Imperial College London
Novel Photoreceptor Cells, Pupillometry and Electrodiagnosis in Orbital, Vitreo-retinal and Refractive Disorders
Farhan Husain Zaidi
A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy of the University of London and the Diploma of Imperial College
Faculty of Medicine Imperial College London, the University of London
Course: Clinical Medicine Research Registered Subject Fields: Vision Science, Ophthalmology and Surgery; Specific Areas:- Vision Science: primarily ganglion cells in the eye and orbit (physiology, disease, psychophysics, anatomy); visual pathways; comea/lens Ophthalmology: primarily clinical subspecialties of oculoplastic/orbital surgery and medical/surgical retina; cataract/refractive, glaucoma, general Surgery: primarily orbital, vitreoretinal and facial plastic particularly ocular adnexal surgery; refractive surgery including cataract and cornea
Full-time postgraduate student Departments of Ophthalmology, 2002, and Visual Neuroscience, 2002-5, after which full-time research concluded, formal thesis writing commenced, with submission in 2007; examined January 2008 Campuses: St Mary's and the Western Eye Hospitals, Hammersmith Hospitals and South Kensington Supervisors and Examiners
Course Supervisors
Dr MJ Moseley. Hon. Senior Lecturer in Ophthalmology, Imperial College London; Senior Lecturer in Ophthalmology, City University; formerly Senior Lecturer in Ophthalmology, Imperial College London. Prof AR Fielder. Professor and Head of Dept. of Ophthalmology, City University; Hon. Consultant Ophthalmologist St Mary's and the Western Eye Hospitals; formerly Professor and Head of Dept. of Ophthalmology, Imperial College London. Prof MW Hankins. Professor of Visual Neuroscience, Wellcome Trust Centre for Human Genetics, University of Oxford; Visiting Professor of Visual Neuroscience, Imperial College London; formerly Professor of Visual Neuroscience, Imperial College London.
Thesis Examiners
The subject fields and the thesis in particular were examined by internal and external examiners for the Doctorate of Philosophy of the University of London and the Diploma of Imperial College by viva voce, 1400 to 1800, at the Western Eye Hospital, Imperial College Healthcare NHS Trust, London, on January 14, 2008. This was successfully passed pending routine minor thesis alterations which are herein complied with as checked and approved by the examiners.
Prof PA Bloom (internal examiner). Consultant Ophthalmic Surgeon, St Mary's and the Western Eye Hospitals, London; Service Director, the Western Eye Hospital, Imperial College Healthcare NHS Trust; Hon. Senior Lecturer in Ophthalmology, Imperial College London; Hon. Prof. Middlesex University, London. Relevant special interests: clinical ophthalmology and surgery; doctorate supervision and publications in peer-reviewed journals e.g. electrophysiology and pupillometry in glaucoma and retinal diseases; ganglion cell function; clinical studies in cataract /refractive surgery. Dr LG Ripley (external examiner). Reader in Electronic Engineering, University of Sussex. Relevant special interests: biomedical engineering applied to basic and clinical vision science including pupillometry (invention of the Sussex pupillometer), theory and testing of colour vision (invention of the automated tritan discrimination test); doctorate supervision and publications in peer-reviewed journals e.g. colour vision and contrast sensitivity in orbital surgery (decompression of compressive optic neuropathy in thyroid eye disease) and vitreo-retinal disease (macular hole surgery and screening of diabetic retinopathy).
Mid-Course Examiners
Prof RG Foster. Professor of Circadian Neuroscience and Head of Departments of Ophthalmology and of Circadian and Visual Neuroscience, Wellcome Trust Centre for Human Genetics, University of Oxford; Visiting Professor of Biology, Imperial College London; formerly Professor of Molecular Neuroscience, Imperial College London. Relevant special interests with doctorate supervision and peer-reviewed publications in circadian biology and non-rod non-cone photoreception, including its discovery in mammals. Dr K Gregory-Evans. Reader in Molecular Ophthalmology, Imperial College London; Hon. Consultant Ophthalmologist, St Mary's and the Western Eye Hospitals. Relevant special interests with doctorate supervision and peer-reviewed publications in genetic and clinical aspects of retinal diseases; retinal electrophysiology; outer photoreceptor dystrophies.
2 Acknowledgements
I acknowledge the award of a Lindo Wing Fellowship from St Mary's and the Western Eye Hospitals to help support presentation of part of this work at international conferences. I am grateful for the support given during my research at Imperial College London by the Department of Ophthalmology (Head, Prof AR Fielder), later merging to become the Department of Visual Neuroscience (Head, Prof C Kennard), and especially by my supervisors — above all to Dr Merrick Moseley as my principle supervisor for the entire duration of this work, to Professor Alistair Fielder of City University (at the time of Imperial College London) who sponsored my work while he was at Imperial College London and continued to support it thereafter, and to Professor Mark Hankins of the University of Oxford (at the time of Imperial College London) who chaired my mid-course examination and later joined in the administrative supervision of this work.
Professor Russell Foster of the University of Oxford (at the time of Imperial College London) and Dr Kevin Gregory-Evans of Imperial College London kindly provided valued suggestions as the mid-course examiners and access to additional facilities in their laboratories. Certain analyses of data were performed using specific computer software in conjunction with Dr. Stuart Peirson and Dr. Katharina Wulff of the University of Oxford, both previously of Imperial College London, and who are specially acknowledged in the relevant chapters.
Clinical collaborators at St Mary's and the Western Eye Hospitals very kindly allowed access to their patients.
There is no conflict of interest declared by the author The work presented in this thesis is my own
Farhan Husain Zaidi
3 Dedication
I am grateful to my family fo7 their support over a period of full-time research as a clinician, especially to my late father for his encouragement during this time. For their patience and encouragement I dedicate this thesis to them and to the light that guides.
4 Abstract
This thesis presents work that quantifies visual potential and associated visual functions in diseased retina and intra-orbital optic nerve. The aim of the first half of this thesis is to employ pupillometry to differentiate topographic regions of diseased eyes and optic nerves in patients with model diseases which offer potential for clinical application. These disease processes are compressive and traumatic orbital lesions affecting the optic nerve, those of the vitreo-retina, and, to a lesser degree, those affecting the refractive interfaces of the eye. Many of these findings are explained by putative non- rod non-cone photoreception in humans. The second half of this thesis investigates the possible existence of this novel receptor in healthy humans and in subjects with disease using a variety of techniques. The final strand to this thesis explores the hitherto unknown visual potential of the receptor. The principal experimental methods are pupillometry, selectively complemented by electrophysiology. In studying the putative photoreceptor, additionally, behavioural experiments are also employed — actigraphy to study circadian rhythms and psychophysical tests to study visual potential. Imaging and other clinical tests are also used. The general format to test most hypotheses comprises the measurement of pupil waveform parameters such as amplitude and re-constriction in assays based on stimulus paradigms designed to elicit spatial and temporal photoreceptor responses. Non-pupillometric tests utilise standard protocols (electroretinogram, 2-alternative forced choice test, visual acuity, visual field, wrist actigraphy, validated sleep questionnaires, optical coherence tomography, computerised tomography (CT), magnetic resonance imaging (MRI), and fundus fluorescein angiography). This thesis finds that full-field pupil luminance and colour responses are useful in determining macula status in subretinal neovascular membranes (SRNs), retinal detachments, and in diagnosing compressive and traumatic optic neuropathies considered for orbital decompression/surgical exploration. Inter-eye comparison (intra-subject) of pupil responses between healthy and diseased eyes is useful in diagnosing disease using the pupil waveform. The effects of stimulus duration, intensity and photopigment bleaching (adapting light, photostress) are also considered. Cataract is found to cause a variety of changes in pupil perimetric sensitivity, attributed to index myopia, chromatic aberration, and scatter. Study of achromatic and chromatic pupil gratings responses in orbital diseases, proliferative vitreoretinopathy, and amblyopia, parallel the visual responses expected in these conditions and are shown to be cortically-influenced. In this thesis a new index of pupil escape-type events is described, the pupil evasion ratio. Its spectral profile suggests rod, cone, and non-rod non-cone activity from a novel fifth photoreceptor. An assay for the putative photoreceptor is developed and incorporates sustained blue light which elicits a previously un-described double pupil constriction in normal subjects. Subsequent study of photoreceptor dystrophies enables direct identification of the novel photoreceptor in humans using a pathological model for rod and cone knock-out. The novel receptor is shown to localise anatomically to the ganglion cell layer, with a peak spectral sensitivity of approximately 480 nm and an action spectrum matching that for the melanopsin molecule. Though previously described in some non- human mammals, this thesis shows that the novel receptor also drives pupil constriction and mediates circadian photoentrainment in humans as well. It is additionally shown to mediate sight, a new role that had hitherto been un-attributed to it any organism. The novel receptor's role in the pupil's reactions to darkness are documented.
5 Table of Contents
Page
Title Page 1
Supervisors and Examiners 2
Acknowledgements 3
Dedication 4
Abstract 5
Table of Contents 6
Lists of Tables, Figures and Inserts 25
Standard Abbreviations 34
6 Chapter One Introduction, Overview, Aims and Literature Survey 35
1.1 Introduction 35
1.2 Overview 35
1.3 Aims 36
1.4 Survey of the Literature 37 1.4.1 Perception and Methods for Studying Visual Perception 37 1.4.2 The Pupil and the Integrity of the Visual System 40 1.4.3 Ancient Literature 41 1.4.3.1 Cataract Surgery and the Pupil 41 1.4.3.2 Discarding Platonic Views: Light Recognised as the Visual and Pupillary Stimulus by the Imam Jafar ibn Mohammad, and other Advances 43 1.4.4 Transmission of Knowledge to Europe 46 1.4.5 The European Renaissance to the Modern Age 47 1.4.5.1 Overview 47 1.4.5.2 Clinical Physiology: the Pupil, Lens and Photoreception 47 1.4.5.3 Optics and the Pupil 50 1.4.5.4 Applications of Mathematics to Sensory Systems, Waveforms and Optics 51 1.4.5.5 Mechanistic Thinking 55 1.4.5.6 Reflexes as Involuntary Motions, Awareness of the Function of the Pupil 56 1.4.5.6.1 Advances in Reflex Autonomic Physiology and Anatomy 56 1.4.5.6.2 Renewed Awareness of the Relationship between the Pupil and Visual Impairment by Physicians 56 1.4.5.6.3 Functions of the Pupil 57
7 1.4.5.6.3.1 Historical Awareness of Basic Pupillary Functions 57 1.4.5.6.3.2 Retinal Illumination and Photopic Homeostasis: Pupil and Eyelid 58 1.4.5.6.3.3 Reducing Spherical Aberration 59 1.4.5.6.3.4 Reducing Chromatic Aberration 59 1.4.5.6.3.5 Overcoming Ametropic Errors with the Eyelids 59 1.4.5.6.3.6 Depth of Focus 59 1.4.5.7 Nineteenth Century Studies of Reflex Physiology in General, and of the Pupil in Particular 60 1.4.5.7.1 Overview 60 1.4.5.7.2 Patients with Good Vision but Pupils Poorly Reactive or Unreactive to Light 61 1.4.5.7.3 Peripheral Field Loss 61 1.4.5.7.4 Patients with Poor Vision and Excellent Pupil Reactions to Light 62 1.4.5.7.5 A Truly Blind Eye with a Pupil Apparently Constricting to Light Shone into it 62 1.4.5.8 Colour Vision, Cellular Concepts and Colour Channels 63 1.4.5.9 Advances in Studying Sensory Receptors 68 1.4.6 Structure of the Retina, Photoreceptors and Photopigments 69 1.4.6.1 General Topography of the Retina 69 1.4.6.2 The Electroretinogram and Classic Photopigments 70 1.4.6.3 Rods and Cones 71 1.4.6.4 Insights from Disease 75 1.4.7 Quantum Theory and Biological Effects of Light on the Eye 76 1.4.7.1 Quantum Theory 76 1.4.7.2 Ramifications for Understanding Visual Perception 78 1.4.7.3 General Effect of Light on Biological Tissues 79 1.4.7.4 Extra-ocular Systemic Effects 79 1.4.7.5 Specific Ocular Effects 79 1.4.7.5.1 Overview 79 1.4.7.5.2 Sight 80 1.4.7.5.3 Accessory Visual Functions I: Circadian Physiology 80
8 1.4.7.5.4 Accessory Visual Functions II: Pupillary Contraction 82 1.4.7.6 Non-Ocular photoreceptors? 84 1.4.8 Non-Rod Non-Cone Photoreception: Comparison with Animals 84 1.4.8.1 Non-Mammals 84 1.4.8.2 Mammals 86 1.4.9 The Photopigments 86 1.4.9.1 Structure 86 1.4.9.2 Functional Characteristics 87 1.4.9.3 Spectral Sensitivity of the Human Eye and Colour Vision 89 1.4.9.4 Absorption Spectra of the Visual Pigments 90 1.4.10 Phototransduction 94 1.4.10.1 The Classic Phototransduction Mechanism 94 1.4.10.2 Phototransduction mechanism in Melanopsin-Containing Ganglion Cells 96 1.4.10.3 Light and Dark Adaptation and Interplay of Photoreceptors 97 1.4.11 Processing within the Retina 99 1.4.12 Projections to the Brain: Anatomical Parallelism and Non- Parallelism between Visual and Accessory Visual Pathways 99 1.4.12.1 Historical Advances 99 1.4.12.2 The Nerve Fibre Layer 100 1.4.12.3 Structure of the Optic Nerve 100 1.4.12.4 The Optic Chiasm 101 1.4.12.5 Post-Chiasmal Pathways 101 1.4.12.5.1 The Retino-Geniculate and Geniculostriate Tracts 101 1.4.12.5.2 The Retino-Tectal (Mesencephalic) Pathway and Pupillary Contraction 102 1.4.12.5.3 The Accessory Optic System 106 1.4.12.6 Ganglion Cell Subtypes 108 1.4.12.7 Blindsight 117 1.4.12.8 The Effector Mechanism of the PLR: Structure of the Iris, Effects of Age and Iris Colour 117 1.4.13 Vision and the Pupil — Functional Similarities and Differences 118 1.4.13.1 Principles of Mathematical Descriptions of the Pupil 118
9 1.4.13.2 Spatial Summation and Ricco's Law 119 1.4.13.3 Visual and Pupillary Thresholds 120 1.4.13.4 Temporal Summation and Bloch's Law 120 1.4.14 Beyond the Pupil Light Reflex: Other Pupillary Oscillations 122 1.4.14.1 Fatigue Waves, Arousal and Pupillary Unrest 122 1.4.14.2 Distinct Channels beyond the Pupil Light Reflex: Overview 123 1.4.14.3 The Response to Pattern 124 1.4.14.3.1 Vision and the Pupillary Response to Gratings (PGR) 124 1.4.14.3.2 Pupil Acuity 127 1.4.14.4 The Pupil and Coloured Stimuli 127 1.4.14.4.1 Pupil Colour Response 127 1.4.14.4.2 Colour Blindness 129 1.4.14.5 Pupil Response to Colour Gratings 130 1.4.14.6 Pupil Response to Motion 131 1.4.15 Afferent Pupillary Defects 131 1.4.16 Clinical Applications - Pupillometry, Electrodiagnosis, Psychophysical and Other Tests 132 1.4.16.1 Introduction 132 1.4.16.2 Assessment of Visual Potential with Opacities of the Refractive Media 134 1.4.16.2.1 Mild to Moderate Ocular Media Opacities 135 1.4.16.2.2 Dense Media Opacities 135 1.4.16.2.2.1 Pupil Studies 135 1.4.16.2.2.2 Electrodiagnosis 136 1.4.16.2.2.3 Entoptic Phenomena 137 1.4.16.3 Orbital Decompression Surgery 138 1.4.16.3.1 Clinical Background 138 1.4.16.3.2 Orbital Trauma and Orbital Cellulitis 138 1.4.16.3.3 Colour Vision in Orbital Lesions — Thyroid Eye Disease, Orbital Haemorrhage and Infection 139 1.4.16.3.4 Contrast Sensitivity Function in Thyroid Eye Disease 139
10 1.4.16.4 Studies of Retinal Ischaemia and Predicting Neovascular Glaucoma 140 1.4.16.5 Pupil Perimetry 141 1.4.16.5.1 Background 141 1.4.16.5.2 Clinical Applications: Benefits and Limitations of Pupil Perimetry 141 1.4.16.5.3 Research Applications 142 1.4.16.6 Vision, the Pupil and Refractive Surgery 142
1.5 Implications of Reviewing the Literature 143
Chapter Two
General Experimental Methodology 144
2.1 Introduction 144
2.2 Photometric and Radiometric Measurement of Light 144 2.2.1 Principles 144 2.2.2 Irradiance or Radiance? 147 2.2.3 Additional Significance of the Photon 148 2.2.4 Illuminance (Illumination) or Luminance? 149
2.3 Presentation of Light Stimuli 150 2.3.1 Overview 150 2.3.2 Achromatic Light 150 2.3.3 Monochromatic Light 152 2.3.3.1 Monochromatic Light Production 152 2.3.3.2 Advantages and Limitations of Isolating Colour Channels 153 2.3.4 Varying the Emission Level (Photon Flux) of a Light Source 154
2.4 Measurements of Pupil Size through the Ages 155
11 2.5 Pupillograms 157 2.5.1 Considerations in Interpreting Results from Pupillometry 157 2.5.2 Applications of Cybernetics to Pupil and Vision Studies 158 2.5.3 Mathematical Approaches to Reduce Noise 159 2.5.4 Latency and Time Sequences 161 2.5.5 Parameters Extracted from the Pupillogram 162
2.6 Effect of Retinal Area 163 2.6.1 The Effect of Stimulus Size and Retinal Area 163 2.6.2 Maxwellian and Non-Maxwellian View 163
2.7 Pupillometry and Recording Hardware 164
2.8 Stimulus Presentation and Pupil Recording Paradigms 166
2.9 Collection, Storage and Analysis of Pupil Data 167
2.10 Electrophysiology 168 2.10.1 Stimuli 168 2.10.2 The Normal Electroretinogram 169 2.10.3 Factors Influencing the Normal Electroretinogram 169 2.10.3.1 Physiological 169 2.10.3.2 Equipment 169
2.11 Subject Recruitment and Exclusion Criteria 170
2.12 Ethics Approval and Informed Consent 171
12 Chapter 3 Subretinal Neovascularisation, Emergence of Concept of Pupil Evasion Ratio and Establishment of a Discriminatory Chromatic Threshold 172
3.1 Pupil Escape 172
3.2 Objectives 172
3.3 Materials and Methods 173
3.4 Calculation of a New Index - Pupil Evasion Ratio (PEVR) 176
3.5 Variation of PEVR 178 3.5.1 PEVR in Healthy Eyes 178 3.5.2 Comparing PEVR between Diseased and Control Eyes 179
3.6 Significance of PEVR 184
3.7 Use of Long, Medium and Short Wavelength Chromatic Stimuli 191
3.8 Establishment of a Chromatic Threshold 192
3.9 Pupil Evasion using Chromatic Stimuli 192
3.10 Significance of Chromatic Stimuli 194
3.11 Findings of Electrophysiology 194
3.12 Effect of a Sustained Stimulus 195
3.13 Significance of Results 195
13 Chapter Four
Retinal Detachment and the Pupil 197
4.1 Core Concepts 197
4.2 Cellular, Histological and Functional Changes in Retinal Detachment 197
4.3 Clinical Significance of the Pupil in Retinal Detachment 198
4.4 Potential for Pupillometric Study to Aid Diagnosis of Selected Retinal Detachments 199
4.5 Materials and Methods 200
4.6 Electrophysiology 201
4.7 Effect of Achromatic Stimuli on Pupil Constriction 204
4.8 Effect of Chromatic Stimuli on Constriction Amplitude 205
4.9 Effect on Pupil Evasion Ratio (PEVR) 207
Chapter Five Orbital Disease: Diagnosing Compressive and Traumatic Optic
Neuropathy 209
5.1 Compressive and Traumatic Optic Neuropathies and the Pupil 209 5.1.1 Overview 209 5.1.2 Chromatic and Achromatic Visual Responses in Optic Neuropathies 209
14 5.1.3 Constriction Amplitude, Latency and Pupil Evasion in Optic Neuropathies 209 5.1.4 Diagnosis of Optic Neuropathies 210 5.1.4.1 General Principles 210 5.1.4.2 Orbital Lesions: Compressive Optic Neuropathy 210 5.1.4.3 Ganglion Cell Function in Glaucoma 211 5.1.4.4 Idiopathic Intracranial Hypertension 212
5.2 Objectives 213
5.3 Materials and Methods 213
5.4 Achromatic Stimuli 215 5.4.1 Effect of Achromatic Stimuli 215 5.4.2 Sensitivity of Pupillometry over Clinical RAPD 216 5.4.3 Achromatic Light and PEVR 216 5.4.3.1 Comparison between Eyes 216 5.4.3.2 Effect of Pupil Size 217 5.4.3.3 Photopic versus Scotopic PEVR 217
5.5 The Effect of Chromatic Stimuli 219 5.5.1 Overview 219 5.5.2 Effect of Long Wavelength Light: Constriction Amplitude with Pooled Data 220 5.5.3 Long Wavelength Light: Inter-Eye Comparison of Constriction Amplitude 221 5.5.4 Effect of Long Wavelength Light on Inter-Eye PEVR 221 5.5.5 Effect of Medium Wavelength Light: Constriction Amplitude with Pooled Data 221 5.5.6 Medium Wavelength Light: Inter-Eye Comparison of Constriction Amplitude 223 5.5.7 Effect of Medium Wavelength Light on Inter-Eye PEVR 225
15 5.5.8 Effect of Short Wavelength Light: Constriction Amplitude with Pooled Data 225 5.5.9 Short Wavelength Light: Inter-Eye Comparison of Constriction Amplitude 226 5.5.10 Effect of Short Wavelength Light on Inter-Eye PEVR 227 5.5.11 Comparing the Effect of Chromatic Stimuli using Pooled Data 227
5.6 Light and Dark Adaptation 228 5.6.1 Purpose of Investigation 228 5.6.2 Methodology 229 5.6.3 Intra-Eye Effect of Achromatic Light on Light and Dark Adaptation 229 5.6.4 Inter-Eye Comparison of Achromatic Stimuli after Light and Dark Adaptation 230 5.6.5 Intra-Eye Effect of Chromatic Stimuli after Light and Dark Adaptation 232 5.6.6 Inter-Eye Comparison of Chromatic Stimuli after Light and Dark Adaptation 233 5.6.7 Implications of Work with Light and Dark Adaptation 234
Chapter Six Refractive Disorders and Neuronal Plasticity in Orbital Trauma, Compressive Orbital Lesions and Scarring with Proliferative Vitreoretinopathy: Use of Alternate Pupillometric Modalities 235
6.1 Introduction 235
6.2 Pupil Perimetry 235 6.2.1 Technical Factors in Pupil Perimetry 235 6.2.2 Strategies in Pupil Perimetry 236 6.2.3 Relevance to Current Work and Objectives 236
16 6.2.4 Optical Effects of Cataracts 237 6.2.5 Experiments using Light Scatter 237 6.2.6 Chromatic Aberration 239 6.2.7 Index Myopia 240 6.2.8 Methods 241 6.2.9 Stimulus and Recording Paradigm 242 6.2.10 Results 243 6.2.10.1 Intra-Eye Comparisons: Emmetropic Eyes with no Cataract 243 6.2.10.2 Intra-Eye Comparisons: Effect of Axial Myopia 243 6.2.10.3 Intra-Eye Comparisons: Effect of Cataract in Emmetropia — Scatter and Chromatic Aberrations 244 6.2.10.4 Effects of Index Myopia in Nuclear Sclerotic Cataract 246 6.2.11 Inter-Eye Comparisons 248 6.2.12 Implications 249
6.3 Pattern as a Stimulus 249 6.3.1 The Pupil's Response to Pattern 249 6.3.2 The Pupil's Response to Coloured Pattern 249 6.3.3 Applications to Disease: Weinstein's Hypotheses 250 6.3.4 Studies in Thyroid Eye Disease 250 6.3.5 Pupil Acuity 250 6.3.5.1 Applications to Disease 250 6.3.5.2 Theoretical Possibility of Measuring Pupil Acuity due to Scarring in Proliferative Vitreoretinopathy Compared with Visual Acuity 251 6.3.6 Testing Weinstein's Hypotheses 252 6.3.6.1 Experimental Methodology 252 6.3.6.2 Proliferative Vitreoretinopathy — Pathology 253 6.3.6.3 Recruitment of Subjects with PVR 254 6.3.6.4 Traumatic Optic Neuropathy 255 6.3.6.4.1 Gross Pathogenesis of Traumatic Optic Neuropathy 255 6.3.6.4.2 Primary Mechanisms of Optic Nerve Injury 256
17 6.3.6.4.3 Secondary Mechanisms of Optic Nerve Injury 256 6.3.6.5 The Pupil in Traumatic Optic Neuropathies 257 6.3.6.6 Prognosis of Traumatic Optic Neuropathy 258 6.3.6.7 Recruitment of Subjects with Intra-orbital Traumatic Optic Neuropathy 258 6.3.6.8 Relevance of Studies in Amblyopia 258 6.3.7 Experiments 259 6.3.7.1 General Experimental Paradigms 259 6.3.7.2 Achromatic Gratings and Colour Substituted Gratings 259 6.3.8 Responses to Achromatic Gratings in Orbital Trauma 261 6.3.9 Responses to Achromatic Gratings in PVR 263 6.3.10 Responses to Achromatic Gratings in Amblyopia 265 6.3.11 Responses to Chromatic Gratings in Orbital Trauma 267 6.3.12 Responses to Chromatic Gratings in PVR 269 6.3.13 Responses to Chromatic Gratings in Amblyopia 271 6.3.14 Significance of Findings 273
6.4 The Photostress Test 274 6.4.1 Photostress Testing 274 6.4.2 Clinical Implications 275 6.4.3 The Pupil Photostress Test 276 6.4.4 Factors Affecting Recovery from Pupil Photostress 276 6.4.5 Technical Considerations 277 6.4.6 Patient Recruitment 278 6.4.7 Methodology 278 6.4.8 Orbital Trauma 279 6.4.8.1 Orbital Trauma — Achromatic Recovery from Pupil Photostress 279 6.4.8.2 Orbital Trauma — Chromatic Recovery from Pupil Photostress 279 6.4.9 Vitreo-retinal Disease 281 6.4.10 Significance of Pupil Photostress Tests 283
18 Chapter 7 A Double Constriction in the Pupil Waveform: Signature of a
Novel Photoreceptor? 286
7.1 Rationale 286
7.2 Subjects 287
7.3 Stimulus Presentation 287
7.4 Response Recording 288
7.5 Results: Routine Findings 288
7.6 Novel Findings: Spectrally Dependent Double Pupil Constriction 289
7.7 Age and Gender Relationship 292
7.8 Potential Explanations: Less Likely Aetiologies 292 7.8.1 Excluding Artefact: Stimulus Generation 292 7.8.2 Excluding Accommodation 292 7.8.3 Fatigue Waves 292
7.9 Potential Explanations: Significance of the Double Constriction 293 7.9.1 Overview 293 7.9.2 Implication for Pupil Evasion Ratio 293 7.9.3 Significance to Melanopsin-Containing Photoreceptors 293 7.9.4 Significance to Research Methodology 294
7.10 Physiological Purpose of Second Constriction 294
7.11 Studies in Retinas of Laboratory Primates 294
19 Chapter 8 Humans Lacking Rods and Cones: Electrophysiology, Imaging, Genetics and Clinical Findings 296
8.1 Introduction 296
8.2 'Classic' Photoreceptor Dystrophies 296
8.3 Genotype-Phenotype Correlation in Autosomal Dominant
Cone-Rod Dystrophy 297
8.4 Effect of Cataract 298
8.5 Retinal Function 299
8.6 Fundus Photography and Ocular Coherence Tomography 299
8.7 Electroretinography 301
8.8 Implications 303
Chapter 9 Direct Evidence for a Melanopsin-Containing Non-Rod Non-Cone Photoreceptor: Irradiance Curves and Pupil Action Spectra 304
9.1 Purpose of Action Spectra: Overview 304
9.2 Pupillometry and Action Spectroscopy 305
20 9.3 The Biological Assay: Pupillometry 305
9.4 Constructing Irradiance Response Curves and Action Spectra: Overview 306
9.5 Construction of Irradiance Response Curves: Principles and Details 307
9.6 Calculating the Relative Sensitivity for the Action Spectrum 312
9.7 Template Fitting 313
9.8 Assumptions in Action Spectra 316
9.9 Conclusions 318
9.10 Special Acknowledgement 319
Chapter 10 Circadian Experiments: Actigraphy, Light Exposure and
Questionnaire Sleep Studies 320
10.1 Assessment of Sleep: Actigraphy, Polysomnography and CCTV 320 10.1.1 Methods of Assessing Sleep 320 10.1.2 Actigraphy and Circadian Rhythms 321
10.2 Methods — Actigraphy 322 10.2.1 Altered Social and Medical Condition of Subject 322 10.2.2 Recording of Movement and Light Exposure 323
21 10.3 Actigraphy and Light Exposure Data Analysis 325 10.3.1 Overview — Results of Actigraphy 325 10.3.2 Rhythm Analysis: Statistical Data 328 10.3.3 Conclusions from Actigraphy 333
10.4 Questionnaire Studies 334
10.5 Implications of Circadian Experiments 335
10.6 Special Acknowledgement 335
Chapter Eleven Visual Perception Originating from a Novel Photoreceptor 336
11.1 Rationale 336
11.2 Methods for Testing Visual Perception 336
11.3 Responses to Forced-Choice Testing 337
11.4 Rationale in Testing for Perception of Colour 339
11.5 Methods for Testing for Perception of Colour 339
11.6 Discrimination of Brightness and Hue 339
11.7 Comments on Findings 340
22 Chapter Twelve
Effect of Darkness on Non-Rod Non-Cone Mediated Pupillary
Constriction 342
12.1 Reactions to Darkness — Overview 342
12.2 The Effects of Immediate Light-Off 342
12.3 Anatomy of the Light-Off Pathway 343
12.4 Paradoxical Pupil Constriction in Darkness 344
12.5 The Delayed Effects of Darkness 345
12.6 The Black Period and Afterimage Contraction 345
12.7 The Dark Pause 346
12.8 The Dark Response in a Human Lacking Rods and Cones 347
12.9 Materials and Methods 348
12.10 Darkness Constriction without Rods and Cones 348
12.11 Interpretation of Results 349
Afterword Overview of Thesis Findings 351
23 Appendix I Example of a Patient Information Sheet 355
Appendix II The Lens Opacities Classification III [LOCSIII] 356
Appendix III The `Morningness-Eveningness' (` Lark/Owl' ) Questionnaire 357
Appendix IV The Pittsburgh Questionnaire 364
References 368
Inserts Publications in Peer-Reviewed Journals (After Page 432)
24 Lists of Tables, Figures and Inserts
Tables
Table 1.1 Classification of ganglion cells. 111 Table 2.1 Electroretinography (ERG) stimuli. 168 Table 3.1 Clinical Features of the 10 participants with unilateral subfoveal subretinal membranes. 175 Table 3. 2 Proposed relationship between pupil dynamics and the cellular processes they may reflect in the afferent arm of the pupil light reflex. 188 Table 3.3 Flash electroretinography of 10 eyes with subretinal membranes (SRNs). 195 Table 5.1 Subjects being considered for orbital decompression. 214 Table 6.1 Classification of proliferative vitreoretinopathy (PVR). 255 Table 6.2 Orbital fracture and visual acuity used to grade traumatic optic neuropathy 257 Table 7.1 Response parameters for all eight chromatic stimuli and the achromatic stimulus (all stimuli of lOs duration) in healthy subjects to show double pupil constriction where present. 289 Table 10.1 Non-parametric circadian rhythm analysis of activity and ambient light per week and across three weeks, in a subject without rods and cones. 333
25 Figures
Figures are original drawings, or adaptations or copies from other sources. The latter are acknowledged in the relevant legend, unless obtained from internet sites purporting to allow duplication and printing of images without permission and acknowledgement for non-commercial use.
Figure 1.1 The anterior and posterior visual systems. 39 Figure 1.2 Relationship of the pupil to surrounding structures. 40 Figure 1.3 Some of the key cells in the retina of special interest. 70 Figure 1.4 The visual portion of the electromagnetic spectrum. 78 Figure 1.5 Peak spectral sensitivity of rods. 90 Figure 1.6 Peak spectral sensitivity of S-, M- and L-cones. 91 Figure 1.7 Retino-geniculate and geniculostriate tracts: the classical pathway for vision. 101 Figure 1.8 Summary of the pupil light reflex pathway. 106 Figure 1.9 Anatomical pot dissection of the anterior visual system and immunofluorescence of the suprachiasmatic nucleas. 108 Figure 1.10 Physiological responses of melanopsin-containing ganglion cells to light. 113 Figure 1.11 Melanopsin-expressing ganglion cells. 116 Figure 2.1 The pupil waveform 160 Figure 3.1 Red-free image of a dense haemorrhagic subfoveal membrane in a subject with age-related macular degeneration. 173 Figure 3.2 Pupillogram showing pupillary diameter as a function of time. 177 Figure 3.3 Relationship of mid-dilation amplitude (MDA) to maximum constriction amplitude and relationship of pupil evasion ratio (PEVR) (%) to constriction amplitude. 179
26 Figure 3.4 Variation in estimates of pupil evasion ratio (PEVR) over seven light intensities in diseased eyes with subretinal membrane (SRN) compared with the healthy fellow eye. 181 Figure 3.5 Variation of mean difference in pupil evasion ratio (PEVR) between eyes with stimulus intensity (10 subjects). 182 Figure 3.6 Pupillogram from a normal eye and from its fellow eye with a subretinal membrane. 190 Figure 3.7 Comparison of effect of long wavelength light and short wavelength light on pupil response threshold in 10 eyes with subretinal membranes (SRNs). 193 Figure 4.1 Fundal view of retinal detachment. 199 Figure 4.2 Flash electroretinography of subjects with Macula-On and Macula-Off retinal detachments. 203 Figure 4.3 Pooled mean pupil constriction amplitudes as a function of log light stimulus intensity: Macula-On retinal detachments (RDs) with fellow healthy eyes and Macula- Off RDs with fellow healthy eyes. 204 Figure 4.4 Effect of red light on pupil constriction amplitude, comparing the mean responses of Macula-Off retinal detachments (RDs) with the mean of their normal fellow eyes. 206 Figure 4.5 Inter-Eye pupil evasion ratio (PEVR) as a function of stimulus intensity for Macula-Off versus Macula-On retinal detachments. 208 Figure 5.1 Range of inter-eye differences in mean pupil constriction amplitude for compressive orbital lesions. 218 Figure 5.2 Mean Inter-eye pupil evasion ratio (PEVR) in compressive orbital lesions. 219 Figure 5.3 Effect of long wavelength light on mean pupil constriction: pooled data for compressive orbital lesions. 220 Figure 5.4 Effect of long wavelength light on pupil constriction amplitude in subjects with compressive orbital lesions. Inter-eye constriction amplitudes. 222
27 Figure 5.5 Mean pupil constriction amplitude as a function of stimulus intensity with a 540 nm light stimulus (pooled data) for subjects with compressive orbital lesions. 223 Figure 5.6 Inter-eye pupil constriction amplitudes as a function of stimulus intensity with a 540 nm light stimulus, comparing healthy eyes with diseased fellow eyes with compressive orbital lesions. 224 Figure 5.7 Mean pupil constriction in eyes with compressive orbital lesions as a function of stimulus intensity using pooled data and a short wavelength light source (440 nm). 225 Figure 5.8 Inter-eye pupil constriction amplitudes as a function of stimulus intensity for a short wavelength stimulus in compressive orbital lesions. 226 Figure 5.9 The statistical significance of detecting optic nerve compression in subjects being assessed for orbital decompression (pooled data). 228 Figure 5.10 Effect of light and dark adaptation on healthy and diseased eyes in compressive optic neuropathy. 230 Figure 5.11 Inter-eye effects of light adaptation and dark adaptation, comparing the effect of achromatic light on the diseased and healthy fellow eyes in compressive optic neuropathy. 231 Figure 5.12 Effect of light and dark adaptation on healthy and diseased eyes in compressive optic neuropathy (short wavelength stimulus). 232 Figure 5.13 Inter-eye effects of light and dark adaptation in compressive optic neuropathy comparing the effect of short wavelength light on the diseased and healthy fellow eyes. 233 Figure 6.1 Longitudinal chromatic aberration. 240 Figure 6.2 Transverse chromatic aberration. 240
28 Figure 6.3 Mean pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in emmetropic eyes with an arcuate scotoma but no cataract. 243 Figure 6.4 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an eye with an arcuate scotoma, axial myopia but no cataract. 244 Figure 6.5 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an emmetropic eye with an arcuate scotoma and dense nuclear sclerosis. 246 Figure 6.6 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an eye with dense nuclear sclerotic cataract, an arcuate scotoma and index myopia. 247 Figure 6.7 Colour fundus photograph of an eye with proliferative vitreoretinopathy (PVR). 253 Figure 6.8 Pupil gratings response (PGR) in orbital trauma. 261 Figure 6.9 Pupil gratings response (PGR) in healthy fellow eyes of eyes suffering orbital trauma. 262 Figure 6.10 Pupil gratings response (PGR) compared between eyes suffering orbital trauma and healthy fellow eyes. 262 Figure 6.11 Pupil gratings response (PGR) in eyes with proliferative vitreoretinopathy (PVR). 263 Figure 6.12 Pupil gratings response (PGR) in fellow healthy eyes to eyes with proliferative vitreoretinopathy (PVR). 264 Figure 6.13 Pupil gratings response (PGR) compared between eyes with proliferative vitreoretinopathy (PVR) and healthy fellow eyes. 265 Figure 6.14 Pupil gratings response (PGR) in an amblyopic eye. 266 Figure 6.15 Pupil gratings response (PGR) in the fellow eye to an amblyopic eye. 266
29 Figure 6.16 Pupil gratings response (PGR) compared between an amblyopic eye and its fellow healthy eye. 267 Figure 6.17 Red/Green pupil gratings response (PGR) in eyes suffering orbital trauma. 268 Figure 6.18 Red/Green pupil gratings response (PGR) in fellow healthy eyes to eyes suffering orbital trauma. 268 Figure 6.19 Red/Green pupil gratings response (PGR) compared between eyes suffering orbital trauma and their healthy fellow eyes. 269 Figure 6.20 Red/Green pupil gratings response (PGR) in eyes with proliferative vitreoretinopathy (PVR). 270 Figure 6.21 Red/Green pupil gratings response (PGR) in opposite healthy eyes to eyes with proliferative vitreoretinopathy (PVR). 270 Figure 6.22 Red/Green pupil gratings response (PGR) compared between eyes with proliferative vitreoretinopathy (PVR) and healthy fellow eyes. 271 Figure 6.23 Red/Green pupil gratings response (PGR) in an amblyopic eye. 272 Figure 6.24 Red/Green pupil gratings response (PGR) in the healthy fellow eye of an amblyopic eye. 272 Figure 6.25 Red/Green pupil gratings response (PGR) compared between one eye with amblyopia and the healthy fellow eye. 273 Figure 6.26 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in five subjects with unilateral orbital traumatic optic neuropathy following achromatic photostress. 280 Figure 6.27 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in five subjects with unilateral orbital traumatic optic neuropathy following chromatic photostress. 282
30 Figure 6.28 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in three subjects with unilateral Macula-Off retinal detachments. 283 Figure 7.1 Mean pupillography of 48 healthy eyes for four of the stimuli. 291 Figure 8.1 Colour fundus photograph and optical coherence tomography (OCT) of central macula in a subject without rods and cones. 300 Figure 8.2 Optical coherence tomography (OCT) of the near retinal periphery in a subject without rods and cones. 301 Figure 8.3 Electroretinographic responses without rods and cones. 302 Figure 9.1 Wavelength dependency of pupil constriction latency in a subject without rods and cones. 306 Figure 9.2 Irradiance response curves (IRCs) in five subjects with advanced photoreceptor dystrophies. 310 Figure 9.3 Irradiance response curves (IRCs) for the second pupil constriction of the right eye in a rodless coneless subject. 311 Figure 9.4 Irradiance response curves (IRCs) for the second pupil constriction of the left eye in a rodless coneless subject. 312 Figure 9.5 Subject SC. Action spectra and template-fitting for right eye (2"d constriction). 315 Figure 9.6 Subject SC. Action spectra and template-fitting for left eye (2nd constriction). 315 Figure 9.7 Action spectrum for subject SC fitted using a variable slope. 316 Figure 9.8 Action spectrum for subject SC fitted using a fixed slope. 317 Figure 10.1 21-Day Single Plot Actogram. 326 Figure 10.2 Double plot actogram for 21 days and radiant light exposure. 327 Figure 10.3 Statistical analysis of circadian photoentrainment. 329 Figure 10.4 Quantitative analysis of hourly activity and ambient light across a three week period. 332
31 Figure 11.1 Psychophysical testing of a subject without rods cones showing conscious perception of light at 481nm (p < 0.05), but inability (p > 0.05) to detect light at 420 nm, 460 nm, 500 nm, 515 nm, 540 nm, 560 nm or 580 nm. 338 Figure 11.2 Discrimination of hue as a function of wavelength without rods and cones. 341 Figure 12.1 Pupil reactions in a human subject lacking rods and cones. 349
32 Inserts (After Page 432)
Zaidi FH, Bremner FD, Gregory-Evans K, Cocker KD, Moseley MJ. Subretinal membranes are associated with abnormal degrees of pupil "evasion": an index of clinical macular dysfunction. The British Journal of Ophthalmology 2006 September; 90(9):1115-1118.
Zaidi FH, Hull JT, Peirson SN, Wulff K, Aeschbach D, Gooley JJ, Brainard GC, Gregory-Evans K, Rizzo JF 3rd, Czeisler CA, Foster RG, Moseley MJ, Lockley SW. Short-wavelength light sensitivity of circadian, pupillary, and visual awareness in humans lacking an outer retina. Current Biology 2007 December 18;17(24):2122-8.
+
33 Standard Abbreviations
In general Le Systeme International d'unites (S.I. units) are used as below unless a more commonly used unit is generally familiar. Distinction between radiometric and photometric units for different experiments is made in the text (chapter two). Non-S.I. units are presented when first mentioned in the relevant text.
Wavelength A, in nanometres (nm) 17 Speed of light in a vacuum c = 3.00 x 10 nm/s Velocity Energy in a photon Millivolts mV Microvolts [tV or microV Candela cd Hour(s) Minute(s) min Second(s) Hertz (frequency per second) Hz Metre(s) Centimetre(s) cm Millimetre(s) mm Micrometre(s) tun or microns Nanometre(s) nm Steradian(s), unit of a solid angle sr Degree(s) ° or deg Cycle(s) cy log logarithm logMAR logarithm of the minimum angle of resolution
34 Chapter One Introduction, Overview, Aims and Literature Survey
1.1 Introduction
In this chapter the aims of this thesis are presented followed by an overview summarising the central themes of the areas of study. There then follows a detailed literature review showing the historical background to the current state of knowledge; more specialised literature is reviewed as necessary in the subsequent chapters. These begin with chapter two where the general experimental methodology and related supplementary results are detailed with a general discussion. The other chapters contain the principal testable hypotheses and experimental details, results and discussions for each group of hypotheses. The principal findings are additionally summarised as an afterword. References are annotated in the text by the name of the first and / or second authors, and by year, and collected near the end of the thesis in alphabetical order. Appendices incorporate information that would otherwise disrupt the flow and sequence of the text. Publications of work in peer-reviewed journals detailing some key fundamental findings in this thesis are inserted at the end.
1.2 Overview
For thousands of years observation of the pupil has been used by physicians as an objective indicator of visual potential in diseases affecting the anterior visual pathway. Classical physiological and anatomical studies showed that the pupil light reflex is a paradigm of reflex arcs in general. Later studies showed that both vision and the pupil light reflex (PLR) originate in rod and cone-based phototransduction mechanisms. Gross macroscopic observation of the pupil's anatomical size and its physiological motions were during the mid-twentieth century supplemented by the technological advance of modern infrared video-pupillometry, allowing spatial and
35 temporal measurements of the pupil up to at least 0.01 mm and in excess of 100 ms respectively. Paralleling the visual pathway, the PLR was found to have multiple components including channels for luminance, colour and pattern, some of which would be studied in human diseases. In the early twenty first century a non-rod non-cone photoreceptor was identified in certain mammals mediating circadian rhythms and contributing significantly to the PLR. A visual influence for this receptor had, hitherto to the work presented herein, not been suggested.
1.3 Aims
In order, the aims of this thesis are as follows. Firstly to present the background to the experimental work by providing a broad and comprehensive review of the existing knowledge that builds to the current work and which may be found in the academic literature. Subsequently it is to provide the general paradigm used in much of this work for subject recruitment, stimulus presentation and experimental recordings. Subsequent chapters present experimental details, most notably using dynamic infrared video-pupillometry, supplemented where necessary by performing electrophysiology. These are studies of the pupil light reflex in diseases affecting photoreceptors and their distal connections, aiming to differentiate diseased from healthy eyes. This is from the perspective of the clinical scope for application of pupillometry, and is reflected in the choice of pathologies studied: diseases of the orbit, retina, vitreous, and refractive disorders. At a basic scientific level analysis of these results aims to develops a hypothesis that the PLR in humans is driven in part by hitherto unknown non-rod non-cone photoreceptor influences from ganglion cells in the retina and optic nerve. The later chapters in this work aim to develop this hypothesis further and to specifically study this receptor's role in mediating pupil constriction, circadian rhythms and vision.
36 1.4. Survey of the Literature
1.4.1 Perception and Methods for Studying Visual Perception
Perceptions (qualia) are beliefs, awareness or inferences arising from sensory data at either a conscious or subconscious level [Gregory 1998]. The central nervous system (CNS), comprising eyes, brain, certain sensory nerves and the spinal cord, have been variously viewed as external receivers and relay stations for perception, or alternatively as generators of the conscious and subconscious mind, or partly as both [Crick 1994; Gregory 1998]. These views have arisen in part as perception has been successfully studied from several perspectives — the physical and biological sciences as well as the arts and humanities, and fields such as philosophy and mysticism [Hochberg 1978 and 1984; Jung 1984; Zeki 1993; Gregory 1998; Velmans 2000]. This is especially true of the study of visual perception, or seeing and sight.
As a percept sight is a subjective experience though it is quantifiable through either subjective measurement (for example, visual acuity, visual field, colour vision, contrast sensitivity, glare testing) or objective measurement of its associated or accessory parameters and the responses that light elicits e.g. pupil movement as an index of acuity and field, or circadian rhythms. Thus while a blind subject cannot appreciate first-hand the concept of ocular vision or its biologically determinable components of luminance, colour, pattern, contrast or visual motion detection, these processes, or their absence, have been quantified by sighted (and with some tests non-sighted observers) by observing the following categories of test: i. Physiological studies, including of biochemical parameters. These are objective studies of function relying primarily on observation by an investigator, e.g. electrical excitation of the visual cortex in the brain, or accessory visual functions such as the pupil's motion
37 in response to light, or patterns of light-induced circadian behaviour. ii. Anatomical structure (biological form and topography) from biopsies and/or imaging of tissue, which can be macroscopic (gross), microscopic, ultrastructural (subcellular) and molecular studies. The term 'optical biopsy' can be used to describe the specific technique of optical coherence topography, used in a key point in this work, and which provides an interferometric image of the micro-architecture of ocular structures as well as other tissues including skin. Alternatively, the term optical biopsy in its generic sense seems to encompass all forms of imaging anatomical form, including computed tomography (CT) with cross-sectional X-rays and magnetic resonance imaging (MRI), though such a use for the term does not seem to have been previously suggested. iii. Psychophysical responses. These are the presence or absence of subjective responses as detected by subjects in response to visual stimulation with light.
Certain types of study of visual perception have relied more heavily on certain disciplines. As an example, the region from the eye to the optic chiasm can be conveniently said from an anatomical, functional and clinical perspective to comprise an anterior visual system which at the chiasm becomes continuous with a posterior visual system (figure 1.1). Studies of the former have relied more heavily on the physical sciences, especially the study of the nature of light, time, energy, matter and space, as the interface of light interacting with living (biological) tissue in the eye is central to study of the anterior visual system [Duke-Elder 1932; Duke-Elder and Scott 1971]. Studies of the latter have been relatively more influenced by the neurosciences. Applied sciences also influence vision science. Cybernetics has facilitated modelling biological responses to light and gave rise to the concept of processing "channels" to describe specific functional nervous pathways such as luminance or colour [Longtin et al 1990; Loewenfeld and Lowenstein 1999]. Studies of clinical diseases have contributed
38 significantly to understanding basic mechanisms and provide an important avenue for practical application of basic scientific findings.
Figure 1.1 The anterior and posterior visual systems. These may be regarded as segregating at the optic chiasm, facilitating different modes of experimental study. Recent evidence presented later in this literature review suggest a retino-hypothalamic tract fundamentally linked to giant ganglion cells, the latter which form the focal point of later study presented in this thesis. The retino-hypothalamic tract may separate from other ganglion cells at the chiasm, which would represent the first anatomical divergence of visual information. Coincidentally this division of the visual system also tends to reflect surgical distinctions in specialties between ophthalmology and orbital surgery (concerned primarily with the anterior visual system) and neurosurgery (the posterior visual system and chiasm). These facts are doubly relevant to work presented in this thesis. First, as subjects were recruited from an eye hospital, compressive neurological lesions affecting the visual system were naturally those affecting the anterior visual system and caused by orbital disease (tumour, trauma and thyroid eye disease). Further, by definition, these subjects all hence had lesions at a point in the visual system before gross segregation of ganglion cell projections had
39 occurred, which is of relevance to the later work presented in this thesis involving specific study of these putative novel giant ganglion cells.
1.4.2. The Pupil and the Integrity of the Visual System
The pupil is the space in the eye within the iris and is slightly off-centre and inferonasal to the centre of the cornea as shown in figure 1.2 [Kardon 2003].
Figure 1.2 Relationship of the pupil to surrounding structures. The right pupil of a subject is shown in the top figure, where it can be appreciated that the pupil lies inferonasally in the iris.
The earliest recorded physiological index of sight has been the study of the motion of the pupil, presumably owing to the accessibility of such study
40 [Thompson 2003]. This is as the pupil is itself found within the central organ of vision of the eye in between the transparent entry port for light of the cornea and the neuronal receptors for light in the retina near the posterior portion of the eye, where both visual and pupillary pathways run concurrently for much of their proximal pathways [Duke-Elder and Scott 1971; Robinson and Fielder 1992; Fielder and Moseley 2000; Thompson 2003]. The study of the pupil has paralleled that of vision for millennia, and numerous scholars, scientists and philosophers have devoted at least some part of their efforts to studying one or both. A formidable corpus of literature has developed devoted to these fields in which the study of the pupil is as ancient as the sources on the study of vision.
1.4.3 Ancient Literature
1.4.3.1 Cataract Surgery and the Pupil
The practical study of the pupil within a civilisation may be traced to the attempts by physicians to treat the common affliction of cataract, or opacification of the lens of the eye [Thompson 2003]. As the leading cause of blindness worldwide cataract surgery is likely to have attracted the interest of surgeons for longer than any other operation given the devastating consequences of blindness for the sufferer [Thompson 2003; Fishman 2003]. Further, most operations on organs could not be performed without the risk of infection in eras without antibiotics, and also without effective anaesthesia and hence excruciating pain. The surgical removal of cataracts was essential for the sufferer in ancient times when support for the blind and partially sighted was especially limited [Fishman 2003]. Cataract is first recorded to have been surgically treated in 600 B.C. by Susruta of India who first introduced the couching operation in which the cataract is surgically dislocated into the vitreous hence crudely clearing the visual pathway for light transmission (this operation is still performed very rarely by 'quack' doctors in parts of the developing world, but nowadays such an occurrence would be regarded as a complication) [Rucker 1965]. After the
41 conquest of northern India by Alexander the Great, his physician brought this knowledge to Alexandria in Egypt. The earliest extant description of the couching operation is given by Celsus in 25 A.D. Interest in cataract surgery is likely to have served to drive interest in the motions of the pupil as an index of visual potential, or what residual vision would exist after surgery on the cataract [Thompson 2003]. The pupil was also studied in the context of pharmacology in ancient India: the alkaloid belladonna (familiar to modern medicine as the drug atropine) was known as a poison, and by causing pupil dilation was also used therapeutically in India for eye diseases [Loewenfeld and Lowenstein 1999].
The Ancient Greeks, even Hippocrates, are not recorded to have contributed to understanding a link between the pupil and vision. The name Claudius Galen is linked to what survives of the knowledge of the ancient world in many medical fields as Galen recorded his thoughts as a teacher of medicine and surgery for posterity and his teachings were also popularised within the historically influential and immense Roman Empire [Thompson 2003]. Born and trained in medicine in Pergamon in Asia Minor, after completing additional clinical work in the main medical cities of his time, Galen moved practice to Rome, which was then the political, commercial and military centre of the world. In Rome he couched cataracts of which there must have been plenty [Thompson 2003]; he died in 199 AD. Galen's reputation and standing demanded a high success rate from his cataract surgery and also good visual outcome so as to avoid disappointed patients [Thompson 2003]. Thompson suggests that Galen noticed that when assessing patients for cataract surgery a visible inequality of static pupil size was of limited prognostic value as the pupil can be of normal size with even severe cataract or many other disease processes in the eye or the opposite fellow eye. Galen also noted what is now called the "consensual" light reflex and drew an association between it and visual potential in the fellow eye. He observed that if a diseased eye is occluded there is less dilation of the fellow eye than otherwise expected, and sometimes there is none. He noted that if positive this test may be associated with a poor visual
42 prognosis for cataract surgery in the affected eye [Galen 2nd century; Thompson 2003]. Galen had appreciated the association between disease in one eye and the movement of the pupil of the fellow eye, and also that pupillary signs may be an observable and objective indicator of visual loss in humans, one independent of the patient's feelings about his visual loss, a sign that "made a statement about the integrity of a part of the visual system that was otherwise entirely invisible and unknowable to the doctor" [Thompson 2003]. However, Galen's proposed mechanism to explain his correct observation was flawed. This was as he used the Platonic schematic inherited from Ancient Greece by the Romans wherein a "pneuma" was believed to emerge from the eye. According to Plato, whose views dominated much of Ancient Greece and Rome, light was not thought of as the stimulus for vision, but sight was regarded as a consequence of the emission of the 'visual spirit' from the brain, optic nerve and finally the eye, passing through its pupil to the outside world [Thompson 2003]. Galen hence mistakenly proposed that the pneuma was responsible for pushing the pupil aside and thereby causing the pupil's movement [Thompson 2003].
1.4.3.2 Discarding Platonic Views: Light Recognised as the Visual and Pupillary Stimulus by the Imam Jafar ibn Mohammad, and other Advances
Galen's "prognostic pupillary signs" were translated and transcribed into Arabic [Thompson 2003]. This preserved them during the European Dark Ages that followed the destruction of Rome [Thompson 2003]. In the Middle East fundamental scientific and philosophical observations on visual perception would be made and were allied with successful clinical outcomes by Muslim and Jewish physicians who discarded some of Plato's more ancient and retarding concepts [Thompson 2003]. These were original observations and were not an incantation of older Greek, Persian or Coptic knowledge nor was it knowledge that was, as some have rather fancifully advanced, secretly saved from being burned in the library of Alexandria [Thompson 2003; Mirza 1997]. Such physicians successfully applied
43 Galen's prognostic pupillary signs in deciding which cataracts carried a good visual prognosis in the 9th and 10th centuries [Hunain ibn Is'haq 9th centry; Ali ibn Isa 10th century; Thompson 2003]. Further, it was apparent to both Ammar ibn Ali and to Ibn Sina (also known as Avicenna) that a light stimulus entering the pupil from outside drove the pupil's movement [Hirschberg 1985; Thompson 2003]. Both men also noted the existence of a direct PLR for the first time in the literature, in distinction to Galen who had noted only the consensual response [Thompson 2003]. Ibn Sina also noted that the iris was substantially composed of muscles which mediated pupillary constriction [Loewenfeld and Lowenstein 1999]. Light was understood to be the stimulus to the contraction of these muscles. This key physiological observation had probably been made even earlier, and would have paralleled the fundamental advance made during this time period wherein light was recognised as the stimulus for vision, as proposed by the Imam Jafar ibn Mohammad (702-765) who taught at the academy started by his father the Imam Mohammad ibn Ali in the city of Medina [Sarton 1927; Mirza 1997]. The Imam Jafar ibn Mohammad stated that for every sensory phenomenon, including sight, there was an organ (in this case the eye) to perceive it, and in the case of vision the stimulus was light, which gave rise to the visual phenomena including colour vision. The Imam Jafar ibn Mohammad is also recorded to have correctly stated the concept that rays of light are emitted, reflected or transmitted by objects, and, further, that it is the entry of these rays of light into the eye that causes vision (as opposed to the Greco-Roman pneuma, which postulated the reverse):
"The rays of light from different objects come to our eyes and enable us to see them. The rays of light from our eyes do not go out and fall on other objects, otherwise we could have seen them in the darkness also. We see only those objects which are luminous. If they are not luminous themselves, they must reflect the light falling upon them by some luminous object."
The first recorded fundamental mathematical concepts in geometric optics were proposed by the Imam Jafar ibn Mohammed on the basis of which he
44 proved that scattering of light can be explained by rays and are the cause of unclear vision for distant objects:
"If we make a device through which all the rays of light coming from the camels grazing at a distance of 3000 Zirah [One Zirah = 40 inches] entered our eyes we would see them grazing at a distance of only 60 Zirah and all other objects would look 50 times nearer to us."
He also stated that since minification of objects occurs as little light is geometrically entering the eye (a small angle object) that the opposite, magnification, would occur as more light enters the eye and the angle increases, hence distant objects could be magnified if the rays of light from the object could be concentrated to increase the angle by a technical instrument. This is the first description of a theory for magnifying instruments, both simple lenses as well as microscopes and telescopes. The Imam Jafar ibn Mohammed's student was the scientist Jabir bin Hayan (also known as Geber, termed the 'father' of chemistry). Reflecting his master's instructions Jabir emphasised experimental methodology in pursuing scientific knowledge [Sarton 1927]. The Imam Jafar ibn Mohammed also stated that light and time were related, that light had a motion, and, further, that this motion was very rapid. Light was said to be capable of moving objects if concentrated, an important principle that while technically unfeasible in ancient times is an important principle associated with the mechanism of lasers.
Relatively recently it has been shown that in approximately 984 Ibn Sahl wrote a treatise called 'On Burning Mirrors and Lenses' in which he set out his understanding of how curved mirrors and lenses bent and focussed light, and on which basis Ibn Sahl has been credited with the first recorded description of what was much later called Snell's law of refraction; Ibn Sahl also used the law to work out the shapes of lenses that focus light with no geometric aberrations, known as anaclastic lenses [Rashed 1993]. Rashed has shown that Ibn Sahl's treatise was used in part by Abu Ali al-Hasan ibn
45 al-Hasan ibn al-Haitham of Basra in Iraq (often called Ibn al-Haitham, also known in the West as Al Hazen) (965-1039) in his Book of Optics which brought together much of the work up to that time, including dispersion of white light into its constituent colours [Rashed 1993; Al Deek 2004]; Al Haitham's works also noted the physical concept of momentum, and included important treatises dealing with anatomy, physiology and medicine including ophthalmology [Al Deek 2004] Similar optical work by Willebrord von Roijen Snell in the Europe of 1621 mathematically described the bending of light when passing through changes of refractive index, published by the philosopher Rene Descartes as the sine law of refraction 11 years later and over the course of time coming to be known as Snell's Law [Gregory 1998]. In addition to his law of refraction Al- Haitham also geometrically described reflection and built the first pinhole camera showing an intra-ocular image of the outside world formed by inversion of the image passing through the cornea in the pupil plane [Sarton 1927].
More general truths about perception were also formulated by other individuals in the ancient Middle East. Ibn-Sing (ca. 980-1037) in Persia explained sensory phenomena in terms of five major senses and other minor ones. He noted that perception occurred in a "common sense" residing uniquely in its full form in the human brain, while the peripheral sensory receptors were considered similar to those of animals whose sense of consciousness differed from that of humans [Jung 1984].
1.4.4 Transmission of Knowledge to Europe
In Spain and in Sicily Muslim and Jewish scholars transmitted these major theoretical and practical advances in knowledge, crucially free of Platonic restrictions, to Europe [Thompson 2003]. In Spain Ibn-Rushd (Averroes) (1126-1198) considered the retina as the tissue containing the receptors for light [Jung 1984]. The German monk Albertus Magnus (1197-1280) became familiar with Arabic works in Padua and disseminated this
46 teaching, amid other places, to Cologne, Strasbourg and Paris [Jung 1984]. Further advances took place especially within numerous monastic communities and new European universities where significant and novel progress was made in understanding human anatomy [Jung 1984]. Magnus's is the first record that postulated the partial decussation at the optic chiasm, a formidable leap in understanding the complexity of the anatomy of the visual system [Jung 1984; Theiss and Griisser 1994; Reeves and Taylor 2004; Taylor 2005]
1.4.5 The European Renaissance to the Modern Age
1.4.5.1 Overview
Modern science accelerated during this period in three relevant areas, and in approximately the following chronological order: i. Optics and further understanding of light as a stimulus. ii. Cellular anatomy and processes, especially in the retina and optic nerve. Most recently this has at a fundamental level included an appreciation of the link between a non-classic (non-rod non-cone) photoreceptor isolated in certain mammals and the pupil's motion. iii. Clinical physiology and surgery. As noted the field of medicine had provided the original context in which visual and accessory visual functions such as pupil constriction were intertwined for study in the ancient world. There was however a very meandering subsequent course of development of key concepts.
1.4.5.2 Clinical Physiology: the Pupil, Lens and Photoreception
During the European Renaissance Leonardo da Vinci (1425 — 1519) in Italy observed that decapitated frogs still moved their bodies in response to certain sensory stimuli, but that these movements were abolished when he
47 destroyed the frog's spinal cord [Jung 1984]. Despite this key observation it did not however immediately translate into an understanding of reflex physiology and anatomy. This ambiguity in thought also occurred in understanding visual perception. An erroneous view of the period that had crept into scientific thought despite earlier advances was that the lens and not the retina was the location of the photoreceptors. Felix Platter started to reverse this view, demonstrating in 1583 that vision continues after the lens has been isolated by cutting its suspensory ligament, suggesting photoreception occurred in the retina and not the lens. His view was slow to catch on more widely as his hypothesis lacked a sufficiently detailed optical basis (the theory of accommodation), as well as experimental verification, and would await later work by Kepler and Scheiner on optics, accommodation and the pupil [Boring 1942; Polyak 1957; Duke-Elder 1958; Duke-Elder and Wybar 1961; Duke-Elder et al 1968; Brindley 1970; Duke-Elder and Scott1971].
In the context of vision and the pupil, three 16th century physician surgeons, Pierro Franco, Felix Platter and Ambroise Pare, as well as the barber surgeon Bartisch, recommended use of the pupils as a predictor of visual success in cataract surgery [Pare A 1585; Bartisch 1585; Koelbing Huldrych 1967; Thompson 2003]. No longer burdened by the concept of a visual pneuma, they disregarded copies of Galen's work and according to Thompson did as the Eastern authors whose works they were familiar with had done before. First they closed both of their patients' eyes, then pressed and rubbed them through the eyelids with their thumbs, giving the surgeon the opportunity to palpate the orbits for prominence, resistance or discomfort. It also helped the pupil test by creating a brief period of dark adaptation and thus bigger pupils which would exert a greater pupillomotor force. The direct pupillary response to light could then be seen when one eyelid was suddenly lifted [Thompson 2003]. Pare's comments are the most detailed to have survived, as subsequently translated in Elizabethan English (given his long-lasting influence in medical circles) on the "ripeness" of a
48 cataract under the title "By what signs ripe and curable cataracts may bee [sic] discerned from unripe and uncurable ones":
"If the sound eye being shut, the pupil of the sore or suffused eye, after it shall be rubbed with your thumbe [sic], bee presently dilated and diffused, and with the like celerity ['celerity', Latin for clarity] returne into the place, color and state, it is thought by some to shew [sic] a ripe and confirmed cataract. But an unripe and not to be couched, if the pupil remain dilated and diffused for a long time after...Cataracts are judged uncurable...whose pupil becometh no broader by this rubbing: for hence you may gather that the stopping or obstruction is in the opticke [sic] nerve, so that how cunningly or well soever the cataract bee couched, yet will the patient remain blind."
The above description by Pare is remarkable for its lucidity and also suggests that it was known from the books transmitted to these surgeons (who as physicians and not barber surgeons were required to read the ancient works during their course of formal study) that the pupil test assesses the integrity of the optic nerve. Pare was undeniably well-read and technically a cut above other surgeons, he was influential and became the most renowned surgeon of the 16th century. For example, his record is the first description to have survived of phantom limb pain following limb amputation, and he rejected the common practice of pouring boiling oil onto chest wounds as it did more harm through the burn than good, though the latter innovation originated in the fact that the defenders ran out of oil at the siege of Turin (Turino) which he observed. He also encouraged other surgeons to handle tissue delicately and respectfully, perhaps owing to a reputation in eye and orbital surgery. He attempted removal of an orbital foreign body from King Henry II of France in 1559 - a wooden splinter from a lance sustained in a jousting accident which had penetrated via the right upper eyelid, fractured the orbital roof lacerating the meninges. The patient however succumbed to meningitis and frontal lobe contusion as a result of which jousting was banned in France [Orcutt 1990]. Nevertheless
49 owing to Pare's repute and influence, all cataract surgeons for the next 300 years considered it good practice to check the direct pupil reaction to light before cataract surgery. The American ophthalmologist Stanley Thompson, who began practicing as a doctor in the mid-twentieth century, states his teachers noted that the said practice had sadly declined in the 19th century, only to be re-kindled in the 20th century.
1.4.5.3 Optics and the Pupil
As a science optics, which understands light as rays, had declined since the time of Al Hazen. This would change with Zacharias Jansenn of Holland who is credited as having built the first microscope simply by putting together two convex lenses at the very end of the 16th century [Jung 1984; Audouze and Israel 1985]. In his early 17th century works Dioptrice and Harmonices Mundi the German Johannes Kepler displayed a formidable appreciation of the optics of the eye, going back to his earlier seventeenth century work on optics the Ad Vitellionem Paralipomena [Kepler 1611 and 1619]. The role of the pupillary plane in understanding the working of lenses, microscopes and telescopes were appreciated with optical diagrams. The telescope was in practical use by at least the early seventeenth century, possibly influenced by Kepler in Prague in the course of producing mathematical laws of planetary motion such as orbital trajectories of planets under the theme of a clockwork universe, another example of this period whereby mathematics was used to increase understanding of physical mechanisms. The telescope was famously developed by Galileo in Italy who discovered the planet Jupiter's satellites using it [Jung 1984; Audouze and Israel 1985]. Kepler may also have introduced glasses for hypermetropia, though spectacles have been found in paintings from centuries before [Milder and Rubin 2004].
Further relevant optical achievements would emerge in this background. In 1619 the Jesuit priest Christoph Scheiner published Oculus, a work
50 including his descriptions of the excision of the sclera of an ox's eye and projection through the cornea, pupillary aperture, lens and vitreous of an inverted image onto the retina [Jung 1984]. A very well-documented and relatively advanced optical experiment with detailed optical illustrations had been performed. Scheiner also provided the first detailed optical description of accommodation, further showing it to be an active process, and noted that the pupil constricts during near vision [Jung 1984]. It is hard not to believe that the obvious convergence that accompanies the near response would also have been observed even if it does not seem to have been found in surviving manuscripts of the time [Jung 1984].
The contributions of Kepler to optics and ScheMer to accommodation and the pupil in many regards mark the beginning of modern visual science, as they directly lead to questions that have subsequently occupied the research field. Notably these are as follows. How is the retinal image kept in focus? Which anatomical structures actually capture the light, and how is such photoreception accomplished? What type of physiological signal does this give rise to? How is this physiological signal processed within the eye? How is the signal transmitted up the optic nerve? What is the signal's destination in the brain? Finally, there is the fundamental underlying question of vision science: how do these physical processes relate to visual perception?
1.4.5.4 Applications of Mathematics to Sensory Systems, Waveforms and Optics
In 1614, in the field of pure mathematics, Napier discovered logarithms. A logarithmic scale is a scale of measurement that uses the logarithm of a physical quantity instead of the quantity itself. This would both aid in more efficiently dealing with data covering a very large range of values by making them more manageable, and also aid thereby in the quantification of the exceptionally broad magnitude of stimuli the sensory system in general was capable of resolving, including both the visual and auditory systems i.e. doubling the input
51 strength adds a constant to the subjective signal strength. Examples such as the Weber-Fechner relationship would use this principle and are discussed later (1.4.5.9). By responding to logarithmic amounts of light (or sound) a dynamic sensory system exists that can receive a great deal of information from the environment favouring survival and success of the organism.
Natural logarithms would tend not to be used in medicine and biology and its associated physical sciences, but logarithms to the base 10 would. In this system the statement 103 =1000, can be expressed instead of '10 exponent 3 equals 1000' as the logarithm (log) of 100 to the base 10 equals 3 (logio100 = 3). Hence each stepwise addition of a 0 to the magnitude of a number represents one log unit change e.g. logio10,000 = 4, log101000 = 3, logio 100 = 2, logio10 = 1, logio 1 = 0, login 0.1= -1. An example of this is in perimetry (measurement of visual field using a curved screen to present stimuli). Values representing light intensity in the non-S.I. units of apostilbs are converted into logarithms. The intensity range of the Humphrey Field Analyzer is 10,000 to 0.1 apostilbs (asb). The term apostilbs pertains to the luminance of a given test target being projected on to the interior of the curved white Ganzfield bowl. Apostilbs (asb) are used in experiments presented in this thesis for quantifying stimulus presentation for electrophysiology and visual fields, areas in which they find routine use in vision science. By converting the values in apostilbs to logarithms and then into another set of units called decibels (which is one tenth of a bel, another non-S.I. unit) the effect is to make these numerical values smaller, easier to deal with, and to give the same results albeit expressed in alternative units. Historically by convention 10,000 apostilbs was by later workers in vision science given a value of zero (0) meaning there is zero retinal sensitivity. This target has the highest intensity - hence if a patient does not respond the scotoma (area of reduced retinal sensitivity) is considered to be an absolute scotoma or blind area. One log unit below the 10,000 asb is 1000 asb and becomes equal to 10 or 10 decibels (dB). The value ten (10) is used instead of one (1) only because it is easier to express changes in whole numbers from zero (0) through ten (10) instead of fractions. The term decibel is hence a relative value which expresses the attenuation in this context from a maximum
52 intensity of 10,000 asb. For these purposes 10,000 asb = 0 dB, 1000 asb = 10 dB, 100 asb = 20 dB, 10 asb = 30 dB, 1 asb = 40 dB and 0.1 asb = 50 dB. It is also possible to hence say that that 1 dB = 0.1 log unit and that 10dB = 1 log unit. There is also a direct relationship with these values used by the Humphrey Field Analyzer and Goldmann field notations. The Goldmann filter steps labelled (a) through (e) are each equal to one (1) dB. Filter steps (1) through (4) are equal to five (5) dB each e.g. step 1112e to II13e is an increase of five dB in brightness and from II12a to 1112b is a one (1) dB increase in brightness. The Humphrey Field Analyzer machine uses a constant target size equal to a Goldmann 'III' and varies the target brightness.
Decibels are also used in other fields including acoustics and electronics. It has great versatility as a logarithmic unit of measurement expressing the magnitude of a physical quantity (usually power or intensity) relative to a specified or implied reference level. As decibels express a ratio of two (same unit) quantities, it is a dimensionless unit. In acoustics it would find use also as there is a massive range of sound pressures to which the ear is sensitive — it helps solve the problem of dealing with very large magnitudes of a stimulus like it does in vision science. However unlike measurements of light intensity in vision science, the decibel used in acoustics to quantify sound levels relative to some zero (0) dB reference implies that with higher values greater loss of hearing occurs — while sensitivity of the eye to light of greater decibels implies less and less visual sensitivity reduction, owing to the convention with regard to apostilbs noted above.
Isaac Newton worked from the mid-17th well into the 18th century in Cambridge and London. Alongside formulating calculus independently alongside Leibniz, and his own mathematical description of gravity, Newton used a prism to show that all spectral colours can be extracted from white sunlight, appearing as in a rainbow [Newton 1730; Sagan 1983; Jung 1984]. Olaus Roemer (1644-1710) of Denmark deduced from telescopic observation of four of Jupiter's satellites that their orbits varied depending on distance from the Earth, the time increasing when the distance increased, and hence that light had a speed which he calculated as velocity =
53 displacement/time [Gregory 1998]. The speed of light is approximately 300,000 km per second in a vacuum and came to be recognised as one of the primary constants of the universe [Einstein 1905 and 1916; Gregor 1998]. Christian Huygens (1629-95) using experiment showed interference of waves (destructive and constructive summation) including light [Gregory 1998]. Waves were described mathematically as sinusoidal with peaks and troughs in their shape algebraically adding or subtracting from each other when they interacted, and with shapes conforming to sine and cosine functions. In addition to the relevance of describing light as a waveform in centuries to come, waveforms would be approximated including that of the pupil's contraction (comprising constriction and dilation phases) to enable analysis as a sinusoidal waveform [Loewenfeld and Lowenstein 1999].
In addition to geometric optics, a second way of understanding the propagation of light was formulated — diffraction integral theory. While the effects of diffraction were observed and recorded by Grimaldi in 1665, it would not be more fully explained till a view became prevalent that light was made not of particles as Newton had believed, but was made of waves as Huygens showed. This occurred after Thomas Young in 1803 performed his double-slit experiments showing diffraction of light by bending through small apertures wherein algebraic addition and subtraction occurred, which is a phenomenon Huygens in the seventeenth century had shown was a property attributable to waves. In the course of time Einstein would show light is in fact understood both as wave and particle - the wave-particle duality of light. In the eye, diffraction at the pupil is a factor contributing to image blur.
Decomposition of a function including those of a square wave in terms of sinusoidal functions (termed basis functions) of different frequencies was enabled by Fourier analysis, as well as their subsequent recombination to obtain the original function. In the course of time application of the mathematics of Fourier analysis in optics enabled a third way of understanding light propagation, termed Fourier optics, which was an
54 addition to diffraction integral theory and geometric optics, and would in time later be complemented by a fourth theory in the form of quantum optics. The plane wave spectrum concept is the basic foundation of Fourier optics and was borrowed from broader electromagnetic theory - the plane wave spectrum is a continuous spectrum of uniform plane waves rather than the curved optical phasefront of geometric optics. In source-free regions (most classical optics pertains to source-free regions), electromagnetic fields may be expressed in terms of a spectrum of propagating, evanescent plane waves. In the twentieth century Fourier optics was applied to lenses, especially how images were effected relative to their focal distances (Fourier transforming property of lenses), yielding a variety of optical devices including spatial optical filters used to clean up lens images produced by lenses, through to computer-generated holograms. Compared with other optical theories there has however been comparatively less application of Fourier optics to practically understanding the functioning of the eye including image formation using the cornea, lens, pupil and retina.
1.4.5.5 Mechanistic Thinking
As previously alluded to, by the seventeenth century mechanistic thinking had grown in its scope for application. This was underscored by the description of the body as a machine in the seventeenth century by Descartes. His Cartesian principles are, partly, still valid. This was that "esprits animaux" or "spiritus animales" (which would in the 19th century be modified to the concept of nerve impulses) originated in the sense organs such as the eye and were conveyed via the nerves to the brain, which would in turn convey messages from the central nervous system to the effector organs [Jung 1984]. Descartes noted that there is multisensory convergence to the brain and conscious attention is essential for perception and action - sensorimotor reflexes and co-ordination involve afferent components, integration in the CNS, and efferent components. In the context of the pupil, in 1664 Descartes noted that the pupil was not under direct voluntary control — mental effort alone could not cause pupil constriction, though
55 shifting gaze from a far point to a near point could, in other words, he had described accommodation [Loewenfeld and Lowenstein 1999]. Also in 1664, Willis traced the optic tract to the thalamus hence anatomically linking the visual output to the brain's relay system. By 1681 Mariotte had noted the physiological blind spot and also articulated a basic trichromatic theory of colour vision, and by 1684 Leeuwenhoek became the first to record an observation of the retina using a microscope noting structures now known as rods and cones [Boring 1942; Polyak 1957; Duke-Elder 1958; Duke-Elder and Wybar 1961; Duke-Elder et al 1968; Brindley 1970; Duke-Elder and Scott1971].
1.4.5.6 Reflexes as Involuntary Motions, Awareness of the Function of the Pupil
1.4.5.6.1 Advances in Reflex Autonomic Physiology and Anatomy
Albrecht von Haller stressed that many movements are involuntary, including autonomous activity of the heart, gut and reproductive system. The Scottish physician William Porterfield (1696-1771) knew of the work of his Edinburgh colleague Robert Whytt (1714-1766) who recognised an afferent and efferent limb in separate nerves for the pupil light response [Thompson 2003]. Porterfield also invented the first optometer, and tested the presence of accommodation after cataract surgery [Porterfield 1759].
1.4.5.6.2 Renewed Awareness of the Relationship between the Pupil and Visual Impairment by Physicians
The pupil became once again a key focus in the clinical examination of the eye. Charles de Saint-Yves wrote in his 1722 textbook 'New Treatise on the Diseases of the Eye', which remained in print for more than 80 years, on how to assess degrees of visual loss from the pupil's mobility:
56 "I have noticed over and over again in my patients that the extent of the visual impairment closely matches the impairment of iris movement. In fact I have found that, without talking to the patient about the visual problem, I have been able to make a fairly good estimate of the quality of the patient's vision, based only on my examination of the pupillary movement." [Saint- Yves 1722].
1.4.5.6.3 Functions of the Pupil
1.4.5.6.3.1 Historical Awareness of Basic Pupillary Functions
It is thereby clear from the aforementioned discoveries and statements in the literature that the function of the pupil had been essentially understood by science and clinical medicine by the eighteenth century, and certainly by the time of Porterfield's work on cataract surgery and accommodation in 1759 [Porterfield 1759]. Biological concepts of the pupil came to be used effectively to model concepts in optics where the f-number (focal ratio, f- ratio, or relative aperture) of an optical system is an expression of the diameter of the entrance pupil in terms of the effective focal length of the lens - the f-number is the focal length divided by the aperture diameter and is a dimensionless number that is a quantitative measure of lens speed, which became an important concept in photography. A small f-number (contracted pupil) is associated with increased depth of field (the tolerance to placement of the object plane). The pupil also increases through its contraction the depth of focus (the tolerance to placement of the image plane - biologically the retina). The following five functions of the pupil are in fact recognised [Westheimer 1964; Campbell and Green 1965; Agarwal 1997; Loewenfeld and Lowenstein 1999; Kardon 2003].
57 1.4.5.6.3.2 Retinal Illumination and Photopic Homeostasis: Pupil and Eyelid
Sight requires transparent ocular media (the precorneal tear film, cornea, aqueous humour, lens, vitreous humour) [Robinson and Fielder 1992; Fielder and Moseley 2000]. In addition to the more static media, light flux is also regulated by the dynamic gross anatomic motion of the eyelids which is predominantly a red pass filter (mean transmissions for 700 nm are 14.5% in the adult and 21.4% in the neonate, and less than or equal to 3% in both adults and neonates below 580 nm) with finer adjustments made by the pupil [Robinson et al 1991; Robinson and Fielder 1992; Fielder and Moseley 2000]. The pupil controls moment-to-moment levels of retinal illumination and thereby influences all light-dependent ocular functions [Robinson and Fielder 1992; Fielder and Moseley 2000]. By responding to changes in retinal illumination per unit area of retina, the pupil's motions represent an example of homeostasis. The latter is a fundamental physiological concept, understood as dynamic auto-regulation by the body to maintain internal parameters at a more or less constant level. Hence the pupil regulates photopic homeostasis through maintaining constant illumination of the retina during steady-state environmental illumination, and during sudden changes in environmental illumination reduces the effect of change in light intensity; it also simultaneously facilitates dark and light adaptation by, respectively, increasing and decreasing in size according to the relative state of retinal illumination. This is a classical negative feedback loop and originates in retinal photochemical and neuronal mechanisms, culminating in contraction or dilation of the iris as necessary. As an example of its importance, patients with aniridia, who all but lack the necessary effector elements in the iris, complain of marked photophobia as, having a huge pupil, their retina is subjected to a greater intensity of light that is also spread across a greater retinal area causing severe discomfort and pain on exposure to bright light.
58 1.4.5.6.3.3 Reducing Spherical Aberration
By reducing the proportion of light rays transmitted through peripheral cornea and lens the pupil reduces optical aberrations due to spherical aberration. Thus in conditions of mydriasis such as during dark adaptation, optical aberrations and their effects such as glare increase, for example during driving in the dark at night.
1.4.5.6.3.4 Reducing Chromatic Aberration
By reducing the proportion of light passing through peripheral cornea and lens the pupil reduces chromatic aberration. The pupil contraction and its associated small f-number increase depth of field producing less chromatic aberration. The pupil may also reduce the effect of chromatic aberration in the optical system of the eye by reducing transverse chromatic aberration (the latter optical effect is discussed more generally in chapter six).
1.4.5.6.3.5 Overcoming Ametropic Errors with the Eyelids
The pupil if markedly constricted can, alongside marked contraction of the orbicularis oculi muscles and narrowing of the palpebral aperture, convert the eye into a pinhole camera. In doing so errors of ametropia can be partially compensated for.
1.4.5.6.3.6 Depth of Focus
The pupil serves to increase depth of focus. The response to a near target (miscellaneously also termed the near response, response to near, or near triad) involves miosis, convergence and accommodation. The pupil increases the distance an object can be moved in space without image blur. This also is part of the pinhole effect.
59 1.4.5.7 Nineteenth Century Studies of Reflex Physiology in General, and of the Pupil in Particular
1.4.5.7.1 Overview
In anatomy Meckel in a landmark paper in 1820 confirmed using the microscope the existence of radial and circular muscles fibre layers in the iris; he also noted the intra-orbital parasympathetic supply to the eye including the pupil, describing the ciliary ganglion and its associated long and short ciliary nerves [Meckel 1820; Loewenfeld and Lowenstein 1999]. Pavlov reinforced study of reflexes in general by his celebrated work on dogs showing conditioned reflexes. Between 1811 and 1827 Charles Bell and F. Magendie, surgeon-anatomists in London and Edinburgh, experimentally proved the existence of spinal reflexes. In the mid-19th century, the most renowned of the 19th century German physiologists, Hermann von Helmholtz, understood modern concepts of reflexes as stereotyped responses to sensory inputs [von Helmholtz 1909-1911; Jung 1984]. In the twentieth century the pupil would be categorised as a prime example [Loewenfeld and Lowenstein 1999].
It is important to note in this context that the direct clinical study of the pupil nevertheless declined in the 19th century [Thompson 2003]. This may have been a price for the surge of interest with the new technology of the direct ophthalmoscope, invented by von Helmholtz in 1850 [Keeler 2002]. A notable exception was Albrecht von Graefe of Germany who cautioned much older ophthalmologists from dilating pupils at the expense of missing pupillary signs, though according to Thompson, von Graefe also was not confident in trusting the pupil signs he elicited [Thompson 2003]. von Graefe's account is also the first record of a physician showing interest in the pupil as an indicator of the conversion state of functional blindness or hysteria, and in malingering [von Graefe 1855; Thompson 2003]. von Graefe also postulated that the pupil response to light would survive in cortical blindness, a fact confirmed by later anatomical studies showing the
60 midbrain to be crucial to the pupil response to light [von Graefe 1855; Loewenfeld and Lowenstein 1999; Thompson 2003]. Thompson has postulated likely reasons for 19th century ophthalmologists treating pupil testing with some caution, and while his reasons are somewhat anecdotal they are based upon his own extensive experience and seem logical. They are discussed further below [Thompson 2003].
1.4.5.7.2 Patients with Good Vision but Pupils Poorly Reactive or Unreactive to Light
Reasonably good vision is possible with a unilateral fixed pupil, although many such cases will complain of imperfect vision. This situation is found with efferent lesions which denervate or mechanically interfere with the movement of the iris, such as trauma (especially traumatic mydriasis), especially if in younger patients, and will usually cause visual complaints at near, but so may iritis with posterior synechiae (inflammatory adhesions between iris and lens). Other causes include third nerve palsies, Holmes- Adie pupils where there is dysfunction in the ciliary ganglion, and pharmacological mydriasis such as atropinisation, all resulting in a pupil with a weak or absent light reaction on the affected side, and hence inter- eye pupillary inequality that increases with brightness of the stimulus as the affected pupil fails to constrict appropriately. By the mid-19th century it was recognised that most such patients had reasonable vision and testing the pupil could be misleading unless the examiner was aware of these factors [Thompson 2003].
1.4.5.7.3 Peripheral Field Loss
A further possibility is excellent visual acuity and poor pupil responses due to gross peripheral field loss. The relationship between area summation and pupil reactions would be discerned in the 20th century in detail. Essentially the pupil responds to changes in the amount of retinal area stimulated, hence peripheral retinal lesions which wipe out large areas of retina can
61 have a marked influence on the pupil response yet leave visual acuity, dependent on central retinal function, unimpaired [Kardon 2003].
1.4.5.7.4 Patients with Poor Vision and Excellent Pupil Reactions to Light
Such eyes may even have been called blind, but were not without perception of light e.g. functional cases, large central scotomata or deep suppression amblyopia. The latter phenomenon while noted was not well- understood by the medical profession till work by Hubel and Wiesel in the basic sciences, and was not subject to a randomised control trial for treatment with patching of the opposite eye till 2004 [Hubel and Wiesel 1962 and 1965; Clarke et a12003; Clarke 2006]. In all these cases the majority of retinal cells were functional and connected to the midbrain.
1.4.5.7.5 A Truly Blind Eye with a Pupil Apparently Constricting to Light Shone into it
Thompson notes this may have been "eyelid-closure pupillary constricton" (a stray near response) that was missed by the examining clinician [Stedman 1995; Thompson 2003]. Alternatively he feels it could be light from a bright light source (taken, the slit lamp was not invented till 1911) shone onto a blind eye reflected off the patient's face and then the doctor's white coat and thence into the opposite eye, which is functional, dark- adapted and thus highly sensitive to light — a consensual response is thus observed but mistaken for a direct response [Thompson 2003]. It may also be that a false positive is observed due to the phenomenon of paradoxical augmentation of the PLR by cataract in the opposite eye [Kardon 2003]. This gives the impression on comparison of the responses between the two eyes that there is greater response in the opposite eye with a cataract, although the visual potential in both eyes is the same.
62 1.4.5.8 Colour Vision, Cellular Concepts and Colour Channels
Understanding light as a waveform means that each colour corresponds to a given frequency [Gregory 1998]. As previously alluded to, in 1681 Mariotte had proposed a trichromatic theory of colour vision, but it was not designed to explaining this implication of light understood as a waveform. A paradox is understanding how a relatively slow-acting nervous system can resolve the different light frequencies, especially the higher frequencies in the blue end of the spectrum. Thomas Young postulated an explanation that in the course of time became established as the trichromatic theory of colour vision [Young 1802; Jung 1984; Gregory 1998; Foster and Hankins 2002]. In 1802 Young published his theory that had been composed the year before, writing:
"Now, as it is almost impossible to conceive each sensitive point of the retina to contain an infinite number of particles, each capable of vibrating in perfect unison with every possible undulation, it becomes necessary to suppose the number limited, for instance, to the principal colours, red, yellow and blue" [Young 1802].
Maxwell in Britain and von Helmholtz in Germany independently resurrected Young's theory in 1850 [MacAdam 1970; Jung 1984; Yellott et al 1984; Schwartz 2004). James Maxwell would write 70 years later:
"It seems almost a truism to say that colour is a sensation; and yet Young, by honestly recognising this elementary truth, established the first consistent theory of colour. So far as I know, Thomas Young was the first who, starting from the well-known fact that there are three primary colours, sought for the explanation of this fact, not in the nature of light but in the constitution of man" [Gregory 1998].
As noted more recently Mariotte has been credited with part of this theory. Young chose three primary colours as experimentally it is possible to
63 produce any colour visible in the spectrum, as well as white, by a mixture of not less than three overlapping lights of different colours, and, further, as the wavelengths required are quite broad, the identity of the three principal colours could thereby be suspected [Gregory 1998]. Half a century later the German physiologist Hermann von Helmholtz refined Young's colour mixing experiments to produce the modern Young-Helmholtz theory of trichromatic vision wherein three groups of retinal receptors are postulated to respond maximally to red, green and blue light [Jung 1984]. Colour solids such as triangles were introduced to predict the resultant colours that colour mixing would produce — such aids rely on the Young-Helmholtz principle representing colour space arranged within the three primary colours [Griisser 1989; Gregory 1998].
At present there is no comprehensive theory of colour vision despite much psychophysical and physiological work [Grosser and Griisser-Cornehls 1973; Griisser 1989; Miller and Newman 1999; Schwartz 2004]. In addition to the trichromatic theory there is the opponent-process theory of colour vision. Originally opponent colours were proposed in the wrong context by Ewald Hering in six "Gebenfarben" (white-black, yellow-blue and red- green) in a re-statement of an older mystical theory that black is a non- chromatic colour [Jung 1984]. The modern theory proposes two chromatic mechanisms (red-green and blue-yellow) and one achromatic mechanism (black-white) [Miller and Newman 1999]. It is said to thereby account for six distinct colour sensations: blue, green, yellow, red, black, and white.
Neither theory is incompatible with the other. For example physiological and psychophysical work have established that the output from the three cone photoreceptors occurs are spectrally tuned to the red, green and blue wavelengths (trichromacy). The output is transmitted on to neurones in the inner retina including ganglion cells according to two opponent mechanisms: blue-yellow cells and red-green cells and one achromatic mechanism [Miller and Newman 1999]. Opponent cells each have centre- surround receptive fields that enhance the efficacy of opponency [Gregory
64 1998]. Spatial and temporal processing of this information occurs at different levels of the visual pathways, including the lateral geniculate nucleus (LGN), but this opponent coding of colour information reaches the higher centres also.
Psychophysical work with colour-deficient subjects suggests that at least two cone receptor classes are needed for colour vision (the Principle of Univariance). Rods are felt to mediate achromatic vision, hence subjects with rod monochromatism see the world in shades of grey. As mentioned, colour perception is not only dependent on stimulus wavelength and the nature of photopigments but on processing by extensive neural systems. The two major colour categories are the very wide range of chromatic valencies (red, green, blue etc) and the significantly more limited achromatic valencies (grey and its shades from darkest black to brightest white) — in total approximately 7 million colour valencies [Grosser 1989]. The chromatic valencies can be described using three phenomenological attributes: hue, saturation and brightness for luminous sources (lightness for surfaces). Hue is closely related to wavelength and is a pure colour, of which a continuum exists from red to yellow to green to blue to purple back to blue, with all colours in-between. However unlike other colours, there is no wavelength for purple, emphasising that hue is not wavelength, but related intimately to it, and is wavelength combined with some aspects of neural processing not yet fully elucidated. Saturation of a shade is a measure of the ratio of chromatic mixing with achromatic valencies. Brightness, or lightness, is determined by the amount of grey, from brightest white to darkest black.
Various chromaticity diagrams taking the form of colour solids (charts) have been designed historically, from spheres to the modern format which is commonly a horse-shoe shape [Norton et al 2002]. They provide a two- dimensional representation of the complete range of colour stimuli; in them rectilinear co-ordinates differ depending on the selected primaries used to make up a stipulated colour [Norton et al 2002]. All trichromatic co-
65 ordinates for red, green and blue lie within and not on any part of the colour diagram. They enable colour matching — the act of mixing colours to achieve the same hue, saturation and brightness as a sample colour [Norton et al 2002]. In the terminology of colour matching, a metamer is one of a set of colour stimuli with identical colour appearance (tristimulus values) but different spectral composition, while an isomer is one of a pair of colour stimuli with identical colour appearance and wavelength composition; note this is different to chemistry where the term isomer refers to compounds with the same molecular formula but different structural formula and where shared properties between isomers only occurs if their molecular structures happen to share certain functional groups [Norton et al 2002; Schwartz 2004]. Using the modern chromaticity diagram values along the red-green axis (abscissa) are determined only by the relative amounts of long-wavelength (L) and medium-wavelength (M) cone activation. It this system it is also theorised that the tritan axis values (ordinate) are determined by short-wavelength (S) cone photoactivation and in this axis red and green responses are constant. In some charts a curved tritan confusion line is used and is said by proponents to offer a more accurate representation of some inadequacies in the psychophysical measurement of blue during colour matching, as discussed extensively by Tregear, Ripley, Knowles and colleagues [Tregear et al 1994].
It should be noted that these models are an imaginary system devised from psychophysical data on colour perception with which it corresponds using the process of colour matching [Wyszecki and Stiles 1982]. Chromaticity diagrams including the widely used Commission Internationale de l'Eclairage (CIE) chromaticity diagram employ imaginary primaries — mathematically transformed real primaries (an unreal colour) that serve to locate primary reference points in the diagram [Norton et al 2002]. Using a mathematical analysis involving calculating the 'dominant wavelength' and `excitation purity' from the diagram, colours can be matched. In general colour-matching functions are usually taken to mean the tristimulus values corresponding to monochromatic stimuli of equal radiant power. The three
66 values of a set of colour-matching functions at a given wavelength are called the colour-matching coefficients. The colour-matching functions may be used to calculate the tristimulus values of a colour stimulus from the colour stimulus function. Hence in typical chromaticity charts as the red- green axis symbolises red-green opponency, movement to the left is taken to correlate with M-cone excitation, and movement to the right L-cone excitation. Likewise, upward movement signifies increased blue cone excitation, and downward movement of the cursor a reduction in S-cone excitation.
The term colour channels was originally introduced to distinguish the pathway for the luminance response from that for the colour response. The term evolved out of cybernetics to describe distinct information-transfer systems with properties such as saturation (maximum channel capacity) [Longtin et al 1990; Loewenfeld and Lowenstein 1999]. Separate colour channels exist for visual perception in the retina, optic nerve and brain. The pupil also responds to colour independently to its luminance response. The response is much smaller. Colour information is carried by the parvocellular pathway which also carries information on high spatial resolution. No information on colour is carried by the magnocellular pathway which primarily codes for contrast and motion detection.
A variety of tests for defects of colour vision exist. The most commonly used clinically are pseudoisochromatic plates (most commonly in the United Kingdom those devised by Ishihara of Japan) though these do not readily test for defects of blue-yellow vision and are mainly a test of red- green colour loss; the test may be followed in more complex cases by Farnsworth-Munsell 100 hue testing. Both the latter stipulate optimum background lighting which is hard to achieve all the time in a clinical situation require though this does not stop them being used routinely since they still provide valuable and usually accurate data on colour blindness. The Lanthony Desaturated 15-Hue test is useful in overcoming many of these problems. The Farnsworth D-15 colour test is also used [Norton et al
67 2002]. A variety of other tests exist [Schwartz 2004]. Tritan discrimination sensitivity allows measurement of loss of blue vision [Tregear et al 1994].
1.4.5.9 Advances in Studying Sensory Receptors
In 1825 Bohemia the physiologist Johannes Purkinje described a spectral shift of peak vision towards shorter (blue) wavelengths in the dark, later named after him as the Purkinje phenomenon [Jung 1984]. In the same year Johannes Muller, titled the "father" of modern physiology, proposed the Law of Specific Energies (or Qualities) that a nerve fibre can only signal one kind of quality [Muller 1826 and 1838; Jung 1984; Gregor 1998]. This meant that specific anatomic channels existed for different primary colours and brightness. In 1838 the London physicist Wheatstone experimentally showed that retinal disparity is the basis of binocular stereoscopic vision [Wheatstone 1838]. Ernst Weber in Leipzig introduced the concept of receptive fields for individual cells [Jung 1984]. In 1834 he proposed Weber's Law which states that the smallest difference in stimulus intensity which can be detected is directly proportional to the background intensity [Jung 1984; Gregor 1998]. This gives a value for residual activity in nerve cell firing in terms of equivalent light intensity called k [Gregory 1998]. This gave rise to the concept of internal residual activity in the absence of stimulus in the sensory system. Thus neuronal firing could occur below stimulus threshold. This effect was compared to the "noise" seen with mechanical recording devices. Such noise was especially observable in the pupil at rest.
Ernst Weber and later Gustav Fechner in Germany effectively founded modern psychophysics by postulating a general logarithmic law of perceptual intensity. Fechner applied and extended Weber's work more specifically to visual perception. By this law it was understood that humans could perceive immense changes in sensory stimuli, including
68 environmental light intensity, the range of which was so vast it even required logarithmic increments to quantify as follows:
S = k. log I
This equation is known as Fechner's logarithmic function, where S = sensation, I = intensity, also known as the Weber-Fechner Law or relationship.
In fact different linear, logarithmic and power functions can be used to describe this phenomenon (e.g. Plateau's and Stevens' Laws for magnitude scaling); the modern expression of this is as Ricco's law of spatial summation [Jung 1984; Fischer and Freund 1970].
During the twentieth century vision science would became an umbrella discipline within academia in its own right, and would become the subject of intricate study by innumerable workers from many fields [Jung 1984; Foster and Hankins 2002; Peirson et al 2005; Foster et al 2007].
1.4.6 Structure of the Retina, Photoreceptors and Photopigments
1.4.6.1 General Topography of the Retina
As long ago as the eighteenth century the macula, owing to its yellow pigment, which is especially intense around the fovea, was noted in human and monkey retinas to form a more or less central pit indenting the vitreal surface, called hence the "yellow spot" or "macula lutea", and later often simply the macula [Polyak 1941; Delori et al 2001; Penfold and Provis 2005]. The mid-nineteenth century invention of the ophthalmoscope facilitated in vivo study of retinal structure in more detail [Jung 1984]. This is summarised in figure 1.3.
69 Cajal used the Golgi impregnation method to considerably further understanding of retinal cellular anatomy and connectivity [Cajal 1892]. By 1924 the fovea had come to be recognised as the site mediating maximal visual acuity as evinced by von Helmholtz's appreciation of this fact in the course of composing his theory of lens accommodation wherein the ciliary muscle's contraction releases tension on the zonular fibers at the lens equator, allowing the lens equatorial diameter to decrease, the lens thickness to increase, and the lens anterior surface to become more steeply curved [von Helmholtz 1909-1911; Vilupuru et al 2004].
The Retina
Nerve Nuclei
Fibres ppm= NM MM. Rods Light
Ganglion Horizontal woods Cells Figure 1.3 Some of the key cells in the retina of special interest to this work. Ganglion cells arise in the retina and traverse the orbit in the optic nerve to the brain. Giant ganglion cells are of special interest to this work as putative photoreceptors in humans.
1.4.6.2 The Electroretinogram and Classic Photopigments
In 1865 the Swedish physiologist Alarik Holmgren discovered the electroretinogram (ERG) thus allowing detailed and relatively non-invasive physiological study of retinal cells by measuring the electrical changes produced by the retina in responses to flashes of light. Bleaching of a chemical by light in the eye was demonstrated by Boll in 1876 [Boring 1942; Polyak 1957; Duke-Elder 1958; Duke-Elder and Wybar 1961; Duke-
70 Elder et al 1968; Brindley 1970; Duke-Elder and Scott1971; Penfold and Provis 2005]. This corresponded to a period of temporary functional blindness followed by restitution of vision [Jung 1984; Gregory 1998]. This chemical was described as the photopigment visual purple (later known as rhodopsin) by Kuhne in Heidelberg in 1877 [Ki.ihne 1877 and 1879; Jung 1984]. Rod visual pigment, or rhodopsin, is called visual purple as in the dark it is dark red as rhodopsin absorbs relatively much more green and blue light than it does red [Grosser 1989].
1.4.6.3 Rods and Cones
Max Schultze (1825-1874) both discerned the structure of rods and cones microscopically and related them functionally as cones mediating colour vision and rods mediating night vision [Jung 1984]. In the late 19th century Johannes von Kries elaborated the anatomical distinction between rods and cones into the duplicity theory of vision with two systems: a cone-mediated perceptual system active predominantly in light and a rod-mediated visual system active predominantly in darkness [Jung 1984]. Cone adaptation is completed in seven minutes, while rod adaptation takes one hour or more [Gregor 1998]. These classic photoreceptors, the rods and cones, are so- named for their shape, with cone density generally increasing towards the macula [Osterberg 1935]. There are approximately 92 million to 120 million rods, 4.6 million to 6 million cones and 1.07 million ganglion cells transmitting their output to the brain in the optic nerve per human eye (range 710,000 to 1.54 million) [Penfold and Provis 2005]. Hence there is a convergence of electrical impulses conveying information upon the ganglion cells [Osterberg 1935; Curcio and Allen 1990; Marieb 2004; Penfold and Provis 2005]. The retina is anatomically subdivided into 10 layers, from backward forwards the retinal pigment epithelium (RPE), then the layers of rods and cones, then the inter-neurones including bipolar cells, amacrine cells and horizontal cells, and glial Muller cells that are supportive, while anteriorly and in first contact with light are the layers of ganglion cells - the ganglion cell layer composed of cell bodies, followed
71 most anteriorly by the nerve fibre layer of ganglion cell axons that are the neuronal outflow projecting in the optic nerve at the optic nerve head (optic disc), to the brain [Gray et al 1995]. Ganglion cells vary dramatically in size and include a subset of morphologically giant ganglion cells, while rod and cone photoreceptors also vary, but less so, generally having widths in the region of 1-311m and lengths of 30-80µm, depending on location in the retina, but there is some normal variation beyond these dimensions also.
The RPE serves to nourish rod and cone outer receptor segments [Penfold and Provis 2005]. Separation of the rods and cones from the RPE along the embryological cleavage plane between the posterior RPE and the anterior neurosensory retina occurs in retinal detachment, causing outer photoreceptor damage and even photoreceptor necrosis by 4 to 6 weeks and according to some sources some damage beginning almost immediately [Hogan and Zimmerman 1962; Apple and Rabb 1998; Marieb 2004]. The fovea centralis (fovea) lies on the temporal side of the optic disc within the macula (area centralis) which is demarcated by the supero- and infero- temporal branches of the central retinal artery and vein that divide at the optic disc [Curcio et al 1990; Hendrickson 1994; Penfold and Provis 2005]. The fovea lies on the visual axis of the eye slightly inferior to the anatomical axis and in the adult human eye it is 4 mm temporal and 0.8 mm inferior to the centre of the optic disc [Penfold and Provis 2005]. It occupies 0.02% of the total retinal area [Curcio et al 1990; Hendrickson 1994; Penfold and Provis 2005]. The fovea is found in almost all primates, including humans, all Old World monkeys, and in all New World monkeys except the nocturnal owl monkey Aotes [Penfold and Provis 2005]. Despite fivefold variations in the dimensions of the eye and retinal area, taking into account slight intra-subject variability, foveal dimensions are relatively constant between primates, suggesting similar basic cellular mechanisms operate in all of them [Hendrickson 1994; Penfold and Provis 2005]. The macula is divided into four concentric zones as described below [Penfold and Provis 2005].
72 i. The centralmost zone or foveola (diameter 250-350µm, symbolising the central 1° 20 s of visual field). This is flattened relative to the rest of the fovea, forming the pit region visible to eighteenth century observers. It contains the highest cone density in the retina, there being no rods, and in the epicentre of the pit the thickness is barely over 100 µm (the thinnest part of the retina) - this centralmost part of the foveloa comprises purely L- and M- cones to the exclusion of S-cones; these L- and M-cones are surrounded by Muller cell processes [Burris et al 2002]. Moving away from the epicentre, but still in the floor of the fovea! pit, scattered ganglion cells are occasionally found [Rohrenbeck et al 1989; Curcio and Allen 1990]. Clinically a white reflex overlying the pit on ophthalmoscopy identifies the pit in healthy young subjects. However cone outer segments are longest in the foveola, forming an inward indentation of the cone nuclear layer, and the other cellular layers, called the fovea externa [Curcio et al 1990; Hendrickson 1994]. The foveola is avascular, comprising the foveal avascular zone (FAZ). It contains all three types of cone: short wavelength (S-), medium wavelength (M-) and long wavelength (L-) cones, but as noted above S-cones are absent from the central 100 microns; peak cone density is not in the centre of this area that is absent of S-cones but slightly eccentric to it [Curcio et al 1991; Hendrickson 1994]. Centralmost absence of S- cones in the retina has two purposes. First it enhances visual acuity while not fully compromising colour vision. This is as the peak spectral sensitivities of L- and M-cones are far closer to each other than to S-cones and the former two classes of receptor also have broader spectral sensitivity profiles. Hence visual acuity (including resolution) is increased while colour vision is still somewhat preserved by L- and M-cone activity, albeit with relatively reduced emphasis on blue light at least as far as the outer retinal elements are concerned. In this context work towards the end of this thesis (chapter 11) showing photoreceptive ganglion cell activity in the
73 inner retina with peak spectral sensitivity to blue light together with associated sight may point to a possible mechanism for compensation for this reduction of blue light sensitivity in the centralmost outer retina by mechanisms in the inner retina. The second effect of absence of S-cones in the centralmost retina is to tend to reduce the effects of chromatic aberration. ii. The foveal slope and immediate surrounds, or non-foveolar fovea (diameter 1.85 mm, or 750 vm adjacent to the edge of the foveola, and symbolising the central 5.5° of visual field) [Curcio and Allen 1990; Curcio et al 1990 and 1991; Hendrickson 1994]. This comprises progressively more of the typical inner retinal layers on moving further out in the retina, and the retinal slope is in fact the thickest part of the retina, with a layer of ganglion cells up to eight deep and a very thick outer plexiform layer. Rods start to appear in the foveal slope. They have the greatest outer segment diameter of any rods in the retina. iii. The parafovea surrounds the fovea and is about 500 irm wide) [Curcio and Allen 1990; Curcio et al 1990 and 1991; Hendrickson 1994]. Rods start to become more numerous relative to cones in this zone, and this trend continues moving out in the retina; ganglion cell density decreases to four cells thick at the parafovea's outer edge. iv. The outermost zone of macula is called the perifovea, which is 1.5 mm wide) [Curcio and Allen 1990; Curcio et al 1990 and 1991; Hendrickson 1994]. At its outer edge, which nasally almost reaches the optic disc, the ganglion cell layer is reduced to just one cell thick.
Despite the marked tendency for cones to be found in greater proportion in the macula, more cones are still found outside the macula. As the fovea is so small, only 0.3% of cones are found in it, though it contains 25% of ganglion cells showing the fovea's massive importance to vision [Curcio and Allen 1990; Curcio et al 1990; Penfold and Provis 2005]. Ganglion
74 cells tend to pass superior or inferior to the foveola so as not to impede the visual responsiveness of the foveola. At the disc they become myelinated. In distinction to cone density, rod density is greatest superior to the optic disc in the "rod hot spot" at approximately 20° where rod density peaks in a concentric ring around the fovea [Curcio and Allen 1990; Curcio et al 1990 and 1991; Hendrickson 1994; Penfold and Provis 2005]. Both rod and cone size vary similarly. Overall total length of foveal cones is 50 — 70 p.m, width 1.2 [tm (outer segments) to 3µm (maximal inner segment diameter). Cone length decreases rapidly and width increases with distance from the foveal centre.
A further consequence of the topography of cones is the Stiles—Crawford effect — a consequence of the directional sensitivity of the cone photoreceptors [Schwartz 2004]. Rays of light passing through the centre of the pupil are less oblique to a cone after refraction and cause greater stimulation than rays passing through the pupil more peripherally. Cones capture that light best which hit it at a narrow angle from its normal. The acceptance angle of a cone is fairly narrow at approximately 5'; rods have larger acceptance angles. The Stiles-Crawford effect reduces the detrimental effects of light scatter on the retina during photopic conditions in which cones are most active.
1.4.6.4 Insights from Disease
Kollner's rule, published in 1912, was an attempt to correlate colour vision defects with anatomical location of known photoreceptors within the retina [Miner H 1912; Schwartz 2004]. Originally stated this was that outer retinal and media disorders produced blue-yellow colour vision defects, while red-green defects in sight were produced by diseases of the inner retina, optic nerve, visual pathways and visual cortex. It has subsequently become adapted over the years to state that in acquired dyschromatopsia, acquired blue-yellow or pure blue (tritan) dyschromatopsias are produced by retinal diseases, and acquired red-green dyschromatopsias (protan and
75 deutan, respectively) by diseases of the optic nerve or beyond [Miller and Newman 1999]. Amongst media opacities nuclear sclerotic cataracts tend to produce blue-yellow defects (by absorbing blue light and transmitting yellow), but other cataracts may not [Schwartz 2004]. Kollner's rule is known to have exceptions [Miller and Newman 1999; Schneck and Haegerstrom-Portnoy 1997]. Nevertheless it is a worthwhile approximation in many cases [Hart 1987]. Exceptions include blue-yellow or tritan defects in many patients with dominant optic atrophy and glaucoma, protan deficits in serous retinal detachments, and a mix of loss of red-green and blue- yellow deficits in acute optic neuritis [Miller and Newman 1999]. Congenital colour vision defects are red-green (affecting 8% of males, 0.4% females), blue-yellow and extremely rarely congenital tritan (blue) deficits.
1.4.7 Quantum Theory and Biological Effects of Light on the Eye
1.4.7.1 Quantum Theory
Further understanding of the nature of photopigments came from advances in quantum physics. Evidence exists that a quantal or particulate nature to light had been known of even in ancient times [Sarton 1927; Jung 1984; Mirza 1997]. By 1758 the philosopher Immanuel Kant had published a pre- Einsteinian idea of relativity of movement and stated that the stationary state was a small motion; in the second edition of his Critique of Pure Reason in 1781 he underlined the necessity to deduce such facts through experiment [Kant 1783; Jung 1984]. The study of quantal phenomena was revitalised in 1900 when Max Planck made the assumption that emission and absorption of energy can only occur in discrete amounts [Chanda 1997; Capra 2000; Halliday et al 2002]. Planck called these amounts quanta or packets of energy; E = hn describes the relationship between energy and light, where E (energy) is described by Planck's constant (h) and the frequency of a wave (n, also written as v). In 1905 Albert Einstein published the Special Theory of Relativity about the nature of light. In addition to the celebrated equation E = mc2 (E = energy, m = mass, c
76 speed of light] he proved by experiment that in its nature light behaved not only like a wave but like a stream of particles (or quanta) later to be called photons [Einstein A 1905 and 1916]. Niels Bohr in 1913 stated that electrons tend to occupy the lowest energy level. Transitions between energy levels can result in the emission of photons (a principle powerfully applied in lasers when this process occurs in mass synchrony during population inversion of the laser gain medium) [Halliday et al 2002]. Einstein later expanded important concepts in his 1915 General Theory of Relativity [Einstein A 1916]. Einstein received the Nobel Prize for a separate topic- his description of the photoelectric effect, which through showing that light shone on a metal surface released electrons from the metal, causing him to resurrect the old Newtonian view that light could be regarded as a particle. The wave-particle duality in the nature of light became recognised. The effect should not be confused terminologically in the medical literature with the piezoelectric effect - the ability of some materials, usually crystals and certain ceramics, to generate an electric potential in response to applied mechanical stress, and which is used in phacoemulsification probes in cataract surgery.
Light also came to be recognised by the early twentieth century as that part of the electromagnetic spectrum that is visible and thus detectable by human photoreceptor systems, which is between approximately 400 nm to 700 nm, though in some human subjects may be from 380 nm and potentially up to 800 nm (figure 1.4). The various parts of the electromagnetic radiation, which differ in wavelength and frequency, all consist of fluctuating electric and magnetic fields. In increasing order of frequency these are radio waves, microwaves, infra-red radiation, visible light (red through violet), ultra-violet, X rays and gamma rays.
77 Figure 1.4 The visual portion of the electromagnetic spectrum.
1.4.7.2 Ramifications for Understanding Visual Perception
Implications of the popularisation of quantum theory included:
i. Pupils came to be regarded as in-built light meters and were used in various areas of visual physiology for objective measurement of function [Loewenfeld and Lowenstein 1999]. Pupillary responses to colour were understood in terms of their fundamental energies.
ii. The chromophore (light-absorbing portion of a photopigment) which drives visual and accessory visual functions became perceived as an absorber of photons [Duke-Elder 1932; Norton et al 2002]. Besides rod and cone opsins, there is the melanopsin- based photopigment. This has been suggested to be an environmental brightness detector [Foster RG, personal communication]. An interaction with photons is cardinal.
iii. Recognition of quanta as a factor that both facilitated and limited appreciation of detail (visual acuity and pattern) [Gregory 1998].
iv. A wide appreciation of the concept that packets of energy and particles are important to biological functions in general. In 1905- 06 Langley wrote of neurochemical transmission in terms of a "receptive substance" [Langley 1905-6; Loewenfeld and Lowenstein 1999]. Later in the twentieth century Sir Bernard Katz
78 of University College London wrote of neurochemical transmission in terms of quantal events [Katz 1952 and 1966]. The word quantum did not just equate with photons but with fixed packets or particles including neurotransmitters like acetylcholine which are crucial to function within the autonomic nervous system.
1.4.7.3 General Effect of Light on Biological Tissues
Quantum theory, especially work by Einstein and Bohr, dictated that short wavelength light possesses high-energy photons, capable of damaging matter including biological systems by stripping electrons from molecules [Chanda 1997; Halliday et al 2002; Peirson et a12005; Foster et al 2007]. The minimum amount of radiant energy that can interact with matter equals the energy of a single photon.
1.4.7.4 Extra-ocular Systemic Effects
Relatively non-specific effects of light include raising melanin levels in melanin-containing external parts of the body exposed to light — the skin and the anterior sclera. Cellular tissue damage occurs if light exposure is excessive by stripping electrons from organic molecules. Energy transformation to heat also occurs when light contacts the body.
1.4.7.5 Specific Ocular Effects
1.4.7.5.1 Overview
Long wavelength light (e.g. red) is of lower energy - it elevates temperature by affecting the vibrational (rotational) energy of electrons. Biophysicists using probability theory prompted work on parallel principles to measure the number of quanta required for the eye to detect light using the Poisson distribution [Hecht et al 1942]. The Poisson distribution is a discrete probability distribution aiming to predict the number of events occurring in
79 a given time if the average of them is known. Retinal photoreceptors are so sensitive that they can be stimulated by an individual quantum (photon) of light. But one photon cannot cause perception - that requires 5 to 8 photons to be incident at the retina [Gregory 1998]. The refractive media absorb many photons, especially with age-related cataract. Nevertheless without media opacities it is still possible to see a single candle placed nearly 20 miles away on a clear day with good contrast [Gregory 1998]. The specific biological functions of light falling on the retina shall be described.
1.4.7.5.2 Sight
This is mediated by rods and cones [Grusser 1989; Loewenfeld and Lowenstein 1999; Peirson et al 2005; Foster et al 2007]. Recent work by Dacey and colleagues has isolated a retino-genicular projection in monkeys originating in giant ganglion cells containing the photopigment melanopsin, raising the possibility in anatomic terms of an alternative mechanism involving non—rod non-cone visual perception [Dacey et al 2005].
1.4.7.5.3 Accessory Visual Functions I: Circadian Physiology
Accessory visual functions are those that are linked with sight in that they depend on photoreceptor activation but do not result in sight, though may influence it (a related term is 'associated' visual functions which are specific to the maintenance of sight, such as ocular motility, but which are not necessarily photoreceptor-driven). A major accessory visual function is melatonin synthesis by the pineal gland, a hormone produced in darkness regulating circadian and/or photoperiodic behaviours, and circadian photoentrainment or biological adjustment to the light/dark or day/night cycle [Utiger 1992; Short 1993; Deacon et al 1994; Arendt 1997; Foster and Hankins 2002; Peirson et al 2005; Foster et al 2007]. Almost all forms of life possess a biological clock, an internal way to keep time by "fine- tuning" the body physiologically to the change in light and dark in the environment [Foster and Hankins 2002]. The biological clock is
80 approximately 24 h or 'circadian', meaning a rhythm of approximately one day [Foster and Hankins 2002]. The cusp period between light and day, the `twilight' transition, has been called the `zeitgeber' or time-giver, and drives the circadian pattern, a process known as photoentrainment or circadian photoentrainment [Roenneberg and Foster 1997; Foster and Hankins 2002]. In 1991 it was found that retinally degenerate mice with barely any rods and cones (rd/rd) and no detectable light perception continued to photoentrain [Foster et al 1991]. This work was repeated [Provencio et al 1994]. Subsequent work published in 1995 with 'retina degeneration slow mice' (rds mice) also gave the same apparent paradox for the level of scientific knowledge of the time [Argamaso et al 1995]. The source of the putative non-rod non-cone photoreceptor transducing light in these animals could only be within the eye as loss of the eye abolished circadian photoentrainment [Foster 1998]. A transgenic mouse model was developed that lacked all functional rods and cones. This was by introducing a synthetic transgene (c1) made of attenuated diptheria toxin driving a part of the human red cone opsin promoter, to knock out the cones, in mice that already lacked rods (the resulting mice were called retinally degenerate or rd/rd cl mice), or with rods knocked out instead by a diptheria toxin-based transgene rdta (called rdta cl mice) [Wang et al 1992; McCall et al 1996; Soucy et al 1998]. These genetically engineered mammals continued to perform circadian photoentrainment (demonstrated by circadian locomotor activity) and regulate pineal melatonin production [Lucas 1999; Freedman 1999]. Other light-dependent responses that were preserved in such mice included pupil contraction and the direct modification of behaviour such as responses of masking behaviour [Lucas et al 2001; Mrosovsky et al 2001]. The latter is an example of circadian locomotor activity inhibited by light presented at night to nocturnal animals like photoentrained mice, a phenomenon known as circadian masking [Borbely and Huston 1974]. In 2000 melanopsin was identified in the inner retina of humans, and the pattern of its distribution seemed to match those of cells projecting to the suprachiasmatic nucleas (the master circadian pacemaker) [Provencio et al 2000]. Anatomically a wide range of
81 techniques, including the rd/rd cl mouse model, combined with calcium imaging, suggest that the non-rod non-cone photoreceptor is a subset of retinal ganglion cells accounting for approximately at least 1% of neurones in the ganglion cell layer which are directly photosensitive acting as brightness detectors [Berson 2002; Sekaran et al 2003]. Very recently the melanopsin photopigment has been shown to be the phototransduction pigment in ganglion cells [Melyan et al 2005; Qiu et al 2005; Foster 2005].
Other specialised effects of light include elevation of levels of the important stress hormone cortisol, increase in heart rate and modulation of sleep, the latter by both circadian and clock-independent means, e.g. non-rapid eye movement (NREM) sleep is increased by light in rats, while darkness delivered in the light phase of a 24 h light/dark cycle triggers rapid eye movement (REM) sleep [Trachsel et al 1986; Alfoldi et al 1991; Miller et al 1998; Benca et al 1998]. Light may also modify alertness, performance, mood and certain behavioural patterns [Foster and Hankins 2002; Hattar et al 2003; Foster 2005; Peirson et al 2005].
1.4.7.5.4 Accessory Visual Functions II: Pupillary Contraction
The classic example of an accessory visual function is the pupil light reflex [PLR]. It is also an associated visual function. It has been known since 1927 that pupil responses persist in animals with a significant loss of rods and cones from inherited retinal disease [Keeler 1927a and b; Trejo and Cicerone 1982; Whiteley et al 1998a and b]. This was also suggested by anecdotal evidence from clinicians. It had been assumed that the persistence of small numbers of classic photoreceptors caused this residual pupil response [Trejo and Cicerone 1982]. This view was challenged more recently from studies in animals and humans. After almost 200 years of building on the notion of a duplex retina, highly significant advances started to be made from the late 1980s and early 1990s with the accumulation of experimental evidence for the possibility of non-rod non-cone photoreception, which was later felt to be using melanopsin as the
82 photopigment [Lucas et al 2001; Foster and Hankins 2002]. Circadian responses in animals were found to survive rod and cone loss, most notably in rodless coneless rd/rd cl mice [Mrosovsky et al 2001]. Further, a type of photoreceptive retinal ganglion cell (equivalent to Type III in rats) containing melanopsin was found in retinally degenerate mice infected with pseudorabies virus (used here as a neuroanatomical tract tracer) to project to the pretectal nucleas (PTN) that is closely associated with the PLR [Provencio et al 1998b]. Contrary to some early viewpoints, the photoreceptors mediating the pupillary contraction were not in the iris [Loewenfeld and Lowenstein 1999; Lucas et al 2001]. An action spectrum for melatonin suppression (measured in blood samples in human subjects) suggests this receptor may be operational in humans, but direct evidence from retinally degenerate humans lacking rods and cones has been lacking [Brainard et al 2001; Thapan et al 2001]. However the PLR in rd/rd cl mice is very small, suggesting that the principle component driving the PLR in mice is from classic photoreceptors. This however has been challenged by evidence from studies in macaque monkeys and humans suggesting a larger component, though again the evidence for such a receptor in humans is indirect [Gamlin et al 2007]. In mice irradiances that cause minimal pupillary contraction in rd/rd cl mice almost saturate the PLR in wild type mice [Lucas et al 2001].
The pupil's response may be considered as an initial constriction (miosis) of the pupil in response to a stimulus, followed by dilation, while the entire cycle of constriction followed by dilation can be considered a contraction of the pupil. It has been suggested that in mice there are two components to the PLR: rods and cones mediating initial adaptation to change in luminance, a task for which they are fundamentally well-suited, and non- rod non-cone photoreceptors which sustain the constriction of the pupil [Foster and Hankins 2002]. In macaque monkeys rendered pharmacologically rodless and coneless through synaptic blockade a sustained component to the PLR has recently been found [Gamlin et al 2007].
83 1.4.7.6 Non-Ocular Photoreceptors?
Early interest in a putative photoreceptor in the popliteal fossa behind the knee apparently shifting circadian rhythms, melatonin and body temperature was unrepeatable in laboratories outside those of Campbell and Murphy who described it, and is largely now thought to have been artefactual [Campbell and Murphy 1998; Foster 1998; Yamazaki et al 1999; Koorengevel et al 2001; Lindblom et al 2000; Lockley et al 1997; Wright and Czeisler 2002; Foster and Hankins 2002].
1.4.8 Non-Rod Non-Cone Photoreception: Comparison with Animals
1.4.8.1 Non-Mammals
Numerous non-mammals have extra-retinal photoreceptors in miscellaneous parts of the brain especially round its base [Foster and Hankins 2002]. The pineal organ has been found to contain photoreceptive tissue in all non-mammalian species studied [Meissl and Yanez 1994]. Most poikilothermic (coldblooded) vertebrates do not require retinal input for the light to be detected by the pineal gland [Dodt 1973; Meissl et al 1986]. These pineal photoreceptors have some structural resemblance to classic rods and cones of the retina [Vollrath 1981]. Insects lack pineal glands, but the compound eyes of locusts were studied and insect melatonin was thereby discovered [Vivien-Roels et al 1984]. It has been found in the tissues of many insects and several other invertebrates since, and has been shown to have a neurohormonal releasing activity in the insect nervous system [Richter et al 2000]. Indeed melatonin is thought to be found in nearly all organisms [Richter et al 2000]. In retrospect this is not surprising as all biological life depends on the Sun. Pineal melatonin in fish and amphibians may play an important role in regulating dermal colour changes, and melatonin receptors were first characterised screening a Xenopus melanophore cDNA library [Ebisawa et al 1994; Shand and Foster 1999]. Dermal cells can be regulated by photoreceptors in the CNS via
84 pituitary melanocyte-stimulating hormone (which in humans is split off from pro-opiomelanocortin the other product being ACTH regulating cortisol levels) or directly by photopigments within dermal cells [Marshall 2000; Foster and Hankins 2002]. Dermal photoreceptors mediate colour changes in chromatophores (pigmented cells) including iridophores (those pigmented cells which reflect) and also aspects of behaviour including locomotion such as tail movement in lampreys; melanosomes inside Xenopus melanophores move to the cell margin when illuminated [Provencio et al 1998a; Rollag et al 2000; Foster and Hankins 2002]. Fish have a para-pineal organ that is likely to contain photoreceptors [Vollrath 1997; Garcia-Fernandez et al 1997; Foster et al 2003]. Some amphibians (frogs — anura) and reptiles (lizards-squamata) and Sphenodon (the only surviving family of the rhynchocephalia or beak-headed reptiles which resemble but are distinct to lizards) even possess an extracranial photoreceptive "third eye" or "parietal eye", complete with lens and retina extending to the brain and which may even be observable as a light- sensitive spot on the superior aspect of the head [Foster and Hankins 2002]. A dorsal bony socket may be found in the skull in these species; some authors believe this socket may have contained a third eye in fossils of now extinct early amphibians and reptiles, though this is controversial [Foster and Hankins 2002]. A parietal eye is also found in some species of fish including pelagic sharks and tuna.
Several non-mammals have photosensitive irides [Loewenfeld and Lowenstein 1999; Foster and Hankins 2002]. In vertebrates like locusts and some beetles, ocelli (simple eyes, which combine functionally to form the typical compound eyes of arthropods such as insects and crustaceans.) with mobile pigment cells expand in light to narrow a central light-permissive aperture [Loewenfeld and Lowenstein 1999]. Teleost fish, especially eels, and amphibians, have irides that constrict after the iris has been excised from the rest of the eye, and the pupils of some mammals constrict extremely slowly (orders of magnitude of tens of seconds) to very bright light, surviving both isolation of iris from eye and application of atropine
85 [Bito and Turansky 1975; Lau et al 1992; Loewenfeld and Lowenstein 1999].
1.4.8.2 Mammals
Important differences exist in the PLR between mammals, in addition to those that are readily predictable like differing response latencies owing to the difference in the length of the central pathways and in the size of the iris. In particular, the sustained component to constriction is reduced in fowl [Barbur et al 2002]. In the rhesus monkey it is reduced to a lesser degree when compared with humans [Gamlin et al 1998]. As some non- mammals contain rhodopsin photopigments extraretinally in their photosensitive irides, workers have asked if the PLR in rd/rd cl mice could originate in the iris and not in the retina — however, this was found to be unlikely as the amplitude and rate of pupil constriction is still much higher than that in isolated or atropinised irides as 0.1% atropine topically abolished the PLR, but the consensual response persisted in the opposite eye [Loewenfeld and Lowenstein 1999; Lucas et al 2001; Foster and Hankins 2002].
1.4.9 The Photopigments
1.4.9.1 Structure
Photopigments are light-absorbing chemicals which as a consequence of their nature are coloured, absorb light energy as photons and therefore act as photon counters [Peirson et a12005; Foster et al 2007]. The human photopigments consist structurally of the 11-cis form of vitamin A retinalaldehyde (the light-absorbing chromophore (a group of light- absorbing atoms)) bonded to an opsin protein that varies in its amino acid sequences between rods and the three subtypes of cone [Wald 1968a and b; Foster and Hankins 2002]. The visual pigments of rods and cones are similar structurally. Chemically rhodopsin is a glycoprotein (opsin) bonded
86 to a chromophore group called 11-cis retinal, the aldehyde of vitamin A (retinal) [Grusser 1989]. The opsin component varies in its nature and relationship to the 11-cis retinaldehyde, and this difference determines the wavelength of light to which the photopigment molecule is maximally sensitive (X, max). There are 5 different photopigments in mammals, including rhodopsin (in rods) and three cone opsins (for S-, M-, and L- cones respectively), and melanopsin which has been shown in certain ganglion cells and more recently in a type of cone [Dkhissi-Benyahya et al 2005]. S-cones additionally regulate melatonin expression in ganglion cells [Dacey et al 2006]. All human photopigment genes have been cloned, sequenced and functionally expressed aside from melanopsin. While the photopigment absorption spectra vary considerably, the basic shape is similar between photopigments, indicative of the numerous common structural features they share. Templates that are in wide use have been constructed that correspond to the absorption profiles of mammalian and non-mammalian photopigments [Lythgoe 1979; Rodieck 1998; Hankins and Foster 2002; Peirson et al 2005].
1.4.9.2 Functional Characteristics
During the mid-twentieth century adaptation to light or dark in the duplex model of the retina was understood as due to regeneration of visual pigments [Gregory 1998]. Rushton and Campbell in the Cambridge Physiological Laboratory proved this using reflection densitometry of rhodopsin in frogs and later in humans in 1954 [Rushton and Campbe111954]. Rushton then obtained three colour-sensitive pigments, also detected in this way, with results broadly confirmed by microscopic spectral absorption measurements of individual cone cells [Gregory 1998]. In the 1950s spectrophometry (radial densitometry) was used to measure the absorption profiles of in vivo retinal tissue, and in the 1960s microspectrophotometry (MSP) was used to accurately measure the absorption profiles for classic photopigments in vitro [Rushton 1956a and b; 1965a-d; 1972; Rushton and Westheimer 1962; Marks et al 1964; Brown
87 and Wald 1964). This involves microscopic projection of a tiny pencil of light through the in vitro photoreceptor outer segments (containing the photopigments) which have often been excised. As a monochromatic beam of light is shone through the photopigment-containing part of a photoreceptor cell, by scanning through the spectrum the absorbance of the pigment contained therein is determined [Foster et al 2007]. The advantage of this approach is that it enables photopigments to be characterized within their native photoreceptors, and enables multiple intact, isolated photoreceptors to be analyzed within a single preparation [Liebman 1972; Bowmaker 1984].
Studies of secondary receptor potentials have further confirmed these results [Grusser 1989]. Relatively minor adjustments to the reference values occur from time to time, taking into account factors such as opacification of the ocular media in the anterior segment of the eye like the crystalline lens. Stockman and Sharpe have produced a standardised correction for pre- retinal absorption of light by the lens [Stockman and Sharpe1998]. Other insights using these techniques have revealed the following features of photoreception, summarised by Rushton in 1977 [Rushton 1977]: i. Rod thresholds show substantial range, and can rise by a factor of three from a background stimulus in which just 1% of the rods have caught one quantum of light. ii. The eye behaves like an automatic camera working to maintain it's average sensitivity about the centre of the working range. iii. Nerve signals are contrast-coded; consequently variations in general illumination do not alter neuronal firing. iv. There are two types of adaptation: a) for backgrounds, which can be described using the Weber-Fechner laws, and b) to bleachings. v. Following bleaching the retinal threshold rises as if now exposed to a bright background. The positive after-image (transient subjective persistence of an image) which occurs after bleaching has the same quantitative properties of this bright background.
88 1.4.9.3 Spectral Sensitivity of the Human Eye and Colour Vision
The peak photopic spectral sensitivity of the eye is much higher than the peak scotopic spectral sensitivity (the Purkinje shift), and, using flicker photometry is in the region of 550 nm. By convention, photopic values are normally quoted unless stated. There is some variation between the different mammalian species studied, but the peak pupil spectral sensitivity on the curve is in the region of 555 nm in humans and parallels that for vision. In darkness the peak of the scotopic spectral sensitivity curve is similar to that for rhodopsin and rods, and is about 498 nm in humans. As with photopic vision, this parallels the scotopic pupil action spectrum - in rodents it is 498 nm, and in humans 507 nm following dark adaptation in humans. These parallelisms between visual and pupillary action spectra have been established by work beginning with that published by Brown- Sequard in 1856 and 1859 (an earlier report of his work was noted (unpublished) by P. Broca in 1855 and 1856, and reported in print by Dr. Sharpey in 1856) and most recently by Berman in 1987, with in-between a multitude of abstracts and papers, and which have been collated by Loewenfeld and Lowenstein [Brown-Sequard 1856a and b, 1859; Berman et al 1987; Loewenfeld and Lowenstein 1999]. The visual threshold is shown by most workers to be lower than the pupil threshold in photopic, mesopic and scotopic conditions [Alpern and Campbell 1962; Loewenfeld and Lowenstein 1999]. Adrian in 2003 using a large (Ganzfield) rather than a small 2° adapting field found that the pupillary action spectrum was less sensitive than visual across the entire spectrum in photopic conditions (contrary to some reports that it was less in the blue region of the spectrum, and which used a small 2° adapting field) [Adrian 2003]. The latter he maintains was suggested to be due to some residual rod activity in photopic conditions, though against this view it may be argued that rods are subject to lateral inhibition from cones in photopic conditions [Bouma 1962 and 1965; Adrian 1973; Alexandridis 1985; Berman 1987; Whitten and Brown 1973; Stabell and Stabell 2002; Adrian 2003]. Interestingly Adrian also reported a greater pupil sensitivity (constriction) in the blue end of the
89 spectrum underscotopicandmesopicconditions.Howevertheabsenceof findings maybeconsistentwitharelativelygreaterpupillomotorresponse found byAdriancontradictsverymanyotherworkers,andisnotreadily explained byadifferenceinfieldsizeoftheadaptinglight.Somethese any differenceinpupilresponsebetweenisoluminantphotopicstimulias 1.4.9.4 AbsorptionSpectraoftheVisualPigments from aputativenovelphotoreceptorinhumans,sinceganglioncellshave The absorptionprofileforrodscorrespondswiththatrhodopsinand- These varyasfollows: been showninsomemammalstoeffectthepupilreactionlight. closely approximatesthescotopicvisualspectralsensitivityprofile.Rods approximately 498nm(figure1.5);anotherspectralabsorptionpeakwhich [Grosser 1989]. mediate scotopicvisionusingrhodopsin.Thepeakspectralsensitivityis is notvisibleexistsforrhodopsinintheultravioletregionatabout350nm
7LeliiII.1I1li1II 400 Figure 1.5Peakspectralsensitivityofrods. Wavelength (um) 90 There are three absorption spectra for cones peaking at 437 nm, 533 nm and 564 nm (figure 1.6).
Higher Frequency
0
5
I 500 600 Wavelength (rim)
Figure 1.6 Peak spectral sensitivity of S-, M- and L-cones from left to right.
These correspond to the three basic cone types classes: S-, M- and L- [De Valois and Jacobs 1968; Hochberg 1979]. In the context of cones the results of absorption profiles and secondary receptor profiles (which are objective methods) hence support the trichromatic theory of colour vision.
The absorption spectrum for a novel ocular photopigment is controversial. In 2001 Lucas, Douglas and Foster at Imperial College London showed that the pupil response in rodless coneless mice was found to match that for the excitation of melanopsin [Lucas et al 2001]. Most reports suggest it lies in the region of about 480 nm [Lucas et al 2001; Hattar et al 2003; Foster 2005]. Other reports using recombinant DNA techniques and, in humans, suppression levels of melanopsin, suggest it is much lower in the region of 420 nm [Thapan et al 2001]. Reasons for this are likely to be as follows. The photoreceptive retinal ganglion cell (pRGC) system system is recruited only under conditions of pharmacological blockade or under continuous
91 illumination, such as that used in in vivo experiments in animals and humans [Foster 2005]. When the pRGC system is not properly recruited by either pharmacological blockade or continuous illumination then misleading data can be produced. The pRGC peak sensitivity has been determined by direct recordings from rat ganglion cells following pharmacological blockade (Berson et al 2002). Action spectra studies using long duration light exposure to suppress pineal melatonin have determined a Amax of 480 nm [Brainard et al 2001]. Further, although Thapan et al in 2001 initially noted a Xmax in the region of 460 nm, when later the template was fitted on a log scale to prevent peak bias, the Xmax was in fact closer to 480 nm. Studies on expressed melanopsin using different cell lines have produced maximum sensitivities from 420-480 nm, with the majority of papers supporting 480 nm [Foster 2005]. Another reason for these discrepancies may come from the fact that in these cells the folding or cellular localisation of the protein may not be comparable to the native protein, both of which may influence the Xmax. The pRGC system hence certainly seems to exhibit a peak sensitivity of 480 nm.
Action spectrum with visual pigment template fitting revealed that the photopigment in these non-classic photoreceptor cells was a previously unknown opsin/vitamin A-based photopigment called opsin photopigment/OP48° after its peak spectral sensitivity [Lucas 2001; Hattar et al 2003]. The gene for the photopigment is likely to be mop4 or melanopsin [Panda et al 2002; Ruby et al 2002; Hattar et al 2003, Lucas et al 2003]. In old world monkeys a projection was demonstrated anatomically from giant ganglion cells to the lateral geniculate nucleus [Dacey et al 2005]. This raises the possibility that photoreception in these cells might influence vision itself, though this has hitherto not been explored. Action spectra for melatonin regulation have been shown in humans using circadian melatonin suppression studies, suggesting that humans possess a non-classic photopigment, probably an ortholog of mouse OP480; this does provide some indirect evidence for the existence of non-rod non-cone photoreception in humans, though the finding of melanopsin in rodent S-
92 cones precludes such a conclusion unless significant activity from the receptor can be found in subjects who are devoid of classic photoreceptors, an avenue also hitherto unexplored [Thapan et al 2001; Brainard et a12001; Hankins and Lucas 2002; Dkhissi-Benyahya et al 2005]. Three papers in the literature have focussed their study on the likely peak spectral sensitivity for a putative melanopsin-driven non-rod non-cone photoreceptor in humans. All groups involved subjects with normal vision. Two produced action spectra for light-induced melatonin suppression in subjects with normal vision [Brainard et a12001; Thapan et al 2001]. In these latter two laboratories experiments were performed at night to investigate the relationship between long duration (30-90 min) light exposure and plasma melatonin at different levels of retinal irradiance and wavelength. The univariant irradiance response curves suggested one photoreceptor regulated melatonin secretion, with a Amax 446-477 nm according to Brainard et al or 457-462 nm according to Thapan et al. In the third laboratory Hankins and Lucas used the cone-specific ERG to demonstrate that the irradiance signal regulating temporal dark adaptation properties of the cone pathway is regulated by a single novel opsin photopigment with a maximum sensitivity of 483 nm, which functions mainly as a brightness detector [Hankins et al 1998; Hankins and Lucas 2002]. Overall, while the region of 480 nm seems to be the most likely peak spectral sensitivity of human melanopsin, the large spread of almost 40nm between groups suggests differences due to artefacts, multiple photopigments (with intact rod and cone photopigments), template fitting, or variation in the corrections applied for short wavelength absorption, e.g. as noted before when the template fitted on a log scale was applied to Thapan et al's data the ,max was closer to 480 nm. In fact comparing all three sets of human data on an equivalent log-normal scale the closeness of points on the long wavelength limb suggests they can all be described by the same photopigment [Foster and Hankins 2002]. These results are similar to the action spectra for mouse OP 481 (originally called OP 479) suggesting that humans and mice may possess an ortholog of the same novel opsin photopigment.
93 1.4.10 Phototransduction
1.4.10.1 The Classic Phototransduction Mechanism
The initial events in phototransduction are common to all image-forming and non-image-forming processes arising from the rods and cones in the retina. At rest sodium ions (Na+) continually enters the outer segment of the rod or cone producing the so-called dark current that maintains a transmembrane potential difference (a dark potential) of about — 40 mV. Rods show greater temporal inertia (longer latency) than cones. An earlier early receptor potential (ERP) that is tiny is also described for both rods and cones and may manifest in the ERG [Grasser 1989; Galloway 1998]. The depolarising dark current keeps calcium ion (Ca2÷) channels at the photoreceptor synaptic endings open, allowing more or less continuous release of the neurotransmitter glutamate by photoreceptors at their synapses with bipolar cells. In the dark, the second messenger protein cyclic guanosine monophosphate GMP (cGMP) binds to sodium channels in the outer segments thus maintaining their patency. The effect of light in triggering photopigment breakdown is to turn off Na+ entry into the cell.
A photon is absorbed in the n-electron region of the conjugated double bonds of the rhodopsin molecule raising the molecule to a higher and less stable energy level where it oscillates with greater frequency [Grasser 1989]. As shown by Bohr and mentioned earlier, electrons may be thought of as existing in orbits with different energies round the nucleus of an atom. This process is so sensitive that even starlight causes some rhodopsin molecules to become bleached [Marieb 2004].The quantum efficiency (probability) of this process is 0.5 — 0.6 to produce the necessary structural molecular change: stereoisomerisation of the 11-cis retinal to the all-trans form. The molecule is degraded in several biochemical steps ending in retinal and opsin. These primary energy processes in phototransduction (bleaching) are converted into a membrane potential change in rods and cones [Aidley 1989; Marieb 2004]. This occurs by an enzyme cascade that
94 ultimately destroys the cGMP that is keeping the sodium gates open in the dark. In this process newly freed opsin (or its earlier intermediate metarhodopsin II) interacts with transducin, a G protein subunit, and activates it. In turn transducin activates PDE (phosophodiesterase) the enzyme that degrades cGMP, and so the sodium channels close. Permeability to Na+ decreases greatly but K+ permeability is unchanged (which is the opposite order to the activation of other receptors and currents); consequently the photoreceptors develop a hyperpolarizing receptor potential of —70 mV that inhibits release of neurotransmitter. This is the opposite of what is expected from general sensory physiology where exposure to a stimulus causes hyperpolarisation and not the expected depolarisation of general sensory systems. Cooling the retina to isolate the various temperature-dependent catalytic enzymatic reactions show various components to the ERP which have important functional correlates (Pak and Ebrey 1966; Grusser).
Also of note, another departure from the general principles of sensory physiology is that rods and cones do not generate action potentials. In fact in the retina only ganglion cells and amacrine cells are known to do so [Aidley 1989; Marieb 2004]. With the former this is as ganglion cells are the output neurones whose axons must travel a large distance to the brain. In almost all other cells in the retina local currents will suffice, and are associated with graded potentials (as opposed to the "all-or-none" response typical of other conduction pathways). The all-or-none action potential is, again, preserved in the ganglion cells — other retinal cells are small, closely packed, and graded potentials serve adequately to regulate neurotransmitter release [Aidley 1989]. Retinal amacrine cells are interneurons making lateral and vertical connections in the inner plexiform layer of the retina, which is where bipolar cells that receive the output from rods and cones, synapse with ganglion cells. One such population, called All (or A2), hitchhikes rod bipolar cells onto the cone bipolar circuitry [Aidley 1989; Van Haesendonck and Missotten 1993]. This subtype connects rod bipolar cell output with cone bipolar cell input, and from there the signal can travel
95 to the respective ganglion cells. The morphological absence of a long axon is the origin of their naming as amacrine cells. The dendrites of amacrine cells function as both presynaptic and postsynaptic sites. Like horizontal cells, amacrine cells work laterally affecting the output from bipolar cells, however, their function is generally more specialized — for example each type of amacrine cell connects with a particular type of bipolar cell, there are about 40 subtypes, and generally each utilises one neurotransmitter. Half the amacrine cells in the retina are sensitive to the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) and most that are not inhibitory GABAergic cells like these are also inhibitory but use glycine as a neurotransmitter; very few amacrine cells are excitatory [Van Haesendonck and Missotten 1993]. Amacrine cell output is in the form of graded voltage changes and regenerative action potentials. There is evidence that the quantal amount of neurotransmitter release from presynaptic sites is increased by spike propagation into the dendrite. Studies using patch clamp techniques of individual rodent amacrine cells show their suppression by GABA and suggest various external factors are likely to cause the nonuniform propagation of the action potential from the soma of an amacrine cell [Yamada et al 2002].
1.4.10.2 Phototransduction Mechanism in Melanopsin-Containing Ganglion Cells
While expression studies of melanopsin have provided generally accepted evidence that melanopsin is the photopigment of the pRGCs, the question of what mechanisms operate to facilitate transduction of light information by melanopsin is still very much open. Signalling pathways that may underlie melanopsin-dependent phototransduction in native pRGCs are under study by various groups [Peirson and Foster 2006]. Spectral opponency may be fundamental to formulating a unifying model [Rea et al 2005].
96 1.4.10.3 Light and Dark Adaptation and Interplay of Photoreceptors
Following quantal exposure, in rods retinal and opsin are resynthesised into rhodopsin in a chain of energy-consuming anabolic reactions that use ATP (adenosine triphosphate) mediated by enzymes (Grusser 1989; Marieb 2004). Equivalent reactions occur in cones. If the photon flux on the retina is stable, a dynamic equilibrium is reached between light-induced bleaching chemical reactions (catabolic) and enzyme-mediated regeneration chemical reactions (anabolic). The less the photon flux (light flux) the greater the rhodopsin levels. This is the basis of dark adaptation as observed by the sighted. During dark adaptation ERP amplitude rises as there is a greater probability that photons will strike unbleached rhodopsin molecules [Griisser 1989]. Night vision is thus mediated principally by rods and vision in bright light (both in the Sun and in a brightly illuminated indoors) primarily by cones.
As long as the light is of low intensity comparatively little rhodopsin is bleached and the retina continues to respond to light stimuli [Marieb 2004]. With high intensity light there is wholesale bleaching of the pigment, and, further, rhodopsin breaks down nearly as fast as it is made [Marieb 2004]. At this level the rods are nonfunctional and cones begin to respond, though there is some overlap. Hence retinal sensitivity adjusts to the rate of photon flux.
Light adaptation occurs when a subject moves from darkness into bright light, for example leaving a dark cinema hall [Marieb 2004]. Here initial dazzling occurs — the subject just sees white (achromatic) light, as the retina is still only sensitive to dim light, mediated by rods which do not distinguish colour. Later both rods and cones become strongly stimulated, large amounts of both rod and cone photopigments are degraded almost instantaneously, producing a "flood of signals" causing the sensation of glare. Compensations also occur shortly after exposure to light: retinal sensitivity decreases massively, and retinal neurones rapidly adapt,
97 switching from the rod to the cone system. Within about 60 s cones are sufficiently excited by bright light to function fully. Hence visual acuity and colour vision (both of which are cone functions) improve over the next 5 — 10 minutes. During light adaptation, retinal sensitivity (rod function) is lost but visual acuity and colour vision is gained.
Dark adaptation is the reverse of light adaptation, for example going from a well-lit room into a dark one [Marieb 2004]. Initially nothing is seen but "velvety blackness" as cones are insensitive to low intensity light and also as the bright light the subject was previously exposed to has bleached the rod pigment, so rods are inhibited. In the dark rhodopsin reforms and retinal sensitivity increases. As rods adapt much more slowly than cones, dark adaptation is much slower than light adaptation and can go on for hours. However by 20 to 30 minutes enough rhodopsin is usually present to allow dim-light vision [Marieb 2004; Schwartz 2004]; this time duration is used in most experiments to dark adapt subjects.
A rod-cone break is hence seen [Rushton 1975]. The great range of environmental light to which the eye is sensitive can be explained using this process. During these adaptations, changes occur in pupil size. Though it was obvious to ancient observers that bright light caused pupil constriction and dim light pupil dilation, it was not altogether apparent to modern workers till the twentieth century that the pupil was under the control of both groups of receptors [Loewenfeld and Lowenstein 1999]. As discussed, in certain mammals it has also been shown to be under the control of melanopsin-containing ganglion cells.
A circadian role of the non-rod non-cone photoreceptor is to sense the temporal position of the organism in the night/day cycle. In this regard it might operate in conjunction with rods and cones. The crucial period for adjustment to night and day is at dawn and at dusk. During the phase of twilight, light changes not just in terms of its brightness, but with regard to the position of the Sun relative to the horizon and also spectrally with a
98 greater proportion of light being of shorter wavelengths less than 500 nm to which the non-rod non-cone receptors are maximally sensitive along with S-cones [Lythgoe 1979; Peirson et al 2005]. Rods and cones may well operate alongside the non-rod non-cone receptors in a highly complex adaptive process to light and day, including operation in twilight [Foster et al 2003; Foster 2005; Peirson et al 2005]. It has recently been suggested that the non-rod non-cone photoreceptor is predominantly a brightness detector that may mediate sensitivity of the eye to light in the mesopic (twilight/dusk) range, and hence in light/dark adaptation, and also in colour constancy, a phenomenon whereby colours are perceived as similar despite changes in brightness over the mesopic range [Foster et al 2005; Foster RG, personal communication]. The constancy of colour-coded nerve signals might indicate that the receptor mediates colour constancy.
1.4.11 Processing within the Retina
As was noted convergence of sensory information occurs in the retina via elaborate processing by retinal inter-neurones (amacrine cells, horizontal cells and bipolar cells). Maximum convergence and concentration of data is present at the level of the ganglion cells. Ganglion cells are amongst the most complex cells in the human. Neuronal firing in ganglion cell action potentials are relayed via anatomic projections to midbrain (pupil luminance response), hypothalamus (circadian rhythms) and thalamus thence cortex (sight and certain pupil responses such as those to pattern).
1.4.12 Projections to the Brain: Anatomical Parallelism and Non- Parallelism between Visual and Accessory Visual Pathways
1.4.12.1 Historical Advances
Visual physiology gained impetus in 1927 with Edgar Adrian's recordings from the optic nerve and from Hartline's work from 1932 onwards that initiated correlation of neuronal recordings with visual perception [Hartline
99 1938 and 1940; Jung 1984]. Hartline studying frog optic nerve fibres recorded electrical discharges in single ganglion cells [Hartline 1938 and 1940]. Three types of fibre could broadly be categorised. ON-fibres discharge in response to a light stimulus and cease when the stimulus is turned off. ON-OFF fibres respond to light-on with discharge and also when the light is switched off. OFF-fibres respond with discharge only to light off. It was later shown to be a release from inhibition [Dayson 1980].
1.4.12.2 The Nerve Fibre Layer
The ganglion cells are the outflow cells of the eye connecting it to the brain and are the principle cells of the optic nerve, chiasm and tracts. Axons course towards the optic disc, those from the macula coursing in the papillo-macular bundle. At the optic disc the ganglion cell axons become myelinated. It was widely thought the phospholipid-protein myelin secreted by oligodendrocytes might, were it to be found in the retina, reduce transmission of light through the retina to the classic rod and cone photoreceptors of the outer retina. However, myelinated retinal nerve fibres detected clinically are not in fact associated with scotomata on visual perimetry, casting doubt on the general clinical view that this is the functional purpose of the anatomical absence of myelin in retinal ganglion cells. Indeed these facts may be relevant to understanding the further role of the ganglion cell photoreceptors studied in this thesis and to the new role for this receptor in sight that I found and which is presented towards the end of this thesis.
1.4.12.3 Structure of the Optic Nerve The optic nerve is CNS tissue composed of white matter with ganglion cell axons and neurone support (glial) cells. Ganglion cell projections within the optic nerve are relatively poorly understood. Fibres tend to course from temporal to nasal from before backward in the nerve [Gray et al 1995; Agarwal 1997].
100 1.4.12.4 The Optic Chiasm
At the chiasm there is partial decussation (hemi-decussation) of ganglion cell axons into the optic tract, and there is also a projection of some ganglion cells to the suprachismatic nucleas in the hypothalamus though most ganglion cell axons project in the optic tract to midbrain, thalamus and other parts of the brain [Gray et al 1995; Foster and Hankins 2002]. The chiasm thus represents the beginning of divergence or diffusion of visual sensory information that had reached a focus of convergence in retinal ganglion cells.
1.4.12.5 Post-Chiasmal Pathways
1.4.12.5.1 The Retino-Geniculate and Geniculostriate Tracts
These are summarised in figure 1.7.
Tamporal
Amygda,a
f ,uldnar rucle,s
geroc.kt:o
°plc ra,liabori
Figure 1.7. Retino-geniculate and geniculostriate tracts: the classical pathway for vision (adapted from Nature Neuroscience Reviews).
101 These are eventually a cortical projection that courses via the thalamus. It accounts for approximately 90% of ganglion cells in the optic tract and arises from multiple ganglion cell classes which are in turn fed by classic (rod and cone) photoreceptors. These ganglion cells project to the dorsal and ventral lateral geniculate nucleas (dLGN and vLGN) of the dorsal thalamus forming the retino-geniculate tract [Gray et al 1995]. The thalamus is the major relay station of the brain. From the LGN, via the optic radiation, there is a major projection to the primary visual cortex (striate cortex), the geniculostriate pathway. The striate cortex is located on the medial aspect of the occipital lobe cortex at the back of the head, in its calcarine fissure. From there there are many projections to several different areas of the brain. The geniculo-striate pathway is concerned with visual perception. It may be a pathway whereby certain pupil responses, such as those to pattern, are also influenced.
The geniculo-striate pathway displays a hierarchical structure. In 1999 it was found that some Type III retinal ganglion cells (RGCs) in rodents project to the ventrolateral preoptic nucleas (VLPO) a region involved in sleep-wake regulation, and it was suggested that these may be influenced by non-rod, non-cone ocular photoreceptors [Lu et al 1999; Foster and Hankins 2002]. Recently in a monkey model it has been shown that melanopsin-containing ganglion cells project to the LGN [Dacey et al 2005]. Whether such pathways mediate vision was unknown; exploring this exciting possibility is the culmination of the current work presented in this thesis.
1.4.12.5.2 The Retino-Tectal (Mesencephalic) Pathway and Pupillary Contraction
This is sub-cortical. It accounts for approximately 10% of ganglion cells in the optic tract [Foster and Hankins 2002]. It is concerned with pupil contraction. Electrical impulses are relayed via the to the distal third of the optic tract where they split off from the retino-genicular fibres and course to
102 the midbrain in the superior brachium of the superior colliculus, shown most recently using autoradiographic techniques in squirrel monkeys [Hutchins and Weber 1985; Alexandridis 1985; Loewenfeld and Lowenstein 1999]. It remains unclear as to what degree the pupillomotor fibres in the superior brachium are ganglion cells devoted to the purpose of subserving the PLR exclusive to visual inputs and to what degree they are collaterals of retinogenicular fibres subserving vision; most works discussing the issue suggest both types of cell are involved [Alexandridis 1985; Loewenfeld and Lowenstein 1999] The fibres terminate in all the nuclei of the pretectal nucleas (PTN) complex, with the possible exception of the anterior nucleas where uncertainty exists as to whether the staining pattern found using autoradiography represents passing nerve fibres or actual axon terminals and hence synapses.
The PTN and oculomotor (Edinger-Westphal (E-W)) nuclei are important midbrain (mesencephalic) relay stations. They lie inferior to the superior colliculus and medial to the brainstem reticular core, the latter comprising structures such as the nucleas cuneiformis. The brainstem reticular core, though more inferiorly in the pons, also contains the parapontine reticular formation that is crucial to maintaining binocular single vision and horizontal conjugate gaze. Wider functions of the core include a crucial role in conscious and unconscious states in conjunction with the cerebral cortex, pain suppression, and receipt of connections from almost all ascending and descending tracts by which it receives information from the whole somatic form [Morruzzi and Magoun 1949; Lindsley et al 1949; Scheibel 1984; Snell 1987; Brodal 1998].
The pretectal nuclei are worth defining as some authors use the term differently (in some cases the same author), as has been pointed out by Keenleyside [Keenleyside 1989]. According to studies by Hutchins and Weber in the squirrel monkey there are 5 pretectal nuclei: the anterior, medial and posterior pretectal nuclei, the nucleas of the optic tract, and the olivary pretectal nucleas [Hutchins and Weber 1985]. Some workers also
103 include the nucleas of the posterior commissure in the PTN [Benevento and Standage 1983]. The sublentiformis nucleas may be used to refer to the anterior pretectal nucleas, both in work describing the anatomy of the visual system, as well as visuo-auditory interaction (the orbital frontal cortex of the macaque monkey (visual system) and the cortex of the superior temporal sulcus (auditory system)) [Benevento et al 1977a and b; Benevento and Davis 1977; Benevento and Standage 1983].
The classic PLR is a purely subcortical reflex arc. In the midbrain the pupillomotor afferent fibres carrying action potentials signalling levels of light in the environment terminate in the ipsilateral pretectal nucleas. According to Alexandridis the PTN neurones send axons in three directions: to the contralateral pretectal nucleas, the ipsilateral oculomotor nucleas (E-W nucleas) and a small number to the contralateral E-W nucleas as well [Alexandridis 1985]. Some workers are of the opinion that the afferent fibres from the optic tract may also supply both pretectal nuclei, thus there are in theory a total of three partial decusssations anatomically in the afferent and inter-neuronal limb of the PLR [Loewenfeld and Lowenstein 1999]. The midbrain pathway causes pupillary constriction via the parasympathetic outflow from the E-W nucleas which joins the efferent oculomotor (III) nerve, the main motor nerve to the eye [Benevento et al 1977a]. After synapsing within the orbit in the ciliary ganglion the terminal nerve fibres in the reflex arc innervate the iris, causing pupillary constriction. This pathway accounts for the direct pupil constriction on the ipsilateral side to stimulation and a moreoless equal consensual reaction in the opposite eye.
The retino-tectal tract includes Type IIII RGCs, though this is perhaps 1% of all Type HI cells. From the retina there are bilateral projections to the PTN which are approximately equal [Benevento et al 1977a]. Furthermore, the bilateral projections from the PTN to the visceral oculomotor nuclei also appear to be approximately equal [Benevento et al 1977a]. This accounts for the more or less equal consensual pupillary response: direct
104 and consensual responses are of almost equal amplitude. However behavioural studies in humans indicate a small degree of asymmetry [Cox et al 1984]. Pupils also constrict in response to initiation of ocular convergence and image blur which also induces accommodation and ocular convergence in the near response triad [Schor and Ciuffreda 1983]. The pathways are distinct to the PLR as shown by the clinical syndrome of Argyll-Robertson pupils [Bannister 1985; Loewenfeld and Lowenstein 1999]. It was noted as far back as 1951 that in infants the PLR and pupil response to a near stimulus can be dissociated [Renard and Massonnet- Naux 1951]. There is abolishment of the PLR but pupils still constrict as part of the response to a near target.
The PTN is also involved in pupil reflex dilation, either via projections to the reticular formation or the hypothalamus [Benevento et al 1977a]. Sympathetic innervation causing dilation of the pupil is via the superior cervical sympathetic ganglion. Multiple sensory stimuli can elicit this as part of sympathetic activation of which pupil dilation is a classic sign. The outflow is in spinal nerves between the 8th cervical and thoracic segments [Ranson and Clarke 1959]. The PTN receives retinal, tectal and cortical information. A pathway from the pretectum to the pulvinar to the cortex has also been described [Benevento and Standage 1983].
The classic PLR behaves as if it were predominantly a response to change in luminance [Young et al 1993]. In the 20th century it became recognised as a paradigm for reflex arcs in the human body in general, producing in ideal conditions a stereotyped response to a given input [Loewenfeld and Lowenstein 1999]. The pathway is summarised in figure 1.8.
105 Afferent / pathway \\44
Cilliary ganglion
Efferent pathway
Edingher Westphal nucleus
Bilateral Pretectal crossed nucteii efferent output
Figure 1.8 Summary of the Pupil Light Reflex Pathway. Some authors also believe there is connection between the two pretectal nuclei [Alexandridis 1985]; some authorities have suggested there may also be afferent pathway termination in both pretectal nuclei [Loewenfeld and Lowenstein 1999].
1.4.12.5.3 The Accessory Optic System
Some authors in the literature have used this term to describe those subcortical areas whose functions pertain to mediating the effect of retinal photoreception on body homeostasis [Loewenfeld and Lowenstein 1999]. Other workers have identified the term with a distinct group of midbrain nuclei with a visual role in vertebrates, including the inferior and superior optic fasciculi [Simpson 1984; Cooper and Magnin 1986; Keenleyside 1989]. The nuclei receive inputs from the superior colliculi (retino- collicular fibres), diverging from them in a medial position in the optic tract [Pasik et al 1973]; it also receives afferents from non-ocular structures, for
106 the nucleus survives enucleation, and proposed candidates include pretectum and cortex. It has been variously proposed to have a role in stabilisation of the eyes and head, and a connection to vestibular nuclei and hence co-ordination. A connection to the inferior olivary pretectal nucleas has also been shown [Hoffmann et al 1988]. On these bases Keenleyside has suggested it may have a role in visual-vestibular integration and stabilising the eyes and head in space [Keenleyside 1989]. The most recent use for the term 'accessory optic system' in the literature is by circadian biologists who use the it to describe all non-image-forming nuclei that are involved in photoreception, including pupillary contraction, which is the context in which it will be used henceforth in this thesis [Foster and Hankins 2002; Peirson et al 2005].
Information reaches the master circadian pacemaker of the suprachiasmatic nucleus (SCN) through a specific monosynaptic pathway arising in retinal ganglion cells called the retinohypothalamic tract (RHT) (Moore and Lenn 1972; Ralph et al 1990). The potential location of this projection is shown in figure 1.9. It may play a role in enabling pupil dilation by virtue of its proximity to the sympathetic pathway in the hypothalamus [Loewenfeld and Lowenstein 1999]. In rodents this projection is not retinotopic [Card et al 1991; Provencio et al 1998a and b; Foster and Hankins 2002]. The thresholds to elicit responses are longer than for vision [Nelson and Takahishi 1991a and b]. These affect numerous circadian rhythms including autonomic and neuroendocrine control mechanisms. The latter are in large part dependent on secretion by the pineal gland of melatonin in darkness, and its suppression in light [Foster and Hankins 2002].
107 Figure 1.9 (a) An old anatomical pot dissection of the anterior visual system in relation to the brain from a cadaver, showing the undersurface of the eyes and brain. The potential location of retinohypothalamic tract, coursing postero-superiorly from the optic chiasm into the hypothalamus is shown by the superimposed arrow; (b) the suprachiasmatic nucleas or master circadian pacemaker shown using immunofluorescence (courtesy of Professor RG Foster, Wellcome Trust Centre for Human Genetics, University of Oxford)
1.4.12.6 Ganglion Cell Subtypes
Ganglion cells are amongst the most complex cells in mammals. Prior to advances in understanding non-rod non-cone photoreception, ganglion cells were thought of purely as integrators and transmitters of information to the brain. It is now generally accepted that, at least in mice and certain non- human primates, a subgroup of giant ganglion cells are intrinsically
108 photosensitive and use melanopsin as their photopigment. These giant ganglion cells are approximately 100 [tm in diameter and project to the LGN and it is assumed the pretectum [Dacey et a12005].
There are 28 types of ganglion cell in some mammals, sharing common features like organisation of their receptive fields, but important inter- species differences exist. The classification of ganglion cells is highly complicated. In 1984 Perry, Oehler and Cowey described the morphological features of two main classes of ganglion cell that project to the dorsal LGN [Perry et al 1984]. Brodal has extended such concepts to provide a clear account of the potentially confusing terminology that research over many years on different species has created with regard to the classification of ganglion cells [Brodal 1998]. Recent work on photosensitive ganglion cells is necessitating further changes to the classification of these very complicated cell types, though the area has not been comprehensively reviewed, partly because of the rapid pace of advances in the field and its relative novelty.
Two basic morphological groups of ganglion cell have been identified on the basis of intracellular staining patterns and functional responses. The cat and monkey have been studied in detail. First, P-alpha cells (1-2% of total ganglion cells, but 10% in the retinal periphery), which have relatively large cell bodies, several stout primary dendrites with extensive secondary branching, and are known to project to the LGN; these are also called alpha-cells, Y cells, or Type 1 cells. Second, P-beta cells (50-60% of total ganglion cells) with generally smaller cell bodies and less branching; they are also known as beta-cells, X-cells, or Type 2 cells. In monkeys the remaining 40% of ganglion cells have been studied further and described as P-gamma cells (also called from other species studies gamma cells, W cells and Type 3 RGCs) with small cell bodies and branched dendritic trees and P-epsilon cells with larger cell bodies; both types project to superior colliculus [Perry and Cowey 1984]. In monkeys P-alpha cells were rarely found projecting to the midbrain, but it is still not known if these were
109 collaterals of P-alpha cells projecting primarily to magnocellular layers of the LGN, or separate P-alpha cells. Examining the proportion of crossed and uncrossed fibres projecting to the midbrain, 70% were found to project contralaterally. Overall, no more than 10% of RGCs project to the midbrain in monkeys, but 50% do so in cats [Wassle and filing 1980; Perry and Cowey 1984]. These facts are somewhat misleading if interpreted to imply that the midbrain projection from retinal ganglion cells is of only minor anatomical importance. On the contrary, perhaps one-third of the brain is specialised for visual functions, while Weiskrantz notes that the midbrain projections of RGCs involves more fibres than in the entire auditory nerve [Weiskrantz 1986].
It has been proposed that P-alpha cells probably correlate with retinal broad band cells which have large receptive fields and rapidly conducting axons as described by de Monasterio in the 1970s [de Monasterio 1978a to e]; these cells mediate responses to motion, pattern and contrast. P-beta ganglion cells functionally correlate with those cells with colour opponent centre surround organisation, small receptive fields, slow conduction velocities and linear summation properties that were described by de Monasterio [de Monasterio 1978a to e; de Monasterio and Gouras 1975]. Work on segregation in the LGN and striate cortex by Hubel and Wiesel led to the formalisation of this correlation between structure and function, dividing geniculate ganglion cells into parvocellular and magnocellular pathways (Table 1.1) [Hubel and Wiesel 1978; Livingstone and Hubel 1988a].
110 Property Geniculate Subdivision of Cells Collicular Cells
Parvocellular Magnocellular Type 2 (II) RGCs Type 1 (I) RGCs Type 3 (III) RGCs P-Beta / X cells P-Alpha / Y cells P-Gamma and P-Epsilon or W cells >50% of 1-2% of ganglion cells 40% of ganglion ganglion cells 10% in the cells retinal periphery Small cells; few Large cells; multiple Small cells; dendrites, small branching dendrites, small but not receptive fields big receptive fields sparse dendrites
Colour Yes No No (colour-opponent) (broadband)
Contrast Low High No sensitivity (threshold > 10%) (threshold < 2%)
Spatial High Lower, by 2- Non-linear resolution 3-fold at a given retinal eccentricity
Temporal Slow conduction Fast conduction velocity Slow conduction resolution velocity and and transient responses velocity and both sustained (phasic) phasic and tonic responses (tonic) responses
Table 1.1 Modification of classification of ganglion cells by Livingstone and Hubel [adapted from Keenleyside 1989]. The latter authors classified the magnocellular and parvocellular pathways for visual perception. In this
111 Table projections not known to subserve vision (those to the superior colliculus) or Type III RGCs are also shown (P-Gamma and P-Epsilon cells (W cells) are closely related to each other). RGCs = retinal ganglion cells. The parvocellular system codes mainly for colour and high spatial frequency; the magnocellular system has no coding for the latter components to the visual world and concerns primarily motion detection and contrast detection in the visual world [Norton et al 2002]. The high contrast sensitivity of the magnocellular pathway neurones is used clinically in frequency doubled perimetry as these cells are damaged early in glaucoma.
Ganglion cells can also be classified as "phasic" (brisk, transient trains of action potentials) or "tonic" (sluggish, prolonged trains of action potentials) [Loewenfeld and Lowenstein 1999]. Type 1 RGCs are phasic, Type 2 are tonic and Type 3 are mixed phasic and tonic.
Melanopsin-containing ganglion cells show different spiking patterns, with slow-onset steady-state depolarising spiking in response to a light flash as opposed to slow but rapid-onset hyperpolarisation of rods and cones. Key physiological differences between this 'alternate' system of photoreception and 'classic' photoreception are shown for mice in figure 1.10.
112
Responses Act-ion spectra
met GC 3
0 cone rti — mcq GC — Rods Green aties\ cones \ light slim 20 ml/i 400 500 600 700 1min Wavelength (run)
Figure 1.10 Physiological responses of melanopsin-containing ganglion cells to light. Comparison of spiking pattern with a green cone in mice (left) and action spectra to the right showing peak spectral sensitivity in the region of 480 nm for melanopsin-containing ganglion cells compared with classic photoreceptors [after Berson et al 2003]
The LGN is a laminated structure of 6 layers, each layer containing a retinotopic map of the contralateral visual field [Livingstone and Hubel 1988a]. P-alpha cells have been found to project to the magnocellular LGN layers (the two most ventral LGN layers) and P-beta cells to the parvocellular layers of the LGN (its four most dorsal layers), the so-called ventral and dorsal processing streams [Perry, Oehler and Cowey 1984; Schwartz 2004]. An input from giant ganglion cells containing melanopsin has been shown [Dacey et al 2005]. Magno- and parvocellular pathways have been found to be segregated up to striate cortex (area 17 according to Brodmann) [Schwartz 2004]. The laminated surface is labelled from layers 1 to 6 from the outside in. Hence, as an example, magnocellular LGN layers have been found to project to layer 4C-alpha while parvocellular LGN layers have been found to project to layer 4 C- beta [Livingstone and Hubel 1988a and b]. Overall the connections between LGN and area 17 are
113 extremely complex. Further projections to area 18 and thence to higher visual areas (V3, V4 and MT) also exist.
Some cells in the cortex respond specifically to bars of specific orientation, length, patterns, specific spatial frequency profiles, binocular disparities and colour configuration. Although most striate cortical neurones are binocular, most are dominated by one eye. Homonymous field defects beyond the LGN tend to be congruous unless the monocular crescent is involved [Duke-Elder and Scott 1971]. In contrast, optic nerve lesions tend to cause depressions in field, or sometimes central or caecocentral scotomas [Duke-Elder and Scott 1971]. Ocular dominance is arranged in a regular alternating pattern of right and left ocular dominance columns for right eye and left eye dominance [Hubei and Wiesel 1965, 1968]. The slabs run through the substance of the cortex, perpendicular to its surface, so an electrode perpendicular to its surface encounters neurones in ocular dominance columns dominated by one eye. The same principle holds for orientation slabs [Hubel and Wiesel 1962, 1974]. A full set of ocular dominance columns (for both eyes) and orientation columns (all orientations) forms a hypercolumn, whose size is approximately 1 mm x 1 mm.
Striate blobs are irregular patches rich in cytochrome oxidase in the superficial layers of the cortex. The adjoining visual area 2 contains a regular pattern of stripes [Wong-Riley 1979]. Blobs contain large numbers of double-colour opponent neurones from parvo input [Ts'o and Gilbert 1988]. Parvo (midget) system neurones also supply the inter-blob areas of superficial striate cortex [Schwartz 2004]. The blob (lamina region: area 17 blobs to thin stripes to V4) is thought to be principally involved with colour processing. The parvocellular system is also specialised for high spatial frequency and orientation discrimination. The magno pathway bypasses these areas. The magno system is thought to be specialised for orientation selectivity, movement direction selectivity, stereopsis and has low spatial resolution.
114 Cells of the parvo ganglion cell system are inferred to receive synapses from either a single, or a small number of cones, either directly or via midget bipolar cells. They are thought to have a role in preserving information on visual acuity and colour [Polyak 1941; Hendrickson 1994].
Extensive studies of properties of RGCs projecting to the superior colliculus were performed by Schiller and Malpeli and by de Monasterio [Schiller and Malpeli 1977; de Monasterio 1978a to e]. Electrophysiology suggests that these cells have large receptive fields, non-linear spatial summation, lack colour opponency, respond predominantly in a transient fashion of neuronal firing, and have low conduction velocities.
In addition to the retinogenicular system, sparse LGN-extrastriate pathways also exist [Benevento and Yoshida 1981; Yukie and Iwai 1981; Bullier and Kennedy 1983; Cowey et al 1989; Cowey and Stoerig 1989]. As it has been shown recently that giant ganglion cells expressing melanopsin project to the dorsal geniculate, anatomically it hence seems plausible that non-rod non-cone photoreception might influence vision, and that this is a major purpose, potentially, of the giant ganglion cells as opposed to midget ganglion cells. To identify the melanopsin-expressing cells in the primate that was studied by Dacey et al, a polyclonal antibody derived from the conceptually translated, full-length complementary DNA for the human melanopsin protein was used to immunostain human and macaque retinas [Dacey et al 2005]. In flat mounts of the entire retina, the melanopsin antisera revealed a morphologically distinct population of 3,000 retinal ganglion cells displaying completely stained cell bodies, dendritic trees and axons (figure 1.11). It has been suggested that since there are 1.5 million ganglion cells in the human retina, the melanopsin-expressing cells may comprise only 0.2% of the total. However in rodents they have been shown to be 1% of all ganglion cells in the work by Dacey et al.
Evidence is accumulating which shows the overlap between non-classical (`alternate') and classical photoreception systems. Melanopsin-
115 expressing ganglion cells project to the LGN and more generally to many classical visual regions (LGN, superior colliculus) [Hattar et al 2002). S- cones are likely to modulate the melanopsin expressing ganglion cells - in rodent a novel type of cone has very recently been shown to express melanopsin [Dkhissi-Benyahya et al 2006]. An interaction between rods, cones and melanopsin-containing ganglion cells has been suggested [Rea et al 2005].
100 gm
Figure 1.11 Melanopsin-expressing ganglion cells [after Dacey et al 2005].
116 1.4.12.7 Blindsight
Blindsight is sight in the absence of conscious perception. Despite, for example, destruction of striate cortex, there may still be perception, though the subject will deny it [Weiskrantz 1986 and 1987; Weiskrantz et al 1995; Barbur et al 1998 and 1999; Trevethan et al 2007]. A forced choice methodology will detect sight despite the subject feeling they cannot see [Schwartz 2004]. Shining a light in one of two temporally or spatially separated intervals will result in the subject indicating the correct panel as being the one with the light in it, though otherwise denying that they can see. Blindsight represents an extrastriate pathway, possibly via the superior colliculus [Weiskrantz 1986; Sahraie et al 1997; Leh et al 2006]. Such patients show intact pupillary responses to light [Weiskrantz et al 1999].
1.4.12.8 The Effector Mechanism of the PLR: Structure of the Iris, Effect of Age and Iris Colour
The iris is composed of anterior and posterior leaves. It is the posterior leaf that contains the iris sphincter and dilator muscles that are responsible for pupil contractions. The rest of the iris is composed of connective tissue elements and supply structures including a highly specialised system of end arteries [Gray et al 1995; Bron et al 1997; Snell and Lemp 1997]. The iris sphincter is located just inside the pupillary border, arranged in approximately 20 motor segments that are individually innervated [Kardon 1998 and 2003]. In healthy irises these function as a syncytium, contracting together. The iris dilator muscle is located more peripherally. Pupil contraction at its extremes of constriction and dilation is nonlinear, owing to mechanical limitation of iris movement owing to compaction of iris tissue in the inner circumference of the iris (the inner collarette) [Lowenstein and Loewenfeld 1999] Thus for a given stimulus intensity, a 3- mm diameter pupil in the dark will undergo less constriction that a 5-mm diameter pupil in the dark. In addition to autonomic fluctuations, this is a factor that causes a large degree of variability in the PLR [Kardon 2003].
117 The iris is enveloped in an autonomic nerve net [Loewenfeld and Lowenstein 1999]. The sphincter pupillae and dilator pupillae muscles each have dual innervation from both sympathetic and parasympathetic nerve endings [Kardon 1998 and 2003; Loewenfeld and Lowenstein 1999].
Aging, contrary to popular belief, results in smaller pupils through loss of central inhibition (and not atrophy of the dilator muscle) [Loewenfeld and Lowenstein 1999]. Robinson, Fielder and Moseley have studied the development of the PLR in healthy neonates and those with immature visual systems [Robinson and Fielder 1990 and 1992; Fielder and Moseley 2000]. Studying the PLR in 50 neonates it was absent in all less than 30 weeks' gestational age, developing slowly thereafter and by 35 weeks it was present in all [Robinson and Fielder 1990]. Before the onset of a discernible PLR mean horizontal pupillary diameter was 3.46 mm compared with 3.02 mm after the PLR was detectable. There is no obvious relationship between the sex of a subject and pupillary size or movement.
Brown irides have greater pupillary contraction amplitude, constriction and dilation velocities, but there is no effect on resting pupil size or latency time [Bergamin et al 1998]. Despite this iris colour is rarely taken into consideration in pupillometric work in the literature, outside of occasional consideration in pharmacological mydriasis where it is possible brown irides may dilate more slowly [Bergamin et al 1998; Loewenfeld and Lowenstein 1999].
1.4.13 Vision and the Pupil — Functional Similarities and Differences
1.4.13.1 Principles of Mathematical Descriptions of the Pupil
The pupil's moments can be described by both linear and non-linear analysis. In response to a flash of light after initially constricting the pupil normally redilates to its resting size. The amplitude of the pupillary
118 response depends on several factors, as well as the state of retinal adaptation [Alexandridis 1985].
It has been conclusively shown across several subjects that as target luminance increases, amplitude of pupil constriction increases [Webster et al 1968]. The shape of this response is however very variable between subjects. In a fairly similar series of experiments, latency was shown to decrease to 120 ms for a 6 log unit rise in stimulus intensity [Lee, Cohen and Boynton 1969].
Area summation effects are very important for the generation of pupillomotor force. In 1956 Schweitzer showed that for targets less than 1 deg, as target size was decreased the luminance of the target needed to be increased by a proportional amount to maintain a threshold pupillary response (area . intensity = constant) [Schweitzer 1956; Bouman and Schweitzer 1956; Schweitzer and Bouman 1958a and b]. For larger target areas incomplete summation occurred and with the largest targets this was eventually zero. Schweitzer did not use pupillometry, and simply calculated the percentage of times a pupil response was observed in a given stimulus configuration. Webster et al measured amplitudes of pupil response evoked by various target sizes over the range 6 to 74 degrees [Webster et al 1968]. They found the largest target was associated with the steepest gradient of luminancy / pupil response function.
1.4.13.2 Spatial Summation and Ricco's Law
Building on the laws concerning perceptual thresholds previously formulated in 19th century Germany, Ricco's Law defined a direct relationship between stimulus area on the retina and visual response — for a central stimulus up to 10 minutes of arc from the fovea (the critical diameter for scotopic conditions; the critical diameter is smaller for photopic conditions as the fovea is cone-rich). Increasing target size up to this critical diameter has no effect on the visual threshold — then suddenly
119 as stimulus size increases beyond this the threshold reduces and continues to do so as stimulus size goes up [Barlow 1958]. The summation area in the retinal periphery is much larger than in the macula and of the order of two degrees [Weiskrantz 2004]. Pupillomotor force seems to obey these same rules as described for perception, a further example of the functional parallelism between vision and the pupil. Pupillary constriction is proportional to the area.intensity product for small areas but this summation is incomplete with larger areas [Loewenfeld and Lowenstein 1999]. However pupillary summation areas are larger than those suggested by perceptual thresholds. Keenleyside has suggested this may be because of influence by post-retinal pathways [Keenleyside 1989]. Large areas of dysfunction in retina, especially large peripheral retinal areas or the macula, might be expected to produce reduction in the PLR. Kardon has recently summarised the reductions in the amplitude of the PLR that are caused by various pathologies of the retina or optic nerve [Kardon 2003].
1.4.13.3 Visual and Pupillary Thresholds
The effect of retinal adaptation on the threshold criterion pupil response has received a lot of attention [Bouman and Schweitzer 1958a and b; Alexandridis and Dodt 1967; Schubert and Thoss 1967; Webster, Cohen and Boynton 1968; Alexandridis and Koeppe 1969; Ohba and Alpern 1972). As known in part from ancient times, dark adaptation was found to increase pupil size [Crawford 1936; Flamant 1948; ten Doesschate and Alpern 1967a and b]. Despite this, it was found that the effect of retinal adaptation on the pupillary threshold varies in a similar way to the visual threshold, but the pupillary threshold is slightly higher.
1.4.13.4 Temporal Summation and Bloch's Law
Bloch's Law of Temporal Summation states that a critical period exists in which two stimuli separated by time cannot be resolved i.e. intensity of stimulus.time of stimulus = contant (It=k). Temporal summation has been
120 studied in relation to the pupil. In 1963, Baker, using very short duration flashes (0.1 to 10 ms) across several intensities, showed a linear relationship between stimulus duration and intensity of pupil constriction (pupil response constriction = product of duration and intensity) [Baker 1963]. Also in the 1960s Lowenstein and Loewenfeld showed that as the duration for a test target above threshold was increased from 0.1 to 1.0 s the longer duration stimulus flash tended to produce pupil contractions which were much longer and also of greater amplitude [Loewenfeld and Lowenstein 1999]. These experiments showed that the pupil follows Bloch's Law for perceptual thresholds (intensity.duration = constant), and this was most likely for durations of test stimuli up to 10 ms. It is a further example of parallelism of visual and pupillary function. There have been no detailed quantitative reports of the pupil response and temporal summation beyond 10ms. However Keenleyside has suggested the results of Lowenstein and Loewenfeld noted above may be used to imply that the pupil continues to summate beyond 10 ms [Keenleyside 1989].
The consensual response was studied with renewed detail in the late 1970s and 1980s [Smith et al 1979; Smith and Smith 1980; Wyatt and Musselman 1981; Cox and Drewes 1984]. All studies suggested that stimulation of temporal retina is a more effective generator of consensual pupillomotor force than stimulation of nasal retina i.e. the magnitude of pupil contraction in the consensual eyes was greater with temporal stimulation. Different laboratories reported different results for the comparison of the effects of nasal and temporal retinal stimulation on the direct and consensual responses. Smith and Smith found from 40 healthy eyes of 20 subjects that nasal retinal stimulation produced a greater direct over consensual response, while temporal retinal stimulation produced equal direct and consensual responses [Smith and Smith 1980]. Cox and Drewes studying 26 healthy eyes of 13 subjects found that nasal retinal stimulation produced a greater consensual response over direct, while pupillary response to temporal stimulation was greater in the direct over consensual reflex [Cox and Drewes 1984]. Known as 'reaction anisocorias' (larger responses in one
121 pupil over the other) they have been attributed to greater ganglion cell projection from the temporal over nasal retina and to inter-individual asymmetries in the decussating projections of the pupillary pathways [Cox and Drewes 1984].
1.4.14 Beyond the Pupil Light Reflex: Other Pupillary Oscillations
1.4.14.1 Fatigue Waves, Arousal and Pupillary Unrest
The PLR effector response, rather than merely being a discernible pupil motion, comprises a period of latency, followed by constriction, then by dilation [Loewenfeld and Lowenstein 1999]. Other pupil responses exist independent to the PLR, and also independent to other specific reflex responses of the pupil such as to colour and pattern [Loewenfeld and Lowenstein 1999]. After a few minutes (faster in less healthy or tired subjects) the pupil starts to undergo oscillatory changes in most people [Loewenfeld and Lowenstein 1999]. The extent of variation between subjects is considerable - some subjects will not show fatigue for hours, and others show fatigue after a couple of minutes. These waves are rather slow (3 s or longer) and vary in depth from barely perceptible waverings to huge waves covering a greater part of the range of pupillary mobility of several millimetres [Loewenfeld and Lowenstein 1999]. The typical shape produced resembles a delta, hence the term delta or D waves has been described. Fatigue waves show a genetic tendency, being very similar in their time course and intensity in identical twins [Loewenfeld and Lowenstein 1999].
Autonomic stimulation associated with arousal may result in W and V waves. These are short, shallow pupillary contractions with long latencies that can occur during recording [Thompson 1966].
Pupillary noise has been classified differently by various workers and authorities with varying nomenclature being ascribed. Even fatigue waves
122 have been classified as forms of noise. However the latter are relatively very large in size, which suggests they are not interference and hence not noise — a definition of noise being interference from without the system of reference (here the biological mechanisms of the body). Small amplitude pupillary unrest is also called hippus and may be related to respiration [Loewenfeld and Lowenstein 1998]. The latter term implies to some in the literature that the condition of small alterations in pupil diameter in steady- state light serves no purpose, which is not necessarily true [Loewenfeld and Lowenstein 1998]. Hippus is therefore best not classified as noise as is sometimes done in the literature as pupillary constrictions that may serve a physiological purpose should not be defined as noise.
1.4.14.2 Distinct Channels beyond the Pupil Light Reflex: Overview
Non-visual stimuli such as sound or emotion can induce pupil dilation or constriction via subcortical and cortical pathways independent of the retina and optic nerve [Loewenfeld and Lowenstein 1999]. Initial work by Barbur et al at City University in London, and built upon by Cocker, Moseley and Fielder at another laboratory, suggested that certain non-luminance visual stimuli can evoke pupil responses mediated predominantly by pathways projecting initially to the cortex, and from there to the brainstem most likely by cortico-pretectal pathways (called cortically-mediated pupil reactions), a concept that has become widely accepted [Barbur and Thomson 1987; Barbur et al 1992; Cocker and Moseley 1992; Cocker et al 1994; Cocker and Moseley 1996; Cocker et al 1998]. The efferent pathway is nevertheless the same as the pupil luminance response.
Non-luminance pupil reactions include those to colour, pattern (structure or form), and motion. While the pupil luminance response is primarily a brainstem reflex, there is evidence that the response to structure, colour and motion may have substantial cortical components to their processing. For example using large numbers of averages in healthy subjects it has been shown that the progressively longer latencies in pupil constriction across a
123 range of stimuli parallel those of visual perception, which progress in terms of increasing latency of threshold responses from luminance to structure to colour to motion stimuli; these suggest a hierarchy of central organisation [Barbur et al 1979]. This observation would be consistent with understanding the functional specialisation of the visual cortex for processing spatial structure, colour and movement which are understood to be both hierarchical and interdependent [Barbur et al 1979; Zeki et al 1991; Barbur et al 1992; Zeki 1993; Barbur et al 1998; Bartels and Zeki 2000]. It should be noted that unlike the pupil response to luminance which can approach 100% change in area, those to colour and structure are only of the order of 5% changes in area, while the response to motion is even smaller and is reminiscent of somewhat paralleling the relatively lesser efficiency of the CNS processing visual motion [Barbur et al 1986; Barbur and Thomson 1987; Barbur et al 1992; Cocker and Moseley 1992; Cocker et al 1994; Cocker and Moseley 1996; Cocker et al 1998; Young and Kennish 1993]. This creates special problems of its own in clinical subjects in whom very large numbers of repeat measurements are difficult and in whom there is greater variability owing to emotional and autonomic liability [Kardon 2003].
1.4.14.3 The Response to Pattern
1.4.14.3.1 Vision and the Pupillary Response to Gratings (PGR)
Responses to achromatic and chromatic gratings have been found in most striate cortex cells, but groups of cells exist that show a preference for luminance or chromatic gratings [Thorell et al 1984]. Pupillary responses to `pure' spatially structured stimuli which did not involve overall changes in light flux started to be reported in 1977 (van der Kraats et al 1977; Slooter and van Norren 1980; Ukai 1985; Barbur and Forsyth 1986; Barbur and Thomson 1987).
124 Van der Kraats et al, Slooter et al and Ukai found the pupil responded to alternation of checkerboard patterns. They also showed that the magnitude of response depended on the size of check. In these studies the size of pupil response reflected the checkerboard contrast sensitivity function, which was psychophysically derived. Slooter et al additionally also noted the high correlation (0.96) between subjective checkerboard visual acuity and pupil acuity.
In 1987, Barbur and Thomson, using sinusoidal grating patterns, showed a similarity between behavioural contrast sensitivity curves and pupil responses. In their work even though the pupil measures were made with high contrast grating patterns (of fixed contrast) and the behavioural contrast sensitivity functions were conventional threshold measurements, the shapes of the functions derived were very similar. Both functions were bandpass (as opposed to broadband) and exhibited high frequency spatial cut-off at similar points.
The pupillary responses to threshold and supra-threshold stimuli have been examined by using patterns of varying contrast. With checkerboard stimuli Ukai and Slooter both measured the effect of contrast on amplitude of pupil constriction (fixed check size patterns). Both laboratories found a linear increase in pupil responsiveness with increasing contrast. Slooter additionally measured the contrast required to obtain a threshold pupillary response for a range of check size patterns. This showed that pupillary measures of sensitivity were three times less sensitive than psychophysical measures. However the shape of the pupil threshold curve was very similar to the curve derived with psychophysical procedures. These experiments suggest a bandpass sensitivity function is obtainable with pupillometry using grating patterns over a range of contrasts.
Slooter and van Norren examined how the amplitude of the pupillary response changed with the size of checkerboard. When the area of patch was increased by a factor of four the magnitude of pupillary response
125 increased by a factor of 1.8. Behavioural contrast sensitivity is also known to increase with grating patch size (up to certain critical limits) [Howell and Hess 1978a and b; Robson and Graham 1981]. This is another example of parallelism between pupillary and psychophysical measures of the integrity of the visual system.
Barbur and Thomson investigated the effect that eccentricity and blurring had on behavioural contrast sensitivity and contrast sensitivity derived using pupillometry. Increasing eccentricity was associated with decreased overall sensitivity and lower cut-off spatial frequencies with both the behavioural and pupillometric contrast sensitivity functions. Progressive blurring of the image using convex (plus) lenses also decreased the overall sensitivity and reduced the cut-off spatial frequencies.
What is the origin of the pupil response to pattern? This question has been given consideration from the early 1980s. Slooter and van Norren, and Ukai, proposed that the pathway was similar to the PLR, but that light reflex-type responses were evoked by local luminance changes differentiated at the retinal or pretectal levels. Barbur and Forsyth said they found no variation in gratings responses in the blind field of a hemianopic subject who appeared to have a functioning midbrain pupil light reflex pathway. It was known from the 1970s that changes in accommodation can be induced by gratings stimuli [Ishikawa et al 1970; Charman and Tucker 1977; Owens 1980]. Ukai in 1985 discussed the possibility that these may be responsible for the pupillary response to gratings. This possibility was unlikely as Ishikawa et al reported in 1970 that the latency of these accommodation responses are of the order of 450 ms, whereas the pupillary responses are typically of shorter latency. Concurrent measures of accommodation and pupil size could clarify this issue.
Moseley, Cocker, Fielder, Bissenden and Stirling performed the first studies of the PGR in neonates [Cocker and Moseley 1992 and 1996; Cocker et al 1994 and 1998]. Unlike the PLR, the PGR is absent at birth and also in
126 delayed visual maturation, and develops some time later. Alongside work by Barbur et al this suggested that unlike the PLR the PGR is partly cortically mediated, even though the final pathway is a midbrain outflow tract.
1.4.14.3.2 Pupil Acuity
Given the functional parallelism between visual and pupillomotor responses, some workers have used the PGR to assess visual acuity [Barbur and Thomson 1987]. Cocker, Moseley, Bissenden and Fielder have used the PGR to assess visual potential in neonates, in whom assessment of visual acuity poses special challenges of lack concordance [Cocker and Moseley 1992; Cocker et al 1994].
1.4.14.4 The Pupil and Coloured Stimuli
1.4.14.4.1 Pupil Colour Response
This has in the literature often come to be called the PRC even though strictly it is a specific response to a variation in the wavelength of light, since colour is perception. Early studies considered whether rods or cones or both contributed to the pupil constriction [Laurens 1923; Wagman and Gullberg 1942; Alpern and Campbell 1962]. Spectral sensitivity was measured using pupil responses as the dependent variable under scotopic and photopic conditions. The intensity of light to give a criterion pupil response at each wavelength was determined. The work of Alpern and Campbell in the Cambridge Physiological Laboratory using flicker photometry went down amongst the classic physiological experiments of the modern age [Alpern and Campbell 1962]. Flicker photometry is a technique used to obtain isoluminant stimuli (coloured stimuli eliciting equal subjective responses in terms of perceptual intensity, which is weighted by the human visual system to green) The criterion response in flicker photometry is 'just noticeable flicker' (approximately 15-20 Hz) for
127 two temporally alternating stimulus panels (e.g. circles), one panel of the reference white and the other of the colour to be matched. This occurs if the intensity of colour is a bit less or a bit above the reference (achromatic lenses are used as chromatic aberration would otherwise alter the size of a blue coloured stimulus). The subject is required to adjust the intensity of the coloured panel. The result is averaged over several recordings. An alternative to flicker photometry is the technique of heterochromatic brightness matching. Two spatially separate but adjacent panels are employed, one being of a reference white colour and the other the colour to be matched. The subject adjusts the intensity of the latter coloured panel till it is perceived by him as the same brightness as the white reference panel. This brightness is then measured with a radiometer. With either technique the process is repeated with all colours that need to be used.
Alpern and Campbell plotted pupillometric spectral sensitivity curves under various conditions and found that scotopic and photopic or mixed scotopic/photopic curves could be drive depending on the level of adaptation. However a photopic curve could only be obtained under conditions of rod saturation by an adapting intense blue background light on at the same time as the target stimulus. This series of experiments clarified that both rods and cones contribute to the pupil's constriction and suggested why previous workers like Bouma had failed to find a cone contribution [Bouma 1962].
Saini and Cohen showed that pupillary changes can be elicited with purely foveal exchanges of monochromatic light at isoluminance, thus eliminating the possibility of rods contributing, though also excluding S-cones [Saini and Cohen 1979]. The responses were of the order of only 0.1mm owing to the tiny area of retina stimulated in order to get a pure cone response. Verdon and Howarth addressed the blue cone contribution to the pupil response, finding that the pupil responded to exchange of two tritanopic metamers which equally stimulated the medium (M) and long (L) wavelength cones, but differentially stimulated the short (s) wavelength
128 cones [Verdon and Howarth 1988]. The contribution of S cone channels to pupil size was small compared to M and L cone channels. This is however not altogether unexpected given the much smaller number of S cones.
It is clear from the literature of the time that during the 1970s the PRC had come to be accepted as an important reflex distinct to the PLR [Thompson et al 1980]. Owing to this a large number of subsequent studies have used the combination of a PLR and PRC with the advantage of boosting the pupil response magnitude, with the disadvantage of losing equiluminance [Loewenfeld and Lowenstein 1999].
Krastel, Alexandridis and Gertz investigated whether an influence from post-receptoral colour opponent mechanisms on pupil size could be observed [Krastel et al 1985]. The influence of colour opponent mechanisms on pupil increment thresholds were examined after choosing stimuli which psychophysically would reveal the influence of colour opponent mechanisms in spectral sensitivity [Krastel et al 1985; King- Smith and Carden 1976]. Similar results were obtained with a pupillometry threshold technique. This suggested that information from colour opponent ganglion cells do not transmit via the midbrain visual centres but via the LGN to cortical centres [de Monasterio 1978a-e]. Colour-specific pupil responses therefore most likely end in cortical pathways.
1.4.14.4.2 Colour Blindness
Pupillometry has had other uses in studying colour channels - the investigation of colour vision deficiencies arising from absence of or deficiencies in one of the three cone pigments. Several workers have measured pupil spectral sensitivities (action spectra) reporting that protanopic observers could be distinguished from colour normal observers owing to reduced pupillary sensitivity to long wavelength light in the protanopes [Cohen and Saini 1978; Glansholm et al 1974; Hedin and
129 Glansholm 1976]. But distinguishing deutan colour vision deficiencies using this method has been less successful [Glansholm et al 1974].
Studying the pupil response by pupillometry to exchanges of chromatic stimuli differing only in spectral composition has also been used to try and identify colour vision deficiencies [Kohn and Clynes 1969; Saini and Cohen 1979; Young and Alpern 1980]. Young, Clavadetscher and Teller reported distinguishing protanopic and deutanopic subjects using this technique [Young et al 1987]. Measuring the pupil response to exchange (substitution) of a red and green stimulus it was found that protan observers had little or no response to a green to red substitution. On the other hand deutan observers had much smaller responses to red to green transitions while maintaining significant pupil responses to the green to red substitution.
Keenleyside has noted two case reports of an infant and toddler [Keenleyside 1989]. It was found that sixteen days postnatally no pupillometric colour responses could be obtained in the infant, though a pupillary response to an ordinary flash of achromatic light could elicit a response, described as being within the adult range (some authorities have commented that this was in fact reduced). In a 28 month old toddler pupillometric responses to colour comparable to the adult range were obtained. These observations suggest potential use of this technique for measurement of colour vision deficiencies in non-compliant and non- concordant adult observers, and perhaps also for the measurement of colour vision in babies.
1.4.14.5 Pupil Response to Colour Gratings
Barbur and colleagues working at City University in London were the first to experimentally study the pupil's response to coloured pattern [Barbur et al 1992]. They used the P_Scan 100 pupillometer and devised settings enabling the presentation of isoluminant gratings stimuli blended into a
130 grey background. Later with a group from the United States they also studied the response in the rhesus monkey [Gamlin et al 1998]. In both series of experiments coloured gratings stimuli were used. Their work suggests that the pupillary response to coloured pattern is also a distinct channel like luminance, colour or gratings as different responses were elicited in response to isoluminant gratings of the same spatial frequency. Aside from these workers independent laboratories had hitherto not performed any work using coloured pattern.
1.4.14.6 Pupil Response to Motion
This has been studied in work at City University by Barbur and co-workers. It was concluded after using a pure motion stimulus that a separate channel exists mediating the pupil's response to motion. While the results have not disputed, no other laboratories have reported such work in the peer- reviewed literature suggesting that most likely such experiments have not yet been repeated elsewhere. Sahraie and Barbur used a changing pattern of dots in random motion without net change in luminance, contrast or colour [Sahraie and Barbur 1997]. Measurements were carried out in normals and in one subject with hemianopia caused by damaged primary visual cortex. Onset of coherent motion triggered stereotyped constrictions of the pupil unaccountable in terms of a pupil light reflex response and given the name pupil motion response (PMR). PMRs had large response latencies and the majority were of small amplitude. PMRs depended on changes in motion parameters such as stimulus speed and direction of motion.
1.4.15 Afferent Pupillary Defects
A large number of pupillary defects have been recorded from a variety of pathologies affecting the afferent visual system — mainly intrinsic optic neuropathies, sometimes orbital lesions compressing the optic nerve, and less frequently retinal and other lesions [Loewenfeld and Lowenstein 1999].
131 Afferent lesions can produce one of the following, analogous to low intensity stimuli [Thompson 19661: i. Reduced constriction amplitude. ii. Shortened contraction time to a brief (one-second stimulus). iii. Lengthening of the latent period. iv. Pupillary escape (more rapid re-dilation) with a continuous light. v. Reflexes with a greater proportion of W and V waves.
Lesions have been observed in the dark but these are more controversial, e.g. paradoxical constriction. They are discussed in detail in chapter twelve.
1.4.16 Clinical Applications - Pupillometry, Electrodiagnosis, Psychophysical and Other Tests
1.4.16.1 Introduction
In one sense these bring this literature review full circle, as this was the earliest use of pupil studies — as an objective correlate of sight. However the functional integrity of the visual system can be assessed using a variety of psychophysical tests also, as well as electrophysiology. For some time ophthalmologists have recognised the "enormous scope" for application of electrodiagnosis to clinical practice [Galloway 1997]. Most regional eye centres have come to offer full electrodiagnostic facilities, while even most large general hospitals offer facilities for measuring visual evoked potentials alongside electroencephalography (EEG) and electromyography (EMG) services. However electrodiagnostic equipment, together with the operators proficient to use it, is still sparse compared with those for physiological investigations such as electrocardiography (ECGs) found in virtually all hospitals. Despite the standardisation of electrodiagnosis by the International Society for the Clinical Electrophysiology of Vision [ISCEV], electrodiagnostic facilities have still not progressed out of regional eye centres [Galloway 1997; Marmor et al 2004]. This is likely to be because
132 electroretinography and visual evoked potentials are partially invasive, requiring technically skilled insertion of wire or contact lenses to touch the cornea, the equipment is not very portable, and relatively few ophthalmologists are confident in reading the data [Galloway 1997; Marmor et al 2004]. Electro-oculography can bypass many of these potential hurdles, but the diagnostic usefulness is limited to fewer diseases than either ERGs or VEPs [Galloway 1997].
By 1999 it was noted that pupillometry was still under-explored in terms of potential for clinical application in assessing retinal function in ophthalmology [Loewenfeld and Lowenstein 1999]. The main drawback of pupillometry is the variability of the technique [Kardon 2003]. Its advantage in comparison with electrodiagnosis is that pupillometry is fully non-invasive and portable machines allowing both static (constricted and dilated pupillary diameters for assessment of certain patients prior to refractive surgery) and dynamic (pupil waveform) measurements are marketed. These can be modified to include diagnostic algorithms. Absolute afferent pupillary defect (or afferent pupillary defect) is the term applied when there is no pupillary response to light stimulation of a completely blind (amaurotic) eye, though as the term can be confused with relative afferent pupillary defect (RAPD) which is compared between eyes, sometimes the term amaurotic pupil is used. By naked eye a difference in pupil constriction amplitude of about 30% can be reliably detected between eyes in measuring the RAPD [Bloom et al 1993]. Electronic personal- computer-driven portable pupillometry has been shown to be effective in detecting large RAPDs, and the use of such a device termed 'portable pupillography' by one group [Volpe et al 2000]. The device used by Volpe and co-workers could only detect RAPDs, both real and those simulated by neutral density filters, of 0.9 log units or more; responses from RAPDs of 0.3 to 0.6 log units could not be distinguished reliably. Further studies using more sensitive pupillometers in a portable capacity have not been published. Portability remains an important potential advantage of pupillometry. The other main advantage (at present) seems to be the use for
133 which the pupil was first used — in assessing visual function behind ocular media opacities. Work later presented in this work suggests use of pupillometry as an assay for a novel photoreceptor in humans, though whether at present this will have major clinical significance is unknown.
1.4.16.2 Assessment of Visual Potential with Opacities of the Refractive Media
It is often very useful in clinical work to be able to assess the integrity of the visual system behind dense ocular media opacities that otherwise render clinical examination difficult by obscuring visualisation of the fundus. Indeed as noted this was of interest to cataract surgeons since the time of Galen if not earlier. In modern-day clinical practice such a situation is of special relevance to assessing the visual potential of surgery for media opacities e.g. cataract surgery, corneal transplants, and vitrectomy for haemorrhage into the vitreous. It is useful for diagnosis and prognosis in all these situations to know the status of the macula, particularly as with vitreous haemorrhage where knowledge of macula status in retinal detachment has very significant vitreoretinal surgical implications, and ultrasound provides little measure of function and is not uncommonly rather equivocal as to whether the macula is involved or not. In this situation the relative afferent pupillary defect (RAPD) becomes an especially important part of assessment, and pupillometry may offer even more sensitivity. Further, these clinical scenarios are very common — for example cataracts are the leading cause of global blindness, the most frequently performed surgical operation in the world, and age-related macular degeneration the leading cause in western countries of macula dysfunction and it can co-exist frequently in patients with cataract. Most methods of retinal imaging cannot penetrate ocular media opacities.
134 1.4.16.2.1 Mild to Moderate Ocular Media Opacities
Indirect ophthalmoscopy can provide an assessment of retinal anatomy through even moderate media opacities, though is regarded as part of the ophthalmic examination rather than as an investigation. Fluorescein angiography commonly employs 30° field fundus cameras with optics similar to the indirect ophthalmoscope and can thus also be used to assess retinal anatomy and also function in this situation though it is invasive and can cause serious allergic reactions so is not used for this specific purpose. Laser interferometry is quantifiable and objective [Datiles et al 1987]. The Potential Acuity Meter and its variations can only be used with focal media opacities and cannot assess the retina through opacities dense enough to obscure fundoscopy with an indirect ophthalmoscope [Spurny et al 1986]. Finally, tests that function with dense media opacities can also be used to assess visual potential behind mild and moderate media opacities, as below.
1.4.16.2.2 Dense Media Opacities
1.4.16.2.2.1 Pupil Studies
Pupil reactions are slightly reduced by most media opacities, but are not absent in even severe media opacities. Cataract can augment the PLR in the opinion of some authorities, presumed to be due to increased light scatter since the PLR is dependent upon the area of retina stimulated [Kardon 2003]. This does not seem to follow from Ricco's Law on visual perception or Schweitzer's work using the pupil, and indeed this effect of scatter was not found in work presented in chapter six. The relative reduction in pupil constriction amplitude has been documented for several media opacities and has recently been tabulated by Kardon enabling accurate compensation for the effect of media opacity when assessing visual potential of the retina and optic nerve [Kardon 2003].
135 It is self-implicit from consideration of the neuroanatomy of the pupil light reflex that reduction in the direct pupil reflex in the same eye or the consensual reflex in the opposite eye implies a lesion in the reflex arc on the same side as the light is shone. In practice it is hard to detect such a change unless gross, as there is little to compare it with, a problem solved in part by the relative afferent pupillary defect (RAPD) test.
The relative afferent pupillary defect (swinging flashlight test) is used on a day-to-day clinical basis. Comparing the PLR between the eyes by alternating a light between the two pupils allows detection of even quite small afferent lesions by looking at the degree of asymmetry. The test is rendered invalid in bilateral afferent lesions. However it offers the advantage of being quantifiable with neutral density filters.
Finally, pupillometry can detect even very small differences in pupil diameter in excess of 0.01 mm with systems such as P_Scan (or similar systems) which have been used to study tiny amplitude reflexes such as those elicited by coloured gratings and motion, and permits use of a variety of stimuli.
1.4.16.2.2.2 Electrodiagnosis
The electroretinogram (ERG) and visual evoked potential (VEP) can both be obtained in the presence of dense media opacities [Galloway 1992]. Some authorities have stated that the ERG is unattenuated by media opacities [Galloway 1999]. Studies in Japan of the ERG prior to cataract surgery in eyes with severe cataracts on one side however showed a reduction in the amplitude of a- and b-waves in the more severely affected eyes compared to the fellow eyes, but the difference was not statistically significant [Cruz and Adachi-Usami 1989]. However it is possible the responses may be significantly attenuated in eyes with cataract were they to be compared to eyes without any cataract, requiring raising the stimulus intensity which is outside the remit of standard protocols from ISCEV. The
136 ERG and VEP response is never augmented by opacities, and reliability would become reduced particularly in cases of unilateral opacities as comparison of inter-eye responses is a part of analysis. Further, media opacities can deform checkerboard stimuli used to elicit pattern ERGs and pattern VEPs [Holder 1987; Holder et al 2007]. The flash ERG is a measure of diffuse retinal damage and is incapable of reliably detecting specific macular dysfunction. However it can be used to assess diffuse retinal damage, and combined with VEPs which assess both retinal and post- retinal pathways. It has however been shown that loss of wavelets in a diabetic cataract may indicate poor outcome owing to ischaemia [Galloway 1992].
1.4.16.2.2.3 Entoptic Phenomena
These are all subjective and cannot be quantified but can be elicited even in dense media opacities [Datiles et al 1987]. In the transilluminator test a fibreoptic transilluminator rotated in a circular fashion over the closed upper or lower eyelid elicits the shadow of peri-macular retinal vessels, which look like a tree. The test is useful but subjective — there is a high incidence of false negatives owing to difficulty in comprehending the test. The Blue-field Entoptic Test involves projecting a 530 nm blue light onto the retina through the pupil. The subject may visualise the perimacular vessels more readily than in the transilluminator test and may also see small dots with a halo around them, which are white blood cells within a column of red blood cells crossing the peri-macular capillaries [Sinclair et al 1989; Thompson 1992]. Other psychophysical tests can be employed. A 4- quadrant light projection is useful as it can readily be performed within the clinical examination, but cannot be objectively quantified. It is less reliable in severe media opacities. It relies upon a useful generalisation that barely no perception of light is rare without retinal or optic nerve disease [Miller and Newman 1998].
137 1.4.16.3 Orbital Decompression Surgery
1.4.16.3.1 Clinical Background
Modern imaging has for some time provided generally adequate assessment of structure in most cases of orbital disease [Dailey et al 1992; Wolfley 1997]. However, as opposed to detecting anatomical lesions, the impact on function, especially in the visual pathway, is harder to detect. Even the combination of a battery of tests up to and including electrodiagnosis can leave vaguaries with regard to function in the intra-orbital optic nerve [Miller and Newman 1998]. The functional status of the optic nerve in early compression from orbital lesions is even less certain in the acute situation, particularly in trauma and in orbital cellulitis.
1.4.17.3.2 Orbital Trauma and Orbital Cellulitis
Patients with orbital trauma are not uncommon, and are frequent in certain ophthalmic practices. Electrodiagnosis is not available as an emergency clinical service in most centres, and may not be feasible as the eyelids are hard to open for prolonged periods on a non-portable machine. Psychophysical techniques are subjective and can be hard to perform in an emergency situation [Zaidi and Moseley 2004]. Retro-bulbar haemorrhage and orbital cellulitis can necessitate urgent decompression by lateral canthotomy or ethmoidectomy (endoscopic or external). Often however the indication to decompress only becomes apparent once the clinical picture is sufficiently progressed to cause a tense, tender orbit, a point at which recovery of optic nerve function even in the hands of sub-specialists in orbital surgery who perform ethmoidectomy is much more limited than had diagnosis and/or referral to such a subspecialist or centre been made earlier [Lund et al 1997; Lund and Rose 2006].
138 1.4.16.3.3 Colour Vision in Orbital Lesions — Thyroid Eye Disease, Orbital Haemorrhage and Infection
Commonly used measures of optic nerve dysfunction are not invariably affected in optic nerve compression [Tanner et al 1995]. These include the presence of an RAPD, Snellen visual acuity, red/green colour deficits using Ishihara plates, the latter suggesting loss of parvocellular system function. In thyroid eye disease up to 50% of patients can have normal optic discs while Ishihara test scores are significantly reduced only with visual acuities less than 20/40. Potts et al have compared chromatic discrimination sensitivity, cortical evoked potentials and pattern ERGs in thyroid-related optic neuropathy, finding that chromatic discrimination sensitivity the most reliable [Potts et al 1990]. Work by Tanner, Tregear, Ripley and Vickers has shown that decreased chromatic discrimination is always associated with optic nerve compression [Tanner et al 1995]. Further, the utility of automated chromatic discrimination was shown as a useful and inexpensive test for the presence of optic nerve compression. The clinical utility of the test was evinced by a change to clinical protocol, with more effective follow-up of patients and where accompanied by a RAPD and/or reduced Snellen acuity, confident intervention with orbital decompression.
1.4.16.3.4 Contrast Sensitivity Function in Thyroid Eye Disease
Tanner, Tregear, Ripley and Vickers also found loss of low and medium frequency in patients with thyroid eye disease both with and without optic neuropathy. They pointed out the latter phenomenon has not yet been satisfactorily explained, but has been noted by other workers [Arden 1978; Jindra and Zemon 1989; Mourits et al 1990; Suttorp-Schulten et al 1993; Tanner et al 1995]. It implies loss of function in the magnocellular system in thyroid eye disease, though loss of colour vision in the disease implies loss of parvocellular system function; quite possibly both are affected in the neuropathy. Intermediate and high spatial frequency loss in thyroid eye
139 disease has been reported, but conflicts with the bulk of the literature [de Marco et al 2000].
1.4.16.4 Studies of Retinal Ischaemia and Predicting Neovascular Glaucoma
Work quantifying the RAPD using neutral density filters showed a strong correlation between density of afferent defect and area of retinal non- perfusion on fundus fluorescein angiography in 40 patients with central retinal vein occlusion [Grey and Bloom 1991]. There was an increasing incidence of rubeosis iridis with more severe grades of afferent defects or ischaemia. Bloom, Papakostopoulos and colleagues have studied the PLR using both RAPD (neutral density filters) and infrared pupillometry in central retinal vein occlusion (CRVO) [Bloom et al 1993]. The PLR was found to correlate with the extent of retinal ischaemia and serves as a predictor of the likelihood of the development of neovascularisation, a significant finding as clinical prediction of the development of neovascularisation is part of management. Papakostopoulos, Bloom and co- workers from this group also reported the electro-oculogram in CRVO showing that the EOG light peak/dark trough ratio (Lp/Dt) x 100 to be significantly lower in eyes affected by CRVO than in unaffected eyes, and for the EOG to be abnormally low in absolute terms in the majority (71%) of affected subjects [Papakostopoulos et al 1992]. The ERG in CRVO has been studied by Morrell et al who found interocular differences in 30 Hz flicker latency, pattern ERGs and ratio of scotopic and photopic a:b wave amplitudes in their subjects; depth of RAPD was a very sensitive indicator of rubeosis [Morrell 1991].
140 1.4.16.5 Pupil Perimetry
1.4.16.5.1 Background
Historically most pupillometric studies have used a full-field stimulus [Loewenfeld and Lowenstein 1999]. An alternative way to present stimuli is a comparatively small focal stimulus eliciting pupil responses from small areas of the retina. These can be recorded like visual responses are in visual perimetry or in campimetry. Regardless of whether stimuli are in one plane or in a Ganzfield bowl, the technique became known as pupil perimetry.
1.4.16.5.2 Clinical Applications: Benefits and Limitations of Pupil Perimetry
Pupil perimetry has been suggested as an alternative to visual perimetry. Visual perimetry relies heavily on subject compliance and concordance with instructions and is of limited used in infants, children and the very sick in particular, necessitating the substantial number of statistical algorithms that are commercially available. Subjects with functional (non-organic) disorders such as malingering and hysterical states can evade detection. Pupil perimetry has the capacity to overcome a major inherent limitation of visual perimetry — notably the subjectivity of the latter [Kardon 1995; Yoshitomi et al 1999]. However there are other problems that are not overcome - inability to maintain fixation during thresholding, loss of concentration, inability to understand the testing process, higher variability in peripheral field locations and in damaged areas of the field [Lewis et al 1986; Heijl et al 1987, 1989a and b; Vingrys and Demirel 1993]. Variability in pupillary responses have been the main limitation to widespread clinical application of pupil perimetry. Further, it predominantly tests the integrity of retina, optic nerve, the chiasm and anterior optic tract — brain and posterior optic tract lesions will not show as distinct lesions, and this differential diagnosis is also important clinically. Pupil perimetry is used in some centres, predominantly to assess functional disorders, though the area
141 is under active research to find algorithms that might reduce the effect of variability [Rajan et al 2002; Kandel GL, personal communication].
1.4.16.5.3 Research Applications
Comparison of visual and pupillomotor perimetric responses may aid defining the receptive field size of ganglion cells [Lowenstein and Loewenfeld 1999]. Pupil field loss may pre-empt certain visual defects, helping to define the anatomy and pathology of optic neuropathies [Kardon et al 1991]. Kardon has studied a large number of CNS lesions in this context [Kardon 1995]. In general, correlation between visual and pupillomotor field loss was very high in compressive optic neuropathies and ischaemic optic neuropathy. Most absolute scotomas showed strong correlation between visual and pupillomotor responses, but less was found in relative scotomas, especially glaucoma and demyelinating diseases. Regarding the latter, optic neuritis was usually associated with relatively greater peripheral field loss compared with visual fields. After an initial episode of optic neuritis areas of visual threshold loss usually recovered faster than pupillomotor field loss; pupillomotor sensitivity recovery was also often incomplete. It is unclear at what point visual and pupillomotor fibres segregate in the optic nerve, which could explain such differences. Patients with isolated occipital lobe infarcts showing homonymous visual field defects also showed some corresponding pupil field defects, providing more evidence of cortical PLR mediation under certain stimulus conditions.
1.4.16.6 Vision, the Pupil and Refractive Surgery
Pupillometry is preferred by most ophthalmic surgeons performing refractive surgery over pupil gauges [Duffey and Learning 2003]. As physiologically there is scotopic pupil dilation, if this is marked the size of optical ablation zone in the corneal flap may have to be relatively reduced for operations like laser-assisted in situ keratomileusis (LASIK) and newer techniques, to prevent glare e.g. for night driving. In this thesis the complex
142 roles later explored for pRGCs both in mediating pupil reactions in humans, and as recorders of environmental brightness, are directly relevant to refractive surgery. So are the explorations of visual roles for pRGCs found towards the end, and the pupil's reactions to darkness detailed at the end.
1.5 Implications of Reviewing the Literature
From ancient to modern times studies of visual perception have used the pupil as an index of functional integrity in the visual system in both basic and clinical science. The current work can be regarded as having two parts: i. It begins with studies performed in the context of clinical disorders where pupillometry may be usefully applied, and which additionally provide suitable disease models for basic scientific study. These are to assess the ability of the pupil waveform used in a variety of full- field stimulus paradigms (achromatic, chromatic, pattern) to detect key orbital and vitreo-retinal pathologies, followed by the effect that disorders of the eye's refracting system (lens and cornea) have on pupil perimetry, such as index myopia and scatter. The work is augmented by electrodiagnostic studies as appropriate. ii. In addition to providing basic and clinical scientific data on orbital, vitreo-retinal and refractive disorders with direct potential for clinical application, these results are also analysed for evidence of melanopsin-containing ganglion cell photoreception in humans. The search for the putative pRGC in humans is then taken further using a variety of methodologies both in healthy subjects and directly in pathologically rodless coneless subjects with intact ganglion cells. These include pupillometry, analytical-photobiological action- spectroscopy, electrodiagnosis, actigraphy and sleep studies to assess circadian photoentrainment. Psychophysical experiments are performed to explore visual roles for the novel photoreceptor cell. Experiments on pupillary reactions to darkness are also conducted.
143 Chapter 2 General Experimental Methodology
2.1 Introduction
In this chapter the general methodology involved in the measurement and the generation of stimuli will be discussed. The advantages and also what limitations there are of the techniques employed will be discussed. More specific detail is provided in the relevant specific chapters, particularly on pupil perimetry, actigraphy, sleep studies, and some psychophysical methodology notably the two-alternative forced choice procedure (2AFC).
2.2 Photometric and Radiometric Measurement of Light
2.2.1 Principles
In one sense light and radiant energy are not the same. Radiant energy can be viewed as all electromagnetic energy, whilst historically light is that part of the electromagnetic spectrum visible to the human eye. The strict definition of light is based upon the human visual system's sensitivity to radiant energy, which as noted in the previous chapter, is a subjective experience with often closely-matching objective correlates, for example in humans the peak spectral sensitivity of the pupil (an objective correlate of vision) parallels that of the visual system, even showing a matching Purkinje shift. Yet the term light is generally used to define radiant energy detected by physiological systems in humans and non-humans, and other than by strict purists is frequently used in photobiology and vision science interchangeably with the term radiant energy [Peirson et al 2005; Foster et al 2007]. Light can be measured:
1. In terms of its physical energy. This is greatest at higher frequencies of the visible spectrum.
144 2. In terms of how it is perceived by the visual system of humans, which sees green (medium wavelengths) as brighter.
Consequently two groups of units have evolved for measuring the effect of light on the body. Physical or radiometric units measure light in terms of photon flux and measure electromagnetic energy within the optical spectrum, which includes ultraviolet radiation, visible light, and infrared radiation. An ideal radiometer has a 'flat' spectral response. A range of radiometric quantities and units are used; the two most commonly used in the biological sciences are irradiance and radiance. These are measures for studying environmental light and are used in calculations concerning fundamental interactions between photons and visual pigments at the level of phototransduction.
However in terms of the effect that light has on subjective visual responses to radiant energy, photometric units are often used. The spectral response of a photometer is hence not flat but attempts to reproduce that of the average human eye; two mean human-eye responses are used: the photopic response has a maximum sensitivity at a wavelength of 555 nm, and the human scotopic response a maximum sensitivity at a wavelength of 498 nm. By convention photometric measurements are considered photopic unless otherwise stated. The most commonly used photometric measures are illuminance and luminance. The basic energy unit for the photometric unit is the lumen (lm) which is defined with reference to a truly monochromatic (single wavelength) light source as follows. For scotopic measurements 1 watt (W) = 1700 scotopic lm at 507 nm. For photopic measurements 1 W = 683 photopic lm at 555 nm.
As photometric measures do not have a flat spectral response, and as the emission spectrum for most light sources is varied, converting photometric units into radiometric units is difficult. It is only possible if the precise emission spectrum of the light source is known by calculating the power output and relating this to the integrated area under the spectra for that
145 wavelength range. Where both photometric and radiometric units need to be measured dedicated detectors are used.
The spectral responses of other organisms differ from humans so it may be considered inappropriate to use photometric measurements when studying non-human photoreceptor systems [Peirson et al 2005; Foster et al 2007]. Since human irradiance detection involves a novel opsin with a Xmax of 460- 480 nm thought important in mammalian circadian responses to light, photometry is not advocated in human circadian studies either [Peirson et al 2005; Foster et al 2007]. Photometric measures have however been used extensively historically.
As photometric measures do not have a flat spectral response if they are used to calibrate chromatic light sources caution must be exercised, e.g. equal photometric measures of 'red' and 'green' light will have different photon content and would not be comparable stimuli. The principle use of photometry is hence in comparisons of subjective psychophysical visual responses to chromatic stimuli.
In most cases where the effect of achromatic light is being studied, unless this is specific to the study of photopigments, either photometric or radiometric units may be used; the ISCEV uses photometry, while most circadian studies use radiometry [Peirson et al 2005; Foster et al 2007]. Since the pupil's motions are not a psychophysical percept, radiometry or photometric measures of achromatic stimuli have both been used [Loewenfeld and Lowenstein 1999]. However where comparison is being made between chromatic stimuli, radiometric measures may be preferable in pupil studies in this thesis where the aim is to specifically study melanopsin- containing ganglion cells, in order to enable comparison within the recent literature as radiometric measures have been used in most recent studies [Peirson et al 2005; Foster et al 2007]. Alternatively where this is not the case, either type of unit may be used.
146 The term 'intensity' is often used to mean an amount of light. Technically however intensity is a measure of the radiant or luminous energy from a given direction in space; units are energy (luminous or radiant) per steradian per unit time. The term in its non-technical meaning remains in frequent use owing to its versatility.
2.2.2 Irradiance or Radiance?
Both are radiometric measures. Irradiance is the amount of light falling upon a known surface area. The unit of irradiance is W/m2. The majority of irradiance detectors such as that used in this work (Tektronix radiometer, Oregon USA) are calibrated in W/cm2, and this measure is used to quantify the incident light coming from all directions over a 180° field of view. Many irradiance detector heads do not provide a complete 180° field of view, and instead of using a cosine diffuser (which integrates light from a 180° field of view), detector heads are usually constructed so that light from a wide field passes through an aperture onto a detector surface, which usually measures 1 2 2 cm . Therefore the most commonly used units for irradiance are RW/cm and photons/cm2/s.
Radiance is the amount of light in a specific region of space, and refers to the amount of light as viewed from a specific direction. Radiance detectors are usually constructed by using an aperture and a plus (positive, or convex) lens in front of the detector which allows light from a defined region of space (unit of solid angle or steradian) to fall onto the detector surface. Therefore radiance is expressed in terms of energy per unit area, per unit time, per steradian and the most commonly used units are pW/cm2/sr and photons/cm2/s/sr.
If the direction or position of the light source is crucial then radiance measures are often used, e.g. measurements of biological responses to the Sun in relation to the horizon in field work. If, however, the position of the
147 light source is not as important, then irradiance measures are often used, e.g. laboratory animals subjected to light pulses scattered using a frosted glass ("opal" screen) [Schwartz 2004; Peirson et al 2005; Foster et al 2007].
2.2.3 Additional Significance of the Photon
As discussed in the previous chapter, photopigments absorb energy as photons and thereby they act as photon 'counters'. A photon's energy is proportional to the reciprocal of its wavelength (1/X). Hence high frequency short wavelength light (blue light) has photons of higher energy than red light. But since photopigments act as photon counters the biological effectiveness of light in terms of effects stemming from photoreception are unrelated to these differences in energy. This contrasts with the general effect of these differences in energy. Short wavelengths of light possess high-energy photons, and if containing enough energy may be potentially damaging to biological systems, stripping electrons from molecules but not elevating temperature (i.e. ionising radiation such as ultraviolet light, x-rays and gamma-rays). On the other hand long wavelength photons contain low energy, ranging from infrared up to microwaves and radio waves, and any energy they impart to a biological system is through effects on the vibrational or rotational energy of electrons, which acts to elevate temperature but not to strip electrons. A practical implication in experiments is that if the biological effectiveness of a hue of blue light, say for example 481 nm, is to be compared with a hue of red light, say for example 620 nm, the two stimuli must contain the same number of photons (photon flux). Since the standard radiometric measures (radiance and irradiance) are quantifications of energy (capacity to do work), if the stimuli are calibrated 2 to contain equal radiometric irradiances (same number of p,W/cm ) then blue stimuli would contain fewer photons than red stimuli. Hence the apparent sensitivity of the photopigment would erroneously shift to longer wavelengths. Therefore to compare the effect of different wavelengths of light it is necessary to convert radiometric measures of energy into photons. The method used here is to convert energy flux in W/cm2 into photon flux in
148 photons/cm2/s (or µWes2/E). As noted in the previous chapter, the amount of energy contained in a photon (E in jWs2/photon) is described by E = hn, where h is Planks's constant (6.626 x 10-34 W•s2) and n is the frequency of the wave (n = c/?) (also notated as v). In the calculation it is necessary to calculate E for the wavelength of the stimulus. As an example for 1 photon of 500 nm light this calculation would proceed as follows: E = (6.626 x 10- 34 W.s2) x [(3.00 x 1017 nm/s) / 500 nm] = 3.976 x 10-19 W•s2 /photon or
3.976 x 10-13 1,tWes2/photon. Having calculated the energy of a photon of light of 500nm, the next step is to convert a light value for the 500 nm stimulus from watts into photons, e.g. a hypothetical irradiance of the 500 nm stimulus is 3 x 10-2 µW.s2 then the measured value in µWs2 is divided by the energy of a 500 nm photon in µWs2. Thus = 3 x 10-2 µWs2 / 3.975 x 10-13 µW•s2/photon = 7.5 x 1010 photons/cm2/s. An alternative unit of photon flux is the Einstein (E/m2/s) or microEinstein (RE/m2/s) used by plant physiologists [Peirson et al 2005; Foster et al 2007].
2.2.4 Illuminance (Illumination) or Luminance?
Illuminance is a measure of the amount of light falling upon a surface of defined area, and is similar conceptually to the radiometric measure of irradiance. Traditionally the lux is a commonly used unit of illuminance, due in large extent to the low cost and availability of lux-meters in comparison with other detectors. Luminance is also known as photometric brightness and is conceptually analogous to the radiometric term radiance. Like measures of radiance, luminance detectors have a defined collection angle for light expressed in steradians. In most cases a lens is used to determine the collection angle. A very large variety of different units are used to express luminance, the most common being lx/sr and in terms of 2 candela (cd) as cd/m . The photometer used in the present work expresses 2 measured light in terms of cd/m (Tektronix Lumacolor TM II photometer, Oregon USA).
149 2.3 Presentation of Light Stimuli
2.3.1 Overview
In general light stimuli can be divided into white (achromatic) or monochromatic (chromatic or coloured) light sources. Light output from stimulus generators were calibrated at Imperial College London (Charing Cross campus), then re-calibrated every few months after transfer to the Vision Function Laboratory at the Western Eye Hospital (171 Marylebone Road, London NW1, U.K.). This laboratory housed the pupillometer and electrophysiology equipment in a sealed dark room on a fifth floor with no audible external sounds (e.g. from traffic), and no reflections of light other than within the experimental set-up. No light was detected (+/- 0.5 cd/m2) in the dark at the subject's head rest, even with a small reading light left on for the investigator; this light was not in the subject's field of view. Instructions to subjects were given in such a way so as to prevent undue anxiety. Arrangements were made to ensure that there was no possibility of outside interruption during experiments.
2.3.2 Achromatic light
Various types of white (achromatic), or more accurately broad spectrum, light sources are used in contemporary studies in vision science. Sources depending upon a heated element may exhibit temperature dependence, a similar effect to `black-body' (or cavity) radiation wherein the output of radiation at any given wavelength is dependent upon the temperature of the radiator. Tungsten - halogen incandescent lights, fluorescent lights, high-pressure xenon arc lamps and sunlight vary greatly in their emission spectra, with sunlight itself showing a variation depending on time of day [Peirson et a12005; Foster et al 2007]. The cathode ray tube (CRT) uses an additive red, green, blue (RGB) system of colour reproduction. Luminant phosphors inside the face of the CRT are excited to glow when electrons directed at them impact them. Individual colour dots of red, green and blue can in fact be seen on the face of a CRT with loupes, spectacle-mounted magnifiers used by operating surgeons to magnify anatomical structures. A pixel, an abbreviation of Picture Element, is a single point in a graphic image. As they aranged in rows and
150 columns, so many can be used on a screen (routinely thousands or even millions) that the impression of connectivity within the image can be created to the eye. The number of bits used to represent each pixel is the factor determining the number of colours or shades of grey that can be represented. In colour monitors the specific colour that a pixel describes is some blend of three components of the colour spectrum - RGB; the relative intensity of the three colours controls the pixel's determined colour — if all colours are set to the highest level the pixel is white, and if all are set to zero the pixel is black. In distinction to the use of the term in vision science where it is a component of visual acuity, the term resolution with regard to electronic graphics systems refers to the number of pixels that can be displayed on the screen, and is often categorised numerically as number of pixels horizontally x the number of pixels vertically displayed on the screen. Screen resolution hence reflects the capacity of the output device or video card (graphics card) which produces the images (specifically the large two-dimensional storage matrix with width and height, called 'memory', each element of the memory containing an item of numerical data representing a pixel), and the monitor where the actual pixels are found. The amount of information stored about a pixel in the graphic card memory determines its colour depth, which controls the precise colour of the pixel. The term bit depth is used to describe colour depth as the precision of colour depth is specified in bits. The more bits used per pixel, the finer the colour detail in the image. Increased colour depths require substantially more computer memory for storage of the image and mean more data for the video card to process, reducing the maximum refresh rate of the system. Up to three bytes of data in the microprocessor graphics card can be allocated for specifying a pixel's colour, one byte for each major colour component. Hence a 'true colour' system (colour depth 24-bit) employs all three (3.0) bytes of storage per pixel and can display up to 16,777,216 colours on a colour monitor; a 'high colour' system (colour depth 16- bit) uses 2.0 bytes of storage per pixel and can produce up to 65,536 colours; a `256-colour mode' system (8-bit colour depth) uses 1.0 bytes of storage per pixel and can produce up to 256 colours; a 'standard VGA' system (colour depth 4-bit) uses 0.5 bits of storage per pixel and can produce up to 16 colours [Sarsam 2002]. However, many colour display systems including the one used in this work use only one byte of memory per pixel (limiting the display to 256 different colours).
151 Mixing of RGB output produces a form of achromatic or chromatic light that is widely used owing to its practicability and is used frequently in the current work [Loewenfeld and Lowenstein 1999].
2.3.3 Monochromatic Light
2.3.3.1 Monochromatic Light Production
There are a large number of ways to produce narrow spectrum or monochromatic light as follows:
(a) Absorption filters made of glass, gelatin or liquids in which coloured compounds are suspended or dissolved. (b) Cut-off filters transmit or block light above or below defined cut-off wavelengths. (c) Interference filters produce high quality monochromatic light using an arrangement of highly reflective surfaces ensuring only a narrow band of wavelengths are transmitted, and are extensively used in this thesis. These filters are made by successively evaporating dielectric and silvered films onto glass. Interference filters are defined on the basis of their transmission spectrum (transmittance — T): the wavelength of maximum transmission (Xmax) and on the basis of their half-maximal bandwidth (AX)0.5. High quality interference filters have a half-maximal bandwidth closer to 10 nm and were used in this work in a variety of Xmax. The angle of incidence of the light falling on these filters greatly modifies their transmission spectra and it is crucial that the incident light strike the filter normally (at 90° to the surface), achieved by a collimated source beam. Further, the mirrored surface should face the light source. The transmittance spectra of interference filters tend to be symmetrical, while the transmittance spectra of gelatin filters are often very asymmetrical. Although able to produce higher quality monochromatic light than gelatin filters, interference filters are expensive, and can only be used on relatively small areas such as being shone full-field through the pupil eye but not to bathe the whole body in light. Interference filters, coupled with powerful tungsten - halogen (or similar) bulbs (quartz halogen used in this work,
152 built to specification California, USA) and high quality optic fibers for light delivery, have formed the basis of most monochromatic light sources for circadian experiments. (d) Diffraction gratings (monochromatic diffraction gratings, monochromaters) are designed to disperse incident light into its spectral components from which the desired narrow band of wavelengths can be isolated. They however produce relatively little monochromatic light and, like interference filters, cannot be used to irradiate large areas. (e) Light-Emitting Diodes (LED) emit monochromatic light of fairly high purity, although their half-maximal bandwidth is usually broader than interference filters. (0 Modulation of RGB output in CRTs can produce monochromatic light, but with broader bandwidth than interference filters and a slight mismatch of coloured components, while it is recognised that the red band commonly shows two spectral peaks, including in the CRT used in this work [liyama, Japan, Vision Master 506: Model MS103DT] [Arden G, personal communication]. CRT output is not spectrally pure but can still be monochromatic,
Psychophysical experiments ideally require the use of flicker photometry to ensure isoluminancy in comparisons of responses above zero between monochromatic stimuli.
2.3.3.2 Advantages and Limitations of Isolating Colour Channels
Complete colour channel isolation is possible by use of colour matched colour substitution. This is theoretically the gold standard to produce a pure PRC. One colour is substituted for another of equal luminance as determined by flicker photometry. Since there is no net change in luminance on alternating the colours, an effector response solely due to the new colour and not to the effect of luminance is produced. For example, on a CRT this can be achieved by blending isoluminant CIE-matched coloured stimuli into an isoluminant grey background, and then alternating the coloured stimuli with no overall change in luminance, a process known as colour substitution. The pupillary responses produced are a pure PRC, but are relatively very small with amplitudes of the order of 0.1 mm at maximal
153 stimulus intensity, owing to the small amplitude PRC response, and they are hence obscured by variability which is greatest with small amplitude responses necessitating up to weeks of recording sessions to average even one normal subject [Barbur et al 1992; Loewenfeld and Lowenstein 1999]. Pupil saturation at 0.3 mm is standard for most work before the stimulus becomes intolerable [Kohn and Clynes 1969; Barbur et al 1992]. Alternately, a larger stimulus can be produced by combining a PRC with the PLR, by presenting a red, green or blue stimulus on a black background [Loewenfeld and Lowenstein 1999; Kardon RH, personal communication]. The pupil response seen in this situation is not a pure PRC, being combined with a PLR, but the trade-off is for a much larger response of the order of lmm with less variability requiring fewer sittings to average the readings. Since the PRC became widely recognised, many studies outside of those investigating the basic nature of the PRC (as opposed to using the PRC as a tool) have used the latter method owing to its implicit advantage of a larger response with less variability requiring less averaging in patients; this offers advantages in many subjects with ophthalmic and neurological disease who, as opposed to young healthy subjects, are less likely to be able to withstand weeks of testing and in whom such prolonged testing may conflict with their clinical care [Thompson et al 1980; Loewenfeld and Lowenstein 1999]. A colour-substituted stimulus remains the gold standard stimulus for much psychophysical work, especially in colourimetry, but is not always the most appropriate stimulus for all work. Another colour substitution technique that may be used is presentation of an alternating checkerboard [Arden GB, personal communication]. Though the checkerboard is a pattern stimulus, net luminance does not change nor does pattern as the checkerboard does not vary in position, only the colours do. However, for pupillometric studies, the response carries the same weakness of colour-matched colour substitution in that the response is still fairly small.
2.3.4 Varying the Emission Level (Photon Flux) of a Light Source
There are two common methods for controlling the irradiance of a light source. The voltage applied to the light source can be varied and/or the light source can be screened using neutral density filters. In almost all cases changing the voltage
154 applied to the light source will change its colour temperature, and hence the spectral emission of the bulb (as described by black body radiation). However, if good quality interference filters are being used to generate monochromatic light, changes in the spectral composition of the light source will not present a problem. In any case, the preferred way of changing the flux of a stimulus is through use of neutral density filters, which (in theory) reduce transmission at all wavelengths equally and is generally favoured wherever possible in this work. A great range of neutral density filters exist varying both in quality and price, ranging from black plastic bags to glass neutral density filters used in this work which are carefully calibrated to reduce radiant flux by specific log unit amounts. Other methods for controlling photon flux include the use of crossed polarisers, not used in this work, and movement of the light-source away from the exit aperture of the device.
2.4 Measurements of Pupil Size through the Ages
In addition to static and dynamic methods of measuring pupil size, techniques may be categorised as: direct naked eye observation which may be augmented by lenses and scales (static techniques), photography (a form of static pupillometry), the dynamic techniques of cinematography and video (both digital video and analogue video in which a video capture or video overlay card is used), and electronic measures of differential iris and pupil reflectivity using digital infrared light systems (and which may be either static or dynamic pupillometry) [Charman and Cronly 1991; Loewenfeld and Lowenstein 1999; Sarsam 2002]. The latter three techniques were combined in the mid-twentieth century, as will be discussed in the paragraph below [Lowenstein and Loewenfeld 1958]. To all these techniques may be applied modern microprocessing by computer-based systems enabling both fast and accurate pupil size measurements to be made. Modern digital (the word originates from the same source as digit or finger (digitus in Latin), as fingers are used for counting) systems use discrete (discontinuous) values to represent information for input, processing, transmission and storage. While digital representations are discrete, the information represented may be either discrete, such as numbers, icons or letters, or continuous, such as sounds or images. On the other hand non-digital (analogue) systems use a continuous range of values
155 representing information. The abacus was an ancient digital invention, and the term is also used to describe the mechanisms of counting used in modern computing and electronics, and popularly where real-world information is converted by computers to binary numeric form as in digital audio and digital photography. Such data- carrying digital signals carry electronic or optical pulses, the amplitude of each representing a logical 1 (pulse present and/or high) or a logical 0 (pulse absent and/or low).
In Ancient Greece Archimedes evolved methods for measuring pupil size, as much later did Galileo during the Renaissance [Schweitzer 1956]. Lambert in 1760 used a pair of compasses to measure pupil size under different lighting conditions while looking in a mirror [Lambert 1760]. Landolt' s pupillometer eliminated parallax by using two Fresnel prisms (a tangency principle - having contact at a single point or along a line without crossing) [Sarsam 2002]. Miscellaneous methods of measuring pupil size have been extensively reviewed by Loewenfeld and Lowenstein [Loewenfeld and Lowenstein 1999] who witnessed first-hand the transition from older methods to application of the cinematograph allowing dynamic measurement of the pupil movement rather than static snapshots, through to introduction of infra- red measurement and dynamic video-pupillometry. Contemporary subjective techniques range from the use of hand-held charts to sophisticated automated techniques that compute pupil area from either the number of pixels vertically and horizontally (both meridians are used to reduce noise) between the iris edges of the pupil, or scan lines of a video image of the eye [Murray and Loughnane 1981; Alexandridis 1985; Longtin et al 1990; Bos 1991; Sarsam 2002]. The limiting factor with video-based systems is the sampling frequency [Sarsam 2002]. The field was revolutionised by Lowenstein and Loewenfeld who introduced electronic infrared video-pupillometry in the mid-twentieth century [Lowenstein and Loewenfeld 1958]. In 1987 Barbur, Thomson and Forsyth at City University in London described an accurate infrared video technique for measuring pupil size and binocular movements and which relies on the detection of the border of pupil and iris along several pre-selected scan lines on a video image of the eye [Barbur et al 1987]. This P_Scan machine though produced in the mid-1980s remains a very accurate model: spatial frequency in excess of 0.01 mm, temporal frequency 20 ms.
156 Most contemporary work uses either some form of static pupillometry (often digital) or dynamic digital infra-red video-pupillometry which provides a recording of the pupil over a longer period of time. Aside from the P_Scan 100 machine, which was designed at City University, other pupillometers in common use at present include the Sussex invented by Ripley (which at present comes in two types including one that is handheld; the Sussex is currently being upgraded under the direction of Ripley at the University of Sussex), Procyon (portable and links to laptop computers), Micromeasurements, Puptrak, and Top-Down, amongst others [Sarsam 2002].
2.5 Pupillograms
2.5.1 Considerations in Interpreting Results from Pupillometry
The most important practical limitation of pupillometry is the variability in the response [Kawasaki et al 1995; Kardon 2003]. This may seem surprising as the PLR is an objective assessment of visual potential. However the reflex fluctuates a great deal on account of moment-to-moment changes in autonomic activity — which influence both the final oculomotor (III) nerve outflow (descending hypothalamic influences) and antagonistic sympathetic influences that also innervate the iris. The iris in addition to containing two antagonistic sets of smooth muscle also contains a very large proportion of connective tissue compared to most muscles [Brodal 1998]. The resting pupil diameter varies in health from an average of 2 mm in the light to 8 mm in the dark, with wide confidence intervals in age-matched normals, and even between eyes in the same subject (15% of the population have physiological anisocoria); pupillary diameter is smaller in the very young and the very old (thought by some authorities to be due to reduced inhibitory descending inputs on the third nerve nuclei in the brainstem); and larger pupils normally exhibit a larger maximum contraction amplitude [Miller and Newman 1999; Loewenfeld and Lowenstein 1999; Kardon 2003]. As noted in chapter one iris colour is not usually taken into consideration even in much pharmacological work despite brown irides having greater contraction amplitude, constriction and dilation velocities; this has even been the case with the workers
157 initially making this finding in their subsequent work despite their initial recommendations that it be taken into consideration [Bergamin et al 1998; Loewenfeld and Lowenstein 1999; Bergamin and Kardon 2002 and 2003; Bergamin et a12003]
Ideally subjects with autonomic neuropathy should not be included in research studies owing to well-documented changes in the PLR (other than those used to study diabetes). Some workers would exclude all diabetics irrelevant of documented autonomic neuropathy affecting the pupil [Bremner FD, personal communication]. However other workers strongly disagree [Bergamin 0, personal communication]. It is difficult to see a reason to exclude all diabetics, and many levels of autonomic neuropathy at most rather than invalidating the data may obscure it. However as a caution in experiments reported in this thesis no diabetics, nor indeed any subjects with potentially relevant systemic illness, were included in any experiments. Most studies in the literature have however not taken this precaution [Loewenfeld and Lowenstein 1999].
The extent of variability can be reduced by inter-eye comparison within an individual, by reducing the time period between measurements to as brief an interval as possible, and by the use of multiple averaging. Shorter testing periods reduces the effect of so-called `D, or 'fatigue waves', recognised by their blunted configurations. Nevertheless confidence intervals are often very wide in pupillometric studies.
2.5.2 Applications of Cybernetics to Pupil and Vision Studies
Cybernetics is a broad field of study, the word being derived from the Greek for government. It aims to understand the functions and processes of systems by focussing on information transfer. A broad sub-area of direct relevance to studying the pupil is servo-analysis. A servo is a mechanism that converts an input into an action. Applications include establishing the latent period in the pupil light reflex, studies of noise, studying colour channels in terms of information transfer (both for the pupil and for vision, as noted in the previous chapter), as well as modelling
158 other aspects of the pupil light reflex in terms of a control system in servo-analysis. The latter have included three main themes. First, a simple servo system (servo- feedback loop, closed loop or negative feedback) where the recorder is the pretectal nucleas (PTN), the action it initiates is a signal to the iris which contracts in response, and which via the retina feeds signals back to the PTN that there is an error in the system (notably light levels in the retina have reduced), to which the PTN responds by reducing the signal to the iris — hence the closed loop system which is physiologically equivalent to homeostasis. If the signal from the iris to the PTN is disrupted, the system becomes open loop in servo-analytic terminology. The second main servo-analytic area is study of this open loop system. This state can be induced pharmacologically by drugs paralysing the iris, or by Maxwellian view optical systems that shine a tiny pencil of light into the eye through the pupil, avoiding intersecting the iris margin which are far outside it — the iris continuously contracts. The latter phenomenon leads to appreciation of a third application of servo-analysis in this context - pupil cycling. These are high gain oscillations or increased signals, described by Stark mathematically in servo-analytic terms by a Nyquist criterion [Loewenfeld and Lowenstein 1999].
2.5.3 Mathematical Approaches to Reduce Noise
The pupil waveform is recognised on the basis of infra-red video pupillography to have the following components as shown in Figure 2.1 and known also as the criterion reflex or threshold reflex, [Loewenfeld and Lowenstein 1999]. A period of latency followed by a minimal dilation leads into the constriction phase. This is followed by dilation to baseline, which takes longer if the stimulus is sustained in duration.
159
Pupil Diameter ram 7.75
7.25
6.75
6.25
5.75 1 11 1111 II II lu 1 111 11 1111 Emir' li 11 111 11 inalimitil nu' 11 111 IHR I trim]] II III HI 11111 11111 11 11 III II III 1111 III II II I III II crill 0 0 0 0 0 0 0 0 0 0 0 Time ms 71- 03 Ctrl CCD 0 .'" CO Ctrl CO 0 Cr) (9 0 CO N-- c), co 1--- c) 1- .. .,- v-- 0 A N (N CO co,
Figure 2.1 The pupil waveform
Departures from these stereotypical response patterns in the waveform can be classified as forms of noise [Loewenfeld and Lowenstein 1999]. Verdon and Howarth studied responses to temporal modulation of coloured stimuli. They used Fourier analysis of pupil fluctuation components that were not at the stimulus temporal modulating frequency [Verdon and Howarth 1988]. This was used to establish pupillary "noise", or pupillary fluctuations unrelated to stimulus temporal frequency modulation. A baseline noise level was derived. Keenleyside has commented that this technique does not provide an estimate of the variability in the amplitude of response if the stimulus temporal modulation frequency is matched [Keenleyside 1989].
Barbur et al studied variability of responsiveness to the repeated presentation of a particular stimulus by showing the standard deviations associated with a particular averaged response trace [Barbur et al 1989]. This technique has the advantage of showing whether the noise is uniformly distributed over the pupillogram for a given stimulus wavelength and can be used as an indication of the number of averages required. It is also applied in the form of a digital filter in the P_Scan 100 system used in this work. The disadvantage is that the computed variance includes any drift in mean pupil size during the experiment.
160 2.5.4 Latency and Time Sequences
Pupil reaction latent periods can be divided into absolute and relative (absolute 180 ms, relative approximately 500 ms for a threshold stimulus). It is worth noting that some servo-analysts believe the minimal dilation after latency is a measurement artefact in digital pupillometers, though it is regarded as part of the biological reflex by other authorities [Loewenfeld and Lowenstein 1999; Sarsam 2002]. Latency is inversely related to amplitude of pupil construction [Loewenfeld and Lowenstein 1999]. However rudimentary studies with normal subjects and early studies with subretinal membranes at the very start of this work showed that pupil constriction latency is prone to greater variability than amplitude. Moreover the relative variability in pupil constriction latency over amplitude is well-known amongst workers [Kardon 2003; Smith SE, personal communication]. It may nevertheless be useful sometimes to derive the latency or a measure of how the pupillary response is temporally related to stimulus parameters, for example recent work by Bergamin and Kardon has used time windows to identify which part of the pupil waveform has greatest sensitivity for diagnosing asymmetric disease of the anterior visual pathway, concluding this was the time measured from when maximum contraction velocity occurs to the time to peak contraction [Bergamin et al 2003; Bergamin and Kardon 2003]. Siegle et al have studied pupillograms as an index of concentration and attention and its variability in subjects with clinical depression and in healthy adults [Siegle et al 2004]. Time windows were suggested as a way to evaluate each subject. Lee et al had previously tried to extract time- dependent parameters in healthy pupillograms [Lee et al 1969]. They showed that averaging latencies from traces that show a lot of variability gives an average latency that does not resemble the latency of any individual trace [Lee et al 1969]. As quoted by Keenleyside, work by Lee et al in 1966 and 1967 measuring latent periods found that the average data of groups of 20 reactions per stimulus condition showed less standard deviation than did smaller groups, but that groups of 25 reactions brought no additional benefit [Keenleyside 1989]. It was also predicted the standard deviation (SD) associated with the averaged trace would show a marked increase around the point of the maximum rate of constriction, though this was not found and in fact is contradicted by later work by Bergamin, Zimmerman
161 and Kardon using using time interval analysis [Bergamin et a12003]. Keenleyside confirmed Lee's findings, and also showed that the average traces were representative of the form of the individual traces [Keenleyside 1989]. Workers seeking to apply servo-analytic techniques to pupillometry have focussed on establishing a cut-off for latency but this interesting concept has yet to find widespread application [Bos 1991; Loewenfeld and Lowenstein 1999]. Bos suggested that the period of latency less than 25 ms is fully artefactual and should be rejected in calculations of latency [Bos 1991].
2.5.5 Parameters Extracted from the Pupillogram
The classic parameters measured in physiological recording techniques are constriction amplitude, followed by latency [Loewenfeld and Lowenstein 1999]. It is well-established that latency and amplitude are inversely related to each other [Loewenfeld and Lowenstein 1999]. Much less commonly used is duration of dilation. A number of statistical techniques have been applied to pupillometric data and most techniques use parametric tests, the reflex being somewhat of a paradigm for a reflex arc [Loewenfeld and Lowenstein 1999]. Recently importance has once again been drawn to mid-dilation amplitude in the context of measuring pupil escape, the dilation of the pupil in response to a sustained stimulus [Bergamin and Kardon 2002]. Pupil escape has evaded measurement by many workers over several years, despite being readily appreciated in pupillograms with the naked eye and also clinically where it indicates afferent visual pathway dysfunction. A more detailed discussion is in chapter three. Related to these are measures of dx/dt or the rate of change of pupil diameter which have been successfully used by Kardon et al to diagnose visual dysfunction in time windows as noted above. Finally, it is theoretically possible to use the surface area under and above the pupillogram when it is regarded as a two-dimensional curve. Calculation of surface area between and/or above the pupillogram as an index of dysfunction was considered early in this work, but literature review showed it has barely been studied [Loewenfeld and Lowenstein 1999]. It would be involved work, for example necessitating time-space permutations, and since this work was not primarily about comparison of different ways to measure the pupil, was not taken further.
162 2.6 Effect of Retinal Area
2.6.1 The Effect of Stimulus Size and Retinal Area
Visual stimuli of 1° or smaller will stimulate only L- and M-cone photoreceptors (the foveola); stimuli larger than this will stimulate rods as well as S-cones [Hart 1987]. A pure cone response can therefore be elicited by a stimulus of 1° or 0.3 mm or less in diameter, which would be likely to produce only a slight pupil constriction without using intensities of light that might cause severe discomfort to subjects. This does pose a theoretic constraint on the significance of any findings, but in pupillometric studies using a large flash of light several degrees wide this factor has not been felt to mitigate against the significance of the findings in terms of elaborating macula dysfunction [Loewenfeld and Lowenstein 1999]. S-cones can be isolated by exposure to a 2 minute yellow stimulus.
2.6.2 Maxwellian and Non-Maxwellian View
Some pupil studies employ Maxwellian view, other do not [Loewenfeld and Lowenstein 1999]. As described above experimentally a pure cone signal can be obtained if the central foveola is stimulated. A Maxwellian optical system (Maxwellian view) enables this by producing a narrow collimated beam of incident light, and can also prevent the effect of pupil cycling; the latter is a phenomenon that can occur with certain sizes of incident light beam wherein the pupil can be subject to repeat oscillations as the incident light is reduced by the pupillary shutter coming down, then as the visual stimulus is also consequently reduced by the iris, the pupil dilating, followed by the edge of the light beam being interrupted again, and so on [Thompson 1987]. Pupil cycling is in fact rare unless a specific experimental design is designed to bring it on [Weinstein et al 1980; Thompson 1987]. The pupil constriction amplitude is naturally much smaller in Maxwellian view, raising some of the problems of this that have been discussed. Further, it is often useful to assess the function of the retina as a whole, a situation for which Maxwellian view is sub-ideal. For this reason the current work did not employ Maxwellian view. Pupil cycling did not occur.
163 2.7 Pupillometry and Recording Hardware
The P_Scan 100 system (Circuit-Solve Ltd., Cambridge, UK) was modified for use. Features of this model have been described in extensive technical detail in the literature [Barbur et al 1987; Alexandridis et al 1992]. Stimuli were either presented using a CRT as the stimulus generator (Iiyama, Japan, Vision Master 506: Model MS103DT) with modulation of the luminant phosphor output by the investigator, or by use of a custom-made high output quartz halogen light box source [California, USA] enabling use of monochromatic interference filters. The computer controlling the generation of stimuli integrated a high resolution Pennants colour graphics system; the graphics board supported 256 simultaneous colours including black, with a maximum size of 1152 x 999 pixels driving the high resolution CRT display.
The system incorporates an infra-red video pupillometer (the response recorder) [Lowenstein and Loewenfeld 1958; Barbur et al 1987; Alexandridis et al 1992; Loewenfeld and Lowenstein 1999]. In addition, the modified P_Scan 100 system incorporated the following equipment: a desktop computer (Hewlett-Packard, USA, Model Vectra 486/33ST) with interactive software for stimulus generation and recording of both pupil and eye movement, computer monitor (Philips, Taiwan, Model Brilliance 1520), leads, combined adaptor / earth box for the pupillometer which also linked to the computer (Circuit Solve Ltd, Cambridge, UK), and closed-circuit television (CCTV) cameras (Hitachi Denshi Ltd, Japan, Model VM-920K) producing magnified views of the subject's eye as recorded by the pupillometer to facilitate adequate positioning of subjects by the investigator. The twin instrument base unit carried the pupillometer comprising infrared IR illuminator (IR diode cells), IR-sensitive CCD sensors, associated optics, and chin/forehead rest with the option of a head clamp to restrain head movements in select subjects.
The IR illumination system comprising one panel of IR-emitting diodes for each eye, were located in a position in front of and below each eye, providing good illumination of the iris with specular images outside the pupil area. A group of
164 infrared-emitting electronic cells comprise an array. Semi-mobile attachment of the IR illuminators allowed optimum quality images to be produced by toggling the illuminator using the clearest image of the pupil that could be seen on CCTV. The IR illuminators provided a continuous IR light source; some investigators have suggested that a pulsed IR illuminator on for 4 ms intervals within each image frame can help reduce image smear caused by rapid eye movement [Koeppe and Alexandridis 1968; Loewenfeld and Lowenstein 1999]. However, the modified P_Scan 100 system contains facilities for eye movement tracking as well as pupil movement, enabling such pupillograms to be selected and discarded. A 45° IR reflecting mirror transmitting 95% of visible light was used to provide a clear field of view of the stimulus while at the same time reflect the IR image of the iris to a video camera. For monocular stimulus presentation with binocular recording of pupil responses, an occluder was used to cover the eye not being stimulated. The modified P_Scan 100 system did not incorporate an IR transmitting filter which can serve to block all visible light when placed in front of the unstimulated eye. The IR mirror-reflected rays of infrared from the iris and thereby pupil were captured by an IR video camera. The video signal was sent to a desktop computer (Hewlett-Packard) containing a purpose-built image analyser.
This extracted 64 equal-spaced pre-set scan lines from each video field signal coming from the video camera at a frequency of 50 Hz. The image analyser would then detect the increment and decrement in signal intensity along the scan lines, which corresponded to the pupil margin (iris-pupil junction), and stored the co- ordinates of these intersections. An algorithm was next applied to filter this data. This discarded erroneous intersection coordinates and then fitted a circle of best-fit (since the pupil is approximately a circle) to the data; the inventor of the P_Scan 100 system reports that small elliptical distortions of the pupil image, as occur in normal subjects, do not affect the accuracy of measurements significantly [Barbur et al 1987; Alexandridis et al 1992]. Image analysis was performed every 20 ms (temporal resolution of the pupillometer) with a spatial resolution of better than 0.01 mm. Purpose-built software enabled the data to be displayed in the form of a pupillogram on the video display unit (VDU).
165 The system incorporated auto-thresholding of pupil size. Calibration was with a 1 x 1 mm artificial pupil.
2.8 Stimulus Presentation and Pupil Recording Paradigms
Subjects were seated at the response-recording side of the system facing the stimulus generator, head movements restrained by use of the forehead and chin rests. Subjects were tested fixating on a small cross throughout and with their appropriate refractive correction. The eye not being tested was occluded to prevent summation of responses from consensual and direct pupil reflexes. The investigator viewed a CCTV image of the subject's eye and initiated the presentation of each stimulus when the eye was aligned with the subject comfortable. An audible cue gave the subjects the opportunity for one blink, after which blinking was discouraged till the next audible cue or the experimenter allowed one purposefully. The audible cue was followed by a random time delay (1 to 2 s) that allowed time for blinking to finish, after which the stimulus was presented. A small adapting stimulus was used before each sequence of stimulus intensities was applied, as the first pupil response to a series of light stimuli is known to poorly represent reflex behaviour in subsequent stimulus presentations [Loewenfeld and Lowenstein 1999].
Many experiments involved the presentation of stimuli to a dark adapted eye. This posed the problem of introducing an element of light adaptation, at least theoretically, after presentation of each stimulus. To eliminate this possibility totally would require time-consuming dark adaptation to be repeated before presentation of each stimulus, and is not commonly used [Marmor et al 2004]. This is estimated to be small for most test sources [Rushton 1963]. However, the possibility remains, and is recognised in the methodology used by ISCEV in this work for visual electrophysiology. Stimuli were hence presented from the lowest intensity to the highest to reduce the overall effect of the stimulus reducing the extent of dark adaptation; any effect of loss of dark adaptation is thus made to follow a similar pattern between different experiments, with any effect being greater with progression of stimulus intensity, and minimal at the lower stimulus
166 intensities. Graded trains of stimuli would be presented after each other followed by repeat dark adaptation.
2.9 Collection, Storage and Analysis of Pupil Data
The same software facilitated data collection, storage and several aspects of analysis. Each individual pupillogram would be stored in a file with identical stimuli. Raw data files were individually processed at the initiation of the investigator. The resulting processed data files contained pupillograms that were viewable on the computer monitor. Each individual pupillogram was then screened by the investigator. Where necessary, a digital filter was applied to remove blink artefacts and/or to improve the signal to noise ratio, though this was required with less than 1% of pupillograms. Eye movement tracking for each pupillogram enabled pupillograms with erratic fixation outside 1° to be eliminated; again, this was rarely required. The screened pupillograms were exported to disc. Pupillograms of the same stimulus intensity within a specific experimental paradigm, and that had been screened for quality, were then averaged to produce the mean pupillogram. The mean pupillogram was converted into human readable data in the form of a Microsoft Office text (.txt) file. The .txt file was exported to disc and as it contained human readable data could be downloaded using a Microsoft Excel software programme for analysis and production of graphics. The Microsoft Excel programme also facilitated extrapolation of pupillograms in those rare situations where the dilation curve was insufficient (see below). In most cases pupil diameter had recovered fully to baseline before presentation of the next stimulus in the series; when full recovery did not occur a regression function was used to extrapolate the time to full recovery. In a few cases the reflex response was superimposed on a shifting baseline and time to full recovery could not be computed: these traces were discarded from the subsequent averaging process. Pupillograms were analysed predominantly using the built-in P_Scan software, but also using Excel. Numerous parameters could be extracted using the P_Scan programme using movement of cross-hairs by the investigator to the required data points to be measured on each pupillogram; however, the P_Scan analysis system
167 is not automated. The process of measuring pupillogram parameters was thus very time-consuming.
2.10 Electrophysiology
2.10.1 Stimuli
Electrophysiology is a very established physiological index of function in the retina and optic nerve [Schwartz 2004]. In stark contrast with pupillometry, the very widespread use of visual electrophysiology including in clinical work has helped recording techniques considerably compared to the past owing to international standard protocols guiding its use for production of the electroretinogram (ERG), pattern ERG, visual evoked potential (VEP) and pattern VEP. Only pattern-onset VEPs are not routinely performed within the context of ISCEV criteria. The standard ISCEV stimuli were presented to subjects dark-adapted for 20 minutes using a Ganzfield bowl calibrated for distance and photometric intensity of the light sources (Roland Consult (Electrophysiological Diagnostic Systems), Brandenburg, Germany) [Brigell et al 2003; Marmor et al 2004]. The standard ERG stimuli are shown in Table 2.1.
Scotopic ERG +/- lmV 25dB flash, fixation on, dark. On 0.5Hz (3 records) Scotopic ERG +/- 1 mV 20dB flash, fixation on, dark. On 0.5Hz (3 records) Scotopic ERG +/- 1 mV 10dB flash, fixation on, dark. On 0.5Hz (3 records) Scotopic ERG +/- lmV 0dB standard flash, fixation on, dark. 0.5Hz (3 records) Photopic ERG +/- lmV 0dB SF, fixation on, BackL. On 30Hz (10 records) Photopic ERG +/- lmV 10dB standard flash, fixation on, BackL. On 0.5Hz (6 records) Photopic ERG +/- 1 mV 0dB flash, fixation on, BackL. On 0.5Hz (6 records)
Table 2.1 ERG stimuli (Roland Consult Electrophysiological Diagnostic Systems, Germany Retiport 32). Ganzfield Q400 stimulus.
168 Normative data for DTL (Dawson Trick Litzkow) ERG corneal electrodes (Diagnosis LLC, USA) in adults was used. Normative data was obtained from 20 healthy eyes (age range 30 to 42, 10 males and 10 females).
2.10.2 The Normal Electroretinogram
The ERG trace consists of a negative a-wave followed by a positive b-wave. Additional components are wavelets on the b-wave (oscillator potentials), a scotopic threshold response, and an early receptor potential (ERP), none of which are routinely found within the above ISCEV standard criteria. The a-, b-, and c- complex is thought to be the sum of three processes arising from different layers of the retina (PI, PII and PIII) [Galloway 1998]. The contribution of novel photoreceptors to these is unknown. These are as follows:
PI is positive and contributes to the c-wave; it arises from the RPE PIT is also positive and contributes to the b-wave; it arises from the region associated with Muller cells. PHI is negative and manifests itself as the a-wave' s leading edge; it arises from rods and cones.
2.10.3 Factors Influencing the Normal Electroretinogram
2.10.3.1 Physiological
(a) During dark adaptation the ERG enlarges and b-wave amplitude increases (b) Amplitude reduces slightly with age; there is no relationship to sex. (c) Pupils are dilated using the ISCEV standard as small pupils reduce the ERG. (d) Myopic eyes tend to having smaller ERGs.
2.10.3.2 Equipment
(a) Stimulus strength and duration have significant effects. Increased stimulus intensity results in increased a- and b-wave amplitude up to saturation, which
169 occurs earlier with the b-wave. The oscillatory potential wavelets become more prominent at higher intensities, and at the highest tolerable intensities the early receptor potential is seen. Flicker with a frequency of 30 Hz isolates cones responses as rods cannot respond to this high rate. The pattern ERG employs instead of a flash a pattern, by ISCEV convention an alternating black and white checkerboard. The response arises from the macula and ganglion cells.
(b) Amplifier settings affect the ERG waveform shape and size therefore ISCEV standardisation is preferred with uniform use of one type of recommended electrode. The position and type of electrode affect the response, e.g. contact lens electrodes tend to give larger and hence more reliable responses than eyelid wires but cause image distortion so are unreliable for the pattern ERG.
2.11 Subject Recruitment and Exclusion Criteria
Subjects were recruited with the aid of clinical consultant collaborators from St Mary's and the Western Eye Hospital NHS Trust. On a practical level the recruitment of subjects was directly in collaboration with senior trainees in vitreo- retinal surgery (fellow) and oculoplastic / orbital surgery (fellow) and for the rarer conditions a consultant with retinal expertise. Patients with diabetes mellitus were excluded (these patients can have autonomic neuropathy) as were those taking systemic or topical drugs known to affect autonomic function as referenced in the British National Formulary. No subjects had serious systemic illnesses. The remainder of subjects were invited to attend for an initial examination to determine whether they met the study's inclusion criteria. Visual function (logMAR acuity, colour discrimination (Ishihara pseudo-isochromatic plates) in good illumination, and perimetry if not available (Humphrey 24-2 and 120-2 full threshold test or Goldmann)), pupillary reactions with swinging flashlight tests in dim illumination and dilated fundoscopy were performed on both affected and fellow eyes. Subjects with efferent lesions were excluded by eliciting clinical signs indicating sympathetic or parasympathetic lesions and by testing the size and reactivity of both pupils in the dark and in the light using pupillometry to exclude non- physiological anisocoria [Miller and Newman 1999]. Pupil examination was
170 always conducted at least 48 hours after any pharmacological mydriasis had been administered. Patients were asked to abstain from caffeine, stimulants, chocolates and undue physical exercise for 24 hours before the recordings were made.
2.12 Ethics Approval and Informed Consent
Local Research Ethics Committee approval was obtained (St Marys NHS Trust). All subjects were provided with an information sheet and gave informed consent according to the Declaration of Helsinki [Appendix I].
171 Chapter 3 Subretinal Neovascularisation, Emergence of Concept of Pupil Evasion Ratio and Establishment of a Discriminatory Chromatic Threshold
3.1 Pupil Escape
A brief light stimulus produces an initial transient or 'phasic' pupillary constriction [Kardon 1998 and 2003]. An uninterrupted prolonged light stimulus produces a sustained pupillary constriction [Kardon 1998 and 2003]. Temporal differences between phasic and sustained pupillary constrictions may be reflected in the RAPD observed in response to different time intervals of alternation of the light source, with differences in the extent of inter-eye constriction amplitude depending on frequency of alternation of the light source [Bergamin and Kardon 2002]. This may support clinical usage of two tests — a conventional RAPD test and a `mini'- RAPD test involving a more rapid frequency of alternation of the light source between eyes.
Preferential reduction in the sustained, as opposed to the initial phasic, pupillary constriction has been called pupil escape [Bergamin and Kardon 2002]. More recently it has been shown that, when expressed as a ratio incorporating pupil escape and maximal pupil constriction to a prolonged (5 second) stimulus (pupil escape ratio), central visual field loss is associated with greater escape than equivalent loss in the peripheral field [Bergamin and Kardon 2002].
3.2 Objectives
This series of experiments primarily investigates whether similar patterns of quantifiable pupil dilation can be reliably estimated in response to a much shorter light stimulus. This was performed in patients with unilateral subretinal neovascular membranes (SRNs). The ability to differentiate a
172 healthy eye from one with a significant macular lesion such as SRN might enable pupillometry to complement existing investigations in situations where it offers conceivable advantages such as assessing visual potential objectively in the presence of ocular media opacities. The study of responses to a shorter duration stimulus may also provide information on subtypes of ganglion cells complementing that revealed when measuring escape from sustained light stimuli.
3.3 Materials and Methods
Ten participants completed this study. Three participants were men, seven women (mean age 63, range 39-82). The digital archives of all patients who had recently undergone fundus fluorescein angiography (FFA) at the Western Eye Hospital in London were queried to identify those patients with unilateral subfoveal subretinal membranes (figure 3.1) [Zaidi et al 2004].
Figure 3.1 Red-free image of a dense haemorrhagic subfoveal membrane in a subject with age-related macular degeneration (in this pilot study all participants had unilateral subfoveal membranes).
173 FFAs were reviewed by senior clinicians (consultant ophthalmologists) with subspecialist expertise in retinal diseases, especially SRNs. Unilateral SRNs with healthy fellow eyes were unusual, primarily as the most common aetiology for this condition is age-related macular degeneration, which has a strong tendency to be bilateral in severity. Subjects attended and underwent general visual function and psychophysical testing as detailed in the general paradigm in chapter two. Patients were included only if this examination confirmed that the SRN was unilateral with normal visual function in the fellow eye. Twelve patients were identified matching these inclusion criteria, but two patients had to withdraw from the study (one from the subject's time constraints and the other due to a hip fracture sustained at home) leaving ten patients available for further assessment. Clinical details for these ten participants are shown in table 3.1.
Patients returned for pupillometry between 2 days and 2 weeks after the initial examination. Patients were asked to abstain from caffeine, stimulants, chocolates and undue physical exercise for 24 hours before the recordings were made. After 20 minutes of dark adaptation, stimulation of the eye and pupil recordings were performed using the P_Scan 100 infrared pupillometer described in chapter two. Stimuli were presented in non- Maxwellian view on a standard cathode ray tube (CRT) display containing luminant phosphors (liyama, Japan, Vision Master 506, Model MS103DT). The stimulus consisted of a square patch of sustained white light subtending an angle of 22 x 14 deg (duration 1100 ms). Viewing distance was 1m and the untested eye was occluded with a patch. Alternatively for chromatic stimuli the interference filters were used. The investigator viewed a CCTV image of the subject's eye and initiated the presentation of each stimulus when the eye was aligned with the camera and the subject comfortable. An audible cue signalled to subjects that they might blink if they wished to, and after which blinking was to be suppressed until the next audible cue or otherwise as instructed. Each audible cue was followed by a small random time delay (1 to 2 s) that allowed time for blinking, after which the stimulus was presented.
174 Participant Aetiology RAPD Age
1 Infective. Presumed ocular No RAPD 39 M
histoplasmosis. Untreated.
2 AMD with predominantly In Diseased Eye 53 F
classic SRN. Treated by three
courses of PDT.
3 AMD with occult membrane. In Diseased Eye 53 F
Untreated.
4 Inflammatory. Suspected No RAPD 55 F
punctate inner choroidopathy
(PIC) complicated by SRN
formation several years
previously. Untreated.
5 AMD. Classic membrane In Diseased 60 F
treated by three courses of PDT. Eye
6 AMD. Occult membrane. In Diseased Eye 68 M
Untreated.
7 AMD. Disciform scar. In Diseased Eye 69 F
Untreated.
8 AMD. Classic membrane In Diseased Eye 71 M
treated by one course of PDT.
9 AMD. Disciform scar. In Diseased Eye 77 F
Untreated.
10 Infective. Presumed Ocular No RAPD 82 F
Hi stoplasmosis.
Table 3.1 Clinical Features of the 10 participants with unilateral subfoveal subretinal membranes
175 A small adapting stimulus was used before each sequence of stimulus intensities. Stimuli were presented at 7 light intensities (range 2 log cd/m2). The duration of each stimulus was 1100 ms, and stimuli were presented five times for a given light intensity before transferring the patch and repeating the experiment on the fellow eye. The time interval between identical stimuli was over 10.5 s. The 7 stimulus intensities were presented in ascending order of intensity to minimise the effect of light adaptation (a protocol similar to that recommended by the International Society for Clinical Electrophysiology of Vision (ISCEV) for electrophysiological recordings described in the previous chapter). Following 20 minutes of dark adaptation, the time taken to record all 70 reflex responses in both eyes was approximately 30 minutes (7 stimuli presented 5 times = 35 stimuli per eye, x2 for each eye = 70). Recording was initiated 500 ms before stimulus onset and continued for a total of 10 s afterwards.
3.4 Calculation of a New Index - Pupil Evasion Ratio (PEVR)
The following parameters were derived from each averaged pupillogram: reflex amplitude (RA) in mm and mid-dilation amplitude (MDA) in mm. The latter was calculated by estimating the total time (t) to recover from maximum constriction back to baseline diameter then measuring the degree of recovery in mm after half of this time (t12), shown in Figure 3.2. An index from these parameters was derived which we have termed the pupil evasion ratio (PEVR), defined as: PEVR = (RA/MDA) * 100%. Pupil evasion represents a similar concept to pupil escape, except the stimulus is much shorter than that for pupil escape which is conventionally a few seconds long. The multiplication by a factor of 100% is in common with previous electrodiagnostic indices such as the electro-oculogram which have used this conversion factor to achieve a readily understandable and easily usable index [Arden et al 1962; Papakostopoulos et al 1992]. For every participant and for each eye tested, PEVR was estimated at all seven stimulus intensities. Pupil constriction data and data for pupil evasion/escape were normally distributed. Mean data points in each
176 comparison were less than 20 in each analysis but more than 20 if raw unaveraged data was used; in view of this both a parametric test (paired t- test) and later in view also of the fundamental significance of the ratio of PEVR to subsequent studies a non-parametric test was applied (Wilcoxon matched-pairs signed-ranks test for paired data rather than a Mann-Whitney U-test).
Diameter mm 7.75 _
7.25
6.75
6.25
5.75 11111.111 oi n iirnrmmiliimmmil miumiiiiii miimnioI1111iimm11111 11 1 III 111 111 111 1 1 1 11 11 1 111 III 111 11 III 1 1 1 o o 0 0 0 0 0 0 0 0 0 Cr CO N (SD 0 t CO N CO 0 Time s CO CD 0 CO N- 0 CO fL 0 N N N CO CO
Figure 3.2 Pupillogram showing pupillary diameter as a function of time. The stimulus was presented at 500 ms of duration 1100ms. After a brief latent period the pupil constricts reaching peak constriction amplitude at 1400 ms. The white arrow shows this reflex amplitude [RA]. The pupil then starts to dilate, reaching maximum dilation by 3400 ms after which there was no significant extra dilation for 10 s from the onset of the stimulus. The time interval between RA and the onset of maximal dilation (shown in light blue) represents the time t for dilation. Dividing t by 50% gives t / 2, which corresponds to a pupil diameter in mm shown by the dark blue arrow and called mid-dilation amplitude [MDA]. Pupil Evasion Ratio or PEVR is then calculated as (RA/MDA)*100%.
177 3.5 Variation of PEVR
3.5.1 PEVR in Healthy Eyes
Measurements of the mid-dilation amplitude (MDA) were linearly correlated with the amplitude of the reflex response (RA) for all subjects. An example is shown in Figure 3.3a in which it is seen that the larger reflex amplitudes produced by higher intensity stimuli were associated with bigger MDA measurements. This covariation masks any independent effect of macular disease on pupil evasion (analogous to escape) and so MDA estimates were normalised with respect to reflex amplitude by calculating the pupil evasion ratio (PEVR = (RA/MDA)*100%). Figure 3.3b shows how this estimate is independent of reflex amplitude and better serves as an index of pupil evasion (analogous to escape), with low values for PEVR indicating greater 'evasion'.
It is worth noting that estimates of PEVR varied considerably between individuals for the same stimulus intensity, ranging in the order of <10% to >1000%.
178 a) -0 z 0.35 :r. 0.3 1:2E" 0.25 ct s 0.2 sc 0.15 - 0.1 g' " 0.05 0 -6 '\ E sz.‘ •cbb rb oq' Cot'' gi c§ cs- o• Q. K. • h., • N. • h. • Constriction Amplitude (mm)
500 450 400 -• 350 2, 300 CC 250 - - w 200 • CL 150 • 100 • 50 0 0 0.5 1 1.5 2 2.5 Constriction Amplitude (mm)
Figure 3.3 (a) Relationship of mid-dilation amplitude (MDA) to maximum constriction amplitude in a participant (top); (b) relationship of PEVR (%) to constriction amplitude for the same participant (bottom). The correlation coefficient is approximately 0 and R2 =0.02. Note for clarity only one regression line (horizontal) is shown; vertical and horizontal lines intersect.
3.5.2 Comparing PEVR between Diseased and Control Eyes
PEVR was consistently smaller in the eyes with SRNs compared with the healthy fellow eyes. Figure 3 shows three representative examples. Figure
179 3.4a was recorded from patient CJ who had a large subfoveal predominantly classic SRN, 3000 tm in diameter, treated with 3 courses of photodynamic therapy. At all stimulus intensities except 0.6 log cd/m2 the PEVR estimates appear grossly reduced in the diseased eye. A paired t-test confirmed that this difference was significant (mean difference in PEVR between SRN and control eye = 1954%; p=0.002) as did a Wilcoxon matched-pairs signed- ranks test (p=0.016) both implying that in this patient the macular disease was associated with significantly greater pupil escape. In most subjects the difference in PEVR was not as significant: a more typical example is illustrated in Figure 3.4b showing measurements from patient LB who had a large subfoveal age-related disciform scar 1500 tm in diameter but untreated.
Despite smaller differences in PEVR estimates between diseased and control eyes (range 5 to 30%) a paired t-test confirms the significance of this result (average difference = 12.9%; p=0.005 paired t-test and p=0.016 using a Wilcoxon matched-pairs signed-ranks test the latter with very little variation in significance between subjects). The least convincing case is nevertheless the subject that was first studied and in whom this phenomenon was detected.
180 3050 2550 - - -4- - - Healthy Fellow t- 2050 4 Eye ce • > 1550 SRN o_ 1050 550 9 • 50 4 44 4,0. .0 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4 Light Intensity log cd m-2
200 180 .4 . - -4- - - Healthy Fellow 160 ..4 Eye _ . - - '4 - - - op. - • ..,...... „.-4 - 0 .e.. 140 Av —4-- SRN 120 100 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4 Light Intensity log cd m-2
195
175 ---4--- Healthy Fellow Eye ce 155 tu --4-- SRN 135
115 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4 Light Intensity log cd m-2
Figure 3.4 Variation in estimates of PEVR (which is reduced by SRNs) over 7 light intensities in the diseased (SRN) eye compared with the healthy fellow eye of three patients: (a) (Top): patient CJ with a large age-related subfoveal SRN previously treated with PDT; (b) (Middle): patient LB with a large but untreated age-related subfoveal SRN; (c) (Bottom): patient MO with a small <500 p.m subfoveal SRN of presumed inflammatory origin.
181 The case is illustrated in Figure 3.4c; this participant (MO) had a small <500 p.m subfoveal SRN of presumed inflammatory origin that had not been treated. PEVR estimates were distinctly smaller in the diseased eye at 4 out of the 7 stimulus intensities and a paired t-test suggests that the average difference was of borderline significance, or of significance if the lowest stimulus intensity is disregarded (range 1 to 100%; mean difference in PEVR = 25%; p = 0.05 (paired t-test) and 0.016 (Wilcoxon matched-pairs signed-ranks test) , or p = 0.04 (paired t-test) and 0.03 (Wilcoxon matched- pairs signed-ranks test) if the lowest intensity stimulus is disregarded).
There is a tendency for the difference in PEVR between diseased and healthy eyes to be most apparent at higher stimulus intensities; this is illustrated in figure 3.5.
600 400 t. 200 ea 0 r • 2 -200 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4- -400 Light Intensity log cd m-2
200 — 150
100
S3) 50 2 0 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4
Light Intensity log cd m-2
Figure 3.5 Variation of mean difference in PEVR between eyes with stimulus intensity (10 subjects). The mean data points shown in 3.5a (top) are magnified in the graph shown in 3.5b (bottom). Error bars signify 95% confidence intervals (mean +/- 2 SD). As PEVR is a ratio in accordance
182 with strict mathematical notation negative values are not shown for the error bars.
It can be appreciated there that variation in PEVR estimates between different individuals result in wide 95% confidence intervals (mean +/- 2 standard deviations (SD)) around the data points for each stimulus intensity. Nevertheless a trend does emerge wherein the largest differences in PEVR were generally observed at the highest stimulus intensities.
Overall, comparing PEVR between diseased and control eyes, a statistically significant reduction in PEVR (signifying greater pupil evasion) was found in the diseased eye in 9 out of 10 patients, and with one subject (MO, Fig. 3.4c) having a difference of borderline significance. Since PEVR is greatest at higher stimulus intensities, if the lowest stimulus intensity is ignored a statistically significant reduction in PEVR is found in the diseased eye in all subjects: range p = 0.002 to p = 0.04 (paired t-test) and p = 0.03 (Wilcoxon matched-pairs signed-ranks test) the lack of variation with the latter suggesting PEVR is a generally robust test.
There was no apparent association between reduction in PEVR and the aetiology of the SRN. There is a possible association with size of SRN at extremes. The extent of SRNs on FFA ranged from 3000 pm to <500 pm in diameter, and these specific cases were associated respectively with the largest and least statistically significant difference in PEVR between fellow eyes, including at all stimulus intensities (p < 0.002 and 0.05 respectively, mean difference in inter-eye PEVR 1954.0% and 25.0% respectively). This suggests an association between PEVR and visual field may be likely. However a Chi-Squared test showed no significant overall association between size of SRN and reduction in PEVR when all the participants were included (p > 0.1). It is to be noted however that fundus fluorescein angiography is predominantly an imaging tool displaying the anatomical extent and morphology of SRN, and while studies have shown excellent correlation with functional indices in retinal ischaemia, relationship to
183 function may be less exact outside ischaemic conditions [Maumenee 1969; Bloom et al 1993]. There was no association between degree of loss of visual acuity (counting fingers or worse in all diseased eyes) and extent of PEVR reduction, presumably as all SRNs involved the fovea in this pilot study. Gender, age and treatment status had no association with reduction in PEVR.
3.6 Significance of PEVR
In healthy subjects, onset of a light stimulus causes an initial rapid constriction of the pupil. If the light stimulus is maintained, the pupil starts to slowly dilate, sometimes as quickly as after a few hundred milliseconds. This phenomenon is known as 'escape' and arises because of both retinal and central adaptation processes [Akamatsu et al 1979; Cox 1992; Young et al 1993; Bergamin and Kardon 2002; Cox 2003]. The precise mechanisms of this adaptation are not known but probably include photopigment bleaching, disappearance of the phasic component and diminution of the tonic component to the 'ON' signal in the afferent limb of the reflex, and re- setting of the gain in the reflex loop at the level of the midbrain [Akamatsu et al 1979; Young et al 1993; Loewenfeld and Lowenstein 1999]. Patients with lesions of the anterior visual pathways show exaggerated pupil escape, as first reported by Marcus Gunn [Gunn 1897 and 1902; Thompson 2003]. A number of methods to quantify escape have been suggested. Kestenbaum measured the difference in pupillomotor input between the eyes [Kestenbaum 1946; Thompson 2003]. This has been found by Jiang and colleagues to roughly correspond to the relative afferent pupillary defect (RAPD) measured in log units of neutral density filter [Jiang et al 1989]. Bergamin and Kardon have logically suggested that variations in the RAPD found with different diseases may be due to the alternating frequency of the swinging flashlight and thus be influenced by whether it is the transient constriction or pupil escape that is being observed [Bergamin and Kardon 2002; Cox 1992, 2003; Zaidi and Moseley 2004].
184 Pupil escape can be clinically observed, but has proved a demanding quantity to put a reliable measure on. Finding an objective, quantifiable and statistically reliable association between pupil escape and disease in the afferent visual pathway disease has been challenging. Measurement of pupil escape in relation to the peak of the sustained contraction (and not as an index of the peak phasic contraction, though the two are to some extent related) was shown by Cox to be ineffective in showing significant pupil escape quantitatively in eyes with afferent damage compared to normal eyes [Cox 1992]. Akamatsu et al showed a shortened time period of less than 12 seconds (normal range 26 s, +1-12.7 s) between the beginning of the stimulus to when pupil area was greater than 63% of baseline in 13 out of 16 eyes with optic neuritis, taken to indicate a greater pupil escape, most likely in association with a central field loss [Akamatsu et al 1979]. It was however uncertain if there was to some extent confounding secondary to a smaller stimulus and thus constriction amplitude in the eyes with optic neuritis. Most recently, Bergamin and Kardon convincingly showed that when pupil escape is expressed as a ratio incorporating the amplitude of the preceding phasic contraction, the resultant pupil escape ratio can be used to indicate greater pupillary escape [Bergamin and Kardon 2002; Cox 2003].
In the experiments that have thus far been presented in this chapter I extended the same principle that was successfully applied by Bergamin and Kardon, namely comparing pupil dilation within an index as a proportion of peak phasic constriction, but also with a much shorter light stimulus than conventionally used to elicit pupil escape. It was found that when mid- dilation amplitude is expressed as a function of peak phasic constriction, the ratio (called to avoid confusion with PER, PEVR) is reduced in eyes with SRNs, denoting greater pupil evasion. We call this phenomenon 'evasion' to avoid confusion with 'escape' which tends to imply the response to a prolonged stimulus. PEVR shows wide fluctuations between eyes, with some overlap between diseased and normal eyes, and very wide 95% confidence intervals (figure 3.5) consequent to a compounding mathematically of the well-described effect of variability in pupillary
185 responses. Nevertheless, as shown in figure 3.5, inter-eye PEVR tends to be greater with higher stimulus intensities, which is a further similarity of PEVR with PER. Unlike mid-dilation amplitude, PEVR was largely unrelated to maximum contraction amplitude, and was thus not artefactual.
For clinical purposes it is important to note that the inter-eye PEVR was for all stimulus intensities greater in normal eyes compared to the fellow eye with SRN (i.e. the eye with SRN displayed greater pupil evasion). Inter-eye comparison of PEVR yielded statistically significant results in all subjects at the higher light intensities, and in 90% of subjects across all light intensities. This is probably as the arousal state of the subjects was the same when comparing inter-eye pupil responses separated by a few minutes as opposed to with pooled groups of normal and healthy eyes — similar levels of autonomic arousal are accompanied by a similar degree of background sympathetic tone in the iris muscles [Loewenfeld and Lowenstein 1999; Bergamin and Kardon 2003]. Further, pupil escape is associated with the brightness of the light stimulus [Akamatsu et al 1979].
There are several reasons therefore to believe that PEVR represents a similar functional index to PER, but instead involving a brief burst of light. First, there is similarity in the derivation of pupil evasion ratio using an index incorporating peak phasic constriction. Second, this ratio has the ability to differentiate greater pupil evasion, which had, like escape, historically been unsuccessful in objective terms until very recently when PER was measured in subjects with perimetric central field loss [Bergamin and Kardon 2002]. Third, the difference in PEVR is greater between fellow eyes at higher light intensities [Akamatsu et al 1979].
What would explain the ability of such ratios as used by Bergamin and Kardon [PER] and in this study [PEVR] to successfully identify eyes with macula dysfunction when simple measures of pupil escape or evasion such as mid-dilation amplitude fail [Bergamin and Kardon 2002]? A key similarity between these ratios is the reliance on comparison with maximum
186 phasic constriction (called reflex constriction in this study). In the 1960s it was suggested that the pupil could be modelled using both a rate-sensitive process filtering temporal information (the transient constriction) and a proportional process dependent on the steady-state quantity of the same retinal image (the sustained component) [Clynes 1961; Kohn and Clynes 1969; Lowenstein and Loewenfeld 1969]. This was later supported by work in monkeys showing that the phasic broad-band pathway (P-Gamma and P- Epsilon cells) projects to the pretectum [Schiller and Malpeli 1977; Levanthal et al 1981; Perry and Cowey 1984]; phasic cells have also been found along with tonic (sustained) cells in the pretectal area of the rat [Trejo and Cicerone 1984]. In 1993 principal component analysis suggested that the sustained component neurones started to fire at about the same time as the transient, though as expected their firing rate peaked later temporally [Young et al 1993; Young and Kennish 1993]. In this sense the phasic- sustained response was similar to the geniculo-cortical response to light [Young et al 1993; Livingstone and Hubel 1988a and b]. The accentuation of pupil escape in larger pupils is felt by Sun and Stark to be related to the relative excess of neuronal firing in tonic (sustained) channels in photopic conditions as opposed to phasic firing in mesopic and scotopic conditions [Sun and Stark 1983]. Studies in rats suggest that firing in tonic-on cells of the pretectal olivary nucleas integrate signals from tonic-on centre retinal ganglion cells whose neuronal firing signals for constriction of the pupil, and eventually to neurones in the oculomotor nucleus, while tonic-off cells, which are much fewer in number, mediate pupillary dilation. [Trejo and Cicerone 1984; Clarke and Ikeda 1985 a and b]. Tonic-on pretectal cells in the rat are also retinotopically organised and are aggregated in a strip running from the dorso-medial tip of the pretectum to the ventro-lateral boundary, suggesting similarity with such organisation within the lateral geniculate body [Trejo and cicerone 1984; Livingstone and Hubel 1988a and b]. It may be postulated that it is therefore the relative loss of phasic to sustained neuronal firing in retinal ganglion cells in the macula that enables pupil evasion to be distinguished when expressed as PEVR. This is as such an index incorporates phasic constriction (representing the sum neuronal
187 firing from both sustained and phasic on-ganglion cells) in a form which allows it to be compared with the subsequent loss of firing from sustained on-ganglion cells (Table 3.2).
Effector Response (Pupil Afferent Response (Ganglion Dynamics) Cell)
Initial 'Phasic' Constriction Combined Transient (Phasic') & 'Sustained' Neuronal Firing Pupil Escape (mid-dilation Sustained Neuronal Firing only amplitude)
Table 3. 2 Proposed relationship between pupil dynamics and the cellular processes they may reflect in the afferent arm of the pupil light reflex
Mathematically such a comparison is not possible when pupil evasion or pupil escape are measured using response parameters which fail to incorporate components from both classes of ganglion cell. This may explain why workers not using phasic constriction amplitude as part of their calculation have been unable to find an effective quantitative way to discriminate between pupil escape in eyes with afferent damage compared to normal eyes [Cox 1992]. The process of comparison with phasic constriction (RA in this study) algebraically produces a simple yet reliable estimate of the ratio or balance of firing between sustained and phasic neuronal elements (phasic and sustained parts of the pupillogram), as well as their disruption, as occurs with optic neuritis [Akamatsu et al 1979], central field loss [Bergamin and Kardon 2002] and in this study central subretinal membranes. The latter pathology is histologically characterised by dense fibrovascular proliferation, often with associated haemorrhage (figure 3.1), and is located anatomically either beneath the retinal pigment epithelium, superficial to it, or in both sites, with consequent extensive
188 atrophy of the adjacent rod and cone photoreceptors accompanied by devastating visual consequences [Gass 1967 and 1994; Green 1985; Bressler et al 1988; Chang et a11994; Schnurrbusch 2001].
With more gross cases of damage to the afferent visual pathway this ratio is likely to be analogous to changes found in the RAPD when measured using short or longer stimulus durations as has been suggested previously with pupil escape [Bergamin and Kardon 2002]. In addition to being objective it is also likely to be more sensitive, for example in this study 70% of patients had clinically detectable RAPDs (Table 3.1) while all had significant differences in PEVR between eyes at higher stimulus intensities. With more subtle cases of damage to the anterior visual system, PEVR can facilitate reliable differentiation of macula function in diseased eyes. For example, figure 3.6 shows the pupil waveforms produced in this study by a stimulus of 14 cd m -2 in the subject with the smallest SRN [Subject MO] and least significant difference in PEVR between eyes.
189 Diarretex nut
• e F r
Diameter nun
Time ms
Figure 3.6 Pupillogram from a normal eye is shown in (a) (top) and that from its fellow eye with a subretinal membrane in (b) (bottom, in bold). These recordings are from Subject MO who had a small subretinal membrane and the least difference in pupil evasion between eyes out of all participants in this study. Despite the suspicion on inspection of the pupillograms that pupil evasion (analogous to escape) is greater in fig 6a, PEVR correctly identified the pupillogram shown in fig 6b as having greater evasion. This subject did not have a clinical RAPD. Note the different constriction amplitudes (much steeper in the healthy eye) between the two pupillograms that can create a misleading impression of evasion that PEVR eliminates by incorporating the reflex amplitude constriction.
On gross inspection of the pupil waveform there appears to be greater evasion (analogous to escape) in the pupillogram shown in figure 3.6a. However, this pupillogram is from the subject's normal eye. The fellow eye, shown in figure 3.6b, has the subretinal membrane. The fact that the
190 constriction amplitude is much less in the eye with the SRN creates the misleading impression with such a small amount of evasion that there is more evasion in the healthy eye, simply as constriction is more rapid in the healthy fellow eye. Taking this phasic reflex amplitude into consideration in a mathematical ratio enables the greater evasion in the diseased eye to be detected. On this basis it is possible that PEVR may be more sensitive to afferent damage than PER. This may be as PEVR is measured in response to a shorter duration stimulus incorporating a greater proportion of transient neuronal firing in the healthy eye, to which the PEVR in the diseased eye is compared. Such a comparison is worth further study.
At a more fundamental level the fact that PEVR is less (and hence evasion is greater) in eyes with severe macula pathology such as subretinal membranes is likely to reflect the activity of photoreceptors that drive ganglion cell output to the brain. The fact that the macula contains a far greater proportion of cones than the retinal periphery suggests they may have an important role in driving sustained pupillary constriction. That would also be a physiologically plausible function as cones are sensitive to light of high intensity and would hence be likely to serve a useful purpose in facilitating sustained pupillary constriction.
It is hoped that clinically this index will help developing means to reliably differentiate diseased from normal eyes using pupillometry of subjects and patients. Further studies are needed beyond this pilot work to assess the reliability of this index in very large numbers of patients and with other pathologies, including the relationship to visual field. This ratio and others like it have the capacity to be used to study retinal processing using pupillometry, including responses to colour.
3.7 Use of Long, Medium and Short Wavelength Chromatic Stimuli
Within the same experimental paradigm, instead of using steps of achromatic light, the same 10 subjects underwent stimulation using separate
191 trains of radiometrically equivalent long, medium and short wavelength stimuli (11-12.5 log photons/cm2/s) using monochromatic interference filters. The frequencies were, respectively 440 nm, 540 nm and 620 nm. These tests were done at separate times sometimes during the same day, sometimes on other days, as per the subjects' time constraints, ability to be compliant with instructions (e.g. abstaining from caffeine), and levels of fatigue.
3.8 Establishment of a Chromatic Threshold
The threshold for pupillary response was elevated with long wavelength light as compared with medium wavelength light as compared with short wavelength light, which was slightly higher than the opposite healthy eyes. For comparison, the responses for long versus short wavelength light are shown in figure 3.7.
3.9 Pupil Evasion using Chromatic Stimuli
There was no relationship between PEVR and the use of short and medium wavelength light. PEVR was consistently reduced (signifying greater pupil evasion) in all the diseased eyes of all subjects stimulated with long wavelength light.
192 Pupil constriction mm 2.5
2
1.5
1
0.5
0 11.0 11_25 113 11.75 12.0 12.25 123 LOG Photon Flux photons cm
Pupil constriction mm 2.5
2
1.5
I
0.5
0
11.0 11_25 11.5 1135 12.0 12.25 12.5 LOG Photon Flux photons / cm
Figure 3.7 Comparison of effect of long wavelength light (a) (top) and short wavelength light (b) bottom, on pupil response threshold in the 10 eyes with SRN.
193 3.10 Significance of Chromatic Stimuli
The raised stimulus threshold with longer wavelengths of light may be a consequence of the relative preservation of both rods and S-cones in SRNs, both of which are proportionately less affected than M- and L-cones which are more greatly concentrated within the macula and which hence have peak spectral sensitivities to higher wavelengths.
The unusual findings with regard to PEVR suggest a factor in operation at lower stimulus wavelengths of light (and not particularly in white light), and that this is probably found in the dilation segment of the pupil waveform. Could this be S-cones? This is possible, but there are very few of them. Of note, ganglion cell photoreceptors in animals have a peak spectral sensitivity to blue light, raising the question as to what degree they might be implicated.
3.11 Findings of Electrophysiology
Electroretinography confirmed the presence of statistically significant cone system dysfunction in SRNs (table 3.3). This was evinced by a statistically significant difference in 'a' wave amplitudes compared with normals (n=20) for photopic stimuli (p = 0.02, paired t-test) as compared with scotopic stimuli (p = 0.4, paired t-test). `b' wave amplitudes were reduced, but were not statistically significant compared with controls (the average p value for the three stimuli, p = 0.08 for photopic stimuli and p = 0.2 for scotopic stimuli, paired t-tests).
194 Normal SRN Normal SRN Stimuli 'a' 'a' 'b' 'b' waves waves waves waves microV microV microV microV scotopic 25 dB 6 5 62 67 scotopic 20 dB 10 11 70 52 scotopic 10 dB 45 36 136 140 scotopic 0 dB 64 70 181 160 flicker 30Hz 16 10 56 42 photopic 10 dB 15 12 32 27 photopic 0 dB 21 17 76 45
Table 3.3 Flash electroretinography of 10 eyes with SRNs. Normal controls (n=20 eyes) aged 30-42, 50% male, 50% female, as documented in chapter two.
3.12 Effect of a Sustained Stimulus
One young subject (#1) consented to have in addition long duration stimulation with a 5 s sustained stimulus, using both achromatic and chromatic stimuli. The same stimulus paradigms already described with achromatic and chromatic light were used, except that stimulus duration was 5 s. Pupil escape was reduced with achromatic stimuli and the 600 nm stimulus, but there was no relationship with the short and medium wavelength stimuli. There were insufficient subjects in this limb of the experiments to reliably test for thresholds to long wavelength stimuli.
3.13 Significance of Results
PEVR is a novel ratio and is seen to very closely parallel PER. However to what degree this is the case, and which of the two is the more sensitive parameter, would need to be addressed in a larger study, ideally using younger subjects who can more readily tolerate the very long stimulus protocols. Subjects with SRNs are not the ideal group to use to investigate this in. However as it is a not uncommon pathology in older subjects with
195 cataracts and other media opacities, PEVR may be used to estimate visual potential in such subjects prior to potential surgery such as corneal transplantation, cataract surgery or vitrectomy for vitreous opacities.
Inter-eye pupil evasion and escape were found to be greater in eyes with SRNs. This relationship was only found with achromatic light and long wavelength stimuli. It could not be elicited with medium and short wavelength light. These findings should ideally be investigated using other disease models for macula dysfunction, as well as ganglion cell dysfunction, to ensure the effect is not specific to SRNs. This investigation is taken further using retinal detachment and orbital pathologies in subsequent chapters.
196 Chapter 4 Retinal Detachment and the Pupil
4.1 Core Concepts
A retinal detachment (RD) is a split in the retina between the anterior neurosensory retina derived from neuroectoderm and the posterior RPE. A RD can be due to a break in the retina, termed a rhegmatogenous RD, or due to other causes, termed non-rhegmatogenous, which includes traction, tumour mass and inflammatory exudate. The resulting visual deficit associated with a RD can be devastating, especially if the macula, which mediates central vision, is affected. Surgery is indicated for RD, and is usually undertaken as soon as reasonably possible, or even as a subacute emergency if the macula is still attached in order to prevent extension of the RD to involve the macula. If, however, the macula detaches visual prognosis is often considerably more limited. An important clinical distinction is thus between macula-on and macula-off RDs, with the former requiring very urgent surgical intervention.
4.2. Cellular, Histological and Functional Changes in Retinal Detachment
There are few studies of the effect of RD in humans, which requires enucleation of an eye. Recently studies of a post-enucleated eye from autopsy confirmed the older findings that dated from RDs induced experimentally in animal eyes [Sethi et al 2005; Wickham et al 2006]. Changes include outer photoreceptor shortening as well as inner retinal disruption occurring within weeks, and according to one authority some damage occurring almost instantaneously, potentially with a strong element of reversibility if the retina can be reattached by surgery; histological recovery has been documented in animal studies and functional recovery
197 can clinically occur within weeks [Hogan and Zimmerman 1962; Anderson and Guerin 1986; Guerin et al 1989; Tack et al 2001].
4.3 Clinical Significance of the Pupil in Retinal Detachment
The cellular deficit caused by a RD is significant enough to causes a deficit in afferent neuronal firing and thus pupil constriction from the affected eye. Clinical studies have shown that detachment of each retinal quadrant reduces pupil constriction by about 0.35 log units; the fractionated log units being quoted are the notation preferred by Folk et al in their paper and correspond to an approximately 20% reduction in pupil constriction [Folk et al 1987]. Detachment of the macula causes a 0.68 log unit reduction in the amplitude of pupillary constriction in the fractionated log units preferred by Folk et al, corresponding to approximately 45% reduction in constriction amplitude. Macula-Off RDs are usually associated with detachment of both retinal periphery and macula, and the resulting pupillomotor asymmetry is considerable enough between eyes to produce a RAPD [Bovino and Burton 1980, Folk et al 1987]. Despite the wide fluctuation in the extent of pupillomotor deficit in RD, a RAPD can still be found in most cases of Macula-Off RD.
Diagnosis of RD is usually by fundoscopy (figure 4.1), but it is not uncommon for the fundal view to be obscured by concurrent vitreous haemorrhage accompanying the RD. The status of the pupil is a key diagnostic and prognostic guide in such cases, especially as diagnostic imaging with ocular ultrasound is, not uncommonly, challenging to interpret in even experienced hands owing to the presence of blood in the vitreous, which is a not infrequent scenario since haemorrhagic RDs are often associated with haemorrhagic posterior vitreous detachment which can produce similar radiological features [Restori M 1998].
198 Figure 4.1 Fundal view of retinal detachment (shown inferiorly in this colour fundus photograph). The macula is the area to the left of the optic nerve head, and is threatened should this "barely Macula-On" RD extend. The bullous nature of this particular rhegmatogenous RD due to a retinal break slightly overestimates the extent of macula involvement if clinical fundoscopy alone is used. Cases such as this could benefit from more accurate diagnosis of anatomic status at the macula, and the presence of a RAPD would strongly suggest macula dysfunction.
4.4 Potential for Pupillometric Study to Aid Diagnosis of Selected Retinal Detachments
There is a greater density of cones in the macula than in the periphery. Work presented in this thesis using SRNs suggested that macula dysfunction manifests as changes in inter-eye PEVR. Like SRNs, macula- Off RDs also are associated with dysfunction in the macula. Could there then be, as with SRNS, an association between Macula-Off RDs and components to the PLR such as amplitude and PEVR? The latter parameter has never been studied in RD. Would decrements in responses to chromatic stimuli still be found as they were with SRNs in the previous studies presented in this thesis? Further, since RDs can be Macula-off or Macula- on, differences in pupil response components could be compared between
199 the two groups of RD, where Macula-Off RDs involve dysfunction in a relatively greater proportion of cones than Macula-On RDs. Any findings would, additionally, be of relevance to understanding PEVR through an additional disease model and be of relevance to understanding the puzzling inability noted in the last chapter of PEVR to resolve macula function unless achromatic and long wavelength stimuli are presented.
Clinically, a potential application of any positive results might be to aid diagnosis of macula-off RD in the presence of ocular media opacities, as well as macula dysfunction more generally in other pathologies, thus augmenting the existing work already presented with SRNs. As this series of experiments was a pilot study, RD was studied in isolation of media opacities.
4.5 Materials and Methods
Twelve subjects with RD were recruited by collaboration with vitreoretinal sub-specialists at the Western Eye Hospital and liaison with the Clinical Fellow in that department. Inclusion criterion was unilateral RD. Diagnosis including of macula status was made on fundoscopy by vitreoretinal subspecialists. Seven subjects had Macula-Off RDs and five subjects had Macula-On RDs. Subjects with other ocular or relevant systemic diseases were excluded. Many cases of RD are emergencies, and a very significant limitation to recruitment was non-experimental pupil dilation, usually undertaken with mydriatics routinely as part of clinical examination. This put considerable restrictions on recruitment which was performed by close collaboration and timing with the Fellow in post at the time at the Western Eye Hospital. Recruitment necessarily therefore spanned a period in excess of one year.
RDs varied in extent from one-half the retina to one-fifth of the retina. Cases that were Macula-On were old chronic inferior RDs where the urgency of surgery is less important as the subretinal fluid that causes
200 extension of a RD is limited somewhat by the effect of gravity and posturing. It was not possible to conduct pupillometry on other acute Macula-On RDs as the interval between surgery and administration of mydriatics for clinical examination did not allow the mydriasis to fully wear off within 48 hours of mydriatic eye medication having being administered to the subject. The experiments were conducted as in the previous chapter except that RDs were studied. As before subjects were dark-adapted and each stimulus was presented seven times in ascending order of intensity. Trains of achromatic stimuli were presented using the CRT, and three radiometrically equivalent chromatic stimuli of 440 nm, 540 nm and 620 nm, using monochromatic interference filters. The standard general experimental paradigm was adhered to as described in chapter two. Owing to subject time constraints, notably that many were due for imminent surgery that morning, it was not possible to study the effects of a prolonged stimulus on pupil escape.
Electrophysiology was undertaken in accordance with the description in chapter two (ISCEV criteria). Hence the first stimulus was presented after 20 minutes dark adaptation, with step-wise increments in intensity to reduce the effect of bleaching of photopigment. The series of scotopic stimuli were presented before the photopic stimuli.
4.6 Electrophysiology
Electrophysiology confirmed the presence of significant cone system dysfunction in Macula-Off RDs compared with Macula-On RDs (figure 4.2). Greatest loss of rod, cone, bipolar and ganglion cell activity was found in macula-off RDs, followed by macula-on RDs. Analysis of scotopic 'a' wave amplitudes below the cone visual threshold shows that rod activity is reduced in both types of retinal detachment, and with incremental increases in stimulus intensity under photopic conditions.
201 The 30 Hz flicker modality, which isolates cone responses, shows mean cone responses for all subjects are reduced in macula-off RDs to a greater extent that in macula-on RDs, with very wide confidence intervals for macula-on RDs in this stimulus paradigm suggesting a wide variation in the extent that some macula-on RDs are associated with widespread loss of cone function. By 0dB 'a' wave amplitudes from macula-off RDs start to saturate, presumably due to relatively greater loss of cone-mediated responses with macula-off RDs. The 0dB photopic stimuli show very wide confidence intervals for macula-on RDs, again suggesting substantially greater preservation of cone activity with macula-on RDs. 'b' wave responses shown at bottom of figure 4.2 essentially parallel those of the 'a' wave except that amplitudes are typically greater. Cone dysfunction was, however, far from total in even Macula-Off RDs presumably as no Macula- Off detachments were older than a few days.
202 50% male,female.Meanofallsubjects'responsesaredisplayed are shown(top)and'b'waveamplitudes(bottom). 'a'waveamplitudesare graphically. scot=scotopic;photphotopic.Errorbarsrepresentmean+/- smaller aspernormal.Tosummarise,bothrodand conefunctionwere Figure 4.2 2 SD.Forclarityonlysuperiorerrorsbarsareshown. 'a'waveamplitudes Macula-Off retinaldetachments.Normalcontrols(n=20eyes)age30-42, detachments responsessaturatedwithphotopicstimuli. Wideconfidence intervals withmacula-off retinaldetachmentswerebecauseofthevery wave reductiongenerallyparalleledeachother.With macula-offretinal reduced bybothmacula-onandmacula-offretinal detachments;a-andb- variable extentofmacular involvement.
'12,'Wave mlcroV 100 150 200 - 50 Flash electroretinographyofsubjectswithMacula-Onand n ISCEV StimulusParadigm e 203
. 0 Normal 0 Macula-Off Macula-On 4.7 Effect of Achromatic Stimuli on Pupil Constriction
There was a reduction in pupil constriction amplitudes in all subjects with RDs compared to their healthy fellow eyes, and this was more marked with Macula-Off status. This of course correlates with the well-known clinical findings of a RAPD in RD, especially with Macula-Off RD. Indeed all subjects with Macula-Off RDs had a RAPD in this study. However, comparing pooled data for the two groups with normals (figure 4.3), the reduction in constriction amplitudes for Macula-On RDs were statistically insignificant, presumably owing to inter-subject variability (p = 0.2, two- tailed paired t-test). Nevertheless Macula-Off RDs continued to elicit a consistent and statistically significant reduction in mean constriction amplitudes using even pooled data (p = 0.04, two-tailed paired t-test).
E 2-5 E 0 —4—Macula-On RDs * - - • - -Normal Eyes O •• • 0.5 0. 3 0
2 Log Light Intensity cd m-2
0.7 E 0.6 5 0.5 Macula-Off RDs 72'° 0.4 • - - •--- Normal Eyes 0.3 ° 0.2 1: 0.1 0
2 Log Light Intensity cd m-2
Figure 4.3 Mean pupil constriction amplitudes as a function of log light stimulus intensity for all subjects, comparing all Macula-On RDs with
204 pooled fellow healthy eyes (top) and all Macula-Off RDs with pooled fellow healthy eyes (bottom). This is to reduce effect of variability, for example it may be anticipated that Macula-Off RDs were certainly apparently more anxious owing to having recently lost central vision and being about to undergo surgery to potentially restore it, and might be reasonably expected to have a greater degree of autonomic unrest manifesting as pupillary instability in both eyes.
4.8 Effect of Chromatic Stimuli on Constriction Amplitude
With Macula-On RDs compared with normal eyes there was no consistent decrement in pupil constriction amplitude and no significant difference in constriction amplitudes to chromatic stimuli of progressive intensity between any of the three wavelengths (p = 0.88 to 0.97, two-tailed paired t- tests). However with Macula-Off RDs there was a considerably reduced pupil constriction amplitude compared to normal fellow eyes in response to light of long wavelength only (p = 0.042, two-tailed paired t-test) (figure 4.4), a borderline reduction in amplitude with medium wavelength light (p = 0.05, two-tailed paired t-test), and no significant difference in response amplitude between Macula-Off RDs and normal eyes for short wavelength stimuli (p = 0.08, two-tailed paired t-test).
Thus in general while achromatic stimuli were associated with reduced responses with both types of retinal detachment, using chromatic stimuli no such reduction could reliably be found with macula-on retinal detachments. However using chromatic stimuli macula-off retinal detachments were associated with reduced responses to long wavelength light, borderline responses to medium wavelength light, but no reduction with short wavelength light.
205 Pupil constriction mm 2.5
2
Macula-Off RDs 1.5 .4, Normal Fellow 1 Eyes
0.5
0 11.0 11.25 11.5 11.75 12.0 12.25 12.5 Log Photon Flux photons / cm2 / s
Figure 4.4 Effect of red light on pupil constriction amplitude, comparing the mean responses of Macula-Off RDs with the mean of their normal fellow eyes. While generally reduced even at low stimulus intensities, the response from Macula-On RDs saturates. Despite these patients being prone to very wide fluctuations in autonomic arousal immediately before surgery, the difference is still detectable. There is clearly greater residual cone function than with SRNs, in whom identical testing conditions enabled establishment of a minimum threshold response. This is presumably as the detachments were at most a few days old. But also, as M- and L-cones are the only photoreceptor found in the fovea, which in the group of SRNs studied was uniformly obliterated.
It is postulated that the discriminatory effect of long wavelength light in Macula-Off RD represents the relatively greater loss of M- and L- cones which are more densely concentrated in the macula, especially the fovea. The difference is less distinct with medium wavelength light, it is postulated, since there is proportionately greater overlap with the spectral sensitivity of the rod-rich retinal periphery, and still less with blue light
206 which is much nearer the spectral sensitivity of rods as well as blue cones. These results are practical evidence of the concept noted earlier that chromatic clinical RAPD estimates or chromatic pupillometry could be used as an index of macula status. Attempts have been made to use chromatic stimuli to elicit an RAPD clinically, specifically to elicit macula and optic nerve dysfunction, but the problem at the time of finding a strong enough portable stimulus was not overcome [Thompson et al 1980].
4.9 Effect on Pupil Evasion Ratio (PEVR)
PEVR was calculated as documented in chapter two. With achromatic light there was no pattern to the difference in PEVR between fellow eyes where one eye had a Macula-On RD. There was however a markedly reduced PEVR, signifying greater pupil evasion, in eyes with Macula-Off RDs (figure 4.5a). Using chromatic stimuli, with the 620 nm stimulus there was no difference in PEVR between fellow eyes with Macula-On RDs, but there was a reduction in PEVR, signifying greater pupil evasion, with Macula-Off RD (figure 4.5b). There was a slight trend towards a reduction in PEVR with the 540 nm stimulus, but only with Macula-Off RDs, and this did not achieve a level of statistical significance. There was no association between PEVR and either type of RD with the 440 nm stimulus. The statistically significant results are shown in figure 4.5. The data in these graphs also suggest a tendency towards greater PEVR with higher stimulus intensities as noted in chapter two with SRNs, and this seems even more marked with RDs.
207 29)0
2000 1500
LU 1000 MaculaRDs-Off 500 --II- Macula-On RDs III IS IF -500 -1000 -1500 11.0 11.25 115 1175 12.0 1225 123 Log Photon Flux photons / ce/ s
2500 - 2L100 - 1500 - ,x 1000 - w .500 - a. - 4— Macula-Off RDs -500 - —s— Macula-On RDs d -1000 -1600 - -2000 - -2500 - -3000 - 11.0 11.25 113 11.75 12.0 12.25 12.5 Log Photon Flux photons / cm' s
Figure 4.5 Inter-Eye PEVR as a function of stimulus intensity for Macula- Off retinal detachments versus Macula-On retinal detachments. Contrary to strict mathematical notation, for clarity in presentation as only one set of error bars can be shown per data point, error bars extending into negative values are shown even though the data deals with PEVR which is a ratio; (a) (top) achromatic white light (p = 0.01, two-tailed paired t-test) and (b) (bottom) long wavelength light (p = 0.04, two-tailed paired t-test).
208 Chapter 5
Orbital Disease: Diagnosing Compressive and Traumatic Optic Neuropathy
5.1 Compressive and Traumatic Optic Neuropathies and the Pupil
5.1.1 Overview
As discussed in chapter one, functional parallelism between visual and pupillomotor function is well-established. In terms of photoreceptor activation and subsequent conduction within the optic nerve, distinct chromatic and achromatic channels are recognised for both visual and pupillary responses. In this section the influence of orbital disease, and more specifically those intra-orbital lesions compressing the optic nerve, are considered.
5.1.2 Chromatic and Achromatic Visual Responses in Optic Neuropathies
A broad variety of achromatic and chromatic visual responses are reduced in optic neuropathies, including colour vision and contrast sensitivity, and in severe cases blindness can result [Miller and Newman 2003; Schwartz 2003].
5.1.3 Constriction Amplitude, Latency and Pupil Evasion in Optic Neuropathies
Reduced pupillary constriction amplitude and increased latency of constriction have been described extensively for optic neuropathies [Loewenfeld and Lowenstein 1999]. Certain diseases are associated with a relative sparing of visual or pupillary responses in relation to each other. Other parameters have also been studied. In addition to their work on
209 central field loss arising from macular lesions, Bergamin and Kardon have shown that pupil escape is greater in subjects with central visual field loss due to optic neuropathies, including in the few cases of compressive optic neuropathy that they studied [Bergamin and Kardon 2002]. It was shown in previous chapters that PEVR is associated with macula lesions. The possibility that, like pupil escape, it is also disturbed by optic neuropathies is clearly worth investigation given the similar effect of both macula disease and optic neuropathies on pupil evasion.
5.1.4 Diagnosis of Optic Neuropathies
5.1.4.1 General Principles
Optic neuropathies can be divided into the following types: compressive/traumatic, glaucomatous, inflammatory, vascular, and other e.g. infective, nutritional, drug-induced, and inherited. A wide range of colour vision defects have been described in optic neuropathies. However with most types of optic neuropathy colour vision defects, while an interesting area of study, offer relatively less specific diagnostic value than other investigations. The latter include assessment of the optic disc and intra-ocular pressure (in glaucoma studies), radiological imaging, and blood tests. Examples of blood testing include DNA analysis, autoantibody levels (especially anti-nuclear antibody and anti-neutrophil cytoplasmic antibody), full blood count, vitamin B12, folic acid, angiotensin converting enzyme, and tests for infection such as treponemal serology and immune status to tubercle bacilli.
5.1.4.2 Orbital Lesions: Compressive Optic Neuropathy
Compressive optic neuropathy may be associated with optic disc swelling due to initial oedema of ganglion cells owing to reduced axoplasmic flow [Miller and Newman 1998]. Congestive symptoms usually precede visual loss, especially with thyroid eye disease which is commonly bilateral,
210 gradual and symmetric, while tumours and traumatic haematomas tend to be unilateral unless there are associated injuries or the lesion is not arising from with the orbit (though even in the latter symptoms may frequently be unilateral). The neuropathy is considered compressive in thyroid eye disease as radiographic studies in the latter show extensive skeletal muscle swelling [Miller and Newman 1998]. Field loss with compressive optic neuropathies of orbital origin may comprise generalised depression, or often central scotomas frequently combined with arcuate scotomas. Eventually if unchecked the optic nerve atrophies. Loss of axons and glial tissue occurs to a variable extent. In compressive optic neuropathy both pupil reactions and colour vision tests are a major part of clinical assessment, as discussed previously in chapter one. However unlike other optic neuropathies, the effect of orbital disease is particularly serious. Unlike many other causes of optic neuropathy, it demands very urgent sight-saving intervention by oculoplastic / orbital surgeons and general ophthalmologists in the setting of the clinic, in inpatients, and in the setting of the accident and emergency department. Failure of diagnosis with appropriate intervention with high- dose systemic steroids and/or surgical decompression has devastating visual consequences, with some authorities claiming that the time from onset of colour vision loss to irreversible visual damage being four hours in cases of retro-bulbar haemorrhage [Lund et al 1997].
5.1.4.3 Ganglion Cell Function in Glaucoma
Pathologically glaucoma is the commonest optic neuropathy, and involves either compressive and/or vascular insufficiency in the optic nerve head [Miller and Newman 1998]. The aetiology and site of damage is different to most other compressive aetiologies in that compression arises within the eye from raised intra-ocular pressure. Various workers have shown both swelling of L/M- cones and ganglion cell death in glaucoma [Pacheco- Cutillas et al 1999; Nork et al 2000; Nork 2000]. However while the latter is uncontroversial findings that indicate classical photoreceptor loss may be artefactual [Quigley 2000]. On a superficial level loss of L/M-cones also
211 may not readily explain loss of blue/yellow colour vision over red/green loss in glaucoma [Miller and Newman 1998]. However the broad spectral sensitivities of L- and M- cones compared with S-cones (as discussed previously in the context of colour perception and the Principle of Univariance) does not preclude such a possibility. Other chromatic defects have been studied in glaucoma. Chromatic and achromatic stimuli have been compared [Pearson et al. 2001]. Stimulus increments of red-on-white, blue-on-white and critical flicker frequency were used to isolate, respectively, the red-green chromatic mechanism, the blue-on-yellow chromatic mechanism, and the high-frequency flicker achromatic mechanism; of note, some authorities feel it is possible for blue colour vision loss to occur separate to blue-yellow loss [Tregear et al 1994]. Using 3.1 deg circular stimuli, Pearson and co-workers found that chromatic defects were commoner in subjects with glaucoma than were achromatic defects. This apparent anatomical and psychophysical discrepancy found by this and other studies was felt to possibly be due to differences in cortical summation of ganglion cell responses for the chromatic and achromatic pathways, though it was conceded that the precise anatomical origin of colour vision losses in glaucoma has never been convincingly explained on the basis of existing knowledge of photoreceptor, ganglion cell and cortical function. Blue-yellow colour perimetry has been used most often to isolate glaucomatous scotomas in colour perimetry [Serra et al 1998; Pacheco- Cutillas et al 1999]. Pupillary responses are reduced in advanced glaucoma and an RAPD is found in markedly asymmetric glaucoma [Kaback et al 1976; Brown et al 1987; Burde 1988; Kalaboukhova et al 2007]. Pupil perimetry and standard threshold perimetry show matching defects in ischaemic and non-glaucomatous compressive optic neuropathies, but not in patients with primary open-angle glaucoma [Kardon 1992].
5.1.4.4 Idiopathic Intracranial Hypertension
Like glaucoma idiopathic intracranial hypertension involves damage to the optic nerve head. The cause of compression is at some point pressure
212 transmitted by the cerebrospinal fluid onto the optic nerve head [Miller and Newman 1998]. This is possible as the neuroretina and optic nerve are in strict anatomical terms parts of the brain composed of oligodendrocyte glial cells and not Schwann cells in their nerve sheaths [Gray et al 1995]. Further unlike the olfactory nerve which also has a similar neuronal environment, they are additionally surrounded by a meningeal sheath. In this work pupillometry was initially suggested as being a worthwhile application to compare with visual fields to assess damage in idiopathic intracranial hypertension. In view of later developments in this work of a more fundamental nature concerning non-rod non-cone photoreception, this was deferred for other projects.
5.2 Objectives
In this series of experiments the theme of PEVR is further developed to see if it is present in orbital disease causing compressive optic neuropathy. More routine measures of pupillary dysfunction such as constriction amplitude are also studied. One young subject's clinical management additionally permitted time to perform a study of the effects of dark adaptation on the related pupillary responses without her becoming fatigued (subject #3).
5.3 Materials and Methods
Seven subjects were recruited for these experiments (table 5.1). Patients with compressive and/or traumatic optic neuropathy were recruited over a period of three years. This involved collaboration with an oculoplastic/orbital subspecialist and on a practical level liaising directly with their subspecialty Fellows during this period. Subjects had unilateral compressive optic neuropathy diagnosed by history, examination and imaging by oculoplastic subspecialists. Damage was felt to be proximal to the entry of the central retinal artery into the optic nerve. All subjects were candidates for orbital surgery to decompress and/or to explore the orbit.
213 Gender: Aetiology Age M=Male F=Female
Traumatic (bone 40 M fragment), RAPD Traumatic 38 M (haematoma), RAPD Compressive 27 F (unknown, large inflammatory component), RAPD Suspected 40 M traumatic (long- standing), RAPD Compressive 84 F (tumour), RAPD Compressive 80 F (thyroid eye disease; virgin orbit), no RAPD Compressive 83 F (thyroid eye disease; previous surgery to orbit), no RAPD
Table 5.1 The seven subjects with compressive optic neuropathy (including compression). RAPD was present in all subjects other than those with thyroid eye disease.
214 The urgent/ emergency nature of certain forms of compressive and/or traumatic optic neuropathy posed the prime challenge in recruitment. For example visual assessment of one subject began while he was an inpatient following discharge from intensive care and high dependency units, with the assessment completed following discharge from the hospital. One other subject was an inpatient for the period of all the tests. All subjects underwent testing according to the general experimental paradigms detailed in chapter one.
Most subjects had clinical RAPDs, suggesting wide variation in extent of damage, shown also on visual fields where reliable. All subjects had some degree of visual field depression consistent with compressive optic neuropathy. Achromatic and chromatic stimuli were presented respectively using CRT display and three monochromatic interference filters (respectively 440 nm, 540 nm and 620 nm) as in work documented previously in chapters two and three.
One subject's clinical management permitted time for research to measure pupillary responses additionally under photopic conditions (in the general protocol subjects are dark-adapted). Standard stimuli were otherwise used but following 'lights-on'.
5.4 Achromatic Stimuli
5.4.1 Effect of Achromatic Stimuli
At equal stimulus intensities there were massive differences in constriction amplitudes and latencies within the group of diseased eyes and the degree of overlap of responses compared with the pool of healthy eyes was considerable. This variability is well-known with pupillometry and seems perhaps compounded in this subgroup of patients by the additional stimulus to autonomic arousal and hence pupillary variability of potential or recent or
215 planned acute surgery and/or trauma, as well as the relatively diverse range and severity of aetiologies [Kardon 2003].
Inter-eye comparison of diseased with healthy eyes once again successfully differentiated healthy from diseased eyes in all cases. Pupillometric constriction amplitudes were reduced even in the eye diagnosed with compression in two subjects without RAPDs (figure 5.1). Latency was, as noted in chapter one, prone to greater variability, though was still inversely related to amplitude.
5.4.2 Sensitivity of Pupillometry over Clinical RAPD
The greater sensitivity of pupillometry using P_Scan facilitates detection of compressive optic neuropathy using achromatic light in subjects, including those without clinical RAPDs. This is a finding of potential direct clinical relevance. Disease was detected using constriction amplitude in response to achromatic light in all subjects. In this section the use of PEVR will be taken further. It is found that even in the subject with the least inter-eye difference in PEVR (and constriction amplitude) the former is 20x more sensitive in terms of statistical significance in detecting dysfunction than constriction amplitude, even though even the latter was accurate in distinguishing the diseased eye in all cases owing to the sensitivity of recording; the eye concerned did not have a clinical RAPD.
5.4.3 Achromatic Light and PEVR
5.4.3.1 Comparison between Eyes
Qualitatively gross examination of recordings did not always permit pupil evasion to be observed in the diseased eyes. There was, further, no statistically significant difference in PEVR between healthy and diseased eyes using the pooled data, which is expected given the results with other diseases noted in preceding chapters. Inter-eye comparison however showed
216 that PEVR was lower in diseased eyes compared with their fellow healthy controls. This was sufficient to enable pupil evasion to be diagnosed in all eyes with optic nerve compression when compared to the healthy fellow eye. This was more convincing than even with SRNs. As before, the subjects with the greatest and least differences in inter-eye PEVR are shown graphically in figure 5.2 (p = 0.001 to 0.002, two-tailed paired t-tests).
5.4.3.2 Effect of Pupil Size
No clear relationship between PEVR and resting pupil size was found, in distinction to the results with SRNs and RDs.
5.4.3.3 Photopic versus Scotopic PEVR
The subject undergoing testing additionally following light-on was found to have a reduction in her inter-eye PEVR to levels below statistical significance (p = 0.6, two-tailed paired t-test), and which was found at the three highest stimulus intensities only. Sun and Stark have suggested pupil escape may be more apparent at higher stimulus intensities [Sun and Stark 1983]. It is plausible that this is related to increased photoreceptor recruitment.
217
2.5 g 2 • 1.s • Diseased Eye - - go- - Healthy FellowEye 0c 1 0.5 I0 -0.4 -0.1 0.2 0.6 0.9 1.2 1.4 Log Light intendty cd itn2
1.2
1 lg. 0.8 • Diseased Eye ca 0 -0.4 -0.1 02 0.6 0.9 1.2 1.4 Log Light Intensity cd )m2 3.5 3 _a- g 2.5 2 ♦ - Healthy Fellow E ye .4C _a c 1.5 • D iseased E ye 6 1 0.5 o u -0.4 -0.1 0.2 0.6 0.9 1.2 1.4 Log Light Intensity cd /m2 Figure 5.1 Range of inter-eye differences in mean pupil constriction amplitude for three subjects: (a) (top) Subject with least inter-eye difference and no clinical RAPD (p < 0.02, two-tailed paired t-test);( b) (middle) subject with a moderate difference in constriction amplitude and a clinical RAPD (p < 0.009, two-tailed paired t-test); (c) (bottom) subject with the greatest difference in inter-eye constriction amplitude and a clinical RAPD (p < 0.0002, two-tailed paired t-test). 218 600 500 400 - •••• - Healthy Fellow Eye > 300 LIJ • Diseased Eye 200 100 0 -0.04 -0.01 0.2 0.6 0.9 1.2 1.4 Log Light Intensity cd /m2 600 500 r.e 400 re - - -•- -- Healthy Fellow Eye uj> 300 13. -- • Diseased Eye 200 100 0 -0.04 -0.01 0.2 0.6 0.9 1.2 1.4 Log Light Intensity cd /m2 Figure 5.2 Mean Inter-eye PEVR in two subjects: (top) subject with greatest inter-eye PEVR difference (p = 0.001, two-tailed paired t-test); (bottom) subject with the least inter-eye PEVR difference (p = 0.002, two- tailed paired t-test). 5.5 The Effect of Chromatic Stimuli 5.5.1 Overview Unlike use of achromatic stimuli, using pooled data certain chromatic stimuli could differentiate diseased eyes from healthy ones. Latency was 219 found to vary in a less consistent manner than amplitude and tended to be inversely related to amplitude, as commented upon before. Once again comparison of PEVR between pooled healthy eyes and diseased fellow eyes with compressive optic neuropathy showed no pattern, presumed a consequence of the great variation of PEVR within groups of diseased and healthy eyes. However inter-eye differences in PEVR were significant with different chromatic stimuli. 5.5.2 Effect of Long Wavelength Light: Constriction Amplitude with Pooled Data There was a difference in mean constriction amplitude between healthy and diseased eyes (figure 5.3). The diseased eyes had lower pupil constriction amplitudes (p = 0.01, two-tailed paired t-test). However 95% confidence intervals overlapped considerably owing to the well-described phenomenon of variability in the PLR. 2 m 1.8 m 1_6 de 1.4 itu l 1.2 -.-•--- Healthy Eyes 1 Amp n Diseased Eyes 0_8 tio ic 0.6 tr s 0.4 Con 0.2 t- 0 11.0 ]] 25 11.5 11.75 12.3 12.25 12.5 LOG Photon Flux photons .1 cm` s Figure 5.3 Effect of long wavelength light on mean pupil constriction: pooled data from healthy eyes and eyes with compressive optic neuropathy. Error bars represent 95% confidence intervals (mean +/- 2 SD). Only one error bar per data point is shown for clarity. 220 5.5.3 Long Wavelength Light: Inter-Eye Comparison of Constriction Amplitude A statistically significant trend towards reduced inter-eye sensitivity to long wavelength light was found in all diseased eyes (figure 5.4). Unlike SRNs, no eyes showed a distinct threshold response over the range of stimuli used, presumably as with Macula-Off RDs the onset of the pathologic lesion is relatively recent and cellular dysfunction more limited than with SRNs which were months to years old. Unlike RDs, there was no saturation response over the stimulus range used, suggesting more intact cellular function; this probably only represents the severity of disease in these particular groups of patients. 5.5.4 Effect of Long Wavelength Light on Inter-Eye PEVR One subject with a major inflammatory component to their optic neuropathy showed a trend to a difference in PEVR between eyes, but the difference was not statistically significant (p = 0.12, two-tailed paired t-test). These are distinct to the findings with vitreo-retinal disease. It is also plausible the subject with the inflammatory compressive aetiology may have had some subclinical retinal involvement in keeping with many inflammatory diseases. 5.5.5 Effect of Medium Wavelength Light: Constriction Amplitude with Pooled Data Diseased eyes had a consistently reduced amplitude of pupil constriction (p = 0.03, two-tailed paired t-test). However there was considerable overlap of 95% confidence intervals (figure 5.5). 221 0.5 0.45 mm 0.4 de 0.35 litu 0.3 • - -•- - Healthy Fellow Eye Amp 0.25 *-- Diseased Eye 0.2 tion ic 0.15 tr s 0.1 Con 0.05 0 11.0 11 25 11.5 11.75 12.3 12:25 12.5 LOGPhoton Flux photons / cm2 s 1.6 6 1.4 • - - - Healthy Fellow Eye • Diseased Eye 0.6 I.; 0.4 8 0.2 0 11.0 11 25 .11.5 11.75 12.3 12.25 12.5 LOGPhoton Flux photons / cm2 s Figure 5.4 Effect of long wavelength light on pupil constriction amplitude in subjects with compressive optic neuropathy. Inter-eye constriction amplitudes are shown for (a) (top) subject with the least inter-eye difference (p = 0.002, two-tailed t-test) and (b) (bottom) subject with the greatest inter- eye difference (p = 0.0007, two-tailed t-test). 222 8 •E 6 a 4 • - Healthy Eyes ..tt I 0-- Diseased Eyes. 0 t 0 »tr, 0 -2 -4 11.0 11 25 11.5 11_75 12.J 1'2.25 12.5 LOGPhoton Flux photons / cm2 Figure 5.5 Mean pupil constriction amplitude as a function of stimulus intensity with a 540 nm light stimulus (pooled data). Error bars represent 95% confidence intervals (mean +/- 2 SD). For clarity only one error bar is shown per data point. 5.5.6 Medium Wavelength Light: Inter-Eye Comparison of Constriction Amplitude An overall trend to reduced inter-eye sensitivity to 540 nm light was found in eyes with compressive optic neuropathy but (figure 5.6). As with the effect of long wavelength light on compressive optic neuropathy, unlike SRNs no eyes showed a distinct threshold response over the range of stimuli used, though saturation was reached sometimes. 223 0_9 m 0.8 m 0.7 de p 0.6 litu 0.5 - - - - Healthy Fellow Eye Amp 0.4 •-- Diseased Eye tion 0.3 ic tr 0.2 ns Co 0.1 • 0 • 11.0 11 25 11_5 11.75 12J 12.25 12.5 LOG Photon Flux photons / cm` s 5 E 4.5 E 4 c 3.5 Z. 3 • - - - - Healthy Fellow Eye < 2.5 • Diseased Eye 0 2 rt" It 1.5 .. e .7 8 0.5 • 0 11.0 H 2 5 11_5 11.75 1.2.J 12.25 12.5 LOGPhoton Flux photons / cm2 ; s Figure 5.6 Inter-eye pupil constriction amplitudes as a function of stimulus intensity with a 540 nm light stimulus, comparing healthy eyes with diseased fellow eyes with compressive optic neuropathy: (a) (top) shows the responses for subject with the least difference (p = 0.07, two-tailed paired t- test]; (b) (bottom) subject with the greatest difference in inter-eye constriction amplitude (p = 0.04, two-tailed paired t-test) 224 5.5.7 Effect of Medium Wavelength Light on Inter-Eye PEVR One subject with a major inflammatory component to the neuropathy showed a trend towards greater PEVR in the healthy eye however this was again not statistically significant (p = 0.57, two-tailed paired t-test). Otherwise there was no pattern to PEVR between fellow eyes. 5.5.8 Effect of Short Wavelength Light: Constriction Amplitude with Pooled Data Comparison of healthy with diseased eyes showed a tendency towards greater pupillary constriction in the healthy eyes. However this was not statistically significant (p = 0.06, two-tailed paired t-test), though it is reasonable to hypothesise that increasing the number of subjects might result in a statistically significant finding (figure 5.7). 4_5 •E 4 41,) 3.5 .12 3 E 2.5 - - - Healthy Eyes c 2 Diseased Eyes 0.5 0 11.0 11 25 11.5 11.75 12..1 1•2.25 12_5 LOG Photon Flux photons cm` s Figure 5.7 Mean pupil constriction in eyes with compressive optic neuropathy as a function of stimulus intensity using pooled data and a short wavelength light source (440 nm). Error bars represent mean+/- 2 SD. For clarity only one error bar is shown per data point. 225 5.5.9 Short Wavelength Light: Inter-Eye Comparison of Constriction Amplitude A trend to reduced sensitivity in all the diseased eyes were found, but this was only statistically significant in one subject (p = 0.007, two-tailed paired t-test) and not in the others (p = 0.05 to 0.08, two-tailed paired t-tests) (figure 5.8). 1 0.9 mm 0.8 de 0.7 litu 0.6 • Diseased Eye Amp 0.5 n - - - - Healthy Fellow Eye 0.4 tio ic 0.3 tr s 0.2 Con 0.1 0 11.0 11:5 11.5 11.75 12.) 12.25 12.5 LOG Photon Flux photons / cm2 / s 3.5 3 • mm de 2.5 itu l 2 - - -*- - - Fellow Healthy Eye Amp 1.5 • Diseased Eye tion ic 1 tr ns 0.5 Co 0 11.0 11 25 11.5 11.75 12.) 12.25 12_5 LOG Photon Flux photons ! cm2 ! s Figure 5.8 Inter-eye pupil constriction amplitudes as a function of stimulus intensity for a short wavelength stimulus: (a) (top) subject with the least inter-eye difference (p = 0.08, two-tailed paired t-test; (b) (bottom) subject #3 with the largest inter-eye difference (p = 0.007, two-tailed paired t-test). 226 5.5.10 Effect of Short Wavelength Light on Inter-Eye PEVR There was no relationship in inter-eye PEVR using short wavelength light. 5.5.11 Comparing the Effect of Chromatic Stimuli using Pooled Data The constriction amplitude was highest in pooled normal eyes for the 540 nm stimulus relative to the 440 nm and 620 nm stimuli, which is consistent with the known pupil spectral sensitivity which peaks at 555nm in the light, and 507 nm in the dark [Loewenfeld and Lowenstein 1999]. The statistical significance of detecting optic nerve involvement in subjects being assessed for orbital decompression is greatest with long wavelength, then medium wavelength then short wavelength light (figure 5.9). These discrepancies are similar to those often observed in compressive optic neuropathy, where red- green loss is found in testing colour vision [Miller and Newman 1998]. Unlike retinal disease where colour vision loss is more stereotyped owing to the concentration of cones in the macula, the results may be due therefore in part to the path of macular fibres through the optic nerve, which is from temporal to nasal [Gray et al 1995]. Unlike the retinal pathologies that were studied, there was no difference in inter-eye PEVR for different chromatic stimuli. The preservation of inter- eye PEVR found previously with long wavelength light was not found, but there is not enough evidence to say differences in PEVR between eyes represent purely a retinal phenomenon since it is not known to what extent colour channels, though clearly damaged, were severely affected in compressive optic neuropathy, unlike the subjects with SRNs and Macula- Off RDs where macula function was very severely affected. Further, inter- eye PEVR (though not pooled results) was reduced in eyes with compressive optic neuropathy using achromatic stimuli. 227 0.009 - p o - 0 . 0.) 0 0 )4 (1 - ll 400 450 600 550 600 660 700 Wavelength nm Figure 5.9 The statistical significance of detecting optic nerve compression in subjects being assessed for orbital decompression surgery (pooled data for constriction amplitudes — PEVR is not shown). This was greatest with long wavelength, then medium then short wavelength light (two-tailed paired t-test). The y-axis is the p value; the area of the circles corresponds to the magnitude of the p value; stimulus wavelength is plotted on the x-axis. 5.6 Light and Dark Adaptation 5.6.1 Purpose of Investigation With progressive dark adaptation the pupil response starts to reflect a greater contribution from summation of action potentials in ganglion cells 228 arising from phototransduction by rods. The work thus far had tested subjects after dark adaptation. It suggested relative preservation of response to short wavelength light in these patients. Do pupil responses under photopic conditions exacerbate these differences? 5.6.2 Methodology One subject, subject #3 was a 27 year old Afro-Caribbean female, an emergency admission with an inflammatory orbital lesion who had undergone pupillometry as already described. She was nevertheless willing to undergo further tests at the Western Eye Hospital where she had originally been seen. This was on the day after discharge from St Mary's Hospital where she had been admitted. The subject was tired but was willing to be tested. Fortunately the testing protocol was (comparatively) briefer than the other stimulus paradigms and perhaps for this reason as well as her young age there were no fatigue waves. The subject underwent testing using trains of achromatic light (CRT tube) and the three monochromatic light filters (440 nm, 540 nm, 620 nm), each with 20 minutes of dark adaptation, as described in the general experimental methodology. However, following completion of the tests using dark adaptation she was exposed to lights-on' in the laboratory and the tests repeated. 5.6.3 Intra-Eye Effect of Achromatic Light on Light and Dark Adaptation Both sets of curves for each eye showed significant increases in pupillomotor force with dark adaptation (p = 0.004 and 0.005, two-tailed paired t-tests). 229 4.5 4 3.5 3 —A— Dark Adapted 2.5 Healthy Eye 2 Light Adapted 1.5 Healthy Eye 1 0.5 0 -1 0.6 0.7 1.2 1.5 1.7 t9 Log Light Intensity cd /m2 4 E E 3.5 -c 3 2.5 —it—Dark Adapted Diseased Eye 2 —46—Light Adapted 2 1.5 Diseased Eye 0 -1 0.6 0.7 1.2 1.5 1.7 1.9 Log Light Intensity cd Im2 Figure 5.10 Effect of light and dark adaptation on (a) (above) the subject's healthy eye (p=0.004, two-tailed paired t-test); (b) (bottom) the diseased eye (p = 0.005, two-tailed paired t-test). 5.6.4 Inter-Eye Comparison of Achromatic Stimuli after Light and Dark Adaptation Pupil constriction was reduced in the diseased dark adapted eye compared with the healthy fellow eye (p = 0.03, two-tailed paired t-test), and also after light adaptation (p = 0.0001, two-tailed paired t-test) (figure 5.11). 230 • Light ..1-daptecl Healthy Feftow E' ArlApte•cl Diseased Eye 0 1_) . Log Photon Flux p h oton 4_5 4 - 3 5 3 a 25 °Dark Adapted Healthy Eye DarkAdapted Diseased Eye 2 1.5 0 05 j 11 0 12.5 Log Photon Flux photonsktn2is Figure 5.11 Inter-eye effects of (a) (top) light adaptation and (b) (bottom) dark adaptation, comparing the effect of achromatic light on the diseased and healthy fellow eyes. 231 5.6.5 Intra-Eye Effect of Chromatic Stimuli after Light and Dark Adaptation The healthy eye showed a large rise in pupil constriction amplitude after dark adaptation with all chromatic stimuli. This also occurred in the diseased eye but not with the short wavelength stimulus (p = 0.08, two- tailed paired t-test). 4 E E 3.5 • 3 11. 2.5 - Dark Adapted Healthy Eye cE 2 —k— Light Adapted 2 1_5 Healthy Eye T.: 1 tn g 0.5 0 11.0 1125 11_5 11.75 12.J 12.25 12.5 LOG Photon Flux photons / cm2 / s 2.5 E E • 2 - Light Adapted Diseased Eye --Ac- Dark Adapted 0 Diseased Eye • 0.5 0 11.0 1125 11.5 11.75 12.J 1225 U.S LOGPhoton Flux photons / cm2 s Figure 5.12 Effect of light and dark adaptation and a short wavelength stimulus on (a) (above) the subject's healthy eye (p = 0.03, two-tailed paired t-test); (b) (bottom) the diseased eye (p = 0.08, two-tailed paired t- test). 232 5.6.6 Inter-Eye Comparison of Chromatic Stimuli after Light and Dark Adaptation Chromatic stimuli were associated with a reduction in inter-eye responses in the dark-adapted diseased eye, which has previously been noted when testing all 7 subjects, while in this subject this had also been with the short wavelength stimulus. While this was again observed here with the dark adapted eye, this reduction in sensitivity was lost in photopic conditions when comparing with the opposite eye but the effect was found only using the blue light stimulus (figure 5.13). 3 - 3.5 Light Adapted Healthy Eye in Light Adapted Diseased Eye 0-5 0 .0 '2- 5 Loci Photon F 1 Li >c arils/4c e 112/s 4 - 3._S - in Dark Adapted Healthy Eye in Dark Adapted Diseased Eye 0.5 O 11.0 1 2 - 5 Loy Priotorl FIcix usl-lcbtorvs/crroz/s Figure 5.13 Inter-eye effects on (a) (top) light adaptation (p = 0.3, two- tailed paired t-test) and (b) (bottom) dark adaptation (p = 0.009, two-tailed 233 paired t-test), comparing the effect of short wavelength light on the diseased and healthy fellow eyes. 5.6.7 Implications of Work with Light and Dark Adaptation The preserved sensitivity of the pupil responses to short wavelength light under photopic conditions suggests, given that this did not occur with the achromatic stimulus, relative preservation of a group of pathways active in light at short wavelengths. Yet the proportion of S-cones is too small to account for these [Hendrickson 1994; Lagreze and Kardon 1998]. As previously described photoreceptive ganglion cells with a peak spectral sensitivity in the blue light region of the visible spectrum have been described and are active in photopic conditions in certain mammals, though hitherto this thesis had not been directly described in humans. 234 Chapter 6 Refractive Disorders and Neuronal Plasticity in Orbital Trauma, Compressive Orbital Lesions and Scarring with Proliferative Vitreoretinopathy: Use of Alternate Pupillometric Modalities 6.1 Introduction In this section additional responses to light and pupil perimetry are considered within the context of this thesis; these include the pupil's responses to gratings stimuli, and the responses to photostress. 6.2 Pupil Perimetry 6.2.1 Technical Factors in Pupil Perimetry The pupillomotor threshold is higher than the visual, so higher intensity stimuli are generally used, with a minimum size equivalent to a Goldmann size IV or at least a 16 mm2 stimulus [Kardon 1995; Bergamin 0, personal correspondence]. Reasonably strong contractions can however be elicited at several points in the horizontal meridian using a 0.25 mm x 0.25 mm square target (0.0625 mm2 ), the shape of the stimulus perhaps representing the more limited technology of the 1970s [Kani et al 1978; Hong et al 2001]. In parallel with visual responses, sensitivity is greatest at fixation, and reduces gradually the further from fixation the stimulus is presented. Cones are less sensitive to perimetric stimuli than rods, but in the fovea, owing to the relative cone density, powerful pupillary responses are nevertheless elicited [Kani et al 1978]. The physiological blind spot is readily mapped out with pupil perimetry. In normal subjects, the temporal field shows greater pupillary contraction and shorter latency than the nasal [Kardon 1995]. This is as the PLR is an index of the amount of retinal area stimulated, and the temporal retina is larger. 235 6.2.2 Strategies in Pupil Perimetry Most work using pupil perimetry is geared towards developing strategies to reduce variability. It has been suggested that dark adaptation might reduce variability in pupil responses, at least in part by increasing the amplitude of constriction [Bergamin 0, personal correspondence]. 6.2.3 Relevance to Current Work and Objectives Unlike in animals a putative human subject without rods and cones might additionally also have cataracts and related refractive problems, such as index myopia and increased chromatic aberration, and also problems like scatter. Not only was study of such refractive disorders on the pupil part of the breadth of this work, cataract is also a common age-related condition also found in younger people with photoreceptor dystrophies, which might provide a model human disease for absence of rods and cones and hence facilitate further detailed study within this thesis. Additionally, the results for SRNs and RDs which presented in chapters three and four both show a similar chromatic cut-off in the inter-eye PEVR difference with lower wavelengths. Likewise the work on compressive orbital disease and the effect of blue light in chapter five suggested a preservation of a photopic pathway sensitive to blue light yet unlikely to be cones. The objective of the following series of experiments was to determine to what extent refractive defects associated with cataract affected the pupillary response to blue light stimuli in comparison with achromatic pupil perimetry. There was no access to a lens opacity meter within the available resources. Pupil perimetry could confirm the presence of back scatter, a particular problem with nuclear cataracts, as is could be evinced by an obscuration of the focal pupil responses from around a scotoma. The PLR reflects cellular function per unit retinal surface area and this is held to account for the augmentation of the full-field PLR in subjects with cataracts [Kardon 2003]. Hitherto there had been no work on the effects that cataracts 236 might have on focal stimuli — notably, while scattering of light augments the PLR in full-field stimulation, this same effect might blur and obscure the focal response. Three relevant questions were to be answered in a pilot study within the context of the principal direction the work was taking. Do cataracts distort pupil perimetric responses? What is the effect of achromatic light, and, independently, what is the effect of blue light? If there is a distortion of perimetric responses, what could be the relative effects of the different refractive components that might be responsible — scatter, chromatic aberration and index myopia? 6.2.4 Optical Effects of Cataracts Cataracts preferentially absorb and block transmission of ultraviolet and blue light to a greater degree than they do longer wavelengths. Both the refractive interfaces of cornea and the crystalline lens normally blocks transmission of ultraviolet light. Another optical effect of cataracts of non- uniform density is that they generally scatter rays of light randomly. Cataracts may sometimes thereby amplify the PLR, which is a measure of the amount of surface area of retina illuminated [Kardon 2003]. In physical optics two types of scatter are theoretically recognised — backward and forward scatter [van Norren and Vos 1974; Schwartz 2003]. Rayleigh scatter preferentially affects shorter wavelengths of light relative to longer wavelengths. The second types of scattering phenomenon is Mie scattering. Rayleigh scatter applies to small particles like small molecules. Larger suspended particles scatter light in a different way. First, the light they scatter tends to be projected forward away from the light source (forward scatter) and second, the light they scatter does not favour any particular wavelength and so is white. 6.2.5 Experiments using Light Scatter Very recent studies of the light scattering effect of cataracts have been published in an in vitro model [Ozolinsh et al 2006]. The effects of various 237 colour contrast stimuli were psychophysically measured with scatter created by fog in Clermont-Ferrand, France, where there is a great deal of fog, and in laboratory conditions using different light scattering eye occluders. Blue light scattered in fog to the greatest extent, especially if stimuli were monochromatic blue. This showed that scattering has the greatest effect on the blue component of screen luminance. Further experiments showed it did not however reduce visual acuity. It had the least effect on the perception of white-blue stimuli. With coloured stimuli presented on a white background, visual acuity in fog for blue Landolt-C optotypes was higher than for red and green optotypes on the white background. Colour Landolt-C optotype luminance was presented on a LCD screen corresponding to the RGB colour contributions in achromatic white stimuli (computer digital R, G, or B values for chromatic stimuli equal to RGB values in the achromatic white background) resulting in greatest luminance contrast for the white-blue stimuli, and hence felt to account for the better visual acuity with the white- blue stimuli. The differential effects of the various morphs of cataract on scatter have also been studied. In vivo studies were presented in 1992 [de Waard et al 1992]. Forward scatter was measured using the psychophysical direct compensation method (ring-shaped flickering light source a certain distance from a dark test field; to measure the amount of stray light variable counter-phase compensation light is presented in the test field — the flicker perception in the test field can be extinguished and measured as an index of intraocular scatter); contrast sensitivity loss was also measured (using a glare tester, Visitech MCT 8000) as it can be due to forward scatter [van den berg 1986; de Waard et al 1992; Franssen et al 2006]. Scatter measured using these techniques was greatest with posterior subcapsular cataracts compared to nuclear and cortical cataracts, and might tend to attenuate the pupil response. Forward scatter exhibited considerable variation among individuals, especially for cortical and posterior subcapsular cataracts, and thus estimates of forward scatter could not be mathematically derived from backscatter (or the slit-lamp image). Backscatter might in fact augment pupil responses. Backscatter (determined directly with 238 the Lens Opacity Meter of Interzeag) was largest for nuclear cataracts, intermediate for posterior subcapsular lens opacities, and almost zero for cortical cataracts. 6.2.6 Chromatic Aberration Being an optical system the eye exhibits chromatic aberration, known as "colour fringing" with non-biologic camera systems such as in photography. It is a consequence of dispersion: the refractive index of different wavelengths of light is wavelength-dependent, as different colours of light propagate at different speeds in a medium [Schwartz 2003]. Classically Newton showed this occurs with a prism splitting an incident beam of white light into the colours of the rainbow. In an optical system, here the cornea and lens, refraction of light is to a focus that is different for the various wavelengths of light (the focal length / focal planes for different wavelengths are different). Chromatic aberrations are those departures from perfect imaging which are due to dispersion. There are two types of chromatic aberration, and in practice they occur concurrently. Longitudinal chromatic aberration is the inability of a lens to bring different wavelengths to bear on the same focal plane and hence axially displaced images occur with dispersion in the visible spectrum greatest for blue light (figure 6.1). Obliquely incident light creates transverse chromatic aberration or lateral colour. Foci are displaced sideways. Colours are in focus in the same plane, but the image magnification depends on the wavelength so images of the same size are magnified differently depending on colour (figure 6.2). Transverse (lateral) chromatic aberration implies focal length depends on wavelength, whereas transverse chromatic aberration in a complex lens does not strictly require a variable focal length. This seems counterintuitive, but in a lens corrected for lateral chromatic aberration the principal planes do not need to coincide for all wavelengths. Since the focal length is determined by the distance from the rear principal plane to the image plane, the focal length may depend on the wavelength even when all images are in the same plane. 239 Figure 6.1 Longitudinal chromatic aberration Figure 6.2 Transverse chromatic aberration. Figures 6.1 and 6.2 distinguish the longitudinal and lateral components, but as these coexist in practice a "polychromatic image" occurs, such as that filled by a person in the image space. The latter is a volume comprised of a multitude of images — indeed a continuum of monochromatic images of various positions and sizes due to the effects of chromatic aberrations. 6.2.7 Index Myopia While dispersion relates to wavelength, the optical density of the refracting medium is an important factor. Certain cataracts, most notably nuclear sclerosis, can cause such a profound refractive effect owing to their refractive density as to cause up to a few dioptres of induced myopia by increased refraction of light. Concurrently with this type of cataract there is likely to be increased backward scatter [de Waard et al 1992]. 240 6.2.8 Methods Subjects with unilateral visual field defects were recruited. A very common visual field defect was studied (to facilitate recruitment) - large horizontal glaucomatous arcuate scotomas in the superotemporal field extending horizontally from at least 15 degrees to at least 24 degrees and between 10 and 20 degrees in height vertically. Further, that the retinal lesion in glaucoma is in ganglion cells is not irrelevant in the context of putative human retinal ganglion cell photoreception. Despite the high prevalence of glaucoma, unilateral examples of this lesion were not very common. Cataracts were classified using a standard validated measurement system, Lens Opacities Classification System (LOCSIII) [Appendix II] applied by the principal observer. Axial length measurements were recorded from ultrasound B-scan measurements. The following subjects with the aforementioned unilateral visual field defects on the same side of the cataract (where present) were studied: 1. Three emmetropes (mean axial length 23.0 mm, range 22.75-23.50 mm), no significant cataracts, mean age 65, all females. 2. A subject with axial myopia (axial length 26.5 mm), no significant cataract, a 64-year old male. 3. An emmetropic subject with cataract (dense nuclear cataract in both eyes, grade N4C4 in LOCSIII, axial length 24.2 mm), a 76 year-old male. 4. A myopic (non-axial) subject with cataract (dense nuclear cataract in both eyes, grade N4C4 in LOCSIII, axial length 24.5 mm, index myopia 3.0 dioptres (D)), an 83 year-old male. Subjects attended at least twice for further tests that have been previously documented under the general methods, which differed only in that visual fields were obtained in the form of two stable, consecutive and reliable Humphrey 24-2 fields that had been performed routinely in the course of their clinical management. 241 6.2.9 Stimulus and Recording Paradigm Subjects were dark-adapted for 20 minutes and a wide-aperture refractive correction used where necessary. Subjects with the opposite eye occluded were instructed to look at a fixation target. A very powerful quartz halogen achromatic light source of intensity 1016 photons/cm2/s was used as the stimulus 20° away from fixation to coincide with the scotoma. From Pythagoras' theorem this, to be equivalent to a Goldmann size IV (16mm2) target at the same target distance as with Goldmann perimetry, was calculated as 2.3 mm (radius of a circle = square root of 16/7c as area = mr2). This 2.3 mm radius source (light pipe variable shutter) was presented 20° away from fixation. The distance was calculated using trigonometry (same as distance from target using Goldmann perimetry = 30 cm, and for 20° away from fixation this is as adjacent (x) = opposite/tangent 0) = 30 cm. Beginning at a 20° radial locus from fixation, 40° along a circle from the scotoma (to enable comparison of responses between scotomatous and non- scotomatous loci) the stimulus would be turned on, and this would progress at 20° radial loci (from fixation) and 40° apart from each other. This was performed five times at each location, with a stimulus duration of 1 s separated by 5 s intervals. By working 360° round the field of vision, the 20° radial field coinciding with a scotoma had been tested using 40° increments. This was then repeated in the opposite healthy eye. This same paradigm was then repeated after dark adaptation for 20 minutes but with blue light of the same radiometric stimulus intensity using a 481nm interference filter of narrow band width (1016 log photons/cm2/s). Responses were recorded using the standard measurement paradigm (P_Scan 100 as documented in chapter two). Mean pupil constriction amplitude was calculated for each of the perimetric locations (as amplitude generally shows less variability than latency and it is generally preferred). 242 6.2.10. Results 6.2.10.1 Intra-Eye Comparisons: Emmetropic Eyes with no Cataract In the three emmetropic eyes without cataract responses to the achromatic stimulus were greater than to the blue stimulus which is expected (figure 6.3). Both achromatic and blue stimuli exhibited a very marked reduction in stimulus intensity in the region of the scotoma, more marked with achromatic light. 2.5 _ —6— Blue Stimulus - . --a— Achromatic Stimulus - 1 / 4C BC 120 160 200 240 280 320 C rcurnferenta I Location 1 degree Figure 6.3 Mean pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in emmetropic eyes with an arcuate scotoma but no cataract. Recordings begin 40 ° outside the scotoma. Error bars represent 95% confidence intervals (mean +/- 2 SD). 6.2.10.2 Intra-Eye Comparisons: Effect of Axial Myopia In the eye with axial myopia and no cataract there was a similar pattern of greater amplitude of responses to the achromatic stimulus as well as a more 243 marked reduction in responses in the region of the scotoma to both achromatic and blue stimuli, again slightly more marked with the achromatic stimulus (figure 6.4). 2.5 2 1b---._ - - --s— Blue Stimulus --•—Actutimatic Stimulus C C SC 120 160 200 240 280 320 Circumferential Location i degree Figure 6.4 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an eye with an arcuate scotoma, axial myopia but no cataract. Recordings begin 40° outside the scotoma. Error bars represent 95% confidence intervals (mean +/- 2 SD). The sensitivity of pupil perimetry does not therefore seem to be affected very much by axial myopia in the absence of cataract, with either achromatic or blue light perimetry 6.2.10.3 Intra-Eye Comparisons: Effect of Cataract in Emmetropia - Scatter and Chromatic Aberrations In the emmetropic eye with cataract there were wide fluctuations of pupil responses. It was barely possible to map out the scotoma with the blue stimulus (figure 6.5) and not at all possible with the achromatic stimulus. 244 Nuclear cataract therefore seems to reduce the sensitivity of detecting a perimetric stimulus whether of blue or white light. However, the result was still significant with the blue stimulus, there being more fluctuations in pupillary responses with the achromatic stimulus. The white light response is also reduced to a greater degree, being otherwise stronger than for blue light when there is no cataract present. This may be due to forward scatter. But it should affect blue light also to a similar degree in the plots. This is as Rayleigh scatter does not occur with cataracts, and Mie scatter affects all wavelengths equally. Chromatic aberration must therefore be considered. A monochromatic light stimulus will be associated with less chromatic aberration, and may account for the mild but statistically significantly better resolution obtained with blue perimetry. It would presumably also lead to better resolution with any monochromatic source, not only blue. An alternative explanation is that as discussed blue-yellow colour loss is documented in glaucoma. However this would not explain the poorer resolution compared to white light found in earlier on in the eyes without cataract, making reduced chromatic aberration with a monochromatic light a more likely explanation. Another explanation is that the results may be explained by relatively greater absorption by cataracts of blue light but in the subject with index myopia where one would have expected this phenomenon to be enhanced it was not. 245 0.4 0.3 g• 0.25 -4— Blue Stimulus Achromatic Stimulus 3 0.15 CL O. 0.05 C 40 80 120 160 200 240 280 320 Orcurriferentia I Location P degree. Figure 6.5 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an emmetropic eye with an arcuate scotoma and dense nuclear sclerosis. Recordings begin 40° outside the scotoma, which is barely seen here with blue pupil perimetry between 40° and 80°, and is not at all resolvable using achromatic pupil perimetry). Error bars represent 95% confidence intervals (mean +/- 2 SD). 6.2.10.4 Effects of Index Myopia in Nuclear Sclerotic Cataract Comparing responses in the eye with dense nuclear sclerosis and index myopia, there was a marked blunting of pupillary responses to both achromatic and blue stimuli (figure 6.6). 246 C . C -4- Blue Stimulus .2 --s—Achiumatic Stimulus 0 C , 2 4C s.c 12C 1ir2C 200 240 2130 320 Circumferential Location! degree Figure 6.6 Pupil constriction amplitude as a function of circumferential stimulus location for achromatic and blue stimuli of equal intensity in an eye with dense nuclear sclerotic cataract, an arcuate scotoma and index myopia. Recordings begin 40° outside the scotoma, which though not resolvable with pupil perimetry with visual perimetry lied in the vicinity of 40°. Error bars represent 95% confidence intervals (mean +1- 2 SD). The subject had 3 D of index myopia and an optically exceptionally dense cataract. However earlier myopia as an independent factor was not found to affect perimetric resolution of a scotoma with pupillometry. What then could explain the blunting of responses to both white and blue light? The blunting of responses to both achromatic and blue stimuli suggests random Mie scatter. It is likely that the scatter is sufficient in a cataract with index myopia to obscure all differences between the two stimuli, including the better resolution that a monochromatic stimulus enabled due to reducing the effect of chromatic aberration. This is not an unreasonable explanation given that scatter is closely related to the density of the refractive medium. 247 6.2.11. Inter-Eye Comparisons As expected there was a reduction in pupil constriction amplitude from areas of diseased retina compared to corresponding areas of opposite eye healthy retina regardless of whether cataract was present or not. More interesting to note is that in the two eyes with cataract there was no significant difference in pupil constriction amplitude between the non- scotomatous areas of retina in eyes with cataracts and corresponding retinal areas of healthy fellow eyes using achromatic light. In one subject the responses tended to be very slightly lower in the eye with cataract, and in the other in the eye without cataract (p = 0.42 and p = 0.14, two-tailed paired t-tests). The absorption of light by nuclear cataract seems therefore not to attenuate the PLR presumably as backward scattering compensates for loss of light energy by absorption. Equally, the phenomenon of augmentation of the PLR by cataract as postulated by some workers, which they presume due to scatter, was not observed with achromatic light despite being described in the literature anecdotally [Kardon 2003]. However the results of this pilot study cannot discount the latter purported phenomenon from the literature for two reasons. First, only two of the six subjects had cataract. Second, this study uses pupil perimetry where the response and overall effect of scatter is reduced compared to full-field stimulation. It is possible the phenomenon could be found using perimetry if enough subjects were tested. The difference in pupil constriction amplitude using blue light between the non-scotomatous areas of retina in eyes with cataracts and healthy fellow eyes was not statistically significant (p = 0.6 and p = 0.5, two-tailed paired t-tests for the two subjects ). This may be as the pupil constriction amplitude overall is less to blue light. However unlike with achromatic light which showed no consistency between the two subjects, in both subjects with blue light there was a tendency to increased pupillary constriction amplitude in eyes with cataract even though this fell short of statistical significance. It is 248 not possible therefore to discount a role for increased scatter with short wavelengths of light in producing this effect. 6.2.12. Implications In addition to providing new data on the effect of refractive disorders on pupil perimetry in the presence of cataract, the results suggest that a human model for non-rod non-cone photoreception should, preferably, not have very significant cataract, and certainly no advanced nuclear opacities. A monochromatic light source may offer benefits compared to achromatic stimuli. This is of relevance to groups investigating the clinical application of pupil perimetry as it suggests that chromatic pupil perimetry may offer benefits in subjects with cataract as it might reduce chromatic aberration. 6.3 Pattern as a Stimulus 6.3.1 The Pupil's Response to Pattern The pupillary response to pattern has been extensively discussed in chapter one. It is a small amplitude response and only came to be accurately recorded with the advent of modern ultra-sensitive infrared pupillometers. The response is cortically-mediated via connections to the midbrain [Barbur et al 1992]. Stimuli have been presented either as gratings (the pupil gratings response or PGR) or as checkerboards. 6.3.2 The Pupil's Response to Coloured Pattern As discussed in chapter one, this has only been studied in work at City University. Whether the pupillary responses represent a distinct channel, or summation of colour and gratings channels, is not fully established. 249 6.3.3 Applications to Disease: Weinstein's Hypotheses As there is considerable functional parallelism between the PLR and visual responses, Weinstein suggested in the 1980s that defects in the PGR might similarly parallel visual responses to [Thompson et al 1980]. In particular he hypothesised that in macula disease the PGR might be reduced at high spatial frequencies, and in optic nerve disease at low spatial frequencies. However no experimental work had hitherto been conducted to confirm or refute Weinstein's original suggestions. 6.3.4 Studies in Thyroid Eye Disease Weinstein had assumed that optic nerve disease tended to produce low frequency contrast sensitivity loss. Studies in dysthyroid eye disease have generally concurred with this. Work by Tanner, Tregear, Ripley and Vickers found loss of low and medium frequency in patients with thyroid eye disease both with and without optic neuropathy [Tanner et al 1995]. Regarding the aetiology of this phenomenon, they pointed out that it was not known with certainty why subjects with dysthyroid eye disease and no optic nerve involvement also showed the changes as well. Most papers state that low frequency contrast sensitivity loss is likely to be found in dysthyroid optic neuropathy, shown either through experiment or in statements in reviews [Arden 1978; Jindra and Zemon 1989; Mourits et al 1990; Suttorp-Schulten et al 1993; Tanner et al 1995]. Intermediate and high spatial frequency loss in thyroid eye disease has also been shown, but conflicts with the bulk of the literature [de Marco et al 2000]. 6.3.5 Pupil Acuity 6.3.5.1 Applications to Disease There was a suggestion early during planning stages of this thesis that pupil acuity might likewise provide an objective measure of visual acuity in 250 patients with proliferative retinopathy (PVR), or other conditions where vision is very low and which may be subject to further clinical interventions, in particular where there is a degree of neuronal plasticity. A comparison could be made with visual responses and hence logMAR levels of vision. At one stage of the work it was thought that PVR might be a good model to provide data on pupil acuity as advanced PVR is a very severe retinal condition where vision approaches perception of light and hand movements. However the first group of experiments using PVR (later documented in this chapter to explore Weinstein's hypotheses) showed the very small amplitudes of the pupil gratings responses elicited in very advanced PVR. This meant the responses would not be very accurate without very lengthy repeat readings, which were by then out of the time frame of the current work and the principle direction it was taking in exploring the aetiology of PEVR. Further, direct clinical application would, at most, be very limited. The basic theoretical experimental set-up that was worked out will be discussed nevertheless as an insight into the versatility of pupillometry as a way to gauge acuity. 6.3.5.2 Theoretical Possibility of Measuring Pupil Acuity due to Scarring in Proliferative Vitreoretinopathy Compared with Visual Acuity To compare pupil acuity with logMAR visual acuity it would make sense to measure in convenient 0.1 log unit steps as these are the increments on the logMAR visual acuity scale (or in a pilot, 0.3 log unit steps — this would be convenient because 0.3 corresponds to a halving of resolution i.e. 6/ 6 to 6/12, 6/12 to 6/24 and so on). Subjects with PVR however might have such poor vision as not to be able to resolve the logMAR chart. Assuming they could not read the top line, they could be moved closer in 0.1 log unit (distance) steps. In other words each step towards the chart would add 0.1 to their logMAR score. For example, a patient who could not read the 1.0 (Snellen equivalent 6/60) line but could read it when one step closer to the chart would have an acuity of 1.1 logMAR. Refining this example for letter- 251 by-letter scoring, if they were brought in one step in distance and they could only read two letters of the top line, then their score would be 1.0 + 0.02 + 0.02 = 1.04 logMAR. The next question asked was what would be the step sizes towards the chart? For a standard (4m distance) logMAR chart they would be: 3.2, 2.5, 2.0, 1.6, 1.25 and 1.0 metres for a chart. To summarise with an example, a subject who cannot read the top line until they are 1.25 metres from the chart would have an acuity 1.0 + 0.5 = 1.5 logMAR. With regard to pupillometry, the viewing distance and spatial frequencies required would be inputted into the P_Scan 100 programme and the correct number of cycles after computation would be displayed automatically on the stimulus presentation monitor. A distance of about a metre was chosen to experiment with, so logMAR 1.0, 1.3, 1.6 = Snellen 6/60, 6/120, 6/240 = cy/deg 3, 1.5, 0.75. To calculate 0.1 log unit equivalents i.e. 1.1, 1.2, 1.4, 1.5 etc is slightly more complex, and would involve multiplying by 1/1.26 e.g. 1.1 logMAR = 3*1/1.26 = 2.38 cy/deg. 6.3.6 Testing Weinstein's Hypotheses 6.3.6.1 Experimental Methodology Subjects were selected with advanced retinal disease (in which a component was clinically felt to be PVR) and severe traumatic orbital disease causing optic neuropathy. These subjects were recruited by liaising with the relevant oculoplastic and vitreo-retinal Fellows working for consultants in these subspecialties at the Western Eye Hospital, as previously described. Both conditions show an element of reversibility through neuronal recovery and plasticity at the level of retina and optic nerve and are hence consistent with Weinstein's hypotheses [Cardillo et al 1997; Miller and Newman 1998; Charteris et al 2002]. As a direct consequence of the results in these two groups of subjects the theme of these experiments was extended to explore if responses were cortically mediated in one additional subject with amblyopia. 252 6.3.6.2 Proliferative Vitreoretinopathy — Pathology Proliferative vitreoretinopathy (PVR) is a non-neoplastic intraocular growth resulting from an excessive inflammatory reaction after a retinal break. PVR is an outcome of retinal detachment, usually when there has been a significant delay to surgery. Further surgery may be required to re-attach the retina. In PVR scar tissue forms in sheets on the retina which contract. This marked contraction pulls the retina toward the centre of the eye, detaching and distorting the retina severely. PVR can occur both posteriorly (as shown in Fig 6.7) and anteriorly with folding of the retina both anteriorly and circumferentially. Figure 6.7 Colour fundus photograph of an eye with PVR On a microscopic level, PVR consists of dense glial membranes, displaced RPE cells, with complex layered architectures and focal glial 'pegs' extending to the inner retina from the membranes. These represent a process of neuronal plasticity as first, second and third-order neurones attempt to re- establish synaptic connections [Charteris et al 2002]. It is a metaplastic process as RPE and glial cells start to cause fibrosis. It occurs in 5-10% of all rhegmatogenous retinal detachments [The Retina Society Terminology Committee 1983], and in case of penetrating ocular trauma and retinal 253 translocation surgery, the incidence of PVR increases to 25% and 18-23% respectively [Cardillo et al 1997]. While the causes of PVR have not been elucidated, certain risk factors for the development of this condition have been identified by multivariate analysis. These risk factors include aphakia, long duration of retinal detachment, vitreous haemorrhage, choroidal detachment, presence of preoperative PVR, larger size of detachment, use of silicone oil, high vitreous protein level, and the use of extensive cryotherapy during retinal detachment repair [Bonnet and Guenoun 1995; Nagasaki et al 1998; Kon et al 2000]. Visual outcome in PVR has been historically poorer than non-PVR cases after repair. Many studies such as the Silicone Oil Study group have reported 11-25% of patients overall will have 20/100 or better vision [Abrams et al 1997; Lewis et al 1991]. It is hypothesized that poor visual outcome after PVR repair may be due to irreversible photoreceptor loss or misalignment, epiretinal and subretinal pathology such as RPE scarring and multi-layering, toxicity to intraocular agents such as silicone oil, or even optic neuropathy [Cardillo et al 1997]. 6.3.6.3 Recruitment of Subjects with PVR Various grades of PVR are clinically described. Two male subjects aged 47 and 53 with previous rhegmatogenous retinal detachments and clinically suspected severe PVR Grade C in 12-clock hours using The Retina Society classification (table 6.1). Vision was counting fingers in the affected eye. The opposite eye was healthy. There was no other medical history. 254 Grade Findings A RPE cell clumps in vitreous B Partial thickness retinal wrinkling, rolled edge of breaks C Full-thickness fixed retinal folds Cl Up to one quadrant of fixed folds C2 Up to two quadrants of fixed folds C3 Up to three quadrants of fixed folds C 'in' x clock hours `D' Redundant term (originally used to describe the morphologies of more than nine clock hours of fixed folds) Table 6.1 Classification of PVR (after The Retina Society). Navigable vision is much more likely in Grade C3 than in cases that are more severe. The Silicone Oil Study grading is less popular. 6.3.6.4 Traumatic Optic Neuropathy 6.3.6.4.1 Gross Pathogenesis of Traumatic Optic Neuropathy Classically these are separated into direct optic nerve injury via penetrating trauma e.g. knife, surgical instrument, or indirect caused by forces applied at a distance such as a severe blow to the head producing shearing forces of deceleration[Miller and Newman 1999]. Lesions can occur to the disc (avulsion), anterior optic nerve and posterior optic nerve, the latter two divided on the basis of where the central retinal blood vessels enter or leave the nerve — approximately lcm behind the disc. Intra-orbital, intra- canalicular and intracranial injuries may occur. Both direct and indirect optic nerve injury results in mechanical and also ischaemic damage. 255 6.3.6.4.2 Primary Mechanisms of Optic Nerve Injury These occur at the time of injury, such as laceration of the nerve or shearing from deceleration. Studies using post-mortem autopsies show extensive haemorrhage associated with the optic nerve [Miller and Newman 1999]. Trauma induces a focal axon deficit rather than a diffuse brain shearing; the resulting impaired axonal transport produces functional separation of the optic nerve into proximal and distal segments usually within 6 to 24 hours of injury. The distal segment far from the brain undergoes Wallerian degeneration while the proximal segment in contact with the brain becomes oedematous and retracts. Apoptosis or programmed cell death may occur in injured neurones. 6.3.6.4.3 Secondary Mechanisms of Optic Nerve Injury Subsequent cell death can occur even adjacent to the primary region of cell necrosis. The most important factor is no longer mechanical but ischaemic. Partial perfusion followed by reperfusion generates oxygen free radicals causing reperfusion damage via lipid peroxidation of axon cell membranes rich in polyunsaturated lipids. Bradykinin and kallidin are activated after injury. Bradykinin activates release of arachidonic acid from neurones. The prostaglandins produced may cause a compartment syndrome and hence ischaemia. Inflammation occurs characterised by polymorphonuclear leukocytes in the first one to two days, then replaced by macrophages, the latter which correlates with delayed demyelination post-injury. Macrophages inhibit the astroglial response, which can nevertheless cause scarring in the long-term. 256 6.3.6.5 The Pupil in Traumatic Optic Neuropathies An RAPD is very common, and some authorities believe traumatic optic neuropathy cannot be diagnosed without it [Miller and Newman 1999]. 6.3.6.6 Prognosis of Traumatic Optic Neuropathy Long-term visual prognosis is highly variable owing to neuronal plasticity [Miller and Newman 1999]. Levin et al in The International Optic Nerve Trauma Study (IONTS) failed to provide sufficient evidence to conclude that either corticosteroids or optic canal decompression surgery should be considered the standard of care for patients with traumatic optic neuropathy [Levin et al 1999]. The study authors recommended that it is "... therefore clinically reasonable to decide to treat or not treat on an individual patient basis." A separate work in the form of a meta-analysis by some of the study members, including Levin and Joseph, found that steroids and extracranial orbital decompression was of benefit in selected patients categorised according to a grading system they produced (table 6.2) [Cook et al 1996]. Grade Findings 1 Visual acuity greater than 20/200 in the affected eye and without a posterior orbit fracture 2 Visual acuity between 20/200 and light perception 3 No light perception or with an un-displaced posterior orbital fracture and remaining vision 4 No light perception and a displaced posterior orbital fracture Table 6.2 Orbital fracture and visual acuity used to grade traumatic optic neuropathy [Cook et al 1996]. The system is used to decide which patients might benefit from steroids or extracranial decompression using an orbital approach. Intracranial decompression was not found to be useful in most cases. 257 6.3.6.7 Recruitment of Subjects with Intra-orbital Traumatic Optic Neuropathy Two subjects with unilateral orbital injuries (Grade 1 according to the classification of Cook et al) from non-penetrating trauma sustained four months and eighteen months previously were recruited. Both were otherwise fit men aged 35 and 42. CT scan confirmed traumatic atrophy of the optic nerve secondary to traumatic optic neuropathy compounded by haematoma (compression). There was no associated injury such as intra- ocular damage, fracture, muscle or motor nerve palsy. 6.3.6.8 Relevance of Studies in Amblyopia Amblyopia is a disorder of reduced visual function from a wide range of abnormal visual experiences such as those caused by strabismus, anisometropia, or visual form deprivation during the critical period of visual development [Moseley and Fielder 2001; Clarke 2006]. Thus far there is no `hard' sign for it on fundoscopy [Clarke 2006]. However, contrary to older concepts amblyopia, while predominantly associated with disordered cortical processing, has also in humans been shown to be associated with disordered processing in the retina and, in animals, the lateral geniculate nucleus [Moseley and Fielder 2001; Clarke 2006]. Amblyopia has been defined if best corrected visual acuity is 6/9 or worse, and not attributable directly to any underlying structural abnormality of the eye or the visual pathway such as anisometropia, strabismus, ametropia, or ocular media or other cause for delayed / abnormal transmission of visual information in the central nervous system (Attebo et al 1998; Moseley and Fielder 2001]. Amblyopia is associated with cortical processing deficits including to gratings that show a degree of plasticity in common with PVR and traumatic optic neuropathies [Anderson et al 1999; Weiss and Kelly 2004; Muckli et al 2006; Moseley et al 2006]. Studies of infants with delayed visual maturation suggested the PGR is also influenced by cortical pathways 258 [Cocker et al 1998]. A sine qua non of the findings in orbital and vitreoretinal disease would be to gauge to what extent the PGR was cortically-mediated — a suitable amblyope was chosen. This was a 40 year old male with unilateral strabismic amblyopia and corrected Snellen visual acuity of 6/9 in the affected left eye. He was otherwise fit and well. 6.3.7 Experiments 6.3.7.1 General Experimental Paradigms Achromatic and chromatic gratings were presented using the P_Scan 100 pupillometer described in chapter two. Subjects attended for the full range of initial tests prior to pupillometry as documented in chapter two. Subjects were dark adapted for each series of experiments and testing was conducted with the healthy eye occluded. 6.3.7.2 Achromatic Gratings and Colour Substituted Gratings Photometric calibration data for the display, on the CRT, was provided by measurement of the spectral output of each luminant phosphor gun inputted into the system, and then automatic calibration of the luminance versus applied voltage relationship for each gun, there being no difference. The experimental programmes employ standard colourimetric transformations to produce any specified luminance and chromaticity within the limits of the luminant phosphors [Wyszecki and Stiles 1982]. This was preferred to combining the PGR with a PLR as the coloured PGR has not been studied outside work by Barbur and colleagues, and the channels involved remain slightly speculative. Random spatiotemporal modulation of the immediate background field for isoluminant chromatic stimuli was used to isolate the chromatic signals by reducing small luminance contrast components or rod contrast signals [Birch et al 1992]. The stimulus was turned on in the middle of uniform rectangular background field of luminance 24 cd.m-2 and angular subtense 26° x 21°. Total absence of residual luminance contrast signal in a 259 predetermined coloured stimulus in a given individual subject is not possible because of inter-subject variation in L- and M-cone photoreceptor densities in the retina [Brainard et al 2000]. Restricting the stimulus to the central portion of the visual field may reduce this effect as L:M photoreceptor ratio increases with retinal eccentricity and may effect the photopic luminance of the eye [Hagstrom et a12000; Barbur et al 2004]. Further, as the PGR threshold is much higher than the psychophysical, residual, small luminance contrast signals are unlikely to affect the PGR [Barbur et al 2004]. The test stimuli were achromatic gratings of pre-set equal space-averaged luminance (as opposed to lower space-averaged luminance which consist of pairs both of 'dark' bars), or pre-set red-green isoluminant gratings calibrated automatically by the P_Scan 100 software and verified independently for the current monitor (assistance of Mr. K.D. Cocker) and equal photometric RGB output to the previous monitor matching the Aydin prototype (green being the brightest phosphor, followed by red and blue at 520 nm, 600 nm and 470 nm respectively, photometrically +/-2 cd/m2 mean after 40 measurements for each phosphor). These were presented within a central disc area of 12° diameter. The viewing distance for the requisite luminancy was calculated using modified P_Scan software. Achromatic gratings were presented 50 times at each of the 12 spatial frequencies beginning with the lowest and working up (3 to 14 cy/deg). They were presented against a background of equivalent space-averaged luminance. A small variable audible clue preceded each stimulus to prevent blinking, as before. This was repeated following repeat dark adaptation with the chromatic gratings, often at a subsequent session if patients were fatiguing. The relatively large number of repeat stimuli being presented is necessitated by previous work showing how small the PGR is, and work suggesting that colour responses to stimuli of this magnitude (unlike those with the interference filters) are even smaller, as are responses to coloured pattern [Barbur et al 1992; Cocker and Moseley 1992; Cocker et al 1994; Cocker and Moseley 1996; Barbur et al 2004]. 260 6.3.8 Responses to Achromatic Gratings in Orbital Trauma Mean responses for two affected eyes (each comprising 50 sets of recordings) are shown in figure 6.8 and corresponding readings from their healthy fellow eyes in figure 6.9. The PGR from the healthy eyes has some very gross resemblance to an S-shaped curve. Eyes with Orbital Trauma no 5 0 01 _ — 6 .. 4 mm 0 — _ ion t o ______„....„...... ______-- -4- ,...... ic tr N • , • C 1 il Cons I _ 6 N Pup z. 6 :r -u.0 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.8 PGR in eyes suffering from orbital trauma. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 261 Healthy Fellow Eyes to Orbits with Trauma 0.6 0.5 - - 0.4 - E E 0.3 - - _ 0.2 ...---- -40- ---°- -f------.----$-___ 7 -1------, .L. 0.1 - e c 4.---..__ — -•-__ - o 0 ---. - U - - _ 1 -0 1 = a• -0 2 - - -0.3 -0 4 Spatial Frequency 0.5 to 5.0 cycles / degree Figure 6.9 PGR in healthy fellow eyes of eyes suffering orbital trauma. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +1- 2 SD). Comparing the responses between the two groups (figure 6.10) clearly shows a greater loss of responses to low spatial frequencies (p = 0.001, two- tailed paired t-test). Pupil Constriction mm 0.25 0.2 0.15 a Healthy Fellow Eyes 0.1 INEyes with Orbital Trauma 0.05 0 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.10 PGR compared between eyes suffering orbital trauma and healthy fellow eyes. 262 6.3.9 Responses to Achromatic Gratings in PVR Mean responses for two affected eyes are shown in figure 6.11 and corresponding readings from their healthy fellow eyes in figure 6.12; the latter show a resemblance to an S-shaped curve. Eyes with PVR 0.8 0.6 E E 0.4 0.2 g o U -0.2 a. -0.4 -0.6 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.11 PGR in eyes with PVR. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 263 Healthy Fellow Eyes to Eyes with PVR 0.8 0.6 E 0.4 t 0.2 • • O o U 1-o.2 a. -0.4 -0.6 Spatial Frequency 0.5 to 5.0 cycles degree Figure 6.12 PGR in fellow healthy eyes to eyes with PVR. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +1- 2 SD). Comparing the responses between the two groups (figure 6.13) suggested a blunting of the PGR in PVR, and revealed a statistically significant loss of responses mainly to higher spatial frequencies (p = 0.039, two-tailed paired t-test). This difference is not as marked as the differences in other disease groups. 264 Pupil Constriction mm 0.15 0.1 0 Healthy Fellow Eyes 0 05 ■Eyes with PVR 0 Spatial Frequency 0.5 to 5.0 cycles i degree Figure 6.13 PGR compared between eyes with PVR and healthy fellow eyes. 6.3.10 Responses to Achromatic Gratings in Amblyopia Mean responses for the subject are shown in figure 6.14 and corresponding readings from their healthy fellow eyes in figure 6.15; the latter show only a subtle resemblance to an S-shaped curve. 265 Amblyopic Eye 1 0.8 0.6 0.4 t• 0.2 • 0 • 0 -02 E. a.• -OA -0.6 -0.8 Spatial Frequency 0.5 to 5.0 cycles degree Figure 6.14 PGR in an amblyopic eye. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). Healthy Fellow Eye to Amblyopic Eye 0.4 0.3 E E 0.2 4- f -0.2 -0.3 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.15 PGR in the fellow healthy eye to an amblyopic eye. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 266 Comparing the responses between the two groups revealed a statistically significant loss of responses mainly to lower spatial frequencies and hence similar to that found in traumatic orbital disease (p = 0.042, two-tailed paired t-test), shown in figure 6.16. This is not as marked as the differences in orbital disease, probably because the changes in amblyopia in this individual were subtle. Pupil Constriction mm DHealthy Fellow Eye otAmblyopic Eye Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.16 PGR compared between an amblyopic eye and its fellow healthy eye. 6.3.11 Responses to Chromatic Gratings in Orbital Trauma Mean responses for the two affected eyes are shown in Fig 6.17 and corresponding readings from their healthy fellow eyes in Fig 6.18; the latter show some slight resemblance to a sigmoid curve. 267 Eyes with Orbital Trauma 0.8 0.6 E 0.4 g 0.2 • • • -0.8 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.17 Red/Green PGR in eyes suffering orbital trauma. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). Healthy Fellow Eyes to Orbits with Trauma 0.2 0.15 E 0.1 t 0.05 0c 0 g. -0.05 a. Spatial Frequency 0.5 to 5.0 cycles degree -0.1 -0.15 Figure 6.18 Red/Green PGR in fellow healthy eyes to eyes suffering orbital trauma. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 268 Comparing the responses between the two groups revealed a statistically significant loss of responses similar to that with achromatic gratings, more diffuse and with smaller amplitudes but nevertheless suggesting a tendency towards loss of preferentially low spatial frequencies (p = 0.0002, two-tailed paired t-test). This is shown in figure 6.19. ❑Healthy Fellow Ey'es Pupil Constriction mEyes ,Alth Orbital Trauma mm 0.08 0.06 0.04 0.02 Spatial Frequency 0.5 to 5.0 cycles degree Figure 6.19 Red/Green PGR compared between eyes suffering orbital trauma and their healthy fellow eyes. 6.3.12 Responses to Chromatic Gratings in PVR Mean responses for the two eyes are shown in figure 6.20 and corresponding readings from their healthy fellow eyes in figure 6.21; the latter barely show a resemblance to a sigmoid contrast sensitivity function curve. 269 Eyes with PVR 0.2 0.15 E 0.1 2 17 0.05 VI 0 0 0. o. -0.05 -0.1 Spatial Frequency 0.5 to 5.0 cycles ! degree Figure 6.20 Red/Green PGR in eyes with PVR. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). Healthy Fellow Eyes of Eyes with PVR 0.25 0.2 E 0.15 E 0.1 t 0.05 °e o 0 0. g -0.05 -0.1 -0.15 Spatial Frequency 0.5 to 5.0 cycles I degree Figure 6.21 Red/Green PGR in opposite healthy eyes to eyes with PVR. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 270 Comparison of the responses between the two groups revealed a relatively small but statistically significant loss of responses, similar to that found with achromatic gratings but with lower amplitudes (figure 6.22). These have a tendency towards preferential loss of higher spatial frequency responses (p = 0.002, two-tailed paired t-test). The discrepancy was not as marked as in the eyes with orbital trauma. Pupil Constriction mm 0.1 0.08 0.06 0Healthy Fellow Eyes 0.04 ■Eyes with PVR 0.02 0 Spatial Frequency 0.5 to 5.0 cycles degree Figure 6.22 Red/Green PGR compared between eyes with PVR and their healthy fellow eyes. 6.3.13 Responses to Chromatic Gratings in Amblyopia Mean responses for the subject are shown in figure 6.23 and corresponding readings from the healthy fellow eyes in figure 6.24; only the latter shows a typical contrast sensitivity function-type function curve. 271 Amblyopic Eye 0.15 0.1 E E 0.05 e -----4,-..___._. 0 ...g 0 I o = -0.05 0. c o. -0.1 -0.15 Spatial Frequency 0.5 to 5.0 cycles ! degree Figure 6.23. Red/Green PGR in an amblyopic eye. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). Healthy Fellow Eye of Amblyopic Eye 0.2 0.15 E E 0.1 g T 0.05 z s• • c 0 o V = ai -0.05 C. -0.1 -0.15 Spatial frequency 0.5 to 5.0 cycles I degree Figure 6.24 Red/Green PGR in the healthy fellow eye of an amblyopic eye. Each data point is the mean of 50 recordings each from 2 subjects. Error bars represent 95% confidence intervals (mean +/- 2 SD). 272 Again response amplitudes were relatively small in comparison for those from achromatic gratings (figure 6.25). There was relatively greater loss of responses to low frequency gratings (p = 0.04, two-tailed paired t-test). Pupil Constriction mm 0 08 0.00 E3 Healthy Fellow Eyes 0.04 ■Amblyopic Eyes 0.02 Spatial Frequency 0.5 to 5.0 cycles / degree Figure 6.25 Red/Green PGR compared between one eye with amblyopia and the healthy fellow eye. 6.3.14 Significance of Findings More caution is needed in interpreting the findings from these groups of experiments than for other sections as patient numbers are relatively small in this section and resolution will also be limited at higher spatial frequencies. The studies herein of the achromatic and chromatic PGR in disease still tend to suggest that there seems to be a parallelism between the PGR in disease and the responses that have been well-documented for some time in psychophysical studies i.e. Weinstein's over 30-year old hypotheses have been upheld experimentally. These findings suggest a tendency towards loss of both achromatic and red/green low spatial frequency responses in traumatic orbital optic 273 neuropathy. As discussed previously, it is generally established that compressive optic neuropathies as in thyroid eye disease are associated with a raised psychophysical stimulus threshold for low spatial frequency contrast stimuli, despite the report by de Marco et al [de Marco et al 2000]. The pupillary responses we found would hence seem to parallel the pattern of loss of psychophysical responses found by the majority of researchers [Arden 1978; Jindra and Zemon 1989; Mounts et al 1990; Suttorp-Schulten et al 1993; Tanner et al 1995]. The loss of response to higher spatial frequencies in PVR with both achromatic and red/green chromatic stimuli parallels the psychophysical response of loss of higher spatial frequencies seen in retinal disease. Finally, the results in amblyopia show a relatively greater loss of low spatial frequency pupillary responses to both achromatic and coloured stimuli, and suggest that the PGR may have a cortical component since ambylopia involves cortical and possibly lateral geniculate nucleas dysfunction. That achromatic gratings are partly cortically mediated has previously been shown first by Barbur et al and supported by results from work by Cocker, Moseley, Bissenden and Fielder [Barbur et al 1992; Cocker and Moseley 1992; Cocker et al 1994; Cocker and Moseley 1996]. The results in these experiments are hence additionally significant in suggesting the red-green coloured PGR is at least partly cortically mediated, and hence may conceivably represent a specific channel. 6.4 The Photostress Test 6.4.1 Photostress Testing A photostress test involves administering a bright light stimulus of fixed intensity to bleach a consistent amount of photopigment in a subject's eye, and then measuring at pre-determined time intervals the subsequent recovery of vision as an index of regeneration of photopigment [Brindley 274 1970; Rushton 1972]. Recovery can be estimated by evaluating visual acuity at successive intervals after bleaching, or less commonly, using pupil constriction [Zabriskie and Kardon 1994]. 6.4.2 Clinical Implications The test is well-established. Recovery from photostress is reduced by advanced age, but is independent of pupil size, ametropia and visual acuity [Margrain and Thomson 2002]. It is occasionally used in clinical practice but a recent literature review by Margrain and Thomson suggests that there are a large number of normal results found in the literature even in patients with macula disease, mainly due to a lack of standardisation of methodology [Brindley 1970; Rushton 1972; Gomez-Ulla et al 1986; Stabell and Stabell 1989; Margrain and Thomson 2002]. Both foveal and extrafoveal macula disease has been found to be associated with prolonged recovery time to photostress, but this was not found with optic neuropathy in some studies [Forsius et al 1964; Severin et al 1967; Glaser et al 1977; Horiguchi et al 1998]. However, other studies, in glaucoma, found that optic neuropathy can reduce recovery from photostress [Zuege and Drance 1967; Goldthwaite et al 1976; Lakowski et al 1976; Jonas et al 1990; Parisi and Bucci 1992; Parisi 2001). Age-related macular degeneration (AMD), particularly the "wet" form associated with subretinal neovascularisation, is associated with poor recovery from photostress [Wu et al 1990; Wolffsohn et al 2006]. Owing to such observations the photostress test is sometimes advocated as a test of differentiating poor vision due to retinal disease or optic neuropathy. However advances in imaging such as Optical Coherence Tomography (OCT), spiral CT and high resolution MRI combined with enhancement and angiography make establishing this difference less challenging than previously. 275 6.4.3 The Pupil Photostress Test In 1994 Zabriskie and Kardon published the only work in humans to date on the use of the pupil as an index of recovery from photopigment bleaching to study pathology in the afferent visual system [Zabriskie and Kardon 1994]. In effect, after bleaching the photopigments with a very bright light source (`photostress'), the pupillary constriction to subsequent lower intensity light flashes was measured using pupillometry. The study was performed as pupil responses can parallel visual responses, but further can be objectively quantified. It was felt that some of the negative results in studies with optic neuropathies might have been due to the use of semi-quantitative methods. Studying the pupil constriction amplitude following photostress in nine subjects with unilateral optic neuropathies due to a variety of compressive and inflammatory aetiologies, there was: i. reduced rate of initial constriction ii. somewhat paradoxically, what was called "less initial sensitivity loss" in the diseased eye compared with the fellow healthy eyes when the pupil constriction was expressed as a ratio of final pupil diameter; it should be noted a derived exponential function was used for that calculation adapted from Naka and Rushton [Naka and Rushton 1966; Zabriskie and Kardon 1994]. iii. Despite the obviously lower pupillary amplitude compared with normal eyes, both resting and post-photostress, pupillary constriction returned to normal magnitude in both diseased and normal eyes in the same period of time. 6.4.4 Factors Affecting Recovery from Pupil Photostress Zabriskie and Kardon suggested that recovery from pupil photostress may be affected by: 276 i. the anatomical extent of involvement by disease, such as retina versus optic nerve, and the type of disease, though the former was not proven by them. ii. duration of photostress, where prolonged photostress increases loss of sensitivity after photostress, but not exponential time to recovery. iii. increasing brightness of photostress increases initial sensitivity loss and also the exponential rate of recovery rate with very bright light. 6.4.5 Technical Considerations Zabriskie and Kardon used a derived exponential function to fit the data points (pupil recovery on the ordinate, time from photostress on the abscissa). This was despite the pupil returning to baseline in all their cases after 200-250 s recording. Clearly response amplitude will tend to be lower in diseased eyes based upon studies already presented in this work with these diseases. Alternative measures to study the specific effect of photostress were used - half-time and time to recovery were calculated; initial sensitivity was expressed as a percentage fraction of the final maximum response amplitude at 50 s; finally, percentage response recovery as fraction of final maximum constriction every 10 s was compared as time frames. Inter-eye PEVR was calculated. Optimum stimulus durations for photostress from 30 to 120 s have been suggested [Margrain and Thomson 2002; Zabriskie and Kardon 1994]. The brightness of a pentorch is commonly used [Margrain and Thomson 2002]. In the present study an indirect ophthalmoscope was used, which is several times brighter (calibrated before and after the study with no significant change in output), and a 60 s photostress duration, with recording for 50 s later, and by which all recordings had returned to baseline (following early trial runs). 277 6.4.6 Patient Recruitment In all there were thirteen subjects in this group of experiments. Five subjects with unilateral compressive and/or traumatic optic neuropathy were recruited; these were the two subjects with traumatic optic neuropathy who underwent testing using gratings and whose injuries were now five months and two years old, aged 35 and 43 respectively; an additional subject, a 37 year old male with severe traumatic optic neuropathy following a suspected penetrating orbital injury through the orbital septum due to a knife 18 months previously with no evidence of haematoma on CT scan; and two subjects aged 40 and 39 respectively at the time of these tests and documented previously in chapter four with blunt orbital trauma and compression of the optic nerve by bone and haematoma respectively. Five subjects with SRNs attended: four with AMD, one an inflammatory aetiology (documented previously). Three subjects with Macula-Off RDs attended (documented previously). 6.4.7 Methodology The experimental protocol documented in chapter two was generally observed, using both achromatic (CRT, 120 cd/m2 square wave patch 22 x 14 deg, duration 1100 ms) and chromatic stimuli of equal photon flux from the quartz halogen light pipe (12.5 log photons/cm2/s, duration 1100ms, produced by monochromatic interference filters of 440 nm, 540 nm and 620 nm). An indirect ophthalmoscope was used to photostress the monocular subject for 60 s keeping it the same distance close to the eye as allowed by its casing and headband. Subjects were warned of the very bright nature of the light. This was done in one eye, and pupillometry performed monocularly. A low intensity test stimulus was used, and applied at 10, 20, 30, 40 and 50 s following each photostress stimulus. A particularity was that averaging could not be done in the same stimulus train owing to the critical time constraint. The photostress-pupillometry process was then repeated on the monocular healthy eye. It was then repeated alternating between eyes to 278 gain a mean, four more times with the same achromatic stimulus (a total of five pupillograms for each 10 s post-photostress). After this the test was repeated five times again using each of the three chromatic stimuli. There was no difference in half-time or time to recovery between diseased eyes and the fellow healthy eyes. 6.4.8 Orbital Trauma 6.4.8.1 Orbital Trauma — Achromatic Recovery from Pupil Photostress There was no difference in half-time or time to recovery between diseased eyes and the fellow healthy eyes. There was less initial constriction in recovery from photostress in the diseased eyes (reduced initial sensitivity as a function of final maximum constriction in each group), in distinction to the results of Zabriskie and Kardon — in my work initial pupil constriction in the healthy fellow eyes was 92% of the final maximum pupil constriction, but this was only 77.6% of the final maximum constriction in diseased eyes. Comparing the recovery profiles (percentage response recovery as fraction of final maximum constriction every 10 s) revealed a statistically significant difference (p < 0.0002, two-tailed paired t-test) (figure 6.26). Inter-eye PEVR showed no trend. 6.4.8.2 Orbital Trauma — Chromatic Recovery from Pupil Photostress The recovery curves are shown in figure 6.27. Red light data had very wide confidence intervals for the first post-photostress measurement; variability in the initial time course of photostress recovery with the pupil has previously been noted by Kadlecova and also by Zabriskie and Kardon though no other data exists on the use of chromatic stimuli following pupil photostress [Kadlecova 1960; Zabriskie and Kardon 1994]. There was reduced initial constriction amplitude in response to all three colours of stimuli. With red light only there was a substantial reduction in subsequent constriction amplitudes as a percentage of final amplitude, a phenomenon 279 not found in response to the blue or green stimuli to the same degree — at 25 seconds post-photostress this difference was barely perceptible using blue light, and was 88% of final maximum pupil constriction in healthy eyes with green light compared to fellow diseased eyes where it was 77% 1.6 1.4 1.2 E E • Healthy Eyes -0 1 ■ Diseased Eyes a. E —Poly. (Healthy t'• 0.8 Eyes) Poly. (Diseased Eyes) 0 0.6 0.4 0.2 10 20 30 40 50 Time s Figure 6.26 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in five subjects with unilateral orbital traumatic optic neuropathy following achromatic photostress. Error bars represent 95% confidence intervals (mean +/- 2 SD). For clarity only one error bar has been shown per data point. Polynomial curves (denoted as Poly.) have been applied to best fit the data. 280 Comparing the rate of recovery overall (percentage response recovery as fraction of final maximum constriction every 10 s), this was statistically very significant with long wavelength light (p > 0.0001, two-tailed paired t- test), and much less significant with green light (p > 0.1, two-tailed paired t- test) and still less with blue light (p > 0.2, two-tailed paired t-test). Inter-eye PEVR showed no trend with any colour. 6.4.9 Vitreo-retinal Disease Aside from the expected reduced amplitudes of constriction in diseased eyes, photostress was unable to differentiate macula dysfunction with SRNs. There was no difference in initial sensitivity loss as a function of maximum response of diseased and healthy eyes respectively, and, no difference comparing the rate of recovery overall (percentage response recovery as fraction of final maximum constriction every 10 s) including half-time. There was no relationship between inter-eye PEVR. With Macula-Off RDs there were positive findings using achromatic light, and findings of borderline statistical significance with green light, and none with long wavelength and blue light. With achromatic light there was reduced initial sensitivity as a function of final maximum constriction in eyes with RD (3.8%) and fellow health eyes (28%). Comparing the rate of recovery overall (percentage response recovery as fraction of final maximum constriction every 10 s) was statistically significant (p = 0.01, two-tailed paired t-test) being less in eyes with Macula-Off RDs. There was no trend in inter-eye PEVRs. All parameters with green light were of borderline significance, and comparing the rate of recovery overall (percentage response recovery as fraction of final maximum constriction every 10 s) these just missed statistical significance (p = 0.052, two-tailed paired t-test). There was no relationship between inter-eye PEVR in any groups. 281 0.9 0.8 0_7 mm de 0.6 litu 0.5 • Healthy Eyes 0.4 Amp n Diseased Eyes 0.3 tio ic tr 0.2 s 0.1 Con 0 -0.1 -Fa 20 30 40 50 Time s 1.6 1.4 mm 1.2 de 1 litu 0.8 ••• • Healthy Eyes Amp 0_6 A A Diseased Eyes tion 0.4 tric 0.2 Cons 0 -0_2 10 20 30 40 Time s 2 Healthy Eyes A Diseased Eyes 1t 5 3 D 4 5 0 1 Time s Figure 6.27 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in five subjects with unilateral orbital traumatic optic neuropathy following chromatic photostress — (top) long wavelength light; (middle) green light; (bottom) blue light. Error bars represent 95% confidence intervals (mean +1- 2 SD). For clarity only one error bar is shown in the top and bottom graphs. 282 Figure 6.28 Pupil photostress: mean achromatic pupil constriction amplitudes for healthy and diseased eyes in three subjects with unilateral Macula-Off retinal detachments. Error bars represent 95% confidence intervals (mean +/- 2 SD). For clarity only one error bar has been shown. 6.4.10 Significance of Pupil Photostress Tests There are some differences with the results of Zabriskie and Kardon with achromatic stimuli, essentially in terms of some paradoxical results in that work where reduced initial sensitivity was found in healthy eyes, and which is counter-intuitive; this suggests that the exponential function used in that work may have been inappropriate to describe a non-exponential process [Zabriskie and Kardon 1994]. Zabriskie and Kardon also noted "The photostress is not simply a form of the relative afferent pupillary defect test. The two pupillary tests call for different stimulus durations and other experimental conditions which may not be equivalent." This may be explained by rod activity. The lack of a significant trend in inter-eye PEVR within any photostress stimulus paradigm used in this work, including in subjects in whom such a difference was found following dark adaptation in earlier experiments, is consistent 283 with the dependence of PEVR as an index on neuronal firing arising in rod- mediated photoreception. The relative reduction in rod-driven neuronal firing following photostress therefore disturbs inter-eye PEVR relationships even in subjects where it showed a trend following dark adaptation. The photostress results for traumatic orbital optic neuropathy show a reduction in all parameters with white light, and long wavelength light, and in most parameters with green light, and least reduction with blue light. These changes are consistent with known reductions in red-green sensitivity in orbital compressive lesions found in chapter five and also known changes in colour vision testing. Loss of ganglion cell firing in neurones driven by rods is very likely in traumatic orbital pathology sufficient to impair function in the optic nerve. The lack of positive findings for photostress with SRNs is not inconsistent with the known photoreceptor deficits in these patients (shown in the ERGs in chapter four). There was found to be little loss of rod-driven neuronal firing in these subjects with purely macula disease, and loss of cone-driven neuronal firing found on electrophysiology in this group may not be being detected by pupil photostress owing to largely intact rod-driven neuronal firing between healthy and diseased eyes. The only mention in the literature of attempts to study the pupil photostress test in macula disease have been by Zabriskie and Kardon who in 1994 wrote of then ongoing and unpublished work on this topic in the context of their paper on the pupil photostress test; to date their work on macula disease has not been published, possibly hinting at a negative result [Zabriskie and Kardon 1994]. The extension of cone-rich macula dysfunction topographically to include the rod-rich retinal periphery in Macula-Off retinal detachments (shown in the electrophysiological findings in chapter four) could explain the reduced pupil photostress parameters with achromatic stimuli in this condition, as well reduction in photostress parameters in response to the long wavelength 284 and to a lesser extent the green stimuli. In this context both L- and M-cone activity in the macula, and rod activity, are relevant to consider. In the immediate future it seems unlikely that pupil photostress testing will result in routine clinical application and the findings are more relevant to understanding cellular mechanisms. This is partly as standardisation of the photostress test is more challenging than most tests. 285 Chapter 7 A Double Constriction in the Pupil Waveform: Signature of a Novel Photoreceptor? 7.1 Rationale Why investigate the fundamental nature of the pupil light reflex? In previous chapters it was shown that pupil evasion ratio (PEVR) can accurately differentiate eyes with macula disease from healthy fellow eyes if the stimulus is achromatic or long wavelength light. Reduced inter-eye PEVR was associated with greater pupil evasion. There was some evidence towards a tendency to this with certain types of intra-orbital inflammation in which the optic nerve was also involved. These relationships were dependent on the spectral composition of the light source, being less reliable with medium wavelength (green) stimuli, and did not exist at all for short (blue) wavelengths. It is proposed that the spectral variation of PEVR, in retinal disease may reflect a background component to the pupil waveform that shorter wavelengths of light elicit, and that thereby disrupt the classical physiological sequence of pupil constriction followed by re-dilation that both white light and longer wavelengths of light elicit. However such a putative component to the PLR has never been described. A literature search showed that all publications using chromatic stimuli to investigate the PLR employed short duration stimuli of no more than a couple of seconds, and all work on the related phenomenon of pupil escape employed achromatic stimuli. This assessment of the literature sounds surprising, but its conclusion was not refuted after seeking the opinion of a pupillometrist with almost a lifetime of experience studying the pupil [Smith SE, personal correspondence]. In this section the results of study of the PRC and PLR in response to sustained stimuli are presented to test the hypothesis that a significant background component to the pupil waveform exists that is 286 specifically elicited by a sustained stimulus and is dependent on stimulus frequency. 7.2. Subjects Forty-eight eyes from 24 healthy subjects were studied. Subjects had a mean age of 36 (range 26 to 48), 12 males and 12 females. Inclusion criteria were normal visual acuity (logMAR), colour vision (Ishihara pseudoisochromatic plates) and visual field (confrontation testing and Humphrey 24-2), normal pupil light reflex (indirect ophthalmoscope, high intensity beam), normal undilated and dilated routine ocular examination (tropicamide 1%) in both eyes. Exclusion criteria were any ocular or systemic diseases or non-physiological anisocoria, determined by testing 48 hours or later after initial examination for absence of efferent sympathetic or parasympathetic iris dysfunction. This was determined by comparison of pupil diameters in light and in dark. Subjects abstained from autonomic stimulants such as caffeine, tea and chocolate and any medications for at least 48 hours to any pupillometric experiments and as described in the general experimental methodology in chapter two. 7.3 Stimulus Presentation Acromatic and chromatic stimulus paradigms consisted of high intensity sustained light output for 10 s. These were generated initially using a CRT display (470 nm, 520 nm and 600 nm) to get an appreciation of the level of intensity required to elicit any putative physiological alteration on the pupil waveform (these particular stimuli were neither isoluminant when used in this context, nor radiometrically equivalent). Changes were noticed using even moderate CRT intensity settings. A strong monochromatic quartz halogen light source became available and was used instead to produce stimuli of equal photon flux as follows: 287 i. monochromatic interference filters of low band width (eight wavelengths: 440 nm, 460 nm, 481 nm, 500 nm, 520 nm, 540 nm, 560 nm, 580 nm) transmitted by fibre-optic light pipe 1 metre in front of the tested eye. Intensity was 12.5 log photons/cm2/s with all stimuli and duration 10 s for each stimulus. ii. achromatic light via the same source of 12.5 log photons/cm2/s for the same duration. Subjects' eyes (n=48) were tested monocularly, swapping the patch around to test the opposite eye when completed. Subjects were given 20 minutes of dark adaptation prior to testing using each monochromatic stimulus and the achromatic stimulus. Each stimulus was presented ten times. In an extension of some positive findings shown later in this chapter a video was used to record pupil constriction to the eight wavelengths plus achromatic light from the fibre-optic light pipe 1 metre in front of the tested eye. Intensity was 12.5 log photons/cm2/s but stimuli were on for 5 minutes. 7.4 Response Recording Pupil response was measured using the P_Scan 100 infrared video- pupillometer already described in chapter two, though measurement was continued for 10 s. Data was processed as described under the general methods documented in chapter two. With the extended recording for 5 minutes P_Scan could not be used so a digital video system was set up and the frames analysed using Microsoft Photoshop. 7.5 Results: Routine Findings A pupil contraction was observed with all stimuli. Latency of contraction onset varied with wavelength of stimulus and was shortest for 500 nm light (mean 1260 ms) varying on either side following the typical dark-adapted 288 pupil spectral sensitivity curve [Alpern and Campbell 1962]. Amplitude varied inversely to latency and was also greatest at 500 nm for all stimulus intensities. The pupil gradually re-dilated in all eyes at all stimulus intensities with all stimuli. These findings are expected and well- documented [Loewenfeld and Lowenstein 1999]. 7.6 Novel Findings: Spectrally Dependent Double Pupil Constriction Pupillography data is shown in table 7.1 (all 8 monochromatic stimuli and white light); mean pupillograms for 4 of these wavelengths are illustrated in figure 7.1. Stimulus 440 460 481 500 520 540 560 580 White nm Light 100 100 100 100 73 42 0 0 0 % of all stimuli eliciting a2'4 constriction Mean l't constriction amplitude elicited 0.70 0.705 1.137 1.14 0.76 0.80 1.0 0.73 2.3 Mean 2'd constriction amplitude elicited 0.88 1.01 0.93 0.76 0.51 0.45 0 0 0 Al /A2 0.86 0.7 1.2 1.5 1.49 L77 NA N/A NIA Two-tailed 0.035 0.04 0.07 0.02 0.24 03 o.oi 0.02 0.021 1-test` Table 7.1 Response parameters for all 8 chromatic stimuli and the achromatic stimulus (all stimuli of lOs duration). Al= amplitude of first constriction; A2 = amplitude of second constriction. The closer Al/A2 is to 1 the greater the similarity between first and second pupillary constrictions if both are present — where there was no second constriction the test is not applicable (denoted as N/A). *Paired t-tests comparing association between first and second pupil constrictions recorded every 20 ms over 10 s within an individual stimulus wavelength — with this test applied to this data the lesser the statistical significance the greater similarity it suggests between the first and second pupillary constrictions. 289 Achromatic light and wavelengths of light of 560 nm and above elicited only an initial contraction of the pupil. With light intensities between 520 nm and 540 nm there was a second temporal contraction of the pupil observed in 73% and 42% of eyes respectively, without appreciable change in latency. The presence of a 'double' constriction was by no means consistent within an eye at these wavelengths, and while discernible in all individuals in at least some pupilllograms, was often very rudimentary, however when it occurred it was fairly stereotyped for the individual (mean intra-eye variability in second constriction amplitude standard deviation 0.175 mm and 0.2 mm respectively for 520 nm and for 540 nm). Second constrictions tended to occur between 3.8 and 6.8 s after stimulus onset. The association between first and second pupil constrictions within different eyes were compared for each stimulus wavelength using a two-tailed paired t-test (table 7.1). A statistically significant difference implies a difference the between the first and second constriction waveforms. This occurs at two extremes — with large double constrictions (e.g. 481nm in figure 7.1) where the second constriction waveform is larger than the first constriction, and where there is no second constriction (e.g. 560 nm in figure 7.1) in which cases comparison of data points was in effect between the first constriction and random fluctuations in the waveform, which was made with sensitivity in excess of 0.01 mm and extracted a recording every 20 ms. Constrictions in between such extremes were more similar in shape and size (e.g. 520 nm in figure 7.1), so the difference is of limited statistical significance. For all stimuli of wavelength 500 nm or less, 100% of pupillograms displayed the second constriction. Still lower wavelength stimuli elicited progressively greater amplitudes of second pupil constriction, which peaked between 460 nm and 481 nm. Despite the change in amplitude with stimulus frequency, and also the inverse relationship that latency had to the amplitude of the first constriction, the onset of the second pupil constriction was between 3890 ms and 6780 ms, and was unrelated to stimulus frequency. Further, it would also vary independently to amplitude within individuals (mean 432ms within an individual eye). 290 Fi gure 7 Pupil Constriction Amplitude mm Pupil Constriction Amplitude mm Pupil Constriction Amplitude mm Pup! Constriction Amplitude M U1 C31 01 cr, CI 01 A A U7 U1 U1 U1 U7 M 01 01 mm CO 01 N :S=. ul is3▪ A 01 CO s4 Ln .1 0 I" P... PI U1 PPPPP 9) Cn -4 'CO cn Ch 0 M 520 0 500 ean 620 1000 1040 580 1240 1500 pu 1560 1160 pill 2000 2080 1860 1740 2500 o 2600 2480 gr 2320 3000 a 3120 3100 2900 ph 3500 y 3640 3720 3480 4000 of 4160 4060 48 -4 4340 —j 4500 4680 3 4640 h CD 4960 • 5000 eal 5200 3• 5220 cr) 5580 to 5500 th 5720 • 5800 6000 y 6200 6240 e 6380 6500 yes f 6760 6820 6960 7000 7280 7440 - cyi or f 7500 7540 0) 7800 8060 o 8000 0 our 8120 8320 8680 8500 8700 of 8840 3 9300 9000 th 9280 9360 9500 e 9920 9860 sti 9880 10000 m uli . Of note, no additional contraction was found beyond the second contraction as described above in those experiments recording for five minutes. 7.7 Age and Gender Relationship There was no obvious relationship to age or to sex, but of note most all subjects were comparatively young. It could incidentally be elicited in older individuals outside of the results displayed using the non-matched CRT stimuli. 7.8 Potential Explanations: Less Likely Aetiologies 7.8.1 Excluding Artefact: Stimulus Generation The spectral change in responses is unlikely to be artefactual. They could be elicited by CRT-generated stimuli. 7.8.2 Excluding Accommodation Accommodation could not be held responsible as there was no accompanying convergence on eye movement tracking. Further, the stimuli in both paradigms were placed at a distance of 1 metre away. 7.8.3 Fatigue Waves Fatigue waves are of a different shape and are very unlikely to have been present in these generally young individuals with such regularity. Further the fairly consistent temporal position of the second constriction is most unlikely to occur due to fatigue in so many subjects within the same time period. 292 7.9 Potential Explanations: Significance of the Double Constriction 7.9.1 Overview The results suggest a fundamental and basic addition to understanding of the PLR, which is a biologically important paradigm for physiological reflexes. In addition to the 'classic' pupil constriction a double constriction is elicited with short wavelength light. 7.9.2 Implication for Pupil Evasion Ratio The greater prevalence of this second constriction at lower stimulus wavelengths can explain the disruption in inter-eye PEVR found with short wavelength stimuli in retinal pathologies and in some types of compressive orbital lesion, and which expedited its study. 7.9.3 Significance to Melanopsin-Containing Photoreceptors The spectral sensitivity of the second constriction is not readily explained using previous models of classic ocular photoreceptors (rods and cones) which are traditionally used to understand the PLR [Alexandridis 1985]. The pupil undergoes a second reflex constriction in response to sustained light of shorter wavelengths, with a peak spectral sensitivity at 481 nm. This is distinct to the peak spectral sensitivity of rods and cones, but is identical to the most widely reported speak spectral sensitivity for the recently discovered mammalian non-rod non-cone photoreceptor [Foster 2005]. Non-rod non-cone photoreception, which is localised to a subtype of giant ganglion cells in the retina and optic nerve of certain mammalian species, uses melanopsin as its photopigment, and is known to contribute 40% of pupillomotor force in rd/rd cl rodless coneless mice, but where only brief stimuli have been studied [Lucas et al 2001 and 2003]. 293 7.9.4 Significance to Research Methodology The results suggest the effect of non-rod non-cone photoreception on the PLR may be observable as a distinctly separate component in the pupil's constriction in healthy humans. Previously the effects of putative photoreception by pRGCs had only been shown using retinally degenerate models in animals. The results suggest the novel pRGC receptor is found in humans and produces a second contraction in the physiological pupil light reflex elicited within the time course of a sustained light stimulus of 10 s. This has been assayed and isolated as a distinct and grossly observable component to the pupil waveform using pupillometry in healthy human subjects. 7.10 Physiological Purpose of Second Constriction The photoreceptive melanopsin-containing ganglion cell may be understood as a brightness receptor [Foster 2005]. The function of a second pupil constriction might be to restrict short wavelength bright light to which the receptor is maximally sensitive from exposing the retina to unduly high levels of brightness, and possibly optimising non-rod non-cone photoreceptor function by preventing excess bleaching of the melanopsin photopigment. 7.11 Studies in Retinas of Laboratory Primates Work from primate models suggest a double pupil constriction mediated by pRGCs but with a slightly different time course to that in humans presented in this thesis. During the undertaking of the experiments presented in this thesis Gamlin and colleagues independently presented work at the Association for Research in Vision and Ophthalmology in 2004 (Fort Lauderdale, Florida, United States) on results from a macaque monkey model that was rendered functionally rodless and coneless by blocking ON and OFF retinal channels with an intravitreal cocktail of L-2-amino-4- 294 phosphobutyrate (L-AP4) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX). The pupil response showed two separate constrictions in response to a sustained light stimulus of 10 s, at stimulus onset and at 10 s with maximal spectral responses in the region of 480 nm. It was unclear to Gamlin who was speaking, and the audience, whether the second constriction was a direct effect of the sustained stimulus which was turned off about 10 s or a response to the stimulus being turned off. Three years later the same group have published some of their work in the peer- reviewed literature. However they have interpreted the second constriction as being possibly due to stimulus offset which was at about 10 s [Gamlin et al 2007]. However presumably as that group's experiments were more or less completed by the time the data was fully analysed, no work has been presented as to whether the second constriction would still be found at 10s in functionally rodless and coneless macaque monkeys subjected to a much longer stimulus than 10 s. That would settle the issue as to whether the responses in macaques were due to stimulus offset or directly to the sustained stimulus. Work from healthy humans in this chapter, and humans without rods and cones in later chapters of this thesis, suggest the latter may be a possibility, but both effects cannot be discounted given work on darkness reactions presented at the end of this thesis. 295 Chapter Eight Humans Lacking Rods and Cones: Electrophysiology, Imaging, Genetics and Clinical Findings 8.1 Introduction The possibility of human melanopsin-expressing ganglion cell photoreception has been suggested in this thesis by the double pupil constriction, and specifically its chromatic sensitivity for short wavelengths below 500 nm. Neither rods nor cones have a peak spectral sensitivity in this region, and only, in some mammals, the melanopsin-expressing photosensitive retinal ganglion cell (pRGC). This line of evidence is, however, indirect. To investigate the existence of a putative pRGC receptor in humans ideally requires that a human model lacking rods and cones should be studied for pupillary, visual and circadian behavioural responses to provide direct unequivocal evidence, rather than inferences. Genetic and chemical techniques used in animal models to establish a functionally and/or structurally rodless coneless retina are self-implicitly unethical and inappropriate for use in humans. Instead establishing the existence of a rodless and coneless state in a human is described in this chapter through study of a suitable group of human retinal diseases. 8.2 'Classic' Photoreceptor Dystrophies Classic photoreceptor dystrophies are the obvious candidate for such a disease model. The retinal dystrophies are an inherited (rarely mutated) group of potentially very serious defects in rod and cone function. In rod- cone dystrophies the rod dysfunction precedes that of cones and is more severe, such as retinitis pigmentosa (RP); in cone-rod dystrophies the converse is true such as in cone-rod dystrophy (CRD). Blindness to 'barely' no perception of light (NPL) by clinical standards can occur, especially by old age in the most severe types, though can be in even childhood, or never 296 - the spectrum of severity is very wide and related in large part to the underlying genetic abnormality. Even eyes that are classified as 'barely' NPL may not exhibit a totally rodless and coneless state outside very elderly sufferers who have additionally been blind from childhood [Gregory-Evans K, personal communication]. In this and some subsequent sections, five subjects with an autosomal dominant cone-rod dystrophy and R.P. were studied. Despite all being barely `NPL', one subject (SC), a sufferer from a well-documented CRD, was considered on the basis of pupillary action spectra as a candidate for being the one most likely without rods and cones. 8.3 Genotype-Phenotype Correlation in Autosomal Dominant Cone- Rod Dystrophy An 87-year old affected member of a family expressing an autosomal dominant cone-rod dystrophy was the focus of study, but other members of this family, as well as subjects with RP, were included. CRD has been described as a severe, early onset phenotype with patients progressing to no perception of light by the 5th decade of life [Evans et al. 1995]. Detailed electroretinography and psychophysical testing in much younger affected members of this and other families have sub-classified the cone-rod dystrophy as seen in the earlier stages of disease as 'type 2': a severe central scotoma, eccentric fixation, cone more than rod function loss in the peripheral field and relatively normal mid-peripheral visual fields, according to the sub-classification of cone-rod dystrophy of Yagasaki and Jacobson and as type lb: less rod than cone dysfunction but with matching areas of cone and rod function threshold elevation in the peripheral field according to the sub-classification of Szlyk et al [Yagasaki and Jacobson 1989; Szlyk et al 1993]. This phenotype has also undergone molecular genetic assessment. The disease gene has been localised to chromosome 19q13.4 (CORD2) [Evans et al 1994]. The actual disease gene has yet to be identified. SC by the time of testing had progressed beyond these earlier stages and was no longer able to perceive light for several decades. 297 This family, based in Liverpool, England, is the one originally used to map the CORD2 region by Evans and colleagues. Subsequent work established that CRX gene mutations were found in many families with CORD2 approximately ten years ago [Freund et al 1997; Tzekov et al. 2000; Itabashi et al. 2003]. However, a number of laboratories have attempted many times to find a CORD2 mutation in this family, and concluded that a CRX mutation is not involved [Hodges et al. 2002]. The absence of a CRX mutation in this family is explained by the nature of the CORD2 region which is 2.4Mb [Gregory-Evans K, unpublished data] and contains 75 genes, many of which are expressed in the retina and are potential alternatives to CRX for mutation in this family. No affected patients aged 65 or over in work performed in the early to mid-1990s by Evans et al (including this subject, who was aged over 65 even in 1995) reported visual perception [Evans et al 1994 and 1995; Gregory-Evans K, personal communication]. This degree of severity was not seen in younger family members and may not be true of families with CRX mutation, but is true of this family whose phenotype was uniquely described by Evans [Evans et al 1994 and 1995; Gregory-Evans K, personal communication]. SC was felt to be the most likely candidate for being rodless and coneless in Britain out of the vast number seen by one of the clinical collaborators who had worked and done research on the genetics of photoreceptor dystrophies at both of London's specialist eye hospitals, in Liverpool, and at Imperial College London [Gregory-Evans K, personal communication]. The family and subject SC were, though based in Liverpool, recruited for experiments by the investigator at the Western Eye Hospital in London, for which they attended many times, in collaboration with the academic unit of Imperial College London at the Western Eye Hospital and the clinical collaborator in this section of the work. 8.4 Effect of Cataract Older patients may have a certain degree of cataract. In view of some of the findings presented in chapter six, no subjects were recruited with optically 298 significant cataracts. No subjects had cataract beyond CO2CL2 by the LOCSIII system (Lens Opacities Classification System III) [Appendix II]. SC had mild cortical lens opacities (Cl, C2) in left and right eyes respectively. The potential effects of cataract on light absorption and pupillometry have been considered in some detail in chapter six. The subjects did not have cataract of type or severity sufficient to cause optical effects. Even so, predictions for retinal light absorption taking into account preretinal lens absorption were performed and failed to show a significant difference in the wavelength of light being absorbed (chapter nine). 8.5 Retinal Function All subjects met all clinical criteria of profound blindness arising from degenerative retinal disease. These included pupils unreactive to light after standard pen torch and indirect ophthalmoscopic examination, and self- reported inability to perceive light. 8.6 Fundus Photography and Ocular Coherence Tomography These are shown in figures 8.1 and 8.2 for subject SC and failed to identify an outer retina. High resolution ocular coherence tomography (OCT) imaging was performed with a third-generation interferometric, non- invasive optical tomographic imaging device (OCT3 Zeiss Humphrey Division, Dublin, CA, USA) providing 2-3 mm tissue penetration and thereby an 'optical biopsy' of the eye in cross-section [Voo et al 2004]. Axial resolution was 10 microns while the diameter of rods and cones is 1-2 microns, and their length is 40-80 microns, hence enabling the existence of all but individual cells to be anatomically excluded. A representative tomogram of SC is shown in figure 8.1 for retinal periphery (C) and fovea (D). 299 Figure 8.1 Fundoscopy of A) SC compared with B) normal age-matched control. C,D) Optical coherence tomography of the left eye of SC. Abnormally thin retina (less than 160 microns) seen in C) the peripheral retina and D) the central macula region. No identifiable outer nuclear layer or photoreceptor layer suggests that photoreceptors are absent. Abnormally high reflectivity is seen in the choroid (Ch). Compare with images from a normal age-matched subject (E,F) where stratification within the neurosensory retina, particularly the outer nuclear layer (ONL), can be seen. By contrast the ganglion cell and nerve fibre layers of the inner retina of SC are of relatively normal thickness, and there is no cellular disruption, allowing clear recognition and delineation of normal histo-architecture in both retinal periphery and macula. G) Comparison of the normal macula profile in an age-matched individual (within green limits, as shown in OCT image in F) illustrates loss of normal macular contour in SC (black line, as derived from D). The normal distribution percentile correlates the color coded areas of the figure to percentages of age-matched people who might possess retina within that region. V = vitreous, NR = neurosensory retina. 300 Figure 8.2 OCT of the near retinal periphery also suggests an amorphous, atrophied outer retina and relatively intact nerve fibre layer. 8.7 Electroretinography Electroretinographic responses were recorded to ISCEV standards in subject SC, as documented in chapter two. Electroretinography demonstrated no detectable rod or cone function. Representative ERGs are shown in figure 8.3. 301 Figure 8.3 electroretinographic responsesforSC.Lightpresentation wasatthestartof and age-matched,normaleyeforcomparison(B,D,F).A,B)darkadapted lower lightintensities);C,D)adapted(mixedphotoreceptor)responses. responses to30Hzflickerstimuli.Whitelightstimuliat3.0cd/m (rod photoreceptorpredominant)responses(nointhesubjectat Note: Cshowsadriftingbaseline;E,F)lightadapted(conepredominant) illumination on were usedforalltests.FiguresC,D,EandFdonewithbackground recordings inallcases. Electroretinographic responsesfromstudypatientSC(A,C,E) 25 microv.oltsidivision 15msecidivisIon (25.0 cd/m Sc 302 2 ). Tracesshownodetectable NORMAL 2 intensity 8.8 Implications The failure to identify an outer retina physiologically or anatomically was commensurate with, and predictive of, almost nine decades of blindness in SC. 303 Chapter Nine Direct Evidence for a Melanopsin-Containing Non-Rod Non-Cone Photoreceptor: Irradiance Curves and Pupil Action Spectra 9.1 Purpose of Action Spectra: Overview Action spectra are a means of establishing the spectral sensitivity of a biological response by calculating the amount of light of different wavelengths required to elicit the same response [Peirson et al 2005; Foster et al 2007]. The classical approach to studying photopigments was by methods such as photopigment absorption spectroscopy including microspectrophotometry (MSP) [Alpern and Campbell 1962; Schwartz 2003]. According to the first law of photochemistry only the radiation absorbed by a substance can produce a photochemical effect, hence direct measurement of absorption provides the clearest measure of photopigment spectral sensitivity [Peirson et al 2005; Foster et al 2007]. The second law of photochemistry (the Stark-Einstein Law) states that for each photon of light only one absorbing molecule is activated. Historically, studies of the visual pigments of the retina were conducted by measuring the absorbance of retina pigment extracts by spectrophotometry, such as radial densitometry and microspectrophotometry (MSP), as detailed in chapter one. These were used to determine the sensitivity of rod opsins and the difference between rhodopsins (containing vitamin A1) and porphyropsins (containing vitamin A2). A problem of retinal extracts is that the resulting extract may contain a mixture of photopigments, and alongside retinal rod pigments the smaller contributions from cone opsins are found, though these can be biochemically isolated [Knowles and Dartnall 1977]. Photopigment extracts are also most suited to tissues containing large amounts of pigment [Peirson et al 2005; Foster et al 2007]. Such approaches are hence limited in the context of studying circadian photopigments as the location of the photoreceptor itself may be unknown, and if located 304 photopigment concentration may be too low for such absorption techniques. Action spectra provide then an indirect approach to analyzing such photopigments, establishing the spectral sensitivity of a biological response at different wavelengths. If relatively large amounts light are required to produce a certain biological response, such as pupil constriction or suppression of pineal melatonin, it implies a weak absorption by the photopigment mediating this response, and conversely, a lower light dose to evoke the same response implies a high absorption. With the reciprocal of the light dose plotted against wavelength (X), the resulting action spectrum directly correlates with the absorption spectrum of the pigment mediating the response [Peirson et al 2005; Foster et al 2007]. 9.2 Pupillometry and Action Spectroscopy On the basis of work with double pupil constriction, it was hypothesised that the five subjects putatively with CRD or RP should also possess some pupil reactivity to bright light - despite the clinical reports that they were unresponsive to the light from a pen torch and indirect ophthalmoscope. The action spectra, using the pupil effector responses as a biological assay, were calculated as follows. 9.3 The Biological Assay: Pupillometry Direct, monocular constriction responses to light stimulation were recorded at 10 irradiance intensities (1011-1016 photons/cm2/s) for each of 8 wavelengths (420 nm, 460 nm, 481 nm, 500 nm, 515 nm, 540 nm, 560 nm and 580 nm) in both eyes of the five subjects. Pupil area was monitored using the P-Scan_100 infrared pupillometer as documented in chapter two. Stimuli were of 10 second duration provided by a quartz halogen light source. Irradiance and wavelength were controlled using neutral density and monochromatic interference filters, respectively (10 nm half-band width). Five recordings were made from each eye. Responses were relatively small amplitude, as is expected given the extent of outer retinal degeneration. A double pupil constriction was found in all eyes at the shorter stimulus 305 wavelengths. The variability of responses was least in subject SC, in whom even measures of latency were highly repeatable without extensive averaging (figure 9.1). 6000 5000 -v7 E 4000 c 0 F.) - - -lo- - - 1ST r; CONSTRICTION c 3000 0 - - -N- - - 2ND 0 0 CONSTRICTION >, 0 c a) 2000 1000 0 420nm 481nm 515nm 560nm Wavelength (nm) Figure 9.1 Wavelength dependency of pupil constriction latency in subject SC (average of 10 responses). 9.4 Constructing Irradiance Response Curves and Action Spectra: Overview For each IRC, the maximum and minimum responses were determined from the data set from both eyes. The photon flux necessary to elicit a 50% maximum response (known as the EC50) was determined by fitting a sigmoid slope using an iterative approach to minimise the sum-of-squares 306 employing a quasi-Newton algorithm (the direction and magnitude of the change were determined by a computer software programme employing differentiation using the Frontline systems website). This technique has been used often and is essentially a least squares approach to nonlinear regression. The spectral sensitivity was determined from equal photon flux. The ?max was then determined by fitting the Govardovskii template using a similar approach to the IRC fitting, changing the 2max to minimise the sum- of-squares to provide the best fit [Govardovskii et al 2000]. R2 values are determined in the standard manner (R2=1-(SSreg/SStotal)). 9.5 Construction of Irradiance Response Curves: Principles and Details For each wavelength, the irradiance response curve (IRC) was plotted as log photon flux (x-axis) versus response magnitude (y-axis). The EC50 was then calculated for each wavelength from this IRC by fitting a sigmoid function to the IRC. A sigmoid curve may be defined by four parameters: (Top — Bottom) Response = 1+1 00EC50-n)A) where 'Top' and 'Bottom' correspond to the maximum saturating response and baseline response, respectively; the EC50 is the photon flux necessary to elicit a 50% maximum response (around which the curve rotates); and k is the Hill slope of the curve. Ideally a saturating response should be obtained for every wavelength, and a four-parameter model used for every IRC. However, in practice this is often not possible, and, as here, it simply was not feasible to produce the photon flux necessary to produce a maximal response from the available light source despite its high intensity for all stimulus wavelengths. If the 307 saturating response is known, then this value may be used as the maximum for all IRCs [Peirson et al 2005; Foster et al 2007]. In many biological systems no detectable response may be obtained with a low photon flux, resulting in the baseline being set to 0. This results in just two remaining parameters to fit, in what is called the 'global' model [Motulsky and Christapoulos 2004]. It is possible to define one complete IRC and then use a single, non-saturating irradiance at other wavelengths [Peirson et al 2005; Foster et al 2007]. Irradiance response curves (IRCs) were generated by plotting pupil constriction (y-axis) against light dose (log photons/cm2/s) and are shown in figure 9.2 for the five subjects (ten eyes) and figures 9.3 and 9.4 for Subject SC. The details of these processes were as follows. IRCs were fitted with a four- parameter sigmoid curve: y = a + (b (1+10(')/d)) where y is the response, a is the baseline, b is the maximum response, c is the irradiance required to evoke a 50% pupil constriction (IR50), x is the photon flux (log photons/cm2/s) for which y is to be derived, and d is the slope of the curve [Peirson et al 2005; Foster et al 2007]. In linear regression the line of best fit through data points can be solved analytically; correlation describes the extent to which the line of best fit matches the data points, and in this the maximal correlation of a line with its data = 1. Unlike linear regression, the best fit of a non-linear model such a sigmoid curve cannot be solved analytically, and numerical methods are required, in this case a quasi-Newton algorithm which represents an iterative approach to root-finding (an alternative which has not become widely used for these specific types of calculations with IRCs is Halley's method) [Peirson et al 308 2005; Foster et al 2007]. An iterative approach is a mathematical technique involving use of successive approximations, in this case to minimise the sum of squares of the model (SSmodel), which is equal to the sum of the difference between the data and the sigmoid model, squared (so as to become sign independent) at each point on the IRC, i.e. (data - model)2. This is compared to the total sum of squares (SStotal)1 ' which assumes a straight line through the mean. The measure of fit, R2, is a value between 0 and 1, and may be thought of as the fraction of the variance explained by the model, and is derived from 1- (SSmodel/SStotal). When R2 = 0, then the model is no better than a straight line through the data, and when R2 = 1, the model describes the data perfectly (R2 is normally used to denote the fit of nonlinear regression, whereas r2 is used to denote linear regression). Statistical packages perform these dose-response curves to give the best possible fit (GraphPad, or within Excel using the Solver Add-in (Microsoft)). Further, the same protocol is used for fitting the resulting action spectra with a photopigment template. For fitting the IRCs, each curve was fitted in Excel (Microsoft), using Solver add-in (Frontline systems) to minimise the sum-of-squares for each IRC, as described above. 2 R values were >0.90 for all IRCs. Pupil constriction was spectrally tuned, peaking (,max) at 476nm. 309 420 nm 460 nm 1.0 1.0 C 0.8 0.8 C C 0. 0.5 g 0.5 411 c°e 0.3 12 0.3 0.0 0.0 8 10 12 14 16 18 8 10 12 14 16 18 Log Photons Log Photons 481 nm 500nm 1.0 1.0 H41) 0.8 2 0.8 C a 0.5 O0. o.5 U) 1k43) 0.3 a) 0.3 CL 0.0 0.0 18 8 10 12 14 16 18 8 10 12 14 16 Log Photons Log Photons 515 nm 540 nm 1.0 1.0 •cn 0.8 101 3 0.8 C 0 0' 5 a 0,5 0. 0 0.3 0.0 0.0 10 12 14 16 18 8 10 12 14 16 18 Log Photons Log Photons 560 nm 580 nm 1.0 1.0 in 0.8 u) 0.8 C C O 0.5 a 0.5 0.3 a) 0.3 0 Or 00 0.0 10 12 14 16 18 8 10 12 14 16 18 Log Photons Log Photons Figure 9.2 Analysis of pupil response to light in five subjects with CRD or RP. Response is amplitude of pupil constriction. Irradiance response curves (IRCs) were conducted at eight wavelengths for both eyes in all subjects. Responses are plotted as percent maximum response obtained. IRCs were fitted with a four-parameter sigmoid function, with R2 values >0.90 in all cases. 310 420 nm 460 nm 1 0 1 .0 a) 0.8 0 8 C C 0.50 a 0 .5 3 • 0 3 0 0 0.0 10 12 14 16 8 10 12 14 16 Log Photons Log Photons 481 nm 500 nm 1 0 8 O. 0 5 a 0 3 O 0 S 10 1 2 1 4 16 18 Log Photons Log Photons 515 nm 540 rim 1.0 1.0 - • • 9 .8 .8 - 5 8 O 5 - 0.5 O. 0 so 0.3 0.3 - 00 0.0 10 12 1 4 16 18 Log Photons Log Photons 560 rim 580 nm 1 .0 1.0 a+ 0 8 . 0 8 C a o 9.s - O.O 9.s O. a a 0 3 0.3 42 & •• 0 0 n 1 0.0 a 10 1 2 1 4 1 6 1 8 8 1 0 1 2 1 4 1 6 1 8 Log Photons Log Photons Figure 9.3 IRCs for 2nd constriction of the right eye in subject SC. 311 420 nm 46 9 nm 1 0 • 0 8 C 0.5 0-3 0 0 O 0 10 12 16 16 18 a 10 12 14 16 1a Log Photons Log Photons 481 nm 590 nm a 0.8 C 0.0 0.3 0 0 10 12 1i 16 18 Log Photons Log Photons 515 nm 540 nm Log Photons Log Photons 560 nm 580 nm 1.0 1 0 O 8 0 8 0.5 O 0.5 N O 3 A 0 .3 0 .0 O 9 a 10 12 1l 16 18 10 12 14 16 18 Log Photons Log Photons Figure 9.4 IRCs for the second constriction of the left eye in subject SC. 9.6 Calculating the Relative Sensitivity for the Action Spectrum Sensitivity is based upon photon flux required to elicit the same response with different wavelengths, so a low photon flux is indicative of high sensitivity to the wavelength concerned, and vice versa (reciprocal relationship between sensitivity and photon flux required to elicit a 50% response). Relative sensitivity was calculating by dividing the lowest photon flux obtained by the photon flux obtained at each wavelength (using a linear scale and not log units). Values in such calculations are theoretically 312 between 0 and 1, with the wavelength of maximum sensitivity having a value of 1 and higher photon fluxes possessing values below 1. The action spectrum was then plotted as relative sensitivity (y-axis) against wavelength (x-axis). These are shown for subject SC in figures 9.5, 9.6, 9.7 and 9.8. As responses reached half maximum at only three wavelengths, relative sensitivity based upon IR50 would result in extrapolation beyond the data range. Consequently, relative sensitivity was determined from responses obtained from an equivalent photon flux in the dynamic range of the IRC (i.e. above background noise and before saturation), as described previously [Provencio and Foster 1995]. Data were analysed based upon responses to 1 x 1015 photons/cm2/s though comparable responses were obtained in the range 1 x 1014 to 1 x 1016 photons/cm2/s, additionally enabling evaluation of pupil responses in bright and dim light. 9.7 Template Fitting Vitamin A based photopigments have a characteristic absorption profile; when expressed on a frequency scale (which is for wavenumber rather than wavelength), the shape of this profile is constant [Dartnall 1953]. While the wavelength of maximum sensitivity (Xmax) may vary, the shape of the absorption spectrum is constant. Indeed a single curve defined by the Xmax will provide the entire absorption profile of a vitamin A based photopigment. An A2 form of retinal in amphibia, teleosts and some reptiles led to a second slightly differing visual pigment template, which is proportionally broader [Bridges 1967]. The template first produced by Dartnall in 1953 has been modified many times [Partridge and deGrip 1991; Govardovskii et al 2000]. Wavenumber (traditionally measured per cm2) may be simply calculated by 1/X x 106 — e.g. 500 nm is equivalent to 20,000 waves / cm. The wavenumber separation from the peak for both Al and A2 visual pigments are provided in the literature [Dartnall 1953; Bridges 1967; Partridge and deGrip 1991; Govardovskii et al 2000; Peirson et al 2005; Foster et al 2007]. 313 When calculating the fit of a visual pigment template, calculations were conducted on the logarithm of relative sensitivity. This is due to the aforementioned calculations for R2. As the SS2model is calculated from the squared difference between data and model, a difference of 10% near the wavelength of maximum sensitivity contributes proportionally more to the SS2model than a 10% difference at any other wavelength. When calculated on a logarithmic scale, a 10% difference however gives the same contribution to the SS2modet at all wavelengths. This therefore ensures that all data points contribute to the best fit of the photopigment template. Fitting a photopigment template relied upon the same non-linear protocol as fitting the IRC. The only variable here is the Xmax, while the SS2model was minimised by changing this parameter to provide the best fit, with a resulting R2 value as above. Absorbance spectra for human visual pigments were generated using the Govardovskii template [Govardovskii et al 2000] based upon the published Xmax values [Bowmaker and Dartnall 1980] of rod (498 nm), blue cone (420 nm), green cone (534 nm) and red cone (563 nm). Free-fitting of the visual pigment template was conducted using the least squares method described above, using log transformed sensitivity and template data to prevent peak-fitting bias (shown in figures 9.5 and 9.6 for subject SC). Action spectrum of pupil responses provided a poor fit to rod and cone photopigments (rod R2=0.35, SW cone, MW cone, LW cone R2<0). An optimum fit to the pupil response to light was provided by an opsin/vitamin A-based template with Amax 476nm (R2=0.892). Note: Data shown were not corrected for preretinal lens absorption. When this standardized correction was applied, the kmax moved from 476 nm to 480 nm [Stockman et al 1999], although the R2 decreased to 0.70 due to the differential filtering effects at 420 nm. 314 A ctiort Spectra (LO G 0.00 -0.50 a -1.00 -1 50 -2.00 350 400 450 500 553 600 653 700 avelength Inm '2 0 1 0 - '5 0 0 7 5 - 4 8 1 d•••••••••• 5 0 0 1 0 50 a 5 1 5 0 .2 5 5 4 0 5 80 0 00 10 12 14 16 18 E P F Log photon flux Figure 9.5 Action spectrum template-fitting: subject SC, 2nd constriction right eye . Xmax 488.7 nm; R2 = 0.534. Res, Respo = Response; Log Res = Log response (prevents peak filtering bias). Templates use wave number. A ctitan Spectra (LO G 0.00 -0.50 0 Ce. -1 00 0 -1.50 -2.00 350 400 450 500 550 600 630 W avelength (nm 4 2 0 1 .00 4 e 0 0 75 0 4 8 1 0. 0.50 411••••• 5 0 0 5 1 5 0.25 5 4 0 500 12 14 16 18 E P F Log photon flux Figure 9.6 Subject SC. Action spectra for left eye (2nd constriction). X,-„,„ 496.7; R2 = 0.110. Res or Respo = Response. 315 9.8 Assumptions in Action Spectra The construction of action spectra is dependent upon four underlying assumptions which if not present can be a cause of inaccuracy: univariance, reciprocity, screening pigments, and quantum efficiency. The principle of univariance should hold i.e. the mechanism of activity should be the same at all wavelengths. This is evident in the data if the IRCs present a similar slope (k value). A fixed slope is used for the different wavelengths. However, some laboratories use a variable-slope, with the trade-off of an extra degree of freedom for a fixed parameter, provided the sum of squares is the same [Motulsky and Christapoulos 2003]. The IRCs were plotted using both techniques, and the sum of squares was the same, and there was no significant difference (475nm fixed slope and 480nm variable slope) (figures 9.7 and 9.8). 1% 400 450 500 550 600 650 700 Figure 9.7 Action spectrum for SC fitted using a variable slope applied to first constrictions. Black curve = SC; Blue curve= S-cone (420 nm); Grey curve = rod (498 nm); Green cone = 534 nm; Red cone = 563 nm. 316 100% 10% - % 400 450 500 550 600 650 700 Figure 9.8 Action spectrum for SC fitted using a fixed slope applied to first constrictions. Black curve = SC; Blue curve= S-cone (420 nm); Grey curve = rod (498 nm); Green cone = 534 nm; Red cone = 563 nm A standard fixed shape sigmoid curve-fit was used, with a fixed minimum IRC of 0, a maximum of 1 and slope fixed at the mean of 2. Were IRCs to have demonstrated dramatically different slopes, it might have indicated a different mechanism of action. IRCs were deemed univariant and showed a high statistical fit to a visual pigment nomogram (R2 = 0.892, compared to R2 = 0.35 for rod and R2 < 0.01 for all three cone classes), suggesting that pupil constriction driven by a single photopigment. Data shown were not corrected for preretinal lens absorption [Stockman et al 1999]. When this correction was applied, the Xi. moved from 476 nm to 480 nm. According to the law of reciprocity photochemical action depends on the product of light intensity and duration of exposure. Hence exposure to low light intensities for longer durations yields the same effect as higher intensities for shorter durations. Reciprocity could break down when saturating stimuli 317 are applied, as addition of extra photons cannot initiate any further reactions. This potential effect is overcome by using the EC50. Other factors are more difficult to control. Photopigments may be screened by other biological pigments, and thereby affect the spectral sensitivity of the biological response to light. The action spectrum of a light-dependent response, e.g. vision or pupil reaction, must match the absorbance spectrum of the photopigment it is acting on taking these into consideration [Lythgoe 1979]. There are many effects, e.g. increased scattering of short-wavelength light (Mie scatter, detailed in chapter six), absorption by haemoglobin in overlying tissues, and by other biological pigments such as melanin [Hartwig and van Teen 1979]. Finally it is important to bear in mind that by the law of photochemical equivalence every atom, ion or molecule involved in a reaction must absorb a photon for action spectra to be valid. The proportion of molecules that undergo a specific reaction following absorption of a photon is termed the quantum efficiency of the reaction. For example absorption of light in opsin-based photopigments has been measured as 0.67, or, 2/3 of the photons absorbed, which lead to an isomerisation from the //-cis to the all-trans form [Dartnall 1968; Le Grand 1970]. 9.9 Conclusions The pupillary spectral profile of both the five subjects collectively and the model subject SC individually show intact but diminished pupillary reactions to light with a double pupil constriction with peak spectral sensitivity using irradiance response curves in the region of 480nm. The pupil action spectrum could be fitted by a single vitamin-A based photopigment template consistent with melanopsin-driven photoreception from intact pRGCs and absence of rod and cone function. 318 9.10 Special Acknowledgement Part of the data analysis reported in this chapter was undertaken using computer software in conjunction with Dr. S. Peirson of the University of Oxford, formerly of the Department of Visual Neuroscience at Imperial College London. 319 Chapter Ten Circadian Experiments: Actigraphy, Light Exposure and Questionnaire Sleep Studies 10.1 Assessment of Sleep: Actigraphy, Polysomnography and CCTV 10.1.1 Methods of Assessing Sleep Actigraphs are devices placed on the wrist (occasionally the ankle or trunk) to record physical movement [Ancoli-Israel et al 2003]. Till the late twentieth century polysomnography (PSG) was used very commonly by sleep researchers and clinical psychiatrists to study sleep/wake patterns [Ancoli-Israel et al 2003]. PSG involves multiple recording techniques (typically 11 channels). These provide electroencephalogram (EEG) recording, EOG recording to determine rapid eye movement (REM) sleep associated with dreams, air flow using a thermistor or pressure probe that fits inside the nostrils, electroymogram (EMG) recording of chin (lower facial) muscles and leg movements, an electrocardiogram (ECG/ECG) recording heart rate and rhythm, oxygen saturation with a pulse oximeter fitted over finger top or ear lobe, chest and abdominal wall movement using a series of belts wrapped round the subject. Actigraphy was introduced over 20 years ago [McPartland et al 1976; Colburn et al 1976]. In 1995 the American Sleep Disorders Association (now the American Academy of Sleep Medicine (AASM)) reporting on a meta-analysis of studies using actigraphy concluded that it is a "cost- effective method for assessing specific sleep disorders" [Sadeh et al 1995]. The Standards of Practice Committee recommended that actigraphy might be useful in the study of circadian rhythm (sleep/wake) disorders [Sadeh et al 1995]. Increased demand led the commercial sector to respond with higher quality devices that can be strapped round the wrist like a wristwatch. Most devices are accelerometers and have memory cells 320 allowing weeks of continuous recording [Ancoli-Israel 2003]. Movement is sampled several times a second and stored for later downloading to a computer for display of actigraphs and data analysis [Ancoli-Israel 2003]. Newer actigraphs can combine recording movement with other functions, for example measurements of ambient light exposure using photocells facing away from the body of the subject. Continuous activity/rest and light exposure monitoring using actigraphy is often hence felt to be the best way of observing objectively and quantitatively the subject's unrestricted behaviour unobtrusively. CCTV monitoring is too intrusive and not very accurate to measure small movements; diaries are subjective and not as accurate as actigraphy. 10.1.2 Actigraphy and Circadian Rhythms Human wrist activity displays a distinct circadian rhythm of rest-activity [Ancoli-Israel et al 2003]. These reflect rhythmicity of neuro-musculo- skeletal activity that is synchronised with the light-dark (day-night) cycle and thus governed by complex neuro-endocrine factors promoting periods of physical activity / inactivity in sleep, waking and non-sleeping periods of rest [Ancoli-Israel et al 2003]. The circadian period of this sleep/wake rhythm accurately predicts the period of PSG-defined sleep/wake rhythm [Pollack et al 2001]. Actigraphy has shown that infants have a circadian rest-activity pattern independent of temperature [Scher et al 1998]. In a non- electric society in Papua New Guinea one-week actigraphy of subjects combined with light exposure measurement revealed synchronisation between the rest-activity and light-dark cycles [Siegmund et al 1998]. In the longest reported group of such experiments, Binkley measured wrist activity of one woman continuously for an entire year, and found entrained circadian rest-activity cycles, changes in sleep length synchronised with the menstrual cycle, and annual phase changes presumed due to summer and winter-time clock changes [Binkley 1994]. Shift work, jet lag, space flight in astronauts, illness, insomnia and its treatment, psychiatric illness and its 321 treatment, and experimentally induced sleep changes, can all change sleep patterns as identified by actigraphy [Ancoli-Israel et al 2003]. A review by Ancoli-Israel et al. for the American Academy of Sleep Medicine have concluded that actigraphy is the "standard marker of circadian rhythms," and that it "highly appropriate" and "more practicable than PSG" in assessing circadian rhythms, which has also been confirmed independently in the review of Sadeh and Acebo [Sadeh and Acebo 2002; Ancoli-Israel et al 2003]. 10.2 Methods - Actigraphy 10.2.1 Altered Social and Medical Condition of Subject Subject SC was approached with a view to her possible participation in continuous wrist actigraphy combined with light intensity measurement for up to several weeks. During other experiments presented in this thesis SC had been an independent 87-year old retired social worker. By the commencement of this final series of experiments she had unfortunately moved to a residential nursing home attached to a hospital, owing to difficulty in coping exacerbated by the onset of hypothyroidism and a fall. Thyroid replacement therapy had been instituted under medical supervision by the commencement of and for the duration of the experiments; she was biochemically euthyroid on thyroid function tests. There was no history of any specific muscular disease or condition affecting the wrist joints or carpal bones other than expected osteoporosis for her age. The subject had a full mini-mental test score, orientated in time, place and person. SC was once again eager to participate, and this time in an especially long series of experiments which I informed her as the principal investigator would carry with them certain limitations on her new and already limited lifestyle. The matter of SC's decline in general health, her increased frailty and mild depression since her hypothyroidism and change in social circumstances was made known by me to the supervising academic staff. This was as I felt 322 that SC might now be making an excessive amount of self-sacrifice to participate in a scientific study, albeit out of her deep-seated desire to aid medical research, and perhaps compounded by increased social vulnerability. It was decided that subject SC was still a potential candidate for study. Informed consent was deliberately obtained from three different individuals (two members of the academic staff in addition to myself as the principal investigator), and from two institutions, including one unfamiliar to the subject (Imperial College London, the University of London as before, and now additionally City University). This was with the express purpose of underlining to SC that the research team would fully understand if she refused further testing. Myself aside, the other two investigators were deliberately chosen as they were either less known to the subject or totally unknown to her so as to reduce the possibility of SC feeling any obligation to participate in further experiments conducted by an investigator familiar to her from previous experiments and who she might not therefore wish to disappoint. The subject gave informed consent to proceed when asked by all three individuals who all counselled her separately. 10.2.2 Recording of Movement and Light Exposure At the insistence of all investigators SC was deliberately given an additional three weeks to settle into her new surroundings prior to commencing the experiments, which ran through late summer into early autumn 2005. Her new environment facilitated co-option of trained nurses into helping supervise compliance with the experiments and thus exceptional concordance with instructions. With the aid of nurses an Actiwatch-L® device was attached to the subject's non-dominant wrist [Cambridge Neurotechnology Ltd 2002]. This was for simultaneous monitoring of wrist muscle and joint movement and integrated also a light sensor; sampling and storing of movement and light exposure 323 information was at one minute epochs. The device is shaped like a wristwatch. It needs to be removed and replaced with a fresh memory every 22 days for data readings from its 64 kB memory. Recording was for a total of 21 consecutive days. Data was subsequently downloaded from the Actiwatch-L® to a personal computer using the Actiwatch Software, version 5.25 (CNT). During this period in the summer of 2005 subject SC was conditionally restricted from wetting the actigraph-light sensor or removing it for washing, bathing and sleeping wherever reasonably possible save in the interests of hygiene or discomfort (which did not occur). Nurses aided the subject in bathing, washing and monitoring use of the sensors at all times including sleep, with the intention of keeping the sensors attached and dry. Monitoring by the subject and nursing staff, combined with communication of instructions to nursing staff to ensure proper usage, led to continuous wear without malfunction and data recording for the entire 21 day recording period. Theoretically, surrounding noise and other periodic interruptions can entrain activity independent of light environment. This was avoided as the subject remained self-caring — there were no timed visits to attend to her by caregivers, washing and meal times were not fixed, the subject self- medicated and threw were no regular noise patterns such as ward rounds. The conditions for rest activity were those of an observational study, hence there were no restrictions on sleep, its timing, or naps. The subject did not use an alarm clock or wake-up call to raise her from sleep. Bed time and morning wake times were self-selected by the subject. All activities of daily living were self-selected by the patient down to self-medicating and recorded by her in the diary we gave her. Daily administration of levothyroxine was given at lunch time and tamoxifen was given regularly in the evening at approximately 21:00. Lactulose was occasionally administered and Ciproxine (ciprofloxacin) was only given on the first two days of recordings. Along with the actigraphy, the subject's carers noted the 324 time when the watch had been removed and when medication was administered. The purpose of lighting was not to entrain the subject like in laboratory animals. The purpose of recording light exposure was to observe daily routine behaviour, which includes exposure to daylight and ambient room light. Lights were on regularly being turned off at night, and the subject slept in the dark at night with no lights on. Quantitative data on light levels for photoentrainment are shown later in this chapter. 10.3 Actigraphy and Light Exposure Data Analysis 10.3.1 Overview — Results of Actigraphy The period length, peak and amplitude of the rhythms of rest-activity and light-dark were determined by applying the periodogram analysis after Sokolove and Bushell [Sokolove and Bushell 1978]. Peak and amplitude of the rhythms of rest-activity and light-dark were determined by applying the Cosinor analysis. Calculations and graphical output were carried out in 'El Temps', version 1.87 (A. Diez-Noguera, Universitat de Barcelona) [Wulff et al 2006]. The resulting period and circadian phase of rest-activity rhythm and light exposure pattern indicated a normally entrained sleep-wake cycle typical of an octogenarian (figure 10.1 and figure 10.2). Actogram shows that the subject went to sleep between 21-00 and 22-00 on most days. There is somewhat disturbed night-time activity which is progressive over the three- week period. The subject wakes between 08-00 and 09-00 regularly. There are some daytime naps, which are not consistent over the three-week period. Light exposure is regular, beginning between 06-00 and 08-00 and ending between 19-00 and 20-00. Chi-square periodogram analysis showed a 24.08 h for both variables. 325 Actogram File Display Options Graphs Markers Age: 87 Sex: F Light On Identity: Hely_1 23-Aug-2005 flue) Inteival: 1.00 07:00 Start date: Stall Time: 00:00 jJJ j Light Off 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 I 23:45 , ' II Activity Scale I i i Tue Wed 1 'ILI 1,141 I 500 jJji Zero WeifTlui k..JJ_.111.Li L AIL ....1 .J.... 1 I ILL. . ../i 1 ki.0 ..._11I.114..11 d I 0 ii ThtuFri —. ..1 11„,-1,1 , Li .0„ L 0 .1.1.J.1,L,1,0.„1 Li .,,hu. . L Light scale .....1 i IL 8 FifSat ..,,Li. ., „...... „, . i 111...,.il „Li Start Day I 1 SatSun ill.. ii „Li 1, „1.11,4„,,..1.., .1. Li, ,ii, ., aL,,,..ii,„11,1 .iiLii.i ....,, Li Start Time I 00:00 I SunMon i i 1 i ..IJ J. h .. IL LAII.1.11IL d1d Redraw Mon:Tue 1, .1 A illi,,,.1ALI 1.,..11J A giii, .1. Tue:Wed k...... 1. I.,.. 11.....1.11.h. . WeitThu L.., Lik,JL. ,.,,i. lit [.. iil, .i. .1111..6kuilh. Thu Ft i ii ,1 , , I, ,, I, 11,11i ,J &Jill. klailll .LL,ii Fri:Sat Ali .1 id . iii . I JAI Ad IAA Sat Sun d . •d1,11., LIU 1.11111a111 IA JI Sun Mon Mon:Tue . .. 1.01111 11.41 Tue:Wed _ .141,01.11 LAIL Thufri —11 FriSat ...Li 4, la SatStin siLL L.LiLLL,d1 i.,1111 1 SunMon I J11iIiiiL411111.,AL1. MoirTue I 11,111 Figure 10.1 21-Day Single Plot Actogram. Activity is shown by black bars, radiant light exposure in yellow. This data is double-plotted in figure 10.2 326 ACTIVITY LIGHT EXPOSURE clock time (h) Clock lime (h) G 12 18 24 30 36 42 48 0 18 24 30 36 42 48 s • day no 0 2F ive t • • 0 C) Consecu NDND NDND Figure 10.2 21-Day Double Plot Actogram (left) and radiant light exposure (right). Entrained period of the circadian rest-activity rhythm and light exposure of SC recorded over 21 days. Data are double-plotted to discern patterns of activity with consecutive days plotted next and beneath each 327 other (D = day, N = night). Rest-activity data (activity = black, rest = blank) are derived from actigraphically recorded sleep-wake behaviour. Illumination shows time of exposure to light (light in lux = black). Activity and Light Exposure (luxes) are shown in black. One 24-hour period of time is shown in each right-hand column and repeated in the subsequent left- hand column (for ease of viewing activity in the early hours). Recordings start at the top of right side of the charts. 10.3.2 Rhythm Analysis: Statistical Data Raw activity values were analysed directly from the actograms and light exposure graphs. Detailed rhythm analysis is presented in figure 10.3. For both sets of data (light exposure and actigraphy), basic calculations of variance (denoted by Actogram), periodogram analysis (after Lomb-Scargle and Leiden University), and Cosinor analysis including variability, are shown [Lomb 1976; Scargle 1982]. Periodograms are concerned with displaying events in biological periods or frequencies and their analysis (using peak detection of the highest value). They give measures of power (intensity) of rhythms in a time series for a range of periods and the corresponding statistical significance [Van Dongen et al 1999a and b]. Neither traditional periodograms based on data folding within the time domain (e.g. the chi-squared periodogram) or classical power spectral analysis (using a Fast Fourier Transform (FFT) computer algorithm) are ideally suited to study biological rhythms as biological rhythm data is usually unequally spaced [Van Dongen et al 1999a and b]. For such calculations a Lomb-Scargle Periodogram may be used, applied from astronomical time series analysis [Lomb 1976, Scargle 1982]. Irregular data distribution is overcome by using regression to a harmonic (cosinor analysis), followed by application of a least squares method. This periodogram has become the modern method of displaying and analysing biological rhythms [Ferraz-Mello 1981; Ruf 1999; Van Dongen et al 1997; 1999a and b; Wulff et al 2007]. 328 •••,-1,-0,,,,#) 1B 7 7•,• ) f., I , •.••••••1117 ,1711,1. ”ri 14.114 iu • I; • Try-rum",Inry-r+r-10,1,11r-r .00(1 1409 IiCL isr.p Atha* (Actegaurn) Light (Aidegown) n=30220 'ratline rroor-124. noir_3011_5. Acinrily (Periodogram -813) tight (Peeedognorn -SB) FigheOtorairm sat Highest maim sat T=1445.08 nin ( 8.144) T=14451110 nal( 1271) signal_ above 4.43 sign& ahem 4.43 Peak T y 4g. P T y 4g_ t 1445.00 0.04 it 1445_00 12.71 Acfmnly (Ooskier) Light (Ootsimar) PeriodT=1445.0 mini(2ll' 5 nil) Period T=1445_0 Iola (24h 5 main) Pooutosm 9227113: 892.201-853.331 Araulose 708.421754_342-778502 (1&22 h (05% 0:14:52-15:53) CI240 h (95% Ct 1234 h — 12:58 h) Figure 10.3 Statistical analysis of circadian photoentrainment. For activity/rest rhythms to be evaluated as "entrained", the period length of the rhythm has to be approximately 24 hours with a power at or above the level of significance. From the periodogram (boxed below the actograms in the above diagram), this has been proven for rest/activity and light for the subject. lA illustrates the actograms and periodograms of activity and 1B for light. The graphs are the results of the rhythm analysis: (Actogram): number of values (sampling 1 min) maximum/minimum levels of activity and light. (Periodogram): period length (T in min) of 24 h 5 min a period which represents one activity-rest cycle, level of significance (above 4.43) 329 and peak of significance [Sokolove and Bushell 1978]. (Cosinor): time of the peak in activity and light (acrophase) with 95% confidence interval (95% CI). The extent of rhythmic variation of both light intensity and actigraphy are shown graphically. Both show very little variability (straight regression lines superimposed upon the entrainment lines) except relatively small spikes at and around 14-45, suspected from observers to be due to afternoon napping. For cosinor analysis a cosine curve with a period of 24 hours was fitted to the data by the least-squares method. This is the most popular method of data analysis for circadian actigraphy and light exposure [Satlin et al 1991; Binkley 1994; Middleton et al 1996 and 1997; Glod et al 1997; Sakurai and Sasaki 1998; Youngstedt et al 2001; Ancoli- Israel et al 2003]. To study rest-activity, light-dark cycle, and their correlation, the following parameters were measured for each 24 h cycle for the six weeks: acrophase (acrofase) (time of peak activity), amplitude (peak-to-nadir difference) and mesor (mean) of the fitted curve. Cosinor analysis is shown in figure 10.2 to ascertain peak of activity (left) or light exposure (right) respectively. Simple averaging of peaks is not used very often as cosinor analysis permits fitting of activity or levels of light exposure of the subject to a cosine curve template, one 24 h cycle corresponding to one wavelength (peak to peak or trough to trough). Acrophase (peak time of activity in rhythm, acrofase) is found to be 15-00 to 16-00 hours for activity (on the left side in figure 10.2) and between 12-00 midday and 13-00 for light exposure (on the right side in figure 10.2). This is very similar to results expected from young people. Mesor is the mean of a rhythmic function similar to the mean of activity but unlike a simple average fitted to the cosine curve for activity (on the left side in figure 10.2) or light exposure (on the right side in figure 10.2) in a 24 h cycle (range 22.2-29.95). Mean cyclical activity (Mesor) is 23.7992 which is low (normal approximately 100-200). In order to detect the period length of the sleep-wake cycles the Lomb- Scargle periodogram was hence applied to the actigraphic and light 330 exposure time series using the integrated package for chronobiology analysis El Temps. The power of each of these periodicities, which is a measure of their contribution to the variance of the time series as a function of frequency, confirmed the significance of the contribution of approximately 24 hour periodicities to the time series (significance threshold of p < 0.05). In figure 10.2 the periodogram for sleep-wake activity (left) and for light exposure (right) is shown produced using the El Temps software. Variance (%V) is calculated by fitting to a model (the archetypal periodogram compiled from miscellaneous studies) to which subject data has been fitted. Variance is determined by how effective the fit to the model is, signified in how closely the data points for the subject's entrained time fits the straight line of significance, or conversely how randomly. The y-axis is entrained time in minutes per day. The peak (commonest) period of rest-activity cycle in the subject over the three-week period is found to be 1445 minutes, or 24.08 h; this length of cycle is considered entrained. Further, as it is the same for sleep-wake activity (left) and light exposure (right) activity and light exposure are positively correlated. Artefacts can occur using certain methods of study. Measurements of nadir of temperature and melatonin peak, because of environmental artificial light exposure, have the capacity to put back the circadian pacemaker. However in the subject under consideration as she slept early between 21-00 and 22- 00 this effect is limited. Further, since this subject did not wake to open curtains till approximately 09-00, any residual effect is likely to have been eliminated. The effect of artificial light can compound measurements of circadian photoentrainment cycles however if people wake much before sunrise while it is still dark and put artificial lights on. The latter can delay the circadian pacemaker and the sleep-wake period prolongs — it will not be free-running, but it may still possibly be a 24 h period, e.g. rhythm of 23.9. Ambient light showed a clear day/night difference with lights on during the day-time and no or little lights at night-time, with corresponding change in 331 activity and is shown in figure 10.4. The simultaneous recording of activity and light hence show that light exposure coincided with the time of activity. Quantitative information on light intensity and its relationship to activity are shown in detail in table 10.1 and figure 10.4. In viewing this it is worth considering that 100 lux of ambient room light (that in the mid-afternoon in the study) has 50% of the circadian timing phase shifting effect of light 10 times as bright [Zeitzer et al 2005]. It was not necessary in this study to resort to further analyses that may sometimes be indicated if rhythmicity and /or an association between activity and light exposure are still suspected despite lack of apparent correlation [Ancoli-Israel et al 2003]. Activity and ambient Light 180.00 160.00 x) 140.00 lu ht ( 120.00 Lig / 100.00 Activity ts) n 80.00 Light cou ( 60.00 ity iv t 40.00 Ac 20.00 0.00 — P. rb 43 A cb Figure 10.4 Quantitative analysis of hourly activity and ambient light cross three weeks. Distribution of average activity (blue) and ambient light (yellow) derived from wrist-actigraphy. 332 Ambient light (lux) 1 1 7 40 01:00 2866 08:00 Activity (count/1 min) 1 1 7 225 00:00 1498 09:00 Ambient light (lux) 2 8 7 11 02:00 1509 08:00 Activity (count/1 min) 2 8 7 787 04:00 2118 13:00 Ambient light (lux) 3 15 7 2 01:00 1036 08:00 Activity (count/1 min) 3 15 7 856 02:00 2961 12:00 Ambient light (lux) 1-3 1 21 18 01:20 1804 08:00 Activity (count/1 min) 1-3 1 21 623 02:00 2192 11:20 Table 10.1 Non-parametric circadian rhythm analysis of activity and ambient light per week and across three weeks (bottom). L5 = weekly levels of average activity/light during the 5 least active or least brightest hours per day; M10 = weekly levels of average activity/light during the 10 most active or brightest hours per day; L5onset = onset of the L5 period; M10 onset = onset of the M10 period.2 The analysis shows a clear day-night difference in levels of activity and ambient light. Lowest levels of average activity and ambient light occur at night (623 counts and 18 lux, respectively), starting 1:00 to 2:00. Levels of highest average activity and ambient light occur during the day-time (2192 counts and 1804 lux, respectively), with lights on starting around 8:00 and the period of highest activity starting around 11:20. 10.3.3 Conclusions from Actigraphy Subject SC was found to have fragmented sleep and a weak circadian rhythm in rest-activity, but despite this remained entrained and showed a diurnal circadian pattern in rest/activity. Period for both rest-activity and light-dark cycle was 24h 5 min. Subject SC is likely to have possessed a functionally intact pRGC pathway mediating circadian photoentrainment. 333 10.4 Questionnaire Studies The assessment of sleep-wake cycles in the setting of residential care homes shows that actigraphy can be conclusive in the assessment of sleep-wake cycles, and can be supported by validated questionnaires used as instruments to study a subject [Hoekert et al 2006]. Limitations are subjects with restless legs syndrome and obstructive sleep apnoea; neither condition affected the subject. Self-rating of the preferential timing of sleep and activity was carried out with the validated morningness-eveningness questionnaire developed by Home and Ostberg [Home and Ostberg 1976]. This questionnaire comprises 19 questions which are scored and summed to generate a total score that ranges between 16 and 86, with high scores indicating early morning chronotype and low scores indicating late evening chronotype [Appendix III]. The Pittsburgh Sleep Quality Index is the other standard validated sleep questionnaire [Buysse et al 1989]. It was administered for the evaluation of self-rated sleep quality and sleep disturbances referring to the days and nights of the previous month. The I 9-item questionnaire comprises seven component scores, including: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and day-time dysfunction [Appendix IV]. The seven component scores are summed to generate a global score that ranges from 0 to 21, with higher scores indicating worse sleep. A global score greater than 5 identifies a clinically significant sleep disturbance with 89% sensitivity and 86.5% specificity [Buysse et al. 1989]. Nurses assisted with the completion of the questionnaires. The subject scored 69 on the morningness-eveningness chronotype questionnaire, which classified her as "moderately morning type" borderline to "definitive morning type". This chronotype score was consistent with her habitual sleep timing between 21:00 h and 05:00 h according to the Pittsburgh Sleep Quality Index (PSQI). The global score on the PSQI was 5, which identified her as borderline "good sleeper". These are all consistent with preserved circadian rhythms. 334 10.5 Implications of Circadian Experiments Subject SC, despite being rodless and coneless, continued to photoentrain with intact circadian sleep-activity rhythms. This suggests human pRGCs are intrinsically photosensitive, and in addition to having a role in initiating pupillary reflexes to light shown in chapter nine, also mediate circadian rhythms. 10.6 Special Acknowledgement The actigraphic investigation reported in this chapter was undertaken using computer software in conjunction with Dr. K. Wulff of the University of Oxford, formerly of the Department of Visual Neuroscience at Imperial College London. 335 Chapter Eleven Visual Perception Originating from a Novel Photoreceptor 11.1 Rationale The recent finding in non-human primates that the pRGCs project to the dorsal lateral geniculate nucleus (dLGN) prompted an exploration into the possibility that these photoreceptors might contribute to an individual's ability to detect or even experience some awareness of light [Dacey et al 2005]. The finding of circadian entrainment and a PLR in a functionally blind subject (SC) provided a subject in whom this hypothesis might be tested, and whether the pRGCs in SC might indeed provide some awareness of light, triggering either a conscious response or, in the absence of a conscious percept, a response analogous to blindsight [Cowey 2004]. 11.2 Methods for Testing Visual Perception Blindsight was discussed in chapter one. Experimentally it is in effect an above-probability ability to detect the presence of a light stimulus in the absence of a conscious percept, and is assessed using a forced-choice testing paradigm [Schwartz 2004]. In forced choice methodology a stimulus is in one of two temporal (or sometimes spatial) intervals and it is mandatory for the subject to decide which interval contains the stimulus (hence the test is called the 2-alternate forced choice test (2-AFC)). Since the subject may well claim they cannot see as they are 'blind', subjects need to be informed before testing that they must give a response even if they think they 'see' nothing. Even then some subjects may raise objections to the test during an experiment. However Weiskrantz showed in the 1980s that performance at locating the stimulus is above chance in blindsight [Weiskrantz 1986] — 336 `nonconscious' vision due to activation of extrastriate cortex [Goebel et al 2001]. In testing SC stimulus duration was 10 s at an intensity of 1.45 x 1016 photons/cm2/s. This was based upon the highest light output that could be radiometrically matched across all stimuli. Pre-determined coin toss determined which of the two temporal intervals that the stimulus would appear in. After some initial hesitancy on being asked to report the presence of visual stimuli of which she was nominally unaware, SC was fully able to comply with the requirement of a forced-choice testing paradigm. Twenty trials were conducted at each wavelength for each eye (420 nm, 460 nm, 481nm, 500 nm, 515 nm, 540 nm, 560 nm and 580 nm). Data were analysed using the Sign test with a probability of level of 0.05 (corresponding to correctly identifying the stimulus interval in 15/20 trials) deemed to be indicative of statistically significant detection. 11.3 Responses to Forced-Choice Testing Under these conditions SC was correctly able to identify the interval in which a 481nm test light appeared (p < 0.05), but was unable (p > 0.05) to detect light at longer or shorter wavelengths (420 nm, 460 nm, 500 nm, 515 nm, 540 nm, 560 nm and 580 nm) (figure 11.1). Furthermore, SC reported that the presence of the detectable (481nm) stimuli elicited in her a percept she described as "brightness". The failure to detect light at other wavelengths close to 480 nm is readily explained by the experimental protocol. Based upon the action spectrum for pupillary constriction, responses at 460 nm would be 1.6x less sensitive and responses at 500 nm 2.9x less sensitive. The lower sensitivity at these wavelengths precludes the detection of graded responses i.e. within these stimulus intensities an 'all or nothing' visual response. 337 100 80 60 o Right eye 40 - • Left eye 20 - 420 460 481 500 506 515 540 560 580 Wavelength (nm) 100 >, 90 80 (.) 70 c 4- Left Eye 0 4c 60 as 2. , cy, —0—Right Eye • .E 50 ti 0 a) Level of 40 cn Significance U .a) 30 17 cn 20 O 10 0 420 460 481 500 506 515 540 560 580 Stimulus Wavelength nm Figure 11.1 Psychophysical testing of SC revealed conscious perception of light at 481nm (p < 0.05), but inability (p > 0.05) to detect light at longer or shorter wavelengths (420 nm, 460 nm, 500 nm, 515 nm, 540 nm, 560 nm 338 and 580 nm). These results mirror the spectrally tuned response of her pupil, and suggest that SC's detection and awareness of light also arise from pRGCs. Each histogram represents percent correct responses out of 20 trials, for both left and right eyes (360 trials in total). The differences between eyes, which is very small in this subject, may be because of asymmetric trans-synaptic atrophy of ganglion cells, some degree of which is inevitable in chronic rod and cone dystrophies. 11.4 Rationale in Testing for Perception of Colour Colour vision is classically regarded as requiring photoreception by a minimum of two photopigments with different action spectra [Schwartz 2004]. Recently, as discussed in chapter one it has been proposed that pRGCs may alter their peak spectral sensitivity such that two peak spectral sensitivities exist [Rea et al 2005]. Subject SC was studied from the perspective of whether pRGCs may influence colour vision. 11.5 Methods for Testing for Perception of Colour A 2-AFC procedure was employed exactly as described earlier in this chapter, but with the question being asked of the subject which temporal interval they "saw colour" in (as opposed previously to "brightness"). The subject was then asked to name the colour. 11.6 Discrimination of Brightness and Hue The subject could only discriminate hue for the 481 nm stimulus (range 420 to 580 nm). For all other stimuli (achromatic and chromatic) the response was perception of brightness only i.e. she reported 'white' light. The final set of stimuli were presented late in the afternoon on 21 June 2005. For stimuli of equal energy the responses were 100% perception of blue in the left eye, and 30% with the right eye. Oddly the subject said the right eye blue was more intense. Somewhat surprisingly she said that the pale blue 339 was more intense in the right eye though she gave considerably more correct responses to stimuli being presented in the left eye. The subject described this as a "pale blue" hue for both eyes, on direct questioning compared by her visual memory (to when she had full sight in childhood) to a sky blue as opposed to a deep sea blue. This is shown in figure 11.2. 11.7 Comments on Findings A subject who was apparently pathologically depleted of rods and cones yet with largely intact pRGCs maintained a rudimentary form of brightness perception. There is also some suggestion that colour vision may be possible for blue light. The latter could be a consequence of some residual cone function, though there is no other evidence anatomically or physiologically for it as shown in chapters nine, ten and eleven. The possibility of residual memory accounting for the observed phenomenon cannot be fully discounted - subjects losing vision in early childhood can many years later recollect colours, especially blue which seems associated with brightness more than red, despite only having one class of photoreceptor [Schwartz 2004] Alternatively the mechanism is more complex, and possibly related to spectral opponency within pRGCs. An effect more pronounced in many non-human mammals is that they not uncommonly respond to brightness changes in daylight in the environment with marked behavioural changes, such as hibernation. The pRGC photoreceptor is principally a brightness detector [Foster 2005]. Visual perception of the brightness of light is potentially advantageous to survival in the wild. Photopic and scotopic photoreceptor systems (respectively cones and rods) are well-described. An exciting possibility is that mesopic (twilight and dusk) vision may also have a novel class of photoreceptor specialised for its photoreception. 340 120 100 e Hu 80 ing n r ce is D MI Left Eye 60 ❑ Right Eye ions t erva 40 f Obs o % 20 0 420 460 481 500 506 515 540 560 580 Stimulus Wavelength nm %of Subject's Observations Discerning Hue E Left Eye 0 0 0 ❑ Right Eye 0 0 0 0 Stimulus Wavelength nm Figure 11.2 Discrimination of hue as a function of wavelength in SC. Where hue was discriminated in all cases it was described as "pale blue" or "sky blue". 341 Chapter Twelve Effect of Darkness on Non-Rod Non-Cone Mediated Pupillary Constriction 12.1 Reactions to Darkness — Overview Several pupillary reactions to darkness have been described [Loewenfeld and Lowenstein 1999]. Rather than signifying simply passive dilation of the pupil owing to the light stimulus being turned off, reactions to darkness are a finely tuned neurological process distinct from spontaneous pupillary oscillations in darkness. The latter include fatigue waves described in chapter one, which reflect moment-to-moment fluctuation in autonomic arousal. They also include other 'autonomic' fluctuations in the pupil that occur with breathing and cardiovascular activity. These latter occur after processing by the reticular activating system which lies not very far from the midbrain pupillary centres noted in chapter one, in which context the reticular activating system is responding to intermittent stimulation of pulmonary afferent fibres, vascular pressure receptors, and firing of central respiratory neurones e.g. pupillary changes during Cheynes-Stokes breathing, Meyer waves, and the Traube-Hering type noted by Borgdoff [Loewenfeld and Lowenstein 1999]. 12.2 The Effects of Immediate Light-Off After full retinal bleaching the pupil dilates forcefully initially, and then more slowly, and after a few seconds begins to waver. The pupil thereafter undergoes small amplitude oscillations. The mechanism for pupillary dilation at a cellular level is thought to involve activation of retinal "off" ganglion cells. This is corroborated by physiological studies simultaneously recording human pupillograms and electroretinograms showing retinal "off" discharges on withdrawal of the activating light; further, at sub-saturation levels the amplitude of retinal 'off activity was directly related to both the 342 intensity of the adapting light and the amplitude of the subsequent pupillary dilation [Kawabata 1960]. 12.3 Anatomy of the Light-Off Pathway light-off involves the following pathways. i. Parasympathetic feedback loops starting in the retina, projecting to the midbrain via superior brachium and thence both inhibition of the constrictor pupillae muscle and activation of the dilator pupillae muscle (chapter one). "Off' discharges have been recorded from single units in the pretectum of cats, some of which are tonically inhibited by light and show sustained firing in darkness [Nisida and Okada 1959; Cavaggioni and Zampollo 1968; Smith et al 1968]. The firing rate of these "off' action potentials, as well as the cessation of "on" firing activity, accompany pupil dilation after a latent period of 200 to 300 ms. The latency is expected owing to the slowness of smooth muscle cells [Aidley 1989]. ii. Ganglion cell axons which at some stage project in the superior brachium also conduct retinal "off' discharges via the suprachismatic nucleus (SCN) in the thalamus. The latter are a group of cells found in the anterior diencephalon just above the chiasm where the first anatomical bifurcation of ganglion cells or their collaterals takes place [Mai 1978]. From the SCN tracts including the anterior and posterior accessory optic tracts and fibres in the median forebrain bundle pass to the mesencephalon. As noted in chapter one this entire system had till recently been called the accessory optic system [Mai 1978; Loewenfeld and Lowenstein 1999]. The fibres have an important role in light-dark circadian cycles as destruction of the SCN abolishes certain neuroendocrine functions totally, at least in part owing to the effect on the pineal gland, while bilateral interruption of the connection from 343 diencephalon to mesencephlaon abolishes the influence of light on hormonal cycles, which cease to be photoentraining and become free-running. These pathways may well connect to the sympathetic nervous system, which is intimately associated with pupillary movement. iii. Key areas of the latter are known to involve the following. Sympathetic nerve impulses have been recorded from the central stump of divided cervical sympathetic chain in response to light-off [Passatore and Pettorossi 1976; Passatore 1976; Nishino et al 1976; Passatore et al 1977]. Incidentally they have also been recorded from the long ciliary nerves [Nisida et al 1960; Nisida and Okada 1960; Okada et al 1960a and b]. These impulses were all inhibited when the light was turned on, and released from inhibition with light off, and corresponded to pupillary dilation. The extensive sympathetic contribution to pupillary dilation in the dark is underlined by observations that the noradrenaline content of the iris and other structures supplied by the superior sympathetic ganglion doubles after the animals have been kept in darkness for two weeks, in comparison with the heart which is supplied by lower sympathetic ganglia [Owman et al 1982]. 12.4 Paradoxical Pupil Constriction in Darkness A paradoxical transient pupillary constriction with light-off has been recorded in some diseases, though it is an inconstant feature even then, reducing its importance to diagnosis [Loewenfeld and Lowenstein 1999; Kardon 2003]. It has been reported in some subjects with congenital achromatopsia (absence of cones) and also in subjects with congenital stationary night blindness (absence of rods) [Price et al 1985]. The 'reflex' at least has the appearance of a standard pupillary contraction. This phenomenon has also been reported in some cases of optic neuritis and 344 other conditions [Price et al 1985]. Paradoxical dark reaction defies ready functional correlation with known retinal cells. 12.5 The Delayed Effects of Darkness According to some authorities there is a delayed pupil constriction at approximately 5 minutes of darkness-onset. Crawford noted this in 1936 as a "curious anomaly" [Crawford 1936]. It is an inconstant phenomenon found in three out of eleven subjects and shows variation in onset and offset. Loewenfeld and Lowenstein are of the opinion that the constriction represents fatigue [Loewenfeld and Lowenstein 1999]. Several other groups have however also observed the phenomenon and advanced possible explanations, that it is either related to the cone-rod break of retinal dark adaptation, or the appearance of "Eigengrau", which are spontaneous sensations of grey, dim clouds apparently drifting in the visual field during dark adaptation [Kadlecova and Peleska 1955; Bouma 1965] 12.6 The Black Period and Afterimage Contraction Following light-off of a very bright and prolonged adapting light stimulus of a few minutes, the pupil redilates, and then within a few seconds starts to reconstrict, becoming almost as small as in the light. In the succeeding minutes the pupils are highly unstable, fluctuating extensively about the mean constricted size in the dark, with oscillations becoming weaker after a few minutes, and after this wavering the pupil redilates within these oscillations. These changes were noted by Campbell at the Cambridge Physiological Laboratory in 1960 [Loewenfeld and Lowenstein 1999]. Lowenstein and Loewenfeld showed that these correlate psychophysically with a black period in vision following light-off, followed by afterimage onset coinciding with pupil reconstriction in darkness [Loewenfeld and Lowenstein 1999]. In 1940 Craick had demonstrated that the photochemical processes needed to produce an afterimage can take place in an eye made blind by pressure while the adapting light is applied, and that light-off 345 combined with removal of pressure results in an afterimage in the absence of perception of the initial image from pressure-on [Craik 1940]. Lowenstein and Loewenfeld have suggested that the phenomenon in this context occurs due to a transient blocking by the intense light of retinal conduction of the impulses arising from photoreceptor stimulation and then their release by light-off to create an afterimage associated with pupil constriction. Newsome in 1971 showed the black period varied linearly with the duration and intensity of the adapting light [Newsome 1971]. 12.7 The Dark Pause The dark pause (also called darkness reaction, dark reflex or dark flash) is the interruption of a bright light by a brief dark pause (e.g. 1 s) followed by the re-introduction of the light stimulus [Lowenstein and Givner 1943; Lowenstein 1954; Thompson 1966; Loewenfeld and Lowenstein 1999]. The normal dark pause reaction is triphasic (the reaction is described with the pupil already constricted from initial light-on). With light-off there is swift dark dilation followed upon light-on by re-constriction to below the previous level, followed by re-dilation and a return to the previous (pre- darkness) level of pupillary constriction. As the intensity of the adapting light is diminished the first phase of the reaction is lost as the pupil is barely constricted to begin with, and there is no recordable dilation during the second in the dark pause, though when the light comes on again the pupil constricts as described in the triphasic reaction. The dark pause can thus describe a biphasic reaction if the stimulating light intensity is low. Since it reflects predictable changes in levels of retinal area, by comparing the ERG with the pupillary responses, the dark pause reaction is found to represent activity in retinal "off" ganglion cells (switching off of retinal-off cells), which are known to be capable of inhibiting neurones in the pretectal region and the visceral area of the oculomotor nucleas in the midbrain. This may hence cause pupil dilation by reducing parasympathetic outflow and causing relaxation in the sphincter pupillae and relaxation of the dilator, 346 while simultaneously there is increased sympathetic outflow in to darkness, causing additional activation of the dilator pupillae and relaxation of the constrictor [Lowenstein and Loewenfeld 1969; Loewenfeld and Lowenstein 1999] . Correspondingly, light-on activity has been postulated to be associated with the switching off of retinal ganglion "off" cells and the turning on of "on" cells. The fact that the retinal "on" discharge is much more vigorous than the "off" neuronal burst and pupil wave would explain why the dark pause reaction loses its first dilation component when the stimulus is of low intensity or the adapting light dim - no effective "on" signal is generated. It would also explain the same finding reported when the dark pause is short - pupil dilation evoked by "off" ganglion cells is overpowered by the contraction evoked by the intense "on" discharge that follows it before the pupil has time to respond to the off signal and dilate. The constriction- redilation followed by constriction to a greater level is due to the fact that the retina gains sufficient sensitivity with dark adaptation to respond to the re-appearing light, as it would to an increment in light intensity. The dark pause has been suggested as a possible important parameter that is highly sensitive to afferent damage [Loewenfeld and Lowenstein 1999]. However from a practical standpoint observations of responses in darkness are rarely used in clinical practice, as the experimental conditions of darkness necessitate a dedicated room that can be plunged into total darkness and quite likely the use of infra-red light and goggles to see. 12.8 The Dark Response in a Human Lacking Rods and Cones Melanopsin-expressing ganglion cells are responsible for circadian photoentrainment, including in a human lacking rods and cones (chapter ten). A response in darkness is hence plausible. Since the pRGC was shown in chapter nine to also effect the PLR in such subjects, a pupillary response to darkness may also be mediated by the receptor. The early and delayed 347 effects of darkness on the pupil are studied in a subject lacking rods and cones. 12.9 Materials and Methods A model subject previously shown to lack rods and cones [Subject SC, documented previously in the preceding chapters in detail] was recruited for study over a period separated by seven months from the previous tests. After the first attendance the laboratory had partly moved location in the Western Eye Hospital to a set-up where a temporary dark-room was used for this experiment. An intense bright achromatic light of energy log 16.5 log photons/cm2/s and duration ls was presented monocularly as described previously in chapter nine and in the general experimental methods. Pupil recording in darkness for durations longer than 15 or so seconds was not possible even with modification of the P_Scan software. Non-infrared video pupillometry was not feasible. So a new P_Scan recording was begun every 15 s. A total of fourteen recordings were made, seven for each eye, which were then averaged. 12.10 Darkness Constriction without Rods and Cones A pupillary constriction was observed to begin at about 3 s following light- off, with maximal amplitude at approximately 4900 s, with full redilation before 8 s (figure 12.1). 348 0.9 0.8 m 0.7 m n io 0.6 t ic 0.5 tr 0.4 0.3 - il Cons 0.2 Pup 0.1 0 T-FTT O c§) ,(§) OO eo ,z0, ()C),\,z 6) 6) 6) 6) NO l),() (0c) 00 CO ?>0 0 Time s Figure 12.1 Pupil reactions in a human subject lacking rods and cones (each data point represents 14 readings, seven runs from each eye). Stimulus was turned off at Os. Pupil constriction is darkness occurs between 3 s and 8 s. Recording was for 15 s. Confidence intervals represent mean +/- 2 SD. 12.11 Interpretation of Results A marked darkness constriction that does not fully fit previous known patterns was elicited. Artefacts or false positives, for example from accommodation, are most unlikely given the extent of visual perception in the subject. The purpose of this response in terms of ocular function is unknown since there is no light, but it may be a sign of a signal to the circadian system of prolonged darkness i.e. retinal-off ganglion cell activity. Recently Gamlin et al found a pupil constriction in functionally rodless coneless Macaque monkeys subject to a sustained light stimulus at approximately the time of light-off following the sustained stimulus [Gamlin et al 2007]. However it was unclear if this was in fact a response to light (the sustained light stimulus) rather than light-off (darkness) or conceivably both. Other darkness reactions may mask the response, so it may not be of physiological significance outside of disease states. 349 Alternatively, it may however account for the 'paradoxical' pupil constriction seen with certain retinal conditions, especially those that affect rod and cone function severely in which activity in the novel photoreceptor could be being 'unveiled'. The relationship to the 'black period' and afterimage formation is also possible, both in terms of time frame, and as seen in figure 12.1, small oscillations are noted after the main constriction. The constriction to darkness seen in a human lacking rod and cone photoreception is a novel finding which can plausibly be attributed to the activity of pRGCs. A related effect, the pupil constriction associated with afterimage formation, and any putative association between it and pRGC function, is beyond the scope of this thesis in which fundamental study of the pRGC and its role in sight are a culmination. 350 Afterword Overview of Thesis Findings The specific findings in this thesis of special relevance to the detection of and/or effects of orbital, vitreo-retinal and refractive disorders have already been considered in detail within the respective chapters. In these pupillometry, complemented where possible by electrophysiology, was found to enhance the detection of orbital and vitreo-retinal disease in selected cases. In general inter-eye comparison of responses, measurement of pupil evasion or escape, and selective use of chromatic stimuli, were the most useful strategies. As is frequent with theses written by clinicians including surgeons such myself, this thesis represents predominantly two conceptual areas: pure basic scientific work, and clinical scientific work lying on the very cusp of basic science with clinical science. In this thesis the latter bridge to direct clinical application in the following fields — orbital/oculoplastic work and vitreoretinal (surgical/medical) work, both of which dominate many chapters, with significant work in chapter six certainly additionally relevant in the context of pupillometry to cataract/refractive disorders and surgery, and to a lesser extent glaucoma (here in the context of its retinal effects and perimetry). There is also relevance to general ophthalmology and hence the wider medical/surgical field. Much of the basic scientific work, while of fundamental importance to the subject fields as a whole, emerges from simultaneous scientific study also within the thesis of the above clinical areas. Hence there is no large clinical trial (associated with subjecting hundreds of patients to each test, for example) as such — smaller, tightly-designed trials conducted by the author sufficed and had the requisite power as the scientific methodology and underlying concepts which prompted the various investigations in this thesis were sufficiently good. This gave time to perform 351 more experiments on related topics instead within the thesis, broadening it and increasing its depth in key areas The purpose of the remainder of this section is providing an overview of those findings of this thesis that have developed the work as a whole and hence bound each chapter to the next. This began in this thesis with work using subretinal membranes that was used to calculate a novel ratio (pupil evasion ratio) representing an analogous parameter to the pupil escape ratio, but measured following a brief rather than sustained light stimulus. Inter- eye pupil evasion and escape were found to be greater in eyes with subretinal membranes. This relationship was only found with achromatic light and long wavelength stimuli. It could not be elicited with medium and short wavelength light. These findings were investigated using other disease models for macula dysfunction, as well as for ganglion cell dysfunction, to investigate whether these findings would be specific to only subretinal membranes. Pupil evasion ratio was significantly altered by Macula-Off retinal detachments, and hence seems to be a generic sign of macula dysfunction, as well as by orbital lesions especially those involving a significant inflammatory component and hence to ganglion cell function also. Subsequent additional study of orbital disease suggested selectively preserved sensitivity of the pupil responses to short wavelength light after light adaptation compared with dark adaptation. Given that this did not occur with achromatic stimuli, relative preservation of a group of pathways active after light adaptation and responsive to short wavelengths was inferred. It was hypothesised that all the aforementioned findings could be explained by non-classic photoreception (non-rod non-cone). Further, this hypothesis would, ultimately, require testing through direct study of a human disease model causing loss of rods and cones. 352 Such subjects might additionally have cataracts which could skew any findings. It was found that a monochromatic light stimulus was associated with less chromatic aberration in subjects with dense cataracts using blue versus achromatic pupil perimetry. However in advanced cataract the effect of scatter was sufficient, especially if the cataract was severe enough to cause index myopia, to obscure all advantages offered in perimetry with chromatic stimuli, including the better resolution that a monochromatic stimulus enabled due to reducing the effect of chromatic aberration. Myopia of its own had no such effect. To explain the chromatic variation in PEVR the possibility of this being due to a previously uncharacterised yet fundamental component in the pupil waveform was investigated using a sustained stimulus. This found that shorter wavelengths of light elicited a double pupil constriction in healthy subjects a few seconds after the initial constriction and between approximately 4 s and 7s following stimulus onset. This component was specific for short wavelength light stimuli and was suggested as an explanation for the spectral variation of PEVR that was found early in this thesis. The spectral sensitivity of the second constriction suggested it was an effect of a novel photoreceptor (peak spectral sensitivity approximately 480nm). To directly investigate the putative receptor subjects with advanced classic photoreceptor dystrophies, intact ganglion cells, but without significant cataract were studied. Actigraphy and questionnaire study over several weeks of a subject suggested that photoentrainment with intact circadian sleep-activity rhythms was present. Ten eyes from all the five subjects recruited in this section suggested preservation of double pupil constrictions with action spectral analysis revealing a peak spectral sensitivity in the region of 480 nm. The responses could be fitted by a template indicating melanopsin as the candidate photopigment driving responses. This was consistent with activity from photoreceptive ganglion cells. 353 A subject without rods and cones was psychophysically tested for perception. The subject maintained a rudimentary form of brightness perception elicitable with very high intensity stimuli of approximately 480 nm. There is also some suggestion that colour vision may be possible, but if so due to a poorly understood and complex mechanism. The novel receptor seems to have a role in brightness detection and in mesopic vision — the centuries-old model of a duplex retina with rigorously segregated light (cone-mediated) and dark (rod-mediated) functions is hence suggested to be an oversimplification. Further, studies of responses in this subject in darkness showed a pupillary constriction beginning at about 3 s following light-off, with maximal amplitude at approximately 5 s, with full redilation before 8 s. This suggests the novel receptor in humans also has some role in darkness and is consistent with its primary role as a brightness detector capable of detecting a wide range of environmental light. The existence of a novel photoreceptor in humans is fundamentally interesting. That it is suggested in this thesis to have a role in vision, which was hitherto undetermined in any organism, potentially opens another exciting avenue in subsequent research here and in related fields. Of note, what role does the receptor have in disease? Can it be a target for medical therapy using surgery, drugs or gene therapies? Further elaboration of the receptor's function using the pupillometric assay for a double constriction would be a reasonable immediate application of the results of this thesis, both clinically to the disorders already studied herein and initially extending this at a basic scientific level to other candidate diseases. Clinically application of several results is worthwhile in orbital/oculoplastic, retinal (surgical/medical) and general ophthalmic practice, the latter in assessing visual potential and outcome after surgery for refractive disorders and for media opacities. 354 Appendix I Example of a Patient Information Sheet PATIENT INFORMATION SHEET Pupillometry at the Western Eye Hospital What is it about? This study will examine the effect of eye disease on the movement of the pupil. The pupil is the round dark space in the coloured part of your eye (the iris). We are investigating the link between certain eye diseases and the pupil's behaviour, so that in future doctors can better diagnose and treat patients with these diseases. We shall measure the size of the pupil using a pupillometer. This is non-invasive technique which means that there is no object that physically enters the body or the eye. All it involves is viewing a flashing light and recording on video the pupil's movement. What will be asked of you? We ask you to come to the Western Eye Hospital for one session of pupillometry. You should abstain from taking excessive amounts of caffeine such as tea, coffee or fizzy drinks for 24 hours before the session. Once there, you will undertake a sight and reading test, of the kind you will routinely have in any visit to an ophthalmologist or optician. You will then be asked to place your chin on a rest and stare at a target on a screen in front of you. A light will repeatedly appear and disappear whilst a video camera records your pupil's movement. The light is not very bright, and is not startling. The testing usually takes no more than half an hour. However, you should set aside 1 hour just in case we need to repeat the test which is unlikely. At the end of the session you will be asked if you want to know what the results show. Your test result is confidential, and will not be disclosed to any person or organisation outside of the study investigators. We will pay for your travelling expenses to and from the hospital if you do not otherwise have a hospital appointment on that day. Are there any risks to you? There are no known risks from undergoing pupillometry. The test will not in any way reduce or change the quality of care that you receive from the doctors in the hospital or from your GP. Are there any benefits to you? Through giving up a small amount of your time and taking part in this study you will be helping doctors to discover if pupillometry can be used to better diagnose and treat eye disease. This test has the potential to help people with many types of eye problem, including those causing blindness. It may also help with the treatment that you receive in the future should the test be found to be better than those currently used. Entry to the study is entirely voluntary and should you wish to withdraw at any time you may do so and will not be asked to give any explanation. The study is funded by Imperial College London who also provide legal indemnity. Contact Should you at anytime wish to contact the principal study investigator, Mr Farhan Zaidi, please do not hesitate. He is a medical doctor and postgraduate research student based at the Department of Ophthalmology, Imperial College London. Telephone 020 8383 3693. 355 Appendix II The Lens Opacities Classification III [LOCSIII] The LOCSIII standard was originally designed for standardising photographic images of cataract, which should be matched for size to the photographic image being used. NO1 to N06 and NC1 to NC6 are standards for, respectively, nuclear opalescence and colour; Cl to C5 for cortical cataract; and P1 to P5 for posterior subcapsular cataract. Rarer morphs of cataract are not covered by LOCSIII. In practice LOCSIII is widely used for grading cataract on direct observation (slit lamp) compared with the photographs. 356 Appendix III The `Morningness-Eveningness' (`Lark/Owl') Questionnaire (First Page) [In the interests of clarity after the front page shown on this side the rest of the layout of this questionnaire is in a spatially condensed format] SUBJECT CODE: DATE: INSTRUCTIONS a) Please read each question very carefully before answering. b) Answer all questions. c) Answer questions in numerical order. d) Each question should be answered independently of others. Do NOT go back and check your answers. e) For some questions, you are required to respond by placing a cross alongside your answer. In such cases, select ONE answer only. f) Please answer each question as honestly as possible. Both your answers and results will be kept in strict confidence. Question 1 Considering your own feelings, at what time would you get up if you were entirely free to plan your day? 04:00- 05:00 05:00— 07:00 Time: 07:00 — 09:00 09:00 — 11:00 11:00 — 13:00 or later Question 2 QUESTION Considering only your own feelings, at what time would you go to bed if you were entirely free to plan your day? 19:00 — 20:00 20:00 — 22:00 22:00 — 00:00 Time: 00:00 — 02:00 02:00 — 04:00 or later 357 Question 3 If there is a specific time you have to get up in the morning, to what extent are you dependent on being woken up by an alarm clock? a. Not at all dependent b. Slightly dependent c. Fairly dependent d. Very dependent Question 4 Assuming adequate environmental conditions, how easy do you find getting up in the morning? a. Not at all easy b. Slightly easy c. Fairly easy d. Very easy Question 5 How alert do you feel during the first half hour after having woken in the morning? a. Not at all alert b. Slightly alert c. Fairly alert d. Very alert Question 6 How is your appetite during the first half hour after having woken in the morning? a. Not at all good b. Slightly good c. Fairly good d. Very good 358 Question 7 During the first half hour after having woken in the morning, how tired do you feel? a. Very tired [ ] b. Slightly tired [ ] c. Fairly refreshed [ [ d. Very refreshed [ [ Question 8 When you have no commitments the next day, at what time do you go to bed compared to your usual bedtime? a. Seldom or never later [ ] b. Less than one hour later [ ] c. 1-2 hours later [ ] d. More than 2 hours later [ ] Question 9 You have decided to engage in some physical exercise. A friend suggests that you do this one hour twice a week and the best time for him is between 07:00 and 08:00h. Bearing in mind nothing else but your own inclinations, how do you think you would perform? a. Would be on good form b. Would be on reasonable form c. Would find it difficult d. Would find it very difficult Question 10 At what time in the evening do you feel tired and in need of sleep? 19:00 — 20:00 20:00 — 22:00 Time: 22:00 — 00:00 00:00 — 02:00 02:00 — 04:00 or later 359 Question 11 You wish to be at your peak for a test which you know is going to be mentally exhausting and lasting for two hours. You are entirely free to plan your day, when would you do this task? a. 08:00 — 10:00 b. 11:00 — 13:00 c. 15:00 — 17:00 d. 19:00 — 21:00 Question 12 If you went to bed at 23:00h at what level of tiredness would you be? a. Not at all tired b. A little tired c. Fairly tired d. Very tired Question 13 For some reason you have gone to bed several hours later than usual, but there is no need to get up at any particular time the next morning. Will you: a. Wake up at the usual time and not go back to sleep b. Wake up at the usual time and doze c. Wake up at the usual time and go back to sleep 1 d. Wake up later than usual 1 Question 14 One night you have to remain awake between 04:00 and 06:OOh. You have no commitments the next day. Which suits you best: a. Not to go to bed until 06:OOh b. Nap before 04:OOh and sleep after 06:OOh c. Sleep before 04:OOh and nap after 06:OOh d. Sleep before 04:OOh and remain awake after 06:OOh 360 Question 15 You have to do two hours physical work. When would you prefer to do it between: a. 08:00 — 10:00 b. 11:00 — 13:00 c. 15:00 — 17:00 d. 19:00 — 21:00 Question 16 You have decided to engage in some physical exercise. A friend suggests that you do this between 22:00 and 23:00h twice a week. How do you think you would perform: a. Would be on good form b. Would be on reasonable form c. Would find it difficult d. Would find it very difficult Question 17 Suppose that you can choose your own work hours, but had to work five hours in the day. Which five consecutive hours would you choose: 20:00 — 01:00 05:00 — 10:00 Hours: 08:00 — 13:00 ..... 11:00 — 16:00 14:00 — 19:00 17:00 — 22:00 361 Question 18 At what time of day do you feel your best? 20:00 - 01:00 05:00 — 10:00 Time: 08:00 — 13:00 ..... 11:00— 16:00 ..... 14:00— 19:00 ..... 17:00 — 22:00 Question 19 One hears of "morning" and "evening" types. Which do you consider yourself to be? a. Morning type h. More morning than evening c. More evening than morning d. Evening type 362 SCORING for student population Definitely morning type 70-86 Moderately morning type 59-69 Neither type 42-58 Moderately evening type 31-41 Definitely evening type 16-30 [Home and Ostberg 1976] SCORING for non-student population (> 21 years) Definitely morning type 70-86 Moderately morning type 64-69 Neither type 53-63 Moderately evening type 31-52 Definitely evening type 16-30 [Taillard et al 2004] 363 Appendix IV The Pittsburgh Questionnaire [In the interests of clarity after the front page shown on this side the rest of the layout of this questionnaire is in a condensed format] Pittsburgh Sleep Quality Index Subject code: STUDY: Date: Instructions: The following questions relate to your usual sleep habits during the past month only. Your answers should indicate the most accurate reply for the majority of days and nights in the past month. Please answer all the questions. 1) During the past month, when have you usually gone to bed at night? Usual bed time 2) During the past month, how long (in minutes) has it usually take you to fall asleep each night? Number of minutes 3) During the past month, when have you usually got up in the morning? Usual getting up time 4) During the past month, how many hours of actual sleep did you get at night? (This may be different from the number of hours spent in bed.) Hours of sleep per night For each of the remaining questions, check the one best response. Please answer all questions. 364 5) During the past month, how often have you had trouble sleeping because you..... a) Cannot get to sleep within 30 minutes Not during the Less than Once or twice Three or more past month once a week a week times b) Wake up in the middle of the night or early morning Not during the Less than Once or twice Three or more past month once a week a week times a week c) Have to get up to use the bathroom Not during the Less than Once or twice Three or more past month once a week a week times a week d) Cannot breathe comfortably Not during the Less than Once or twice Three or more past month once a week a week times a week e) Cough or snore loudly Not during the Less than Once or twice Three or more past month once a week a week times a week f) Feel too cold Not during the Less than Once or twice Three or more past month once a week a week times a week g) Feel too hot Not during the Less than Once or twice Three or more past month once a week a week times a week h) Had bad dreams Not during the Less than Once or twice Three or more past month once a week a week times a week 365 i) Have pain Not during the Less than Once or twice Three or more past month once a week a week times a week j) Other reason(s), please describe How often during the past month have you had trouble sleeping because of this? Not during the Less than Once or twice Three or more past month once a week a week times a week 6) During the past month, how would you rate your sleep quality overall? Very good Fairly good Fairly bad Very bad 7) During the past month, how often have you taken medicine (prescribed or "over the counter") to help you sleep? Not during the Less than Once or twice Three or more past month once a week a week times a week 8) During the past month, how often have you had trouble staying awake while driving, eating meals, or engaging in social activity? Not during the Less than Once or twice Three or more past month once a week a week times a week 366 9) During the past month, how much of a problem has it been for you to show enthusiasm to get things done? No problem at all Only a very slight problem Somewhat of a problem A very big problem 10) Do you have a bed partner or roommate? No bed partner or roommate? Partner/roommate in other room Partner in same room, but not same bed Partner in same bed If you have a roommate or bed partner, ask him/her how often in the past month you have had)... a) Loud snoring Not during the Less than Once or twice Three or more past month once a week a week times a week b) Long pauses between breaths while asleep Not during the Less than Once or twice Three or more past month once a week a week times a week c) Legs twitching or jerking while you sleep Not during the Less than Once or twice Three or more past month once a week a week times a week d) Episodes of disorientation or confusion during sleep? Not during the Less than Once or twice Three or more past month once a week a week times a week e) Other restlessness while you sleep; please describe Not during the Less than Once or twice Three or more past month once a week a week times a week 367 References Abrams GW, Azen SP, McCuen BW 2nd, Flynn HW Jr, Lai MY, Ryan SJ (1997). Vitrectomy with silicone oil or long-acting gas in eyes with severe proliferative vitreoretinopathy: results of additional and long-term follow- up. Silicone Study report 11. Archives of Ophthalmology 1997; 115: 335- 344. Adrian W (2003). Spectral sensitivity of the pupillary system. Clinical & Experimental Optometry 2003; 86: 235-238. Adrian WK (1973). Uber den Zusammenlang der Spektralen Empfindlichkeit des Auges und der Pupillenkonstriktion. In: Dodt E, Schrader KE, Bergmann JF (editors). Die normale und gestorte Pupillen — bewegung. Symposium der DOG 10-12 Mtirz 1972 Mi,inchen. Verlag, 1973. [German] Agarwal LP (1997). B.M. Chatterjee's Handbook of Ophthalmology. 6th edition. New Delhi: CBS Publishers & Distributors, 1997. Aidley DJ (1989). The Physiology of Excitable Cells. 3rd edition. Cambridge. Cambridge University Press, 1989. Akamatsu T, Enomoto H, Tabuchi A (1979). Pupillographic studies on the pupillary constriction during 6 seconds of light exposure. Nippon Ganka Gakkai Zasshi 1979; 83: 1530-1537. [Japanese] Al Deek M (2004). Ibn Al-Haitham: Master of Optics, Mathematics, Physics and Medicine. Al Shindagah 2004; November-December. Al Majlisi MB (1994). Bihar al Anwar. Translation. vol. 111, pp. 57-151. London, 1994. Originally published in the I600s. [Cited by Mirza KL, 1997]. 368 Alexandridis E, Dodt E ( 1967). The light reflexes of the pupil width of a rod- monochromatic female. Albrecht von Graefes Archiv fur klinische und experimentelle Ophthalmologic. Albrecht von Graefe's archive for clinical and experimental ophthalmology 1967; 173: 153-61. Original article in German. [Cited by Keenleyside MS (1989)]. Alexandridis E, Koeppe ER (1969). The spectral sensitivity of retinal photoreceptors conducting the pupillary light reflex in humans. Albrecht von Graefes Archiv fur klinische und experimentelle Ophthalmologic. Albrecht von Graefe's archive for clinical and experimental ophthalmology 1969; 177: 136- 151. Original article in German. [Cited by Keenleyside MS (1989)]. Alexandridis E, Leendertz JA, Barbur JL (1992). Methods for studying the behaviour of the pupil. Journal of Psychophysiology 1992; 5: 223-239. Alexandridis E (1985). The Pupil. Translated by Telger T. New York: Springer- Verlag, 1985. Alfoldi P, Franken P, Tobler I, Borbely AA (1991). Short light-dark cycles influence sleep stages and EEG power spectra in the rat. Behavioural Brain Research 1991; 43: 125-31. Erratum in: Behavioural Brain Research 1991; 44: 233. Ali Ibn Isa (10th century). Memorandum Book of a Tenth-Century Ophthalmologist. Translated by Wood CA. Chicago: Northwestern University, 1936. Originally published in the tenth century. Alpern and Campbell (1962). The spectral sensitivity of the consensual light reflex. Journal of Physiology 164; 478-507. Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP (2003). The role of actigraphy in the study of sleep and circadian rhythms. Sleep 2003; 26: 342-392. 369 Anderson D, Guerin C (1986). Morphological recovery in the reattached retina. Investigative Ophthalmology and Visual Science 1986; 27: 168-183. Anderson SJ, Holliday IE, Harding GF (1999). Assessment of cortical dysfunction in human strabismic amblyopia using magnetoencephalography (MEG). Vision Research 1999; 39: 1723-1738. Apple DJ, Rabb MF (1998). Ocular Pathology. Clinical Applications and Self- Assessment. 5th edition. St Louis: Mosby, 1998. Arden GB (1978). The importance of measuring contrast sensitivity in cases of visual disturbance. The British Journal of Ophthalmology 1978; 62: 198-209. Arden G B, Barrada A, Kelsey JH (1962). New clinical test of retinal function based upon the standing potential of eye. British Journal of Ophthalmology 1962; 46: 449-467. Arendt J, Deacon S (1997). Treatment of circadian rhythm disorders — melatonin. Chronobiology International 1997; 14: 185-204. Argamaso SM, Froehlich AC, McCall MA, Nevo E, Provencio I, Foster RG (1995). Photopigments and circadian systems of vertebrates. Biophysical Chemistry 1995; 56: 3-11. Attebo K, Mitchell P, Cumming R, Smith W, Jolly N, Sparkes R (1998). Prevalence and causes of amblyopia in an adult population. Ophthalmology 1998; 105: 154-159. Audouze J, Israel G (1985) (editors). The Cambridge Atlas of Astronomy; Cambridge: Cambridge University Press, 1985. 370 Baker FH (1963). Pupillary response to double-pulse stimulation; a study of nonlinearity in the human pupil system. Journal of the Optical Society of America 1963; 53: 1430-1436. Bannister R (1985). Brain's Clinical Neurology. 6th edition. Oxford: Oxford University Press, 1985. Barbur JL, Forsyth PM (1986). Can the pupil response be used as a measure of the visual input associated with the geniculo-striate pathway? Clinical Vision Science; 1: 107-111. Barbur JL, Thomson WD (1987). Pupil response as an objective measure of visual acuity. Ophthalmic & Physiological Optics 1987; 7: 425-429. Barbur JL, Dunn GM, Wilson JA (1986). The perception of moving comets at high retinal illuminance levels: a rod-cone interaction effect. Biological Cybernetics 1986; 55: 145-158. Barbur JL, Forsyth PM, Thomson WD (1987). A new system for simultaneous measurement of pupil size and two-dimensional eye movements. Clinical Vision Science 1987; 2: 131-142. Barbur JL, Harlow AJ, Sahraie A (1992). Pupillary responses to stimulus structure, colour and movement. Ophthalmic & Physiological Optics 1992; 12: 137-141. Barbur JL, Keenleyside MS, Thomson WD (1989) Investigation of central visual processing by means of pupillometry. In: Kulikowski JJ, Dickenson CM, Murray IJ (editors.) Seeing Colour and Contour. Oxford: Pergammon Press, 1989. Barbur JL, Ruddock KH, Waterfield VA (1979). A colour-dependent abnormality in human visual detection of stimulus motion and spatial structure. Neuroscience Letters 1979; 15: 307-312. 371 Barbur JL, Weiskrantz L, Harlow JA (1999). The unseen color aftereffect of an unseen stimulus: insight from blindsight into mechanisms of color afterimages. Proceedings of the National Academy of Sciences of the United States of America 1999; 96: 11637-11641. Barbur JL, Sahraie A, Simmons A, Weiskrantz L, Williams SC (1998). Residual processing of chromatic signals in the absence of a geniculostriate projection. Vision Research 1998; 38: 3347-3353. Barbur JL, Moro S, Harlow JA, Lam BL, Liu M (2004). Comparison of pupil responses to luminance and colour in severe optic neuritis. Clinical Neurophysiology 2004; 115: 2650-2658. Barbur JL, Prescott NB, Douglas RH, Jarvis JR, Wathes CM (2002). A comparative study of stimulus-specific pupil responses in the domestic fowl (Gallus gallus domesticus) and the human. Vision Research 2002; 42: 249-255. Barlow HB (1958). Temporal and spatial summation in human vision at different background intensities. The Journal of Physiology 1958; 141: 337- 350. Bartels A, Zeki S (2000). The architecture of the colour centre in the human visual brain: new results and a review. The European Journal of Neuroscience 2000; 12: 172-193. Bartisch G (1585). Ophthalmodouleia. Translated by Blanchard DL. Oostende, Belgium: Wayenborgh, 1995. Originally published in 1585. Benca RM, Gilliland MA, Obermeyer WH (1998). Effects of lighting conditions on sleep and wakefulness in albino Lewis and pigmented Brown Norway rats. Sleep 1998; 21: 451-460. 372 Benevento LA, Davis B (1977). Topographical projections of the prestriate cortex to the pulvinar nuclei in the macaque monkey: an autoradiographic study. Experimental Brain Research 1977; 30: 405-424. Benevento LA, Rezak M, Santos-Anderson R (1977a). An autoradiographic study of the projections of the pretectum in the rhesus monkey (Macaca mulatta): evidence for sensorimotor links to the thalamus and oculomotor nuclei. Brain Research 1977; 127: 197-218. Benevento LA, Fallon J, Davis BJ, Rezak M (1977b). Auditory--visual interaction in single cells in the cortex of the superior temporal sulcus and the orbital frontal cortex of the macaque monkey. Experimental Neurology 1977; 57: 849-872. Benevento LA, Standage GP (1983). The organization of projections of the retinorecipient and nonretinorecipient nuclei of the pretectal complex and layers of the superior colliculus to the lateral pulvinar and medial pulvinar in the macaque monkey. The Journal of Comparative Neurology 1983; 217: 307- 336. Benevento LA, Yoshida K (1981). The afferent and efferent organization of the lateral geniculo-prestriate pathways in the macaque monkey. The Journal of Comparative Neurology 1981; 203: 455-474. Bergamin 0, Schoetzau A, Sugimoto K, Zulauf M (1998). The influence of iris color on the pupillary light reflex. Graefe's archive for clinical and experimental ophthalmology 1998; 236: 567-570. Bergamin 0, Kardon RH (2002). Greater pupillary escape differentiates central from peripheral visual field loss. Ophthalmology 2002; 109: 771-780. 373 Bergamin 0, Kardon RH (2003). Latency of the pupil light reflex: sample rate, stimulus intensity, and variation in normal subjects. Investigative Ophthalmology and Visual Science 2003; 44: 1546-1554. Bergamin 0, Zimmerman MB, Kardon RH (2003). Pupil light reflex in normal and diseased eyes: diagnosis of visual dysfunction using waveform partitioning. Ophthalmology 2003; 110: 106-114. Berman SM (1987). Pupillary size differences under incandescent and high pressure sodium lamps. The Journal of the Illuminating Engineering Society 1987; 16: 3-20. Berson DM, Dunn FA, Takao M (2002). Phototransduction by retinal ganglion cells that set the circadian clock. Science. 2002; 295: 1070-1073. Berson DM (2003). Strange vision: ganglion cells as circadian photoreceptors. Trends in Neurosciences 2003; 26: 314-320. Binkley S (1994). Wrist motion rhythm phase shifts in travellers may differ from changes in time zones. Physiology & Behavior 1994; 55: 967. Birch .1, Barbur JL and Harlow AJ (1992). A new method based on random luminance masking for measuring isochromatic zones using high resolution colour displays. Ophthalmic & Physiological Optics 1992; 12: 133-136. Bito LZ, Turansky DG (1975). Photoactivation of pupillary constriction in the isolated in vitro iris of a mammal (Mesocricetus auratus). Comparative biochemistry and physiology. A, Comparative physiology 1975; 50: 407-413. 374 Bloom PA, Papakostopoulos D, Gogolitsyn Y, Leenderz JA, Papakostopoulos S, Grey RH (1993). Clinical and infrared pupillometry in central retinal vein occlusion. British Journal of Ophthalmology 1993; 77:75-80. Bonnet M, Guenoun S (1995). Surgical risk factors for severe postoperative proliferative vitreoretinopathy (PVR) in retinal detachment with grade B PVR. Graefes Archive for Clinical & Experimental Ophthalmology 1995; 233: 789- 791. Borbely AA, Huston JP (1974) .Effects of two-hour light-dark cycles on feeding, drinking and motor activity of the rat. Physiology & Behavior 1974; 13: 795-802. Boring EG (1942) Sensation and perception in the History of Experimental Psychology. New York: Appleton-Century, 1942. Bos JE (1991). Detection of the pupil constriction latency. Medical & biological engineering & computing 1991; 29: 529-534. Bouma H (1962). Size of the static pupil as a function of wavelength and luminosity of the light incident on the human eye. Nature 1962; 193: 690-691. Bouma H (1965). Receptive Systems Mediating Certain Light. Reactions of the Pupil of the Human Eye. Philips Research Reports Supplement No. 5; Eindhoven: Philips' Gloeilampenfabrieken, 1965. Bouman MA and Schweitzer NM (1956). Threshold measurements on the light reflex of the pupil. Ophthalmologica 1956; 132: 286-288. Bovino JA, Burton TC (1980). Measurement of the relative afferent pupillary defect in retinal detachment. American Journal of Ophthalmology 1980; 90: 19-21. 375 Bowmaker JK (1984). Microspectrophotometry of vertebrate photoreceptors. Vision Research 1984; 24: 1641-1650. Bowmaker JK, Dartnall HJ (1980). Visual pigments of rods and cones in a human retina. The Journal of Physiology 1980; 298: 501-511. Brainard DH, Roorda A, Yamauchi Y, Calderone JB, Metha A, Neitz M, Neitz J, Williams DR, Jacobs GH (2000). Functional consequences of the relative numbers of L and M cones. Journal of the Optical Society of America. A, Optics, Image Science, and Vision 2000; 17: 607-614. Brainard GC, Hanifin JP, Greeson JM, Byrne B, Glickman G, Gerner E, Rollag MD (2001). Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor. The Journal of Neuroscience 2001; 21: 6405-6412. Bressler NM, Bressler SB, Fine SL (1988). Age-related macular degeneration. Survey of Ophthalmology 1988; 32: 375-3413. Bridges CD (1967). Spectroscopic properties of porphyropsins. Vision Research 1967; 7: 349-369. Brigell M, Bach M, Barber C, Moskowitz A, Robson J; Calibration Standard Committee of the International Society for Clinical Electrophysiology of Vision (2003). Guidelines for calibration of stimulus and recording parameters used in clinical electrophysiology of vision. Documenta Ophthalmologica. Advances in Ophthalmology 2003; 107: 185-193. Brindley GS (1970). Physiology of the Retina and Visual Pathway. Baltimore: Williams & Wilkins Company, 1970. Brodal P (1998). The Central Nervous System. 2nd edition. Oxford: Oxford University Press, 1998. 376 Bron AJ, Tripathi RC, Tripathi BJ (1997). Wolff's Anatomy of the Eye and Orbit. 8th edition. London: Hodder Arnold, 1997. Brown PK, Wald G (1964). Visual pigments in single rods and cones of the human retina. Direct measurements reveal mechanisms of human night and color vision. Science 1964; 144: 45-52. Brown RH, Zilis JD, Lynch MG, Sanborn GE (1987). The afferent pupillary defect in asymmetric glaucoma. Archives of Ophthalmology 1987; 105: 1540- 1543. Brown-Sequard CE (1856a). Proprietes et Fonctions de la Moelle Epiniere. Paris: Bailliere & Fils, 1856. (According to Loewenfeld and Lowenstein (1999) this work by Brown-Sequard was reported via Broca P in 1855 and 1856). [French] Brown-Sequard CE (1856b). Researches of the action of certain parts of the solar spectrum upon the iris (communicated by Dr. Sharpey). Proceedings of the Royal Society of London 1856; 8: 233-234. Brown-Sequard CE (1859). Recherches experimentales sur l'influence excitatrice de la lumiere, du froid et de la chaleur sur l'iris, dans les cinq classes d'animaux vertebras. Th Journal of Physiology 1859; 2:281-294; 451- 460. [French] Bullier J, Kennedy H (1983). Projection of the lateral geniculate nucleus onto cortical area V2 in the macaque monkey. Experimental Brain Research 1983; 53: 168-172. Burde RM (1988). Afferent pupillary defect in asymmetric glaucoma. Archives of Ophthalmology 1988; 106: 721. 377 Burris C, Klug K, Ngo IT, Sterling P, Schein S (2002). How Muller glial cells in Macaque fovea coat and isolate the synaptic terminals of cone photoreceptors. The Journal of Comparative Neurology 2002; 453: 100-111. Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ (1989). The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Research 1989; 28: 193-213. Cajal S Ramon y (1892) La retine des vertebras. La Cellule 1892; 9: 121-133. [French] Campbell FW, Green DG (1965). Optical and retinal factors affecting visual resolution. The Journal of Physiology 1965; 181: 576-593. Campbell SS, Murphy PJ (1998). Extraocular circadian phototransduction in humans. Science 1998; 279: 396-399. Capra F (2000). The Tao of Physics. Random House Inc., 2000. Card JP, Moore RY (1991). The organization of visual circuits influencing the circadian activity of the suprachiasmatic nucleas. In: Klein DC, Moore RY, Reppert SM (editors). Suprachiasmatic nucleas: the mind's clock. pp. 51-76. New York: Oxford University Press, 1991. Cardillo JA, Stout JT, LaBree L, Azen SP, Omphroy L, Cui JZ, Kimura H, Hinton DR, Ryan SJ (1997). Post-traumatic proliferative vitreoretinopathy. The epidemiologic profile, onset, risk factors, and visual outcome. Ophthalmology 1997; 104: 1166-1173. Cavaggioni A, Zampollo A (1968). The pupillary reflex and the pre-tectal region of the cat. Electroencephalography and Clinical Neurophysiology 1968; 25: 516. 378 Chanda R (1997). Quantum Mystery. India: National Book Trust, India, 1997. Chang TS, Freund KB, de la Cruz Z, Yannuzzi LA, Green WR (1994). Clinicopathologic correlation of choroidal neovascularization demonstrated by indocyanine green angiography in a patient with retention of good vision for almost four years. Retina 1994; 14: 114-124. Charman WN, Tucker J (1977). Dependence of accommodation response on the spatial frequency spectrum of the observed object. Vision Research 1977; 17: 129-139. Charman WN, Cronly JR (1991). Vision and Visual Dysfunction: Visual Optics and Instrumentation. pp. 392-393. Macmillan Press, 1991. Charteris DG, Sethi CS, Lewis GP, Fisher SK (2002). Proliferative vitreoretinopathy-developments in adjunctive treatment and retinal pathology. Eye 2002; 16: 369-374. Clarke M (2006). Modern treatment of amblyopia. In: Lorenz B, Moore AT (editors). Paediatric Ophthalmology, Neuro-ophthalmology, Genetics. Berlin and Heidelberg: Springer-Verlag, 2006. Clarke MP, Wright CM, Hrisos S, Anderson JD, Henderson J, Richardson SR (2003). Randomised controlled trial of treatment of unilateral visual impairment detected at preschool vision screening. British Medical Journal 2003; 327: 1251. Clarke RJ, Ikeda H (1985a). Luminance and darkness detectors in the olivary and posterior pretectal nuclei and their relationship to the pupillary light reflex in the rat. I. Studies with steady luminance levels. Experimental Brain Research 1985; 57: 224-232. 379 Clarke RJ, Ikeda H (1985b). Luminance detectors in the olivary pretectal nucleus and their relationship to the pupillary light reflex in the rat. II. Studies using sinusoidal light. Experimental Brain Research 1985; 59: 83-90. Clynes M (1961). Unidirectional rate sensitivity: a biocybernetic law of reflex and humoral systems as physiologic channels of control and communication. Annals of the New York Academy of Sciences 1961; 92: 946-969. Cocker KD, Moseley MJ (1992). Visual acuity and the pupil grating response. Clinical Vision Science 1992; 7: 143-146. Cocker KD, Moseley MJ (1996). Development of pupillary responses to grating stimuli. Ophthalmic & Physiological Optics 1996; 16: 64-67. Cocker KD, Moseley MJ, Bissenden JG, Fielder AR (1992). Visual acuity and pupillary responses to spatial structure in infants. Investigative Ophthalmology and Visual Science 1994; 35: 2620-2625. Cocker KD, Moseley MJ, Stirling HF, Fielder AR (1998). Delayed visual maturation: pupillary responses implicate subcortical and cortical visual systems. Developmental Medicine and Child Neurology 1998; 40: 160-162. Cohen GH, Saini VD (1978). The human pupil response as an objective determination of colour vision deficiency. Modern Problems in Ophthalmology 1978; 19: 197-200. Colburn TR, Smith BM, Guarini JJ, Simmons NN (1976). An ambulatory activity monitor with solid state memory. ISA Transactions 1976; 15: 149-154. Cook MW, Levin LA, Joseph MP, Pinczower EF (1996). Traumatic optic neuropathy. A meta-analysis. Archives of Otolaryngology -- Head & Neck Surgery 1996; 122: 389-392. 380 Cooper HM, Magnin M (1986). A common mammalian plan of accessory optic system organization revealed in all primates. Nature 1986; 324: 457-459. Cowey A, Stoerig P, Perry VH (1989). Transneuronal retrograde degeneration of retinal ganglion cells after damage to striate cortex in macaque monkeys: selective loss of P beta cells. Neuroscience 1989; 29: 65-80. Cowey A, Stoerig P (1989). Projection patterns of surviving neurons in the dorsal lateral geniculate nucleus following discrete lesions of striate cortex: implications for residual vision. Experimental Brain Research 1989; 75: 631- 638. Cowey A (2004). The 30th Sir Frederick Bartlett lecture. Fact, artefact, and myth about blindsight. The Quarterly Journal of Experimental Psychology 2004; 57: 577-609. Cox TA, Drewes CP (1984). Contraction anisocoria resulting from half-field illumination. American Journal of Ophthalmology 1984; 97: 577-582. Cox TA (1992). Pupillary escape. Neurology 1992; 42: 1271-1273. Cox TA (1992). Pupillary escape. Ophthalmology 2003; 110: 245-248. Craik K J W (1940). Origin of visual after-images. Nature 1940; 145: 512. Crawford BH (1936). The dependence of pupil size upon external light stimulus under static and variable conditions. Proceedings of the Royal Society London. Series B, Containing papers of a Biological character.1936; 121: 376-395. Crick F (1994). The astonishing hypothesis: The scientific search for the soul. New York: Charles Scribner's Sons, 1994. 381 Cruz RD, Adachi-Usami E (1989). Quantitative evaluation of electroretinogram before cataract surgery. Japanese Journal of Ophthalmology 1989; 33: 451-457. Curcio CA, Allen KA (1990). Topography of ganglion cells in human retina. The Journal of Comparative Neurology 1990; 300: 5-25. Curcio CA, Sloan KR, Kalina RE, Hendrickson AE (1990). Human photoreceptor topography. The Journal of Comparative Neurology 1990; 292: 497-523. Curcio CA, Allen KA, Sloan KR, Lerea CL, Hurley JB, Klock IB, Milam AH (1991). Distribution and morphology of human cone photoreceptors stained with anti-blue opsin. The Journal of Comparative Neurology 1991; 312: 610- 624. Dacey DM, Liao HW, Peterson BB, Robinson FR, Smith VC, Pokorny J, Yau KW, Gamlin PD (2005). Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 2005; 433: 749-754. Dalley RW, Gown AM, Orcutt JC (1992). Evaluation of Orbital Disease. In: Walsh TJ (editor). Neuro-ophthalmology: Clinical Signs and Symptoms. 3rd edition. Philadelphia: Lea & Febiger, 1992. Dartnall H (1953). The interpretation of spectral sensitivity curves. British Medical Bulletin 1953; 9: 24-30. Dartnall HJ (1968). The photosensitivities of visual pigments in the presence of hydroxylamine. Vision Research 8: 339-358. 382 Datiles MB, Edwards PA, Kaiser-Kupfer MI, McCain L, Podgor M (1987). A comparative study between the PAM and the laser intereferometer in cataracts. Graefe's Archive for Clinical and Experimental Ophthalmology 1987; 225: 457. Dayson H (1980). Physiology of the Eye. 4th edition. Edinburgh, London and New York: Churchill Livingstone, 1980. Deacon S, English J, Arendt J (1994). Acute phase-shifting effects of melatonin associated with suppression of core body temperature in humans. Neuroscience 1994; 178: 32-34. Delori FC, Goger DG, Hammond BR, Snodderly DM, Burns SA (2001). Delori Macular pigment density measured by autofluorescence spectrometry: comparison with reflectometry and heterochromatic flicker photometry. Journal of the Optical Society of America. A, Optics, image science, and vision 2001; 18: 1212-1230. De Marco R, Ambrosio G, Ferrara G (2000). Measuring contrast sensitivity in patients with Graves' ophthalmology complicated by ocular hypertension and suspect glaucoma or dysthyroid optic neuropathy. Documenta Ophthalmologica 2000; 100: 165-177. de Monasterio FM (1978a). Macular pigmentation and the spectral sensitivity of retinal ganglion cells of macaques. Vision Research 1978; 18: 1273-1277. de Monasterio FM (1978b). Spectral interactions in horizontal and ganglion cells of the isolated and arterially-perfused rabbit retina. Brain Research 1978; 150: 239-258. de Monasterio FM (1978c). Properties of concentrically organized X and Y ganglion cells of macaque retina. Journal of Neurophysiology 1978; 41: 1394- 1417. 383 de Monasterio FM (1978d). Center and surround mechanisms of opponent- color X and Y ganglion cells of retina of macaques. Journal of Neurophysiology 1978; 41: 1418-1434. de Monasterio FM (1978e). Properties of ganglion cells with atypical receptive field organization in retina of macaques. Journal of Neurophysiology 1978; 41: 1435-1449. de Monasterio FM, Gouras P (1975). Functional properties of ganglion cells of the rhesus monkey retina. The Journal of Physiology 1975; 251: 167-195. De Valois RL, Jacobs GH (1968). Primate color vision. Science 1968; 162: 533-540. de Waard PW, IJspeert JK, van den Berg TJ, de Jong PT (1992). Intraocular light scattering in age-related cataracts. Investigative Ophthalmology & Visual Science 1992; 33: 618-625. Dkhissi-Benyahya 0, Rieux C, Hut RA, Cooper HM (2006). Immunohistochemical evidence of a melanopsin cone in human retina. Investigative Ophthalmology & Visual Science 2006; 47: 1636 - 1641. Dodt, E (1973). In: Jung R (editor). Handbook of Sensory Physiology. pp. 113-140 Berlin: Springer, 1973. Duffey RJ, Learning D (2003). US trends in refractive surgery: 2002 ISRS survey. Journal of Refractive Surgery 2003; 19:357-363. Duke-Elder WS (1932). Text-book of Ophthalmology. vol. I. The development, form and function of the visual apparatus. London: Henry Kimpton, 1932. Duke-Elder S (1958). System of Ophthalmology. vol. I. The Eye in Evolution. St. Louis: C. V. Mosby Company, 1958. 384 Duke-Elder S, Wybar KC (1961). System of Ophthalmology. vol. II. The Anatomy of the Visual System. St. Louis: C. V. Mosby Company, 1961. Duke-Elder S, Gloster J, Weale R (1968). System of Ophthalmology. vol. IV. Physiology of the Eye and of Vision. St. Louis: C. V. Mosby Company, 1968. Duke-Elder S, Scott GI (1971). System of Ophthalmology. vol. XII. Neuro- ophthalmology. London: Henry Kimpton, 1971. Ebisawa T, Karne S, Lerner MR, Reppert SM (1994). Expression cloning of a high-affinity melatonin receptor from Xenopus dermal melanophores. Proceedings of the National Academy of Sciences of the United States of America 1994; 91: 6133-6137. Einstein A (1905). Uber einen die erzeugung and verwandlung des lichtes betreffenden heuristischen gesichtspunkt; Annalen der Physik (Leipzig) 1905; 17: 132. [On a heuristic viewpoint concerning the production and transformation of light. Translated from the German by Arons AB, Peppard MB. The American Journal of Physiology 1965; 33: 367] Einstein A ( 1916). Relativity: The Special and the General Theory (1916). Translated by Lawson RW. Three Rivers Press, 1995. [Translated in 1920 from the German work of 1916 incorporating the Special Theory (originally published 1905) and the General Theory (originally published 1915)] Evans K, Fryer A, Inglehearn C, Duvall-Young J, Whittaker JL, Gregory CY, Butler R, Ebenezer N, Hunt DM, Bhattacharya S (1994). Genetic linkage of cone-rod retinal dystrophy to chromosome 19q and evidence for segregation distortion. Nature Genetics 1994; 6: 210-213. 385 Evans K, Duvall-Young J, Fitzke FW, Arden GB, Bhattacharya SS, Bird AC (1995). Chromosome 19q cone-rod retinal dystrophy. Ocular phenotype. Archives of Ophthalmology 1995; 113: 195-201. Ferraz-Mello S (1981). Estimation of periods from unequally spaced observations. The Astronomical Journal 1981; 86: 619-624. Fielder AR, Moseley MJ (2000). Environmental light and the preterm infant. Seminars in Perinatology 2000 Aug; 24: 291-298. Fischer B, Freund HJ (1970). Eine mathematische Formulierung fur Reiz- Reaktionsbeziehungen retinaler Ganglienzellen. Kybernetik 1970; 7: 160-166. [German] Fishman GA (2003). When your eyes have a wet nose: the evolution of the use of guide dogs and establishing the seeing eye. Survey of Ophthalmology 2003; 48: 452-458 Flamant F (1948). Variation du diametre de la pupille de l'oeil en fonction de la brilliance. Rev. d'Opt. 1948; 34: 433-459. [French] Folk JC, Thompson HS, Farmer SG, O'Gorman TW Dreyer RF (1987). Relative afferent pupillary defect in eyes with retinal detachment. Ophthalmic Surgery 1987; 18: 757-759. Forsius H, Eriksson AW, Krause U (1964). The dazzling test in diseases of the retina. Acta Ophthalmologica 1964; 42: 55-63. Foster RG (1998). Shedding light on the biological clock. Neuron 1998; 20: 829-832. Foster RG (2005). Neurobiology: Bright blue times. Nature 433, 698 — 699. 386 Foster RG, Hankins MW (1998). Non-rod, non-cone photoreception in the vertebrates. Progress in Retinal and Eye Research 2002; 21: 507-527. Foster RG, Hankins MW, Peirson SN (2007). Light, photoreceptors, and circadian clocks. Methods in Molecular Biology 2007; 362: 3-28. Foster RG, Provencio I, Hudson D, Fiske S, De Grip W, Menaker M (1991). Circadian photoreception in the retinally degenerate mouse (rd/rd). Journal of Comparative Physiology. A, Sensory, Neural, and Behavioral Physiology 1991; 169: 39-50. Foster RG, Hankins M, Lucas RJ, Jenkins A, Munoz M, Thompson S, Appleford JM, Bellingham J (2003). Non-rod, non-cone photoreception in rodents and teleost fish. Novartis Foundation Symposium 2003; 253: 3-23; discussion 23- 30, 52-55, 102-109. Franssen L, Coppens JE, van den Berg TJ (2006). Compensation comparison method for assessment of retinal straylight. Investigative Ophthalmology and Visual Science 2006; 47: 768-776. Freedman MS, Lucas RJ, Soni B, von Schantz M, Munoz M, David-Gray ZK and Foster RG (1999). Regulation of mammalian circadian behavior by non- rod, non-cone, ocular photoreceptors. Science 284: 502-504. Freund CL, Gregory-Evans CY, Furukawa T, Papaioannou M, Looser J, Ploder L, Bellingham J, Ng D, Herbrick JA, Duncan A, Scherer SW, Tsui LC, Loutradis-Anagnostou A, Jacobson SG, Cepko CL, Bhattacharya SS, McInnes RR (1997). Cone-rod dystrophy due to mutations in a novel photoreceptor- specific homeobox gene (CRX) essential for maintenance of the photoreceptor. Cell. 1997; 91: 543-553. 387 Galen C (2nd century). On the Usefulness of the Parts. Book 10, 11/72. Translated by May MT. p. 476. Ithaca, New York: Cornell University Press, 1968. Originally written in the second century. Galloway NR (1997). Electrodiagnosis. In: Walsh TJ (editor). Neuro- ophthalmology: Clinical Signs and Symptoms. 4(11 edition 1997. Baltimore: Williams & Wilkins, 1997. Galloway N (1998). Electrodiagnosis. In: Hamilton AMP, Gregson R, Fish GE (editors). Text Atlas of Retina. London: Martin Dunitz, 1998. Gamlin PD, Zhang H, Harlow A, Barbur JL (1998). Pupil responses to stimulus color, structure and light flux increments in the rhesus monkey. Vision Research 1998; 38: 3353-3358. Gamlin PD, McDougal DH, Pokorny J, Smith VC, Yau KW, Dacey DM (2007). Human and macaque pupil responses driven by melanopsin-containing retinal ganglion cells. Vision Research 2007; 47: 946-954. Garcia-Fernandez JM, Jimenez AJ, Gonzalez B, Pombal MA, Foster RG (1997). An immunocytochemical study of encephalic photoreceptors in three species of lamprey. Cell and Tissue Research 1997; 288: 267-278. Gass JDM (1967). Pathogenesis of disciform detachment of the neuroepithelium. 1. General concepts and classification. American Journal of Ophthalmology 1967; 63: 573-585. Gass JDM (1994). Biomicroscopic and histopathologic considerations regarding the feasibility of surgical excision of subfoveal neovascular membranes. American Journal of Ophthalmology 1994; 118: 285-298. 388 Glansholm A, Hedin A, Tengroth B (1974). Pupillography in colour normals and colour defectives with chromatic stimuli of equal physical power. Modern Problems in Ophthalmology 1974; 13: 185-189. Glaser JS, Savino PJ, Sumers KD, McDonald SA, Knighton RW (1977). The photostress recovery test in the clinical assessment of visual function. American Journal of Ophthalmology 1977; 83: 255-260. Glod CA, Teicher MH, Polcari A, McGreenery CE, Ito Y (1997). Circadian rest-activity disturbances in children with seasonal affective disorder. Journal of the American Academy of Child and Adolescent Psychiatry 1997; 36: 195- 199. Goebel R, Muckli L, Zanella FE, Singer W, Stoerig P (2001). Sustained extrastriate cortical activation without visual awareness revealed by fMRI studies of hemianopic patients. Vision Research 2001; 41: 1459-1474. Goldthwaite D, Lakowski R, Drance SM (1976). A study of dark adaptation in ocular hypertensives. Canadian Journal of Ophthalmology 1976; 11 : 55-60. Gomez-Ulla F, Louro 0, Mosquera M (1986). Macular dazzling test on normal subjects. The British Journal of Ophthalmology 1986; 70: 209-213. Govardovskii VI, Fyhrquist N, Reuter T, Kuzmin DG, Donner K (2000). In search of the visual pigment template. Visual Neuroscience 2000; 17: 509-528. Gray H, Bannister LH, Berry MM, Williams PL (1995). Gray's Anatomy: The Anatomical Basis of Medicine and Surgery. 38th edition. Edinburgh: Churchill Livingstone, 1995. Green WR, McDonnell PJ, Yea JH (1985). Pathologic features of senile macular degeneration. Ophthalmology 1985; 92: 615-627. 389 Gregory RL (1998). Eye and Brain. The Psychology of Seeing. 5th edition. Oxford: Oxford University Press, 1998. Grey RH, Bloom PA (1991). Retinal ischaemia and relative afferent pupil defects in central retinal vein occlusion. European Journal of Ophthalmology 1991; 1: 85-88. Griisser O-J, Griisser-Comehls U (1973). In: Jung R (editor). The Handbook of Sensory Physiology. vol. 3a. pp. 333-429. Berlin-Heidelberg-New York: Springer, 1973. Griisser OJ (1989). Vision and Eye Movements. In: Schmidt RF, Thews G (editors). Human Physiology. New York: Springer-Verlag, 1989. Guerin CJ, Anderson DH, Fariss RN, Fisher SK (1989). Retinal reattachment of the primate macula. Photoreceptor recovery after short-term detachment. Investigative Ophthalmology & Visual Science 1989; 30: 1708-1725. Gunn RM (1897). Discussion on retro-ocular neuritis. Transactions. Ophthalmological Society of the United Kingdom 1897; 17: 107-108. Gunn RM (1902). On functional or hysterical blindness. Ophthalmic Review 1902; 21: 271-280. Hart WM Jr (1987.) Acquired dyschromatopsias. Survey of Ophthalmology 1987; 32: 10-31. Hagstrom SA, Neitz M, Neitz J (2000). Cone pigment gene expression in individual photoreceptors and the chromatic topography of the retina. Journal of the Optical Society of America. A, Optics, Image, Science, and Vision 2000; 17: 527-537. 390 Halliday D, Resnick R, Walker J (2002). Fundamentals of Physics. 6th edition. New York: John Wiley & Sons, 2002. Hankins MW, Jones RJ, Ruddock KH (1998). Diurnal variation in the b-wave implicit time of the human electroretinogram. Visual Neuroscience 1998; 15: 55-67. Hankins MW, Lucas RJ (2002). The primary visual pathway in humans is regulated according to long-term light exposure through the action of a nonclassical photopigment. Current Biology 2002; 12: 191-198. Hartline HK (1938). The response of single optic nerve fibers of the vertebrate eye to illumination of the retina. American Journal of Physiology 1938; 121: 400-415. Hartline HK (1940). The effects of spatial summation in the retina on the excitation of the fibers of the optic nerve. American Journal of Physiology 1940; 130: 700-711. Hartwig HG, van Veen T (1979). Spectral characteristics of visible radiations penetrating into the brain and stimulating extra-retinal photoreceptors. Journal of Comparative Physiology. A, Sensory, Neural, and Behavioral Physiology 1979; 120: 277-282. Hattar S, Liao HW, Takao M, Berson DM, Yau KW (2002). Melanopsin- containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 2002; 295: 1065-1070. Hattar S, Lucas RJ, Mrosovsky N, Thompson S, Douglas RH, Hankins MW, Lem J, Biel M, Hofmann F, Foster RG, Yau KW (2003). Melanopsin and rod- cone photoreceptive systems account for all major accessory visual functions in mice. Nature 2003; 424: 76-81. 391 Hecht S, Shlaer S, Pirenne MH (1942). Energy, quanta and vision. The Journal of General Physiology; 1942; 25: 819-840. Hedin A, Glansholm A (1976). Pupillary spectral sensitivity in normals and colour defectives. Modern Problems in Ophthalmology 1976; 17: 231-236. Heill A, Lindgren G, Olsson J (1987). Normal variability of static perimetric threshold values across the central visual field. Archives of Ophthalmology 1987; 105: 1544-1549. Heijl A, Lindgren G, Olsson J (1989a). The effect of perimetric experience in normal subjects. Archives of Ophthalmology 1989; 107: 81-86. Heril A, Lindgren A, Lindgren G (1989b). Test-retest variability in glaucomatous visual fields. American Journal of Ophthalmology 1989; 108: 130-135. Hendrickson AE (1994). Primate foveal development: a microcosm of current questions in neurobiology. Investigative Ophthalmology and Visual Science 1994; 35: 3129-3133. Hirschberg J (1985). The History of Ophthalmology. vol. 2. The Middle Ages. p. 145, note no. 427 on Ammar. Translated by Blodi FC and Bonn JP. Wayenborgh Verlag, 1985. Hochberg J (1978). Perception. 2nd edition. Englewood Cliffs, New Jersey: Prentice Hall, Inc. 1978. Hochberg J (1984). Perception. In: Darian-Smith I (editor). Handbook of Physiology. Section I, The Nervous System, Vol. III. Sensory Processes, Part I. Bethesda, Maryland: The American Physiological Society, 1984. 392 Hoekert M, der Lek RF, Swaab DF, Kaufer D, Van Someren EJ (2006). Comparison between informant-observed and actigraphic assessments of sleep- wake rhythm disturbances in demented residents of homes for the elderly. The American journal of Geriatric Psychiatry 2006; 14: 104-111. Hoffmann KP, Distler C, Erickson RG, Mader W (1988). Physiological and anatomical identification of the nucleus of the optic tract and dorsal terminal nucleus of the accessory optic tract in monkeys. Experimental Brain Research 1988; 69: 635-644. Hogan MJ, Zimmermann LE (1962). Ophthalmic Pathology. An Atlas and Textbook, 2nd edition. Philadelphia: W B Saunders, 1962. Holder GE (1987). Significance of abnormal pattern electroretinography in anterior visual pathway dysfunction. The British Journal of Ophthalmology 1987; 71: 166. Holder GE, Brigell MG, Hawlina M, Meigen T, Vaegan, Bach M; for the International Society for Clinical Electrophysiology of Vision (2007). ISCEV standard for clinical pattern electroretinography-2007 update. Documenta Ophthalmologica 2007; 114: 111-116. Hong S, Narkiewicz J, Kardon RH (2001). Comparison of pupil perimetry and visual perimetry in normal eyes: decibel sensitivity and variability. Investigative Ophthalmology and Visual Science 2001; 42: 957-965. Horiguchi M, Ito Y, Miyake Y (1998). Extrafoveal photostress recovery test in glaucoma and idiopathic central serous chorioretinopathy. The British Journal of Ophthalmology 1998; 82: 1007-1012. Home JA, Ostberg 0 (1976): A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. International Journal of Chronobiology 1976; 4: 97-110. 393 Hodges MD, Vieira H, Gregory-Evans K, Gregory-Evans CY (2002). Characterization of the genomic and transcriptional structure of the CRX gene: substantial differences between human and mouse. Genomics 2002; 80: 531- 542. Hubel DH, Wiesel TN (1962). Receptive fields, binocular interaction and functional architecture in the cat's visual cortex. The Journal of Physiology 1962; 160: 106-154. Hubei DH, Wiesel TN (1965). Binocular interaction in striate cortex of kittens reared with artificial squint. Journal of Neurophysiology 1965; 28: 1041-1059. Hubei DH, Wiesel TN (1968). Receptive fields and functional architecture of monkey striate cortex. The Journal of Physiology 1968; 195: 215-243. Hubel DH, Wiesel TN (1974). Sequence regularity and geometry of orientation columns in the monkey striate cortex. The Journal of Comparative Neurology 1974; 158: 267-294. Hubel DH, Wiesel TN (1978). Anatomical demonstration of orientation columns in macaque monkey. The Journal of Comparative Neurology 1978; 177: 361-380. Hunain ibn Is'haq (9th century). The Book of the Ten Treatises of the Eye Ascribed to Hunain Ibn Is'haq (809-877 AD). Oculist in Cairo. Translated by Max Meyerhof. Government Press, 1928. Originally written in the ninth century. Hutchins B, Weber JT (1985). The pretectal complex of the monkey: a reinvestigation of the morphology and retinal terminations. The Journal of Comparative Neurology 1985; 232: 425-442. 394 Ishikawa S, Naito M, Inaba K (1970). A new videopupillography. Ophthalmologica. 1970; 160: 248-259. Itabashi T, Wada Y, Sato H, Kunikata H, Kawamura M, Tamai M (2003). Ocular findings in a Japanese family with an Arg41Trp mutation of the CRX gene. Graefe's Archive for Clinical and Experimental Ophthalmology 2003; 241: 535-540. Jiang MQ, Thompson HS, Lam BL (1989). Kestenbaum's number as an indicator of pupillomotor input asymmetry. American Journal of Ophthalmology 1989; 107: 528-530. Jindra LF, Zemon V (1989). Contrast sensitivity testing: a more complete assessment of vision. Journal of Cataract and Refractive Surgery 1989; 15: 141-148. Jonas JB, Zach FM, Naumann GO (1990). Dark adaptation in glaucomatous and nonglaucomatous optic nerve atrophy. Graefe's Archive for Clinical and Experimental Ophthalmology 1990; 228: 321-325. Jung R (1984). Sensory Research in Historical Perspective: some Philosophical Foundations of Perception. In: Darian-Smith I (editor). Handbook of Physiology. pp. 1-74. Bethesda, Maryland: The American Physiological Society, 1984. Kaback MB, Burde RM, Becker B (1976). Relative afferent pupillary defect in glaucoma. American Journal of Ophthalmology 1976; 81: 462-468. Kadlecova V (1960). The pupillomotorial adaptation of the eye. Ophthalmologica. 1960; 140: 379-387. 395 Kadlecova V, Peleska M (1955). Infrascopic examination of the diameter of the pupil in darkness adaptation; infrascopic pupilometry. Ceskoslovenskci oftalmologie.1955; 11: 260-266. [Czech with English summary] Kalaboukhova L, Fridhammar V, Lindblom B (2007). Relative afferent pupillary defect in glaucoma: a pupillometric study. Acta Ophthalmologica Scandinavica. 2007; 85: 519-525. Kani K, Aoyama T, Maruo T (1978). Objective perimetry. Nippon Rinsho 1978; Supplement: 1832-1833. [Japanese, abstract and selections of text in English] Kant I (1783). Critik der Reinen Vernunft. 1781; 2nd edition. Riga: Hartknoch, 1783. [Critique of Pure Reason. Translated from the German by Smith MK. London: Macmillan, 1929] Kardon RH (1992). Pupil perimetry. Current Opinion in Ophthalmology 1992; 3: 565-570. Kardon RH (1995). Pupillary light reflex. Current Opinion in Ophthalmology 1995; 6: 20-26. Kardon RH (1998). Anatomy and Physiology of the Pupil. In: Miller NR, Newman NJ (editors). Walsh & Hoyt's Clinical Neuro-ophthalmology. 5th edition. vol.] . chapter 20. Baltimore: Williams & Wilkins, 1998. Kardon RH (2003). The Pupil. In: Kaufman PL, Alm A (editors). Adler's Physiology of the Eye. 10th edition. chapter 32. St. Louis: Mosby, 2003. Kardon RH, Kirkali PA, Thompson HS (1991). Automated pupil perimetry. Pupil field mapping in patients and normal subjects. Ophthalmology 1991; 98: 485-495; discussion 495-496. 396 Kawasaki A, Moore P, Kardon RH (1995). Variability of the relative afferent pupillary defect. American Journal of Ophthalmology 1995; 120: 622-633. Katz B (1952). The Nerve Impulse. Scientific American 1952; 11: 55. Katz B (1966). Nerve, Muscle and Synapse. McGraw-Hill, 1966. Kawabata H (1960). Course of the potential change in the human electroretinogram during light adaptation. Journal of the Optical Society of America 1960; 50: 456-461. Keeler CE (1927a). Le reflexe irien a la lumiere chez la souris a refine sans batonnets. Comptes rendus des seances de la Societe de biologic et de ses filiales 1927; 96:10 - Abstract in Zentralblatt fur die Gesamte Ophthalmologic and ihre Grenzgebiete 1927;18:252. [French and German] Keeler CE (1927b). Iris movements in blind mice. The American Journal of Physiology 1927; 81: 107-112. Keeler CR (2002). The ophthalmoscope in the lifetime of Hermann von Helmholtz. Archives of Ophthalmology 2002; 120: 194-201. Keenleyside MS (1989). Pupillometry and Assessment of Visual Function. Oxford: University of Oxford; D.Phil thesis, 1989. Kepler J (1611). Dioptrice. Augsburg, Germany: Frank, 1611. In: von Dyck W, Caspar M (editors). Gesammelte Werke. vol. 4. Mi,inchen: Beck, 1937-1963. [Latin] Kepler J (1619). Harmonices Mundi. Linz, Austria: Tampach, Plancus, 1619. In: von Dyck W, Caspar m (editors). Gesammelte Werke. Vol. 6. Miinchen: Beck, 1937-1963. [Latin] 397 Kestenbaum A (1946). Clinical Methods of Neuro-ophthalmic Examination. pp. 289-291. New York: Grune and Stratton, 1946. King-Smith PE, Carden D (1976). Luminance and opponent-color contributions to visual detection and adaptation and to temporal and spatial integration .Journal of the Optical Society of America 1976; 66: 709-717. Knowles A, Dartnall H (1977). The Photobiology of Vision. New York: Academic Press Inc, 1977. Koelbing Huldrych M (1967). Renaissance der Augenheilkunde, 1540-1630. pp. 71-99. Bern and Stuttgart: Verlag Hans Huber, 1967. [German] Koeppe ER, Alexandridis E (1968). Pupillenlichtreflex und Purkinje- Verschiebung. Bericht iiber die Zusammenkunft. Deutsche Ophthalmologische Gesellschaft 1968; 68: 278-281. [German] Kohn M, Clynes M (1969). Color dynamics of the pupil. Annals of the New York Academy of Sciences 1969; 156: 931-950. Miner H (1912). Die StOrungen des Farbensinnes. ihre Klinische Bedeutung und ihre Diagnose. Berlin: Karger, 1912. [German] Kon CH, Asaria RH, Occleston NL, Khaw PT, Aylward GW (2000). Risk factors for proliferative vitreoretinopathy after primary vitrectomy: a prospective study. The British Journal of Ophthalmology 2000; 84: 506-511. Koorengevel KM, Gordijn MC, Beersma DG, Meesters Y, den Boer JA, van den Hoofdakker RH, Daan S (2001). Extraocular light therapy in winter depression: a double-blind placebo-controlled study. Biological Psychiatry 2001; 50: 691-698. Erratum in: Biological Psychiatry 2002; 51: 194. 398 Kraats J van der, Smit EP, Slooter JA (1977). Objective measurements by the pupil balance method. Documenta Ophthalmologica 1977; 14:213. Krastel H, Alexandridis E, Gertz J (1985). Pupil increment thresholds are influenced by color opponent mechanisms. Ophthalmologica 1985; 191: 35-38. Ki,ihne W (1877). Uber den Sehpurpur. Untersuch. Physiol. Inst. Univ. Heidelberg 1877; 1: 15-103. [German] Kiihne W (1879). Chemische Vorgiinge in der Netzhaut. In: Hermann L (editor). Handbuch der Physiologie. Sinnesorgane. vol 3, pt 1, pp. 235-342. Leipzig: Vogel, 1879. [German] Lagreze WD, Kardon RH (1998). Correlation of relative afferent pupillary defect and estimated retinal ganglion cell loss. Graefe's Archive for Clinical and Experimental Ophthalmology 1998; 236: 401-404. Lakowski R, Drance SM, Goldthwaite D (1976). Chromatic extrafoveal dark adaptation function in normal and glaucomatous eyes. Modern Problems in Ophthalmology 976; 17: 304-310. Lambert JH (1760). Photometria, sive de mensura et gradibus luminis, calorum et umbrae. part IV, chapter II, pp. 370. [Translated from Latin into German by Anding E. Leipzig: W. Engelmann, 1892] Langley JN (1905-1906). On the reaction of cells and of nerve endings to certain poisons. The Journal of Physiology 1905-1906; 33: 374-413. Lau KC, So KF, Campbell G, Lieberman AR (1992). Pupillary constriction in response to light in rodents, which does not depend on central neural pathways. Journal of the Neurological Sciences 1992; 113: 70-79. 399 Laurens H (1923). Studies on the relative physiological value of spectral lights: III. The pupillomotor effects of wave-lengths of equal energy content. The American Journal of Physiology 1923; 64: 97-119. Le Grand Y (1970). An Introduction to Photobiology: the Influence of Light on Life. London: Faber and Faber, 1970. Lee RE, Cohen GH, Boynton RM (1969). Latency variation in human pupil contraction due to stimulus luminance and/or adaptation level. Journal of the Optical Society of America 1969; 59: 97-103. Leh SE, Mullen KT, Ptito A (2006). Absence of S-cone input in human blindsight following hemispherectomy. European Journal of Neuroscience 2006; 24: 2954-2960. Leventhal AG, Rodieck RW, Dreher B (1981). Retinal ganglion cell classes in the Old World monkey: morphology and central projections. Science 1981; 213: 1139-1142. Levin LA, Beck RW, Joseph MP, Seiff S, Kraker R (1996). Traumatic optic neuropathy. A meta-analysis. Archives of Otolaryngology -- Head & Neck Surgery 1996; 122: 389-392. Lewis RA, Johnson CA, Keltner JL, Labermeier PK (1986). Variability of quantitative automated perimetry in normal observers. Ophthalmology 1986; 93: 878-881. Lewis H, Aaberg TM, Abrams GW (1991). Cause of failure after initial vitreoretinal surgery for severe proliferative vitreoretinopathy. American Journal of Ophthalmology 1991; 111: 8-14. 400 Liebman PA (1972). Microspectrophotometry of photoreceptors. In: Dartnall HJA (editor). Handbook of Sensory Physiology. vol VII, part 1, pp. 481-528. Berlin: Springer, 1972. Lindblom N, Hatonen T, Laakso M, Alila-Johansson A, Laipio M, Turpeinen U (2000). Bright light exposure of a large skin area does not affect melatonin or bilirubin levels in humans. Biological Psychiatry 2000; 48: 1098-1104. Lindsley DB, Bowden JW, Magounm HW (1949). Effect upon the EEG of acute injury to the brain stem activating system. Electroencephalography and Clinical Neurophysiology 1949; 1: 475-486. Livingstone MS, Hubel DH (1988a). Segregation of form, color, movement, and depth: anatomy, physiology, and perception. Science 1988; 240: 740-749. Livingstone MS, Hubel DH (1988b). Do the relative mapping densities of the magno- and parvocellular systems vary with eccentricity? The Journal of Neuroscience 1988; 8: 4334-4339. Lockley SW, Skene DJ, Arendt J, Tabandeh H, Bird AC, Defrance R (1997). Relationship between melatonin rhythms and visual loss in the blind. The Journal of Clinical Endocrinology and Metabolism 1997; 82: 3763-3770. Loewenfeld IE, Lowenstein 0 (1999). The Pupil. vols. I and 2. Oxford: Butterworth-Heinemann, 1999. Lomb NR (1976). Least-squares frequency analysis of unequally spaced data. Astrophysics and Space Science 1976; 39: 447-462. Longtin A, Milton JG, Bos JE, Mackey MC (1990). Noise and critical behavior of the pupil light reflex at oscillation onset. Physical review. A. 1990; 41: 6992-7005. 401 Lowenstein 0 (1954). Clinical pupillary symptoms in lesions of the optic nerve, optic chiasm and optic tract. Archives of Ophthalmology 1954; 52: 385-403. Lowenstein 0, Givner I (1943). Pupillary reflex to darkness. Archives of Ophthalmology 1943; 30: 603. Lowenstein 0, Loewenfeld IE (1958). Electronic pupillography; a new instrument and some clinical applications. A.M.A. Archives of Ophthalmology 1958; 59: 352-363. Lowenstein 0, Loewenfeld IE (1969). The Pupil. In: Dayson H (editor). The Eye. vol 3, Muscular Mechanisms, pp. 274. New York: Academic Press, 1969. Lu J, Shiromani P, Saper CB (1999). Retinal input to the sleep-active ventrolateral preoptic nucleus in the rat. Neuroscience 1999; 93: 209-214. Lucas RJ, Freedman MS, Munoz M, Garcia-Fernandez JM, Foster RG (1999). Regulation of the mammalian pineal by non-rod, non-cone, ocular photoreceptors. Science 1999; 284: 505-507. Lucas RJ, Douglas RH, Foster RG (2001). Characterization of an ocular photopigment capable of driving pupillary constriction in mice. Nature Neuroscience 2001; 4: 621-626. Lucas RJ, Hattar S, Takao M, Berson DM, Foster RG, Yau KW (2003). Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 2003; 299: 245-247. Lund VJ, Larkin G, Fells P, Adams G (1997). Orbital decompression for thyroid eye disease: a comparison of external and endoscopic techniques. The Journal of Laryngology and Otology 1997; 111: 1051-1055. 402 Lund VJ, Rose GE (2006). Endoscopic transnasal orbital decompression for visual failure due to sphenoid wing meningioma. Eye 2006; 20: 1213-1219. Lythgoe J (1979). The Ecology of Vision. Oxford: Oxford University Press, 1979. MacAdam DL (1970). Sources of Color Science. Cambridge MA: MIT Press, 1970. Mai JK (1978). The accessory optic system and the retino-hypothalamic system. A review. Journal fair Hirnforschung. 1978; 19: 213-288. [German] Margrain TH, Thomson D (2002). Sources of variability in the clinical photostress test. Ophthalmic and Physiological Optics 2002; 22: 61-67. Marieb EN (2004). Human Anatomy and Physiology, 6th edition. San Francisco: Pearson Benjamin Cummings, 2004. Marks WB, Dobelle WH and MacNichol EF (1964). Visual pigments of single primate cones. Science 1964; 143: 1181-1183. Marmor MF, Holder GE, Seeliger MW, Yamamoto S (2004). International Society for Clinical Electrophysiology of Vision. Standard for clinical electroretinography (2004 update). Documenta Ophthalmologica 2004; 108: 107-114. Marshall W (2000). Clinical Chemistry. 4th edition. St Louis: C.V. Mosby, 2000. 403 Maumenee AE (1969). Doyne Memorial Lecture. Fluorescein angiography in the diagnosis and treatment of lesions of the ocular fundus. Transactions of the Ophthalmological Societies of the United Kingdom 1969; 88: 529-556. McCall MA, Gregg RG, Merriman K, Goto Y, Peachey NS, Stanford LR (1996). Morphological and physiological consequences of the selective elimination of rod photoreceptors in transgenic mice. Experimental Eye Research 1996; 63: 35-50. McPartland RJ, Foster FG, Kupfer DJ, Weiss BL (1976). Activity sensors for use in psychiatric evaluation. IEEE Transactions on Bio-Medical Engineering 1976; 23: 175-8. Meckel JF (1820). Handbuch der Menschlichen Anatomie. Halle and Berlin, Germany, 1820. [German] Meissl H, Nakamura T, Thiele G (1986). Neural response mechanisms in the photoreceptive pineal organ of goldfish. Comparative Biochemistry and Physiology. A, Comparative Physiology 1986; 84: 467-473. Meissl H, Yanez J (1994). Pineal photosensitivity. A comparison with retinal photoreception. Acta Neurobiologiae Experimentalis 1994; 54 Supplement: 19- 29. Melyan Z, Tarttelin EE, Bellingham J, Lucas RJ, Hankins MW (2005). Addition of human melanopsin renders mammalian cells photoresponsive. Nature 2005; 17; 433: 741-745. Middleton B, Arendt J, Stone BM (1996). Human circadian rhythms in constant dim light (8 lux) with knowledge of clock time. Journal of Sleep Research 1996; 5: 69-76. 404 Middleton B, Arendt J, Stone BM (1997). Complex effects of melatonin on human circadian rhythms in constant dim light. Journal of Biological Rhythms 1997; 12: 467-477. Milder B, Rubin ML (2004). The fine art of prescribing glasses without making a spectacle of yourself. 3rd edition. Gainesville, Florida: Triad Publishing Company, 2004. Miller AM, Obermeyer WH, Behan M, Benca RM (1998). The superior colliculus-pretectum mediates the direct effects of light on sleep. Proceedings of the National Academy of Sciences of the United States of America 1998; 95: 8957-8962. Miller NR, Newman NJ (1999). The Essentials: Walsh and Hoyt's Clinical Neuro-ophthalmology. 5th edition 1999. Baltimore: Williams and Wilkins, 1999. Mirza KA (1997) and the Research Committee of the University of Strasbourg. Imam Jafar Ibn Muhammad As-Sadiq A.S.: The Great Muslim Scientist and Philosopher. 2nd edition. Ontario: Willowdale, 1997. Moore RY, Lenn NJ (1972). A retinohypothalamic projection in the rat. The Journal of Comparative Neurology 1972; 146: 1-14. Morrell AJ, Thompson DA, Gibson JM, Kritzinger EE, Drasdo N (1991). Electroretinography as a prognostic indicator or neovascularisation in CRVO. Eye 1991; 5: 362-368. Morruzzi G, Magoun HW (1949). Brain stem reticular formation and activation of the EEG. Electroencephalography and Clinical Neurophysiology 1949; 1:455-473. 405 Moseley M, Fielder AR (2001). Improvement in amblyopic eye function and contralateral eye disease: evidence of residual plasticity. Lancet 2001; 357: 902-904. Moseley MJ, Stewart CE, Fielder AR, Stephens DA; MOTAS cooperative (2006). Intermediate spatial frequency letter contrast sensitivity: its relation to visual resolution before and during amblyopia treatment. Ophthalmic and Physiological Optics 2006; 26:1-4. Motulsky H, Christapoulos A (2004). Fitting Models to Biological Data Using Linear and Nonlinear Regression. New York: Oxford University Press, 2004 Mourits MP, Suttorp-Schulten MS, Tijssen RO, Apkarian P (1990). Contrast sensitivity and the diagnosis dysthyroid optic neuropathy. Documenta Ophthalmologica 1990; 74: 329-335. Mrosovsky N, Lucas RJ, Foster RG (2001). Persistence of masking responses to light in mice lacking rods and cones. Journal of Biological Rhythms. 2001; 16: 585-588. Muckli L, Kiess S, Tonhausen N, Singer W, Goebel R, Sireteanu R (2006). Cerebral correlates of impaired grating perception in individual, psychophysically assessed human amblyopes. Vision Research 2006; 46: 506- 526. Muller J (1826). Zur vergleichenden Physiologie des Gesichtssinnes des Menschen und der Thiere, nebst einem Versuch fiber die Bewegungen der Augen und den menschlichen Blick; Leipzig: Knobloch, 1826. [German] Muller J (1838): Handbuch der Physiologie des Menschen fur Vorlesungen. 2"1 edition. Coblenz, Germany: Holscher, 1838. [German] 406 Murray RB, Loughnane LH (1981). Infrared video pupillometry: a method used to measure the pupillary effects of drugs in small laboratory animals in real time. Journal of Neuroscience Methods 1981; 3: 365-375. Nagasaki H. Shinagawa K. Mochizuki M (1998). Risk factors for proliferative vitreoretinopathy. Progress in Retinal & Eye Research 1998; 17: 77-98. Naka KI, Rushton WA (1966). An attempt to analyse colour reception by electrophysiology. Journal of Physiology 1966; 185: 556-586. Nelson DE, Takahashi JS (1991a). Comparison of visual sensitivity for suppression of pineal melatonin and circadian phase-shifting in the golden hamster. Brain Research 1991; 554: 272-277. Nelson DE, Takahashi JS (1991b). Sensitivity and integration in a visual pathway for circadian entrainment in the hamster (Mesocricetus auratus). Journal of Physiology 1991; 439: 115-145. Newsome DA (1971). Afterimage and pupillary activity following strong light exposure. Vision Research 1971; 11: 275-288. Newton I (1730). The Mathematical Principles of Natural Philosophy; Newton I. Opticks: Or a treatise of the Reflexions, Refractions, Inflexions and Colours of Light. Also Two Treatises of the Species and Magnitude of Curvilinear Figures. London: S. Smith & B. Walford, 1704. Reprint of the 1730 edition, the last corrected by the author. New York: McGraw-Hill, 1931; New York: Dover, 1952. Nishino H, Kiyomi K, Brooks CM (1976). The role of suprachiasmatic nuclei of the hypothalamus in the production of circadian rhythm. Brain Research 1976; 112: 45-59. 407 Nisida I, Okada H (1959). On the role of the pupillo-constrictory centers in the adrenalin-mydriasis and -miosis. The Japanese Journal of Physiology 1959; 9: 348-358. Nisida I, Okada H (1960). The activity of the pupillo-constrictory centers. The Japanese Journal of Physiology 1960; 10: 64-72. Nisida I, Okada H, Nakano 0 (1976). The activity of the ciliospinal centers and their inhibition in pupillary light reflex. The Japanese Journal of Physiology 1960; 10: 73-84. Nork TM, Poulsen GL, Nickells RW, Ver Hoeve JN, Cho NC, Levin LA, Lucarelli MJ (2000). Protection of ganglion cells in experimental glaucoma by retinal laser photocoagulation. Archives of Ophthalmology 2000; 118: 1242- 1250. Nork TM (2000). Acquired color vision loss and a possible mechanism of ganglion cell death in glaucoma. Transactions of the American Ophthalmological Society 2000; 98: 331-363. Norton TN, Corliss DA, Bailey JE (2002). The Psychophysical Measurement of Visual Function. r edition. Woburn, MA, United States of America: Butterworth-Heinemann, 2002. Ohba N, Alpern M (1972). Adaptation of the pupil light reflex. Vision Research 1972; 12: 953-967. Okada H, Nakano 0, Nisida I (1960a). Effects of sciatic stimulation upon the efferent impulses in the long ciliary nerve of the cat. The Japanese Journal of Physiology 1960; 10:327-339. 408 Okada H, Nakano 0, Okamoto K, Nakayama K, Nisida I (1960b). The central path of the light reflex via the sympathetic nerve in the cat. The Japanese Journal of Physiology 1960; 10: 646-658. Orcutt JC (1990). Orbital Foreign Bodies. In: Linberg JV (editor).Oculoplastic and Orbital Emergencies, pp. 183-197. Norwalk, Connecticut: Appleton & Lange, 1990. Osterberg G (1935). Topography of the layer of rods and cones in the human retina, Acta Ophthalmologica 1935; 6 (Supplement): 1-103. Owens DA (1980). A comparison of accommodative responsiveness and contrast sensitivity for sinusoidal gratings. Vision Research 1980; 20: 159-167. Owman C, Edvinsson L, Lindvall M, Sjoberg NO (1982). Influence of external light conditions on norepinephrine levels in organs innervated by sympathetic nerves from different levels. Brain Research Bulletin 1982; 9: 777-779. Ozolinsh M, Colomb M, Ikaunieks G, Karitans V (2006). Color stimuli perception in presence of light scattering. Visual Neuroscience 2006; 23: 597- 601. Pacheco-Cutillas M, Edgar DF, Sahraie A (1999). Acquired colour vision defects in glaucoma-their detection and clinical significance. The British Journal of Ophthalmology 1999; 83: 1396-1402. Pak WL, Ebrey TG (1965). Visual receptor potential observed at sub-zero temperatures. Nature 1965; 205: 484-486. Pak WL, Ebrey TG (1966). Early receptor potentials of rods and cones in rodents. The Journal of General Physiology 1966; 49: 1199-1208. 409 Panda S, Sato TK, Castrucci AM, Rollag MD, DeGrip WJ, Hogenesch JB, Provencio I, Kay SA (2002). Melanopsin (Opn4) requirement for normal light- induced circadian phase shifting. Science 2002; 298: 2213-2216. Papakostopoulos D, Bloom PA, Grey RH, Dean Hart JC (1992). The electro- oculogram in central retinal vein occlusion. The British Journal of Ophthalmology 1992; 76: 515-519. Pare A (1585). The Apologie and Treatise of Ambroise Pare. Translated and published in England in 1634, version edited by Keynes G. pp. 184-185. Chicago: University of Chicago Press, 1952. Originally published in 1585. Parisi V, Bucci MG (1992). Visual evoked potentials after photostress in patients with primary open-angle glaucoma and ocular hypertension. Investigative Ophthalmology and Visual Science 1992; 33: 436-442. Parisi V (2001). Electrophysiological evaluation of the macular cone adaptation: VEP after photostress. A review. Documenta Ophthalmologica 2001; 102: 251-262. Partridge JC, De Grip WJ ( 1991). A new template for rhodopsin (vitamin Al based)visual pigments. Vision Research 1991; 31: 619-630. Pasik T, Pasik P, Hamori J (1973). Nucleus of the accessory optic tract. Light and electron microscopic study in normal monkeys and after eye enucleation. Experimental Neurology 1973; 41: 612-627. Passatore M (1976). Physiological characterization of efferent cervical sympathetic fibers influenced by changes of illumination. Experimental Neurology 1976; 53: 71-81. Passatore M, Pettorossi VE (1976). Efferent fibers in the cervical sympathetic nerve influenced by light. Experimental Neurology 1976; 52: 66-82. 410 Passatore M, Pettorossi VE, Casoni RP (1977). Sympathetic preganglionic pupillodilator fibres in the light reflex. Experientia 1977; 33: 218-219. Pearson P, Swanson WH, Fellman RL (2001). Chromatic and achromatic defects in patients with progressing glaucoma. Vision Research 2001; 41: 1215-1227. Peirson S, Foster RG (2006). Melanopsin: another way of signalling light. Neuron 2006; 49: 331-339. Peirson SN, Thompson S, Hankins MW, Foster RG (2005). Mammalian photoentrainment: results, methods, and approaches. Methods in Enzymology 2005; 393: 697-726. Penfold PL, Provis JM (2005). Macular Degeneration. Berlin-Heidelberg: Springer-Verlag, 2005 Perry VH, Cowey A (1984). Retinal ganglion cells that project to the superior colliculus and pretectum in the macaque monkey. Neuroscience 1984; 12: 1125-1137. Perry VH, Oehler R, Cowey A (1984). Retinal ganglion cells that project to the dorsal lateral geniculate nucleus in the macaque monkey. Neuroscience 1984; 12: 1101-1123. Polyak SL (1941). The Retina. Chicago: University of Chicago Press, 1941. Polyak S (1957). In: Kiiver H (editor). The Vertebrate Visual System. Chicago: University of Chicago Press, 1957. 411 Potts MJ, Fells P, Falcao-Reis F, Bucetti S, Arden GB (1990). Colour contrast sensitivity, pattern ERGs and cortical evoked potentials in dysthyroid optic neuropathy. Investigative Ophthalmology and Visual Science 1990; 31:189. Price MJ, Thompson HS, Judisch GF and Corbett JJ (1985). Pupillary constriction to darkness. The British Journal of Ophthalmology 1985; 69: 205- 211. Provencio I, Foster RG (1995). Circadian rhythms in mice can be regulated by photoreceptors with cone-like characteristics. Brain Research 1995; 694: 183- 190. Provencio I, Kay SA (2002). Melanopsin (Opn4) requirement for normal light- induced circadian phase shifting. Science 2002; 298: 2213-2216. Provencio I, Cooper HM, Foster RG (1998b). Retinal projections in mice with inherited retinal degeneration: implications for circadian photoentrainment. The Journal of Comparative Neurology 1998; 395: 417-439. Provencio I, Jiang G, De Grip WJ, Hayes WP, Rollag MD (1998a). Melanopsin: An opsin in melanophores, brain, and eye. Proceedings of the National Academy of Sciences of the United States of America 1998; 95: 340- 345. Provencio I, Wong S, Lederman AB, Argamaso SM, Foster RG (1994). Visual and circadian responses to light in aged retinally degenerate mice. Vision Research 1994; 34: 1799-1806. Provencio I, Rodriguez IR, Jiang G, Hayes WP, Moreira EF, Rollag MD (2000). A novel human opsin in the inner retina. The Journal of Neuroscience 2000; 20: 600-605. 412 Qiu X, Kumbalasiri T, Carlson SM, Wong KY, Krishna V, Provencio I, Berson DM (2005). Induction of photosensitivity by heterologous expression of melanopsin. Nature 2005; 433: 745-749. Quigley HA (2001). Selective citation of evidence regarding photoreceptor loss in glaucoma. Archives of Ophthalmology 2001; 119: 1390-1391. Rajan MS, Bremner FD, Riordan-Eva P (2002). Pupil perimetry in the diagnosis of functional visual field loss. Journal of the Royal Society of Medicine 2002; 95: 498-500. Ralph MR, Foster RG, Davis FC, Menaker M (1990). Transplanted suprachiasmatic nucleus determines circadian period. Science 1990; 247: 975- 978. Ranson SW, Clarke SL (1959). The Anatomy of the Nervous System: Its Development and Function. 10th edition. Philadelphia: Saunders, 1959. Rashed R (1993). Geometrie et dioptrique au Xe siecle: Ibn Sahl, al-Quhi et Ibn al-Haytham. Paris: Les Belles Lettres 1993. Rea MS, Figueiro MG, Bullough JD, Bierman A (2005). A model of phototransduction by the human circadian system. Brain Research Reviews 2005; 50: 213-228. Reeves C, Taylor DSI (2004). A history of the optic nerve and its diseases. Eye 2004; 18: 1096-1109. Renard G, Massonnet-Naux ( 1951). La synergie pupillaire a la convergence. Arch.Ophtal.Rev.gen.Ophtal., Paris 1952; 11:137-145 [as cited by Loewenfeld IE, Lowenstein 0, 1999]. 413 Restori M (1998). Diagnostic Ultrasound in Vitreoretinal Disease. In: Hamilton AMP, Gregson R, Fish GE (editors). Text Atlas of Retina. London: Martin Dunitz, 1998. Richter K, Peschke E, Peschke D (2000). A neuroendocrine releasing effect of melatonin in the brain of an insect, Periplaneta americana (L.). The Journal of Pineal Research 2000; 28: 129-135. Robinson J, Fielder AR (1960). Pupillary diameter and reaction to light in preterm neonates. Archives of Disease in Childhood 1990; 65: 35-38. Robinson J, Bayliss SC, Fielder AR (1991). Transmission of light across the adult and neonatal eyelid in vivo. Vision Research 1991; 31: 1837-1840. Robinson J, Fielder AR (1992). Light and the immature visual system. Eye 1992; 6: 166-172. Robson JG, Graham N (1981). Probability summation and regional variation in contrast sensitivity across the visual field. Vision Research 1981; 21: 409- 418. Rodieck RW (1998). The First Steps in Seeing. Sunderland, MA: Sinauer, 1998. Roenneberg T, Foster RG (1997). Twilight times: light and the circadian system. Photochemisty and Photobiology 1997; 66: 549-561. Rohrenbeck J, Wassle H, Boycott BB (1989). Horizontal Cells in the Monkey Retina: Immunocytochemical staining with antibodies against calcium binding proteins. The European Journal of Neuroscience 1989; 1: 407-420. Rollag MD, Provencio I, Sugden D, Green CB (2000). Cultured amphibian melanophores: a model system to study melanopsin photobiology. Methods in Enzymology 2000; 316: 291-309. 414 Ruby NF, Brennan TJ, Xie X, Cao V, Franken P, Heller HC, O'Hara BF (2002). Role of melanopsin in circadian responses to light. Science 2002; 298: 2211 -2213. Rucker CW (1965). Cataract: a historical perspective. Investigative Ophthalmology 1965; 4: 377-383. Ruf T (1999). The Lomb-Scargle periodogram in biological rhythm research: Analysis of incomplete and unequally spaced time-series. Biological Rhythm Research 1999; 30: 178-201. Rushton WA (1956a). The difference spectrum and the photosensitivity of rhodopsin in the living human eye. The Journal of Physiology 1956; 134: 11- 29. Rushton WA (1956b). The rhodopsin density in the human rods. The Journal of Physiology 1956; 134: 30-46. Rushton WA (1963). Effect of instantaneous flashes on adaptation of the eye. Dark adaptation after exposing the eye to an instantaneous flash. Nature 1963; 199: 971-972. Rushton WA (1965a). A foveal pigment in the deuteranope. The Journal of Physiology 1965; 176: 24-37. Rushton WA (1965b). Cone pigment kinetics in the deuteranope. The Journal of Physiology 1965; 176:38-45. Rushton WA (1965c). Stray light and the measurement of mixed pigments in the retina. The Journal of Physiology 1965; 176: 46-55. 415 Rushton WA (1965d). The red-sensitive pigment in normal cones. The Journal of Physiology 1965; 176: 56-72. Rushton WA (1972). Pigments and signals in colour vision. The Journal of Physiology 1972; 220: 1. Rushton WAH (1975). Visual Adaptation. In: Moses RA (editor). Adler's Physiology of the Eye. 6th edition. Saint Louis: The C. V. Mosby Company, 1975. Rushton WA (1977). Visual adaptation. Biophysics of Structure and Mechanism 1977; 3: 159-162. Rushton WAH, Campbell FW (1954). Measurement of rhodopsin in the human eye. Nature 1954; 174: 1096-1097. Rushton WA, Westheimer G (1962). The effect upon the rod threshold of bleaching neighbouring rods. The Journal of Physiology 1962; 164: 318-329. Sadeh A, Hauri PJ, Kripke DF, Lavie P (1995). The role of actigraphy in the evaluation of sleep disorders. Sleep 1995; 18: 288-302. Sadeh A, Acebo C (2002). The role of actigraphy in sleep medicine. Sleep Medicine Reviews 2002; 6: 113-124. Sagan C (1983). Cosmos. Random House, 1983 Sahraie A, Barbur JL (1997). Pupil response triggered by the onset of coherent motion. Graefe's archive for clinical and experimental ophthalmology 1997; 235: 494-500. 416 -Sahraie A, Weiskrantz L, Barbur JL, Simmons A, Williams SC, Brammer MJ (1997). Pattern of neuronal activity associated with conscious and unconscious processing of visual signals. Proceedings of the National Academy of Sciences of the United States of America; 94: 9406-9411. Saini VH, Cohen GH (1979). Using color substitution pupil response to expose chromatic mechanisms. Journal of the Optical Society of America 1979; 69: 1029-1035. Saint-Yves C de (1722). Nouveau Traite des Maladies des Yeus, etc. pp. 37-38. Paris: chez Pierre-Augustin Le Mercier, 1722. [French] Sakurai N, Sasaki M (1998). An activity monitor study on the sleep-wake rhythm of healthy aged people residing in their homes. Psychiatry and Clinical Neurosciences 1998; 52: 253-255. Sarsam DE (2002). The Sussex Pupillometer. Sussex: University of Sussex; PhD thesis, 2002. Sarton G (1927). Introduction to the History of Science. vol. 1. Baltimore: Williams and Wilkins, 1927. Satlin A, Teicher MH, Lieberman HR, Baldessarini RJ, Volicer L, Rheaume Y (1991). Circadian locomotor activity in Alzheimer's disease. Neuropsychopharmacology 1991; 5: 115-126. Scargle JD (1982). Studies in astronomical time series analysis: II. Statistical aspects of spectral analysis of unevenly spaced data. The Astrophysical Journal 1982; 268: 835-853. 417 Scheibel AB (1984). The Brain Stem Reticular Core and Sensory Function. In: Darian-Smith I (editor). Handbook of Physiology. Section I, The Nervous System, Vol. III. Sensory Processes, Part I. Bethesda, Maryland: The American Physiological Society, 1984. Scher A, Tirosh E, Lavie P (1998). The relationship between sleep and temperature revisited: evidence for 12-month-olds: a research note. Journal of Child Psychology and Psychiatry, and Allied Disciplines 1998; 39: 785-788. Schiller PH, Malpeli JG (1977). Properties and tectal projections of monkey retinal ganglion cells. Journal of Neurophysiology 1977; 40: 428-445. Schneck ME, Haegerstrom-Portnoy G (1997). Color vision defect type and spatial vision in the optic neuritis trial. Investigative Ophthalmology and Visual Science 1997; 38: 2278-2289. Schnurrbusch UE, Welt K, Horn LC, Wiedemann P, Wolf S (2001). Histological findings of surgically excised choroidal neovascular membranes after photodynamic therapy. The British Journal of Ophthalmology 2001; 85: 1086-1091. Schor CM, Ciuffreda KJ (1983). Vergence Eye Movements: Basic and Clinical Aspects. Oxford: Butterworth-Heinemann Limited, 1983. Schubert E, Thoss F (1967). Influence of different adaptation on the threshold as well as the stimulus strength dependency of latency and amplitude in the human consensual pupillary reflex. Pfliigers Archiv fiir die gesamte Physiologic des Menschen and der Tiere 1967; 294: 28-39. [ German] Schwartz SH (2004). Visual Perception: a Clinical Orientation. 3rd edition. USA: McGraw-Hill, 2004 418 Schweitzer NM (1956). Threshold measurements on the light reflex of the pupil in the dark adapted eye. Documenta Ophthalmologica. Proceedings series 1956; 10: 1-78. Schweitzer NM, Bouman MA (1958a). Differential threshold measurements on the light reflex of the human pupil. A.M.A. Archives of Ophthalmology 1958; 59: 541-550. Schweitzer NM, Bouman MA (1958b). Differential threshold measurement on the light reflex of the human pupil. Ophthalmologica 1958; 135: 298-300. Sekaran S, Foster RG, Lucas RJ, Hankins MW (2003). Calcium imaging reveals a network of intrinsically light-sensitive inner-retinal neurons. Current Biology 2003; 13: 1290-1298. Serra A, Zucca I, Tanda A, Piras V, Fossarello M (1998). Blue-yellow perimetry in patients with ocular hypertone. Acta ophthalmologica Scandinavica. Supplement 1998; 227: 24-27. Sethi CS, Lewis GP, Fisher SK, Leitner WP, Mann DL, Luthert PJ, Charteris DG (2005).Glial remodeling and neural plasticity in human retinal detachment with proliferative vitreoretinopathy. Investigative Ophthalmology and Visual Science 2005; 46: 329-342. Severin SL, Tour RL, Kershaw RH (1967). Macular function and the photostress test 1. Archives of Ophthalmology 1967; 77: 2-7. Shand J, Foster RG (1999). The Extraretinal Photoreceptors of Non- Mammalian Vertebrates. In: Archer SN, Djamgoz MBA, Loew ER, Partridge JC, Valerga S (editors). Adaptive Mechanisms in the Ecology of Vision. Boston: Kluwer Academic, 1999. 419 Short RV (1993). Melatonin: hormone of darkness. British Medical Journal 1993; 307: 952-953. Siegmund R, Tittel M and Schiefenhovel W (1998). Activity monitoring of the inhabitants in Tauwema, a traditional Melanesian village: rest/activity behaviour of Trobriand islanders (Papua New Guinea). Biological Rhythm Research 1998; 2949-2959. Simpson JI (1984). The accessory optic system. Annual Review of Neuroscience 1984; 7: 13-41. Sinclair SH, Azar-Cavanagh M, Soper KA, Tuma RF, Mayrovitz HN (1989). Investigation of the source of the blue field entoptic phenomenon. Investigative Ophthalmology and Visual Science 1989; 30: 668-673. Slooter J, van Norren D (1980). Visual acuity measured with pupil responses to checkerboard stimuli. Investigative Ophthalmology and Visual Science 1980; 19: 105-108. Smith JD, Ichinose LY, Masek GA, Watanabe T, Stark L (1968). Midbrain single units correlating with pupil response to light. Science 1968; 162: 1302- 1303. Smith SA, Smith SE (1980). Contraction anisocoria: nasal versus temporal illumination. The British Journal of Ophthalmology 1980; 64: 933-934. Smith SA, Ellis CJ, Smith SE (1979). Inequality of the direct and consensual light reflexes in normal subjects. The British Journal of Ophthalmology 1979; 63: 523-527. Snell RS (1987). Clinical Neuroanatomy for Medical Students. 2nd edition. Boston and Toronto: Little, Brown, and Company, 1987. 420 Snell RS, Lemp MA (1997). Clinical Anatomy of the Eye. 2nd edition. Malden, Massachusetts: Blackwell Scientific Incorporated, 1997. Sokolove PG, Bushell WN (1978). A chi square periodogram: its utility for the analysis of circadian rhythms. Journal of Theoretical Biology 1978; 72: 131- 160. Soucy E, Wang Y, Nirenberg S, Nathans J, Meister M (1998). A novel signaling pathway from rod photoreceptors to ganglion cells in mammalian retina. Neuron 1998; 21: 481-493. Spurny RC, Zildivar R, Belcher CD, Simmons RJ (1986). Instruments for predicting visual acuity. Archives of Ophthalmology 1986; 104: 196. Stabell U, Stabell B (1989). Dark-adaptation mechanisms of the long-wave foveal cones. Scandinavian Journal of Psychology 1989; 30: 207-219. Stabell B, Stabell U (2002). Effects of rod adaptation on color perception with light adaptation. Journal of the Optical Society of America 2002; 19: 1249- 1258. Stedman TL (1995). Reaction, eye-closure pupil, pp. 1503, Stedman's Medical Dictionary. Baltimore: Williams and Wilkins, 1995. Stockman A, Sharpe LT (1998). Human cone spectral sensitivities: a progress report. Vision Research 1998; 38: 3193-3206. Stockman A, Sharpe LT, Fach C (1999). The spectral sensitivity of the human short-wavelength sensitive cones derived from thresholds and color matches. Vision Research 1999; 39: 2901-2927. Sun F, Stark L (1983). Pupillary escape intensified by large pupillary size. Vision Research 1983; 23: 611-615. 421 Suttorp-Schulten MS, Tijssen R, Mourits MP, Apkarian P (1993). Contrast sensitivity function in Graves' ophthalmopathy and dysthyroid optic neuropathy. The British Journal of Ophthalmology 1993; 77: 709-712. Szlyk JP, Fishman GA, Alexander KR, Peachey NS, Derlacki DJ (1993). Clinical subtypes of cone-rod dystrophy. Archives of Ophthalmology 1993; 111: 781-788. Taek K, Seung MH, Kyung CY (2001). ElectroretMographic studies in rhegmatogenous retinal detachment before and after reattachment surgery. Korean Journal of Ophthalmology 2001; 15: 18-127. Taillard J, Philip P, Chastang JF, Bidelac B (2004). Journal of Biological Rhythms 2004; 19: 76-86. Tanner V, Tregear SJ, Ripley LG, Vickers SF (1995). Automated achromatic contrast and chromatic discrimination sensitivity testing in dysthyroid optic neuropathy. Eye 1995; 9: 352-357. Taylor D (2005). Optic nerve axons: life and death before birth. Eye 2005; 19: 499-527. ten Doesschate J, Alpern M (1967a). Effect of photoexcitation of the two retinas on pupil size. Journal of Neurophysiology 1967; 30: 562-576. ten Doesschate J, Alpern M (1967b). Influence of asymmetrical photoexcitation of the two retinas on pupil size. Journal of Neurophysiology 1967; 30: 577-585. The Retina Society Terminology Committee (1983). The classification of retinal detachment with proliferative vitreoretinopathy. Ophthalmology 1983; 90: 121- 125. 422 Thapan K, Arendt J, Skene DJ (2001). An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans. The Journal of Physiology 2001; 535: 261-267. Theiss P, Grusser OJ (1994). Vision and cognition in the natural philosophy of Albert the Great (Albertus Magnus). Documenta Ophthalmologica 1994; 86: 123-151. Thompson HS (1966). Afferent pupillary defects. Pupillary findings associated with defects of the afferent arm of the pupillary light reflex arc. American Journal of Ophthalmology 1966; 62: 860-873. Thompson HS (1987). The pupil cycle time. Journal of Clinical Neuro- ophthalmology 1987; 7: 38-39. Thompson HS (2003). The vitality of the pupil: a history of the clinical use of the pupil as an indicator of visual potential. Journal of Neuro-ophthalmology 2003; 23: 213-224. Thompson HS, Watzke RC, Weinstein JM (1980). Pupillary dysfunction in macular disease. Transactions of the American Ophthalmological Society 1980; 78: 311-317. Thorell LG, De Valois RL, Albrecht DG (1984). Spatial mapping of monkey V1 cells with pure color and luminance stimuli. Vision Research 1984; 24: 751 - 769. Trachsel L, Tobler I, Borbely AA (1986). Sleep regulation in rats: effects of sleep deprivation, light, and circadian phase. American Journal of Physiology 1986; 251: 1037-1044. 423 Tregear SJ, Ripley LG, Knowles PJ, Gilday RT, de Alwis DV, Reffin JP (1994). Automated tritan discrimination sensitivity: a new clinical technique for the effective screening of severe diabetic retinopathy. International Journal of Psychophysiology 1994; 16: 191-198. Trejo LJ, Cicerone CM (1982). Retinal sensitivity measured by the pupillary light reflex in RCS and albino rats. Vision Research 1982; 22: 1163-1171. Trejo LJ, Cicerone CM (1984). Cells in the pretectal olivary nucleus are in the pathway for the direct light reflex of the pupil in the rat. Brain Research 1984; 300: 49-62. Trevethan CT, Sahraie A, Weiskrantz L (2007). Form discrimination in a case of blindsight. Neuropsychologia 2007; 45: 2092-2103. Ts'o DY, Gilbert CD (1988). The organization of chromatic and spatial interactions in the primate striate cortex. Journal of Neuroscience 1988; 8: 1712-1727. Tzekov RT, Sohocki MM, Daiger SP, Birch DG (2000). Visual phenotype in patients with Arg41Gln and a1a196+ 1 by mutations in the CRX gene. Ophthalmic Genetics 2000; 21: 89-99. Ukai K (1985). Spatial pattern as a stimulus to the pupillary system. Journal of the Optical Society of America 1985; 2: 1094-1100. Utiger RD (1992). Melatonin — the hormone of darkness. The New England Journal of Medicine 1992; 327: 1377-1379. Van den Berg (1986). Importance of pathological intraocular light scatter for visual disability. Documenta Ophthalmologica 1986; 61: 327-33. 424 Van Dongen HPA, Olofsen E, VanHartevelt JH and Kruyt EW (1997). Periodogram analysis of unequally spaced data: The Lomb method. Leiden: Department of Physiology, Leiden University, 1997. Van Dongen HPA, Olofsen E, VanHartevelt JH and Kruyt EW (1999a). Searching for biological rhythms: peak detection in the periodogram of unequally spaced data. Journal of Biological Rhythms 1999; 14: 617-620. Van Dongen HPA, Olofsen E, VanHartevelt JH, Kruyt EW (1999b). A procedure of multiple period-searching in unequally spaced time-series with the Lomb-Scargle method. Biological Rhythm Research 1999; 30: 149-177. Van Haesendonck E, Missotten L (1993). .A subgroup of bipolar cells in human retina is GABA-immunoreactive. Neuroscience Letters 1993; 161: 187-190. van Norren D, Vos JJ (1974). Spectral transmission of the human ocular media. Vision Research 1974; 14: 1237-1244. Velmans M (2000). Understanding Consciousness. London: Routledge, 2000. Verdon W, Howarth PA (1988). The pupil's response to short wavelength cone stimulation. Vision Research 1988; 28: 1.119-1128. Vivien-Roels B, Pevet P, Beck 0, Fevre-Montange M (1984). Identification of melatonin in the compound eyes of an insect, the locust (Locusta migratoria), by radioimmunoassay and gas chromatography-mass spectrometry. Neuroscience Letters 1984; 49:153-157. Vilupuru AS, Roorda A, Glasser A (2004). Spatially variant changes in lens power during ocular accommodation in a rhesus monkey eye. Journal of Vision 2004; 4: 299-309. 425 Vingrys AJ, Demirel S (1993). False-response monitoring during automated perimetry. Optometry and Vision Science 1998; 75: 513-517. Vollrath L (1981). The Mammalian Epiphysis Cerebri. In: The Pineal Organ. pp. 5-320. Berlin: Springer-Verlag, 1981. Vollrath L, Spessert R, Kratzsch T, Keiner M, Hollmann H (1997). No short- term effects of high frequency electromagnetic fields on the mammalian pineal gland. Bioelectromagnetics 1997; 18: 376-387. von Graefe A (1855). Ueber ein einfaches Mittel, Simulation einseitiger Amaurose zu entdecken, nebst Bemerkungen ueber die Pupillar-Contraction bei Erblindeten. Albrecht von Graefe's Archiv fiir Ophthalmologic 1855; 2:266-72. [German] von Helmholtz H (1909-1911). Handbuch der Physiologischen Optik. 2nd edition, 1909-1911. Originally published in 1859-1867. [Handbook of Physiological Optics. Translated and edited from the 2nd German edition by Southall JPC. New York: The Optical Society of America, 1924-1925 (reprinted in New York: Dover, 1962)] Voo I, Mavrofrides EC, Puliafito CA (2004). Clinical applications of optical coherence tomography for the diagnosis and management of macular diseases. Ophthalmology Clinics of North America 2004; 17: 21-31. Wagman IH, Gullberg JE (1942). The relationship between monochromatic light and pupil diameter. The low intensity visibility curve as measured by pupillary measurements. American Journal of Physiology 1942; 137: 769-778. Wald G ( I 968a). The molecular basis of visual excitation. Nature 1968; 219: 800-807. 426 Wald G (1968b). Molecular basis of visual excitation. Science 1968; 162: 230- 239. Wang Y, Macke JP, Merbs SL, Zack DJ, Klaunberg B, Bennett J, Gearhart J, Nathans J (1992). A locus control region adjacent to the human red and green visual pigment genes. Neuron 1992; 9: 429-440. Wassle H, Wing RB (1980). The retinal projection to the superior colliculus in the cat: a quantitative study with HRP. The Journal of Comparative Neurology 1980; 190: 333-356. Webster JG, Cohen GH, Boynton RM (1968). Optimizing the use of the criterion response for the pupil light reflex. Journal of the Optical Society of America; 58: 419-424. Weinstein JM, Van Gilder JC, Thompson HS (1980). Pupil cycle time in optic nerve compression. American Journal of Ophthalmology 1980; 89: 263-267. Weiskrantz L. (1986). Blindsight: A case study and implications. Oxford: Clarendon Press, 1986. Weiskrantz L (1987). Residual vision in a scotoma. A follow-up study of form' discrimination. Brain 1987; 110: 77-92. Weiskrantz L, Barbur JL, Sahraie A. (1995). Parameters affecting conscious versus unconscious visual discrimination with damage to the visual cortex (V1). Proceedings of the National Academy of Sciences of the United States of America 1995; 92: 6122-6126. Weiskrantz L, Cowey A, Barbur JL (1999). Differential papillary constriction and awareness in the absence of striate cortex. Brain 1999; 122: 1533-1538. Weiskrantz L (2004). Roots of blindsight. Progress in Brain Research 2004; 144: 229-241. 427 Weiss AH, Kelly JP (2004). Spatial-frequency-dependent changes in cortical activation before and after patching in amblyopic children. Investigative Ophthalmology and Visual Science 2004; 45: 3531-3537. Westheimer G (1964). Pupil size and visual resolution. Vision Research 1964; 4: 39-45. Wheatstone C (1838). Contributions to the physiology of vision. I. On some remarkable and hitherto unobserved, phenomena of binocular vision. Philosophical Transactions of the Royal Society of London 1838; 128: 371- 394. Whiteley SJ, Young MJ, Litchfield TM, Coffey PJ, Lund RD (1998a). Changes in the pupillary light reflex of pigmented royal college of surgeons rats with Age. Experimental Eye Research 1998; 66: 719-730. Whiteley SJ, Sauve Y, Aviles-Trigueros M, Vidal-Sant M, Lund RD (1998b). Extent and duration of recovered pupillary light reflex following retinal ganglion cell axon regeneration through peripheral nerve grafts directed to the pretectum in adult rats. Experimental Neurology 1998; 154: 560-572. Whitten DN, Brown KT (1973). Photopic suppression of monkey's rod receptor potential, apparently by a cone-initiated lateral inhibition. Vision Research 1973; 13: 1629-1658. Wickham L, Sethi CS, Lewis GP, Fisher SK, McLeod DC, Charteris DG (2006). Glial and neural response in short-term human retinal detachment. Archives of Ophthalmology 2006; 124: 1779-1782. Wolffsohn JS, Anderson SJ, Mitchell J, Woodcock A, Rubinstein M, Ffytche T, Browning A, Willbond K, Amoaku WM, Bradley C (2006). Effect of age related macular degeneration on the Eger macular stressometer photostress recovery time. British Journal of Ophthalmology 2006; 90: 432-434. 428 Wolfley D (1997). Evaluation of Orbital Disease. In: Walsh TJ (editor). Neuro- ophthalmology: Clinical Signs and Symptoms. 4th edition. Baltimore: Williams & Wilkins, 1997. Wong-Riley MTT (1979). Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Research 1979; 171: 11-28. Wright KP Jr, Czeisler CA (2002). Absence of circadian phase resetting in response to bright light behind the knees. Science 2002; 297: 571. Wu G, Weiter JJ, Santos S, Ginsburg L, Villalobos R (1990). The macular photostress test in diabetic retinopathy and age-related macular degeneration. Archives of Ophthalmology 1990; 108: 1556-1558. Wulff K, Joyce E, Middleton B, Dijk DJ, Foster RG (2006). The suitability of actigraphy, diary data, and urinary melatonin profiles for quantitative assessment of sleep disturbances in schizophrenia: a case report. Chronobiology International 2006; 23: 485-495. Wyatt HJ, Musselman JF (1981). Pupillary light reflex in humans: evidence for an unbalanced pathway from nasal retina, and for signal cancellation in brainstem. Vision Research 1981; 21: 513-525. Wyszecki G, Stiles WS (1982). Color Science — Concepts and methods. Quantitative Data and Formulae. 2nd edition. New York: John Wiley and Sons, 1982. Yagasaki K, Jacobson SG (1989). Cone-rod dystrophy. Phenotypic diversity by retinal function testing. Archives of Ophthalmology 1989; 107: 701-708. 429 Yamada Y, Koizumi A, Iwasaki E, Watanabe S, Kaneko A (2002). Propagation of action potentials from the soma to individual dendrite of cultured rat amacrine cells is regulated by local GABA input. Journal of Neurophysiology 2002; 87: 2858-2866. Yamazaki S, Goto M, Menaker M (1999). No evidence for extraocular photoreceptors in the circadian system of the Syrian hamster. Journal of Biological Rhythms 1999; 14: 197-201. Yellott JI, Wandell B, Cornsweet T (1984). The Beginnings of Visual Perception: The Retinal Image and Its Initial Encoding. In: Darian-Smith I (editor). Handbook of Physiology, Section 1: The Nervous System, Vol. III, Part 2. I. Bethesda, Maryland: The American Physiological Society, 1984. Yoshitomi T, Matsui T, Tanakadate A, Ishikawa S (1999). Comparison of threshold visual perimetry and objective pupil perimetry in clinical patients. Journal of Neuro-ophthalmology 1999; 19: 89-99. Young RS, Han BC, Wu PY. (1993). Transient and sustained components of the pupillary responses evoked by luminance and color. Vision Research 33: 437- 446. Young RS, Alpern M (1980). Pupil responses to foveal exchange of monochromatic lights. Journal of the Optical Society of America 1980; 70: 697-706. Young RS, Kennish J (1993). Transient and sustained components of the pupil response evoked by achromatic spatial patterns. Vision Research 1993; 33: 2239-2252. Young T (1802). On the theory of light and colours. Philosophical Transactions of the Royal Society of London 1802; 92: 18-21. 430 Young R, Clavadetscher J and Teller D (1987). Screening of red-green color deficient observers using the chromatic pupillary response. Clinical Vision Science 1987; 2: 117-222. Youngstedt SD, Kripke DF, Elliott JA, Klauber MR (2001). Circadian abnormalities in older adults. Journal of Pineal Research 2001; 31: 264-272. Yukie M, Iwai E (1981). Direct projection from the dorsal lateral geniculate nucleus to the prestriate cortex in macaque monkeys. The Journal of Comparative Neurology 1981; 201: 81-97. Zabriskie NA, Kardon RH (1994). The pupil photostress test. Ophthalmology 1994; 101: 1122-1130. Zaidi FH, Cheong-Leen R, Gair EJ, Weir R, Sharkawi E, Lee N, Gregory- Evans K (2004). The Amsler chart is of doubtful value in retinal screening for early laser therapy of subretinal membranes. The West London Survey. Eye 2004; 18: 503-508. Zaidi FH and Moseley MJ (2004). Use of pupil size and reaction to detect orbital trauma during and after surgery. Clinical Otolaryngology and Allied Sciences 2004; 29: 288-289. Zeitzer JM, Khalsa SB, Boivin DB, Duffy JF, Shanahan TL, Kronauer RE, Czeisler CA (2005). Temporal dynamics of late-night photic stimulation of the human circadian timing system. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 2005; 289: 839-844. Zeki S, Watson JD, Lueck CJ, Friston KJ, Kennard C, Frackowiak RS (1991). A direct demonstration of functional specialization in human visual cortex. Journal of Neuroscience 1991; 11: 641-649. 431 Zeki S (1993). A Vision of the Brain. Oxford: Blackwell, 1993. Zuege P, Drance SM (1967). Studies of dark adaptation of discrete paracentral retinal areas in glaucomatous subjects. American Journal of Ophthalmology 1967; 64: 56-63. 432 Downloaded from bjo.bmj.com on 30 April 2007 Subretinal membranes are associated with abnormal degrees of pupil "evasion": an index of clinical macular dysfunction F H Zaidi, F D Bremner, K Gregory-Evans, K D Cocker and M J Moseley Br. J. Ophthalmol. 2006;90;1115-1118 doi:10.1136/bjo.2005.083279 Updated information and services can be found at: http://bjo.bmj.com/cgi/content/ful1/90/9/1115 These include: References This article cites 14 articles, 4 of which can be accessed free at: http://bjo.bmj.com/cgi/content/ful1/90/9/1115#BIBL Rapid responses You can respond to this article at: http://bjo.bmj.com/cgi/eletter-submitJ90/9/1115 Email alerting Receive free email alerts when new articles cite this article - sign up in the box at the service top right corner of the article Topic collections Articles on similar topics can be found in the following collections Vision Research (671 articles) Other ophthalmology (2377 articles) Notes To order reprints of this article go to: http://www.bmjjournals.com/cgi/reprintform To subscribe to British Journal of Ophthalmology go to: http://www.bmjjournals.com/subscriptions/ Downloaded from bjo.bmj.com on 30 April 2007 1115 SCIENTIFIC REPORT Subretinal membranes are associated with abnormal degrees of pupil "evasion": an index of clinical macular dysfunction F H Zaidi, F D Bremner, K Gregory-Evans, K D Cocker, M J Moseley Br J Ophthalmol 2006;90:1115-1118. doi: 10.1136/bjo.2005.083279 committee approval was obtained ( St Mary's NHS Trust). An Aim: To assess whether macular dysfunction caused by information sheet was provided and informed consent unilateral subretinal neovascular membranes (SRNs) is obtained (Declaration of Helsinki). Patients with diabetes associated with pupil "evasion" (that is, increased initial mellitus or medication affecting autonomic function were rate of re-dilation following a brief light stimulus). excluded. Methods: Comparative observational series. 20 eyes of 10 Digital recordings of fundus fluorescein angiography (FFA) participants, all with unilateral SRNs and healthy fellow eyes. of suspected subretinal membranes were screened with Dynamic infrared pupillography at seven stimulus intensities diagnoses confirmed by clinicians with special retinal (duration 1100 ms, intensities over 2 log unit range). Pupil expertise at the Western Eye Hospital in London.' evasion ratio (PEVR; defined as the ratio of light response Unilateral SRNs were unusual as most were attributable to amplitude to amount of recovery at the mid-time point of re- age related maculopathy (ARM). Participants underwent a dilation expressed as a percentage) was calculated for each battery of psychophysical and clinical tests including 1ogMAR stimulus intensity (mean of five recordings). acuity, colour discrimination (Ishihara pseudo-isochromatic Results: Inter-eye PEVR is significantly reduced in eyes with plates), perimetry (Humphrey 24-2 and 120-2 full threshold SRN (that is, greater pupil evasion in SRN eyes: range p = test or Goldmann), pupil reactions and swinging flashlight 0.002 to p = 0.05 (paired t test)) and is most apparent at tests, and dilated funduscopy. Efferent lesions were excluded higher stimulus intensities. by clinical signs of autonomic dysfunction using pupillo- Conclusions: PEVR is a novel parameter that is analogous to metric measurements of size, and reactivity in dark and light. the pupil escape ratio, but measured following a short rather Patients abstained from caffeine, stimulants, chocolates, and than a sustained light stimulus. PEVR is significantly altered undue physical exercise for 24 hours before the recordings by macular disease. Clinically PEVR may be used to detect were made, 2 days to 2 weeks after initial tests. Following 20 minutes of dark adaptation the eye was stimulated and occult unilateral or asymmetric maculopathy in situations the pupil recorded with a modified binocular P_Scan 100 such as ocular media opacities like cataract, when pupil infrared pupillometer (Circuit Solve, Cambridge, UK) pre- reactions are unaffected or augmented, while other tests of viously described in the literature (spatial resolution retinal function are diminished. PEVR represents altered >0.01 mm, temporal resolution 20 ms neuronal firing in cones and macular ganglion cells. Following an audible cue, the tested eye was stimulated in non-Maxwellian view using a cathode ray tube display (Iiyama, Japan, Vision Master 506, Model MS103DT). The light stimulus was a square achromatic patch brief light stimulus produces an initial transient or (22x14 degrees). The untested eye was occluded with a "phasic" pupillary constriction.' ' An uninterrupted covering patch. prolonged light stimulus produces a sustained pupil- A An adapting stimulus was presented before each series of lary constriction, followed sometimes as quickly as after a stimuli, as first pupil response to light series can poorly few milliseconds by slow dilation called escape.' ' Escape is represent subsequent reflex behaviour. Seven stimuli in amplified with central field loss.' The present study describes ascending intensity were employed (range 2 log cd m-2 ), a novel form of pupil escape (pupil "evasion") which for the duration 1100 ms each, each presented five times and first time is elicited and measured using a very short repeated with the fellow eye. Presentation of stimuli in stimulus. Evasion was determined in patients with unilateral ascending order of intensity reduces the effect of light subretinal neovascular membranes ( SRNs) allowing assess- adaptation (International Society for Clinical ment of both retinal ganglion cell and grading of visual Electrophysiology of Vision).' Time interval between identical potential.' This may be clinically useful as pupil escape stimuli was >10 seconds. The time taken to record all 70 correlates with the relative afferent pupillary defect (RAPD) reflex responses in both eyes was approximately 30 minutes. and, unlike comparative investigations of retinal and optic nerve function, is undiminished or even augmented by media Recording began 500 ms before stimulus onset and opacities, including common conditions such as cataract that continued for 10 seconds afterwards. Pupil diameter had usually recovered fully to baseline before presentation of a restrict visualisation of the fundus." subsequent stimulus; when full recovery did not occur within the 10 seconds a regression function was used to extrapolate MATERIALS AND METHODS Inclusion and exclusion criteria Abbreviations: ARM, age related maculopathy; FFA, fundus fluorescein Ten consecutive patients with unilateral SRNs and no other angiography; MDA, mid-dilation amplitude; PEVR, pupil evasion ratio; ocular pathology in either eye were recruited. There were PDT, photodynamic therapy; RA, reflex amplitude; RAPD, relative seven women and three men aged 39-82 (mean 73). Ethics afferent pupillary defect; SRNs, subretinal neovascular membranes www.blophthalmol.com Downloaded from bjo.bmj.com on 30 April 2007 1116 Zaidi, Bremner, Gregory-Evans, et al 7.75 E 7.25 r, E 6.75 0 6.25 5.75 CV 0 0 0 V0 43 c‘e 'o 63 P 4' Li? O X43-... cl..... n) -\ r° ri? c'\ r') lime (seconds) 9 26 0 6 6 0 5 5 (3(3 CP QP OP CP 03.1CPSP-503 o ap .fur rp no bCPq 0°rk. 4) (1, r33 O "3' 1 v Neb- ri;,:" 42' rir' 0 lime (ms) lime (ms) Figure 1 Pupil diameter versus time. (Top) 1100 ms stimulus presented at 500 ms. After reflex latency the pupil constricts, peak constriction amplitude (reflex amplitude, RA) reached at 1400 ms (black arrow). The pupil then dilates (max 3400 ms, indicated by vertical green line), with no further dilation for at least 10 seconds from stimulus onset (not shown). Time interval, t, from time of RA (maximum constriction) to onset of maximal dilation is shown by the horizontal green arrow; half time, t V2, is shown by the blue arrow, and corresponds to the mid-ditation amplitude (MDA). PEVR = RA/ MDA x 100. Bottom left, normal pupillogram of participant 4 (stimulus 14 cd m-2). Bottom right, pupillogram from fellow eye of participant 4, but with SRN, same stimulus intensity. PEVR (derived using RAs) correctly identified greater evasion in the pupillogram, bottom right, with the subretinal membrane, despite gross observation suggesting that evasion might be greater in the pupillogram at bottom left. This is because the higher RA in the normal eye, bottom left, in participant 4 at this stimulus intensity creates the false visual impression of evasion, while mathematical calculation of PEVR using the above formula correctly identified the eye at bottom right as the one with the subretinal membrane. the initial resting diameter. In a few cases the reflex response the pupil evasion ratio. Figure 2 (second plot from top) shows was superimposed on a shifting baseline and time to full how this estimate is independent of reflex amplitude and recovery could not be computed—such traces were excluded better serves as an index of pupil evasion (analogous to from the subsequent averaging. Excessive eye movement escape), with low values for PEVR indicating greater evasion. (erratic fixation >1 degree) was rare but the involved Figure 2 (lower two plots) indicates how PEVR estimates pupillograms were excluded. Digital data filtering removed varied considerably between individuals for the same blink artefact and improved signal to noise ratio. A mean stimulus intensity (range <10% to >1000%). pupillogram (average of five recordings) for each of the seven PEVR was consistently smaller in the eyes with SRNs stimulus intensities was computed for each eye of each compared with the healthy fellow eyes ( fig 3). Figure 3 (top) participant. (participant 2) represents a large subfoveal predominantly Pupil waveform parameters were measured with P_Scan classic SRN, 3000 Am in diameter, treated with three courses software using cross hairs, and/or transfer as text (.txt) files of PDT. At all stimulus intensities except 0.6 log cd M-2 the to Microsoft Excel. Reflex amplitude (RA, mm) and mid- PEVR is grossly reduced in the diseased eye (paired t test dilation amplitude (MDA, mm) were calculated from each significant (mean difference in PEVR between SRN and mean pupillogram of each stimulus intensity (fig 1(top)). A control eye = 1954.0%; p = 0.002)) showing macular disease novel index was derived which we term the pupil evasion is associated with significantly greater pupil evasion. In most ratio (PEVR), defined as: PEVR = (RA/MDA) x 100. participants the difference in PEVR was not as significant: a Multiplication by a factor of 100 is in common with previous more typical example is illustrated in figure 3 (middle plot) electrodiagnostic indices facilitating a readily understandable (participant 6) with a large untreated subfoveal age related percentage index.' disciform scar 1500 nm across. Inter-eye PEVR difference was smaller (range 5-30%) but significant (mean difference RESULTS = 12.9%, p = 0.005 (paired t test)). The least convincing case The relation to the RAPD is described (table 1). is nevertheless the participant in whom PEVR was first Mid-dilation amplitude (MDA) was linearly correlated detected ( fig 3, bottom). This participant (4) has a small with the amplitude of the reflex response (RA). Larger reflex (<500 pm) untreated subfoveal SRN of presumed inflam- amplitudes produced by higher intensity stimuli were matory origin. PEVR estimates were distinctly smaller in the associated with greater MDA measurements (fig 2, top). diseased eye at four out of the seven stimulus intensities This co-variation masks any independent effect of macular (mean difference in PEVR = 25%; p = 0.05 (paired t test), or disease on pupil evasion and so MDA estimates were p = 0.04 (paired 1 test) excluding the lowest intensity normalised with respect to reflex amplitude by calculating stimulus). www.bjophthalmol.com Downloaded from bjo.bmj.com on 30 April 2007 Pupillometrey and subretinal membranes 1 1 1 7 Table 1 Clinical features of the 10 participants with unilateral subfoveal subretinal membrane? " Participant Aetiology RAPD Age Sex 1 Infective. Presumed ocular histoplasmosis. No RAPD 39 M Untreated 2 ARM with predominantly classic SRN. In diseased eye 53 F Treated by three courses of PDT 3 ARM with occult membrane. Untreated In diseased eye 53 F 4 Inflammatory. Suspected punctate inner No RAPD 55 F choroidopathy complicated by SRN formation several years previously. Untreated 5 ARM. Classic membrane treated by three In diseased eye 60 F courses of PDT 6 ARM. Occult membrane. Untreated In diseased eye 68 M 7 ARM. Disciform scar. Untreated In diseased eye 69 F 8 ARM. Classic membrane treated by one In diseased eye 71 M course of PDT 9 ARM. Disciform scar. Untreated In diseased eye 77 F 10 Infective. Presumed ocular histoplasmosis No RAPD 82 F ARM, age related maculopathy; PDT, photodynamic therapy; RAPD, relative afferent pupillary defect. Difference in PEVR between diseased and healthy eyes is extreme sizes of SRN. SRNs on FFA ranged from 3000 [inn to slightly more apparent at higher stimulus intensities (fig 2, <500 i.tm in diameter, and the largest and smallest were lower two plots). Greatest inter-eye PEVR differences were associated respectively with the largest and least statistically generally observed at higher stimulus intensities. significant difference in PEVR between fellow eyes, including Compared to fellow control eyes, statistically significant at all stimulus intensities (p<0.002 and 0.05, respectively, reduction in PEVR (greater pupil evasion) was found in all mean difference in inter-eye PEVR 1954.0% and 25.0%, diseased eyes, including participant 4 with a tiny SRN (fig 3). There was no apparent association between reduction in PEVR and SRN aetiology. There is a possible association with 3050 `., 04 2050 I 00.4 LciL2 1050 o3 50 I -0.4 -0.1 0.2 0.6 0.9 1.1 1.4 TT, 73 0.2 Light intensity (log cd M-2) 0.1 - I I I a 0.0rr I I I 200 0.073 0.871 0.930 1.615 1.638 1.881 1.978 180 Constriction amplitude (mm) 160 140 - 500 120 - • 1 1 I I I 1 1 400 -0 100 -0.4 -0.05 0.2 0.6 0.9 1.1 1.4 300 200 Light intensity (log cd m-2) a- • 100 • --•-- Healthy fellow eye 00 • 1 I 1 195 0.5 1.0 1.5 2.0 2.5 -•- SRN 175 Constriction amplitude (mm) 155 600 135 E 024000 115 -0.4 -0.1 0.2 0.6 0.9 1.1 1.4 a_ a 0 - • -31E • Light intensity (log cd m-2) •w -200 -400 I I I I I I 1 -0.4 -0.1 0.2 0.6 0.9 1.1 1 4 Light intensity (log cd m-2) 200 ce 150 100 a 50 0 Light intensity (log cd m2) Figure 3 PEVR versus light intensity. (Top) Participant 2 with a large Figure 2 (Top) MDA versus constriction amplitude (RA) in apatient, age related subfoveal SRN previously treated with PDT. (Middle) (second from top) PEVR (%) versus RA (same patient); (seconal from Participant 6 with a large but untreated age related subfoveal SRN. bottom) mean difference in inter-eye PEVR versus stimulus intensity (Bottom) Participant 4 with a small <500 vtm subfoveal SRN of (mean 10 patients), magnified at the bottom. Error bars signify 95% presumed inflammatory origin, with fundus photograph and FFA confidence intervals (mean (+2 SD). showing very small SRN. www.blophthalmol.com Downloaded from bjo.bmj.com on 30 April 2007 1118 Zaidi, Bremner, Gregory-Evans, et al respectively). This suggests an association between PEVR and ACKNOWLEDGEMENTS visual field may be likely, but x2 analysis of all 10 subjects The authors arc grateful to Professor Alistair Fielder for facilitating showed no significant overall association between size of this work. Professor Stephen Smith provided valued comments. We SRN and reduction in PEVR (p>0.1). There was no appreciate the contribution of the Undo Wing of St Mary's and the association between the degree of loss of visual acuity Western Eye Hospitals who provided a fellowship to the principal investigator (FHZ) helping to support part of this research. We arc (counting fingers or worse in all diseased eyes) and extent also grateful to Mr Karl Southerton, medical photographer, St Mary's of PEVR reduction, presumably as all SRNs involved the and the Western Eye Hospitals, in accessing and processing the fovea. Sex, age, and treatment status had no association with digital files on patients with subretinal membranes. reduction in PEVR in this study. This work was presented at the Association for Research in Vision and Ophthalmology (ARVO), Fort Lauderdale, Florida, United States, DISCUSSION 2004 under the title "Macula dysfunction in subretinal neovascular Pupil escape arises because of both retinal and central membranes results in significant pupillary escape"; The Oxford adaptation processes, and probably includes photopigment Ophthalmological Congress 2004; and some data were included in a presentation the European Association for Vision and Eye Research bleaching, disappearance of the phasic component and (EVER), Vilamoura, Algarve, Portugal, 2004. diminution of the tonic component to the "ON" signal in the afferent limb of the reflex, and resetting of the gain in the reflex loop at the level of the midbrain.' ' Patients with Authors' affiliations lesions of the anterior visual pathways show exaggerated F H Zaidi, K Gregory-Evans, K D Cocker, M 1 Moseley, Department of pupil escape.' Ophthalmology and Visual Neuroscience, Imperial College London, UK Measurement of pupil escape in relation to the peak of the F H Zaidi, K Gregory-Evans, The Western Eye Hospital, St Mary's NHS sustained contraction is ineffective in detecting pupil Trust, London, UK escape.' A shortened time period between stimulus onset F D Bremner, The National Hospital for Neurology and Neurosurgery, to when pupil area was greater than 63% of baseline in 13 out London, and University College London Hospitals, London, UK of 16 eyes with optic neuritis has been taken to indicate K D Cocker, M J Moseley, Department of Optometry and Visual Science, City University, London, UK greater pupil escape, perhaps with central field loss, though confounding may have occurred." When pupil escape is Competing interests: none. expressed as a ratio incorporating the amplitude of the Local regional ethics committee approval was obtained for this study (St preceding phasic contraction, the resultant pupil escape ratio Mary's NHS Trust, London). can be used to indicate greater pupillary escape.' When mid-dilation amplitude is expressed as a function of Correspondence to: Farhan H Zaidi, Department of Ophthalmology, peak phasic constriction with a short duration stimulus, the King's College Hospital, London, UK; [email protected] ratio, called PEVR, is reduced in eyes with SRNs, denoting Accepted for publication 30 November 2005 greater pupil evasion. Concerns about PEVR are that (i) it shows wide variability between eyes (inter-test variability) and fluctuation (intra-test variability), (ii) overlap of PEVR REFERENCES between normal and abnormal eyes; (iii) wide confidence 1 Kardon RH. The pupil. In: Kaufman PL, Alm A, eds. Adler's physiology of the intervals that diminish its differentiating value ( fig 2). eye.Chapter 32.10th ed. St Louis: Mosby, 2003. However, these problems are overcome by the strategy of 2 Bergamin 0, Kardon RH. Greater pupillary escape differentiates central from peripheral visual field loss. Ophthalmology 2002;109:771-80. inter-eye comparison of PEVR between diseased and healthy 3 Bloom PA, Papakostopoulos D, Gogolitsyn Y, et al. Clinical and infrared eyes, where PEVR is greater in normal eyes (less evasion): pupillometry in central retinal vein occlusion. Br J Ophthalmol statistically significant in all participants at all light 1993;77:75-80. 4 Grey RH, Bloom PA. Retinal ischaemia and relative afferent pupillary defects intensities. This is because inter-eye comparison takes into in central retinal vein occlusion. Ear J Ophthalmol 1991;1:85-8. account the similar autonomic arousal and degree of back- 5 Papakostopoulos D, Bloom PA, Grey RH, et al. The electro-oculogram in ground sympathetic tone in the iris muscles of both eyes.' central retinal vein occlusion. Br J Ophthalmol 1992;76:515-19. PEVR correlates with and is more sensitive than the 6 Thompson HS. The vitality of the pupil: a history of the clinical use of the pupil as an indicator of visual potential. J Neuroophthalmol 2003;23:213-24. RAPD—for example, only 70% of patients, all with detectable 7 Zaidi FH, Cheong-Leen R, Gair EJ, et al. The Amster chart is of doubtful value differences in PEVR, had clinically detectable RAPDs in retinal screening for early laser therapy of subretinal membranes. The West (table 1). PEVR aids diagnosis in cases where the RAPD is London Survey. Eye 2004;18:503-8. 8 Barbur JL, Forsyth PM, Thomson WD. A new system for simultaneous equivocal ( fig 1, bottom)." measurement of pupil size and two-dimensional eye movements. Clin Vis Sci Inter-eye PEVR and PER incorporate maximum phasic 1987;2:131-42. constriction (reflex constriction). In the 1960s the pupil was 9 Marmor MF, Holder GE, Seeliger MW, et al. Standard for clinical modelled using both a rate sensitive process filtering electroretinography 12004 update). Doc Ophthalmol 2004;108:107-14. 10 Arden GB, Barrada A, Kelsey JH. New clinical test of retinal function based temporal information (transient phasic constriction) and a upon the standing potential of eye. Br J Ophthalmol 1962;46:4,49-67. proportional process dependent on steady state processes 1 1 Zaidi FH, Moseley Ml. Use of pupil size and reaction to detect orbital trauma (sustained constriction), which is neuronally correlated with during and after surgery. Clin Otolaryngol Allied Sci 2004;29:288-9. ganglion cell firing.' PEVR algebraically incorporates 12 Cox TA. Pupillary escape. Neurology 1992;42:1271-3. 13 Akamatsu T, Enomoto H, Tabuchi A. Pupillographic studies on the pupillary phasic constriction (representing the sum neuronal firing constriction during 6 secinds of light exposure. Nippon Ganka Gakkai Zasshi from both sustained and phasic ON ganglion cells) compared 1979;83:1530-7. with the subsequent loss of firing from sustained ON 14 Bergamin 0, Kardon RH. Latency of the pupil light reflex: sample rate, stimulus intensity, and variation in normal subjects. Invest Ophthalmol Vis Sci ganglion cells only (MDA), making inter-eye PEVR a power- 2003;44:1546-54. ful tool in selected cases. Functionally the PEVR may also 15 Clynes M. Unidirectional rate sensitivity: a biocybemetic law of reflex and represent the activity of the cone rich macula, which partly humoral systems as physiologic channels of control and communication. drives sustained pupillary constriction—this is physiologi- Ann N Y Acad Sci, 1961 28, 92:946-69. 16 Bergamin 0, Zimmerman MB, Kardon RH. Pupil light reflex in normal and cally plausible as cones are biological sensors of high diseased eyes: diagnosis of visual dysfunction using waveform partitioning. intensity light and augment pupillary constriction. Ophthalmology 2003;110:106-14. www.bjophthalmol.com Current Biology Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Short-Wavelength Light Sensitivity of Circadian, Pupillary, and Visual Awareness in Humans Lacking an Outer Retina Farhan H. Zaidi18, Joseph T. Hull28, Stuart N. Peirson38, Katharina Wulff3, Daniel Aeschbach2'4, Joshua J. Gooley2'4, George C. Brainard°, Kevin Gregory-Evansl, Joseph F. Rizzo6, Charles A. Czeisler2'4, Russell G. Foster3*, Merrick J. Moseley' **, and Steven W. Lockley2,4*** 1Division of Neuroscience and Mental Health, Faculty of Medicine, Imperial College London, London W6 8RF, United Kingdom. 2Division of Sleep Medicine, Brigham and Women's Hospital, Boston, Massachusetts 02115. 3Nuffield Laboratory of Ophthalmology, University of Oxford, Wellcome Trust Centre for Human Genetics, Roosevelt Drive, Oxford, OX3 7BN, United Kingdom. 4Division of Sleep Medicine, Harvard Medical School, Boston, Massachusetts 02115. 5Department of Neurology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107. 6Department of Ophthalmology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts 02114. 'Department of Optometry and Visual Science, City University, Northampton Square, London EC1 V OHB, United Kingdom. Summary As the ear has dual functions for audition and balance, the eye has a dual role in detecting light for a wide range of behavioral and physiological functions separate from sight [1-11]. These responses are driven primarily by stimulation of photosensitive retinal ganglion cells (pRGCs) that are most sensitive to short-wavelength (-480 nm) blue light and remain functional in the absence of rods and cones [8-10]. We examined the spectral sensitivity of non-image-forming responses in two profoundly blind subjects lacking functional rods and cones (one male, 56 yr old; one female, 87 yr old). In the male subject, we found that short-wavelength light preferentially suppressed melatonin, reset the circadian pacemaker, and directly enhanced alertness compared to 555 nm exposure, which is the peak sensitivity of the photopic visual system. In an action spectrum for pupillary constriction, the female subject exhibited a peak spectral sensitivity (X,„tom ) of 480 nm, matching that of the pRGCs but not that of the rods and cones. This subject was also able to correctly report a threshold short- wavelength stimulus (-480 nm) but not other wavelengths. Collectively these data show that pRGCs contribute to both circadian physiology and rudimentary visual awareness in humans and challenge the assumption that rod- and cone-based photoreception mediate all "visual" responses to light. © 2007 Elsevier Ltd. *Corresponding author [email protected]. **Corresponding author [email protected]. ***Corresponding author [email protected]. These authors contributed equally to this work. This document was posted here by permission of the publisher. At the time of deposit, it included all changes made during peer review, copyediting, and publishing. The U.S. National Library of Medicine is responsible for all links within the document and for incorporating any publisher-supplied amendments or retractions issued subsequently. The published journal article, guaranteed to be such by Elsevier, is available for free, on ScienceDirect. Zaidi et al. Page 2 Keywords HUMDISEASE Results and Discussion Two blind subjects (one male, 56 yr old; one female, 87 yr old) without light perception were studied in parallel experiments. The female subject was a member of a family expressing an autosomal-dominant cone-rod dystrophy, which is described as a severe, early-onset phenotype with patients progressing to no perception of light by the fifth decade of life [12, 13]. The male subject had retinitis pigmentosa, a progressive disease of the retinal photoreceptors, and he reported losing light perception in his mid-30s. He had bilateral posterior subcapsular cataracts. Both subjects met all clinical criteria of blindness arising from degenerative retinal disease. These include pupils that are unreactive to light after standard penlight examination and self- reported inability to perceive light. Fundus photography and ocular coherence tomography failed to identify an outer retina in the female subject (an absence consistent with blindness), and electroretinography demonstrated no detectable rod or cone function (Figure 1). A fundoscopic examination of the male subject also revealed atrophy of the retinal pigment epithelium layer throughout the fundi, and visually evoked potentials were negative, again consistent with total visual loss. Both subjects reported having no sleep disorders and normal age-appropriate 24-hr sleep/wake patterns, as confirmed by quantitative assessments of circadian rest-activity behavior carried out with wrist actigraphy while they lived at home [14, 15]; these results are consistent with a functionally intact retinohypothalmic tract [1, 16] (Figure 2). A normal circadian phase was further confirmed using urinary 6-sulphatoxymelatonin (aMT6s) rhythms in the male subject [3] (Figure 2; also, Supplemental Data available online). In experiment 1, conducted with the male subject, we aimed to test the spectral sensitivity of the circadian, neuroendocrine, and neurobehavioral axes (Figure 3 and Figure S1). First, we confirmed that he retained a normal melatonin-suppression response to bright-white light exposure [1] on two separate occasions three years apart (see Supplemental Data). We then conducted a 14 day inpatient study to compare the effects of 6.5 hr exposure to 460 nm and 555 nm monochromatic light on circadian phase resetting, melatonin suppression, and enhancement of arousal [17, 18]. In order to compare the relative contribution of the photosensitive retinal ganglion cells (pRGCs) and classical (rod/cone) photoreceptors, we chose two light sources that would differentially stimulate these systems: a monochromatic "blue" light source with a peak emission (Xmax) at 460 nm and hence close to the Amax of human pRGCs (-480 nm) [11, 19], and a monochromatic light source with a Xmax at 555 nm corresponding to the peak of human photopic vision. Given that this subject exhibited a 24-hr sleep-wake pattern and an entrained aMT6s rhythm, we predicted that the pRGC/melanopsin- driven system would be intact and that the short-wavelength stimulus would elicit full circadian, neuroendocrine, and neurobehavioral responses, whereas the lack of classical photoreception would preclude any response to mid-wavelength 555 nm light. In a randomized, single-blind design, we exposed the subject to an equal photon density (2.8 x 1013 photons/cm2/s) of 555 nm and 460 nm monochromatic light for 6.5 hr, timed to start 1.25 hr before the prestudy bedtime [17, 18]. The subject was seated 90 min prior to and during light exposure, and for 60 min afterward, and was administered a pupil dilator (1 drop per eye, 0.5% cyclopentolate HCI; Cyclogyl, Alcon Laboratories, Texas) and kept in darkness for 15 min prior to lights on (see Supplemental Data). As hypothesized because of the absence of a functional cone response, ocular exposure to 555 nm light had no effect on plasma melatonin, Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 3 whereas 460 nm light suppressed melatonin by 57% (Figure 3A). Exposure to 460 nm light also caused a —1.2 hr phase delay in the timing of the circadian melatonin rhythm, whereas 555 nm light caused a minimal phase shift (-0.4 hr). In addition, the blue light preferentially increased alpha activity (8-10 Hz) in the waking electroencephalogram (EEG) recordings, indicating a more alert state [18, 20] (Figure 3B), and appeared to decrease subjective sleepiness and improve auditory performance during the latter half of the light exposure (Figure S1), consistent with the short-wavelength sensitivity for the acute effects of light in sighted subjects under similar conditions [17, 18, 21]. It is interesting to note that the blue light did not cause a suppression of delta and theta activity in the waking EEG, as we have previously observed in sighted subjects [18], and it is tempting to suggest that the lack of rod-cone photoreception in this subject may account for the altered EEG response at those particular frequencies, as we recently speculated [22]. Further data are required, however, to confirm this hypothesis. Nevertheless, the short-wavelength near-maximal sensitivity to light at this photon density for a range of responses indicates that this blind subject has a fully functional non-rod, non-cone photoreceptor system mediating the circadian, neuroendocrine, and neurobehavioral effects of light, presumably via intact melanopsin-containing pRGCs. In experiment 2, we investigated the spectral sensitivity of pupil construction in the female subject by using analytical-photobiological action-spectroscopy techniques. On the basis of her 24 hr sleep/wake pattern and our previous studies on rodents [9, 23], we reasoned that she should also possess some pupil reactivity to bright light, despite the clinical reports that she was unresponsive to the brief light exposure from either a penlight or indirect ophthalmoscopic examination. Quantitative pupillometry, employing monochromatic light at a broad range of wavelengths and irradiances (1011-1016 log photons/cm2/s) with an exposure duration of 10 s, showed that the subject possessed a functioning pupillomotor system responsive to bright light. The pupil-constriction response was spectrally tuned, peaking (k,„„) at 476 nm. Irradiance-response curves showed a high statistical fit of their derived half-saturation constants to a vitamin A opsin-pigment nomogram (R2 = 0.89, compared to R2 = 0.35 for rod and R2 < 0.01 for all three cone classes), suggesting that pupil constriction was being driven by a single photopigment (Figure 4). The spectral maxima of 476 nm corresponds well to the action spectra for pRGCs in both human (483 nm) and nonhuman primates (482 nm) [10, 24], but not the Xrnax of human rods (-498 nm) or short, medium, and long-wavelength cones (A.max —420, 534, and 563 nm, respectively) [25] (Figure 4). When the pupil-action spectrum was corrected for preretinal lens absorption [26], the peak spectral sensitivity shifted slightly from 476 nm to 480 nm. Consistent with the results from experiment 1, these data show that this subject possesses both an intact retinopretectal projection (pupillary constriction) and a retinohypothalamic projection (circadian entrainment), and that these responses to light are driven exclusively by short-wavelength-sensitive pRGCs in subjects lacking rods and cones and do not require input from the photopic system [24]. Notably, the confirmation of a pupil response following longer-duration exposure than typically used in brief penlight examinations questions the relevance of this technique, given that unreactive pupils are considered clinically to be a sine qua non of profound blindness of retinal origin despite earlier evidence for short- wavelength sensitivity in human pupil responses [27, 28]. The recent finding in primates that the pRGCs project to the dorsal lateral geniculate nucleus (dLGN) [10]—the thalamic relay that provides a direct input to the visual cortex—led us to explore the possibility that these photoreceptors might contribute to an individual's ability to detect or even experience some awareness of light. We therefore tested whether the female subject could report whether a given light stimulus was present in the first or second of two temporal intervals in a two-alternative forced-choice paradigm (2AFC). After some initial hesitancy about being asked to report the presence of visual stimuli of which she was nominally unaware, she was able to correctly identify the interval in which a 481 nm test light appeared (p < 0.001) but failed (p > 0.05) to detect light at longer or shorter wavelengths (420, 460, 500, Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 4 515, 540, 560, and 580 nm) (Figure 4). These detection probabilities remained unchanged when corrected for multiple testing (Bonferroni). Furthermore, she reported that the presence of the detectable stimuli (481 nm) elicited in her a percept that she described as "brightness." Although superficially these responses resemble cortical blindsight in that she was able to detect a stimulus with a rate of success above chance [29], these data represent a markedly different phenomenon because subjects with damage to the primary visual cortex (VI) have no conscious perception of the stimulus presented [29]. Could these responses to light have arisen from a small number of surviving rods and/or cones rather than from the pRGCs? Although visually evoked potentials (VEP), electroretinogram (ERG), and ocular coherence tomography (OCT) analysis cannot preclude the persistence of a residual population of rods and/or cones, there was no functional evidence of any significant rod or cone involvement. Both the Xmax of —480 nm and the correspondence of the action spectrum to a single opsin- and vitamin A-based photopigment template strongly implicate phototransduction by the pRGC subsystem alone. The question remains, however, which neuronal pathways and brain structures mediate these "nonvisual" effects of light. Neuroanatomical investigations in rodents show that melanopsin- containing ganglion cells project to a range of retinorecipient nuclei, including major projections to (1) the hypothalamic suprachiasmatic nuclei (SCN), the site of endogenous circadian pacemaker; (2) the intergeniculate leaflet of the thalamus, an area that is closely linked to normal circadian function and conveys photic and nonphotic signals to the SCN; (3) the ventrolateral preoptic area, an area that controls the switch between sleep and wake states; (4) the olivary-pretectal nucleus, implicated in the pupillary constriction response; and (5) the superior colliculus, which mediates visual and auditory sensorimotor reponses [30, 31]. As indicated previously, a subset of melanopsin-containing ganglion cells also project to the dLGN [10, 31] and in primates have a peak spectral sensitivity (?max) of 482 nm [10], thereby possibly providing the neuroanatomical substrate in support of the identical short-wavelength sensitivity for the visual awareness response observed in the female subject. Moreover, recent imaging studies in humans are beginning to identify brain regions associated with light-induced improvements in performance and cognition [32-34] and show preferential short-wavelength activation of the thalamus and the anterior insula, structures strongly implicated in arousal and memory function [34]. Our data strengthen the conclusion that the clinical diagnosis of "complete" blindness (i.e., visual and circadian) should assess the state of both the image-forming and the non-image- forming photoreceptive systems [1]. If blind individuals are found to be light sensitive, this knowledge will help ensure that they expose their eyes to sufficient daytime light to maintain normal circadian entrainment and sleep/wake rhythmicity. This evaluation is particularly critical prior to bilateral enucleation because, if light-responsive eyes are removed or individuals do not expose their eyes to a robust light-dark cycle, the patients may develop a debilitating circadian-rhythm sleep disorder [3, 14]. Patients with diseases of the inner retina that result in retinal ganglion cell death (e.g., glaucoma) are at particular risk and should be counseled about the effects of pRGC loss. Where complete blindness results, appropriately timed melatonin treatment may be warranted in order to establish entrained circadian rhythmicity [35, 36]. Conclusions We have shown that circadian, neuroendocrine, and neurobehavioral responses to light, and even visual awareness of light, are retained in visually blind subjects lacking functional outer retinae, confirming in humans the recent remarkable discovery of a novel photoreceptor system in the mammalian eye. These findings question the traditional view that rod- and cone-based Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 5 photoreception mediate all "visual" responses to light (such as pupillary constriction and visual awareness) and suggest that these and "nonvisual" circadian and neuroendocrine responses to light in humans are driven primarily by a non-rod, non-cone, short-wavelength-sensitive photoreceptor system located in the ganglion cell layer. Supplemental Data Rcfer to Web version on PubMed Central for supplementary material. Acknowledgments This paper is dedicated to the memory of the female subject, who completed these studies with good humor and unfailing enthusiasm. We would like to thank the technical, dietary, and clinical staff of the Division of Sleep Medicine and the General Clinical Research Centre at Brigham and Women's Hospital. This work was supported by the Wellcome Trust, UK (grant 069714 to R.G.F.) and the National Institutes of Health, USA (National Institute of Neurological Disorders and Stroke grants ROI NS040982 to C.A.C. and R01 NS36590 to G.C.B., National Center for Complimentary and Alternative Medicine grant ROI AT002129 to S.W.L. and the National Center for Research Resources MOl RR02635). Drs. G.C.B., C.A.C., R.G.F., and S.W.L. were supported in part by NASA Cooperative Agreement NCC9-58 with the National Space Biomedical Research Institute. Dr. Acschbach has received research support from Cephalon Inc. Dr. Brainard's laboratory has received research support from Apollo Health, OSRAM, and Philips Lighting By. Dr. Brainard holds one active and two pending patents for the effects of light on humans. Dr. Czeisler has received consulting fees from or served as a paid member of scientific advisory boards for: Actelion, Inc., Cephalon, Inc., Fedex Kinko's, Hypnion, Inc., Rcspironics, Inc., Takeda Pharmaceuticals, Inc., and Vanda Pharmaceuticals, Inc. Dr. Czeisler holds a number of process patents in the field of sleep and circadian rhythms, including the biological effects of light, all of which arc assigned to the Brigham and Women's Hospital. Since 1985, Dr. Czcisler has served as an expert witness on various legal cases related to sleep and/or circadian rhythms. Dr. Czeisler has never served as an expert witness for a commercial research sponsor. Dr. Lockley has received research support from ResMed Inc., and holds a process patent on the biological effects of light which is assigned to the Brigham and Women's Hospital. The remaining authors report no financial conflict of interest. References Czeisler C.A. Shanahan T.L. Klerman E.B. Martens H. Brotman D.J. Emens J.S. Klein T. Rizzo J.F. Suppression of melatonin secretion in some blind patients by exposure to bright light. N. Engl. J. Med. 1995;332:6-11. [PubMed: 7990870] Ruberg F.L. Skene D.J. Hanifin J.P. Rollag M.D. English J. Arendt J. Brainard G.C. Melatonin regulation in humans with color vision deficiencies. J. Clin. Endocrinol. Metab. 1996;81:2980-2985. [PubMed: 8768862] Lockley S.W. Skene D.J. Tabandeh H. Bird A.C. Defrance R. Arendt J. Relationship between melatonin rhythms and visual loss in the blind. J. Clin. Endocrinol. Metab. 1997;82:3763-3770. [PubMed: 9360538] Freedman M.S. Lucas R.J. Soni B. von Schantz M. Munoz M. David-Gray Z. Foster R. Regulation of mammalian circadian behavior by non-rod, non-cone, ocular photoreceptors. Science 1999;284:502- 504. [PubMed: 10205061] Lucas R.J. Freedman M.S. Munoz M. Garcia-Fernandez J.M. Foster R.G. Regulation of the mammalian pineal by non-rod, non-cone, ocular photoreceptors. Science 1999;284:505-507. [PubMed: 10205062] Brainard G.C. Hanifin J.P. Greeson J.M. Byrne B. Glickman G. Gerner E. Rollag M.D. Action spectrum for melatonin regulation in humans: Evidence for a novel circadian photoreceptor. J. Neurosci. 2001;21:6405-6412. [PubMed: 11487664] Thapan K. Arendt J. Skene D.J. An action spectrum for melatonin suppression: Evidence for a novel non- rod, non-cone photoreceptor system in humans. J. Physiol. 2001;535:261-267. [PubMed: 11507175] Berson D.M. Dunn F.A. Takao M. Phototransduction by retinal ganglion cells that set the circadian clock. Science 2002;295:1070-1073. [PubMed: 11834835] Hattar S. Lucas R.J. Mrosovsky N. Thompson S. Douglas R.H. Hankins M.W. Lem J. Biel M. Hofmann F. Foster R.G. Yau K.W. Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice. Nature 2003;424:76-81. [PubMed: 12808468] Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 6 Dacey D.M. Liao H.W. Peterson B.B. Robinson F.R. Smith V.C. Pokorny J. Yau K.W. Gamlin P.D. Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN. Nature 2005;433:749-754. [PubMed: 15716953] Foster R.G. Hankins M.W. Circadian vision. Curr. Biol. 2007;17:R746—R751. [PubMed: 17803920] Evans K. Fryer A. Inglehearn C. Duvall-Young J. Whittaker J.L. Gregory C.Y. Butler R. Ebenezer N. Hunt D.M. Bhattacharya S. Genetic linkage of cone-rod retinal dystrophy to chromosome 19q and evidence for segregation distortion. Nat. Genet. 1994;6:210-213. [PubMed: 8162077] Evans K. Duvall-Young J. Fitzke F.W. Arden G.B. Bhattacharya S.S. Bird A.C. Chromosome 19q cone- rod retinal dystrophy. Ocular phenotype. Arch, Ophthalmol. 1995;113:195-201. [PubMed: 7864751] Lockley S.W. Skene D.J. Butler L.J. Arendt J. Sleep and activity rhythms are related to circadian phase in the blind. Sleep 1999;22:616-623. [PubMed: 10450596] Wulff K. Joyce E. Middleton B. Dijk D.J. Foster R.G. The suitability of actigraphy, diary data, and urinary melatonin profiles for quantitative assessment of sleep disturbances in schizophrenia: A case report. Chronobiol. Int. 2006;23:485-495. [PubMed: 16687321] Klerman E.B. Shanahan T.L. Brotman D.J. Rimmer D.W. Emens J.S. Rizzo J.F. Czeisler C.A. Photic resetting of the human circadian pacemaker in the absence of conscious vision. J. Biol. Rhythms 2002;17:548-555. [PubMed: 12465888] Lockley S.W. Brainard G.C. Czeisler C.A. High sensitivity of the human circadian melatonin rhythm to resetting by short wavelength light. J. Clin. Endocrinol. Metab. 2003;88:4502-4505. [PubMed: 12970330] Lockley S.W. Evans E.E. Scheer F.A. Brainard G.C. Czeisler C.A. Aeschbach D. Short-wavelength sensitivity for the direct effects of light on alertness, vigilance, and the waking electroencephalogram in humans. Sleep 2006;29:161-168. [PubMed: 16494083] Brainard G.C. Hanifin J.P. Photons, clocks, and consciousness. J. Biol. Rhythms 2005;20:314-325. [PubMed: 16077151] Aeschbach D. Matthews J.R. Postolache T.T. Jackson M.A. Giesen H.A. Wehr T.A. Two circadian rhythms in the human electroencephalogram during wakefulness. Am. J. Physiol. 1999;277:R1771— R1779. [PubMed: 10600925] Cajochen C. Munch M. Kobialka S. Krauchi K. Steiner R. Oclhafen P. Orgul S. Wirz-Justice A. High sensitivity of human melatonin, alertness, thermoregulation, and heart rate to short wavelength light. J. Clin. Endocrinol. Metab. 2005;90:1311-1316. [PubMed: 15585546] Lockley S.W. Gooley J.J. Circadian photoreception: Spotlight on the brain. Curr. Biol. 2006;16:R795— R797. [PubMed: 16979545] Lucas R.J. Douglas R.H. Foster R.G. Characterization of an ocular photopigment capable of driving pupillary constriction in mice. Nat. Neurosci. 2001;4:621-626. [PubMed: 11369943] Gamlin P.D.R. McDougal D.H. Pokorny J. Smith V.C. Yau K.-W. Dacey D.M. Human and macaque pupil responses driven by melanopsin-containing retinal ganglion cells. Vision Res. 2007;47:946- 954. [PubMed: 17320141] Bowmaker J.K. Dartnall H.J. Visual pigments of rods and cones in a human retina. J. Physiol. 1980;298:501-511. [PubMed: 7359434] Stockman A. Sharpe L.T. Fach C.C. The spectral sensitivity of the human short-wavelength cones. Vision Res. 1999;39:2901-2927. [PubMed: 10492818] Bouma H. Size of the static pupil as a function of wavelength and luminosity of the light incident on the human eye. Nature 1962;193:690-691. [PubMed: 13871842] Alexandridis E. Kocppc E.R. [The spectral sensitivity of retinal photoreceptors conducting the pupillary light reflex in humans]. Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 1969;177:136-151. [PubMed: 5305734] Cowey A. The 30th Sir Frederick Bartlett lecture. Fact, artefact, and myth about blindsight. Q. J. Exp. Psychol. A 2004;57:577-609. [PubMed: 15204125] Gooley J.J. Lu J. Fischer D. Saper C.B. A broad role for melanopsin in nonvisual photoreception. J. Neurosci. 2003;23:7093-7106. [PubMed: 12904470] Hattar S. Kumar M. Park A. Tong P. Tung J. Yau K.W. Berson D.M. Central projections of melanopsin- expressing retinal ganglion cells in the mouse. J. Comp. Neurol. 2006;497:326-349. [PubMed: 16736474] Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi at al. Page 7 Perrin F. Peigneux P. Fuchs S. Verhaeghe S. Laureys S. Middleton B. Degueldre C. Del Fiore G. Vandewalle G. Balteau E. Nonvisual responses to light exposure in the human brain during the circadian night. Curr. Biol. 2004;14:1842-1846. [PubMed: 15498492] Vandewalle G. Balteau E. Phillips C. Degueldre C. Moreau V. Sterpenich V. Albouy G. Darsaud A. Desseilles M. Dang-Vu T.T. Daytime light exposure dynamically enhances brain responses. Curr. Biol. 2006;16:1616-1621. [PubMed: 16920622] Vandewalle G. Gais S. Schabus M. Balteau E. Carrier J. Darsaud A. Sterpenich V. Albouy G. Dijk D. Maquet P. Wavelength-dependent modulation of brain responses to a working memory task by daytime light exposure. Cereb. Cortex. 2007;17:2788-2795. [PubMed: 17404390]Published online April 2,2007 Lockley S.W. Skene D.J. James K. Thapan K. Wright J. Arendt J. Melatonin administration can entrain the free-running circadian system of blind subjects. J. Endocrinol. 2000;164:R1—R6. [PubMed: 10607943] Sack R.L. Brandes R.W. Kendall A.R. Levey A.J. Entrainment of free-running circadian rhythms by melatonin in blind people. N. Engl. J. Med. 2000;343:1070-1077. [PubMed: 11027741] Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 8 A B SUBJECT NORMAL A B ion is D /div lts rovo ic 25 m E F Alp(1‘ 100 , Microns G I I I I I I Normal 1 distribution 15msecidivision 600 percentile 400 200 100 Figure 1. Neuroophthalmology and Ocular Anatomy of the Blind Female Subject and a Normal Control The left panel shows fundoscopy findings of the 87-year-old blind female subject (A) and a representative ocular-coherence tomogram for the peripheral retina (C) and central macula region (D) of the left eye, compared with a normal age-matched sighted control (B, E, and F). Her retina is abnormally thin (less than 160 microns) and there is no identifiable outer nuclear layer or photoreceptor layer, suggesting that photoreceptors are absent, and the choroid has abnormally high reflectivity (Ch) in contrast to the normal age-matched subject (E and F), where stratification within the neurosensory retina, particularly the outer nuclear layer (ONL), can be seen. By contrast, the ganglion cell and nerve fiber layers of the inner retina of the blind woman are of normal thickness, and there is no cellular disruption, allowing clear recognition and delineation of normal histo-architecture in both retinal periphery and macula. In (G), comparison of the normal macula profile in an age-matched individual (within green limits, as shown in OCT image in [F]) illustrates loss of normal macular contour in the blind subject (black line, as derived from [D]). The normal distribution percentile correlates the color-coded areas of the figure to percentages of age-matched people who might possess retinae within that region. V = vitreous, NR = neurosensory retina. The right panel shows electroretinographic responses from the female subject (A, C, and E) and an age-matched, normal eye (B, D, and F) for dark-adapted (rod-photoreceptor predominant) responses (A and B); dark-adapted, light-adapted (mixed photoreceptor) responses (C and D); and light-adapted (cone predominant) responses (E and F) to 30 Hz flicker stimuli. White-light stimuli at 3.0 cd s/m2 intensity were used for all tests and began at the start of recordings in all cases. The traces for the blind subject show no detectable electroretinographic responses (Note: [C]shows a drifting baseline.). Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 9 A B C 2.5 1 . - , id 1•••••ialmt ablik.614.0A1 2.0 AL., 1 • LA•11•AllAilidail 15 ithillAmaArriu . LAM . 10 . - .1. iiii.akallulik• I - /h) 5 410, Lia, lk g 5 id Li. . . .11 0.5 ..1 (u 0.0 in ..6116.6.6.A441A+.1 tigabh•isaral ton 10 . la A alli•LIAILL•Add me ) 1- , 611.64,1666.6u1Ata. 6 iharldhiligALLika -1446.64.11..64/616,. toxy 15 ha a A lp .._L. . . 4.141•Liwan 6-su c.A• IA11111,11. t • ,16111 ' r 20 _ 15 :1114•11.1L.6.iiica .0 11,- AA- • ..., IL L l ....1•111.1•1•AL•Aa• inary 2.0 &IMMO/1440:J Ur .„..J 1.5 25 _ . ...AL, 1.0 J J Whit. .1k64.11 0 1.164 0.5 I • 0.0 0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18 0 0 6 12 18 0 6 12 18 0 Clock time (h) Clock time (h) Clock time (h) Figure 2. Entrained Rest-Activity and Urinary 6-Sulphatoxymelatonin Rhythms in Two Blind Subjects The daily activity rhythm (black) and light (lux) exposure (yellow) patterns of the female (A) and male (B) subjects, recorded at home for 3-4 weeks with wrist actigraphy (Actiwatch-L, Minimitter, New York). Data are double-plotted, with consecutive days plotted next to and beneath each other. The gray bars represent an arbitrary "night" from 23:00-6:00 hr for visual reference. Analysis of actigraphy data indicated that both the female and male subject had sleep onset (mean ± standard deviation [SD] sleep onset — 21:50 + 1:09 hr and 23:22 ± 0:24 hr, respectively) and sleep offset (8:38 ± 1:29 hr and 6:31 + 0:26 hr, respectively) times that fell within the range of actigraphically derived sleep times for blind subjects with previously confirmed normally phased circadian sleep and urinary 6-sulphatoxymelatonin rhythms (mean + 2SD sleep onset = 23:31 + 2:26 hr, sleep offset = 7:11 ± 2:24 hr) [3, 14]. The urinary 6-sulphatoxymelatonin (aMT6s) rhythm peak time (a) in the male subject confirmed the presence of a normally phased nighttime 24 hr rhythm (mean SD = 3:00 + 1:17 hr) that exhibited a normal phase angle (3:38 hr) with respect to the sleep/wake cycle based on previous studies in entrained blind subjects (mean +2SD phase angle, sleep onset — aMT6s peak = 4:38 2:28 [3, 14]). The raw urinary data are shown in [C] with the normal peak-time range for the aMT6s rhythms shown in gray (1:42-6:36 hr) [3]. Published as: Curr Biol. 2007 December 18; 17(24): 2122 2128. Zaidi et al. Page 10 A 120 - 1) 100 /m 80 (pg in ton 60 la me 40 ma Plas 20 -2 -1 0 1 2 3 4 5 6 7 8 9 10 12 C 13 0.5 I V N 10. & W of 0.4 0 2 4 6 8 10 12 14 16 18 20 Frequency (Hz) • N ill 0 °I) 3o -ca' 0.3 o. 0.2 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 Time since lights on (h) Figure 3. Short-Wavelength Light Sensitivity for Melatonin Suppression and Enhancement of EEG Alpha Power in a Blind Man The direct effects of exposure to green (555 nm) and blue (460 nm) monochromatic light on the male subject for melatonin suppression (A) and waking-EEG power density (B) as an objective correlate of alertness. Exposure to 555 nm light caused no suppression of melatonin as compared to the corresponding clock time the previous day, whereas exposure to 460 nm light suppressed melatonin (total suppression by AUC = 57%) and maintained the suppression effect throughout the entire 6.5-hr exposure (A). The 460 nm light also caused an elevation of alpha activity (8-10 Hz) in the waking EEG, indicative of a more alert state (B). Only alpha frequencies exhibited a wavelength-dependent difference during the second half of the light exposure (C). These data are consistent with the short-wavelength sensitivity for the acute effects of light in sighted subjects under similar conditions [17, 18, 21]. Published as: Curr Biol. 2007 December 18; 17(24): 2122 2128. Zaidi et al. Page 11 A 100% 420nm 100% 460nm 50% 50% 0% 11 13 15 13 15 17 l°°% 481nm 50% 0 13 15 17 100% - 515nm 50% 0% 1% 13 15 17 11 13 15 400 450 500 550 600 650 700 100% 560nm 100% 1 580nm Wavelength (nm) 50% 50% 0% 0% 11 13 15 17 11 11 Photons/cm2/s B 100 *** ) c/o 80 ( s 60 - onse ❑ Right eye ■ Left eye resp t 40 ec 20 Corr 0 420 460 481 500 506 515 540 560 580 Wavelength (nm) Figure 4. Short-Wavelength Light Sensitivity for Pupillary Constriction and Light Detection in a Blind Woman Irradiance-response curves (IRCs) were conducted at eight wavelengths for both eyes (squares indicate left eye, triangles indicate right eye) (A, left panel). Responses are plotted as percentage of maximum response obtained. IRCs were fitted with a four-parameter sigmoid function, with R2 values >0.90 in all cases. The resulting action spectrum of pupil responses (A, right panel) provided a poor fit to rod and cone photopigments (rod R2 = 0.35; SW cone, MW cone, LW cone R2 = 0). An optimum fit to the pupil response to light was provided by an opsin/vitamin A-based template with km,„ 476 nm (R2 = 0.89), corresponding closely to the pRGC system. Note: Data shown were not corrected for preretinal lens absorption. When this correction was applied, the 2max shifted from 476 nm to 480 nm. Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Zaidi et al. Page 12 (B) shows the results of the psychophysical testing in the same subject that indicated conscious perception of light at 481 nm (***p < 0.001) but failure (p > 0.05) to detect light at longer or shorter wavelengths (420, 460, 500, 515, 540, 560, and 580 nm). These results mirror the spectrally tuned response of the pupil, and suggest that the subject's detection and awareness of light also arise from pRGCs. Each histogram represents the percentage of correct responses out of 20 trials for both left and right eyes (360 trials in total). Published as: Curr Biol. 2007 December 18; 17(24): 2122-2128. Supplemental Data Short-Wavelength Light Sensitivity of Circadian, Pupillary, and Visual Awareness in Humans Lacking an Outer Retina Farhan H. Zaidi, Joseph T. Hull, Stuart N. Peirson, Katharina Wulff, Daniel Aeschbach, Joshua J. Gooley, George C. Brainard, Kevin Gregory-Evans, Joseph F. Rizzo Ill, Charles A. Czeisler, Russell G. Foster, Merrick J. Moseley, and Steven W. Lockley Supplemental Experimental Procedures circadian photoreception system by prior light history [S3-S6], the protocol was designed to ensure that light conditions were identical Ophthalmological Examinations before each experimental light exposure. Ocular coherence tomography (OCT) was conducted in the female For the first 3 days, the subject was scheduled for 8 hr sleep subject with a third-generation interferometric, noninvasive optical episodes in darkness (timed according to the prestudy sleep sched- tomographic imaging device (OCT3 Zeiss Humphrey Division, Dub- ule) and 16 hr wake episodes in -90 lux white light (measured at lin, California) providing 2-3 mm tissue penetration. Axial and lateral a height of 137 cm in the horizontal angle of gaze). Eight hours after resolution was < 10-20 microns. Electroretinographic responses waking on day 3, ambient light levels were reduced to 4 lux. Upon were recorded to International Society for Clinical Electrophysiology waking on day 4, the subject began a 26 hr constant routine (CR) of Vision (ISCEV) standards [S1] with RETI-port basic electrophysio- procedure [S7] to measure the endogenous circadian rhythm of mel- logical diagnostic systems (Roland Instruments, Brandenburg, Ger- atonin. During the CR procedure, the subject was continually super- many) and DTL (Dawson Trick Litzkow) fiber corneal electrodes. vised while remaining semirecumbant and awake in bed under dim Visually evoked potentials (VEPs) were assessed in the male subject light (- 4 lux) and provided identical hourly snacks in order to abol- via both pattern reversal (2/s) and strobe (2/s) with a high-contrast ish or distribute equally environmental circadian time cues. After an black-white checkerboard. EEG recordings were made from 8 hr sleep episode, the subject began a 16 hr wake episode in dim Fz-Oz, Pz-Oz, Fz-01, and Fz-02. No P100 responses were obtained, light, except during the 6.5 hr monochromatic light exposure indicating no light perception in either eye. centered in the wake episode. Following an 8 hr sleep, the subject completed a second CR procedure for 30 hr before an 8 hr sleep Experiment la: White-Light Melatonin-Suppression Test at his habitual bedtime. This 7 day sequence was then repeated Two white-light melatonin-suppression tests were conducted three under identical conditions. years apart during two different inpatient studies in the Intensive The experimental light-exposure conditions were identical for the Physiology Monitoring (IPM) unit of the General Clinical Research white light studies except that the subject was exposed to mono- Center (GCRC) at Brigham and Women's Hospital. The subject chromatic light continuously for 6.5 hr via a modified Ganzfeld (22CS) was exposed to 10,000 lux white fluorescent light (4100K, source [S8, S9], 15 min after administration of a mydriatic. The order Philips Lighting, The Netherlands) for 6.5 hr during the biological of the monochromatic light exposure (555 nm, 460 nm) was ran- night (21:75-04:25 hr), centered in a 16 hr wake episode in dim light domly determined but not revealed to the subject. Light wavelength (..4 lux). The subject was seated 90 min prior to and 60 min after the and irradiance were measured with a PR-650 SpectraScan Colorim- light exposure and was asked to fix his "gaze" toward the light for eter with a CR-650 cosine receptor (PhotoResearch, California) and 50% of the exposure duration in an alternating "fixed:free" gaze pat- an IL1400 radiometer/powermeter with an SEL-033/F/W detector tern. During the free gaze, he was encouraged to avoid photophobic (International Light, MA), respectively. Plasma melatonin was drawn behavior. Light measurements were made at the level of the eye with every 20-60 min, subjective sleepiness was assessed every 30 min an IL1400 radiometer/powermeter with an SEL-033N/W detector with the Karolinksa Sleepiness Scale, and auditory performance (International Light, Massachusetts). Plasma melatonin was sam- was measured with a 10 min Psychomotor Vigilance Test (PVT) every pled every 20-60 min via an indwelling catheter kept patent with hour while the subject was awake. Waking-EEG recordings were a heparinized saline drip (5 IU heparin/mL 0.45% NaCI). Melatonin conducted during most wake episodes and included a 3 min Karo- suppression was calculated by dividing the percent difference linska Drowsiness Test each hour with a portable digital polysomno- between the area under the curve (AUC) of the plasma-melatonin graphic recorder (Vitaport-3 digital recorder, TEMEC Instruments profile during the 6.5 hr light exposure by the AUC during the B.V., The Netherlands). Recordings consisted of EEG, electrooculo- same clock time while the subject was awake under constant routine gram, and a 2-lead electrocardiogram (ECG), and electrodes were conditions in dim light (--4 lux) the previous day. positioned according to the International 10-20 System, with linked Exposure to 10,000 lux white light for 6.5 hr caused a 90% and mastoid references (Ax) from the z-line, Fz-Ax, Cz-Ax, Pz-Ax, and 86% reduction in plasma-melatonin levels, respectively, consistent Oz-Ax. Power-density spectra were calculated for 2 s epochs with with the normal response in fully sighted subjects studied under a Fast Fourier Transform routine (Vitagraph, TEMEC, The Nether- similar conditions [S2]. lands) and averaged within the 3 min KDT intervals, after visually identified artifacts were excluded. The data presented were re-refer- enced offline using the Cz-Pz derivation because of ECG artifact in Experiment lb: Circadian, Neuroendocrine, and the Ax recordings. For further details, see [S10]. Neurobehavioral Response to Monochromatic Light Exposure Plasma melatonin rhythms during the inpatient baseline days and The effects of monochromatic light exposure were studied during initial circadian phase assessment confirmed the normal circadian a 14 day inpatient protocol in an environment free of time cues (indi- entrainment indicated by the home-based urinary 6-sulpatoxymela- vidual windowless, soundproofed suite) in the GCRC IPM at Brig- tonin rhythms (Figure 2 in the main text), with an average dim-light ham and Women's Hospital. The subject was healthy except for melatonin onset (DLMO) time of 20.70 ± 0.08 hr (n = 3 days). The his visual impairment, as determined from physical and psycholog- circadian phase of the two experimental light exposures was also ical examinations as well as blood and urine toxicology. For 4 weeks consistent, as confirmed by the DLMO phase assessments prior to prior to admission, he maintained a consistent 8 hr sleep episode, as each light exposure (20.28 hr and 19.78 hr, respectively). confirmed with wrist actigraphy (Actiwatch-L, Minimitter, New York) and sleep-wake logs, and refrained from any drug use, including caf- feine, alcohol, nicotine, dietary supplements, and prescription or Experiment 2a: Pupillary-Constriction Action Spectroscopy nonprescription medications, as confirmed by admission toxicol- Direct, monocular constriction responses to light stimulation were ogy. The 14 day inpatient study consisted of two consecutive recorded at ten irradiance intensities (1011-1016 log photons/cm2/s) identical 7 day schedules with the experimental light exposures for each of eight wavelengths (420, 460, 481, 500, 515, 540, 560, occurring 6 days apart. Given the potential for adaptation of the and 580 nm). Pupil area was monitored with a PSCAN_l 00 (Circuit S2 Figure S1. Short-Wavelength Sensitivity for the Acute Effects of Light on Sleepiness and 9 Auditory Performance in a Blind Man The direct effects of exposure to green (555 nm) and blue (460 nm) monochromatic light on the male subject for subjective sleep- iness (A), auditory reaction time (B), and lapses of attention (C) plotted according to the same convention as that for melatonin suppression in Figure 3 (main text). Subjective sleepiness appeared to decrease during the latter half of the 460 nm exposure, as compared to both the previous day and directly with 555 nm exposure (A). Similarly, auditory median reac- tion time (B) and lapses of attention (C) also 4 showed greater improvement during 460 nm light exposure over the second half of the exposure, as compared to 555 nm light expo- sure, consistent with the changes in EEG alpha -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 activity at the same time (Figure 3 in the main text). These data are consistent with the short- wavelength sensitivity for the acute effects 600 of light in sighted subjects under similar conditions. ms) ( e im 500 t tion c 400 - rea ditory u a 300 - dian Me 200 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 C 50 40 500 ms > 30 - es s lap f 20 - o ber 10 Num 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 Time since lights on (h) Solve, UK) infrared pupillometer with a spatial resolution of 0.05 mm. pupil constriction (IR5o), x is the photon flux (log photons/cm2/s) Stimuli were of 10 s duration and provided by a quartz halogen light for which y is to be derived, and d is the slope of the curve. Each source delivered in non-Maxwellian view. Irradiance and wavelength curve was fitted in MS Excel (Microsoft), with Solver add-in (Front- were controlled with neutral density and monochromatic line systems) used to minimize the sum of squares for each IRC, interference filters, respectively (10 nm half-peak bandwidth). as described previously [S111. R2 values were > 0.90 for all IRCs Irradiance-response curves (IRCs) were generated by plotting (Figure 4A in the main text). An action spectrum was then con- pupil constriction (y axis) against light dose (log photons/cm2/s). structed on the basis of relative sensitivities derived from IRCs at IRCs were fitted with a four-parameter sigmoid curve: y = a + (b/ the eight wavelengths investigated (Figure 4A in the main text). [1 +1 0(c-x)/d]), where y is the response, a is the baseline, b is the Given that responses reached half-maximum at only three wave- maximum response, c is the irradiance required to evoke a 50% lengths, relative sensitivity based upon IR50 would result in S3 extrapolation beyond the data range. Consequently, relative sensi- resetting by short wavelength light. J. Clin. Endocrinol. Metab. tivity was determined from responses obtained from an equivalent 88, 4502-4505. photon flux in the dynamic range of the IRC (i.e., above background S10. Lockley, S.W., Evans, E.E., Scheer, F.A., Brainard, G.C., noise and before saturation), as described previously [S12]. Data Czeisler, C.A., and Aeschbach, D. (2006). Short-wavelength were analyzed on the basis of responses to 1 x 10Th photons/cm2/s, sensitivity for the direct effects of light on alertness, vigilance, although comparable results were obtained within the range of and the waking electroencephalogram in humans. Sleep 29, 1 x 1014-1 x 1016 photons/cm2/s. 161-168. Absorbance spectra for human visual pigments were generated S11. Peirson, S.N., Thompson, S., Hankins, M.W., and Foster, R.G. using the Govardovskii template [S13], based upon the published (2005). Mammalian photoentrainment: results, methods, and amp values [S141 of rod (498 nm) and of short- (420 nm), mid- approaches. Methods Enzymol. 393, 697-726. (534 nm), and long-wavelength cone (563 nm). Free-fitting of the S12. Provencio, I., and Foster, R.G. (1995). Circadian rhythms in visual pigment template was conducted via the least-squares mice can be regulated by photoreceptors with cone-like method described above, with log-transformed sensitivity and tem- characteristics. Brain Res. 694, 183-190. plate data used to prevent peak-fitting bias. Given that preretinal S13. Govardovskii, V.I., Fyhrquist, N., Reuter, T., Kuzmin, D.G., and absorption measurements were not available for the subject, the Donner, K. (2000). In search of the visual pigment template. Vis. data shown were not corrected for preretinal lens absorption. Neurosci. 17, 509-528. When standardized corrections were applied [S15], the shifted S14. Bowmaker, J.K., and Dartnall, H.J. (1980). Visual pigments of from 476 nm to 480 nm, although the R2 decreased to 0.70 as a result rods and cones in a human retina. J. Physiol. 298, 501-511. of the differential filtering effects at 420 nm. S15. Stockman, A., Sharpe, L.T., and Fach, C.C. (1999). The spectral sensitivity of the human short-wavelength cones. Vision Res. 39, 2901-2927. Experiment 2b: Forced-Choice Paradigm for Light Detection The subject's ability to detect and/or perceive light stimuli of differ- ent wavelengths was determined by a two-alternative forced-choice (2AFC) procedure. In this procedure, a light stimulus is presented in one of two temporal intervals, and it is mandatory for the subject to state at which interval he or she believes the stimulus to have been presented. Stimuli lasted for 10 sat an intensity of 1.45 x 1016 pho- tons/cm2/s at all wavelengths. Twenty trials were conducted for each of nine wavelengths for each eye. Data were analyzed with the Sign test, with a probability of level of 0.05 (corresponding to cor- rectly identifying the stimulus interval in 15/20 trials) deemed to be indicative of statistically significant detection. Note: These methods were relatively lengthy and taxing on this elderly subject. We there- fore selected a photon flux, based upon pilot experiments on the subject, that produced a significant but not saturating discrimina- tory response. Responses to the 2AFC procedure are not graded, and there is a tendency with this experimental paradigm to exagger- ate above-threshold stimuli and ignore near-threshold stimuli. The results obtained reflect this tendency. The approach allowed us to determine the wavelength of light that would allow a threshold re- sponse rather than a complete response profile. References S1. Marmor, M.F., Holder, G.E., Seeliger, M.W., and Yamamoto, S. (2004). Standard for clinical electroretinography (2004 update). Doc. Ophthalmol. 108, 107-114. S2. Zeitzer, J.M., Dijk, D.J., Kronauer, R., Brown, E., and Czeisler, C. (2000). Sensitivity of the human circadian pacemaker to nocturnal light: Melatonin phase resetting and suppression. J. Physiol. 526, 695-702. S3. Hankins, M.W., and Lucas, R.J. (2002). The primary visual path- way in humans is regulated according to long-term light expo- sure through the action of a nonclassical photopigment. Curr. Biol. 12, 191-198. S4. Smith, K.A., Schoen, M.W., and Czeisler, C.A. (2004). Adapta- tion of human pineal melatonin suppression by recent photic history. J. Clin. Endocrinol. Metab. 89, 3610-3614. S5. Wong, K.Y., Dunn, F.A., and Berson, D.M. (2005). Photorecep- tor adaptation in intrinsically photosensitive retinal ganglion cells. Neuron 48, 1001-1010. S6. Jasser, S.A., Hanifin, J.P., Rollag, M.D., and Brainard, G.C. (2006). Dim light adaptation attenuates acute melatonin suppression in humans. J. Biol. Rhythms 21, 394-404. S7. Duffy, J.F., and Dijk, D.J. (2002). Getting through to circadian oscillators: Why use constant routines? J. Biol. Rhythms 17, 4-13. S8. Brainard, G.C., Hanifin, J.P., Greeson, J.M., Byme, B., Glickman, G., Gemer, E., and Rollag, M.D. (2001). Action spectrum for melatonin regulation in humans: Evidence for a novel circadian photoreceptor. J. Neurosci. 21, 6405-6412. S9. Lockley, S.W., Brainard, G.C., and Czeisler, C.A. (2003). High sensitivity of the human circadian melatonin rhythm to