1) Review current state-of-the-art MR imaging techniques for diagnosis and management of brain tumors. 2) Describe recent progress and advances in molecular genetics of brain tumors and illustrate how these advances impact imaging interpretation. 3) Present strengths and pitfalls of advanced physiologic MR imaging techniques in the assessment of tumor activity following therapy.

To create an imaging genomic biomarker signature in order to identify those Glioblastoma patients (GBM) with 1p/19q deletion. Recent prospective randomized clinical trials have validated correlations between 1p/19q codeletion and increased overall survival of patients treated with radiation therapy with or without chemotherapy

Using The Cancer Genome Atlas (TCGA),we identified 99 treatment naive GBM patients for whom both gene and miRNA expression profiles including the 1p/19q codeletion status, and pretreatment brain MR Imaging from The Cancer Imaging Archive(TCIA) were available. The VASARI feature set and 3D Slicer software 3.6 (http://www.slicer.org) were used for image analysis and image review was done in consensus by 2 neuroradiologists. Fluid Attenuated Inversion Recovery (FLAIR) was used for segmentation of the edema/cellular infiltration and Post GD T1-weighted imaging (T1WI) for segmentation of tumor enhancement and necrosis. Imaging parameters were then correlated with 1p/19q deletion status and gene expression profiles. Multiple complex biomarker signatures based on gene profiling and survival were created.

A novel imaging biomarker signature using multiple imaging parameters predicted 1p/19q co-deletion in patients with GBM. These were also associated with overall survival and progression-free survival.

Imaging genomic signatures can be expected to promote a more robust personalized approach to patient care and accelerate drug development and help stratify patients in clinical trials. An imaging biomarker signature was created using both qualitative and quantitative imaging parameters that predicted 1p/19 deletion status and expression.

Prediction of 1p/19q status promotes a more effective personalized therapy and help stratify patients in clincial trials

The hallmark metabolic alteration of mutant IDH gliomas is the production of 2-hydroxyglutarate (2HG) which may play a central role in downstream effects. Hence, 2HG may be an ideal biomarker for both diagnosing IDH mutations and monitoring response to treatment. 2HG can be measured in-vivo by magnetic resonance spectroscopy and there is significant interest in developing methodology that performs reliably in patients. Here we present results obtained with a new 3D MR spectroscopic imaging (MRSI) sequence that maps 2HG over the entire volume of the tumor during treatment.

A robust 3D MRSI sequence for 2HG imaging was newly developed by integrating adiabatic J-difference spectral editing, spiral imaging, and real-time motion correction. The acquisition parameters were: TR=1.6s, TE=68ms, FOV=200x200x200 mm3, acquisition matrix 10x10x10, NA=20, acquisition time TA=9:55 min:s. Spectra were fitted with LCModel software. Measurements were performed on a 3T MR scanner. 3D MRSI was performed in 20 patients with mutant IDH1 gliomas (WHO grades II-IV) consented with an approved IRB protocol. A baseline scan was done after surgery and before start of adjuvant treatment. At the moment 9 patients have completed a second post-treatment scan. Adjuvant treatment included radiotherapy and/or chemotherapy. The post-treatment scan was done in a time interval of 1-3 months after treatment.

Detectable levels of 2HG were measured in all patients that did not have gross total resection of tumor. 3D metabolic maps were obtained for 2HG, choline, N-acetyl-aspartate, glutamate-glutamine, and lactate. In 9 patients who have undergone both pre- and post-treatment scans, 4 demonstrated marked decrease (30-50%) in the levels of 2HG after completion of adjuvant therapy as shown in Figure 1. The remainder showed partial reduction of 2HG, with no patients showing increased 2HG levels.

We demonstrate for the first time that 3D imaging of 2HG is clinically feasible in patients with IDH1 mutated gliomas. Quantification of 2HG levels in a cohort of mutant IDH glioma patients shows measurable changes during treatment.

2HG imaging could be used to answer clinically important questions of true-/pseudo-response and true-/pseudo-progression in mutant IDH glioma patients. 3D mapping of 2HG and other metabolites is important to capture tumor heterogeneity and reduce variability in longitudinal studies.

To analyse whether apparent diffusion coefficient (ADC) values correlate with survival and with methylguanine-DNA-methyltransferase (MGMT) promoter methylation status and epidermal growth factor receptor (EGFR) amplification on glioblastoma multiforme (GBM).

72 patients with untreated GBM before surgery were analysed (mean time MRI-Surgery=6 days). Patients were followed-up for at least 12 months or until death. A ROI were drawn on ADC-map in the highest restriction region of the tumor and on the normal-appearing contralateral white matter (NCWM). ADCmin-values and ADC-index defined as a ratio between tumoral ADCmin and NCWM-ADCmean were evaluated. MGMT-status(n=60), EGFR amplification(n=53), KPS, tumoral and residual volume, progression-free survival (PFS) and overall survival (OS) were analysed. Kaplan-Meier and Cox-regression model were performed.

53 patients had complete resection. Presurgical and post-surgical mean tumoral volume were 42.4cm3 and 0.57cm3 respectively. Non methylated-MGMT-status(n=27) and low ADC values (<0.7) correlates with a decrease in PFS and OS (p<0.001). EGFR amplification (n=19) was correlated with a decrease in PFS (p=0.029) only when low ADC values and non--methylated-MGMT-status were present. EGFR amplification was not correlated with a poor outcome in the group of higher ADC values and MGMT methylated status (p<0.001). On Kaplan-Meier analyses MGMT-status correlated better with PFS (p=0.002), while ADC values correlate better with OS (p 0.001). In the multivariate analysis low ADC values and MGMTNM status were significant predictors of prognosis when they were adjusted by clinical variables (p= 0.001 and p=0.037, respectively).

The combined use of ADC values and MGMT-status are stronger predictors than using separated in GBM and could modulate outcome in patients with EFGR amplification.

ADC values in GBM correlates significantly with survival, independently of the MGMT and EGFR status .Therefore, ADC values could be used as independent predictors of survival in those patients.

Resting state functional MRI (rs-FMRI) enables clinicians to define critical areas and margins for pre-surgical planning of brain tumor resections without requiring the active participation of the patient. While task-based FMRI has gained utility in the clinical environment, rs-FMRI needs to be automatized and verified in tumor patients to be useful as a reliable clinical tool.

Data were acquired from 48 patients (24 with brain tumors, 24 epilepsy and vascular lesions) including rs-FMRI, task-based FMRI, diffusion tensor imaging (DTI) and structural MRI on 1.5T and 3T MRI scanners. Data were preprocessed (Allen EA, 2011) using AFNI (NIH, Bethesda, MD) and FSL (Oxford, UK) and decomposed into individual functional network components using independent component analysis (ICA) implemented in the GIFT toolbox (MRN, Albuquerque, NM) calculated for 28 and 75 components. ICA components were both manually identified by a trained radiologist overlaid on the anatomical and DTI images and compared by spatial correlation to published template components from healthy subjects (Calhoun, 2008). Predictive values from radiologist vs. automation where generated as well as ranked cross-correlation values.

Reproducible ICA components could be identified from both the 28 and 75 component analyses. Higher component numbers resulted in higher spatial detail and higher classifier values, but occasionally led to functional networks distributed across several components. The median classifier achieved better than 80% agreement. Using the non-deformable MNI registration to warp templates into subject space, templates showed considerable overlap with the tumor in some instances. Calculated ICA components, however, followed the outline of the tumor highlighting functional gray matter as classified by a clinician.

Our automated classification allows extraction of functional network components quickly with good agreement to the manual reader and with seamless integration into the existing clinical FMRI workflow. A larger functional component template library for use with clinical patient populations is currently underway for further validation and improvement of classification accuracy.

Task-free functional MRI can aid in identification of eloquent brain tissue in tumor resections by outlining functional networks and critical margins where active patient participation is not possible.

1) Discuss the differences between image-centric and patient-centric perspectives in clinical Neuroradiology and presurgical brain mapping. 2) Cite the utility of clinical assessments and the electronic medical record in presurgical brain mapping. 3) Discuss the impact of presurgical brain mapping on surgical decision making.

1) Participants will comprehend the current RANO criteria and its limitations in practice. 2) Participants will comprehend and be able to apply digital T1 subtraction for quantification of tumor response. 3) Participants will gain an appreciation for how to use T2/FLAIR to measure response, challenges associated with T2/FLAIR, and potential solutions for measuring nonenhancing tumor response. 4) Participants will comprehend basic and advanced diffusion MR biomarkers to treatment response. 5) Participants will comprehend basic and advanced perfusion MR biomarkers to treatment response. 6) Participants will comprehend basic pH-weighted MR response using CEST imaging. 7) Participants will comprehend basic and advanced PET imaging response.

Depths of novel imaging biomarkers are now available for evaluating biological response to new therapies in CNS tumors. The current course will briefly outline the current Response Assessment in Neuro-Oncology (RANO) criteria, including limitations when implementing in multicenter trials. The use of digital T1 subtraction maps as a method for measuring enhancing tumor volume in the presence of agents that reduce vascular permeability will be discussed. T2/FLAIR response, challenges associated with interpreting T2/FLAIR response, and a potential solution for measuring nonenhancing tumor response using T2 relaxometry will be described. This course will outline simple and advanced diffusion MR biomarkers for patient stratification and response assessment, including ADC histogram analysis, functional diffusion mapping, voxel-wise proliferation and invasion modeling, and DREAM-MRI. Simple and advanced perfusion MR biomarkers, including DSC-MRI, DCE-MRI, and a new pharmacokinetic perfusion-diffusion model will be described. The use of pH-weighted MR response to therapy will also be discussed using CEST imaging. Lastly, basic and advanced PET imaging techniques will be described in the context of response assessment and drug target efficacy.

To provide standardized evaluation of Diffusion Tensor Imaging (DTI) tractography algorithms for mapping white matter pathways during glioma resection

Nineteen tractography teams reconstructed the corticospinal (CST) tract on a series of 11 cases presenting with a glioma near the motor cortex area (high-grade n=7, low-grade n=4), in the three editions of the DTI Tractography Challenge at the MICCAI 2011,2012 and 2013 conference. The datasets included DTI scans (20 and 30 gradient directions, b-value=1000 s/mm2) acquired on a 3T scanner, and co-registered T1-weighted and FLAIR scans with segmented tumor and edema. Participating teams were required to submit part of their tractography results prior to the workshop, and to process two cases in a limited time at the event. Five neurosurgeons and four DTI experts evaluated and discussed the tractography reconstructions using a web-based questionnaire with standardized views of the tractography results. Variability among methods was quantified based on the Dice coefficient of bundle overlap of the voxelized tracts.

The 232 corticospinal tracts submitted to the three editions of the DTI Challenge workshop showed a large inter-algorithm variability (average Dice coefficient of overlap: 0.23(tumor), 0.22(contralateral)). Standardized review of the results demonstrated that most algorithms could reconstruct the CST projection to the cortical motor foot area and identified limitations in the ability of some methods to track the lateral projections to the face and hand areas, as well as false-negative and false-positive tracts in both hemispheres. Improvements of the tractography reconstructions from year 1 to year 3 indicate this collaborative effort is a learning experience for the community.

DTI tractography reconstructions are complex geometric models of white matter anatomy that can provide clinically relevant information for the planning of glioma resection in eloquent areas. By providing a benchmark for the standardized evaluation of tractography algorithms on a common series of clinical data, the DTI tractography challenge initiative aims to accelerate the translation of novel tractography tools from research to the clinics.

Standardized evaluation of DTI tractography techniques can help establish the validity of tractography-derived information to assist in neurosurgical decision-making.

Amide proton transfer (APT) imaging is a novel molecular imaging approach that generates MRI contrast based on endogenous cellular proteins in tissue. The purpose of this study was to determine whether APT imaging can distinguish pseudoprogression from true progression or recurrence in patients with malignant glioma.

Total 53 patients with pathologically confirmed high-grade gliomas (anaplastic astrocytoma or glioblastoma) were assessed. All patients provided written informed consent as required. Eligibility criteria included: treated with concurrent chemotherapy and radiation therapy (CCRT) after surgical resection, developed new or enlarged contrast enhanced lesions after CCRT, and had standard clinical MRI before and after CCRT. APT-MRI scanning was performed at 3T (3D sequence; 15 slices; 4.4 mm thickness). APT-weighted MRI signals were calculated using magnetization transfer ratio asymmetry at 3.5ppm with respect to water. MRI analysis was made, blinded to pathologic diagnosis, based on longitudinal signal changes in T2W, FLAIR, DWI and gadolinium enhancement on T1W, lasting at least six months.

Longitudinal radiological analysis showed that 39 patients had true progression or recurrence and 14 patients had pseudoprogression. The true progression or recurrence is associated with APT hyperintensity, compared to contralateral normal-appear white matter, while pseudoprogression is associated with APT isointensity to mild hyperintensity. The average APT signal intensity was significantly higher in the true progression/recurrence group (2.76%  0.55%) than in the pseudoprogression group (1.19%  0.40%; P < 0.001). Based on the receiver operating characteristic (ROC) analysis, the cutoff APT signal intensity value was 1.89%, with a sensitivity of 100% and a specificity of 92.9%.

The APT-MRI signal may be a valuable imaging biomarker to distinguish between tumor progression or recurrence and pseudoprogression whose diagnosis typically needs repeated surgery or longitudinal MRI scanning over several months.

APT image can help distinguish pseudoprogression from true progression or recurrence. Such a distinction may avoid the time-consuming longitudinal MRI analysis and repeated craniotomy or biopsy.

ACRIN 6677/RTOG 0625 is a multi-center randomized phase II trial of bevacizumab with irinotecan or temozolomide in recurrent GBM. Pseudoresponse in patients receiving VEGF blockade has raised concerns that conventional MRI may not predict overall (OS) and progression-free survival (PFS). We compared the ability of relative cerebral blood volume (rCBV) from DSC-MRI and post-Gd 2D-T1 MRI after 2 weeks of treatment to predict OS and PFS.

37/123 patients enrolled consented to DSC-MRI plus conventional MRI, 13 (mean age 54±14 years, 7 men) with DSC-MRI at baseline plus 2 weeks after start of treatment. Two central readers determined response status at 2 weeks using 2D-T1 enhancement and Macdonald threshold criteria with adjudication if necessary. Enhancing ROIs were also defined semi-automatically from thresholded 2D-T1 difference images and used to extract mean GRE (TE=30-40ms) or SE (TE=60-105ms) rCBV (EPI, pre-load, 90° flip angle, post-processing leakage correction) normalized to normal-appearing white matter. Kaplan-Meier survival estimates and log rank test (2-sided) were used to determine if response status on 2D-T1 MRI and rCBV changes on DSC-MRI are predictive of PFS and OS, respectively. Fisher's exact test (2-sided) was used to determine association between change in rCBV and response status on 2D-T1 MRI.

At 2 weeks, there were 3 responders and 10 non-responder/non-progressors (NR-NPs) on 2D-T1, and 4 positive and 9 negative changes from baseline in rCBV. One patient (NR-NP, positive rCBV change) had progressed clinically before week 2 and was excluded from PFS analyses. PFS was significantly worse for patients with increasing vs. decreasing rCBV (p=0.0034), but not for responders vs. NR-NPs (p=0.44). Similarly, survival time was significantly shorter for patients with increasing vs. decreasing rCBV (p=0.0015) but not for responders vs. NR-NPs (p=0.92). There was no significant association between positive vs. negative change in rCBV and responders vs. NR-NPs on 2D-T1 MRI (p=1.0).

After 2 weeks of anti-VEGF therapy, change in rCBV from baseline has highly significant prognostic value for PFS and OS, whereas 2D-T1 response status does not.

Early increase in rCBV may be a useful MRI biomarker for the failure of anti-VEGF therapy, permitting a timely switch to alternative trials when necessary.

Funded through NCI U01-CA079778 and U01-CA080098.

Acidosis is a hallmark of the tumor extracellular microenvironment. Additionally, studies have shown that tumor regions have increased amino acid uptake in order to meet high metabolic demands. Chemical exchange saturation transfer (CEST) MRI is a non-invasive imaging technique that can provide molecular information about the functional groups of molecules. The CEST signal is sensitive to many factors that affect chemical exchange between molecules, including metabolite concentration and pH. In the current study, we develop and test CEST MRI targeted to the amino acid amine group as a pH-weighted imaging biomarker for identifying cancer tissue in patients with various brain tumors.

Samples of glutamine in water at varying pH (4.0 to 8.6 in units of 0.2) were created at varying concentration. Additionally, samples of phenylalanine and glycine were created for the same pH range. CEST data for these samples were collected at 3T on a Siemens Trio scanner (B1=2μT, 15 100-ms RF saturation pulses, 51 spectral points, ± 5.0 ppm). A normalization image was acquired by setting B1=0. Additionally, serial CEST data for a cohort of 12 GBM patients before, during, and after radiochemotherapy. Image-guided biopsies were obtained in an additional two patients with suspected tumor recurrence.

Results show high CEST asymmetry in low pH values between 5.0-7.0 pH and with increasing amino acid concentration. In GBM patients, changes in elevated CEST signal during radiotherapy provided early, independent information regarding the status of the tumor. Some patients showed continual increase in CEST positive regions during therapy, which was followed by early tumor progression (Fig. 1A). In cases of confirmed pseudoprogression, no elevated CEST asymmetry was noted despite an increase in tumor volume on anatomical images (Fig. 1B). Image-guided biopsies of CEST positive locations confirmed tumor, whereas CEST negative regions showed gliosis and little tumor activity.

CEST MRI targeted to the amine protons may provide a pH-weighted imaging biomarker for identifying regions of active tumor proliferation in patients with brain tumors.

A non-invasive imaging method for obtaining tissue pH information would be invaluable as a tool for detecting human cancers and characterizing tumor response to therapy.

1) Review the role of F-18 FDG in brain tumor imaging. 2) Discuss metabolic brain tumor imaging with amino acids and proliferation markers and learn the complimentary information provided to MRI techniques. 3) Introduce novel alkylphosphocholine analogues, CLR1404 and CLR1502, that can be used for PET imaging, in vivo optical imaging, and therapy of brain tumors.