Glioblastoma and Radiotherapy: a multi-center AI study for Survival Predictions from MRI (GRASP study)

BACKGROUND

The aim was to predict survival of glioblastoma at eight months after radiotherapy (a period allowing for completing a typical course of adjuvant temozolomide), by applying deep learning to the first brain MRI after radiotherapy completion.

METHODS

Retrospective and prospective data were collected from 206 consecutive glioblastoma, IDH-wildtype patients diagnosed between March 2014-February 2022 across 11 UK centers. Models were trained on 158 retrospective patients from three centers. Holdout test sets were retrospective (n=19; internal validation), and prospective (n=29; external validation from eight distinct centers).Neural network branches for T2-weighted and contrast-enhanced T1-weighted inputs were concatenated to predict survival. A non-imaging branch (demographics/MGMT/treatment data) was also combined with the imaging model. We investigated the influence of individual MR sequences; non-imaging features; and weighted dense blocks pretrained for abnormality detection.

RESULTS

The imaging model outperformed the non-imaging model in all test sets (area under the receiver-operating characteristic curve, AUC p=0.038) and performed similarly to a combined imaging/non-imaging model (p>0.05). Imaging, non-imaging, and combined models applied to amalgamated test sets gave AUCs of 0.93, 0.79, and 0.91. Initializing the imaging model with pretrained weights from 10,000s of brain MRIs improved performance considerably (amalgamated test sets without pretraining 0.64; p=0.003).

CONCLUSIONS

A deep learning model using MRI images after radiotherapy, reliably and accurately determined survival of glioblastoma. The model serves as a prognostic biomarker identifying patients who will not survive beyond a typical course of adjuvant temozolomide, thereby stratifying patients into those who might require early second-line or clinical trial treatment.

Clinical, Imaging and Neurogenetic Features of Patients with Gliomatosis Cerebri Referred to a Tertiary Neuro-Oncology Centre

INTRODUCTION

Gliomatosis cerebri describes a rare growth pattern of diffusely infiltrating glioma. The treatment options are limited and clinical outcomes remain poor. To characterise this population of patients, we examined referrals to a specialist brain tumour centre.

METHODS

We analysed demographic data, presenting symptoms, imaging, histology and genetics, and survival in individuals referred to a multidisciplinary team meeting over a 10-year period.

RESULTS

In total, 29 patients fulfilled the inclusion criteria with a median age of 64 years. The most common presenting symptoms were neuropsychiatric (31%), seizure (24%) or headache (21%). Of 20 patients with molecular data, 15 had IDH wild-type glioblastoma, with an IDH1 mutation most common in the remainder (5/20). The median length of survival from MDT referral to death was 48 weeks (IQR 23 to 70 weeks). Contrast enhancement patterns varied between and within tumours. In eight patients who had DSC perfusion studies, five (63%) had a measurable region of increased tumour perfusion with rCBV values ranging from 2.8 to 5.7. A minority of patients underwent MR spectroscopy with 2/3 (66.6%) false-negative results.

CONCLUSIONS

Gliomatosis imaging, histological and genetic findings are heterogeneous. Advanced imaging, including MR perfusion, could identify biopsy targets. Negative MR spectroscopy does not exclude the diagnosis of glioma.

Diverse imaging features of adolescent glioblastoma

We highlight an unusual case of multifocal glioblastoma in an adolescent patient, manifesting as four discrete brain lesions, each distinct in appearance. Familiarity with the diverse imaging features of glioblastoma can reduce misdiagnosis and avoid treatment delays.

Imaging characteristics of H3 K27M histone-mutant diffuse midline glioma in teenagers and adults

Background

To assess anatomical and quantitative diffusion-weighted MR imaging features in a recently classified lethal neoplasm, H3 K27M histone-mutant diffuse midline glioma [World Health Organization (WHO) IV].

Methods

Fifteen untreated gliomas in teenagers and adults (median age 19, range, 14-64) with confirmed H3 K27M histone-mutant genotype were analysed at a national referral centre. Morphological characteristics including tumour epicentre(s), T2/FLAIR and Gadolinium enhancement patterns, calcification, haemorrhage and cyst formation were recorded. Multiple apparent diffusion coefficient (ADC_min, ADC_mean) regions of interest were sited in solid tumour and normal appearing white matter (ADC_NAWM) using post-processing software (Olea Sphere v2.3, Olea Medical). ADC histogram data (2^nd, 5^th, 10^th percentile, median, mean, kurtosis, skewness) were calculated from volumetric tumour segmentations and tested against the regions of interest (ROI) data (Wilcoxon signed rank test).

Results

The median interval from imaging to tissue diagnosis was 9 (range, 0-74) days. The structural MR imaging findings varied between individuals and within tumours, often featuring signal heterogeneity on all MR sequences. All gliomas demonstrated contact with the brain midline, and 67% exhibited rim-enhancing necrosis. The mean ROI ADC_min value was 0.84 (±0.15 standard deviation, SD) ×10^-3 mm²/s. In the largest tumour cross-section (excluding necrosis), an average ADC_mean value of 1.12 (±0.25)×10^-3 mm²/s was observed. The mean ADC_min/NAWM ratio was 1.097 (±0.149), and the mean ADC_mean/NAWM ratio measured 1.466 (±0.299). With the exception of the 2^nd centile, no statistical difference was observed between the regional and histogram derived ADC results.

Conclusions

H3 K27M-mutant gliomas demonstrate variable morphology and diffusivity, commonly featuring moderately low ADC values in solid tumour. Regional ADC measurements appeared representative of volumetric histogram data in this study.

World Health Organization Grade II/III Glioma Molecular Status: Prediction by MRI Morphologic Features and Apparent Diffusion Coefficient

Background A readily implemented MRI biomarker for glioma genotyping is currently lacking. Purpose To evaluate clinically available MRI parameters for predicting isocitrate dehydrogenase (IDH) status in patients with glioma. Materials and Methods In this retrospective study of patients studied from July 2008 to February 2019, untreated World Health Organization (WHO) grade II/III gliomas were analyzed by three neuroradiologists blinded to tissue results. Apparent diffusion coefficient (ADC) minimum (ADC_min) and mean (ADC_mean) regions of interest were defined in tumor and normal appearing white matter (ADC_NAWM). A visual rating of anatomic features (T1 weighted, T1 weighted with contrast enhancement, T2 weighted, and fluid-attenuated inversion recovery) was performed. Interobserver comparison (intraclass correlation coefficient and Cohen κ) was followed by nonparametric (Kruskal-Wallis analysis of variance) testing of associations between ADC metrics and glioma genotypes, including Bonferroni correction for multiple testing. Descriptors with sufficient concordance (intraclass correlation coefficient, >0.8; κ > 0.6) underwent univariable analysis. Predictive variables (P < .05) were entered into a multivariable logistic regression and tested in an additional test sample of patients with glioma. Results The study included 290 patients (median age, 40 years; interquartile range, 33-52 years; 169 male patients) with 82 IDH wild-type, 107 IDH mutant/1p19q intact, and 101 IDH mutant/1p19q codeleted gliomas. Two predictive models incorporating ADC_mean-to-ADC_NAWM ratio, age, and morphologic characteristics, with model A mandating calcification result and model B recording cyst formation, classified tumor type with areas under the receiver operating characteristic curve of 0.94 (95% confidence interval [CI]: 0.91, 0.97) and 0.96 (95% CI: 0.93, 0.98), respectively. In the test sample of 49 gliomas (nine IDH wild type, 21 IDH mutant/1p19q intact, and 19 IDH mutant/1p19q codeleted), the classification accuracy was 40 of 49 gliomas (82%; 95% CI: 71%, 92%) for model A and 42 of 49 gliomas (86%; 95% CI: 76%, 96%) for model B. Conclusion Two algorithms that incorporated apparent diffusion coefficient values, age, and tumor morphologic characteristics predicted isocitrate dehydrogenase status in World Health Organization grade II/III gliomas on the basis of standard clinical MRI sequences alone. © RSNA, 2020 Online supplemental material is available for this article.

Endogenous Chemical Exchange Saturation Transfer MRI for the Diagnosis and Therapy Response Assessment of Brain Tumors: A Systematic Review

Purpose

To generate a narrative synthesis of published data on the use of endogenous chemical exchange saturation transfer (CEST) MRI in brain tumors.

Materials and Methods

A systematic database search (PubMed, Ovid Embase, Cochrane Library) was used to collate eligible studies. Two researchers independently screened publications according to predefined exclusion and inclusion criteria, followed by comprehensive data extraction. All included studies were subjected to a bias risk assessment using the Quality Assessment of Diagnostic Accuracy Studies tool.

Results

The electronic database search identified 430 studies, of which 36 fulfilled the inclusion criteria. The final selection of included studies was categorized into five groups as follows: grading gliomas, 19 studies (area under the receiver operating characteristic curve [AUC], 0.500-1.000); predicting molecular subtypes of gliomas, five studies (AUC, 0.610-0.920); distinction of different brain tumor types, seven studies (AUC, 0.707-0.905); therapy response assessment, three studies (AUC not given); and differentiating recurrence from treatment-related changes, five studies (AUC, 0.880-0.980). A high bias risk was observed in a substantial proportion of studies.

Conclusion

Endogenous CEST MRI offers valuable, potentially unique information in brain tumors, but its diagnostic accuracy remains incompletely known. Further research is required to assess the method's role in support of molecular genetic diagnosis, to investigate its use in the posttreatment phase, and to compare techniques with a view to standardization.Keywords: Brain/Brain Stem, MR-Imaging, Neuro-OncologySupplemental material is available for this article.© RSNA, 2020.

Challenges in glucoCEST MR body imaging at 3 Tesla

BACKGROUND

The aim of this study was to translate dynamic glucose enhancement (DGE) body magnetic resonance imaging (MRI) based on the glucose chemical exchange saturation transfer (glucoCEST) signal to a 3 T clinical field strength.

METHODS

An infusion protocol for intravenous (i.v.) glucose was optimised using a hyperglycaemic clamp to maximise the chances of detecting exchange-sensitive MRI signal. Numerical simulations were performed to define the optimum parameters for glucoCEST measurements with consideration to physiological conditions. DGE images were acquired for patients with lymphomas and prostate cancer injected i.v. with 20% glucose.

RESULTS

The optimised hyperglycaemic clamp infusion based on the DeFronzo method demonstrated higher efficiency and stability of glucose delivery as compared to manual determination of glucose infusion rates. DGE signal sensitivity was found to be dependent on T2, B1 saturation power and integration range. Our results show that motion correction and B0 field inhomogeneity correction are crucial to avoid mistaking signal changes for a glucose response while field drift is a substantial contributor. However, after B0 field drift correction, no significant glucoCEST signal enhancement was observed in tumour regions of all patients in vivo.

CONCLUSIONS

Based on our simulated and experimental results, we conclude that glucose-related signal remains elusive at 3 T in body regions, where physiological movements and strong effects of B1 + and B0 render the originally small glucoCEST signal difficult to detect.

Diagnostic accuracy of dynamic contrast-enhanced perfusion MRI in stratifying gliomas: A systematic review and meta-analysis

Background

T1‐weighted dynamic contrast‐enhanced (DCE) perfusion magnetic resonance imaging (MRI) has been broadly utilized in the evaluation of brain tumors. We aimed at assessing the diagnostic accuracy of DCE‐MRI in discriminating between low‐grade gliomas (LGGs) and high‐grade gliomas (HGGs), between tumor recurrence and treatment‐related changes, and between primary central nervous system lymphomas (PCNSLs) and HGGs.

Methods

We performed this study based on the Preferred Reporting Items for Systematic Reviews and Meta‐Analysis of Diagnostic Test Accuracy Studies criteria. We systematically surveyed studies evaluating the diagnostic accuracy of DCE‐MRI for the aforementioned entities. Meta‐analysis was conducted with the use of a random effects model.

Results

Twenty‐seven studies were included after screening of 2945 possible entries. We categorized the eligible studies into three groups: those utilizing DCE‐MRI to differentiate between HGGs and LGGs (14 studies, 546 patients), between recurrence and treatment‐related changes (9 studies, 298 patients) and between PCNSLs and HGGs (5 studies, 224 patients). The pooled sensitivity, specificity, and area under the curve for differentiating HGGs from LGGs were 0.93, 0.90, and 0.96, for differentiating tumor relapse from treatment‐related changes were 0.88, 0.86, and 0.89, and for differentiating PCNSLs from HGGs were 0.78, 0.81, and 0.86, respectively.

Conclusions

Dynamic contrast‐enhanced‐Magnetic resonance imaging is a promising noninvasive imaging method that has moderate or high accuracy in stratifying gliomas. DCE‐MRI shows high diagnostic accuracy in discriminating between HGGs and their low‐grade counterparts, and moderate diagnostic accuracy in discriminating recurrent lesions and treatment‐related changes as well as PCNSLs and HGGs.

CD8+ encephalitis: a severe but treatable HIV-related acute encephalopathy

Rapidly progressive encephalopathy in an HIV-positive patient presents a major diagnostic and management challenge. CD8+ encephalitis is a severe but treatable form of HIV-related acute encephalopathy, characterised by diffuse perivascular and intraparenchymal CD8+ lymphocytic infiltration. It can occur in patients who are apparently stable on antiretroviral treatment and probably results from viral escape into the central nervous system. Treatment, including high-dose corticosteroids, can give an excellent neurological outcome, even in people with severe encephalopathy and a very poor initial neurological status. We report a woman with CD8+ encephalitis, with a normal CD4 count and undetectable serum viral load, who made a good recovery despite the severity of her presentation.

Venous infarction mimicking a neoplasm in spontaneous intracranial hypotension: an unusual cause of Parinaud's syndrome

We present a case of longstanding, undiagnosed spontaneous intracranial hypotension (SIH) with an acute presentation of Parinaud's syndrome, in whom serial imaging demonstrated development of a midbrain mass. The patient was ultimately diagnosed with tumefactive venous infarction secondary to SIH. However, this patient underwent a brainstem biopsy, which in retrospect may have been avoidable. This case demonstrates the imaging features of tumefactive venous infarction in SIH and highlights the risk of misinterpretation as a neoplasm with potentially catastrophic consequences.

Vascular anomalies of the head and neck in children

Sixty percent of vascular anomalies in children are found in the head and neck. These lesions can present throughout antenatal, perinatal and childhood development. They broadly fall into two categories: vascular tumours and vascular malformations. Their clinical and, often, psychological impact is determined by both pathological type and location: many lesions follow an uncomplicated natural course and other more complex, extensive or progressive lesions can present a threat to life from mass effect, haemorrhage or large volume arteriovenous shunting. Vascular tumours include infantile haemangioma (IH), congenital haemangioma (CH) and kaposiform hemangioendothelioma (KH); of which IH is the most common. Management options for vascular tumours include conservative approaches, oral medications and surgical intervention as determined by tumour type, location and associated complications. Vascular malformations can be categorised into low flow and high flow lesions. Low flow lesions include capillary, venous and lymphatic malformations (LMs). High flow lesions describe the arteriovenous malformations (AVMs), a highly heterogeneous group of lesions which can present in a variety of ways-the mainstay of treatment for these dynamic lesions is endovascular or surgical obliteration. We provide a practical framework for clinical classification of vascular anomalies of the head and neck in children. We also explore principles of their clinical and radiological assessment along with management, highlighting the importance of a multi-disciplinary approach.

Trypsin-specific acyl-4-guanidinophenyl esters for alpha-chymotrypsin-catalysed reactions computational predictions, hydrolyses, and peptide bond formation

The function of acyl-4-guanidinophenyl esters as substrate mimetics for the serine protease alpha-chymotrypsin was investigated by protein-ligand docking, hydrolysis, and acyl transfer experiments. On the basis of protein-ligand docking studies, the binding and hydrolysis properties of these artificial substrates were estimated. The predictions of the rational approach were confirmed by steady-state hydrolysis studies on 4-guanidinophenyl esters derived from coded amino acids (which alpha-chymotrypsin is not specific for), noncoded amino acids, and even simple carboxylic acid moieties. Enzymatic peptide syntheses qualify these esters as suitable acyl donors for the coupling of acyl components far from the natural enzyme specificity, thus considerably expanding the synthetic utility of alpha-chymotrypsin.

Protease-catalyzed hydrolysis of substrate mimetics (inverse substrates): A new approach reveals a new mechanism

Contrary to common protease substrates, the hydrolysis of 4-guanidinophenyl esters of the Boc-Xaa-OGp type by trypsin and trypsin-like proteases performs easily and independently of the structure and chirality of the acyl moiety. The hydrolysis of this new class of substrate mimetics, previously called inverse substrates, is enabled by the highly specific leaving group. However, the mechanism cannot be explained on the basis of the conventional binding model which defines the interactions between the protease and its substrate. Hydrolysis and aminolysis kinetics, protein-ligand docking, and molecular dynamics studies have been carried out in order to get insight into the catalytic mechanism which holds for these substrate mimetics. The experimental and theoretical results obtained for the serine protease trypsin suggest a novel extended kinetic model. It explains the hydrolysis of these types of protease substrates and accounts for the structural consequences for their aminolysis.