New Frontiers in Glaucoma Imaging

Figure 1 (above). A 70-year-old patient with primary open angle glaucoma seen in the Centre for Eye Health Glaucoma Management Clinic. The patient demonstrated glaucomatous structural progression over time on both the retinal nerve fibre layer (RNFL) and ganglion cell inner plexiform layer (GCIPL) thickness measurements, as measured using the Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA). By exam, eight of the instrument’s guided progression analysis, there was marked thinning of both RNFL (black arrow) and GCIPL (orange arrow), with very little measurable thickness remaining. In contrast, the radial peripapillary capillary (RPC) density measurement obtained using the angiography model of the same Cirrus HD-OCT device shows appreciable remaining vascularity (red arrows), demonstrating that OCTA allows further detection and monitoring of progression where the measurement floor is reached with conventional structural OCT parameters.

Glaucoma is the leading cause of irreversible blindness worldwide, and remains a significant public health problem, as half of all cases remain undiagnosed.¹ Early detection is key in the diagnosis and management of the disease. Current clinical guidelines for the assessment of glaucoma comprise a battery of tests, which include imaging of the neural structures of the eye.² Two imaging technologies that are prolific and key in current clinical paradigms for diagnosing and managing glaucoma are colour fundus photography and optical coherence tomography (OCT).³ Despite the usefulness and indispensable nature of these imaging modalities, there remain several issues in assessing the eye for glaucoma. Some of these issues and questions are:

• Is measuring blood flow using OCT Angiography(OCTA) useful in diseases such as glaucoma?
• What is the usefulness of imaging deeper ocular structures like the lamina cribrosa and the anterior chamber angle?
• Are there any emerging technologies which allow cellular resolution and the possibility of detecting cellular dysfunction before cell death?
• What is the usefulness of resolving ocular details at the cellular level and detecting cellular dysfunction before cell death?

OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY

OCTA is a non-invasive imaging modality allowing qualitative and quantitative assessment of the ocular vasculature. Initial use of OCTA was in the qualitative assessment of the retinal vasculature as a non-invasive alternative to fluorescein angiography. More recently, quantification (and therefore normative comparison) of vascular parameters, such as density and regularity, has become more accessible. Therefore, like structural OCT data, quantifiable OCTA parameters have been studied for their potential in identifying glaucoma and its progression.

Aside from the potential benefits of distinguishing between glaucomatous and healthy eyes, OCTA has also been studied as a technique for monitoring glaucoma over time. A specific advantage of OCTA over conventional structural OCT is its ability to measure structural change when retinal nerve fibre layer (RNFL) and macular thickness measurements have reached the measurement floor in moderate to advanced glaucoma. At that stage in the disease, vessel density measured using OCTA has been shown to be useful when continuing to monitor for structural change due to its greater dynamic range.⁴ An example of the application of OCTA in advanced glaucoma is shown in Figure 1. In this example, structural OCT parameters are unable to further quantify glaucomatous change after reaching the floor, but remaining measurable vascular change can be quantified using OCTA, making it a potentially useful tool for measuring future structural progression.

In the earlier stages of glaucoma, some studies have suggested that OCTA may be able to detect subtle changes in retinal microvasculature prior to the onset of visual field damage.^5,6 The reduced capillary density visualised using OCTA is thought to indicate vascular stress that may be a pathophysiological pathway in the development of glaucoma.⁵

Figure 2: Three patients seen in the Glaucoma Management Clinic referred for diagnosis and management and their respective structural imaging results (top row: colour fundus
photography; middle row: optical coherence tomography (OCT) thickness map; bottom row: superficial vascular plexus of OCT angiography (OCTA)). Left column: a patient diagnosed with primary open angle glaucoma (POAG). The patient has clear inferior notching on fundus examination (blue arrow) and adjacent retinal nerve fibre layer (RNFL) loss (white arrow). The deep structural loss is also seen on OCTA (right column, red dashed outline), with a gradient of change in vascular density. Middle column: a patient diagnosed with ischaemic RNFL loss. The patient has an intact neuroretinal rim (blue arrow) and focal, well-defined RNFL loss (white arrows). The vascular loss was also clearly focal and located within the superficial retina when measured using OCTA (bottom row, red dashed outline). Right column: a patient diagnosed with a branch retinal vein occlusion (BRVO). The patient has clearly attenuated retinal veins and collateral vessels across most of the superior hemifield on fundus examination (blue arrows) and broad, deep loss of the RNFL in the corresponding area (white dashed border). The OCTA result highlights the extent of loss of perfusion extending from the 10 o’clock to the 4 o’clock position (bottom row, red dashed outline).

In addition to its dynamic range for measuring disease progression in glaucoma,⁷ the application of OCTA for the differential diagnosis and management of optic nerve disease focuses on recognising the pattern (location and depth) of the vascular loss. OCTA allows the clinician to visualise the extent and the layers of non-perfusion. In glaucoma, there is a gradual reduction in the superficial vessel density within the peripapillary and macular areas (Figure 2, left column). The superficial vascular plexus is of main interest in glaucoma, as it is the region of vasculature that primarily supplies the neural elements relevant to glaucoma.^6,7 In comparison, a primary ischaemic cause of optic atrophy, such as non-arteritic anterior ischaemic optic neuropathy, leads to a deep, focal area of vascular loss in the absence of neuroretinal rim changes (Figure 2, middle column). Similarly, a retinal cause of neural loss, such as in branch retinal vein occlusion, leads to deep widespread loss in the affected retinal area (Figure 2, right column). Thus, OCTA may provide the clinician with an additional tool with which to differentially diagnose non-glaucomatous causes of RNFL loss from glaucoma.

Limitations 

A main limitation of OCTA is its susceptibility to imaging artefacts, more so than conventional structural OCT. Qualitative and quantitative measurements are particularly confounded by the presence of movement, segmentation, and projection errors. Examples of such artefacts are shown in Figure 3. OCTA is typically conducted using high-speed OCT devices, but because it relies on the motion of red blood cells to interpolate vascular ‘flow’, even subtle movements can cause interruptions to the resultant scan (Figure 3, red arrow). Larger truncation artefacts can occur similarly to structural OCT (Figure 3, yellow arrow). A major issue in OCTA is the presence of media opacities or location-specific interruptions to the scan, which may cause artificial ‘reductions’ in vascular perfusion (Figure 3, blue arrows). A sufficiently prominent media opacity, such as cataract, can also manifest with a reduction in signal strength (Figure 3, green arrow and border; this scan had a signal strength of 6/10), which results in the appearance of a generalised loss of perfusion. The future steps for OCTA include advanced algorithms that compensate for these artefacts to increase the fidelity of the result. For now, clinicians need to be aware of the presence of these artefacts and, like in conventional structural OCT, exercise caution when interpreting these results.

SWEPT SOURCE OCT

A known limitation of conventional spectral domain OCT (SD-OCT) is the problem with imaging deeper anatomical structures, such as the angle recess, lamina cribrosa and choroid (Figure 4). Swept source OCT (SS-OCT), using a longer wavelength of light than SD-OCT, has the potential to acquire images at a higher speed and with better contrast and resolution compared with SD-OCT.⁸ With the longer wavelength, the SS-OCT signal has been shown to be able to penetrate deeper ocular structures, such as facilitating better visualisation of the choroid and potentially choroidal neovascularisation.⁹

Two potential applications of SS-OCT in glaucoma include imaging of the anterior chamber angle and the lamina cribrosa. The greater, deeper penetration compared to SDOCT allows the anterior chamber angle to be more accurately characterised,¹⁰ such as the scleral spur, which is critical for quantitative analysis of the anterior chamber angle.¹¹ The diagnostic accuracy of SS-OCT is similar to that of SD-OCT when considering conventional structural parameters like retinal thickness values. In one study, SSOCT was found to show higher diagnostic accuracy when compared to SD-OCT in detection of RNFL and ganglion cell analysis (GCA) defects in myopic glaucomatous eyes,¹² whereas in another study, SS-OCT and SD OCT were found to have similar glaucoma diagnostic abilities, with the SD-OCT having possibly higher detection ability of the ganglion cell inner plexiform layer.^13,14 However, the advantage of SSOCT in the posterior segment is its ability to better resolve the detail of the lamina cribrosa, which is an important structure in glaucoma. Currently, SD-OCT permits a view of the optic cup and parts of the anterior lamina cribrosa, but deeper details, such as the middle to posterior portions of the lamina cribrosa, are more challenging to visualise (Figure 4). Parameters, such as lamina orientation and pore size, have been suggested to be different in glaucomatous eyes compared to healthy eyes.¹⁵ 

The main drawback of SS-OCT is the cost associated with the light source used. SSOCT necessitates the use of a narrow beam with higher speed, which can lead to worse signal to noise ratio and may be more prone to motion interferences.¹⁶ 

Figure 3. Examples of artefacts seen on OCT on the radial peripapillary capillary density result (left) and macular superficial slab (middle and right). The red arrow indicates a motion artefact. The yellow arrows indicate truncation artefacts (due to blinking). The blue arrows indicate small localised media opacities that have led to the appearance of localised loss of perfusion (as no signal is returned). The green outline and arrow delineate an area obscured by cataract, leading to a signal strength of 6/10. The resultant area, delineated by the green outline, appears to have reduced perfusion due to the low signal strength.

HYPERPARALLEL OCT

Hyperparallel (HP) OCT technology splits the incident beam into beamlets, allowing higher incident power. The result is better signal strength and resolution, and the minimisation of artefacts caused by eye movement. Leading the way is the Australian team behind the Cylite HP OCT, the only hyperparallel OCT currently commercially available.¹⁷ The ultrahigh scan speed of more than 300,000 A-scans per second facilitates full volumetric anterior and posterior imaging. An example of its application is full assessment of the anterior chamber and the angle configuration in a single snapshot. Another potential advantage is its ability to provide an overall impression of ocular biometry. Parameters, such as axial length and eye shape, play an important role in understanding the integrity of the retina in glaucoma, and the HP OCT can provide clinicians with this important information.

As an emerging technology, there is still a need for further research in the effectiveness of HP OCT and its applications in glaucoma.

ADAPTIVE OPTICS OCT

One of the main barriers to capturing high resolution images of ocular tissues, using conventional imaging modalities, is the optical aberrations naturally present in human eyes.¹⁸ Adaptive optics (AO) technology has long been used in astronomy to reduce granular, blur, and speckle artefacts to enhance image quality. AO technology has been applied to various imaging technologies of the eye, as early as retinal cameras in the nineties with the development of optical aberrometers.^19,20 AO OCT is one of the latest technologies to integrate adaptive optics.

AO OCT has been shown to be able to visualise RNFL bundles, retinal ganglion cells, and retinal vasculature at much higher resolution compared to conventional OCT. Deeper retinal layers, such as the photoreceptor mosaic and retinal pigment epithelium, can also be imaged.²¹ This capability opens the possibility of identifying loci of structural change and potentially even distinguishing phenotypes of affected retinal ganglion cells.22 In the anterior segment, AO OCT has been demonstrated to enhance views of the anterior chamber angle and the trabecular meshwork.²¹ 

Figure 4. An example of a limitation of scan depth using conventional OCT scan
protocols illustrated using a patient with pachychoroid neovasculopathy with
both pigment epithelium and neurosensory retinal detachment. (A) A standard
high-definition OCT line scan through the fovea clearly shows the detachments, but
the underlying choroid and its thickness is difficult to image (yellow outline). (B) An
enhanced depth imaging (EDI) scan which increases penetrance through the same
area highlights additional choroidal detail, including its depth (red arrow), and
allows visualisation of its posterior border (red outline).

Currently, the integration of AO OCT into clinical practice is limited by the small field of view. Many scans are required to compose an image, meaning that longer acquisition time is necessary. Additionally, due to a restricted depth of focus, a compilation of scans with adjustment of the focal plane is required to create retinal structures and volumetric data. Also, as an imaging modality that examines the optical properties of the eye, the high resolution of AO OCT only provides information on the reflectance properties of the ocular tissues, and is not a true histological section of the eye.

DETECTION OF APOPTOSING RETINAL CELLS IMAGING

Programmed retinal ganglion cell death, or apoptosis in glaucoma, is triggered through a disease process such as retinal ischemia, axonal damage, or change in the lamina cribrosa structure. Since OCT provides information regarding the reflectance properties of the neural layers, it has been hypothesised that an earlier part of the cell death process in glaucoma could be detected with an alternative, suitable biomarker. Such a marker would be clinically useful in potentially identifying an earlier point for intervention to prevent further loss. Detection of Apoptosing Retinal Cells (DARC) technology aims to detect apoptosis of retinal ganglion cells prior to the event of programmed cell death. By using a fluorescent biomarker and confocal scanning microscope, DARC imaging allows in-vivo quantification of apoptosing cells. It can visualise apoptosis at its initial stages, thus management of glaucoma can be commenced prior to the visualisation of perimetric damage. Results of both Phase 1 and Phase 2 clinical trials using DARC imaging and AI show promising results with researchers able to prognosticate glaucomatous damage 18 months prior to that noted on OCT.^23,24 

CONCLUSION

The future is here, but we just don’t know it yet. Just over a decade ago, OCT was scarce in primary care practices. Today, it is prolific and highly accessible to clinicians. OCTA and SS-OCT have slowly increased in usage where they may be used as complementary tools to stereophotography, OCT and visual fields in glaucoma diagnosis. In the same way, we can look forward to HP OCT, AO OCT and DARC imaging gradually becoming more common in glaucoma diagnosis and management, with the key goal being early detection.

The authors thank Professor Michael Kalloniatis for useful and scholarly discussions on the subject matter. 

Elizabeth Wong BOptom (Hons), MOptom, GradCertOcTher, FAAO is a senior staff optometrist at the Centre for Eye Health, University of New South Wales. She has worked in a variety of clinical settings including ophthalmology practice. She received her Bachelor of Optometry from UNSW and subsequently completed her Graduate Certificate in Ocular Therapeutics and Master of Optometry. Ms Wong has been involved in continuing education events both locally, through presentation of CFEH webinars, as well as abroad, at American Academy conferences. She has a special interest in glaucoma and collaborative care, and administers the CFEH collaborative care glaucoma management clinic. Ms Wong has authored several papers published in peer reviewed journals and is a Fellow of the American Academy of Optometry. 

Dr Jack Phu BOptom (Hons), BSc, MPH, PhD, FAAO, OGS Ezell Fellow, Diplomate (Glaucoma) is a clinician-scientist and the head of the glaucoma/ neuro-ophthalmology unit at the Centre for Eye Health, and a Lecturer at the School of Optometry and Vision Science, University of New South Wales. His clinical, research and teaching duties are devoted almost exclusively to the care of patients with glaucoma. Dr Phu’s teaching duties include ocular diseases, therapeutics and visual neuroscience at the undergraduate and postgraduate levels. His research focuses on glaucoma, retinal disease, structure-function relationships in the eye and visual psychophysics. He has published his research extensively in scientific journals, and he has been awarded numerous prizes for his research, including the Optometric Glaucoma Society Ezell Fellowship in 2016.