Abstract
Gliomas are the most common primary of brain tumors in
adults with Glioblastoma (GBM), a World Health Organization grade
IV glioma, being the most common and aggressive. Although a strict
combination of surgery, radiation therapy, and chemotherapy is
standard for GBM, the disease invariably leads to death over months
or years. This is in part due to the incredibly invasive nature of
the tumor, a quality that prevents its complete surgical removal
and shrouds the extent of its infiltration on conventional imaging.
However, the exciting and fast growing field of molecular imaging
is gaining popularity in glioma management and may offer
substantial insight into this elusive disease. Optical and magnetic
resonance molecular imaging techniques are allowing clinicians to
probe the biological underpinnings of GBM on the molecular level
and use this information to guide therapy and stratify patient
populations for targeted therapy. For example, 5-aminolevulinic
acid (5-ALA) is under investigation as an optical technique for
identifying the infiltrative margins of GBMs intraoperatively while
spectroscopic magnetic resonance imaging (sMRI) is allowing the
metabolic profiling of GBM tissue regions non-invasively. It is
with the application of these two techniques that this work is
primarily concerned. Here, sophisticated image analysis and
biostatistical techniques are developed and used to evaluate the
impact of adding 5-ALA and sMRI to the clinical management of
patients with GBM, from surgery to response determination.
Specifically, the efficacy of 5-ALA fluorescence-guided surgery in
tumor resection is evaluated in a cohort of GBM patients,
ex vivo 5-ALA fluorescence and objective histology
image analysis are used to validate sMRI's identification of tumor
infiltration, and metabolic abnormalities are used to judge the
feasibility of including sMRI into radiation therapy planning. In
its parts, this work describes a number of incremental technical
advances in applying molecular imaging techniques to GBM
management. As a whole, this work represents a paradigm shift in
the image-based management of patients with GBM away from
conventional clinical imaging to more quantitative molecular
imaging modalities that non-invasively describe a number of
intrinsic and substantive biological phenomena.
Table of Contents
Table of Contents. v
Table of Figures. xi
Chapter 1. xvii
Introduction. 1
1.1 Gliomas. 1
1.2 Glioblastoma. 6
1.3 Presentation, Diagnosis, and Clinical Management of GBM...
15
1.3.1 Presentation. 15
1.3.2 Diagnosis. 16
1.3.3 Therapy and Clinical Management. 17
1.3.4 Response to Therapy. 20
1.4 Imaging in GBM Management. 21
1.4.1. Important Physical Concepts in MRI. 22
1.4.2. Conventional Imaging Sequences. 40
1.4.3 MRI in GBM Standard-of-Care Therapy. 53
1.4.4 MRI in Glioma Response Assessment. 60
1.5 Molecular Imaging in Glioma. 63
1.5.1 Intraoperative Optical Imaging. 65
1.5.2 Magnetic Resonance Spectroscopy. 71
1.5.3 Spectroscopic Magnetic Resonance Imaging. 73
1.5.4 Overview of Spectroscopic Data Processing. 91
1.5.5 Spectroscopic Methods Used in this Work. 99
1.5.6 MR Spectroscopy in Glioma Management. 102
1.6 Organization of The Dissertation. 109
1.7 References. 111
Chapter 2. 144
Quantitative tumor segmentation for the evaluation of extent of GBM
resection.. 144
2.1 Author's Contribution and Acknowledgement of Reproduction..
144
2.2 Abstract. 145
2.3 Introduction. 145
2.4 Methods. 148
2.4.1 Preoperative and Postoperative Imaging. 148
2.4.2 Image Analysis. 149
2.4.3 Computer-Assisted, Manual Contouring. 149
2.4.4 Semi-automated Segmentation Method. 149
2.4.5 Statistical Methods. 153
2.5 Results. 154
2.5.1 Fuzzy3 Shows Greatest Volume Agreement with Manual
Contouring. 154
2.5.4 Fuzzy3 Algorithm Performs Well Per Diagnostic Performance
Metrics. 164
2.6 Discussion.. 164
2.7 References. 170
Chapter 3. 179
Semi-Automated Volumetric and Morphological Assessment of
Glioblastoma Resection with Fluorescence-Guided Surgery. 179
3.1 Author's Contribution and Acknowledgement of Reproduction.
179
3.2 Abstract. 180
3.3 Introduction. 180
3.4 Materials and Methods. 183
3.4.1 Patient Selection. 183
3.4.2 Fluorescence-Guided Surgery. 183
3.4.3 Image Acquisition and Analysis. 184
3.4.4 Statistical Methods. 187
3.5 Results. 188
3.5.1 Study Accrual 188
3.5.2 Primary endpoints: EOR and RTV.. 188
3.5.3 Secondary Endpoint: PFS and OS. 193
3.5.4 Adverse events (AEs). 196
3.6 Discussion.. 200
5.6.1 Limitations and Strengths. 202
3.6.2 Conclusions. 203
3.7 References. 204
Chapter 4. 211
Whole-Brain Spectroscopic MRI Biomarkers Identify Infiltrating
Margins in Glioblastoma Patients 211
4.1 Author's Contribution and Acknowledgement of Reproduction..
211
4.2 Abstract. 212
4.3 Introduction. 212
4.4 Materials and Methods. 215
4.4.1 Study Design.. 215
4.4.2 Image Acquisition and Processing. 216
4.4.4 SOX2 Immunohistochemistry. 217
4.4.5 Automated Histology Slide Analysis. 218
4.4.6 sMRI-SOX2 Analysis. 219
4.4.7 Statistical Methods. 221
4.5 Results. 221
4.5.1 sMRI shows metabolic abnormalities beyond anatomical MRI.
221
4.5.2 Integration of sMRI into neuronavigation system... 226
4.5.3 Automated histology image analysis gives objective marker of
tumor infiltration 226
4.5.4 sMRI markers exhibit significant correlations with SOX2
density. 229
4.5.5 Ex vivo tissue fluorescence correlates with sMRI markers and
SOX2 density. 231
4.5.6 Cho/NAA identifies regions at high-risk for tumor recurrence.
231
4.5.7 Cho/NAA ratio in T2-hyperintense regions correlates with PFS.
234
4.6 Discussion.. 234
4.6.1 Study Strengths and Limitations. 237
4.6.1 Conclusions. 238
4.7 References. 240
Chapter 5. 245
Impact of Integrating Whole-Brain Spectroscopic MRI into Radiation
Treatment Planning for Glioblastoma. 245
5.1 Author's Contribution and Acknowledgement of Reproduction..
245
5.2 Abstract. 246
5.3 Introduction.. 247
5.4 Methods and Materials. 248
5.4.1 Patients. 248
5.4.2 sMRI data acquisition and registration.. 249
5.4.3 Target and sMRI volume definitions. 250
5.4.4 Data Analysis, Re-planning, and Recurrence Evaluation.
251
5.5 Results. 252
5.5.1 Volumetric and Spatial Analysis. 255
5.5.2 Replanning. 259
5.5.3 Preliminary Recurrence analysis. 263
5.6 Discussion.. 263
5.6.1 Study Strengths and Limitations. 265
5.6.2 Conclusions. 266
5.7 References. 267
Chapter 6. 270
General Discussion and Future Directions. 270
6.1 General Discussion. 270
6.2 Future Directions. 276
6.2.1 sMRI for Lower Grade Glioma Targeting. 276
6.2.2 Automated Spectral Quality Analysis. 279
6.2.3 RT Target Volume Segmentation and Automatic Metabolic
Profiling. 283
6.4 References. 286
Chapter 7. 290
Some Final Thoughts. 290
About this Dissertation
Rights statement
- Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
School |
|
Department |
|
Subfield / Discipline |
|
Degree |
|
Submission |
|
Language |
|
Research Field |
|
Keyword |
|
Committee Chair / Thesis Advisor |
|
Committee Members |
|