Machine learning and glioma imaging biomarkers

Machine learning-based radiomic evaluation of treatment response prediction in glioblastoma

AIM

To investigate machine learning based models combining clinical, radiomic, and molecular information to distinguish between early true progression (tPD) and pseudoprogression (psPD) in patients with glioblastoma.

MATERIALS AND METHODS

A retrospective analysis was undertaken of 76 patients (46 tPD, 30 psPD) with early enhancing disease following chemoradiotherapy for glioblastoma. Outcome was determined on follow-up until 6 months post-chemoradiotherapy. Models comprised clinical characteristics, O⁶-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, and 307 quantitative imaging features extracted from enhancing disease and perilesional oedema masks on early post-chemoradiotherapy contrast-enhanced T1-weighted imaging, T2-weighted imaging (T2WI), and apparent diffusion coefficient (ADC) maps. Feature selection was performed within bootstrapped cross-validated recursive feature elimination with a random forest algorithm. Naive Bayes five-fold cross-validation was used to validate the final model.

RESULTS

Top selected features included age, MGMT promoter methylation status, two shape-based features from the enhancing disease mask, three radiomic features from the enhancing disease mask on ADC, and one radiomic feature from the perilesional oedema mask on T2WI. The final model had an area under the receiver operating characteristics curve (AUC) of 0.80, sensitivity 78.2%, specificity 66.7%, and accuracy of 73.7%.

CONCLUSION

Incorporating a machine learning-based approach using quantitative radiomic features from standard-of-care magnetic resonance imaging (MRI), in combination with clinical characteristics and MGMT promoter methylation status has a complementary effect and improves model performance for early prediction of glioblastoma treatment response.

Challenges and opportunities for artificial intelligence in oncological imaging

Imaging plays a key role in oncology, including the diagnosis and detection of cancer, determining clinical management, assessing treatment response, and complications of treatment or disease. The current use of clinical oncology is predominantly qualitative in nature with some relatively crude size-based measurements of tumours for assessment of disease progression or treatment response; however, it is increasingly understood that there may be significantly more information about oncological disease that can be obtained from imaging that is not currently utilized. Artificial intelligence (AI) has the potential to harness quantitative techniques to improve oncological imaging. These may include improving the efficiency or accuracy of traditional roles of imaging such as diagnosis or detection. These may also include new roles for imaging such as risk-stratifying patients for different types of therapy or determining biological tumour subtypes. This review article outlines several major areas in oncological imaging where there may be opportunities for AI technology. These include (1) screening and detection of cancer, (2) diagnosis and risk stratification, (3) tumour segmentation, (4) precision oncology, and (5) predicting prognosis and assessing treatment response. This review will also address some of the potential barriers to AI research in oncological imaging.

Identification of Tumor-Specific MRI Biomarkers Using Machine Learning (ML)

The identification of reliable and non-invasive oncology biomarkers remains a main priority in healthcare. There are only a few biomarkers that have been approved as diagnostic for cancer. The most frequently used cancer biomarkers are derived from either biological materials or imaging data. Most cancer biomarkers suffer from a lack of high specificity. However, the latest advancements in machine learning (ML) and artificial intelligence (AI) have enabled the identification of highly predictive, disease-specific biomarkers. Such biomarkers can be used to diagnose cancer patients, to predict cancer prognosis, or even to predict treatment efficacy. Herein, we provide a summary of the current status of developing and applying Magnetic resonance imaging (MRI) biomarkers in cancer care. We focus on all aspects of MRI biomarkers, starting from MRI data collection, preprocessing and machine learning methods, and ending with summarizing the types of existing biomarkers and their clinical applications in different cancer types.

Unique challenges for glioblastoma immunotherapy-discussions across neuro-oncology and non-neuro-oncology experts in cancer immunology. Meeting Report from the 2019 SNO Immuno-Oncology Think Tank

Cancer immunotherapy has made remarkable advances with over 50 separate Food and Drug Administration (FDA) approvals as first- or second-line indications since 2015. These include immune checkpoint blocking antibodies, chimeric antigen receptor-transduced T cells, and bispecific T-cell-engaging antibodies. While multiple cancer types now benefit from these immunotherapies, notable exceptions thus far include brain tumors, such as glioblastoma. As such, it seems critical to gain a better understanding of unique mechanistic challenges underlying the resistance of malignant gliomas to immunotherapy, as well as to acquire insights into the development of future strategies. An Immuno-Oncology Think Tank Meeting was held during the 2019 Annual Society for Neuro-Oncology Scientific Conference. Discussants in the fields of neuro-oncology, neurosurgery, neuro-imaging, medical oncology, and cancer immunology participated in the meeting. Sessions focused on topics such as the tumor microenvironment, myeloid cells, T-cell dysfunction, cellular engineering, and translational aspects that are critical and unique challenges inherent with primary brain tumors. In this review, we summarize the discussions and the key messages from the meeting, which may potentially serve as a basis for advancing the field of immune neuro-oncology in a collaborative manner.

Correlation of Metabolic Profiles of Plasma and Cerebrospinal Fluid of High-Grade Glioma Patients

This work compares the metabolic profiles of plasma and the cerebrospinal fluid (CSF) of the patients with high-grade (III and IV) gliomas and the conditionally healthy controls using the wide-range targeted screening of low molecular metabolites by HPLC-MS/MS. The obtained data were analyzed using robust linear regression with Huber’s M-estimates, and a number of metabolites with correlated content in plasma and CSF was identified. The statistical analysis shows a significant correlation of metabolite content in plasma and CSF samples for the majority of metabolites. Several metabolites were shown to have high correlation in the control samples, but not in the glioma patients. This can be due to the specific metabolic processes in the glioma patients or to the damaged integrity of blood-brain barrier. The results of our study may be useful for the understanding of molecular mechanisms underlying the development of gliomas, as well as for the search of potential biomarkers for the minimally invasive diagnostic procedures of gliomas.

Artificial intelligence applied to neuroimaging data in Parkinsonian syndromes: Actuality and expectations

Idiopathic Parkinson's Disease (iPD) is a common motor neurodegenerative disorder. It affects more frequently the elderly population, causing a significant emotional burden both for the patient and caregivers, due to the disease-related onset of motor and cognitive disabilities. iPD's clinical hallmark is the onset of cardinal motor symptoms such as bradykinesia, rest tremor, rigidity, and postural instability. However, these symptoms appear when the neurodegenerative process is already in an advanced stage. Furthermore, the greatest challenge is to distinguish iPD from other similar neurodegenerative disorders, "atypical parkinsonisms", such as Multisystem Atrophy, Progressive Supranuclear Palsy and Cortical Basal Degeneration, since they share many phenotypic manifestations, especially in the early stages. The diagnosis of these neurodegenerative motor disorders is essentially clinical. Consequently, the diagnostic accuracy mainly depends on the professional knowledge and experience of the physician. Recent advances in artificial intelligence have made it possible to analyze the large amount of clinical and instrumental information in the medical field. The application machine learning algorithms to the analysis of neuroimaging data appear to be a promising tool for identifying microstructural alterations related to the pathological process in order to explain the onset of symptoms and the spread of the neurodegenerative process. In this context, the search for quantitative biomarkers capable of identifying parkinsonian patients in the prodromal phases of the disease, of correctly distinguishing them from atypical parkinsonisms and of predicting clinical evolution and response to therapy represent the main goal of most current clinical research studies. Our aim was to review the recent literature and describe the current knowledge about the contribution given by machine learning applications to research and clinical management of parkinsonian syndromes.

Development and Multi-Data Set Verification of an RNA Binding Protein Signature for Prognosis Prediction in Glioma

Objective: Increasing evidence emphasizes the clinical implications of RNA binding proteins (RBPs) in cancers. This study aimed to develop a RBP signature for predicting prognosis in glioma.

Methods: Two glioma datasets as training (n = 693) and validation (n = 325) sets were retrieved from the CGGA database. In the training set, univariate Cox regression analysis was conducted to screen prognosis-related RBPs based on differentially expressed RBPs between WHO grade II and IV. A ten-RBP signature was then established. The predictive efficacy was evaluated by ROCs. The applicability was verified in the validation set. The pathways involving the risk scores were analyzed by ssGSEA. scRNA-seq was utilized for evaluating their expression in different glioma cell types. Moreover, their expression was externally validated between glioma and control samples.

Results: Based on 39 prognosis-related RBPs, a ten RBP signature was constructed. High risk score distinctly indicated a poorer prognosis than low risk score. AUCs were separately 0.838 and 0.822 in the training and validation sets, suggesting its well performance for prognosis prediction. Following adjustment of other clinicopathological characteristics, the signature was an independent risk factor. Various cancer-related pathways were significantly activated in samples with high risk score. The scRNA-seq identified that risk RBPs were mainly expressed in glioma malignant cells. Their high expression was also found in glioma than control samples.

Conclusion: This study developed a novel RBP signature for robustly predicting prognosis of glioma following multi-data set verification. These RBPs may affect the progression of glioma.

A Position Statement on the Utility of Interval Imaging in Standard of Care Brain Tumour Management: Defining the Evidence Gap and Opportunities for Future Research

OBJECTIV E

To summarise current evidence for the utility of interval imaging in monitoring disease in adult brain tumours, and to develop a position for future evidence gathering while incorporating the application of data science and health economics.

METHODS

Experts in 'interval imaging' (imaging at pre-planned time-points to assess tumour status); data science; health economics, trial management of adult brain tumours, and patient representatives convened in London, UK. The current evidence on the use of interval imaging for monitoring brain tumours was reviewed. To improve the evidence that interval imaging has a role in disease management, we discussed specific themes of data science, health economics, statistical considerations, patient and carer perspectives, and multi-centre study design. Suggestions for future studies aimed at filling knowledge gaps were discussed.

RESULTS

Meningioma and glioma were identified as priorities for interval imaging utility analysis. The "monitoring biomarkers" most commonly used in adult brain tumour patients were standard structural MRI features. Interval imaging was commonly scheduled to provide reported imaging prior to planned, regular clinic visits. There is limited evidence relating interval imaging in the absence of clinical deterioration to management change that alters morbidity, mortality, quality of life, or resource use. Progression-free survival is confounded as an outcome measure when using structural MRI in glioma. Uncertainty from imaging causes distress for some patients and their caregivers, while for others it provides an important indicator of disease activity. Any study design that changes imaging regimens should consider the potential for influencing current or planned therapeutic trials, ensure that opportunity costs are measured, and capture indirect benefits and added value.

CONCLUSION

Evidence for the value, and therefore utility, of regular interval imaging is currently lacking. Ongoing collaborative efforts will improve trial design and generate the evidence to optimise monitoring imaging biomarkers in standard of care brain tumour management.

GLIMPSE: a glioblastoma prognostication model using ensemble learning-a surveillance, epidemiology, and end results study

Purpose

Glioblastoma is one of the most common and aggressive brain tumors in the world with a poor prognosis. A glioblastoma prognostication model has the potential to improve the cancer's standard of care. No other paper has looked at using ensemble learning with a population database to predict multiple binary glioblastoma survival outcomes.

Methods

We utilized ensemble learning to design, build, and test a prognostication system for glioblastoma for short-, intermediate- and long-term survival, based on various clinical features. We used the population database SEER which covers 17 different registries. The most important prognostic features were identified and used as a clinical feature set. The statistical feature set was determined using Random Forests. The accuracy, sensitivity, specificity, area under the receiver operating characteristic (AUROC), positive predictive value (PPV), and negative predictive value (NPV) were reported.

Results

Statistically-determined feature sets had the best performance. All the top models for short, intermediate, and long-term survival were random forests. With regards to short-term survival, top model had metrics AUROC = 0.937, accuracy = 86%, specificity = 88%, sensitivity = 85%, NPV = 85%, and PPV = 87%. For long-term survival, the top model had AUROC = 0.893, accuracy = 81%, specificity = 79%, sensitivity = 83%, NPV = 82%, and PPV = 79%. The top intermediate-term survival prediction had AUROC ≥ 0.780 and the other metrics were at least 70%.

Conclusions

Our ensemble models were high-performing and achieved AUROCs as high as 0.94, highlighting the importance of balancing, using ensemble techniques and statistical feature selection. Our models can potentially be used by clinicians after external validation.

Favorable role of IDH1/2 mutations aided with MGMT promoter gene methylation in the outcome of patients with malignant glioma

Aim

The implications of molecular biomarkers IDH1/2 mutations and MGMT gene promoter methylation were evaluated for prognostic outcome of glioma patients.

Materials & methods

Glioma cases were analyzed for IDH1/2 mutations and MGMT promoter methylation by DNA sequencing and methylation-specific PCR, respectively.

Results

Mutations found in IDH1/2 genes totaled 63.4% (N = 40) wherein IDH1 mutations were significantly associated with oligidendrioglioma (p = 0.005) and astrocytoma (p = 0.0002). IDH1 mutants presented more, 60.5% in MGMT promoter-methylated cases (p = 0.03). IDH1 mutant cases had better survival for glioblastoma and oligodendrioglioma (log-rank p = 0.01). Multivariate analysis confirmed better survival in MGMT methylation carriers (hazard ratio [HR]: 0.59; p = 0.031). Combination of both biomarkers showed better prognosis on temozolomide (p < 0.05).

Conclusion

IDH1/2 mutations proved independent prognostic factors in glioma and associated with MGMT methylation for better survival.

Machine Learning Model to Predict Pseudoprogression Versus Progression in Glioblastoma Using MRI: A Multi-Institutional Study (KROG 18-07)

Simple Summary

Even after the introduction of a standard regimen consisting of concurrent chemoradiotherapy and adjuvant temozolomide, patients with glioblastoma multiforme mostly experience disease progression. Clinicians often encounter a situation where they need to distinguish progressive disease from pseudoprogression after treatment. We tried to investigate the feasibility of machine learning algorithm to distinguish pseudoprogression from progressive disease. In multi-institutional dataset, the developed machine learning model showed an acceptable performance. This algorithm involving MRI data and clinical features could help making decision during patients’ disease course. For the practical use, we calibrated the machine learning model to offer the probability of pseudoprogression to clinicians, then we constructed the web-based user interface to access the model.

Abstract

Some patients with glioblastoma show a worsening presentation in imaging after concurrent chemoradiation, even when they receive gross total resection. Previously, we showed the feasibility of a machine learning model to predict pseudoprogression (PsPD) versus progressive disease (PD) in glioblastoma patients. The previous model was based on the dataset from two institutions (termed as the Seoul National University Hospital (SNUH) dataset, N = 78). To test this model in a larger dataset, we collected cases from multiple institutions that raised the problem of PsPD vs. PD diagnosis in clinics (Korean Radiation Oncology Group (KROG) dataset, N = 104). The dataset was composed of brain MR images and clinical information. We tested the previous model in the KROG dataset; however, that model showed limited performance. After hyperparameter optimization, we developed a deep learning model based on the whole dataset (N = 182). The 10-fold cross validation revealed that the micro-average area under the precision-recall curve (AUPRC) was 0.86. The calibration model was constructed to estimate the interpretable probability directly from the model output. After calibration, the final model offers clinical probability in a web-user interface.

Manual segmentation versus semi-automated segmentation for quantifying vestibular schwannoma volume on MRI

Purpose

Management of vestibular schwannoma (VS) is based on tumour size as observed on T1 MRI scans with contrast agent injection. The current clinical practice is to measure the diameter of the tumour in its largest dimension. It has been shown that volumetric measurement is more accurate and more reliable as a measure of VS size. The reference approach to achieve such volumetry is to manually segment the tumour, which is a time intensive task. We suggest that semi-automated segmentation may be a clinically applicable solution to this problem and that it could replace linear measurements as the clinical standard.

Methods

Using high-quality software available for academic purposes, we ran a comparative study of manual versus semi-automated segmentation of VS on MRI with 5 clinicians and scientists. We gathered both quantitative and qualitative data to compare the two approaches; including segmentation time, segmentation effort and segmentation accuracy.

Results

We found that the selected semi-automated segmentation approach is significantly faster (167 s vs 479 s, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p<0.001$$\end{document}p<0.001), less temporally and physically demanding and has approximately equal performance when compared with manual segmentation, with some improvements in accuracy. There were some limitations, including algorithmic unpredictability and error, which produced more frustration and increased mental effort in comparison with manual segmentation.

Conclusion

We suggest that semi-automated segmentation could be applied clinically for volumetric measurement of VS on MRI. In future, the generic software could be refined for use specifically for VS segmentation, thereby improving accuracy.

Electronic supplementary material

The online version of this article (10.1007/s11548-020-02222-y) contains supplementary material, which is available to authorized users.