Pleomorphic xanthoastrocytoma (PXA) is a rare astrocytic neoplasm that commonly affects children and young adults, and presents with seizures. PXA is typically supratentorial with a predilection to the temporal lobe, and often involves the cortex and the meninges. PXAs have a favorable prognosis with a 10-year survival probability of >70%, and are WHO grade II neoplasms. Recent observations and studies demonstrate that PXAs are clinically, histologically and genetically distinct. Some PXAs recur and exhibit aggressive clinical behavior. In such cases, certain histological and clinical factors could account for the aggressive behavior. However, the histological features that predict adverse outcome are poorly defined. In the current WHO classification of CNS tumors, there is no option for a high-grade PXA, even if the tumor had numerous recurrences and poor outcome. In this review, we focus on aggressive clinical behavior and anaplasia in PXA, and discuss how our current experience suggests modifications in the current WHO classification. We also review recent discoveries on the molecular characteristics of PXA that could help us better understand their biological behavior.

Pilocytic and pilomyxoid astrocytomas are some of the most common gliomas in children and young adults. These gliomas are indolent neoplasms with long overall survival probability. The genetic characteristics of these neoplasms are well known, and our deepened understanding of their associated molecular alterations has led to the development of novel treatment strategies and approaches. Currently, we can account for some of the unusual behavior, such as oncogene-induced senescence, associated spontaneous regression, anaplastic transformation, and cerebrospinal dissemination, of these gliomas. Nevertheless, enigmatic issues continue to surround these chronic tumors. Here, we review the classical and uncommon clinical pathological and genetic features of these indolent gliomas.

G2 → M transition is a strategic target for glioma chemotherapy. Key players in G2 → M transition include CDC2 and glycogen synthase kinase 3β (GSK3β), which are highly regulated by posttranslational phosphorylation. This report is a morphological analysis of CDC2 and GSK3β phosphorylation using immunohistochemistry in gliomas with different biological properties. GBM showed a 2.8-fold and 5.6-fold increase in number of cells positive for pThr161CDC2 and a 4.2- and 6.9-fold increase in number of cells positive for pTyr15CDC2 relative to oligodendroglioma and ependymoma, respectively. Elevated labeling for inhibited phospho-CDC2 (pTyr15CDC) correlates with elevated levels of phosphorylated glycogen synthase kinase 3β (GSK3β). 71% of the GBM cases showed intermediate to high intensity staining for pSer9SGK3β 53% of oligodendroglioma, and 73% of ependymoma showed low intensity staining. CDC2 gene amplification correlates with increased survival in glioblastoma multiforme (GBM) and astrocytoma WHO grades II-III, but not in oligodendroglioma WHO grades II-III.

Purpose

We applied voxelwise apparent diffusion coefficient (ADC) histogram analysis in addition to structural magnetic resonance imaging (MRI) findings and patients' age for differentiation of intraaxial posterior fossa tumors involving the fourth ventricle.

Participants and methods

Pretreatment MRIs of 74 patients with intraaxial brain neoplasm involving the fourth ventricle, from January 1, 2004 to December 31, 2015, were reviewed. The tumor solid components were segmented and voxelwise ADC histogram variables were determined. Histogram-driven variables, structural MRI findings, and patient age were combined to devise a differential diagnosis algorithm.

Results

The most common neoplasms were ependymomas ( n = 21), medulloblastoma ( n = 17), and pilocytic astrocytomas ( n = 13). Medulloblastomas followed by atypical teratoid/rhabdoid tumors had the lowest ADC histogram percentile values; whereas pilocytic astrocytomas and choroid plexus papillomas had the highest ADC histogram percentile values. In a multivariable multinominal regression analysis, the ADC 10th percentile value from voxelwise histogram was the only independent predictor of tumor type ( p < 0.001). In separate binary logistic regression analyses, the 10th percentile ADC value, tumor morphology, enhancement pattern, extension into Luschka/Magendie foramina, and patient age were predictors of different tumor types. Combining these variables, we devised a stepwise diagnostic model yielding 71% to 82% sensitivity, 91% to 95% specificity, 75% to 78% positive predictive value, and 89% to 95% negative predictive value for differentiation of ependymoma, medulloblastoma, and pilocytic astrocytoma.

Conclusion

We have shown how the addition of quantitative voxelwise ADC histogram analysis of the tumor solid component to structural findings and patient age can help with accurate differentiation of intraaxial posterior fossa neoplasms involving the fourth ventricle based on pretreatment MRI.

Background and purpose

We aimed to develop a diagnostic algorithm for differentiation of cerebellar hemisphere tumors, combining Apparent Diffusion Coefficient (ADC) histogram analysis and structural imaging features.

Methods

Pretreatment MRI of patients with pathologically proven cerebellar hemisphere neoplasms were reviewed. Voxel-wise volumetric ADC histograms of tumor solid components were determined. Histogram variables, patients' age, and structural imaging features were applied to develop a differential diagnosis algorithm.

Results

Among 142 patients, the most common tumors were metastasis (n = 54), hemangioblastoma (n = 39), pilocytic astrocytoma (n = 18), and medulloblastoma (n = 9). On ADC histogram analysis, medulloblastomas had the lowest, and pilocytic astrocytomas had the highest ADC values. An ADC 15th percentile value < 580 × 10^-6 mm² /s could differentiate medulloblastomas from other cerebellar hemisphere tumors with receiver operating characteristic (ROC) area under the curve (AUC) of .98 (P < .001). Comparing the two most common adult cerebellar hemisphere tumors (metastases and hemangioblastomas), hemangioblastomas had higher ADC values; and an ADC 90th percentile value > 2000 × 10^-6 mm² /s could identify hemangioblastomas with ROC AUC of .82 (P < .001). Along with ADC histogram and patients' age, structural imaging features including enhancement pattern, prominent vascular flow voids, surrounding FLAIR hyperintensity, and solid component T2 signal hyperintensity were independent differentiation variables in the diagnostic algorithm.

Conclusion

In pretreatment differentiation of cerebellar tumors, the ADC histogram analysis is particularly helpful for distinction of medulloblastomas from other subtypes, and differentiation of the two most common adult tumors (ie, metastases and hemangioblastomas) from each other. Combination of structural imaging features and ADC histogram analysis can help with pretreatment differentiation of cerebellar tumors.

Genetic profiling is an increasingly useful tool for sub-classification of gliomas in adults and children. Specific gene mutations, structural rearrangements, DNA methylation patterns, and gene expression profiles are now recognized to define molecular subgroups of gliomas that arise in distinct anatomic locations and patient age groups, and also provide a better prediction of clinical outcomes for glioma patients compared to histologic assessment alone. Understanding the role of these distinctive genetic alterations in gliomagenesis is also important for the development of potential targeted therapeutic interventions. Mutations including K27M and G34R/V that affect critical amino acids within the N-terminal tail of the histone H3 variants, H3.3 and H3.1 (encoded by H3F3A and HIST1H3B genes), are prime examples of mutations in diffuse gliomas with characteristic clinical associations that can help diagnostic classification and guide effective patient management. These histone H3 mutations frequently co-occur with inactivating mutations in ATRX in association with alternative lengthening of telomeres. Telomere length can also be maintained through upregulation of telomerase reverse transcriptase (TERT) expression driven by mutation within the TERT gene promoter region, an alteration most commonly found in oligodendrogliomas and primary glioblastomas arising in adults. Interestingly, the genetic alterations perturbing histone and telomere function in pediatric gliomas tend to be different from those present in adult tumors. We present a review of these mutations affecting the histone code and telomere length, highlighting their importance in prognosis and as targets for novel therapeutics in the treatment of diffuse gliomas.

We applied machine learning algorithms for differentiation of posterior fossa tumors using apparent diffusion coefficient (ADC) histogram analysis and structural MRI findings. A total of 256 patients with intra-axial posterior fossa tumors were identified, of whom 248 were included in machine learning analysis, with at least 6 representative subjects per each tumor pathology. The ADC histograms of solid components of tumors, structural MRI findings, and patients' age were applied to construct decision models using Classification and Regression Tree analysis. We also compared different machine learning classification algorithms (i.e., naïve Bayes, random forest, neural networks, support vector machine with linear and polynomial kernel) for dichotomized differentiation of the 5 most common tumors in our cohort: metastasis (n = 65), hemangioblastoma (n = 44), pilocytic astrocytoma (n = 43), ependymoma (n = 27), and medulloblastoma (n = 26). The decision tree model could differentiate seven tumor histopathologies with terminal nodes yielding up to 90% accurate classification rates. In receiver operating characteristics (ROC) analysis, the decision tree model achieved greater area under the curve (AUC) for differentiation of pilocytic astrocytoma (p = 0.020); and atypical teratoid/rhabdoid tumor ATRT (p = 0.001) from other types of neoplasms compared to the official clinical report. However, neuroradiologists' interpretations had greater accuracy in differentiating metastases (p = 0.001). Among different machine learning algorithms, random forest models yielded the highest accuracy in dichotomized classification of the 5 most common tumor types; and in multiclass differentiation of all tumor types random forest yielded an averaged AUC of 0.961 in training datasets, and 0.873 in validation samples. Our study demonstrates the potential application of machine learning algorithms and decision trees for accurate differentiation of brain tumors based on pretreatment MRI. Using easy to apply and understandable imaging metrics, the proposed decision tree model can help radiologists with differentiation of posterior fossa tumors, especially in tumors with similar qualitative imaging characteristics. In particular, our decision tree model provided more accurate differentiation of pilocytic astrocytomas from ATRT than by neuroradiologists in clinical reads.

Optic pathway gliomas account for 3-5% of all pediatric CNS tumors and represent the most common intrinsic optic nerve tumors. These tumors occur preferentially during the first decade of life and are particularly frequent in children with neurofibromatosis type 1. Although optic pathway gliomas are low-grade tumors, their behavior can be aggressive, and their management is often challenging. Their management includes observation, surgery, chemotherapy and radiation. The role of each modality is discussed as well as current and future developments in treatment, in particular targeted therapies that are currently being investigated.

We applied machine learning algorithms for differentiation of posterior fossa tumors using apparent diffusion coefficient (ADC) histogram analysis and structural MRI findings. A total of 256 patients with intra-axial posterior fossa tumors were identified, of whom 248 were included in machine learning analysis, with at least 6 representative subjects per each tumor pathology. The ADC histograms of solid components of tumors, structural MRI findings, and patients age were applied to construct decision models using Classification and Regression Tree analysis. We also compared different machine learning classification algorithms (i.e., naïve Bayes, random forest, neural networks, support vector machine with linear and polynomial kernel) for dichotomized differentiation of the 5 most common tumors in our cohort: metastasis (n = 65), hemangioblastoma (n = 44), pilocytic astrocytoma (n = 43), ependymoma (n = 27), and medulloblastoma (n = 26). The decision tree model could differentiate seven tumor histopathologies with terminal nodes yielding up to 90% accurate classification rates. In receiver operating characteristics (ROC) analysis, the decision tree model achieved greater area under the curve (AUC) for differentiation of pilocytic astrocytoma (p = 0.020); and atypical teratoid/rhabdoid tumor ATRT (p = 0.001) from other types of neoplasms compared to the official clinical report. However, neuroradiologists interpretations had greater accuracy in differentiating metastases (p = 0.001). Among different machine learning algorithms, random forest models yielded the highest accuracy in dichotomized classification of the 5 most common tumor types; and in multiclass differentiation of all tumor types random forest yielded an averaged AUC of 0.961 in training datasets, and 0.873 in validation samples. Our study demonstrates the potential application of machine learning algorithms and decision trees for accurate differentiation of brain tumors based on pretreatment MRI. Using easy to apply and understandable imaging metrics, the proposed decision tree model can help radiologists with differentiation of posterior fossa tumors, especially in tumors with similar qualitative imaging characteristics. In particular, our decision tree model provided more accurate differentiation of pilocytic astrocytomas from ATRT than by neuroradiologists in clinical reads.