Projekt

Zurück zur Übersicht

Establishing Novel MR Criteria for the Assessment of Malignant Glioma Progression

Titel Englisch Establishing Novel MR Criteria for the Assessment of Malignant Glioma Progression
Gesuchsteller/in Slotboom Johannes
Nummer 140958
Förderungsinstrument Projektförderung (Abt. I-III)
Forschungseinrichtung Universitätsinstitut für Diagnostische, Interventionelle & Pädiatrische Radiolog Inselspital
Hochschule Universität Bern - BE
Hauptdisziplin Klinische Krebsforschung
Beginn/Ende 01.10.2012 - 31.10.2016
Bewilligter Betrag 270'730.00
Alle Daten anzeigen

Alle Disziplinen (4)

Disziplin
Klinische Krebsforschung
Informatik
Technische Physik
Andere Gebiete der Ingenieurwissenschaften

Keywords (11)

classification; glioma; pseudo progression; Perfusion; MR-spectroscopy; texture analysis; pseudo response; MR-Imaging; RANO criteria; image processing; Diffusion

Lay Summary (Englisch)

Lead
Lay summary

Brain-Tumors

In the human brain there two major types of cells: on one hand there are neurons, which are the electrically active cells which control our thinking, moving, breathing, heart beating, body temperature etcetera, and on the other hand there are glial cells which provide support and protection for the neurons in the brain. These glial cells may change their nature during our life, and may turn into tumor cells, which than rapidly increase in number by uncontrolled cell division. The tumors of glial cells are called gliomas. Since there are various glial-cells sub-types, there are also various glioma tumor types (e.g. ependymomas, astrocytomas, oligodendrogliomas). Further, there are less aggressive gliomas, the so-called low grade gliomas, and rapid growing, aggressive gliomas, the so-called high-grade gliomas. The World Health Organization (WHO) made a classification system for glioma distinguishing based on their aggressiveness: low grade glioma are those of WHO grade I and II; the very aggressive, malignant, glioma are of WHO grade III and IV. This SNF study focuses on malignant [WHO grade-III-IV] brain tumors for which have an overall incidence of 5 per 100000 people per year, and are the most common brain tumors.

Diagnostics

For the diagnostics of brain tumors, magnetic resonance imaging (MRI) is the neuroimaging method of first choice since it enables not only the measurement of the tumor location and size, but also allows insights regarding tumor aggressiveness by the application of contrast agents, tumor sub-typing using magnetic resonance spectroscopy (MRS), the blood-flow through the tumor tissue, and quantification of the severity of brain swelling (edema) by measuring the diffusion of the water molecule in the tumor. Nevertheless, the gold standard for initial diagnosis of brain tumors is the histological examination of biopsies. In a image guided surgical intervention the neurosurgeon removes a tumor tissue to be examined by a pathologist. If the location of the tumor does not allow to take a biopsy, for instance in tumors of the brain stem, diagnostics by neuroimaging methods is of utmost importance.

Criteria for Tumor-Progression

During the course of the disease, glioma change their size and aggressiveness. This process is called tumor progression. Neuroimaging criteria have been defined in the beginning of the nineties of last century that define imaging criteria for progression. These so-called Macdonald criteria also define whether the tumor is said to be stable, or shows progression, i.e. increases size and changes its level of aggressiveness. Since the first definition of these Macdonald criteria the treatment of glioma has changed a great deal. The tumors are nowadays treated with a combination of radiation therapy and chemo-therapy with temozolomide. Very recently, new anti-angiogenic drugs came on the market that slows down or prevent the growth of new blood-vessels that feed the brain-tumor, and by this, to stop further tumor growth. Radiation therapy changes the characteristics of the blood-brain-barrier and the irradiated tissue will show signal enhancement in T1-weighted MR-images. Signal enhancement however is in untreated tumors associated with high grade gliomas. This means that an irradiated low-grade tumor may appear as an high-grade tumor after therapy. This effect is recognized, and is called pseudo-progression of the tumor. The effect of anti-angiogenic drugs is opposite: the T1-signal enhancement” disappears due to restoration of the blood-brain-barrier, but the tumor is still there. This effect is called pseudo-response. These effects were recognized by the Response Assessment in Neuro-Oncology (RANO) Working Group that adapted the Macdonald criteria for glioma progression classification.

At the time of the first formulation of the Macdonald criteria at the beginning of the nineties of the last century, a standard MR-examination consisted predominantly of T1- and T2-weighted MR-imaging. In the meantime MR neuro-imaging techniques have developed tremendously. As mentioned above, diffusion weighted imaging (DWI-MRI), perfusion weighted imaging (PWI-MRI) have been established today, and are frequently applied for initial tumor diagnosis and tumor-progression evaluation. Also MRS has established itself as a valuable tool for initial tumor diagnostics and progression evaluation. Despite the fact that these techniques have been established, their use in tumor-progression evaluation in the RANO criteria is still very limited.

Image Analysis

During the last 30 years the analysis of digital medical MR-images has considerably improved. There are novel computer aided methods available now for automatic tumor detection within the images, and automatic tissue classification. This study will focus on the use of so-called “texture parameters” in combination with advanced statistical classification methods.

Aims of this Study

The primary aim of our proposal is to improve the reliability of tumor progression evaluation. To do so, this study foresees the acquisition of T1-, T2-weighted, perfusion-, diffusion-weighted MR-images, as well as localized MR-spectra of de novo glioma patients before, during and after treatment by neurosurgery, radio-chemotherapy and/or anti-angiogenic therapy. With the aid of a tumor progression analysis tool that will be further developed during the study, the clinician can analyze simultaneously all MRI/MRS data and can extract relevant tumor progression information from the tool. At least 80 so-called image texture analysis will be derived from the MRI an MRS datasets for 90 high grade glioma patients as a function of time. MR-measurements will be made at first diagnosis, after neuro-surgery, and during the standard regular 3 month interval follow-up examinations for a period of 2 years. We hypothesize that by retrospective analysis of the collected data, we will be able to distinguish true tumor-progression from pseudo-progression, and true tumor-response to therapy, from pseudo-response, enabling targeted patient treatment in the future.

Direktlink auf Lay Summary Letzte Aktualisierung: 21.02.2013

Verantw. Gesuchsteller/in und weitere Gesuchstellende

Mitarbeitende

Publikationen

Publikation
A machine learning pipeline for supporting differentiation of glioblastomas from single brain metastases
Mocioiu Victor, Pedrosa de Barros Nuno, Ortega-Martorell Sandra, Slotboom Johannes, Knecht Urspeter, ArúsCarles, Vellido Alfredo, Julià-Sapé Margarida (2016), A machine learning pipeline for supporting differentiation of glioblastomas from single brain metastases, in European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Bruges, BelgiumESANN 2016 proceedings, Bruges, Belgium.
Automatic quality control in clinical 1H MRSI of brain cancer
Pedrosa de Barros Nuno, McKinley Richard, Wiest Roland, Slotboom Johannes (2016), Automatic quality control in clinical 1H MRSI of brain cancer, in NMR in Biomedicine, 29(5), 563-575.
Characterization of Enhancing MS Lesions by Dynamic Texture Parameter Analysis of Dynamic Susceptibility Perfusion Imaging.
Verma Rajeev Kumar, Slotboom Johannes, Cäcilia Locher, Heldner Mirjam, Weisstanner Christian, Abela Eugenio, Kellner-Weldon Frauke, Zbinden Martin, Kamm Christian, Wiest Roland (2016), Characterization of Enhancing MS Lesions by Dynamic Texture Parameter Analysis of Dynamic Susceptibility Perfusion Imaging., in Biomed Research International, 2016, 9 pages.
Clinical Evaluation of a Fully-automatic Segmentation Method for Longitudinal Brain Tumor Volumetry.
Meier Raphael, Knecht Urs, Loosli Tina, Bauer Stephan, Slotboom Johannes, Wiest Roland, Reyes Mauricio (2016), Clinical Evaluation of a Fully-automatic Segmentation Method for Longitudinal Brain Tumor Volumetry., in Scientific reports, 6, 23376.
Fully Automated Enhanced Tumor Compartmentalization: Man vs. Machine Reloaded.
Porz Nicole, Habegger Simon, Meier Raphael, Verma Rajeev, Jilch Astrid, Fichtner Jens, Knecht Urspeter, Radina C, Schucht Philippe, Beck Jürgen, Raabe Andreas, Slotboom Johannes, Reyes Mauricio, Wiest Roland (2016), Fully Automated Enhanced Tumor Compartmentalization: Man vs. Machine Reloaded., in PLoS One, 11:e016530.
The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)
Menze BH, Jakab A, Bauer S, Kalpathy-Cramer J, Farahani K, Kirby J, Burren Y, Porz N, Slotboom J, Wiest R, et al. (2015), The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS), in IEEE Transactions on Medical Imaging, 34(10), 1993-2024.
5-ALA complete resections go beyond MR contrast enhancement: shift corrected volumetric analysis of the extent of resection in surgery for glioblastoma.
Schucht Philippe, Knittel Sonja, Slotboom Johannes, Seidel Kathleen, Murek Michael, Jilch Astrid, Raabe Andreas, Beck Jürgen (2014), 5-ALA complete resections go beyond MR contrast enhancement: shift corrected volumetric analysis of the extent of resection in surgery for glioblastoma., in Acta Neurochir (Wien), 156(2), 305-312.
Intraoperative monopolar mapping during 5-ALA-guided resections of glioblastomas adjacent to motor eloquent areas: evaluation of resection rates and neurological outcome.
Schucht Philippe, Seidel Kathleen, Beck Jürgen, Murek Michael, Jilch Astrid, Wiest Roland, Fung Christian, Raabe Andreas (2014), Intraoperative monopolar mapping during 5-ALA-guided resections of glioblastomas adjacent to motor eloquent areas: evaluation of resection rates and neurological outcome., in Neurosurgery Focus, 37, E16.
Multi-Modal Glioblastoma Segmentation: Man versus Machine.
Porz Nicole, Bauer Stephan, Pica Alessia, Schucht Philippe, Beck Jürgen, Verma Rajeev, Slotboom Johannes, Reyes Mauricio, Wiest Roland (2014), Multi-Modal Glioblastoma Segmentation: Man versus Machine., in PLoS One, e96873.
Early re-do surgery for glioblastoma is a feasible and safe strategy to achieve complete resection of enhancing tumor.
Schucht Philippe, Murek Michael, Jilch Astrid, Seidel Kathleen, Hewer Ekkehard, Wiest Roland, Raabe Andreas, Beck Jürgen (2013), Early re-do surgery for glioblastoma is a feasible and safe strategy to achieve complete resection of enhancing tumor., in PLoS One, 8(11), e79846.
Improving labelling efficiency in automatic quality control of MRSI data
Pedrosa de Barros Nuno, McKinley Richard, Wiest Roland, Slotboom Johannes, Improving labelling efficiency in automatic quality control of MRSI data, in Magnetic Resonance in Medicine.
Quality Management in in vivo proton MRS
Pedrosa de Barros Nuno, Johannes Slotboom, Quality Management in in vivo proton MRS, in Analytical Biochemistry.

Zusammenarbeit

Gruppe / Person Land
Formen der Zusammenarbeit
Max Planck Institute for Human Cognitive and Brain Sciences / Leipzig / Professo Harald Möller Deutschland (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
Radboud University Nijmegen Medical Centre / Department of Radiology/ Nijmegen, Prof A. Heerschap Niederlande (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Austausch von Mitarbeitern
Institute of Scientific Instruments Brno / Brno / Dr. Zenon Starcuk, Dr. Jana Starcuková Tschechische Republik (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Austausch von Mitarbeitern
- Industrie/Wirtschaft/weitere anwendungs-orientierte Zusammenarbeit
Universitat Autònoma de Barcelona, Barcelona / Spain / Prof. Miquel Cabañas, Prof. Carles Arús Spanien (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Austausch von Mitarbeitern
icoMetrix (company) Leuven / Scientist-in-charge: Dr. Dirk Loeckx Belgien (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Austausch von Mitarbeitern
- Industrie/Wirtschaft/weitere anwendungs-orientierte Zusammenarbeit
University Clinic of Neurosurgery / University Hospital Bern / Bern / Professor Raabe Schweiz (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Forschungsinfrastrukturen
- Austausch von Mitarbeitern
Department of Clinical Research, AMSM, University of Bern, Bern, Professor Kreis Schweiz (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Forschungsinfrastrukturen
- Industrie/Wirtschaft/weitere anwendungs-orientierte Zusammenarbeit
The University of Manchester, Imaging Science and Biomedical Engineering / Professor Steve Williams Grossbritannien und Nordirland (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
École Polytechnique Fédérale de Lausanne / Dr. Cristina Cudalbu Schweiz (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
Laboratoire CREATIS / Université Claude Bernard Lyon / Prof. Dominique Sappey-Marinier Frankreich (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
CIBER-BBN, Barcelona, Spain, Dr. Margarida Julià-Sapé Spanien (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Austausch von Mitarbeitern
- Industrie/Wirtschaft/weitere anwendungs-orientierte Zusammenarbeit
Department of Electrical Engineering ESAT-STADIUS / KU Leuven, Leuven, Prof. Sabine Van Huffel Belgien (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
ISTB / University Bern / Professor Mauricio Reyes Schweiz (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Industrie/Wirtschaft/weitere anwendungs-orientierte Zusammenarbeit
Department of Imaging & Pathology, Biomedical MRI Unit / KU Leuven, Leuven, / Prof. Himmelreich Belgien (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
Universtity Clinic of Radiooncology / University Hospital Bern / Professor Aebersold Schweiz (Europa)
- vertiefter/weiterführender Austausch von Ansätzen, Methoden oder Resultaten
- Publikation
- Forschungsinfrastrukturen

Wissenschaftliche Veranstaltungen

Aktiver Beitrag

Titel Art des Beitrags Titel des Artikels oder Beitrages Datum Ort Beteiligte Personen
International Society for Magnetic Resonance in Medicine (ISMRM) 2016 Vortrag im Rahmen einer Tagung Investigating Machine Learning Approaches for Quality Control of Brain Tumor Spectra. 07.05.2016 Singapur, Singapur Pedrosa de Barros Nuno Miguel; Slotboom Johannes;
International Society for Magnetic Resonance in Medicine (ISMRM) 2016 Vortrag im Rahmen einer Tagung Automatic quality assessment of short and long-TE brain tumour MRSI data using novel Spectral Features 07.05.2016 Singapur, Singapur Knecht Urspeter; Pedrosa de Barros Nuno Miguel; Wiest Roland; Slotboom Johannes;
ESMRMB - European Society for Magnetic Resonance in Medicine and Biology - 2015 Vortrag im Rahmen einer Tagung Highlighting differences between GBM and brain metastasis using a blind source separation method applied to MRSI data 01.10.2015 Edinburgh, Grossbritannien und Nordirland Slotboom Johannes; Pedrosa de Barros Nuno Miguel; Knecht Urspeter;
ESMRMB - European Society for Magnetic Resonance in Medicine and Biology - 2015 Vortrag im Rahmen einer Tagung Differentiating Glioblastomas from brain Metastases by means of Convex non-negative Matrix Factorization 01.10.2015 Edinburgh, Grossbritannien und Nordirland Slotboom Johannes; Knecht Urspeter; Pedrosa de Barros Nuno Miguel;
International Society for Magnetic Resonance in Medicine (ISMRM) 2015 Poster Computer Aided Radiological Diagnostics: Random Forest Classification of Glioma Tumor Progression Using Image Texture Parameters Derived from ADC-Maps. 30.05.2015 Toronto, Kanada Knecht Urspeter; Porz Nicole Bernadette; Pica Alessia; Wiest Roland; Raabe Andreas; Slotboom Johannes; Pedrosa de Barros Nuno Miguel;
International Society for Magnetic Resonance in Medicine (ISMRM) 2015 Vortrag im Rahmen einer Tagung Impact of Co-Registration on the Histogram Analysis of ADC maps in MRI/MRS Brain Tumor Diagnostics 30.05.2015 Toronto, Kanada Slotboom Johannes; Wiest Roland; Pedrosa de Barros Nuno Miguel; Knecht Urspeter;
European Association of Neuro-Oncology - EANO - 2014 Poster Target Delineation in Glioblastoma: is pre-operative automatic comparable to Expert based Segmentation? 09.10.2014 Torino, Italien Pica Alessia; Slotboom Johannes; Wiest Roland; Porz Nicole Bernadette;
European Association of Neuro-Oncology - EANO - 2014 Poster Unifying Clinical Routine Brain Tumor MR-Spectroscopy and MR-image analysis: Novel JMRUI Plug-ins for Brain Tumor Analysis 09.10.2014 Torino, Italien Pedrosa de Barros Nuno Miguel; Pica Alessia; Wiest Roland; Knecht Urspeter; Slotboom Johannes;
European Association of Neuro-Oncology - EANO - 2014 Poster The Effects of Bevacizumab on MR-imaging and 1H-MR-Spectroscopy (MRS) in Patients with high Grade Gliomas 09.10.2014 Torino, Italien Pica Alessia; Wiest Roland; Knecht Urspeter; Slotboom Johannes;
European Association of Neuro-Oncology - EANO- 2014 Poster Towards Computer Aided Neuroradiological Diagnostics of Brain Tumors 09.10.2014 Torino, Italien Pica Alessia; Porz Nicole Bernadette; Raabe Andreas; Slotboom Johannes; Wiest Roland;


Veranstaltungen zum Wissenstransfer

Aktiver Beitrag

Titel Art des Beitrags Titel des Artikels oder Beitrages Datum Ort Beteiligte Personen
Nacht der Forschung // Science Night of the University of Bern Performances, Ausstellungen (z.B. für Bildungsinstitute) 03.09.2014 Bern, Schweiz


Verbundene Projekte

Nummer Titel Start Förderungsinstrument
159648 RESURGE - Randomized controlled comparative phase II trial on surgery for glioblastoma recurrence 01.07.2015 Projektförderung (Abt. I-III)

Abstract

Introduction - Due to recent changes in glioma treatment, e.g. combined radiotherapy with temozolomide (RT/CT), and anti-angiogenic therapy (AAT), the traditional Macdonald criteria for tumor response evaluation are increasingly recognized to be not adequate any more, since effects of pseudo-progression and pseudo-response are not covered by these criteria. In order to overcome these problems, the Response Assessment in Neuro-Oncology (RANO) Working Group is an international effort to develop new standardized response criteria for malignant glioma, but according to their recent 2010 paper with updated response criteria, should to be regarded as ”a work-in-progress”, rather than definitive criteria. Since the formulation of the Macdonald criteria in the beginning of the nineties, many novel MR-techniques have become available, now widely used in clinical routine, that allow for the non-invasive retrieval of morphological, functional and metabolic properties of brain tumors. In the same 2010 paper of the RANO group, it was pointed out that there is an urgent need to investigate whether these contemporary state-of-the-art MR-techniques (i.e. diffusion weighted imaging, dynamic susceptibility contrast enhanced perfusion, imaging, MR-spectroscopy (MRS)) are able to fill the gap in the proper evaluation of tumor progression. The underlying neuroradiological study should be seen as a structured systematic multimodal attempt to implement these state-of-the-art MR-techniques as indicated in the 2010 RANO paper.Aims - The goal of our proposal is to identify those MRI/MRS-parameters that are complementary indicators of tumor progression status, thus investigating the added value of functional and metabolic MR-information to the currently proposed RANO criteria. To do so, this study foresees the acquisition of T1-, T2-weighted, perfusion, diffusion-weighted MR-images, as well as localized MR-spectra of de novo glioma patients before, during and after treatment by neurosurgery, radio-chemotherapy and/or anti-angiogenic therapy. We intend to investigate the tumor progression in time by determination of 82 MRI/MRS-related parameters that characterize the morphological, functional and metabolic fingerprint of the tumor during a two years follow-up period. The collected MR-data will be analysed by a special software tool that has already partially been developed in house for this purpose, a so called MRI-MRS brain tumor progression analyzer, allowing for both MRI and MRS data analysis within one application. Large parts of the application, i.e. DSCE-perfusion (including Dynamic Histogram Analysis (DHA) and image texture analysis), quantitative MRS, and image segmentation have been established yet in our institution. Algorithms supporting the MR-parameter extraction will be developed in order to automate its retrieval as much as possible, simultaneously minimizing operator dependent biases. The tumor progression analysis tool will also enable the overlay of RT-dose maps onto the MR-images, allowing for studying the exact spatial dependency of dose effects of RT on the 82 MR-parameters.Methodology - Patients with de novo gliomas will undergo routine structural MRI according to the RANO guidelines. Additionally perfusion- and diffusion-MRI, as well as MR spectroscopy will be performed on a 1.5T MR-scanner. Patients with high grade glioma (grade III or grade-IV, proven by biopsy) will be prospectively enrolled in the study. Structural and functional MRI-parameters will be determined before therapy and during a follow up of twenty four months: (a.) T1-enhancing tumor volume; (b.) peritumoral volume (derived from T2w images); (c.) perfusion parameters: cerebral blood volume CBV, tracer transfer coefficient Ktrans, dynamic histogram analysis (DHA) center- and width-parameters (d.) apparent diffusion coefficient (ADC); (e.) spectroscopic parameters: choline, creatine, N-acetyl aspartate, myo-inositol, lactate, glycine and glutamate/glutamine concentrations as well as mobile lipids and macromolecular baseline. The T1, T2, perfusion and diffusion images will be analyzed by means of texture analysis determining 11 parameters per MR- modality. A total of 82 MR-derived parameters will be put into “disease progression state vectors” p and will be examined and correlated to the clinical parameters. For 90 patients the values of these progression state vectors p will be determined at first diagnosis, together with changes in these vectors during and after therapy. In a first descriptive step, parameters originating from the different individual MR-modalities will be compared separately using longitudinal data analysis, identifying the time dependency of those parameters that significantly change during the course of the disease. It is assumed that the course of the disease can retrospectively being classified correctly into one of the extended RANO therapy response states after radio-chemotherapy (RT/CT): progressive disease (PD), complete/partial response (CR/PR), stable disease (SD), pseudo-progression (PsP), and after further progression and renewed ATT and/or CT the pseudo-response (PsR) state. All data at first diagnosis will be used to further verify the observations from our previous work on tumor classification by DHA analysis. In a second step, besides standard survival analysis (Kaplan-Meier survival curves and hazard models) advanced statistical classification methods, using support vector machines, investigating the possibilities to distinguish PD from PsP, and CR/PR from PsR from the 82-dimensional disease progression state vectors p will be performed. Potential significance - The availability of MR-derived markers that discriminate between PD and PsP or CR/PR from PsR is assumed to have a big impact on therapeutic treatment of malignant glioma: ineffective treatment could be modified at an early stage, and the error of stopping apparently non effective treatment could be prevented. The tumor analysis and classification software to analyze the MRI/MRS tumor data, developed during this study, will be made publicly available.
-