McVerry, Ferghal (2014) Multimodal CT imaging in acute ischemic stroke. MD thesis. http://theses.gla.ac.uk/4868/ Copyright and moral rights for this thesis are retained by the author A copy can be downloaded for personal non-commercial research or study, without prior permission or charge This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the Author The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the Author When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given Glasgow Theses Service http://theses.gla.ac.uk/ [email protected] Multimodal CT Imaging in Acute Ischemic Stroke Dr Ferghal McVerry BA, MB, MRCP. A thesis which fulfills the requirements of the University of Glasgow for the degree of Doctorate of Medicine Institute of Neuroscience and Psychology University of Glasgow January 2014 2 Introduction Options for imaging in acute stroke are expanding with the potential to select therapy based on imaging targets, as well as providing additional diagnostic and prognostic information. Multimodal CT has been used to image the ischemic penumbra, infarct core, and to detect leptomeningeal collateral flow although the optimum way to image these variables is not clear. Methods In addition to a systematic literature review of imaging for leptomeningeal collaterals, Data from observational studies of acute stroke which employed multimodal CT imaging on admission and follow up was used to evaluate feasibility of acute stroke imaging with CT and MRI, Perfusion thresholds for core and ischemic penumbra, methods to quantify leptomeningeal collateral flow and sensitivity of non contrast CT for detecting infarct core pixels. Results Advanced imaging in acute stroke and at follow up was more feasible with CT compared to MRI with the possible suggestion that imaging with MRI alone could introduce a bias regarding age and clinical severity for patients entered into clinical studies Heterogeneity in grading and detecting collateral flow was found in the literature providing an opportunity to devise a novel assessment method. Well developed collaterals were associated with imaging and clinical markers for good outcome as well as some potential biomarkers including atrial fibrillation and blood fibrinogen level. Relative cerebral blood flow and delay time were found to be the best predictors on infarct core and ischemic penumbra after derivation of optimum perfusion thresholds and subsequent validation in independent patient groups. 3 Pixel based comparison of infarct core on CT perfusion and non contrast CT highlighted the lack of sensitivity of CT for detecting infarct core based on Hounsfield unit value alone. Conclusion Multimodal CT for acute stroke assessment offers the potential for measuring infarct core, ischemic penumbra and leptomeningeal collateral flow status rapidly according to novel grading scales and thresholds and provides information on tissue viability which cannot be detected on non-contrast CT. Further evaluation on the impact additional imaging should have in clinical practice is needed. 4 Table of Contents Table of Contents .......................................................................... 4 Publications and Presentations.......................................................11 Author's declaration.......................................................................13 Chapter 1. Introduction ..................................................................16 1.1 Ischemic Stroke Pathophysiology.............................................17 1.1.1 The Ischemic Cascade ....................................................17 1.1.2 Evidence for Reversible and Permanent Ischemic Damage .........20 1.1.3 Normal Response to Ischemia ...........................................21 1.1.4 Identification of Penumbra in Humans with PET .....................22 1.1.5 Factors Affecting the Fate of the Penumbra..........................24 1.2 Treatment Options in Acute Ischemic Stroke...............................27 1.3 CT Imaging for Acute Stroke..................................................31 1.3.1 Non Contrast CT...........................................................31 1.3.2 CT Angiography............................................................33 1.3.3 CT Perfusion ...............................................................37 1.4 Magnetic resonance Imaging in Stroke ......................................45 1.4.1 DWI..........................................................................45 1.4.2 FLAIR........................................................................45 1.4.3 Gradient Echo Imaging ...................................................46 1.4.4 Magnetic Resonance Angiography.......................................47 1.4.5 Perfusion Weighted Imaging.............................................47 1.5 Conclusion.......................................................................48 Chapter 2. Collateral Circulation .......................................................49 2.1 Introduction .....................................................................49 2.2 The Cerebral Blood Supply....................................................49 2.3 The Circle of Willis.............................................................52 2.4 Leptomeningeal vessels........................................................54 2.4.1 Historical Context.........................................................54 2.4.2 Collateral Circulation in Animal Models of Stroke....................55 2.4.3 Collateral Therapeutics in Animal Stroke Models.....................59 2.4.4 Collateral Therapeutics in Human Stroke..............................60 2.5 Systematic Review of Methods for Assessing Leptomeningeal Collateral Flow........................................................................................62 2.5.1 Methods.....................................................................63 2.5.2 Results......................................................................64 2.5.3 Discussion...................................................................84 2.5.4 Conclusion..................................................................90 Chapter 3. Materials and Methods ......................................................92 3.1 Introduction .....................................................................92 3.2 The Multicentre Acute Stroke Imaging Study (MASIS) .....................92 3.2.1 Data Recording and Transfer ............................................98 3.3 POSH- POst Stroke Hyperglycaemia..........................................99 3.4 Combined Studies Image Analysis...........................................103 3.4.1 Image Transfer, Anonymization and Storage.........................104 3.4.2 Image Analysis............................................................104 3.5 Conclusion......................................................................107 Chapter 4. Feasibility of Imaging Based Study Recruitment .......................109 4.1 Introduction ....................................................................109 4.2 Methods.........................................................................113 4.3 Results...........................................................................114 5 4.3.1 Acute Imaging.............................................................118 4.3.2 Subacute Imaging.........................................................121 4.3.3 Day 30 Imaging............................................................121 4.4 Discussion.......................................................................122 4.4.1 Conclusion.................................................................127 Chapter 5. Quantifying Collateral Flow Extent in Acute Ischemic Stroke........128 5.1 Introduction ....................................................................128 5.2 Quantifying Collateral Flow Using Multimodal CT Imaging. .............130 5.3 Inclusion Criteria ..............................................................131 5.3.1 Clinical Data ..............................................................132 5.3.2 Image Analysis............................................................132 5.3.3 Collateral Flow Grade ...................................................138 5.3.4 Statistical Analysis .......................................................142 5.4 Results...........................................................................142 5.4.1 Collateral Flow and Outcome ..........................................153 5.5 Discussion.......................................................................163 5.6 Conclusion......................................................................169 Chapter 6. Derivation and Evaluation of Thresholds for Core and Tissue at Risk of Infarction Using CT Perfusion...........................................................171 6.1 Introduction ....................................................................171 6.2 Materials and methods........................................................172 6.2.1 Patient selection.........................................................172 6.2.2 Image Acquisition ........................................................175 6.2.3 Image analysis ............................................................175 6.2.4 Derivation dataset image analysis.....................................177 6.2.5 Validation datasets image analysis ....................................181 6.2.6 Statistical analysis .......................................................182 6.3 Results...........................................................................183 6.4 Discussion.......................................................................193 6.5 Conclusion......................................................................197 Chapter 7. Pixel Based Identification of Infarct Core on Non-Contrast CT ......199 7.1 Introduction ....................................................................199 7.2 Materials and methods........................................................200 7.3 Results...........................................................................205 7.4 Discussion.......................................................................211 Chapter 8. Conclusion ...................................................................215 8.1 Summary........................................................................215 8.2 Clinical Importance and Future Research Considerations ...............215 8.3 Concluding remarks ...........................................................217 Appendices ................................................................................218 Gradient Echo.............................................................................219 DWI.........................................................................................220 Circle of Willis MRA ......................................................................220 Perfusion...................................................................................221 Post Contrast FLAIR ......................................................................222 List of References........................................................................223 6 List of Tables Table 2-1 Assessment Methods for Measuring Collateral Flow Using Different Modalities...................................................................................65 Table 2-2 Angiographically Defined Collaterals.......................................71 Table 2-3CT based Collateral Grading Methods CT based Collateral Grading Methods.....................................................................................79 Table 2-4MRI Based Collateral Assessments...........................................83 Table 2-5 Original and Modified TIMI and TICI Scales................................87 Table 3-1 MASIS Study Calendar.........................................................98 Table 3-2 POSH Study Calendar ........................................................103 Table 4-1 Recruitment Rates of Trials and Observational Studies of Thrombolysis using CT or MRI ...........................................................................111 Table 4-2 Screening Failures for Study Participation in MASIS.....................117 Table 4-3 Comparison of Clinical Characteristics Between Patients Imaged Using CT or MRI on Admission for MASIS......................................................120 Table 5-1 Baseline Clinical Characteristics of Subjects with Good and Reduced Collaterals.................................................................................145 Table 5-2 Collateral Status and Associated Imaging Findings on Admission .....149 Table 5-3 Univariate Logistic Regression Summary..................................153 Table 5-4 Collateral Flow and Clinical Outcome.....................................155 Table 5-5 Clinical and Imaging Factors Associated with Clinical Outcome ......158 Table 5-6 Univariate Logistic Regression Analysis Summaries for Clinical Outcome Prediction (Modified Rankin Scale ≤2 at one month.................................162 Table 6-1 Perfusion Parameter Ranges and Increments Evaluated for Tissue at Risk.........................................................................................179 Table 6-2 Perfusion Parameter Range and Increments Evaluated for Core......179 Table 6-3 Characteristics of Tissue at Risk Evaluation Group......................184 Table 6-4 Clinical Characteristics of Infarct Core Groups..........................189 Table 7-1 Sensitivity and Specificity of HU Thresholds for Detecting Infarct Core in Grey Matter ............................................................................210 Table 7-2 Sensitivity and Specificity of HU Thresholds for Detecting Infarct Core in White Matter...........................................................................210 7 List of Figures Figure 1-1 Cascade of Damaging Events in Cerebral Ischemia......................20 Figure 1-2 Consequences of Reduced Cerebral Perfusion Pressure ................22 Figure 1-3 Interaction Between CBF Threshold for Infarction and Time..........25 Figure 1-4 NCCT, CTA MIP and CTA-SI in patient with right MCA occlusion.......37 Figure 1-5 The Maximal Slopes Method.................................................40 Figure 1-6 The Impulse Residue Function..............................................41 Figure 2-1 Anatomy of the Middle Cerebral Artery...................................51 Figure 2-2 Illustration of Collateral Scoring Based on Catheter Angiography.....68 Figure 2-3 Hyperintense Vessels on FLAIR. ............................................81 Figure 3-1 MIStar Interface..............................................................107 Figure 4-1 Flowchart for Recruitment to MASIS Study ..............................116 Figure 4-2 Imaging Acquired at Each Study Timepoint..............................122 Figure 5-1 DT Summary Map with ROI Measuring Pixels with DT>2 Seconds.....134 Figure 5-2 Fusion Tool Before and After Co-registration ...........................136 Figure 5-3 Reformatted Structural Image After Co-Registration ..................137 Figure 5-4 4D Angiography Demonstrating Retrograde Collateral Flow ..........139 Figure 5-5 CTA After Application of High-Low HU Threshold ......................141 Figure 5-6 Flowchart for Inclusion in Collateral Flow Evaluation .................143 Figure 5-7 Collateral Status and NIHSS................................................146 Figure 5-8 Collateral Grade and Fibrinogen Level...................................146 Figure 5-9 Collateral Grade and Atrial Fibrillation..................................147 Figure 5-10 Collateral grade and Volume with DT ≥ 2 Seconds....................150 Figure 5-11 Collateral grade and Volume with DT≥ 4 seconds.....................150 Figure 5-12 Collateral grade and DT≥ 6 seconds.....................................151 Figure 5-13 Collateral grade and DT≥ 8 Seconds.....................................151 Figure 5-14 Collateral grade and Infarct Core Volume..............................152 Figure 5-15 Collateral Grade and Penumbra Volume................................152 Figure 5-16 Collateral Grade and Infarct Volume....................................156 Figure 5-17 Collateral Grade and Infarct Growth....................................156 Figure 5-18 Admission NIHSS and Outcome...........................................159 Figure 5-19 24hr NIHSS and Outcome..................................................159 Figure 5-20 Infarct Volume and Clinical Outcome..................................160 Figure 5-21 Proportion Penumbra Salvage and Clinical Outcome.................160 Figure 5-22 ASPECTS Score on CBV Maps and Clinical Outcome...................161 Figure 5-23 ASPECTS Score on NCCT and Clinical Outcome........................161 Figure 5-24 Collateral Flow and Clinical Outcome ..................................162 Figure 6-1 ROI Surrounding Final Infarct Volume....................................177 Figure 6-2 Penumbragram with Infarct ROI...........................................180 Figure 6-3 True & False Positive/Negative Pixels....................................181 Figure 6-4 Effect of Varying Perfusion Threshold on Sensitivity and Specificity183 Figure 6-5 Youden’s Index for Each Perfusion Threshold Evaluated for Tissue at Risk.........................................................................................185 Figure 6-6 Frequency Distribution for Optimum DT, Relative MTT and Absolute MTT Thresholds for Tissue at Risk......................................................186 Figure 6-7 Bland Altman Plot for DT >2 Seconds, Relative MTT 145 %, Relative MTT 125 % and Absolute MTT 7 Seconds for Detecting Tissue at Risk ............187 Figure 6-8 Youden’s Index for Each Perfusion Threshold Evaluated for Core ...190 Figure 6-9 Frequency Distribution for Optimum Relative (A) and Absolute (B) MTT Thresholds for Core ......................................................................191 Figure 6-10 Bland Altman Plot for Relative CBF ≤45%, Absolute CBV 2ml/100g, Absolute MTT> 8 Seconds, and Relative MTT125%, for Detecting Infarct Core..192 Figure 7-1 Core Grey & White Matter Segmentation................................202 8 Figure 7-2 ROC Curve for Presence of Grey Matter Core Pixels on CT............206 Figure 7-3 ROC curve for presence of White Matter core pixels on CT...........207 Figure 7-4Frequency Distribution for Grey Matter Cut Points .....................208 Figure 7-5 Frequency Distribution for White Matter Cut Points ...................209 9 Acknowledgments This thesis, along with my training in stroke medicine was made possible by the help of my supervisor Professor Keith W. Muir to whom I will be forever grateful. I was given every opportunity during my fellowship to ask questions, explore areas of particular interest and was given lots of encouragement and helpful advice. The skills learnt during this time are invaluable both academically and clinically, and for his role as a supervisor I owe Professor Muir a debt of gratitude. In addition I acknowledge the support of the other members of the acute stroke team based at the Southern General Hospital; Dr Tracey Baird, Dr Ian Reeves, Dr George Duncan, Dr Phil Birschel, Dr Julie McManus, Dr Fozia Nazir, and Dr Margaret Roberts from whom I learnt a great deal in acute stroke management and who welcomed me into the acute stroke team. The radiography and Nursing staff were excellent in facilitating the research projects. Dr Evelyn Teasdale and Dr Celestine Santosh were sources encouragement and set an example for work ethic and attention to detail in stroke imaging. My research colleagues Dr Niall MacDougall and Dr Krishna Dani are good friends and provided excellent insights into acute stroke research and imaging, and both made me feel welcome in Glasgow. Christine Aiken, Sandra Greenshields, Angela Welch, Sally Baird and Wilma Smith were always sources of assistance. Dr Mark Parsons was a welcoming host during my visiting fellowship to the John Hunter Hospital, Newcastle Australia and provided excellent training and insight into image analysis in stroke. I am grateful to both the Translational Medicine Research Collaboration and Chest Heart and Stroke, Scotland for providing funding for my fellowship and conference attendances. I also wish to express thanks to Professor Jean-Claude Baron and Professor I Mhairi Macrae for taking the time to examine the manuscript and for helpful suggestions on how to improve it following the viva examination. Finally I must acknowledge my family, my parents who were always at hand to help out especially when travelling over and back from Glasgow, and in