO₂ values did not differ significantly between GB and BM. This demonstrates that the combination of radiomic features and DL algorithms is more efficient for class differentiation than the comparison of mean or median values. Oxygen metabolic radiomics and deep neural networks provide insights into brain tumor phenotype that may have important diagnostic implications and helpful in clinical routine diagnosis.

Radiophysiomics: Brain Tumors Classification by Machine Learning and Physiological MRI Data

The precise initial characterization of contrast-enhancing brain tumors has significant consequences for clinical outcomes. Various novel neuroimaging methods have been developed to increase the specificity of conventional magnetic resonance imaging (cMRI) but also the increased complexity of data analysis. Artificial intelligence offers new options to manage this challenge in clinical settings. Here, we investigated whether multiclass machine learning (ML) algorithms applied to a high-dimensional panel of radiomic features from advanced MRI (advMRI) and physiological MRI (phyMRI; thus, radiophysiomics) could reliably classify contrast-enhancing brain tumors. The recently developed phyMRI technique enables the quantitative assessment of microvascular architecture, neovascularization, oxygen metabolism, and tissue hypoxia. A training cohort of 167 patients suffering from one of the five most common brain tumor entities (glioblastoma, anaplastic glioma, meningioma, primary CNS lymphoma, or brain metastasis), combined with nine common ML algorithms, was used to develop overall 135 classifiers. Multiclass classification performance was investigated using tenfold cross-validation and an independent test cohort. Adaptive boosting and random forest in combination with advMRI and phyMRI data were superior to human reading in accuracy (0.875 vs. 0.850), precision (0.862 vs. 0.798), F-score (0.774 vs. 0.740), AUROC (0.886 vs. 0.813), and classification error (5 vs. 6). The radiologists, however, showed a higher sensitivity (0.767 vs. 0.750) and specificity (0.925 vs. 0.902). We demonstrated that ML-based radiophysiomics could be helpful in the clinical routine diagnosis of contrast-enhancing brain tumors; however, a high expenditure of time and work for data preprocessing requires the inclusion of deep neural networks.

Metabolic Tumor Microenvironment Characterization of Contrast Enhancing Brain Tumors Using Physiologic MRI

The tumor microenvironment is a critical regulator of cancer development and progression as well as treatment response and resistance in brain neoplasms. The available techniques for investigation, however, are not well suited for noninvasive in vivo characterization in humans. A total of 120 patients (59 females; 61 males) with newly diagnosed contrast-enhancing brain tumors (64 glioblastoma, 20 brain metastases, 15 primary central nervous system (CNS) lymphomas (PCNSLs), and 21 meningiomas) were examined with a previously established physiological MRI protocol including quantitative blood-oxygen-level-dependent imaging and vascular architecture mapping. Six MRI biomarker maps for oxygen metabolism and neovascularization were fused for classification of five different tumor microenvironments: glycolysis, oxidative phosphorylation (OxPhos), hypoxia with/without neovascularization, and necrosis. Glioblastoma showed the highest metabolic heterogeneity followed by brain metastasis with a glycolysis-to-OxPhos ratio of approximately 2:1 in both tumor entities. In addition, glioblastoma revealed a significant higher percentage of hypoxia (24%) compared to all three other brain tumor entities: brain metastasis (7%; p < 0.001), PCNSL (8%; p = 0.001), and meningioma (8%; p = 0.003). A more aggressive biological brain tumor behavior was associated with a higher percentage of hypoxia and necrosis and a lower percentage of remaining vital tumor tissue and aerobic glycolysis. The proportion of oxidative phosphorylation, however, was rather similar (17–26%) for all four brain tumor entities. Tumor microenvironment (TME) mapping provides insights into neurobiological differences of contrast-enhancing brain tumors and deserves further clinical cancer research attention. Although there is a long roadmap ahead, TME mapping may become useful in order to develop new diagnostic and therapeutic approaches.

Differentiating Glioblastomas from Solitary Brain Metastases: An Update on the Current Literature of Advanced Imaging Modalities

Differentiating between glioblastomas and solitary brain metastases proves to be a challenging diagnosis for neuroradiologists, as both present with imaging patterns consisting of peritumoral hyperintensities with similar intratumoral texture on traditional magnetic resonance imaging sequences. Early diagnosis is paramount, as each pathology has completely different methods of clinical assessment. In the past decade, recent developments in advanced imaging modalities enabled providers to acquire a more accurate diagnosis earlier in the patient's clinical assessment, thus optimizing clinical outcome. Dynamic susceptibility contrast has been optimized for detecting relative cerebral blood flow and relative cerebral blood volume. Diffusion tensor imaging can be used to detect changes in mean diffusivity. Neurite orientation dispersion and density imaging is an innovative modality detecting changes in intracellular volume fraction, isotropic volume fraction, and extracellular volume fraction. Magnetic resonance spectroscopy is able to assist by providing a metabolic descriptor while detecting variable ratios of choline/N-acetylaspartate, choline/creatine, and N-acetylaspartate/creatine. Finally, radiomics and machine learning algorithms have been devised to assist in improving diagnostic accuracy while often utilizing more than one advanced imaging protocol per patient. In this review, we provide an update on all the current evidence regarding the identification and differentiation of glioblastomas from solitary brain metastases.

Surveillance for subsequent neoplasms of the CNS for childhood, adolescent, and young adult cancer survivors: a systematic review and recommendations from the International Late Effects of Childhood Cancer Guideline Harmonization Group

Exposure to cranial radiotherapy is associated with an increased risk of subsequent CNS neoplasms among childhood, adolescent, and young adult (CAYA) cancer survivors. Surveillance for subsequent neoplasms can translate into early diagnoses and interventions that could improve cancer survivors' health and quality of life. The practice guideline presented here by the International Late Effects of Childhood Cancer Guideline Harmonization Group was developed with an evidence-based method that entailed the gathering and appraisal of published evidence associated with subsequent CNS neoplasms among CAYA cancer survivors. The preparation of these guidelines showed a paucity of high-quality evidence and highlighted the need for additional research to inform survivorship care. The recommendations are based on careful consideration of the evidence supporting the benefits, risks, and harms of the surveillance interventions, clinical judgment regarding individual patient circumstances, and the need to maintain flexibility of application across different health-care systems. Currently, there is insufficient evidence to establish whether early detection of subsequent CNS neoplasms reduces morbidity and mortality, and therefore no recommendation can be formulated for or against routine MRI surveillance. The decision to start surveillance should be made by the CAYA cancer survivor and health-care provider after careful consideration of the potential harms and benefits of surveillance for CNS neoplasms, including meningioma.

Predictive diagnostic and/or prognostic biomarkers obtained from routine blood biochemistry in patients with solitary intracranial tumor

Background

Radiological and/or laboratory tests may be sometimes inadequate distinguishing glioblastoma from metastatic brain tumors. The aim of this study was to find possible predictive biomarkers produced from routine blood biochemistry analysis results evaluated preoperatively in each patient with solitary brain tumor in distinguishing glioblastoma from metastatic brain tumors as well as revealing short-term prognosis.

Methods

Patients admitted to neurosurgery clinic between January 2015 and September 2018 were included in this study and they were divided into GLIOMA (n=12) and METASTASIS (n=17) groups. Patients' data consisted of age, gender, Glasgow Coma Scale scores, duration of stay in hospital, Glasgow Outcome Scale (GOS) scores and histopathological examination reports, hemoglobin level, leukocyte, neutrophil, lymphocyte, monocyte, eosinophil, basophil and platelet count results, neutrophil-lymphocyte ratio and platelet-lymphocyte ratio values, C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) levels were evaluated preoperatively.

Results

The CRP levels of METASTASIS group (143.10 mg/L) were higher than those of GLIOMA group (23.90 mg/L); and it was 82% sensitive and 75% specific in distinguishing metastatic brain tumor from glioblastoma if CRP value was >55.00 mg/L. A positive correlation was determined between GOS score and hemoglobin level and between ESR and CRP values. However, GOS scores were negatively correlated with the ESR level and duration of stay in hospital.

Conclusions

Study results demonstrated that CRP values could be predictive biomarker in distinguishing metastatic brain tumor from glioblastoma. In addition, ESR, CRP, hemoglobin levels and duration of stay in hospital could be prognostic biomarkers in predicting short-term prognosis of patients with solitary brain tumor.

A Critical Appraisal of the Quality of Glioma Imaging Guidelines Using the AGREE II Tool: A EuroAIM Initiative

Background: Following the EuroAIM initiative to assess the quality of medical imaging guidelines by using the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument, we aimed to evaluate the quality of the current imaging guidelines in patients with gliomas.

Methods: A literature search was conducted to identify eligible imaging guidelines considered in the management of adult patients with gliomas. The selected guidelines were evaluated using the AGREE II instrument by four independent appraisers. The agreement among the four appraisers was estimated using the intraclass correlation coefficient (ICC) analysis.

Results: Seven guidelines were selected for the appraisal. Six out of the seven guidelines showed an average level of quality with only one showing a low quality. The highest scores were found in Domain 1 “Scope and purpose” (mean score = 81.2%) and Domain 4 “Clarity of presentation” (mean score = 77.6%). The remaining domains showed a low level of quality and, in particular, Domain 5 “Applicability” was the most critical with a mean score of 41.7%, mainly related to a minor attention to barriers and facilitators as well as costs and resources implications of applying the guidelines. The ICC analysis showed a very good agreement among the four appraisers with ICC values ranging from 0.907 to 0.993.

Conclusions: The available guidelines on glioma imaging emerged as of average quality according to the AGREE II tool analysis. Based on these results, further efforts should be made in order to involve different professional bodies and stakeholders and increase patient and public involvement in any future guideline drafting as well as to improve the applicability of these guidelines into the clinical practice.

Machine learning for semi-automated classification of glioblastoma, brain metastasis and central nervous system lymphoma using magnetic resonance advanced imaging

Background

Differentiating glioblastoma, brain metastasis, and central nervous system lymphoma (CNSL) on conventional magnetic resonance imaging (MRI) can present a diagnostic dilemma due to the potential for overlapping imaging features. We investigate whether machine learning evaluation of multimodal MRI can reliably differentiate these entities.

Methods

Preoperative brain MRI including diffusion weighted imaging (DWI), dynamic contrast enhanced (DCE), and dynamic susceptibility contrast (DSC) perfusion in patients with glioblastoma, lymphoma, or metastasis were retrospectively reviewed. Perfusion maps (rCBV, rCBF), permeability maps (K-trans, Kep, Vp, Ve), ADC, T1C+ and T2/FLAIR images were coregistered and two separate volumes of interest (VOIs) were obtained from the enhancing tumor and non-enhancing T2 hyperintense (NET2) regions. The tumor volumes obtained from these VOIs were utilized for supervised training of support vector classifier (SVC) and multilayer perceptron (MLP) models. Validation of the trained models was performed on unlabeled cases using the leave-one-subject-out method. Head-to-head and multiclass models were created. Accuracies of the multiclass models were compared against two human interpreters reviewing conventional and diffusion-weighted MR images.

Results

Twenty-six patients enrolled with histopathologically-proven glioblastoma (n=9), metastasis (n=9), and CNS lymphoma (n=8) were included. The trained multiclass ML models discriminated the three pathologic classes with a maximum accuracy of 69.2% accuracy (18 out of 26; kappa 0.540, P=0.01) using an MLP trained with the VpNET2 tumor volumes. Human readers achieved 65.4% (17 out of 26) and 80.8% (21 out of 26) accuracies, respectively. Using the MLP VpNET2 model as a computer-aided diagnosis (CADx) for cases in which the human reviewers disagreed with each other on the diagnosis resulted in correct diagnoses in 5 (19.2%) additional cases.

Conclusions

Our trained multiclass MLP using VpNET2 can differentiate glioblastoma, brain metastasis, and CNS lymphoma with modest diagnostic accuracy and provides approximately 19% increase in diagnostic yield when added to routine human interpretation.

Connection between Cell Phone use, p53 Gene Expression in Different Zones of Glioblastoma Multiforme and Survival Prognoses

The aim of this paper is to investigate p53 gene expression in the central and peripheral zones of glioblastoma multiforme using a real-time reverse transcription polymerase chain reaction (RT-PCR) technique in patients who use cell phones ≥3 hours a day and determine its relationship to clinicopathological findings and overall survival. Sixty-three patients (38 males and 25 females), diagnosed with glioblastoma multiforme (GBM), underwent tumor resection between 2008 and 2011. Patient ages ranged from 25 to 88 years, with a mean age of 55. The levels of expression of p53 in the central and peripheral zone of the GBM were quantified by RT-PCR. Data on p53 gene expression from the central and peripheral zone, the related malignancy and the clinicopatholagical findings (age, gender, tumor location and size), as well as overall survival, were analyzed. Forty-one out of 63 patients (65%) with the highest level of cell phone use (≥3 hours/day) had higher mutant type p53 expression in the peripheral zone of the glioblastoma; the difference was statistically significant (P=0.034). Results from the present study on the use of mobile phones for ≥3 hours a day show a consistent pattern of increased risk for the mutant type of p53 gene expression in the peripheral zone of the glioblastoma, and that this increase was significantly correlated with shorter overall survival time. The risk was not higher for ipsilateral exposure. We found that the mutant type of p53 gene expression in the peripheral zone of the glioblastoma was increased in 65% of patients using cell phones ≥3 hours a day.

The role of imaging in the management of progressive glioblastoma : a systematic review and evidence-based clinical practice guideline

QUESTION

Which imaging techniques most accurately differentiate true tumor progression from pseudo-progression or treatment related changes in patients with previously diagnosed glioblastoma?

TARGET POPULATION

These recommendations apply to adults with previously diagnosed glioblastoma who are suspected of experiencing progression of the neoplastic process.

RECOMMENDATIONS LEVEL II

Magnetic resonance imaging with and without gadolinium enhancement is recommended as the imaging surveillance method to detect the progression of previously diagnosed glioblastoma.

LEVEL II

Magnetic resonance spectroscopy is recommended as a diagnostic method to differentiate true tumor progression from treatment-related imaging changes or pseudo-progression in patients with suspected progressive glioblastoma.

LEVEL III

The routine use of positron emission tomography to identify progression of glioblastoma is not recommended.

LEVEL III

Single-photon emission computed tomography imaging is recommended as a diagnostic method to differentiate true tumor progression from treatment-related imaging changes or pseudo-progression in patients with suspected progressive glioblastoma.

In silico analysis of nuclei in glioblastoma using large-scale microscopy images improves prediction of treatment response

In this paper, we present a complete and novel workflow for quantitative nuclear feature analysis of glioblastoma using high-throughput whole-slide microscopy image processing as it relates to treatment response and patient survival. With a complete suite of computer algorithms, large numbers of micro-anatomical structures, in this case nuclei, are analyzed and represented efficiently from whole-slide digitized images with numerical features. With regard to endpoints of treatment response, the computerized analysis presents a better discrimination than traditional neuropathologic review. As a result, this analysis method shows potential to facilitate a better understanding of disease progression and patients' response to therapy for glioblastoma.

Diagnostic examination performance by using microvascular leakage, cerebral blood volume, and blood flow derived from 3-T dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging in the differentiation of glioblastoma multiforme and brain metastasis

INTRODUCTION

Conventional magnetic resonance (MR) imaging has limited capacity to differentiate between glioblastoma multiforme (GBM) and metastasis. The purposes of this study were: (1) to compare microvascular leakage (MVL), cerebral blood volume (CBV), and blood flow (CBF) in the distinction of metastasis from GBM using dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging (DSC-MRI), and (2) to estimate the diagnostic accuracy of perfusion and permeability MR imaging.

METHODS

A prospective study of 61 patients (40 GBMs and 21 metastases) was performed at 3 T using DSC-MRI. Normalized rCBV and rCBF from tumoral (rCBVt, rCBFt), peri-enhancing region (rCBVe, rCBFe), and by dividing the value in the tumor by the value in the peri-enhancing region (rCBVt/e, rCBFt/e), as well as MVL were calculated. Hemodynamic and histopathologic variables were analyzed statistically and Spearman/Pearson correlations. Receiver operating characteristic curve analysis was performed for each of the variables.

RESULTS

The rCBVe, rCBFe, and MVL were significantly greater in GBMs compared with those of metastases. The optimal cutoff value for differentiating GBM from metastasis was 0.80 which implies a sensitivity of 95%, a specificity of 92%, a positive predictive value of 86%, and a negative predictive value of 97% for rCBVe ratio. We found a modest correlation between rCBVt and rCBFt ratios.

CONCLUSION

MVL measurements in GBMs are significantly higher than those in metastases. Statistically, both rCBVe, rCBVt/e and rCBFe, rCBFt/e were useful in differentiating between GBMs and metastases, supporting the hypothesis that perfusion MR imaging can detect infiltration of tumor cells in the peri-enhancing region.