Bioelectric source characterization of acute myocardial ischemia

Despite a century of research and practice, the clinical accuracy of the electrocardiogram (ECG) to detect and localize myocardial ischemia remains less than satisfactory. Myocardial ischemia occurs when the heart does not receive adequate oxygen-rich blood to keep up with its metabolic requirements, and severe ischemia can lead to myocardial infarction and life-threatening arrhythmias. Early and accurate detection is an essential component of managing this condition. Ischemia is known to be a dynamic condition that reflects a changing imbalance between blood supply and metabolic demand so that it is natural that examination under physical stress conditions or exercise testing (ET) is in widespread clinical use. However, ET is characterized by poor sensitivity (68%) and specificity (77%), limiting its diagnostic usefulness and providing the motivation to address some gaps in our understanding of myocardial ischemia and its ECG signature. This dissertation is composed of three studies. The aim of the first study was to evaluate the conventionally held mechanisms for nontransmural ischemia using intramural electrodes to measure three-dimensional potential distributions in the ventricles of animals exposed to acute ischemia. We demonstrated that contrary to accepted dogma, the electrocar- diographic response of acute myocardial ischemia originated throughout the ventricular wall, i.e., in the subendocardium, midmyocardium, or the subepicardium, under various conditions. Our goal in the second study was to evaluate whether acute myocardial ischemia follows a similar pattern of spatial and temporal evolution as seen in myocardial infarction. Our findings show that the spatial and temporal evolution of acute ischemia is characterized by multiple distinct regions that expand in all three directions, with maximal expansion in the circumferential direction, especially in the early stages of ischemic development. Furthermore, with increased stress, these regions continue to expand and eventually merge into one another, and in the extreme become transmural. The progression of myocardial infarction, by contrast, was very quickly transmural in extent and formed a cohesive block of affected tissues. The aim of the third study was to evaluate the sensitivity of epicardial electrical markers of acute ischemia relative to direct evidence of ischemia derived from intramural electro- grams. The key finding from this study is that the epicardial T-wave is a more sensitive index of acute ischemia than epicardial ST segment changes, especially in the early stages of acute ischemia development.

BIOELECTRIC SOURCE CHARACTERIZATION OF ACUTE MYOCARDIAL ISCHEMIA by Kedar Kirtikumar Aras A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Bioengineering The University of Utah August 2015 Copyright © Kedar Kirtikumar Aras 2015 All Rights Reserved The Univ e r s i t y of Utah Graduat e School STATEMENT OF DISSERTATION APPROVAL The dissertation of Kedar Kirtikumar Aras has been approved by the following supervisory committee members: Robert S. MacLeod Alexey V. Zaitsev Scott T. Youngquist Edward W. Hsu Edward V. R. DiBella Chair Member Member Member Member 06/02/2015 Date Approved 06/02/2015 Date Approved 06/02/2015 Date Approved 06/02/2015 Date Approved 06/08/2015 Date Approved and by Patrick A. Tresco the Department of Chair of Bioengineering and by David B. Kieda, Dean of The Graduate School. ABSTRACT Despite a century of research and practice, the clinical accuracy of the electrocardiogram (ECG) to detect and localize myocardial ischemia remains less than satisfactory. Myocardial ischemia occurs when the heart does not receive adequate oxygen-rich blood to keep up with its metabolic requirements, and severe ischemia can lead to myocardial infarction and life-threatening arrhythmias. Early and accurate detection is an essential component of managing this condition. Ischemia is known to be a dynamic condition that reflects a changing imbalance between blood supply and metabolic demand so that it is natural that examination under physical stress conditions or exercise testing (ET) is in widespread clinical use. However, ET is characterized by poor sensitivity (68%) and specificity (77%), limiting its diagnostic usefulness and providing the motivation to address some gaps in our understanding of myocardial ischemia and its ECG signature. This dissertation is composed of three studies. The aim of the first study was to evaluate the conventionally held mechanisms for nontransmural ischemia using intramural electrodes to measure three-dimensional potential distributions in the ventricles of animals exposed to acute ischemia. We demonstrated that contrary to accepted dogma, the electrocar­diographic response of acute myocardial ischemia originated throughout the ventricular wall, i.e., in the subendocardium, midmyocardium, or the subepicardium, under various conditions. Our goal in the second study was to evaluate whether acute myocardial ischemia follows a similar pattern of spatial and temporal evolution as seen in myocardial infarction. Our findings show that the spatial and temporal evolution of acute ischemia is characterized by multiple distinct regions that expand in all three directions, with maximal expansion in the circumferential direction, especially in the early stages of ischemic development. Furthermore, with increased stress, these regions continue to expand and eventually merge into one another, and in the extreme become transmural. The progression of myocardial infarction, by contrast, was very quickly transmural in extent and formed a cohesive block of affected tissues. The aim of the third study was to evaluate the sensitivity of epicardial electrical markers of acute ischemia relative to direct evidence of ischemia derived from intramural electro­grams. The key finding from this study is that the epicardial T-wave is a more sensitive index of acute ischemia than epicardial ST segment changes, especially in the early stages of acute ischemia development. iv I dedicate this dissertation to my parents, Kirtikumar and Padma. I hope that this achievement will complete the dream that you had for me all those many years ago when you chose to give me the best education you could. CONTENTS A B STRA C T .................................................................................................................... iii A CK N OW L ED GM EN T S ........................................................................................ ix CH A PT E R S 1......IN T R O D U C T IO N ................................................................................................ 1 1.1 Background and Significance...................................................................................... 2 1.1.1 Myocardial Ischemia P ro file ............................................................................. 2 1.1.2 Coronary Perfusion and Myocardial Ischemia............................................... 3 1.1.3 Pathophysiology of Myocardial Ischemia ...................................................... 4 1.1.4 ECG Markers for Myocardial Ischemia .......................................................... 4 1.1.5 ECG-Based Detection of Ischemia ................................................................. 6 1.1.6 Clinical Modalities for Detecting Ischemia ................................................... 7 1.2 Study Aims ..................................................................................................................... 8 1.2.1 Aim 1 - Spatial Organization of Ischemia ...................................................... 8 1.2.2 Aim 2 - Spatio-Temporal Evolution of Ischemia .......................................... 9 1.2.3 Aim 3 - Sensitivity of Electrical Markers to Ischemia ................................. 9 1.3 Organization of the Dissertation ............................................................................... 9 2. BACKGROUND .................................................................................................. 11 2.1 Cardiac Anatomy ......................................................................................................... 11 2.1.1 Pericardium and Heart Wall ............................................................................. 11 2.1.2 Cardiac Cell Types ............................................................................................. 12 2.1.3 Cardiac Chambers................................................................................................ 14 2.1.4 Cardiac Skeleton and Cardiac Valves ............................................................. 16 2.1.5 Coronary Arteries and Veins ............................................................................. 18 2.1.6 Electrical Conduction System ........................................................................... 20 2.1.7 Comparative Cardiac Anatomy ...................................................................... 20 2.2 Cardiac Electrophysiology ........................................................................................... 21 2.2.1 Cardiomyocyte .................................................................................................... 21 2.2.2 Cardiac Cell Membrane .................................................................................... 24 2.2.3 Resting Potential ................................................................................................ 26 2.2.4 Action Potential .................................................................................................. 29 2.2.5 Excitation-Contraction Coupling .................................................................... 33 2.2.6 Gap Junctions ....................................................................................................... 35 2.2.7 Myocardial Anisotropy ...................................................................................... 36 2.2.8 Cardiac Propagation ........................................................................................... 37 2.2.9 Cardiac Activation Sequence............................................................................. 38 2.2.10 Electrocardiogram (ECG) .................................................................................. 40 2.3 Cardiovascular Physiology........................................................................................... 45 2.3.1 Cardiac Cycle ....................................................................................................... 45 2.3.2 Hemodynamics .................................................................................................... 48 2.3.3 Coronary Circulation ......................................................................................... 49 2.3.4 Integrated Control of Cardiovascular Physiology ........................................ 51 2.4 Coronary Heart Disease (CHD) and Ischemia........................................................ 53 2.4.1 Definition and Classification ............................................................................. 55 2.4.2 Metabolic Consequences .................................................................................... 57 2.4.3 Structural Consequences .................................................................................... 58 2.4.4 Ionic Consequences ............................................................................................. 58 2.4.5 Neurohumoral Consequences............................................................................. 61 2.4.6 Electrical Consequences .................................................................................... 62 2.4.7 Functional Consequences .................................................................................. 64 2.4.8 Diagnosis................................................................................................................ 65 2.4.9 T re a tm en t.............................................................................................................. 71 2.4.10 Prevention.............................................................................................................. 72 2.5 Literature Review ......................................................................................................... 73 2.5.1 Pre-1950s................................................................................................................ 74 2.5.2 1950s - 1990s ......................................................................................................... 74 2.5.3 2000s - Present .................................................................................................... 78 3. D E S IG N OF EX P E R IM EN T S ...................................................................... 79 3.1 Animal Models .............................................................................................................. 79 3.2 Experimental Setups .................................................................................................... 80 3.2.1 In Situ Setup......................................................................................................... 81 3.2.2 Isolated Heart Perfused by Support Animal................................................. 81 3.3 Study Protocols .............................................................................................................. 82 3.4 Data Acquisition ........................................................................................................... 83 3.4.1 Time Signal Acquisition .................................................................................... 83 3.4.2 Image Acquisition ................................................................................................ 85 3.4.3 Registration Data Acquisition........................................................................... 86 3.5 Data Processing.............................................................................................................. 86 3.5.1 Time Signal Processing ...................................................................................... 87 3.5.2 Image Processing (Segmentation) .................................................................... 88 3.5.3 Geometry Processing ......................................................................................... 88 3.6 Data Visualization......................................................................................................... 89 4. SPATIAL O RGA N IZA T IO N OF M YO CA RD IA L ISCH EM IA . . 91 4.1 Abstract ......................................................................................................................... 91 4.2 Introduction .................................................................................................................. 92 4.3 Methods ......................................................................................................................... 94 4.3.1 Experimental Preparation .................................................................................. 94 4.3.2 Experimental Protocols and Data Acquisition ............................................ 95 4.3.3 Postexperiment Imaging and Signal Processing .......................................... 96 4.3.4 Statistical Analysis ............................................................................................. 97 4.4 Results .............................................................................................................................. 98 4.5 Discussion ....................................................................................................................... 102 4.6 Conclusion ....................................................................................................................... 105 vii 5. S PA T IO -T EM PO RA L EV O LU TION OF A CU T E M YO CA RD IA L ISCH EM IA .................................................................................................................. 107 5.1 Abstract ......................................................................................................................... 107 5.2 Introduction .................................................................................................................. 108 5.3 Methods ......................................................................................................................... 109 5.3.1 Experimental Preparation..................................................................................109 5.3.2 Experimental Protocols and Data Acquisition............................................ 111 5.3.3 Postexperiment Imaging and Signal Processing .......................................... 111 5.3.4 Statistical Analysis............................................................................................. 112 5.4 Results.............................................................................................................................113 5.5 Discussion......................................................................................................................116 5.6 Conclusion......................................................................................................................118 6. E P IC A R D IA L S EN S IT IV ITY TO M Y O CA RD IA L ISC H EM IA 119 6.1 Abstract ......................................................................................................................... 119 6.2 Introduction .................................................................................................................. 120 6.3 Methods ......................................................................................................................... 121 6.3.1 Experimental Preparation .................................................................................. 121 6.3.2 Experimental Protocols and Data Acquisition ............................................ 122 6.3.3 Postexperiment Imaging and Signal Processing .......................................... 123 6.3.4 Statistical Analysis ............................................................................................. 123 6.4 Results .............................................................................................................................. 124 6.5 Discussion ....................................................................................................................... 127 6.6 Conclusions ..................................................................................................................... 129 7. CONCLUSIONS AND FU T U R E W O RK ...............................................130 7.1 Conclusions ..................................................................................................................... 130 7.2 Future Work .................................................................................................................. 133 7.2.1 Spatial Organization of Ischemia .................................................................... 134 7.2.2 Spatio-Temporal Evolution of Ischemia ........................................................ 136 7.2.3 Sensitivity of Electrical Markers to Ischemia ............................................... 137 7.2.4 Design of Experiments ...................................................................................... 137 R E F E R EN C E S ..............................................................................................................140 viii ACKNOWLEDGMENTS I 'd like to express my gratitude to my colleagues, friends, and family. Their unwavering support has been instrumental in completing my dissertation. First and foremost, I would like to thank my advisor, Dr. Rob Macleod, for giving me the opportunity to pursue this doctoral research. He has been my teacher, mentor, father figure, and role model. Thank you for providing an incredibly supportive learning environment and helping me become a better scientist. I could not have imagined having a better advisor for my doctoral study. I would like to thank Dr. Alexei Zaitsev, Dr. Edward Hsu, Dr. Edward DiBella, and Dr. Scott Youngquist for their guidance, patience, and suggestions for this research. In addition, I'd like to thank Dr. Andrews Michaels, Dr. Elizabeth Shiu, and Dr. Bonnie Punske for also serving on my PhD committee. I am also grateful to the other students and postdocs in our research lab including Darrell Swenson, Josh Blauer, Jess Tate, Brett Burton, Wilson Good, and Moritz Dannhauer for their time, support, and advise. This work would not have been possible without contri­butions from everyone at the Cardiovascular Research and Training Institute (CVRTI). I am grateful to Jayne Davis, Nancy Allen, Alicja Booth, Phil Ershler, and Bruce Steadman for their invaluable help in conducting the ischemia experiments. I'd also like to thank members of Dr. Ed Hsu's small animal imaging group including Osama Abdullah and Samer Merchant for their expertise in generating CT, MRI, and DT-MRI scans. I'd also like to thank the research and development (Ayla Khan, Dan White) and also the media development staff (Nathan Galli, Chems Touti ) at the Scientific Computing and Imaging (SCI) Institute for resources, including support software used in this research. Finally, I thank my family for their unyielding support and encouragement. I would like to express my eternal gratitude to my parents, Kirtikumar Narayan Aras and Padma Aras, my brother, Kaushal Aras, and my sister, Radhika Aras Kharpate, for their numerous sacrifices, and their love, patience, and support. I would also like to thank my partner, Holly Majszak, for her patient support and constant encouragement. CHAPTER 1 INTRODUCTION Coronary artery disease (CAD) is one of the leading causes of mortality in the United States, affecting over 13 million people, with half suffering from myocardial (cardiac) ischemia [1]. Cardiac ischemia is a pathological condition that occurs when the heart does not receive enough oxygen-rich blood to keep up with its metabolic requirements. Severe ischemia leads to a heart attack, cell damage, and life-threatening arrhythmias so that early and accurate detection is an essential component of managing this condition. The electrocardiogram (ECG) records the electrical activity of the heart and is used as a central tool to establish the diagnosis of myocardial ischemia. ECG evidence of myocardial ischemia includes shifts in the ST segment, the portion of the ECG between the end of the QRS complex and the beginning of the T-wave, all of which are described below in more detail. The use of ECG for ischemia diagnosis is most critical in two clinical settings and the consequences of errors in interpretation most costly. The first is the emergency room (ER), in which an ECG is often recorded in patients with symptoms of chest pain. Unfortunately, a relatively high percentage of the cases of myocardial ischemia occur without chest pain symptoms [1], a syndrome known as silent ischemia. Moreover, the ECG may also be normal or nonspecific in a patient with myocardial ischemia, leading to diagnostic errors of 30-50% [1]. The second setting is exercise stress testing (ET), which is a noninvasive tool to evaluate cardiovascular response to exercise under controlled conditions. If the heart does not receive enough oxygenated blood during the test, it may produce abnormalities in the ECG suggestive of myocardial ischemia and underlying CAD. However, ET has poor sensitivity (68%) and specificity (77%) [1]. This poor performance of the ECG for detecting ischemia provides powerful motivation to address some gaps in our understanding of myocardial ischemia and its electrocar­diographic signature. Additional motivation comes from the fact that the ECG remains the most widely available diagnostic tool that is painless, noninvasive, safe, and easy to perform. Accurate and robust characterization of the heart's electrocardiographic response 2 to acute episodes of cardiac ischemia would provide the basis for improving the accuracy of myocardial ischemia detection and result in more effective treatments. The research in this dissertation focused on advancing this goal through a series of acute ischemia studies using high-resolution mapping of cardiac potentials. We first investigated the spatial origins of acute myocardial ischemia through qualitative and quantitative analysis of different types and degrees of cardiac ischemia that we could induce in animal models. We began with an evaluation of the current dogma regarding the spatial organization and distribution of especially nontransmural ischemia and suggested an alternative formulation that better supports the available data. Next, we characterized the spatio-temporal progression of acute cardiac ischemia, again with the goal of establishing a robust description of the sequence of ischemia development over the acute time frame. The third and most clinically relevant component of the research was to extend these results to the cardiac surface with the goal of establishing realistic estimates of the degrees to which ischemic changes are detectable from the heart surface. The results from the third project provided valuable insights about information available on the cardiac surface relative to the location of nontransmural ischemia. The cumulative results from our studies have provided new understanding of the underlying bioelectric sources, their spatial organization and temporal progression, and the extent to which the cardiac surface ECG is sensitive to them. 1.1 Background and Significance 1.1.1 M y o c a rd ia l I s c h em ia P ro file Each year in the Western world, there are about 5.8 million new cases of coronary artery disease (CAD), and about 40 million individuals with prevalent CAD are alive today [2]. In the United States, CAD is the single leading cause of death in adults, accounting for one in five deaths [2]. It is estimated that over 13 million Americans have CAD, about one half of whom experience episodes of myocardial ischemia [2]. The estimated annual health care burden in the US associated with CAD is about $142.5 billion USD [2]. Myocardial ischemia can be defined as an imbalance between the supply of oxygenated blood and the oxygen requirements of the heart [3]. Myocardial ischemia resulting predominantly from increased myocardial oxygen demand in the presence of critical narrowing of a coronary artery is termed demand ischemia. By contrast, myocardial ischemia caused predominantly by reduction of coronary blood flow is termed supply ischemia. Demand ischemia occurs in most episodes of stable angina (chest pain), whereas supply ischemia is associated with 3 most episodes of unstable angina. Myocardial ischemia can also arise from a combination of both an increase in oxygen demand and a reduction in supply, i.e., supply an d demand ischemia. According to its location, cardiac ischemia is also classified as transmural ischemia if it affects the full thickness of the ventricular wall or nontransmural ischemia if the injury does not span the entire thickness of the ventricular wall. 1.1.2 C o r o n a ry P e r fu s io n a n d M y o c a rd ia l I s c h em ia Since ischemia reflects an imbalance between blood supply and metabolic demand, it is important to understand the many parameters and regulatory mechanisms that control this central homeostatic system. On the side of metabolic demand, the main parameters dictating cardiac oxygen consumption are heart rate (chronotropy), cardiac contractility (inotropy), and left ventricular (LV) wall stress. At maximal levels, heart rate can increase threefold, contractility three- to fivefold [1]. The main mechanism to bring more oxygen to the myocardium upon increased demand is to increase coronary blood flow, which is regulated to continuously and dynamically match variation in metabolism. In fact, at maximal metabolic load (e.g., maximal exercise), coronary blood flow can reach values up to three to five times those at rest [1]. This difference between values at rest and maximal levels of coronary flow represents the coronary flow reserve. What is known as "coronary autoregulation" is the most powerful mechanism by which coronary flow reserve is recruited to maintain the desired perfusion level. Autoregulation is a local mechanism, i.e., it requires no control from the brain or nervous system, that has the ability to maintain a suitable blood flow despite changes in metabolic demand and local perfusion pressure. When the blood pressure falls, arterial resistance is reduced as the small arteries and arterioles dilate. The reduction in resistance causes blood flow to remain stable despite the presence of reduced perfusion pressure. Myocardial ischemia represents a breakdown in this homeostasis and has a variety of consequences depending on the severity of the reduction in coronary blood flow, the length of the ischemic insult, the area of the myocardium served by the occluded artery, and the presence of collateral vessels from other adjacent beds within the heart. Under the most severe instances of complete occlusion of blood flow, myocardial ischemia can progress to myocardial infarction (heart attack) characterized by tissue death and, if left untreated, necrosis and scar. 4 1.1.3 P a th o p h y s io lo g y o f M y o c a rd ia l Is c h em ia While the origins of ischemia are hemodynamic, there are rapid and profound elec-trophysiological consequences, changes that often represent the most lethal clinical im­pacts of this condition. Carmeliet [4], in his exhaustive review of the electrophysiologi-cal changes during myocardial ischemia, described a series of metabolic, ultrastructural, mechanical, and electrical changes within myocardial cells and tissues that occurs during acutely compromised coronary blood flow. The myocardium becomes cyanotic because of the consumption of freely diffusible oxygen, causing decreased tissue oxygen tension. With increasing tissue hypoxia, intracellular respiration shifts from aerobic to anaerobic form. Adenosine triphosphate (ATP) stores are rapidly depleted, and characteristic metabolic changes occur in the ischemic tissue, including accumulation of tissue lactate, H+ ions, phosphate, and potassium. Ultrastructural changes consist of reduction in the size and number of glycogen granules, intracellular edema, swelling and distortion of mitochondria, and eventually margination of nuclear chromatin and relaxation of myofibrils. Diminished ATP stores and alteration of Ca++ lead to impairment of cardiac contractile function. The most important consequences of ischemia from the perspective of the ECG and this research are those related to altered action potential shape and amplitude. The electrical changes in ischemic cells include a marked elevation in resting membrane potential due to increased extracellular K+ concentration, reduced action potential (AP) amplitude and a reduced rate of AP upstroke due to decreased inward Na+ current, and shortened AP duration due to reduced inward Ca++ current. These changes do not occur uniformly in space, and the resulting variations in action potential amplitude and morphology in different parts of the heart lead to what are known as "injury currents," which ultimately cause important changes in the ECG that we describe in the next section. 1.1.4 E C G M a rk e r s fo r M y o c a rd ia l Is c h em ia ECG records the electrical activity of the heart from the body surface and can detect noninvasively the occurrence of myocardial ischemia. The waves in the ECG reflect the sequence of electrical excitation and recovery as it moves through the atria and the ven­tricles. Fig. 1.1(a) shows schematically the relationship between the activation sequence and the resulting phases/waves of the ECG. The P-wave represents the progression of atrial depolarization as it moves from the sino-atrial (SA) node across the right and then left atria. The QRS complex reflects the corresponding rapid depolarization of the right and left ventricles initiated when activation passes from the atria through the specialized conduction system of the atrio-ventricular (AV) node and the Bundle of His into the Purkinje 5 Fig. 1.1. Clinical model of ischemia: (a) Electrocardiogram (ECG), (b) ventricular ac­tion potentials (AP) under normal (blue) and ischemic (orange) conditions, (c) classical electrocardiography theory. ST segment elevation is diagnosed as ST segment elevation myocardial infarction (STEMI) or transmural ischemia, whereas ST segment depression is diagnosed as non-ST segment elevation myocardial infarction (NSTEMI) or unstable angina or nontransmural ischemia. fiber network. The T-wave represents the repolarization of the ventricles, i.e., the gradual return of ventricular action potentials to resting values. Of special importance in the setting of ischemia is the "ST segment," which connects the QRS complex and the T-wave and represents the period when the ventricles are depolarized. Under normal conditions, the ST segment is isoelectric, i.e., the action potentials all reach approximately the same value and so there are no potential differences across the myocardium and hence no current. During cardiac ischemia, regional changes in the AP morphology of ischemic cells bring about alterations in the ECG. Best documented and used for clinical diagnosis are changes in the ST segment (described in more detail below). However, there are also changes in the QRS complex that include changes in amplitude, inversion of complex polarity, loss of some 6 constituent waves, and in the case of conduction slowing, widening of the QRS complex. Abnormal repolarization during ischemia may also change the amplitude and polarity of the T-wave. The most important ECG diagnostic marker for ischemia detection remains the upward or downward shifts in the ST segment, which occur within 15-30 seconds after the onset of ischemia [5]. ST segment shifts are caused by localized electrical changes that occur in the action potentials of ischemic cells, especially shortening of duration, diminished amplitude, and elevation of resting membrane potential. Fig. 1.1(b) shows these changes schematically, with the blue line indicating a normal action potential and the orange line a typical ischemic action potential. Regional differences in membrane action potentials generate injury currents and produce deflections from the isoelectric potentials of a healthy heart. Fig. 1.1(c) shows the resulting changes in ST segments, the orientation of which depends on a set of geometric factors, both of the ischemic region and the location of the sensing electrode relative to the ischemic zone. For example, a positive deflection (upward shift) occurs if the injury current is flowing toward the recording electrode and a negative deflection (downward shift) if the injury current is oriented away from the electrode. For electrodes located over the heart, ST segment elevation (upward shift) is often thought to represent transmural ischemia, usually leading, if untreated within 1-2 hours, to an infarction (heart attack), whereas ST segment depression is considered a sign of nontransmural, typically transient ischemia. This diagnostic model for ischemia detection is based in part on the dogma that myocar­dial ischemia arises first in the subendocardium, thus generating intracellular injury current flowing towards the endocardium and away from precordial ECG electrodes. Over time or under increased stress, the ischemic region expands towards the epicardium, eventually becoming transmural, and the injury current begins to flow towards the recording electrode and thus produce ST segment elevation. 1.1.5 E C G -B a s e d D e te c tio n o f Is c h em ia The use of ECG for ischemia diagnosis is critical in two clinical settings: the emergency room (ER) and exercise testing (ET). In the ER, a 12-lead ECG with the patient at rest is often recorded in patients with symptoms suggestive of ischemia, e.g., chest pain (angina), shortness of breath, or dizziness; however, the ECG is normal in up to 50% of the patients who actually have chronic stable angina [1]. A further source of confounding error is the finding that in approximately 70% of cases of true cardiac ischemia, there are no clear symptoms of chest pain, a syndrome known as silent ischemia [1]. Exercise testing (ET) 7 is a very broadly used method to evaluate the functional reserve present in a patient with suspected CAD. It is based on the simple idea that by increasing metabolic demand through physical or pharmacological stress, it is possible to exceed the coronary flow reserve and elicit electrocardiographic symptoms of ischemia. While simple and safe to carry out in the carefully controlled conditions of the hospital, ET has poor accuracy (sensitivity of approximately 70% and specificity of only 80%) [6], which has motivated the development of other, typically slower and/or more costly approaches to ischemia diagnosis. 1.1 .6 C lin ic a l M o d a litie s fo r D e te c tin g Is c h em ia In large part because of the limited accuracy of ECG-based diagnosis, there are other clinical modalities available, both noninvasive and invasive, to establish the diagnosis of ischemia with varying degrees of accuracy. Echocardiography (Echo) is a noninvasive approach based on ultrasound to evaluate cardiac structure and function with images that have limited resolution and high noise levels but can capture two and even three dimensions as functions of time. In this way, Echo can establish diagnosis of ischemia based on regional systolic or diastolic wall motion abnormalities. Echo has a mean sensitivity of 70% and mean specificity of 80%, which is similar to the ECG. Myocardial perfusion imaging (MPI) using any one of several imaging modalities is another noninvasive approach available for ischemia diagnosis. Single-photon emission computed tomography (SPECT) MPI is performed using a scintillation camera and an intravenously injected nuclear tracer (e.g., isotope thallium-201) that over time, distributes itself through the heart in proportion to the regional myocardial perfusion, thus providing a map of blood flow to the heart. MPI has higher sensitivity (90%) and lower specificity (70%) compared to ECG and also requires tens of minutes to acquire and so cannot capture rapid changes in ischemia. Biochemical assays that detect cardiac markers such as creatine kinase (CK) and cardiac troponins T (cTnT) and I (cTnI) released into the blood by damaged cardiomyocytes are also used for diagnosis of ischemia. These markers have high sensitivity to detect cardiac injury, but they are also very slow (typically tens of minutes) and not specific for the etiology of cardiac injury, i.e., they may be the result of any number of illnesses, including diabetes, hypertrophy, etc. Perfusion can also be traced using coronary angiography, which is a marginally invasive procedure that entails passing a catheter through an artery into the heart and uses a special dye and x-rays to detect blood flows through the heart and into the myocardium. The dye highlights any restrictions or blockages in the coronary blood flow and remains the gold standard against which the accuracy of all other CAD diagnostic tools is measured. A 8 related, novel approach that eliminates the burden of ionizing radiation is angiography and perfusion imaging based on MRI [7, 8]. Conceptually similar to X-ray-based approaches, MRI perfusion imaging also has poor temporal resolution and high cost and functions poorly as a monitoring approach. 1.2 Study Aims Early and accurate detection of myocardial ischemia remains a key component of manag­ing this condition. ECG remains the most widely available diagnostic tool that is painless, noninvasive, safe, and easy to perform. Moreover, the equipment is readily available, inexpensive, and portable. However, the clinical performance of ECG in detecting and localizing the extent of cardiac ischemia remains unsatisfactory, providing the motivation to address some gaps in our understanding of myocardial ischemia and its ECG signature. The goal of this research was to characterize the electrocardiographic response of the heart during acute myocardial ischemia through a series of acute ischemia studies using high-resolution mapping of cardiac potentials. The ischemia studies included canine and swine animal models. The study protocol was designed to simulate stress testing and entailed inducing demand or supply ischemia by controlling the heart rate and coronary perfusion. During the ischemia study, electrograms were recorded from the cardiac surface as well as the intramural regions. These recorded electrical potentials were used to generate high-resolution potential maps and analyzed to determine the spatial origins and distribution of ischemia, determine the spatio-temporal progression of nontransmural ischemia, and finally evaluate the sensi­tivity of cardiac surface electrograms to variations in location and extent of nontransmural ischemia. To address these research goals, we have carried out studies focused around the following three specific aims. 1.2.1 A im 1 - S p a tia l O r g a n iz a tio n o f I s c h em ia Current literature suggests that myocardial ischemia originates from the subendocardium. However, studies done in the past have relied on measurements primarily on the epicardial and endocardial surfaces. Moreover, those relatively few reported cases of intramural measurements have been limited by sparse spatial resolution. The goal of this study was to characterize the electrocardiographic response of the heart during acute myocardial ischemia through a series of acute ischemia studies using high-resolution mapping of cardiac potentials. We investigated the spatial origins of different types and degrees of acute myocardial ischemia. Despite widespread use, the link, especially quantitatively, between 9 the direction and extent of ST segment shifts and the putative mechanisms of ischemia is equivocal and thus the cause of persistent clinical error. Based on our findings, we suggest an alternative formulation that better supports the available data and provides a better understanding of bioelectric sources during cardiac ischemia. 1.2.2 A im 2 - S p a tio -T em p o r a l E v o lu tio n o f Is c h em ia Reimer et al. [9] in their study on canine models induced irreversible ischemic injury characterized by necrosis and suggested that ischemia progresses as a transmural wavefront from the subendocardium to subepicardium. We profiled the spatio-temporal progression of acute ischemia in order to characterize the way ischemia develops over the acute time frame and determine whether it follows a similar pattern of spatial and temporal evolution as is seen in myocardial infarction. The high-resolution mapping of cardiac potentials enabled us to provide a much more refined description of the progress of nontransmural acute to transmural ischemia. 1.2.3 A im 3 - S e n s itiv ity o f E le c tr ic a l M a r k e r s to Is c h em ia Classical electrocardiographic theory suggests that ST segment elevation on the body surface ECG reveals transmural ischemia, whereas ST segment depression is the result of subendocardial ischemia [10]. However, numerous clinical and experimental studies suggest that this theory is, at best, incomplete. Some studies have suggested that ST segment depression is "reciprocal" to the presence of ST segment elevation [11]. Li et al. used a sheep model in their acute ischemia studies and showed that reducing flow in two different major coronary arteries produced similar epicardial ST segment potential distribution, even though intramural recordings showed that occlusion of different arteries resulted in distinctly different ischemic zones [12]. We evaluated the sensitivity of cardiac surface electrograms to localize acute myocardial ischemia. The goal of this study was to evaluate different electrical markers for their potential to detect the earliest phases of acute myocardial ischemia. Our findings indicate that a combination of markers may provide a more reliable index of acute ischemia. 1.3 Organization of the Dissertation The dissertation is divided into seven chapters. Chapter 1 provides an overview of the dissertation. Chapter 2 includes background on cardiac anatomy, physiology, electrophysiol­ogy, pathophysiology of myocardial ischemia, and patient evaluation tools to diagnose CAD and a review of relevant literature to appreciate the subsequent chapters. Chapter 3 provides 10 an overview of the study design including the data acquisition and processing pipeline. Chapter 4 is a journal paper that has been submitted to the Journal of Electrocardiology and presents the results from spatial organization of myocardial ischemia. Chapter 5 is a journal paper that has been submitted to the Journal of Electrocardiology and deals with the spatio-temporal progression of ischemia. Chapter 6 is a paper published in the Journal of Electrocardiology 2014 and presents the results of a sensitivity study of cardiac surface electrograms. Chapter 7 concludes the dissertation and outlines future directions for this research. CHAPTER 2 BACKGROUND This chapter provides the necessary background to appreciate the research behind the electrocardiographic characterization of acute myocardial ischemia. Accordingly, we cover elements of cardiac anatomy and physiology, include an overview of cardiac electrophysiol­ogy, describe the pathophysiology of myocardial ischemia, highlight the clinical diagnostic tools and techniques used to evaluate and treat patients with myocardial ischemia, and finally include a summary of the literature relevant to this research. 2.1 Cardiac Anatomy The heart is a hollow, cone-shaped muscular organ that serves as the blood circulatory pump for the body. It is located in the thoracic cavity medial to the lungs and behind the sternum. The heart has four chambers: the right atrium, left atrium, right ventricle, and left ventricle. The atria act as receiving chambers for blood and pass blood through one-way valves to the ventricles. They are the pumping chambers that carry blood through arteries to the rest of the body. To prevent blood from flowing backwards or regurgitating back into the heart, a system of four one-way valves is present in the heart: atrioventricular valves (tricuspid, mitral) and semilunar valves (pulmonary, aortic). We briefly cover the structural and functional anatomy of the heart below. 2.1.1 P e r ic a r d iu m a n d H e a r t W a ll The heart sits within a fluid-filled sac-like membrane known as the pericardium, as shown in Fig. 2.1. The pericardium is a type of two-layer membrane that produces fluid to lubricate the heart and prevent friction between the beating heart and the surrounding organs. The two layers of the pericardium are the outer fibrotic layer and the inner, serous layer, which, in turn, consists of parietal and inner visceral layers [1]. The fibrous parietal pericardium envelopes the heart and attaches to the great vessels and to the delicate parietal layer of the serous layer. The inner visceral layer of the serous pericardium forms the outer 12 Fig. 2.1. Structure of heart wall. Reprinted with permission from Wikimedia and Bruce Blausen. Retrieved from h ttp://commons.wikimedia.org. Used under creative commons attribution-noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). lining of the heart and great vessels. Besides lubrication, the pericardium also serves to hold the heart in position and maintain a hollow space for the heart to expand into when it is full. The heart wall contains three loosely defined concentric regions: the epicardium, midmyocardium, and endocardium. The epicardium is the outermost layer of the heart and is just another name for the visceral pericardium [1]. The myocardium is the muscular middle layer of the heart and makes up the majority of the thickness and the mass of the heart wall. The endocardium forms the inner lining of the heart. 2 .1 .2 C a rd ia c C e ll T y p e s The heart is composed of different types of cells including epicardial cells, working cardiomyocytes, smooth muscle cells, cardiac fibroblasts, endothelial cells, pacemaker cells, Purkinje fibers, and the other elements of the conduction system [13], as shown in Fig. 2.2. All these cell types play a role in the structural, biochemical, and electrical functioning 13 Fig. 2.2. Cardiac cell types. Reprinted with permission from Nature Publishing Group. M. Xin, E. Olson, and R. Bassel-Duby, "Mending broken hearts: Cardiac development as a basis for adult heart regeneration and repair," Nature Reviews Molecular Cell Biology, vol. 14, pp. 529-541, 2013. of the heart. The epicardium is derived from a cluster of mesothelial cells, called the proepicardium [1]. The proepicardium gives rise not only to the epicardium but also to epicardium-derived cells: smooth muscle cells, cardiac fibroblasts, and possibly endothelial cells [13]. The myocardium is composed of cardiomyocytes, which are the cells that generate contractile force in the heart required for pumping blood. The atrial and ventricular myocytes are rod shaped and contain a large amount of contractile proteins, organized in clearly periodic myofibrils, a large number of mitochondria, and glycogen deposits [14]. Smooth muscle cells contribute to the formation of coronary arteries, as well as the inflow and outflow vasculature [13]. Cardiac fibroblasts play an important role in maintaing normal cardiac function (e.g., synthesis and deposition of extracellular matrix, cell-cell communication with cardiomyocytes), as well as in cardiac remodeling during pathological conditions such as myocardial infarction [15]. Endothelial cells line the endocardial region, the interior lining of the blood vessels, and cardiac valves [13]. Pacemaker cells (SA and AV nodal cells) and Purkinje fibers are specialized cardiac cells that form the electrical conduction system in the heart. The SA and AV nodal cells are small and generally spindle 14 shaped with little contractile material (myofibrils) but a high density of mitochondria. Moreover, in lieu of a less-developed T-tubule system, they have many caveolae that contain a high density of receptor and channel proteins [16]. The Purkinje cells are similar in shape to contractile myocytes but are larger, have less contractile material, and contain a high concentration of glycogen deposits [14]. The pacemaker cells generate and conduct electrical impulses but due to the lack of contractile machinery do not contract like the atrial and ventricular cardiomyocytes. The cardiac chambers are shown in Fig. 2.3. The right atrium (RA) is the first cardiac structure to receive blood returning from the body. It has a prominent internal muscle ridge, the crista terminals, which is nonconductive and separates the right atrial free wall into a smooth-walled posterior region that receives the venae cavae and coronary sinus [1]. There is also a muscular anterior region lined by parallel pectinate muscles and from which 2 .1 .3 C a rd ia c C h am b e r s Aorta ventricle Septum Fig. 2.3. Cardiac chambers. Reprinted with permission from leavingcertbiology.net (John Loughlin). Retrieved from h t tp :// le a v in g c e r tb io lo g y .n e t, 2015. 15 the right atrial appendage emanates. The right atrial appendage abuts the right aortic sinus and overlies the right coronary artery [1]. The atrial septum (AS) divides the right and left atrial chambers. It is composed of interatrial and atrioventricular regions. The interatrial portion is highlighted by the fossa ovalis [1]. During the fetal development of the heart, the fossa ovalis is known as the foramen ovale and serves as an opening between the right and left atria, allowing blood to pass from the right to left atrium and into the left ventricle, effectively bypassing the lungs [1]. After birth, the opening closes as the blood flow shifts through the right ventricle into the lungs and enters the left atrium by way of pulmonary veins. The atrioventricular (AV) portion of the atrial septum is made up of major muscular and minor membranous components and separates the right atrium from the left ventricle. It corresponds roughly to the "triangle of Koch," which contains the AV node and the proximal portion of the AV (His) bundle [1]. The left atrium (LA) usually receives four pulmonary veins that connect to the posterior wall of the chamber. The atrial appendage arises anterolaterally and lies atop the proximal portion of the left circumflex coronary artery [1]. The left atrial appendage is smaller, more tortuous, and less pyramidal than its right atrial counter part and is often multilobed [1]. In contrast to the right atrial free wall, the left has no crista terminals and no pectinate muscles outside its appendage. The coronary sinus travels along the posterior wall of the left atrium. The right ventricle (RV) receives blood from the right atrium. It is composed of an inlet and trabecular and outflow segments [1]. The inlet component extends from the tricuspid annulus to the insertions of the papillary muscles. An apical trabecular zone extends inferiorly beyond the attachments of the papillary muscles toward the ventricular apex [1]. The outflow portion, also known as the conus or the infundibulum, is a smooth-walled muscular subpulmonary channel. A prominent arch-shaped muscular ridge known as the crista supraventricularis separates the tricuspid and pulmonary valves and is made up of four distinct components: parietal band, infundibular septum, septal band, and the moderator band [1]. These components encircle the main region of the right ventricle. The left ventricle (LV) is the largest and the strongest chamber of the heart and like the right ventricle is made of an inlet portion composed of the mitral valve apparatus, a subaortic outflow portion, and a finely trabeculated apical zone [1]. The upper aspect of the ventricle, the area where the mitral and aortic valves are located, is referred to as the base of the chamber. At the far end of the chamber is the ventricular apex. The LV free wall is thickest towards the base and thinnest towards the apex. The LV free wall is nearly 16 three times thicker than the right ventricular free wall [1]. The LV apex shows much less trabeculation than its counterpart on the right. LV false tendons are discrete, thin, cordlike fibromuscular structures that connect two walls, the two papillary muscles, or a papillary muscle to a wall, usually the ventricular septum [1]. The ventricular septum (VS) is a complex intracardiac partition separating the left and right ventricles. It is slanted backwards and to the right, and it also curves to the right following the course of an inverted S (moving from apex to aortic valve) [1]. The VS can also be divided into muscular and membranous portions. The upper portion, which separates the aortic vestibule from the lower part of the right atrium and the upper part of the right ventricle, is much thinner and fibrous and is termed the membranous VS [1]. The greater part of the VS is thick and muscular and constitutes the muscular VS. The basal half of the VS is smooth-walled, whereas the apical half is characterized by numerous small and irregularly arranged trabeculation [1]. 2 .1 .4 C a rd ia c S k e le to n a n d C a rd ia c V a lv e s The cardiac valves are anchored to their annuli or valve rings. These fibrous rings, at the base of the heart, join to form the fibrous skeleton of the heart. The centrally located aortic valve forms the cornerstone of the cardiac skeleton, and its fibrous extensions abut each of the other three valves. The four rings are mutually supported and held together by the right and left fibrous tissues and by the conus tendon [1]. From the right side of the aortic ring, the membranous portion of the interventricular septum extends downward to meet the muscular portion of the septum. The central fibrous body joins the aortic, mitral, and tricuspid valve annuli [1]. The aortic and pulmonary valve rings are joined together by a stout band of fibrous tissue, the tendon of the conus [1]. Other than the atrioventricular Bundle of His, the fibrous cardiac skeleton serves to electrically isolate the atria from the ventricles. The tricuspid valve allows the blood to flow from the right atrium into the right ventricle. It is so called because it is made of three cusps (leaflets) that separate to allow the blood to pass through and connect to block regurgitation of blood. The tricuspid valve is composed of five components, i.e., annulus, leaflets, commissures, chordae tendinae, and papillary muscles [1]. The anterior tricuspid leaflet is the largest and most mobile. The posterior leaflet is usually the smallest. The septal leaflet is the least mobile because of its many direct chordal attachments to the ventricular septum [1]. The commissures are cleftlike splits in the leaflet tissue that represent the sites of separation of the leaflets, as shown in Fig. 2.4. The tricuspid valve is attached on the ventricular side to tough strings called 17 Posterior Anterior Fig. 2.4. The four cardiac chamber valves including tricuspid valve, bicuspid valve, aortic valve, and pulmonary valve. Reprinted with permission from Wikimedia and OpenStax College. Retrieved from http://commons.wikimedia.org. Used under creative commons attribution-noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). the chordae tendinae. The chordae tendinae anchor and support the leaflets and keep them from folding backwards (prolapse) and allowing blood to regurgitate past them. The chordae tendinae are attached to papillary muscles, which are fingerlike projections arising from the ventricular wall. Papillary muscle contraction pulls the leaflets toward one another and thereby promotes valve closure [1]. The mitral valve allows blood to flow from the left atrium into the left ventricle. It is composed of the same five components as the tricuspid valve. Unlike the other cardiac valves, the mitral valve has only two leaflets. The anterior leaflet is large and semicircular. The posterior mitral leaflet is rectangular and usually divided into three scallops [1]. The aortic valve prevents blood from regurgitating back into the left ventricle. It is composed of three components: annulus, cusps or leaflets, and commissures. In contrast to the mitral and tricuspid valves, the aortic valve has no tensor apparatus (i.e., chordae tendinae or papillary muscles) [1]. The commissures form tall, peaked spaces between the attachment of adjacent cusps. The three half-moon shaped (semilunar) aortic cusps - left, right, and posterior - form pocketlike tissue flaps tha t are avascular [1]. The pulmonary 18 valve prevents the back flow of blood from the pulmonary trunk into the right ventricle. 2.1 .5 C o ro n a ry A r te r ie s a n d V e in s The patterns of coronary distribution are highly variable and as such the correlations between coronary vessel architecture and the regions of the heart they supply are not precise [1]. The right and left coronary arteries arise from the right and left aortic sinuses, respectively. The right coronary artery (RCA) arises nearly perpendicular from the aorta and is embedded in the adipose tissue throughout its course within the right atrioventricular groove, as shown in Fig. 2.5. The first branch of the RCA is typically the conus artery, which supplies the right ventricular outflow tract and forms an important collateral anastomosis, just below the pulmonary valve, with an analogous branch from the left anterior descending coronary artery (LAD) [1]. Among the numerous marginal branches of the RCA that supply the remainder of the RV wall, the largest branch travels along the acute margin from base to apex. The posterior descending and distal posterolateral branches of what is known as a "dominant" RCA supply the basal and middle inferior wall, basal inferior septum, right bundle branch, AV node, AV bundle, posterior portion of the left bundle branch, and Fig. 2.5. Coronary arteries and veins. Reprinted with permission from Springer. J. Lee and N. Smith, "The multi-scale modeling of coronary blood flow," Annals of Biomedical Engineering, vol. 40(11), pp. 2399-2413, 2012. 19 posteromedial mitral papillary muscle [1]. The left main coronary artery (LCA) travels a very short distance along the epicardium between the pulmonary trunk and left atrium before dividing into the anterior descending and circumflex arteries. The LAD courses within the epicardial fat of the anterior inter­ventricular groove and travels a variable distance along the inferior interventricular groove toward the cardiac base [1]. Its septal perforating branches supply the anterior septum and apical septum. The epicardial diagonal branches of the LAD supply the anterior LV free wall, part of the anterolateral mitral papillary muscle, and the medial one-third of the anterior RV free wall [1]. The left circumflex coronary artery (LCX) courses within the adipose tissue of the left atrioventricular groove and commonly terminates just beyond its large obtuse marginal branch. It supplies the lateral LV free wall and a portion of the anterolateral mitral papillary muscle [1]. The coronary veins typically run parallel to the entire course of the coronary arteries. The coronary venous circulation is composed of the coronary sinus, cardiac veins, and thespian venous systems. The great cardiac vein travels in the anterior interventricular groove beside the LAD and in the left anterior interventricular groove beside the LCX [1]. The great cardiac vein and other cardiac veins, such as the left posterior and middle cardiac veins, drain into the coronary sinus, which courses along the posteroinferior aspect of the left atrioventricular groove and empties into the right atrium [1]. Collateral channels provide blood flow between the major coronary arteries and their branches. If stenosis of an epicardial coronary artery produces a pressure gradient across such a vessel, the collateral channel can dilate with time and provide a bypass avenue for blood flow beyond the obstruction. Such functional collaterals can develop between the terminal extensions of two coronary arteries, between the side branches of two arteries, between branches of the same artery, or within the same branch [1]. These collaterals are common in the ventricular septum, ventricular apex, anterior RV free wall, anterolateral LV free wall, and along the atrial surfaces [1]. The microcirculation is composed of arterioles (10-150 ^m in diameter), capillaries (0.5­0.8 ^m), and venules (10-40 ^m). The independent vasoactivity of different-sized arterioles affects the flow of blood to the tissue locally [1]. The flow of blood through the arterioles is usually rapid, continuous, and unidirectional, whereas capillary flow can be highly variable. Oxygen and nutrients diffuse through the capillaries into surrounding tissues, whereas venules are involved in the transvascular exchange of fluid and macromolecules [1]. The veins collect blood that is eventually returned to the heart. 20 2 .1 .6 E le c tr ic a l C o n d u c tio n S y s tem The cardiac conduction system consists of the sinus node, internodal tracts, AV node, AV (His) bundle, and the right and left bundle branches, as shown in Fig. 2.6. The sinus node is located subepicardially in the terminal groove, close to the junction between the superior vena cava and right atrium, whereas the AV node is a subendocardial structure that is located within the triangle of Koch [1]. The AV (His) bundle arises from the distal portion of the AV node and travels along the ventricular septum. The right bundle branch emanates from the distal portion of the AV bundle and forms a cordlike structure that travels towards the anterior tricuspid papillary muscle. The left bundle branch represents a broad fenestrated sheet of subendocardial conduction fibers that spread along the septal surface of the left ventricle [1]. 2 .1 .7 C om p a r a tiv e C a rd ia c A n a tom y As most mechanistic research studies depend on animal models of healthy and diseased organs, it is critical to appreciate variations in structure and function between humans and the typical animal species. All mammals (including dogs, pigs, and sheep) have principally ------------------ Bachmann's Sinoatrial node------ j/y \ y bundle His bundle Purkinje fibres Fig. 2.6. Cardiac conduction. Reprinted with permission from Wikimedia and Madhero. Retrieved from http://commons.wikimedia.org. Used under creative commons attribu-tion- noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). 21 the same atrial and ventricular architecture. However, they do not have the same degree of trabeculation, which is coarser compared to adult human hearts [17]. Differences in aortic valve anatomy have also been reported in the literature relative to the shape of the leaflets [17]. Additional differences exist with respect to coronary structure and perfusion. Dogs and sheep typically have a left coronary type of supply, such that the majority of the myocardium is supplied via branches arising from the LCA. In contrasts, pigs and humans typically have a coronary supply balanced between the LCA and the RCA, and in some cases even a supply dominated by the RCA [18]. In addition, human and pig hearts have only a sparse coronary collateral network located subendocardially. In contrast, an extensive collateral network located near the epicardial surface can be seen in dogs [19]. However, with CAD and over time, collateral vessels become prevalent in humans [20]. All mammalian hearts have a similar conduction system except for differences in the transmural penetration of the Purkinje network between pigs and other species (e.g., dog and human). All Purkinje system fibers originate in subendocardium, but in pigs they penetrate more deeply than in other species and thus alter the time of activation in different depths of the ventricles [21]. 2.2 Cardiac Electrophysiology The heart can be construed as mechanical pump driven by an internal electrical system that generates the power and determines the timing necessary for the heart to eject blood effectively. Myocytes represent the functional unit of this system, and collectively myocytes form tissues that behave as an electrical syncytium [14]. In this section, we explore the basic mechanisms behind the electrical activity of the cardiomyocytes including action potentials, excitation-contraction coupling, cardiac excitation, repolarization, and automaticity. 2 .2 .1 C a rd iom y o c y te The working myocytes of the atria and ventricles are roughly rectangular in shape, 50-200 ^m long and 10-40 ^m wide; they generate force only along the long axis [22]. Individual cells often bifurcate, with the result that a single cell may connect with more than one neighboring cell. Myocytes contain a specialized cisternal system, the sarcoplasmic reticulum (SR), in which Ca++ is actively accumulated by ATP-driven transport. Most of the internal volume is devoted to a cytoskeletal lattice of contractile proteins that gives rise to a striated appearance. The contractile proteins are organized into sarcomeres, which are contractile functional units, bordered on each end by a protein matrix known as the "Z line," as shown in Fig. 2.7. 22 I band A band 1 band Fig. 2.7. Cardiomyocte. Reprinted with permission from American Society for Biochemistry and Molecular Biology. Y. Peng, Z. Gregorich, S. Valeja, H. Zhang, W. Cai, Y. Chen, et al., "Top-down proteomics reveals concerted reductions in myofilament and Z-disc protein phosphorylation after acute myocardial infarction," Molecular Cell Proteomics, vol. 13(10), pp. 2752-2764, 2014. The Z lines are primarily composed of the protein a-actinin. Within each sarcomere, there is an interdigitating lattice of thick and thin protein filaments. The thin filaments extend from the Z line for about 1 ^m toward the center of the sarcomere and are polymeric assemblies of globular subunits of protein actin [22]. The thick filaments are bipolar assemblies of the protein myosin. Myosin molecules have long a helical tails that form the backbone of thick filaments, and each has two globular head domains, which possess the ability to form cross-bridges with 23 the actin filaments. Upon binding with actin, the cross-bridges act as molecular motors responsible for contraction [22]. The region of the sarcomere in which the myosin filaments reside is known as the "A band." The area between the A bands of adjacent sarcomeres is known as the "I band." This area is bisected by the Z lines and is traversed by actin thin filaments that extend from the Z line toward the center of both sarcomeres [22]. The thin actin filaments also carry regulatory proteins tropomyosin and troponin. Tropomyosin is a double-stranded, a helical, coiled-coil protein that spans seven actin monomers [22]. Troponin is a globular protein complex with three subunits: 1) TnC, a calcium binding subunit, 2) TnI, a subunit that inhibits muscle contraction, and 3) TnT, a subunit that connects the troponin complex to tropomyosin and actin [22]. Tropomyosin molecules are aligned end to end around the helical coil of the thin filament with the troponin complex attached to each tropomyosin molecule. In relaxed muscle, tropomyosin binds to actin and impedes the binding of myosin heads to actin binding sites. However, on muscle activation and subsequent increase in myoplasmic calcium concentrations, free calcium binds to troponin, inducing a conformational change that is transmitted to tropomyosin. It shifts the position on the actin thin filament to reveal the site on actin required for strong myosin binding. Myosin can then bind to the thin filament in a manner conducive to force production. Force production occurs as follows. First, during muscle relaxation, myosin can bind to ATP and hydrolyze it, but cannot use the energy released during hydrolysis to create force because of the inhibition of its binding to the thin filament by tropomyosin and troponin [22]. Next, after calcium binding to troponin has released the inhibition of the tropomyosin-troponin complex on the thin filament, an energized myosin crossbridge can attach to the thin filament. This association with actin catalyzes the release of the products of hydrolysis (ADP and inorganic phosphate) and a concomitant conformational change of myosin head occurs while it is bound to actin [22]. This conformational change pulls the actin thin filament past the thick filament. Once this is completed, myosin can rebind ATP, which reduces the affinity of myosin for actin and allows for crossbridge detachment. The subsequent hydrolysis of ATP in turn reenergizes the myosin crossbridge and prepares it for the next force generating cycle. As long as the calcium concentration is high enough to keep tropomyosin-troponin complexes from blocking the myosin binding sites on actin, the cycle continues [22]. There is a direct connection between the overlap of thick and thin filaments and the resultant force developed by cardiac muscle cells. Sarcomere length is defined as the distance 24 between Z lines. In general, maximal isometric force can be elicited when this distance is approximately 2.2 ^m [22]. Force generation is decreased when the sarcomere length is reduced as the contractile proteins are too crowded to work optimally. Likewise, force is reduced if the sarcomere length is increased because of the decrease in the overlap of thick and thin filaments, reducing the potential for possible crossbridge formation. This association between myocyte length and amount of force that can be generated is called the length-tension relationship and plays an important role across scales to explain the response of the heart to changes in ventricular filling [22]. Sarcomeres are linked end to end into assemblies known as myofibrils, which run the length of the long axis of the cardiac cell and are also placed side to side to fill most of the internal volume of the cell. The nucleus is found on the periphery of the cell along with the sarcoplasmic reticulum (SR), which is a vesicular structure that acts as an internal calcium store [22]. Cardiac cells are connected and communicate with one another by junctions of two types. First, intercalated discs form strong mechanical bonds between myocytes that allow force to be transmitted across the myocardium. These structures are formed by the protein-protein associations at the membrane surface of the neighboring cells [22]. Second, gap junctions form an electrical connection between the cardiac cells. They provide direct electrical and chemical communication between the cytoplasmic spaces of the adjoining cells. The electrical communication facilitates the coordinated contractions of the cardiac muscle [22]. 2 .2 .2 C a rd ia c C e ll M em b r a n e While contraction occurs mostly inside a cardiac myocyte, the electrical function of that cell is largely determined by the membrane that surrounds it, hence the overwhelming focus in electrophysiology on membrane structure and function. The plasma membrane defines the exterior of the cell and controls the movement of molecules between the cytosol and the extracellular medium. It also facilitates cell-to-cell signaling and cell adhesion [22]. The basic structural component of the cell membranes is a phospholipid bilayer, formed by molecules in opposing orientation that contain a hydrophilic head and a hydrophobic tail, as shown in Fig. 2.8. The hydrocarbon chains of the phospholipids in each layer form a hydrophobic core within a lipid bilayer that has two important properties: 1) The hydrophobic core is an impermeable barrier that prevents diffusion of water soluble solutes across the membrane. Moreover, this simple barrier is modulated by the presence of membrane proteins that mediate transport of specific molecules across this otherwise impermeable bilayer [22]. 2) 25 Fig. 2.8. Plasma membrane structure and function. Reprinted with permission from Wiki­media and Mariana Ruiz (LadyofHats). Retrieved from http://commons.wikimedia.org. Used under creative commons attribution-noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). The bilayer is stable. The bilayer structure is maintained by hydrophobic van der Waals interactions between lipid chains. Even though the exterior aqueous environment can vary widely in ionic strength and pH, the bilayer has the strength to retain its characteristic architecture [22]. A typical biomembrane is assembled from phosphoglycerides (at high concentrations in cardiac cells), sphingolipids, and steroids (Cholesterol). All three classes of lipids are amphiphatic molecules with a polar head and hydrophobic tail [22]. Natural biomembranes have fluidlike consistency, which is decreased by sphingolipids and cholesterol and increased by phosphoglycerides [22]. The sarcolemma forms multiple invaginations perpendicular to the long cell axis, named "T-tubules," which serve as extensions of the cell surface and host a large density of transport proteins. The T-tubules are well developed in ventricular myocytes but poorly so in atrial myocytes [14]. The cell membrane usually contains both integral (transmembrane) and peripheral membrane proteins, which do not enter the hydrophobic core of the bilayer. Most integral proteins contain membrane-spanning hydrophobic a helices and hydrophilic domains that extend from the cytosolic and exoplasmic faces of the membrane [22]. The binding of a water soluble enzyme (e.g., phospholipase, kinase or phosphatase) to a mem­brane surface brings the enzyme close to its substrate and in some cases activates the substrate. Such interfacial binding is due to the attraction between positive charges on basic residues in the protein and negative charges on phospholipid head groups in the 26 bilayer [22]. The selective permeability of the plasma membrane allows the cell to maintain a con­stant internal environment. The cell membrane is permeable to gases, small hydrophobic molecules, and small uncharged polar molecules via passive diffusion. It is slightly permeable to water due to specialized aquaporins embedded in the membrane and urea, and essentially impermeable to ions and to large polar molecules [23]. Thus, the cell membrane, due to its ability to keep ions separated across its phospholipid bilayer, functions as an electrical capacitor. The specific capacitance is approximately constant across all cell membranes at 1^F/cm2 [14, 22]. The transport of most molecules into and out of cells requires the assistance of special­ized transport proteins (transmembrane proteins) including ATP powered pumps, channel proteins, and transporters. ATP powered pumps are ATPases that use energy to move ions or small molecules across the membrane against a chemical gradient, electrical gradient, or both [22]. This process is referred to as active transport. Channel proteins form a hydrophilic passageway and transport water or specific types of ions and hydrophilic small molecules down their concentration or electrical gradient [24]. Such protein-assisted transport is referred to as facilitated diffusion. Some ion channels are nongated (open much of the time), but most are open only in response to specific chemical or electrical signals (gated). Transporters (carriers) move a wide variety of ions and molecules across cell membranes. Three types of transporters have been identified: uniporters, antiporters, and symporters [22]. Uniporters transport a single type of molecule down its concentration gradient via facilitated diffusion (e.g., glucose, amino acids). In contrast, antiporters and symporters couple the movement of one type of ion or molecule against its concentration gradient with the movement of one or more different ions down its concentration gradi­ent [22]. These proteins, often called cotransporters, transport two solutes simultaneously using energy stored in the electrochemical gradient. This latter process is sometimes called the secondary active process [22]. 2 .2 .3 R e s tin g P o te n t ia l The movement of ions across membranes is influenced by two energetic factors: chemical gradient and electrical gradient. Consider a two-compartment system, with both compart­ments containing KCl, but in compartment 1 at a higher concentration. If the membrane allowed K + and Cl- ions to cross, they would diffuse from compartment 1 to compartment 2. If, however, the membrane were permeable only to K +, it would tend to diffuse from compartment 1 to 2, but Cl- ions would be left behind. As soon as this happens, there 27 will be a net transfer of positive charge from compartment 1 to 2 (carried by K + ions), and consequently compartment 2 will become electrically positive with respect to compartment 1. The resulting electrical gradient will push K + ions from compartment 2 to compartment 1. Very quickly an equilibrium will be established where the electric gradient will be just large enough to move K + ions to the left at the same rate as they tend to diffuse to the right due to the concentration gradient. The potential at which the electrical gradient and the concentration gradient balance each other is called the equilibrium potential or the Nernst potential for that specific ion. In this instance, the membrane potential is equal to potassium equilibrium potential, because only K+ ions can cross the membrane and thus influence the electrical gradient. The Nernst potential for a specific ion is given by E _ > ( f t l ■ ( 2 .) where R is the universal gas constant, T is the absolute temperature, z is the valance, i.e., the charge of the ion, F is the Faraday constant, [C]o is the concentration of ions outside the cell, and [C]j is the concentration of ions inside the cell. The resting membrane potential of a cell is the voltage difference between the interior and exterior of the cell under resting conditions. It is calculated as a weighted average of the equilibrium potentials of the individual ions. The size of each weight is the relative permeability (Pi) of each ion and the result is given by the Goldman-Hodgkin-Katz equation. E _ , S,N pm+ M +U + S,N PA- [a- M m " S f PM+ [M+] in + SN P a - [A - ] o j ' Neutral ions do not contribute to the membrane potential. In a cardiomyocyte, four ions- potassium (K), sodium (Na), calcium (Ca), and chloride (Cl)-contribute to the membrane potential, as shown in Fig. 2.9. Thus, the resting potential is given by e _ + e i PNa+ e i PCa++ Em _ P ---- EK+ + P -----E Na+ + -5 ------ EeC a++ +i PP-C--l --EeC l- . ((22 .33)) P total P total P total P total A central concept in cellular electrophysiology is that of "driving force," a measure of the potential energy for a particular ion. Driving force describes the impetus for an ion to move across the cell. The driving force for an ion increases or decreases depending upon how 28 Fig. 2.9. Cardiac cell ions: (a) Cardiac cell model, (b) ionic concentrations of the major ions and the corresponding equilibrium potentials. Reprinted with permission from Circulation Research. R. Shaw and Y. Rudy, "Electrophysiologic effects of acute myocardial ischemia," Circulation Research, vol. 35, pp. 256-272, 1997. close its equilibrium potential is to the membrane potential and is given by the equation Vd = Vm - Veq, (2.4) where Vm is the membrane potential and Veq is the equilibrium potential. At rest, the membrane potential of a cardiomyocyte is close to EK+ (-94 mV), as potassium has the highest permeability compared to other ions. The equilibrium potential for Na+, is +70 mV, for Ca++, it is +132 mV and for Cl- it is -90 mV [25]. The resulting cardiac membrane potential is approximately -85 mV. Thus, the driving force for K + (9 mV) is smaller than that for Cl- (5 mV), which in turn is much smaller than the driving force 29 for Na+ (-160 mV), which is smaller than that for Ca++ (-222 mV). The sign indicates the direction of the driving force. Outward (-) and inward (+). The resulting current due to a specific ion is given by Ohm's law rri. on -- Vm jy- Veq t ((22 .55)) Rm where Rm is the membrane resistance. One consequence of this behavior is that if the membrane is not permeable to the ion, there will be no ionic current generated regardless of the driving force. 2 .2 .4 A c tio n P o te n t ia l Action potentials (AP) are the result of transient changes in the cellular permeability to sodium, calcium, and potassium-charged ions capable of facilitated diffusion through ion channels. A brief increase in Na+ permeability depolarizes the cell and drives the membrane potential toward the Na+ equilibrium potential, which activates the voltage-gated calcium and potassium channels. The subsequent opening of calcium channels allows calcium to enter the myocyte and sustain the depolarized state. The opening of potassium channels allows potassium efflux from the cell and thus drives the membrane potential back toward the potassium equilibrium potential. The timing of these changes depends on isoforms of channel proteins present in each cell. For example, the most numerous ventricular myocytes, have durations of approximately 250 milliseconds, whereas sinoatrial and atrial muscle action potentials last approximately 150 milliseconds and Purkinje cells about 300 milliseconds. Another morphological feature of action potentials that differs across cell types and has direct impact on macroscopic tissue behavior is the rate of rise, also known as "Phase 0" of the action potential. Atrial and ventricular muscle and Purkinje cells have an extremely rapid Phase 0, as shown in Fig. 2.10. As a result, they also conduct excitation more rapidly than other myocyte types. Before discussing additional specific mechanisms and features, we outline the nomencla­ture and descriptions of the phases of the action potential: • P h a s e 0: Phase 0 is supported by activation of two inward (depolarizing) currents, INa and IcaL. Membrane depolarization rapidly activates these channels and with a delay of several milliseconds for INa and tens of milliseconds for ICaL inactivates them [14]. 30 Fig. 2.10. Cardiac action potentials: Different action potentials seen within the my­ocardium. Reprinted with permission from Springer. A. Amin, A. Asghari-Roodsari, and H. Tan, "Cardiac sodium channelopathies," European Journal of Physiology, vol. 460(2), pp. 223-37, 2010. P h a s e 1: As the sodium channels begin to close, an initial, brief repolarization occurs that is labeled Phase 1. This transient event is primarily supported by Ito, a K+ current that is activated and quickly inactivated by depolarization. P h a s e 2: The opening of L-type calcium channels and voltage-gated potassium channels results in a calcium influx that balances potassium efflux, which results in the positive plateau (Phase 2) of the action potential profile. This phase is perhaps the most varied across different myocytes types, lasting from just a few to several hundred milliseconds. This variation arises because the net current flow in this phase is the result of a dynamic balance of inward and outward currents. The inward currents are mainly carried by and ICaL [26]. The outward currents include the rapid (IKr) and slow (Iks) components of delayed rectifier K + currents [27]. As long as the currents in both directions are balanced, there is an approximately constant transmembrane 31 potential, but this balance is tenuous and varies across cell types and also across the regions of the heart and even in time through physiological changes, like ischemia. • P h a s e 3: Phase 3 is the fast repolarization phase. As the calcium channels close and inward currents diminish, the potassium currents begin to dominate, and full repolarization of the cells occurs. IK1 takes over during the final part of Phase 3 and returns the membrane potential back to its resting (diastolic) value [28]. • P h a s e 4: Phase 4 represents the membrane resting potential during diastole. After repolarization, the N a+ /K + ATPase extrudes the accumulated intracellular Na+ ions and pumps the extracellular K + into the cell. The Na+/Ca++ exchanger and the Ca++ ATPase also contribute to the ionic balance across the cell membrane. The expression of channel proteins is not uniform throughout even regions containing otherwise identical cell types. As a result, there are associated regional variations in action potential morphology. For example, Ito is expressed more in the subepicardial region compared to subendocardial region, and as a result, Phase 1 is more prominent in the subepicardial cardiomyocytes [29]. A lower expression of IKs in the midmyocardial layers [30] may also contribute to electrical heterogeneity across the ventricular wall. The cells of the working myocardium have stable resting potentials, i.e., the cell has to be stimulated to induce an action potential. In contrast, cells of the specialized conduction systems, such as sinoatrial and atrioventricular nodal cells, have unstable resting potentials (-60 to -40 mV), which allows them to serve as pacemakers. The mechanism of this behavior is that they lack IK1 expression [31]. Moreover, at these relatively positive resting potential values, INa cannot fully recover and so becomes inactive, so that depolarization is mainly driven by ICaL [14]. The resulting sequence of events is as follows: I f is activated by IKACh-induced [32] membrane hyerpolarization [33], which causes the membrane potential to begin depolarizing, thereby initiating Phase 4. At -60 mV, the I f inward current (funny currents) carried by Na+ begins to flow [34]. The resulting gradual rise in resting potential crosses the threshold (-50 mV) for opening T-type calcium channels and the resulting inward ICaT. A further depolarization to about -40 mV causes the inward ICaL to activate and an action potential to fire. The resulting Phase 0 depolarization is primarily induced by ICaT. The other inward currents including I f and ICaT, decline as their respective channels close. The influx of calcium is slow, resulting in a slower rate of depolarization compared to other contractile myocytes and Purkinje cells. No initial repolarization or plateau occurs, 32 so Phases 1 and 2 are said to be absent. Repolarization (Phase 3) is accomplished through the opening of voltage-gated potassium channels (IKi). During Phases 0, 1, 2 and part of Phase 3, the cell is refractory to the activation by depolarizing signals. This stability is described by the wavelength (A) as WL = E R P x CV, (2.6) where CV is the conduction velocity. This is termed the "effective refractory period" (ERP) and is attributed to the slow recovery of the inactivated Na+ channels [34]. The ERP acts as a protective mechanism and ensures cardiac electrical stability and protection against premature reexcitation. Table 2.1 shows a summary of the key currents in the cardiac action potential and their contributions to each phase, including those responsible for maintaining homeostatic ion concentrations. Membrane depolarization rapidly activates the INa and ICaL channels. As the calcium and sodium channels close, the potassium currents begin to dominate, and full repolarization of the cells occurs. The Na+/Ca++ exchanger and the Ca++ ATPase also contribute to the ionic balance across the cell membrane. Table 2.1. Cardiac ion currents for a ventricular action potential. Ion C u rre n t T ra n s p o rta tio n P h a s e /ro le Na+ INa Facilitated diffusion via ion channels Phase 0 Ca++ ICa(L) Facilitated diffusion via ion channels Phase 0-2 K + Ito1 Facilitated diffusion via ion channels Phase 1, notch K + IKs Facilitated diffusion via ion channels Phase 2-3 K + IKr Facilitated diffusion via ion channels Phase 3 K + IK1 Facilitated diffusion via ion channels Phase 3-4 Na+, Ca++ INaCa Na-Ca exchanger (cotrans­porter). 3:1 ratio favoring Na ion homeostasis + + Na K INaK Na-K ATP pump 3:2 ratio in favor of Na ion homeostasis Ca++ IpCa Ca ATP pump ion homeostasis 33 2 .2 .5 E x c i ta t io n -C o n tr a c tio n C o u p lin g Excitation contraction (EC) coupling describes the process from electrical excitation of the myocyte to its eventual contraction [35], as shown in Fig. 2.11. During the cardiac AP, Ca++ enters the cell through activated L-type calcium channels (LTCC), also called dihydropyridine receptors (DHPRs). These channels are located in the T-tubules at the sarcolemmal-SR junctions, where the SR calcium release channels, also called ryanodine receptors (RyRs), are also found [36]. RyRs are arranged in large organized arrays at the junctions between SR and sarcolemma beneath DHPR (on the surface of and in T-tubules) and form a large functional Ca release complex at the junction called the couplon [37]. SR calcium release is triggered by the Ca++-induced Ca release (CICR) mechanism mediated by IcaL [35]. Activation of one local DHPR channel can trigger the calcium release for all the RyRs at the couplon. Typically, there are 10-25 DHPR for every 100 RyR, which provides the safety margin for the couplon to function properly [35]. Even so, only a small fraction of DHPRs and RyRs in a cell or couplon needs to fire to generate measured Ca++ fluxes. The calcium release from the sarcoplasmic reticulum(SR) increases 3Na 2K 2QQ ms Fig. 2.11. Mechanism of EC coupling in a contractile myocyte. Reprinted with permission from Nature Publishing Group. D. Bers, "Cardiac excitation-contraction coupling," Nature, vol. 415, pp. 198-205, 2002. 34 the intracellular calcium concentration from about 10-4 to 10-2 mM [34]. A high load of SR Ca++ directly increases the amount of Ca++ released for a given ICaL because of the potentiating effect of high [Ca]SR on the open probability of RyRs [38]. SR Ca++ content can be increased by increasing Ca++ influx, decreasing Ca++ efflux, or enhancing Ca++ uptake into the SR [35]. The combination of calcium influx and release increases the cytosolic calcium concentration, allowing Ca++ to bind to myofilament protein troponin C, which activates the contractile machinery. The strength of contraction, which is graded, can be regulated by altering the amplitude or duration of Ca++ transient and also by altering the sensitivity of myofilaments to calcium [35]. Myofilament Ca++ sensitivity is enhanced dynamically by stretching the myofilaments (due to preload), resulting in stronger contraction. On the other hand, myofil­ament sensitivity is reduced by acidosis, elevated phosphate, and magnesium concentrations (all of which occur during ischemia), and P adrenergic activation, enhanced by caffeine and certain inotropic drugs [35]. The SR calcium release contributes to calcium-dependent inactivation of ICaL [39]. Moreover, Ca++-dependent inactivation at the cytosolic side also limits the amount of Ca++ entry during the AP. When there is high concentration of cytosolic Ca++, further influx of Ca is suppressed. Thus, SR Ca++ release and ICaL provide a local negative feedback on Ca++ influx [35]. Terminating calcium release is needed for diastolic refilling of the heart and is achieved by inactivation of RyR [35]. During relaxation, cytosolic calcium declines, allowing Ca++ to unbind from troponin. Calcium is then transported out of the cytosol primarily through the SR Ca++ ATPase. Phospholamban is an inhibitor of the SR Ca++ ATPase and its phosphorylation by cAMP-dependent or calmodulin-dependent protein kinases (PKA or CaMKII) removes this inhibition, allowing more rapid SR uptake of Ca++ and subsequent decline of [Ca]j [35]. There are also other pathways for calcium transport out of the cytosol: Na+/Ca++ exchanger, sarcolemmal calcium ATPase, and mitochondrial Ca++ uniport. In human, canine, and rabbit ventricle myocytes, SR Ca ATPase removes 70% of activator Ca++ and the Na+/Ca++ exchanger removes 28% and 1% each for the sarcolemmal Ca++ ATPase and the mitochondrial Ca++ uniporter [35]. For homeostasis, the amount of calcium influx must be equal to the calicum efflux for each beat. 35 2 .2 .6 G a p J u n c tio n s The propagation of action potentials from cell to cell occurs through specialized, non-selective large-conductance ion channels packed in arrays called gap junctions [40], as shown in Fig. 2.12. Each cardiac gap junction channel is composed of 12 connexin molecules, with 6 embedded in the cell membranes on each side of the junction. The heart expresses three connexin isoforms, Cx40, Cx43, and Cx45, which are available to form heteromeric channels [41]. Six connexins form a hemichannel called the connexon, which faces another connexon in the neighboring myocyte. In combination, they function as a complete channel enabling electrical continuity between neighboring myocytes. The gap junctions are located within the intercalated disks connecting myocytes along the ends of the cell long axis, but are also present albeit in lower density along the sides of the cells, to provide some side-to-side connections. These intercalated disks also provide the physical framework for mechanical force of contraction generated by each myocyte to be transmitted [14]. Fig. 2.12. Gap junctions: Structure of a gap junction. Reprinted with permission from Wiki-media and Mariana Ruiz (LadyofHats). Retrieved from http://commons.wikimedia.org. Used under creative commons attribution-noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). 36 2 .2 .7 M y o c a rd ia l A n is o tro p y Myocardial anisotropy refers to directionally dependent variation in properties such as conduction velocity [42] and repolarization [43]. The structural determinants of anisotropy include cell geometry, cell size, and distribution of gap junction and ion channels. More­over, since gap junctions behave as voltage- and chemically-gated ion channels, there is dynamic regulation of gap junction conductance, which contributes to transient anisotropy modulation [44]. The myocardial fiber orientation within myocardial tissue is anisotropic, as shown in Fig. 2.13. In ventricles, the fiber inclination angle ranges from 90-180 degrees of rota­tion. Moreover, the rate of rotation is slightly more rapid near ventricular walls. This three-dimensional spatial organization of myocardial fibers influences electrical propaga­tion [44] and mechanical contraction [45]. Alterations in fiber architecture can occur during pathological conditions such as cardiac infarction [46], which may predispose the heart to arrhythmias. z Fig. 2.13. Myocardial anisotropy. Reprinted with permission from Wolters Kluwer Health, Inc. D. Streeter, H. Spotnitz, D. Patel, J. Ross, and E. Sonnenblick, "Fiber orientation in canine LV during diastole and systole," Circulation Research, vol. 24, pp. 339-347, 1969. 37 2 .2 .8 C a rd ia c P r o p a g a tio n When a cardiomyocyte is depolarized to a positive membrane voltage, it creates a difference in membrane potential with respect to the neighboring myocytes, which are still at the negative resting potential. This potential difference causes the intracellular depolarizing currents to flow across the gap junctions into the unexcited cells. Thus, the positive charge moves from the depolarized cell to the resting cells. At the same time, in the extracellular space, the local current flows in the opposite direction, as the extracellular site is made negative by the flow of Na+ ions into the depolarized cell, thus forming a complete propagation circuit, as shown in Fig. 2.14. As the positive charge continues to accumulate in the resting cell, its membrane potential starts rising and when the excitation threshold is reached, triggers an action potential. The newly excited cell now serves as a source for exciting the next neighboring cell. This wave of propagation from cell to cell is called the "activation wavefront," which eventually spreads through the heart. The spread of activation wavefront continues as long as the depolarizing charge supplied by the excited cell is enough to activate the neighboring resting cells. Cardiac propagation is influenced by many factors, including the action potentials of Cell Membrane Fig. 2.14. Propagation circuit: Spread of excitation from one cardiac cell to another. 38 the individual cells, the tissue structure and connectivity, and can be captured by the value of the conduction velocity. One way to capture electrical connectivity in the myocardium is the "space constant" (A), which describes how far along the tissue current can travel before it dissipates into the extracellular space. The space constant is defined by the equation RRri - (2.7) where Rm is the specific membrane resistance per cross-sectional area and Ri is the specific axial resistance per cross-sectional area. Conduction velocity is faster along the longitudinal direction than in the transverse direction of the myocardium as the current crosses more cell borders and thus gap junctions to cover the same distance [47]. This difference in conduction velocities is referred to as "electrical anisotropy." Conduction velocity is also influenced by the amount of electrical source (higher INa density in contractile myocytes and relatively lower ICaL density in pacemaker cells) [14]. Conduction velocity is approximated using the equation K D dV V = , (2.8) V RiCm.m where K is a constant, D is the diameter of the fiber, dV. is the rate of rise of action potential, Ri is the intracellular resistivity, and Cm is the membrane capacitance. A related parameter known as the "propagation safety factor" is determined by the electrical source and electrical load ratio. Under normal conditions, the current size in a single cell is large enough to depolarize a number of connected cells [14]. Source/load mismatch, perhaps due to depressed charge (INa or ICaL) or expanded load (e.g., patchy fibrosis), can trigger arrhythmias resulting from unidirectional block (reentry) (described in more detail below). 2 .2 .9 C a rd ia c A c tiv a tio n S e q u e n c e The cardiac electrical impulse originates in the SA node and propagates through the entire heart, as shown in Fig. 2.15. Through connections from the SA node fibers, the action potentials travel outwardly into adjacent atrial muscle fibers and spread rapidly over the entire atria with conduction velocities of 0.3 m/s [22]. The SA node is also believed to be connected to the AV node via internodal pathways. The conduction of impulse across A 39 Fig. 2.15. Spread of excitation through the heart. The cardiac electrical impulse originates in the SA node and propagates through the entire myocardium. Reprinted with permission from Wikimedia and OpenStax College. Retrieved from http://comm ons.w ikim edia.org. Used under creative commons attribution-noncommercial-sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). these specialized fibers is more rapid, at about 1 m/s [22]. The electrical impulse from the SA node reaches the AV node in about 30 milliseconds [22]. In the AV node, the electrical impulse is delayed by another 90 milliseconds before an impulse enters the distal portion of the AV bundle and into the ventricles [22]. The slow conduction in the AV node is attributed to the low density of gap junctions, resulting in higher intracellular resistance to conduction. This delay allows the atria to contract following excitation and pump a small amount of blood into the ventricles. Normally, the only pathway for electrical impulses to travel from the atria into the ventricles is through the AV node due to the electrical insulation provided by the cardiac skeleton. Moreover, under normal physiological conditions, the electrical impulse cannot travel back from the ventricles to the atria. 40 The cardiac impulse exits the AV node and travels through the Bundle of His, the right and left bundle branches, and the Purkinje system for another 30 milliseconds before reaching the ventricles [22]. The Purkinje fibers have higher density of gap junctions and as result, the conduction velocity reaches 1.5-4 m/s [22]. As the Purkinje fibers penetrate into a large portion of the endocardial myocardium and become continuous with cardiac muscle, they provide for almost instantaneous propagation of cardiac impulse throughout the remainder of the ventricles. The conduction velocity of propagation within the ventricles is between 0.3 to 0.5 m/s [22]. The cardiac impulse spreads generally from the apex to the base, anterior to posterior, and from endocardium to the epicardium [48]. The transmission of cardiac impulse from the endocardium to the epicardium takes another 30 milliseconds [22] so that the total time of propagation from the SA node to the last of the ventricular muscle fibers is approximately 190 milliseconds. The penetration and spread of Purkinje fibers into the ventricular my­ocardium varies across species, affecting the ventricular activation sequence. For example, in pigs, the endocardium and the epicardium get activated more simultaneously than in other species [49]. 2 .2 .1 0 E l e c tr o c a r d io g r am (E C G ) The ECG represents the electrical activity of the heart recorded using body-surface electrodes as potential differences over time. To understand the complex relationship between the ECG and the electrical activity of heart, we first look at the potential field generated by a single cell, then a single fiber, a tissue composed of many fibers, and finally the integrated heart and torso volume conductor model. For a single cell model as shown in Fig. 2.16, the potential, 0 ( r ') at some external observation point, r ' is given by the equation 0 ( r ') = j 1 -Oi{Vd - VrM r ' ), (2.9) 4n & e where Vd and Vr denote the transmembrane potentials over the depolarized and resting regions of the cell and & and ae denote the homogeneous isotropic electrical conductivity inside and outside the cell. The term Q (r') represents the solid angle subtended by the cross-section A at the external observation point r ' . Moreover, the expression Ui(Vd - Vr) represents the equivalent current density double layer and thus a distributed current dipole source [50]. If the cell is completely depolarized or completely repolarized, the 41 Fig. 2.16. Cardiac cell model: Cell model to evaluate a theoretical external potential field. transmembrane voltage Vm over the entire closed surface, S , is a constant, and thus the potential field in the exterior region is zero. It is only when the part of the cell is depolarized and the other part is at rest that we see an external potential field. Expanding this same concept to the single fiber model, the depolarizing wave produces a current entering the extracellular space just ahead of the depolarizing wavefront. There is also a current sink just behind it. The intracellular current Ii travels down the gradient from the location of the (extracellular) sink to the (extracellular) source. Outside the fiber, a passive return current travels from the source to the sink, taking more than one path, and thus generating an external potential field throughout the external volume by passive current flow. For the observation points at some distance from the fiber, the potential field is indistinguishable from the one created by lumping the inward currents together into a single monopole - I o (sink) and all the outward currents into a single current monopole Io (source) separated by a distance d. Moreover, at distances from the monopole pair that are much greater than d, the external potential field can be approximated as one generated by a dipole (D = Iod) [14], as shown in Fig. 2.17. The external potential field can now be expressed as 42 Fig. 2.17. Cardiac fiber model: (a) Single fiber model, (b) dipole source. 1 Dcosp 4naP R 2 (2.10) where D is the magnitude of the dipole, p is the angle between the dipole direction (from sink to source) and the line from dipole location to the field point, and R is the distance from the dipole location to the field point. Expanding the concept further to many fibers or a tissue model, the first simplification is to sum the potentials generated by each fiber by viewing the collection of elementary dipoles distributed over the surface S(t) of the wavefront at time t. Another simplification step is to assume that the direction of dipole density is aligned with the direction of the propagation wavefront and has a uniform strength, thus forming a uniform double layer (UDL) [14]. The external field can then be expressed as 0e(r ; t) = Ms 4nae Q(r ; t), (2.11) where Ms is the dipole density and Q (r'; t) is the solid angle subtended by the entire wavefront S(t) from the observer point. The expression of external fields assumes a uniform medium of homogeneous conductiv­ity around the active cell, fiber, and tissue [51]. In actuality, the cardiac sources are in a volume conductor with different conductivities attributed to lungs, blood, skeletal muscle, fat, etc. Moreover, instead of an infinite medium, the volume conductor has a sharp torso-air boundary, as shown in Fig. 2.18. The distance between the cardiac current source location 43 Fig. 2.18. Cardiac tissue model: External potential field generated by distributed sources. and observation point (electrode location) plays a critical role in determining the magnitude of potentials recorded on the body surface. Since the potential field from a dipole source declines as 1/R2, the ability of the body surface ECG to measure potential fields from smaller current sources is diminished greatly [52, 14]. Moreover, the ECG measures the potential difference between two recording sites (electrodes), with one being used as a reference. However, this assumption is inaccurate for a bounded volume conductor [53]. The normal ECG waveform is shown in Fig. 2.19, and the key features include the P-wave, QRS complex, ST segment, T-wave, and the U-wave. The P-wave represents atrial depolarization. The QRS complex reflects the ventricular repolarization, and the T-wave indicates ventricular repolarization. The U-wave is believed to represent isovolumetric relaxation [14]. The P-wave has a relatively a small amplitude (60-120 ^V), which can be attributed to smaller current sources (as the atria are smaller in size with thin walls). The P-wave is followed by an interval of approximately 70 milliseconds referred to as the "PQ segment" and corresponds to atrial repolarization [54] as well as the excitation wave progressing through the AV node and the His-Purkinje system [55]. Because of the small magnitudes of current generated during this interval, it registers minor, if any, deflection on the ECG. QRS complex deflections are the result of ventricular depolarization, and since the RV and LV are excited almost simultaneously, the ECG reflects the superposition of the effects 44 QRS Complex R S QT Interval Fig. 2.19. Electrocardiogram (ECG) features including the ST segment, and the T-wave. Reprinted with permission from Wikimedia and Anthony Atkielski (Agateller). Retrieved from h ttp://commons.wikimedia.org. Used under creative commons attribution-noncom-mercial- sharealike 3.0 generic license (h ttp ://c re a tiv e c om m o n s .o rg /). of both wavefronts, which often results in electrical cancellation [56]. Moreover, since the RV wall is thinner, its excitation wavefront terminates earlier than the LV wavefront. Consequently, the early portion of the QRS complex receives contributions from both ventricles, whereas the latter portion of the QRS complex waveform is driven primarily by LV excitation. Following the QRS complex deflection, the ECG reflects a relatively quiet period rep­resented by the ST segment. The J-point is marked as the time point when the tracing changes slope sharply at the end of the S-wave [14] and is believed to represent the end of ventricular depolarization and the beginning of ventricular repolarization. As such, there is a period of overlapping potentials between regions at the end of depolarization and those at the beginning of repolarization. Moreover, this potential difference is relatively small, 45 resulting in the ECG returning to a voltage near baseline [14]. The primary reason for the isoelectric potentials during the ST segment in the normal ECG is that all action potentials have such similar plateau potentials that there is no current flow in the heart and hence no body-surface voltage. The T-wave is produced by the repolarization of ventricular myocytes and is attributed to transmural dispersion of repolarization from the epicardium (shorter APD) to endo­cardium (longer APD) [14]. Moreover, the repolarization appears to occur earlier at the base than at the apex [57]. The U-wave is a low amplitude, usually monophasic ECG feature following and often partly merged with the T-wave [14]. The origins of U-wave remain controversial, although a negative U-wave, if present, is believed to signify the presence of cardiac diseases: ischemia, hypertension, and aortic regurgitation [58]. 2.3 Cardiovascular Physiology The cardiovascular system is designed to deliver oxygen and nutrients to the cells of the body and remove carbon dioxide and metabolic wastes from the cells. It is made up of two major circulatory systems working in tandem: pulmonary circulation and systemic circulation. The low resistance pulmonary circulation constitutes the right side of the heart pumping the deoxygenated blood to the lungs through the pulmonary artery and pulmonary capillaries and then returning the oxygen-rich blood to the left atrium through the pulmonary veins. The high-resistance systemic circulation is composed of the left side of the heart pumping oxygenated blood to the rest of the body through the aorta, arteries, arterioles, and systemic capillaries and subsequently returning deoxygenated blood to the right atrium through the venules and veins. We briefly cover the key concepts of cardiovascular physiology in this section. 2.3.1 C a rd ia c Cycle The cardiac cycle describes the pressure, volume, and flow in the atria and ventricles as a function of time, as shown in Fig. 2.20. A single cycle of cardiac activity is composed of two phases: diastole and systole. Diastole constitutes the time period when the ventricles are relaxed, whereas systole represents the time during which the ventricles contract. The cardiac cycle diagram highlights the changes in aortic pressure, left ventricular pressure, left atrial pressure, and left ventricular volume and their relation to the heart sounds and the morphology of the electrocardiogram. The cardiac cycle is typically divided into five phases: atrial contraction, iso-volumetric contraction, ventricular ejection, iso-volumetric relaxation, and ventricular filling. The phases are detailed in the following: 46 Fig. 2.20. Cardiac cycle: Wiggers diagram highlighting various events of a cardiac cycle. Reprinted with permission from John Wiley and Sons. B. Smith, "Classifying processes: An essay in applied ontology," Ratio, vol. 25(4), pp. 463-488, 2012. • Atrial Contraction Phase: The atrial contraction phase is initiated by the P-wave on the ECG, reflecting the atrial activation, which causes the atria to contract. The pressure in the atrial chambers increases, forcing rapid flow of blood into the ventricles across the open atrio-ventricular valves. Following atrial contraction, the atrial pressure begins to fall, which causes the AV valves to float upwards before closure [34]. The ventricular volume is maximal at this stage and is called the end-diastolic volume (EDV), representing the ventricular preload. At resting heart rates, the atrial contraction accounts for 10% of the left ventricular filling as most of the ventricular filling happens before atrial contraction through the passive flow of blood through the open valves. However, at high heart rates, it can rise to as much as 40% of ventricular filling [34]. 47 • Iso-volumetric Contraction Phase: The iso-volumetric contraction phase coin­cides with the QRS complex on the ECG representing ventricular activation, which triggers EC coupling, ventricular contraction, and a rapid increase in intraventricular pressure corresponding to maximal dP/dt. The AV valves close as the intraventricular pressure exceeds the atrial pressure, resulting in the first heart sound (Si). During this time period, there is an increase in pressure without a change in volume as the ventricles contract, hence the iso-volumetric contraction. However, individual myocytes and fibers may contract isotonically (shortening), isometrically (no change), or eccentrically (lengthening), resulting in the heart becoming more spheroid with circumference increases and atrial base-to-apex shortening [34]. • Ventricular Ejection Phase: The ventricular ejection phase represents the ejection of blood from the ventricles. The intraventricular pressure exceeds the aortic and pulmonary pressures, causing the semilunar valves to open, and blood is rapidly ejected into the aorta from the left ventricle and into the pulmonary artery from the right ventricle. Approximately 200 milliseconds after the QRS complex, the ventricles begin to depolarize, which coincides with the T-wave on the ECG [34]. Ventricular pressure begins to fall, whereas atrial pressure begins to rise due to continued venous return from the lungs and systemic circulation. • Iso-volumetric Ventricular Relaxation Phase: The iso-volumetric ventricular relaxation coincides with the closure of the semilunar valves, which causes the second heart sound (S2). Valve closure is associated with a small back flow of blood into the ventricles characterized by the dicrotic notch [34]. Although the ventricular pressures decrease during this phase, the volumes remain constant as all the valves are closed. The volume of blood that remains in the ventricle is called the end-systolic volume (ESV). Moreover, the difference between EDV and ESV represents the stroke volume (SV). The atrial pressure continues to rise. • Ventricular Filling Phase: The ventricular filling phase represents the opening of AV valves as the atrial pressure exceeds the intraventricular pressure. Despite the inflow of blood, the intraventricular pressure continues to fall briefly as ventricles are still undergoing relaxation. Once the ventricles are completely recovered, their pressure begins to rise slowly. As the ventricles continue to fill with blood and expand, the intraventricular pressure rises. During this time period, the aortic and pulmonary 48 arterial pressure continues to fall. In normal hearts, 90% of the ventricular filling occurs by the end of this phase [34]. 2.3.2 H em o d y n am ic s Hemodynamics constitutes the physical factors that govern cardiac output and blood flow: blood flow velocity, pressure, and resistance. Cardiac output (CO) is the amount of blood pumped by the ventricles in 1 minute and is given by the equation: Cardiac Output (Q) = SV x Heart Rate (HR). Blood flow refers to the bulk flow of fluid during circulation. Blood flow velocity refers to the speed with which the blood moves during circulation and is directly related to blood flow and inversely related to the cross-sectional area, as shown in Fig. 2.21. The systolic blood pressure decreases as the blood flows from the aorta (120 mmHg) through the arteries (100 mmHg), arterioles (80 mmHg), capillaries (35 mmHg), and venules (5 mmHg), and falls to near atmospheric pressure in the veins (1 mmHg) [1]. Moreover, the oscillations of blood pressure are abolished in the arteriolar portion of the systemic Precapillary Arteriole sphincters Artery Small venule Fig. 2.21. Circulatory system. Reprinted with permission from Vascular Cell. M. Scioli, A. Bielli, G. Arcuri, A. Ferlosio, and A. Orlandi, "Ageing and microvasculature," Vascular Cell, vol. 16, pp. 6-19, 2014. © 2014 BioMed Central Ltd. 49 circulation. The arterial pulse is altered by such factors as heart rate (increased diastolic pressure), stroke volume (increased systolic pressure), and arterial compliance (decreased systolic pressure) [1] Resistance (R) to blood flow is affected by vessel radius (r) and fluid viscosity (n) such that R is given by Poiseuille's law R = 8n L /n r 4. Thus, when the vessel radius is halved, resistance increases by a factor of 16. Viscosity is affected by temperature, red blood cell mass, and protein concentration [1]. 2.3.3 C o ro n a ry C irc u la tio n Coronary circulation is of special interest in the setting of this research as it is closely tied to ischemia. The energy requirements of the heart include basal metabolism (20%), electrical activation (5%), and mechanical contraction (75%). The major determinants of local metabolic rate and oxygen delivery are shown in Fig. 2.22. Myocardial oxygen demand is directly related to heart rate, wall tension, and contractile state. Wall tension is related to ventricular volume (preload) and ventricular pressure. Contractility is related to the maximal rate of increase in systolic pressure. There is also a transmural gradient of wall tension, with greater tension in the subendocardium relative to subepicardium. Moreover, the basal oxygen demand is believed to be higher in the subendocardial region [59]. Major oxygen supply determinants include local perfusion and the amount of oxygen extraction from the blood. Moreover, oxygen extraction is influenced by oxygen content in the arterial blood, oxygen diffusion gradient, and blood flow velocity through the capillaries. At rest, oxygen extraction is near 75% and can rise to 90% under Fig. 2.22. Myocardial oxygen determinants: Major determinants of oxygen demand and supply to the heart. 50 stress conditions. Since the oxygen extraction is already high, any increase in metabolic demand must be met by a corresponding increase in coronary blood flow. The heart relies primarily on continual resynthesis of mitochondrial ATP for its energy supply. At rest, the energy substrates include oxidation of free fatty acids (60-70%) and metabolism of carbohydrates (30%) [1]. During exercise, oxidation of free fatty acids is inhibited, and carbohydrates become the primary mode for ATP production. The heart has reserves of high-energy phosphates (HEP) and phospocreatine (PCr), which can be used to generate ATP, but they can buffer the ATP levels for only a short period (15-30 seconds) in the absence of coronary blood flow [1]. Coronary flow is regulated variously by metabolic, mechanical, autonomic, and endothelial controls. "Autoregulation" refers to the phenomenon in which a change in aortic pressure is met by the adjustment of coronary vascular resistance to ensure constant blood flow and is effective only in the range of 60-140 mmHg of systemic arterial pressure [1]. "Autoregulatory reserve" is a related term that refers to the maximal degree of vasodilation in the coronary perfusion bed and determines the perfusion pressures over which coronary flow is maintained. Thus, if a vascular bed is already maximally vasodilated, the capacity to autoregulate in the face of a further decrease in perfusion pressure is severely limited. Autoregulation is largely a local phenomenon mediated by myogenic tone (a change in tone in response to changes in pressure and flow) and metabolic means (washout of vasoactive metabolites such as adenosine) [1]. Most coronary flow occurs during diastole, when the perfusion pressure is greater than the intramyocardial pressure [60]. During ventricular systole, the intramyocardial tissue pressure gradient (decreasing towards the epicardium), through extravascular compression, causes a parallel gradient in systolic coronary vascular resistance [61]. Moreover, as the intramyocardial tissue pressure is greater than the coronary driving pressure, the coronary flow decreases and may cease in the subendocardium during systole. During ventricular diastole, the intramyocardial tissue pressure is similar across the myocardium, and the gra­dient in diastolic coronary vascular resistance exists (decreasing towards the endocardium) due to dilation of subendocardial coronary vessels [62]. The result is greater diastolic flow in the subendocardium over subepicardium. Thus, the subepicardium receives 90% of flow during systole and 10% during diastole. The midmyocardium receives an equal amount of blood flow in systole and diastole while the subendocardium is perfused primarily during diastole [1]. Moreover, the ratio of subendocardial flow to subepicardial flow is approximately 1.1:1, as the intrinsic coronary vascular resistance in the subendocardial arteries is lower. 51 2.3.4 In te g r a te d C o n tro l o f C a rd io v a s c u la r P h y sio lo g y The autonomic nervous system (ANS) influences the smooth muscle tone of coronary arteries to modulate coronary flow, as shown in Fig. 2.23. The epicardial coronary arter­ies have both a-adrenergic (vasoconstriction) and ^-adrenergic receptors (vasodilation) to mediate coronary flow. The a-adrenergic receptors mediate the release of norepinephrine (NE) during sympathetic stimulation, which can cause vasoconstriction of arteries. The stimulation of ^2-adrenergic receptors by circulating catecholamines results in coronary vasodilation. That said, the role of ANS in coronary flow regulation is often overshadowed by metabolic and mechanical factors [1]. An endothelium-dervied relaxing factor (EDRF) such as nitrous oxide (NO) is a powerful vasodilator and its release is triggered by a number of stress signals (e.g., hypoxia) and distending forces in the vascular wall [1]. The ANS influences vasomotor tone and cardiac function through its sympathetic (SNS) and parasympathetic (PNS) divisions to achieve an integrated and dynamic response, which entails integrating inputs from the cerebral cortex and specialized sensors (e.g., mechanoreceptors) into brain regions (hypothalamus, pons, medulla) and transmitting Fig. 2.23. Integrated control of cardiovascular physiology: Schematic of the CNS control. Reprinted with permission from American College of Cardiology Foundation. D. Linz, C. Ukena, F. Mahfoud, H. Neuberger, and M. Bohm, "Atrial autonomic innervation: A target for interventional anti arrhythmic therapy?" Elsevier (2014). Journal of American College of Cardiology, vol. 63(3), pp. 215-224, 2014. 52 efferent nerve impulses to the periphery over sympathetic and parasympathetic systems [1]. The parasympathetic system plays a relatively minor role in arterial pressure mediation, but an important role in modulating heart rate. The heart rate is maintained at a constant value (typically approximately 70 bpm in humans) via a continuous firing of the vagus nerve. Parasympathetic discharge increases potassium ion permeability in cardiac myocytes, thus increasing the threshold for depolarization to occur spontaneously in the SA node. As a result, the heart rate declines (negative chronotropic effect) and, with it, the cardiac output. This autonomic neural input predominates during sleep and other sedentary states, eliciting an increase in cardiac cycle time and therefore enabling the heart to expend less energy. Parasympathetic stimulation also slows conduction velocity (negative dromotropic effect) and decreases contractility (negative inotropic effect) although its impact on heart rate is much more potent. Increased sympathetic activity increases heart rate by releasing norepinephrine, which causes a rise in the SA node depolarization rate (positive chronotropic effect). The increase in heart rate is due to an increase in activated calcium channels, which increases the speed at which depolarization occurs. This increased sympathetic outflow can be initiated by a large array of internal and external stimuli: exercise, increase in body temperature, trauma, or stress. Cardiac output is also increased with increased SNS activity by increasing stroke volume. The underlying mechanism for this increased SV is the enhanced cardiac myocyte contractility. Myocytes usually increase in length in proportion to their preload, and because they become more elongated, they also have the capability to shorten over a greater distance. This increased shortening leads to enhanced strength of contraction. The sympathetic excitation also facilitates a larger and more rapid calcium influx, which further augments the degree of overall contraction during systole. The time necessary for the heart to contract and relax fully decreases under sympathetic stimulation due to a larger proportion of the cardiac cycle that is made available for filling. Although an increase in heart rate makes the total duration of the cardiac cycle shorter, there is a decrease in the amount of time necessary for contraction of the heart. Thus, the heart is relaxed for a greater portion of the cycle, enabling enhanced filling to provide a greater volume of blood ejected for each contraction. Epinephrine also increases calcium influx and uptake by SR by stimulating the P receptors and thus acts a positive inotropic agent. The stimulation and withdrawal of SNS activity are also powerful regulators for periph­eral coronary circulation. Changes in blood supply to local tissue or organs are mediated by changes in resistance of peripheral vasculature in response to the need for blood elsewhere. 53 Control of the vascular peripheral resistance is achieved by varying the firing frequency with these sympathetic fibers. More specifically, postganglionic fibers release norepinephrine, which binds to ai-adrenergic receptors within the smooth muscle cells in the arteriolar walls and causes an overall decrease in the diameter of arterioles (vasoconstriction). In contrast, lowering the basal tonic activity causes vasodilation since less norepinephrine is available for binding, causing the smooth muscle cells to relax. Sympathetic stimulation of a subset of fibers originating in the cerebral motor cortex releases epinephrine (instead of norepinephrine), which s