OPTOMETRY REVIEW. Key words: drusen of optic nerve head, multimodal imaging, optic nerve head, papilloedema, pseudopapilloedema

Transcription

1 C L I N I C A L A N D E X P E R I M E N T A L OPTOMETRY REVIEW The usefulness of multimodal imaging for differentiating pseudopapilloedema and true swelling of the optic nerve head: a review and case series Clin Exp Optom 2015; 98: Jaclyn Chiang * MOptom PGCertOcTher Elizabeth Wong * MOptom PGCertOcTher Andrew Whatham * DPhil BOptom PGCertOcTher Michael Hennessy * MBiomedE FRANZCO Michael Kalloniatis * PhD MScOptom PGCertOcTher Barbara Zangerl * DVM PhD * Centre for Eye Health and School of Optometry and Vision Science, University of New South Wales, Kensington, New South Wales, Australia Ophthalmology, Prince of Wales Hospital, Randwick, Australia bzangerl@cfeh.edu.au Submitted: 12 February 2014 Revised: 27 March 2014 Accepted for publication: 1 April 2014 DOI: /cxo Ophthalmic practitioners have to make a critical differential diagnosis in cases of an elevated optic nerve head. They have to discriminate between pseudopapilloedema (benign elevation of the optic nerve head) and true swelling of the optic nerve head. This decision has significant implications for appropriate patient management. Assessment of the optic disc prior to the advanced imaging techniques that are available today (particularly spectral domain optical coherence tomography and fundus autofluorescence), has mainly used diagnostic tools, such as funduscopy and retinal photography. As these traditional methods rely on the subjective assessment by the clinician, evaluation of the elevated optic nerve head to differentiate pseudopapilloedema from true swelling of the optic nerve head can be a challenge in clinical practice with patients typically referred for further neuroimaging investigation when the diagnosis is uncertain. The use of multimodal ocular imaging tools such as spectral domain optical coherence tomography, short wavelength fundus autofluorescence and ultrasonography, can potentially aid in the differentiation of pseudopapilloedema from true swelling of the optic nerve head, in conjunction with other clinical findings. By doing so, unnecessary patient costs and anxiety in the case of pseudopapilloedema can be reduced, and appropriate urgent referral and management in the case of true swelling of the optic nerve head can be initiated. Key words: drusen of optic nerve head, multimodal imaging, optic nerve head, papilloedema, pseudopapilloedema Optic nerve head elevation results from a number of aetiologies which can be divided into two broad groups: pseudopapilloedema (benign elevation of the optic nerve head) and true swelling of the optic nerve head, such as occurs in papilloedema due to raised intracranial pressure. Figure 1 gives a clinical classification of optic nerve head elevation, 1 which is an overlapping feature between the two groups, and hence it is imperative that these conditions are accurately differentiated in clinical practice, as misdiagnosis has significant implications for patient management. The management of pseudopapilloedema encompasses routine review and urgent medical attention is not commonly required. In contrast, true swelling of the optic nerve head secondary to increased intracranial pressure, for example, demands accurate diagnosis and urgent management as it is potentially sight and life threatening. 2 Correct diagnosis of pseudopapilloedema is important to ensure appropriate management, avoid medical costs on unnecessary neurologic procedures and prevent any needless patient anxiety. 2 Conditions that give rise to pseudopapilloedema include small crowded discs, tilted discs and optic nerve head drusen (Figure 1). Conditions that cause bilateral true swelling of the optic nerve head include those that are secondary to raised intracranial pressure (papilloedema). The vast majority of cases presenting with unilateral optic nerve head swelling and normal intracranial pressure, such as inflammatory, ischaemic and infiltrative neuropathies, direct optic nerve compression, traumatic optic neuropathy, toxicity induced optic neuropathy and ocular hypotony, have systemic signs or symptoms which either precede ocular manifestation or have ophthalmoscopic signs other than elevation of the optic disc pointing to its diagnosis. 3 A careful diagnostic work-up, including visual acuities, pupil responses, colour vision, ocular motility and visual fields, is imperative in the assessment of an elevated optic nerve head as characteristic changes in these clinical features, such as pain on eye movement, flame haemorrhages or cotton wool spots, can indicate optic nerve head pathology. It can be challenging to differentiate pseudopapilloedema from true swelling of the optic nerve head with standard clinical tests, such as slit lamp examination and fundus photography in the absence of other clinical signs and symptoms suggesting optic nerve head pathology. Where clinical results are uncertain, recently developed ocular imaging techniques such as spectral domain optical coherence tomography (OCT) and fundus autofluorescence, as well as the more established technique of B-scan ultrasonography, are relatively non-invasive and have improved the ability to differentiate pseudopapilloedema from true swelling of the optic nerve head. The remainder of this article will give an overview of pseudopapilloedema and true swelling of the optic nerve head, followed by 12 Clinical and Experimental Optometry 2014 Optometry Australia

2 Elevated ICP an overview of current optic nerve head imaging techniques and their ability to differentiate pseudopapilloedema from true swelling of the optic nerve head. Clinical use will be illustrated with clinical cases seen at the Centre for Eye Health, University of New South Wales, Kensington, Australia. Table 1 presents the clinical information on the patients. Consent was obtained from all patients in accordance with the Declaration of Helsinki and approved by the Biomedical Human Research Ethics Advisory Panel of the University of New South Wales. CLINICAL FEATURES OF PSEUDOPAPILLOEDEMA Optic disc elevation Pseudopapilloedema describes conditions which mimic papilloedema with elevation of the optic nerve head being the main clinical observation. The degree of optic nerve head elevation is easily appreciated with the OCT three-dimensional view (Figure 2). Optic disc elevation in pseudopapilloedema occurs secondary to a usually benign process, such as a small crowded optic nerve head, a tilted optic disc or drusen of the optic nerve head. ONH swelling Normal ICP Pseudopapilloedema Papilloedema Inflammatory Ischaemic Infiltrative Direct optic nerve Micropapilla Tilted optic discs Optic disc drusen Idiopathic intracranial hypertension Intracranial masses Meningitis compression (e.g. tumor) Traumatic optic neuropathy Toxicity induced optic neuropathy Ocular hypotony Optic neutritis Papillitis Neuro-retinitis Papillophlebitis CRVO CRAO GCA NAION Diabetic Leukaemia Lymphoma Optic nerve head changes likely secondary to other signs/symptoms papillopathy Figure 1. Differential diagnosis of optic nerve head elevation. Schematic flow chart of causes of optic disc elevation. Modified from Sadun and Wang. 1 ICP: intracranial pressure, CRVO: central retinal vein occlusion, CRAO: central retinal artery occlusion, GCA: giant cell arteritis, NAION: non-arteritic ischaemic optic neuropathy. Small crowded disc Small crowded discs are a physiological variation and can be a challenge in the differential diagnoses of optic nerve head elevation. The size of a small disc that can give rise to pseudopapilloedema is in the order of 1.95 ± 0.33 mm 3,4 compared to a normal optic disc of 2.69 ± 0.70 mm. 3,5 This anatomical variation creates the appearance of a crowded optic disc with indistinct margins and no apparent cupping; this is likely a result of a normal number of ganglion cell/optic nerve axons converging at a small optic disc to course through a smaller than average sclera foramen. 6 As these clinical results are caused by an unusually small and crowded disc, commonly associated with hyperopia due to shortened axial length, 7 no medical intervention is indicated unless there is coexistent pathology. Tilted optic disc The tilted optic disc is a congenital condition, thought to be associated with incomplete closure of the embryonic fissure within the spectrum of ocular colobomata. 8 The optic nerve enters the eye at an oblique angle, estimated by the disparity between the maximum height differences of the surface elevation of the disc and causes optic disc torsion, which is rotation of the disc about the sagittal axis of the optic nerve. 9 These optic discs are most commonly tilted in the inferonasal direction, with elevation and indistinct disc margins at the superotemporal aspect of the optic disc. 9,10 The associated blurring and elevation of the disc margins can imitate the appearance of a swollen optic disc. 9 Tilted optic discs have a prevalence of 0.4 to 3.5 per cent in the general population and present bilaterally in 37.5 to 80 per cent of cases. 9 They are often an incidental finding with increased prevalence in myopic and astigmatic eyes. 11 Visual field defects of tilted optic discs can occur in any location of the peripheral field as well as paracentrally, although defects in the superotemporal quadrant are most frequently noted. 12 Temporal visual field depressions, which do not respect the horizontal or vertical midline have also been observed. In cases where the field defects present bitemporally, chiasmal or postchiasmal compressive lesions need to be considered as a differential diagnosis; however, these defects usually respect the vertical midline. 8 Patients are usually asymptomatic except when associated with retinal findings such as macular changes from inferior staphylomata. 13 As tilted discs are generally a non-progressive condition, they require no treatment. Drusen of the optic nerve head Optic nerve head drusen are colloid bodies representing calcified deposits caused by a combination of abnormal vasculature, disturbed axonal metabolism and a small scleral canal. 14 Optic nerve head drusen has a prevalence of 3.4 to 24 per 1,000 population. 15 They may appear at the surface or buried within the disc and present bilaterally in 75 per cent of clinical cases. 15 Superficial drusen of the optic nerve head are visible as yellow highly reflective areas at the nerve head. Buried optic nerve head drusen can mimic papilloedema, if they cause the optic nerve to elevate with indistinct, irregular disc margins. Patients diagnosed with drusen of the optic nerve head usually do not describe symptoms; however, transient visual obscuration, chronic peripheral visual loss and rarely, acute visual loss, have been reported. 14 Optic nerve head drusen can also lead to ischaemic complications such as vascular occlusions due to direct compression on the blood vessels. 14 No effective treatment is established for optic nerve head drusen. Respective patients should be reviewed periodically to monitor any changes of the optic disc appearance, intraocular pressures and progression in visual field loss. 15 The progressive visual field loss that can accompany drusen of the optic nerve head may be indistinguishable from Clinical and Experimental Optometry 2014 Optometry Australia 13

3 Sex Age Associated signs and symptoms Refraction Visual acuity Vertical optic disc diameter (mm) Visual field Patient I: bilateral small crowded discs Patient II: bilateral tilted discs Patient III: unilateral buried optic nerve head drusen Patient IV: late stage bilateral optic nerve head drusen Patient V: papilloedema M 54 Asymptomatic -0.25/ R 6/4.5 R 1.20 R Normal plano/ L 6/4.5 L 1.45 L F 37 Asymptomatic -6.25/ R 6/6 +2 R 1.35 R Normal -5.00/ L 6/6 +2 L 1.30 L F 25 Asymptomatic +0.25/ R 6/6 +2 R 1.30 R Normal +0.25/ L 6/6 +2 L 1.40 L M 80 Asymptomatic +1.50/ R 6/7.6-2 R 1.65 R Marked constriction L 6/9.5 L 1.75 L F 35 Headaches, nausea, transient visual obscuration +0.50/ R +0.25/ L 6/6-1 R 6/6 L 1.50 R 1.40 L Scattered points of reduced sensitivity Table 1. Summary of patient characteristics that due to chronic glaucoma (Figure 3). In such patients, erring on the side of caution, a diagnosis of glaucoma may be assumed and intraocular pressure-lowering medication prescribed as a result in an attempt to slow or halt visual field loss. 15 CLINICAL FEATURES OF TRUE SWELLING OF THE OPTIC NERVE HEAD Bilateral true swelling of the optic nerve head as a result of elevated intracranial pressure is known as papilloedema. In this condition optic nerve head elevation is a clinical manifestation of oedema of the ganglion cell axons and interstitial fluid accumulation within nerve head tissue 16 due to mechanical or ischaemic factors. 17 While papilloedema is a bilateral condition, optic nerve head swelling may be present asymmetrically in this condition, postulated to be due to unequal numbers of arachnoid trabeculations lining the subarachnoid space, which may lead to differences in the resistance of the cerebrospinal fluid pressure reaching the optic disc. 18 Papilloedema can develop idiopathically or secondary to disorders such as meningitis, intracranial masses, shunt obstruction and subarachnoid haemorrhage. 3,19 Idiopathic intracranial hypertension predominantly affects overweight women of childbearing age with an estimated incidence in the general population of one to three per 100,000 per year. 20 Papilloedema may indicate the presence of chronically elevated intracranial pressure or correspond to a combination of the level and time of increase in intracranial pressure. 21 The most commonly reported symptom is a global headache, typically worst in the morning upon waking and being recumbent. 22 Visual signs and symptoms include transient visual loss, photophobia, dyschromatopsia, diplopia and an afferent pupillary defect. Patients may also experience tinnitus, nausea or vomiting. While visual field loss is usually progressive and visual acuity is affected in the later stages of typical idiopathic intracranial hypertension, an acute form, known as fulminant idiopathic intracranial hypertension, can present with rapid constriction of the visual fields (over a few days to less than four weeks after the initial onset of symptoms) and severe signs and symptoms of intracranial hypertension; 23 however, severity of papilloedema does not necessarily correlate with the degree of symptoms. 16 Mechanical features of the optic disc in papilloedema include blurring of the optic disc margins, congestion of the physiological optic disc cup, presence of choroidal and/or retinal folds and thickening of the peripapillary retinal nerve fibre layer (RNFL). 24 Prominent vascular changes encompass enlarged and tortuous peripapillary vessels, obscuration of the major optic disc vessels, soft or hard exudates, haemorrhages and hyperaemia of the optic disc. 24 Ophthalmoscopic features can be subtle in mild papilloedema. 25 On the other hand, optic nerve head swelling in the presence of normal intracranial pressure is predominantly unilateral, associated with systemic changes and presents additional ocular signs to the swollen optic disc, such as retinal haemorrhages, dilated tortuous veins and macular oedema (for example, in a central retinal vein occlusion). These conditions are heterogeneous and include inflammatory and ischaemic neuropathies, infiltrative processes, such as leukaemia and lymphoma, direct optic nerve compression, traumatic optic neuropathy and ocular hypotony or toxic effects induced by certain drugs, for example amiodarone optic neuropathy. 26 ASSESSING THE OPTIC NERVE HEAD WITH IMAGING INSTRUMENTATION Spectral domain OCT High-definition (spectral domain) OCT is a non-invasive imaging technique using Fourier analysis of echo time delay from a light source to obtain in vivo images of the anterior and posterior structures of the eye. Advances in OCT have improved the speed of the scan and the resolution of crosssectional images of the retina and optic disc to as low as three microns 27 (spectral domain) compared to early models of 12 to 15 μm 28 (time domain), allowing enhanced 14 Clinical and Experimental Optometry 2014 Optometry Australia

4 and these images correlate well with histology. 30,31 As results showed good repeatability, OCT provides a relatively accurate mode of detection and monitoring progression of a number of suspected optic nerve head disease processes, 29 including showing a subretinal hyporeflective space adjacent to the optic nerve head in true swelling of the optic nerve head (Figure 5, patient V) and increases in nerve fibre layer thickness parameters (Figure 6). Figure 2. Spectralis optical coherence tomography three-dimensional views highlighting optic nerve elevations and corresponding fundus appearance. Patient I: bilateral, small, crowded optic discs causing elevation of the optic nerve head due to space limitations of a small optic disc. Fundus appearance reveals indistinct disc margins with no disc cupping and temporal peripapillary atrophy. Patient II: bilateral, tilted discs as a result of the optic nerve entering the eye at an oblique angle. Patient III: unilateral, buried drusen of the right optic nerve head, indicating the elevation of the optic nerve head due to buried drusen, while only moderate elevation is present in the left eye. Patient IV: late-stage bilateral optic nerve head drusen with markedly raised optic nerve head and visible yellow reflective areas at the optic discs. Patient V: Papilloedema highlighting raised optic nerve heads bilaterally extending beyond the neuroretinal rim to involve the peripapillary area. Optic disc margins are indistinct with increased tortuosity and obscuration of the blood vessels. visualisation of the retinal structures, including the optic nerve head tissue. The emergence of OCT has allowed objective quantification and documentation of the retinal nerve fibre layer thickness by directly identifying this highly reflective layer in a 3.4 mm diameter circle centred on the disc which is compared to an instrument normative database. 29 Other optic disc parameters such as cup volume and rim area can also be quantified. 29 Optic disc morphology, including optic nerve axons and blood vessels, as well as termination of the retinal pigment epithelium (RPE) and Bruch s membrane and adjacent retinal tissue can be visualised with the OCT (Figure 4, panel A) Heidelberg retinal tomography The first Heidelberg retinal tomograph (HRT) introduced in the 1990s has now evolved into its third version: HRT3, which has a more extensive normative database than that of earlier models. 32 In contrast to cross-sectional differentiation of the retinal tissue obtained from the spectral domain OCT, which can be applied across the posterior pole, the HRT uses multiple consecutive focal planes, scanning the vitreo-retinal interface to produce surface topography of the optic nerve head and peripapillary region. Nerve fibre layer measurements of the optic disc are taken at an arbitrary reference plane of 50 microns posterior to the retinal surface at the temporal margin of the optic nerve head, with the area above the reference plane defined as the rim and below as the cup. 33 Quantitative assessment of the optic disc with the HRT, based on confocal scanning laser ophthalmoscopic (cslo) principles, 34 is performed with the Moorfields regression analysis (MRA), which provides a comparison of the patient s neuroretinal rim area with the predicted neuroretinal rim area in reference to the optic disc area and age in the normative database. 33 While the optic disc margins are automatically delineated on OCT instruments, the clinician is required to outline the disc contour on the HRT. The use of HRT3 has primarily been to detect and monitor the progression of glaucoma through analysis of its topographic change by evaluating surface height changes within the optic disc margin and stereometric trend analysis showing changes to the topographic parameter over time. 35 The threedimensional topographic construction showing optic nerve head height and retinal surface contour (Figure 4, panel B, Patients I and II) also has a useful role in assessing other conditions which affect optic nerve head morphology and cause optic nerve head elevation. 36 Clinical and Experimental Optometry 2014 Optometry Australia 15

5 Figure 3. Visual field restriction as a consequence of late-stage bilateral drusen of the optic nerve head. Humphrey visual fields analyser 30-2 central threshold reveal markedly constricted visual fields in each eye. Short wavelength fundus autofluorescence Short wavelength fundus autofluorescence (SW-FAF) involves stimulating the fundus with short wavelength light and assessing the light emitted by fundus fluorophores at longer wavelengths. This technique is regarded as a valuable diagnostic tool to provide insight into the metabolic status of the RPE by revealing the distribution and accumulation of lipofuscin (the chief fundus fluorophore in the healthy eye). There are several instruments which allow photography of SW-FAF; more commonly used are the confocal scanning laser ophthalmoscope (for example, Spectralis Heidelberg Retina Angiograph-HRA) and fundus cameras (for example, Topcon TRC-NW8F, Canon CR-2 PLUS, Optos 200Tx). Initial clinical use of the confocal scanning laser ophthalmoscope for SW-FAF imaging was described by von Ruckmann, Fitzke and Bird 37 in The excitation wavelength in the confocal scanning laser ophthalmoscopic system of the Heidelberg Retina Angiograph is 488 nm 38 and at longer wavelengths in the fundus cameras. For improved image contrast and to reduce background noise, the angiograph records a series of several single fundus autofluorescent images, aligns these images to correct for eye movements during acquisition, normalises the pixel values and calculates a mean image. 38 In contrast to a series of images, the fundus camera obtains a single image of a specific retinal area stimulated with one flash of short wavelength light. 39 In comparing SW-FAF images from the two systems, the fundus camera shows less intensity of autofluorescence at the blood vessels and optic nerve head 38 and less visualisation of SW-FAF patterns. 40 Different ocular structures display varying degrees of SW-FAF and can be termed as being hyperautofluorescent, which represents higher intensities than the background level (the homogenous level of autofluorescence generated in the posterior pole fundus of a healthy eye), and hence appears brighter, and vice versa hypoautofluorescent areas show lower intensities (than the background level) and appear darker. 39 In a normal eye, the optic nerve head is hypoautofluorescent, as there is inherently no RPE layer and therefore, absence of lipofuscin or other fluorophores (Figure 4, panel D, Patients I and II). 37 In contrast, the adjacent and surrounding fundus displays a homogenous level of moderate autofluorescence (the background level), except for the blood vessels and fovea, which are also darker than the background level in the normal eye. 37 This is due to masking of the SW-FAF signal by the blood (which contains no fluorophores) and the macular pigments lutein and zeaxanthin. 37 Variations from the normal SW-FAF distribution resulting in hyperautofluorescence arise from accumulation of the lipofuscin from the RPE or the presence of intrinsically autofluorescing deposits anterior or posterior to the RPE 39 or within the optic nerve head tissue (Figure 4, panel D, patient III, right eye [OD] and patient IV, OD and left eye [OS]). In contrast, signal-absorbing material such as intraretinal fluid (Figure 4, panel D, patient V) or loss of the retinal pigment epithelium result in a decreased SW-FAF signal. 39 Ultrasonography Initial applications of ophthalmic ultrasonography date back to with standardised echography developed in the early 1970s, which includes A-scan and B-scan ultrasonography. 42 Using a frequency of 16 Clinical and Experimental Optometry 2014 Optometry Australia

6 Figure 4. Findings of fundus examinations on five patients (A) Spectralis optical coherence tomography (OCT) scanning laser ophthalmoscopic images and horizontal scans taken through the centre of the optic discs; a detailed description of these scans are presented in Figure 5, (B) Heidelberg retinal tomography, (C) B-scan ultrasonography and (D) Fundus autofluorescence Patient I: bilateral, small, crowded optic discs. (A) Spectralis OCT scanning laser ophthalmoscopic image clearly outlines the peripapillary atrophy as bright crescents (arrows). (B) Retinal tomographic scans show the contour of the optic nerve head with peripapillary atrophy also highlighted (arrows). (D) Fundus autofluorescence is unremarkable. Patient II: bilateral, tilted discs. (A) Spectralis OCT scanning laser ophthalmoscopic image highlights the peripapillary atrophy, shown as bright crescents (arrows). (B) Retinal tomographic scans reveal the contour of the disc with a reduction in the contour height demonstrated to be the cup and adjacent peripapillary atrophy in a lighter area (arrows). (D) Fundus autofluorescence is unremarkable. Patient III: unilateral, buried drusen of the right optic nerve head. (A) The Spectralis OCT scanning laser ophthalmoscopic image shows a diffuse darkened peripapillary area with visible optic disc blood vessels in the right eye highlighting the elevated contour of the optic disc in contrast with the left eye. (C) The optic disc drusen in the right eye are clearly highlighted by a hyper-reflective region (arrow) in the ultrasonic B-scan and (D) hyper-autofluorescent areas (arrow). Results for the left eye are unremarkable on (C) B-scan ultrasonography and (D) fundus autofluorescence. Patient IV: late stage bilateral drusen of the optic nerve head. (A) There are round hyper-reflective deposits highlighted on the Spectralis OCT scanning laser ophthalmoscopic images (arrows). (C) B-scan ultrasonography reveals hyper-reflective areas at the optic nerves suggestive of optic disc drusen (arrows). (D) Autofluorescent imaging reveals hyper-autofluorescent areas within each optic disc. The left is denser than the right, indicating the presence of optic disc drusen (arrows). Patient V: papilloedema. (A) Spectralis OCT scanning laser ophthalmoscopic images show a diffuse darkening of the peripapillary area with less visible optic disc blood vessels highlighting the elevation of the optic disc. (C) Elevation of the optic disc is demonstrated on B-scan ultrasonography. (D) Fundus autofluorescence reveals dense hypo-autofluorescence extending beyond the peripapillary area. 10 Mhz, A-scans are useful in measuring the width of the optic nerve head, axial length and size of intraocular masses. 43 B-scan ultrasonography produces a two-dimensional cross-section of the globe, optic nerve and the orbit. 43 The normal optic nerve is described as a regular tubular structure with low homogenous reflectivity surrounded by highly reflective peri-neural sheaths and orbital fat. 44 As different ocular tissues display varying acoustic reflectivity of internal structures, tissue differentiation as well as topographic information can be gained with these methods (Figure 4, panel C, Patients III and IV). 44 Clinical and Experimental Optometry 2014 Optometry Australia 17

7 Figure 5. Comparison and description of horizontal spectral domain optical coherence tomography (OCT) scans through the central optic disc. Patient I: bilateral, small, crowded optic disc scans show no visible cupping and reveal a pseudo lumpy-bumpy contour due to blood vessel shadowing artefact (asterisk). Patient II: scans of bilateral tilted discs show the presence of optic disc cupping and highlight the height differences between the surface elevation of the temporal and nasal aspect of the optic disc. Patient III: unilateral, buried drusen of the optic nerve head. The right reveals a lumpy-bumpy appearance of the subretinal hyporeflective space (asterisk). Note that the contralateral left eye (OS) mimics that appearance due to vessel shadowing (arrow head). Patient IV: late-stage bilateral drusen optic nerve heads correspond with the drusenoid deposit contour with well-delineated hyper-reflective margins on the OCT scan (asterisk). Patient V: papilloedema reveals the characteristic V contour describing the subretinal hyporeflective space above the retinal pigment epithelium and intraretinal cystic spaces in the adjacent retinal architecture (asterisk). MULTIMODAL IMAGING OF THE ELEVATED OPTIC NERVE HEAD Studies have not yet shown how to distinguish with certainty, using optic nerve head imaging techniques, between a small crowded optic disc and subtle true swelling of the optic nerve head. In a study conducted by Karam and Hedges, 45 the authors found that even with OCT it was difficult to distinguish congenitally small crowded optic discs from mild papilloedema, as the nerve fibre layer thicknesses, although thicker than normal, were not statistically different from small crowded discs and mild papilloedema. The authors 45 suggested this could be due to a similar underlying mechanism of slowed axoplasmic flow leading to swelling of the 18 Clinical and Experimental Optometry 2014 Optometry Australia

8 Figure 6. Peripapillary retinal nerve fibre layer thickness measurements from the spectral domain optical coherence tomography. Patient I: bilateral, small, crowded optic discs cause sectoral thickening of the retinal nerve fibre layer due to the space limitations of the small disc. The inferotemporal retinal nerve fibre layer regions are thicker than normal limits in this patient. Patient II: bilateral, tilted discs reveal sectoral retinal nerve fibre layer thickening, in particular the temporal areas of each eye due to the obliquely inserted and tilted nature of the optic discs resulting in a temporal shift on the retinal nerve fibre layer thickness profile. There are four contiguous retinal nerve fibre layer clock hours thicker temporally and two contiguous retinal nerve fibre layer clock hours thicker nasally (right eye: OD). There are three contiguous retinal nerve fibre layer clock hours thicker temporally (left eye: OS). Patient III: unilateral, buried drusen of the optic nerve head. The right eye (OD) reveals five contiguous retinal nerve fibre layer clock hours of thickening nasally and one retinal nerve fibre layer clock hour of thinning superotemporally. The left eye (OS) mimics partial thickening of three contiguous retinal nerve fibre layer clock hours inferonasally and two retinal nerve fibre layer clock hours superiorly due to a tilted disc. Patient IV: late-stage bilateral drusen of the optic nerve heads cause extensive areas of retinal nerve fibre layer thinning with eight contiguous retinal nerve fibre layer clock hours of retinal nerve fibre layer thinning in the right eye and up to four retinal nerve fibre layer clock hours of thinning in the left. Patient V: papilloedema presents with severe and extensive thickening of the retinal nerve fibre layer with 10 retinal nerve fibre layer clock hours of thickening in the right disc and all retinal nerve fibre layer clock hours thickened in the left. axons. In patients with suspected crowded optic discs, OCT results have been used to monitor the thickness of the nerve fibre layer over time to confirm stability (Figure 6, Patient I). 46 The small crowded optic disc appears elevated on OCT with no apparent cupping and posterior shadowing of the blood vessels at the optic disc (Figure 5, Patient I). Heidelberg Retinal Tomography can measure the size of the optic disc, thus confirming a small disc (Figure 4, panel B, Patient I); however, it is difficult to exclude co-existing true swelling of the optic nerve head with this technique. Due to the limitations of the HRT normative database, optic Clinical and Experimental Optometry 2014 Optometry Australia 19

9 disc size was found to influence the sensitivity of the HRT (poorer sensitivity for small and large discs), and hence results for a small crowded disc need to be evaluated with caution. 47 Echographic (ultrasonic) features of the small crowded optic disc are non-specific. Overall, the use of multimodal imaging to differentiate between small crowded optic discs and true swelling of the optic nerve head, from the imaging techniques described, involves confirmation of a small optic nerve head (OCT or HRT) and (a diagnosis of exclusion) absence of other findings suggestive of optic nerve head oedema, such as peripapillary nerve fibre layer thickening with OCT or adjacent subretinal hyporeflective space. A tilted disc can also give the appearance of an elevated optic nerve head (Figure 2, Patient II). Studies conducted to quantify the difference in thickness of the retinal nerve fibre layer of the tilted disc compared to normal eyes have shown that there is a significant reduction of the superior nerve fibre layer (Figure 6, Patient II) 48 and a blunted superior peak in inferiorly tilted discs; 49 however, results have also demonstrated that there is no association between optic disc tilt and altered pattern of the nerve fibre layer 50 and mixed results in terms of average nerve fibre layer thickness. 49,51 As the optic disc margins are automatically delineated on some OCT instruments, peripapillary atrophy can be incorrectly included in the measurements, hence overestimating the disc area compared to the confocal scanning laser ophthalmoscope. 52 Aside from abnormal distribution of the thickness of the retinal nerve fibre layer, choroidal thickness is thinner adjacent to the depressed sectors of the optic disc. 10 There are several methods to quantify optic nerve head tilt, such as that based on surrogate parameters of ovality indices 53 and more recently, with high-resolution crosssectional images of the optic disc on spectral domain OCT (Figure 5, Patient II) to objectively measure the horizontal 51 and vertical 54 tilt angle with a defined reference plane using anatomical landmarks, such as Bruch s membrane. With the original HRT, all parameters of the tilted disc identified in Singaporean children were outside normal limits based on the HRT parameters 55 and it is suggested that this is due to an inaccurate placement of the reference plane in the tilted disc. 9 Features of the tilted disc measured with the original HRT include a reduction in these parameters, namely disc size which is inversely proportional to the degree of tilt, 52 rim and cup area, cup-to-disc area ratio, cup volume and cup depth and an increase in the following parameters: rim volume, rim-todisc area ratio, nerve fibre layer thickness and volume. 55 Echographic (ultrasonic) features of the tilted optic disc described by Singh 56 included increased size of the dural optic nerve width on A-scan and up to a doubling in size of the optic nerve shadow detected on B-scan as well as exaggeration of the B-scan ultrasonic signal; however, these characteristics are yet to be confirmed by other studies. 9 The role of SW-FAF in detecting tilted optic discs is not relevant, as there is no optic nerve head tissue differentiation (Figure 4, panel D, Patient II). The use of multimodal imaging in differentially diagnosing and monitoring tilted optic discs is similar to that for small, crowded optic discs described earlier, in that optic disc imaging can confirm an abnormal nerve head, in this case a tilted disc, but a diagnosis of pseudopapilloedema is a diagnosis of exclusion of other findings, such as peripapillary nerve fibre layer thickening and adjacent subretinal fluid. The use of imaging for this purpose is somewhat restricted due to the anomalous nature of tilted discs and the lack of a validated normative database for tilted discs in instruments such as the OCT and HRT. While the use of multimodal imaging has limitations in its use in the differential diagnosis of small crowded discs and tilted discs from true swelling of the optic nerve head, B-scan ultrasonography and SW-FAF can be used to definitively detect and diagnose optic disc drusen. B-scan ultrasonography is the gold standard in the diagnosis of disc drusen, first introduced for this role in the 1970s. 14 It is found to be superior in detecting optic nerve head drusen than orbital computed tomography scan and preinjection control photography. 57 This method is more sensitive for detecting buried drusen than funduscopy or SW-FAF; however, a skilled practitioner is important in detecting smaller drusen. 14 The role of SW-FAF in differentially diagnosing conditions of optic nerve head elevation arises from its ability to detect and confirm the presence of optic disc drusen based on hyperautofluorescence of the calcific properties of drusen. SW-FAF had the highest level of sensitivity and specificity compared to red and green filters on retinal photography (sensitivity 88 per cent, specificity 100 per cent) when differentiating optic disc drusen from optic nerve head oedema. 58 As fundus autofluorescence is a relatively non-invasive technique; its role in detecting optic nerve head drusen in children is found to be particularly effective. 59 Similar to B-scan ultrasonography, the calcified drusen appear as hyper-reflective round deposits at the optic disc on SW-FAF (Figure 4, panel C and D, Patient III, OD and Patient IV, OD and OS). With spectral domain OCT, optic disc drusen were morphologically characterised as hyper-reflective subretinal masses associated with hyporeflective spaces or cysts with calcified hyper-reflective walls. 60 The height of the drusen can be quantified by comparing its perpendicular height to the RPE. 60 Johnson and colleagues 61 and Flores-Rodriguez and colleagues 62 have discussed the use of OCT to differentiate papilloedema from drusen of the optic nerve head. One of the main qualitative observations from these studies was the characterisation of the subretinal hyporeflective space located between the neurosensory retina and RPE/choriocapillaris complex at the margin of the optic nerve head in papilloedema. This is referred to as a lazy V contour (Figure 5, Patient V). 61 It has been proposed that the subretinal hyporeflective space is attributed to leakage of fluid from the optic nerve head; however, the exact mechanism underlying the development of the subretinal hyporeflective space remains unclear. Arguably, this space is also hypothesised to be merely posterior shadowing due to the hyper-reflectivity of the retinal nerve fibre layer. 63 Swelling of the optic nerve head can also be associated with cystic spaces in the adjacent retinal architecture (Figure 5, patient V). In contrast, drusen of the optic disc were found to have an irregular lumpy-bumpy contour with an abrupt decrease in the subretinal hyporeflective space on OCT scans, which may correspond with the contour of drusenoid deposits (Figure 5, Patient III, OD and Patient IV, OD and OS). 61 We found it occasionally difficult to distinguish the proposed lazy V contour or lumpy-bumpy appearance of the OCT scan of the optic nerve head. It is also important to note that other observations on OCT caused by anatomical anomalies, such as tilted optic discs or a small crowded optic disc, or features inherent to the imaging modalities can 20 Clinical and Experimental Optometry 2014 Optometry Australia

10 complicate clinical interpretation. For example, superficial blood vessels can cast a shadow onto underlying tissues and mimic hyporeflectivity of the subretinal layer (Figure 5, Patient I, OD and OS and Patient III, OS). In this subset of patients, it can be challenging to distinguish whether the resulting OCT pattern is attributed to pathology or merely noise due to limitations of the equipment. Looking at other OCT characteristics to quantitatively assess the disc for papilloedema, a key feature of papilloedema is the presence of at least seven contiguous clock hours of increased thickness of the retinal nerve fibre layer on OCT, yielding a sensitivity of 98 per cent and specificity of 77 per cent relative to a population of healthy optic nerve heads (Figure 6, Patient V). 64,65 Increased thickness of the nasal nerve fibre layer, with thresholds ranging from 78 μm 63 to 108 μm 62 and the subretinal hyporeflective space to be more than 127 μm in height (sensitivity 80 per cent, specificity 70 per cent) were also noted as differential markers in patients with true swelling of the optic nerve head. 61 Thickening of the retinal nerve fibre layer was more prominent in patients with severe papilloedema; however, measurement of nerve fibre layer thickness was less sensitive in detecting mild papilloedema than the peripapillary total retinal thickness, which measures the total retinal thickness from the internal limiting membrane to the pigment epithelium. 66 Changes in the nerve fibre layer thickness are dependent on the chronological stage of the disease and the affected individual, if no previous comparative measurements are available. Congenital anatomical variations, such as tilted discs, can also mimic focal thickening and bias measurements, if changes due to disease occur (Figure 6, Patients I and II). In comparison, a reduced nerve fibre layer on OCT indicates the presence of retinal ganglion cell atrophy and is usually observed in inflammatory conditions, such as optic neuritis albeit in the later stages 67 and late optic disc drusen (Figure 6, Patient IV). 68 There is limited evidence to support the use of HRT in differentially diagnosing papilloedema from pseudopapilloedema. Trick and colleagues 69 investigated the quantification of disc elevation by measuring the volume of the disc above the retinal surface in pseudopapilloedema and papilloedema using the original HRT. They found that the volume is greater in patients with papilloedema, with most of the tissue elevation extending beyond the disc rim and peripapillary retina. Several case reports have illustrated the role of retinal tomography to objectively monitor papilloedema and assess success of treatment using the three-dimensional reconstruction of the optic disc. 70,71 Typically as papilloedema resolves, there is a marked decrease in surface elevation of the optic nerve head seen on the HRT along with changes in the disc volume, which may signify a decrease in intracranial pressure or optic atrophy. 71,72 In terms of SW-FAF, in an optic disc with papilloedema, the hypo-autofluorescence pattern involves not only the disc itself but also extends beyond the peripapillary area due to the marked elevation and swelling of the optic nerve (Figure 4, panel D, Patient V). The efficacy of optic nerve ultrasonography for differentiating papilloedema from pseudopapilloedema has also been explored. 73 Methods to assess the optic nerve for papilloedema included A-scan optic nerve width measurements, cross-sectional B-scans revealing a doughnut or a crescent sign, which signifies a ring of fluid around the optic nerve and the 30 test performed with A-scan. 44 The elevation of the swollen optic disc can also be appreciated on B-scan ultrasonography as a protrusion into the vitreous cavity (Figure 4, panel C, Patient V). 74 Ultrasonography is less effective in detecting papilloedema (85 per cent sensitivity and 83 per cent specificity) at a normal optic nerve width of up to 3.3 mm; however, a higher degree of sensitivity and specificity was achieved when the normal optic nerve width was set at 3.0 mm or above. 73 A similar protocol used in a study of children suggested that ultrasound was as conclusive as lumbar puncture for diagnosis of intracranial hypertension but far less invasive. 75 A positive result on the 30 test describes a net reduction of optic nerve width of greater than 10 per cent between maximum measurements of the optic nerve width at primary gaze, signifying a distended optic nerve, where subarachnoid fluid is increased and when measured again at 30 lateral gaze correlating with redistribution of the subarachnoid fluid due to stretching of the nerve. 44 Causes of a positive 30 test are not limited to intracranial hypertension but also include optic neuritis, uveal effusion syndrome and an apical mass. 44 Where a negative 30 test is described, these results indicate solid sheath or nerve thickening caused by conditions such as optic nerve glioma and nerve sheath meningioma. 44 As the resolution of crosssectional B-scans is fairly low and measurements of the optic nerve width are reliant on the skill of the ophthalmic practitioner, mild papilloedema may be difficult to diagnose with these methods and hence a challenge to decisively exclude. 74 SUMMARY AND CONCLUSION Differentially diagnosing between pseudopapilloedema and true swelling of the optic nerve head is an important task for ophthalmic practitioners in deciding who will be referred for other urgent assessment, such as neuroimaging procedures, and who can be monitored routinely. If clear signs and symptoms associated with optic disc oedema, medical history and a thorough eye examination are not sufficient to point to a definitive diagnosis, then OCT, SW-FAF and B-scan ultrasonography are adjunctive tests which can help assess and/or precisely quantify the disc and retinal nerve fibre layer features. Table 2 gives a summary of the overall utility of multimodal imaging. In any eye with optic nerve head elevation, these imaging tests may confirm for example, true swelling of the optic nerve head through a lazy V contour and significant peripapillary nerve fibre layer thickening. The absence of findings suggesting true swelling of the optic nerve head and the presence of a small optic disc, tilted optic disc or optic nerve head drusen point to pseudopapilloedema. With regards to optic disc drusen, OCT imaging and thinning of the nerve fibre layer may suggest their presence, but B-scan and/or SW-FAF are key to confirming the diagnosis. The misdiagnosis of true swelling of the optic nerve head carries potential sight- and life-threatening outcomes. We conclude that current commercially available optic nerve head imaging techniques such as spectral domain OCT, SW-FAF and ultrasonography can be useful in improving the differential diagnosis of optic nerve head elevation in clinical practice. Pseudopapilloedema remains a diagnosis of exclusion, as in certain cases the possibility of true swelling of the optic nerve head superimposed on a small or tilted optic disc needs to be considered. Therefore, while multimodal imaging can assist in differentially diagnosing pseudopapilloedema, this is still in part a diagnosis of exclusion. Aside from definitively diagnosing optic disc drusen with fundus autofluorescence and Clinical and Experimental Optometry 2014 Optometry Australia 21

11 Disc pathology Role of multimodal imaging Spectral domain OCT HRT SW-FAF Ultrasonography Conclusion Small crowded disc Can confirm a small ONH Monitor whether nerve fibre layer is thicker than normal and to confirm stability Limitations Difficult to exclude co-existing true swelling of the ONH Retinal nerve fibre layer thicknesses not statistically different between small crowded discs and mild papilloedema Tilted optic disc Can quantify ONH tilt angle using a defined reference plane Retinal nerve fibre layer thickness can show a reduction of superior fibre layer and a blunted superior peak in inferiorly tilted optic discs Limitations Peripapillary atrophy can be incorrectly included hence overestimating disc area on spectral domain OCT Variable results for retinal nerve fibre layer thickness on spectral domain OCT Lack of validated normative database for tilted discs Optic disc drusen Lumpy bumpy contour of the subretinal hyporeflective space of drusen Thicker nerve fibre layer indicates early ONH drusen and thinned nerve fibre layer shows late stage disc drusen Limitations Can be difficult to distinguish a lazy V contour from a lumpy bumpy contour Anatomical ONH anomalies such as a small crowded disc or tilted optic disc and superficial blood vessels can mimic hyporeflectivity of the subretinal layer True swelling of the ONH Lazy V contour of the subretinal hyporeflective space of true swelling of the ONH Cystic spaces in the adjacent retinal architecture indicates ONH swelling At least seven contiguous clock hours of increased RNFL thickness indicates papilloedema Limitations Can be difficult to distinguish a lazy V contour from a lumpy bumpy contour Anatomical ONH anomalies such as a small crowded disc or tilted optic disc, and superficial blood vessels can mimic hyporeflectivity of the subretinal layer Measurements of nerve fibre layer thickness are less sensitive in detecting mild papilloedema. Confirms a small ONH Not relevant as there is no ONH tissue differentiation Limitations of normative database for small discs Parameters are generally outside normal limits Inaccurate placement of the reference plane in the tilted disc on HRT Lack of validated normative database for tilted discs Limited use in the diagnosis and detection of ONH drusen as the HRT measures surface topography Limited evidence in detecting and diagnosing ONH drusen Volume of the disc is greater in papilloedema. Monitor and assess success of treatment using three-dimensional reconstruction of optic disc. Marked decrease in surface height elevation signifies decrease in intracranial pressure or optic atrophy. Limited evidence to detect and diagnose true swelling of the ONH Limited use in diagnosing a small crowded disc and differentiating between a small crowded disc and true ONH swelling Not relevant as there is no ONH tissue differentiation Limited use in diagnosing a tilted optic disc and differentiation between true swelling of the ONH Non-invasive method to definitively detect and diagnose drusen of the ONH based on hyper-autofluorescence. Less sensitive than ultrasonography for detecting buried drusen Hypo-autofluorescence pattern extends beyond the peripapillary area due to marked elevation and swelling of the optic nerve. Limited evidence in the diagnosis of true swelling of the ONH. Not relevant as there is no ONH tissue differentiation Ultrasonographic characteristics are non-specific for clinical diagnosis An increased size of the dural optic nerve width on A-scan Doubling in size of the optic nerve shadow on B-scan. Ultrasonography characteristics are yet to be confirmed with further studies Gold standard for diagnosing and detecting ONH drusen based on acoustic hyper-reflectivity. More sensitive that funduscopy or SW-FAF for detecting buried drusen Requires a skilled practitioner to detect buried, smaller-sized drusen due to the low resolution of cross-sectional B-scans. A-scans with normal optic nerve width set at 3.0 mm. B-scan shows a doughnut or crescent sign A positive 30 test performed with A-scan Ultrasonographic results, in particular optic nerve width measurements, are reliant on the skill of the ophthalmic practitioner and resolution of B-scans are fairly low. A small crowded disc is a diagnosis of confirmed small disc and absence of other findings suggestive of ONH oedema. A tilted disc is a diagnosis of exclusion in the absence of other findings suggestive of oedema of the ONH. ONH drusen can be definitively detected and confirmed with SW-FAF and ultrasonography. True swelling of the ONH can in many cases be detected using multimodal imaging, but should be interpreted in conjunction with other clinical findings. OCT: optical coherence tomography, HRT: Heidelberg retinal tomograph, SW-FAF: Short wavelength fundus autofluorescence; ONH, optic nerve head, RNFL: retinal nerve fibre layer Table 2. Summary of overall utility of multimodal imaging for differentiating pseudopapilloedema and true swelling of the ONH 22 Clinical and Experimental Optometry 2014 Optometry Australia