How to optimise glaucoma detection in everyday practice

Raised intraocular pressure is one of many risk factors for primary open-angle glaucoma. The higher the IOP, the greater the risk of glaucoma. An intraocular pressure greater than 22 mmHg without any glaucomatous damage classifies as ocular hypertension. Patients with this condition do not have glaucoma, but they need closer monitoring because of their increased risk of developing it. At the opposite end of the intraocular pressure scale, it is possible to have a pressure within the normal range (10-22 mmHg) associated with signs of glaucomatous damage. This classifies as normal-tension glaucoma.

Three theories of glaucomatous damage help explain why some patients have primary open-angle glaucoma despite having low intraocular pressure. And why others do not have glaucoma despite having high intraocular pressure.

• Vascular theory: Retinal ganglion cells (RGC) are damaged by insufficient ocular blood flow due to either raised intraocular pressure or low perfusionof the capillaries that supply the optic nerve head. The resulting ischemia damages the retinal ganglion cell axons (the retinal nerve fibre layer – RNFL).  This theory helps explain the existence of normal tension glaucoma with low blood pressure and poor circulation associated with an increased risk of glaucoma.  But ischemic damage of the axons can also occur with high systemic blood pressure by causing haemorrhages – effectively creating a mini stroke within the optic nerve head, limiting blood supply to the axons. Read more here.

• Mechanical theory: This theory proposes that the retinal ganglion cells are damaged by increased intraocular pressure, either directly or indirectly.
  Raised intraocular pressure exerts mechanical pressure and distorts the lamina cribrosa (the connective tissue structure within the optic nerve head). The retinal ganglion cell axons pass through the pores of this structure as they leave the globe. The axons may be damaged directly by the high pressure or indirectly as a result of lamina cribrosa distortion and compression. Read more here.
• Biochemical theory: Various biochemical mechanisms may also play a role in glaucomatous neurodegeneration. These include damage by nitric oxide and oxygen free radicals (Ahmad 2016).

In reality, it is likely that the pathogenesis of glaucoma is multifactorial and varies between patients. For any glaucoma patient, one, two or all three of the above theories can help explain their glaucoma pathogenesis. Understanding this better may lead to improvements in glaucoma treatment in the future.  For now, however, the only real treatment we have for primary open-angle glaucoma is eye drops or procedures (lasers/surgery) to lower the intraocular pressure.

Do you remember when you were learning about the measurement of intraocular pressure? The applanation tonometry formula is based on the assumption that the eye is a dry, thin-walled sphere. Well, you will probably have gathered by now that it isn’t dry and it isn’t a thin-walled sphere. Some corneas are easier to deform than others, potentially leading to an underestimation of intraocular pressure. This ocular hypertensive study shows that the risk of developing primary open-angle glaucoma is inversely correlated with central corneal thickness. In addition to the measurement artefact associated with the fact that the eye is not a dry, thin-walled sphere, the researchers could not rule out additional biomechanical reasons why subjects with a thinner cornea are at greater risk of glaucoma. They discovered that participants with a corneal thickness of 555 µm or less have a three times greater risk of developing primary open-angle glaucoma than participants, who have a corneal thickness of more than 588µm.

Also, you must remember that some patients will have a thinner and more easily deformable cornea due to pathologies such as keratoconus (assuming no cross-linking treatment) or corneal refractive surgery such as LASIK. Your threshold for concern about measured IOP should be adjusted accordingly. For example, a reading of 21mmHg may indicate an IOP of 22.5 or even 24mmHg, depending on how thin the cornea is. This article helps describe how corneal thickness can affect the reliability of IOP measurement.

Keeping this in mind, make sure you ask about refractive surgery during history taking and consider measuring corneal thickness. It will help to identify patients who you should monitor more closely. Anterior segment OCT is a fast way to measure the central corneal thickness (CCT) but air-puff tonometers – such as the TRK-2P – also provide a quick measurement of CCT and automatically adjust intraocular pressure values for corneal thickness.

According to a study by Michael W. Marcus and colleagues, medium and high myopes are up to 60% more likely to develop primary open-angle glaucoma than an emmetrope. That is one of the many reasons why clinicians should offer myopia management to slow the development of myopia in children. The greater risk of glaucoma is probably multifactorial; axial elongation of the globe due to myopia certainly plays a role. The retina and optic nerve are under more biomechanical stress. And the characteristics of the myopic eye – a large optic disc, peripapillary atrophy and a thin retina – make it harder to differentiate early glaucoma from a healthy eye. The retinal nerve fibre layer thickness of a high myope tends to lie at the thin end of the reference population. Therefore, myopes are more likely to show ‘red disease’ on OCT (thickness measurements that dip into the concerning red area of the reference database but are nevertheless normal for that individual).

So, is OCT still useful in myopes? Absolutely! But progression data may be better than OCT data from a single visit. Therefore, you may want to perform a ‘wellness’ scan on your myopic patients under 40 years, as it is useful to have previous scans to compare with. Add ocular parameters into the patient profile in your OCT software, such as axial length, or, at least, add the refractive error details so that thickness measurements are corrected for the effect of ocular magnification.

At the other end of the scale, don’t forget that hyperopes tend to have a shorter than average axial length and a shallow anterior chamber. That puts them at a greater risk of suffering from primary closed-angle glaucoma. A quick anterior segment OCT scan or van Herrick will provide information about whether the anterior chamber angle is open/closed or narrow and therefore at risk of closure.

Be aware of the risk factors for primary open-angle glaucoma. A careful history can identify low or high systemic blood pressure – both of which increase the risk of primary open-angle glaucoma based on the vascular theory of glaucomatous damage. It is worth remembering that there is a balance between the measured intraocular pressure and the pressure that the ocular structures can withstand. One person’s eye may be able to withstand quite a high pressure without any damage occurring, whereas another person may suffer from glaucomatous damage at a pressure well within the ‘normal’ range. Make sure you are familiar with the definitions of normal-tension and high-tension glaucoma, along with ocular hypertension.

A measurement of intraocular pressure enables you to identify patients with raised values, who are at greater risk of developing glaucoma. Additional measurements such as refractive error and corneal thickness can help you further refine the level of glaucoma risk for an individual. The higher the refractive error, the greater the risk, with myopes at greater risk of developing primary open-angle glaucoma and hyperopes at greater risk of developing primary closed-angle glaucoma.

Use an air puff tonometer that incorporates corneal thickness measurement such as the TRK-2P or capture an anterior segment scan on your OCT to assess the corneal thickness and identify patients with thinner corneas who are at greater risk of primary closed-angle glaucoma. Since primary open-angle glaucoma is generally asymptomatic, early diagnosis of glaucoma is critical for long term visual prognosis. Anything we can do to detect glaucoma earlier should be done, including performing a more detailed and frequent examination of those with a higher risk of glaucoma.

The key retinal layers of interest concerning glaucoma are the retinal nerve fibre layer (RNFL) and the ganglion cell layer (GCL). The OCT analysis uses the following definitions for each layer:

• Retinal nerve fibre layer: Tissue thickness between the inner limiting membrane and the boundary between the RNFL and the GCL.
• GCL+ is between the RNFL/GCL boundary and the inner plexiform layer/inner nuclear layer boundary.
• GCL++ is effectively the RNFL and GCL+ layers added together (inner limiting membrane to inner plexiform layer/inner nuclear layer boundary).

The scan speed of modern OCT devices is around 50,000 A-scans per second, with a reachable 100,000 A-scans per second for swept-source OCT devices. These speeds allow dense data sets to be collected and greatly reduce the chance of missing information that falls between scan lines. The Topcon 3D OCT-1 Maestro2 offers a 12x9mm scan pattern covering the disc and macular regions in a single scan. This consists of 128 B scans, each containing 512 A-scans. That’s a very dense dataset. While it is possible to scan the macula and the disc separately, a wide field scan has significant benefits, both in terms of capture time and when it comes to reviewing the data. Not only is it quicker to review a single report but it makes more sense clinically since both macular and disc areas are affected by glaucoma and damage tends to be continuous between the two regions.

OCT thickness measurements are very repeatable but always dependent on the quality of the scan. Before you review the OCT results, check the quality score. There are various reasons for a low scan quality: Cataract reducing the amount of light reaching the retina. Dry eye distorting the light entering the eye. Poor patient instruction. It is important to ask all patients to take a few good blinks immediately prior to the OCT capture. Clear instruction is even more important with patients with a poor-quality tear film, and those with very dry eyes may benefit from the instillation of a drop of artificial tears immediately prior to capture to create a smooth refracting surface for the OCT beam entering and leaving the eye.

In addition to scan quality, you should also check for artefacts associated with eye movements or blinks. These artefacts show up in the scan preview as black lines or jagged blood vessels. When they are present, it is worth repeating the scan to see if the quality can be improved by asking the patient to keep their eyes wide open during capture and reminding them throughout capture to keep looking at the green target.

Click here for a capture-optimisation training or visit Topcon Healthcare University online for a range of OCT-related courses.

All commercially available OCTs include a database based on a large reference population. The option allows RNFL, GCL+ and GCL++ thickness measurements for an individual patient, which you can compare with the measurements expected for the patient´s age group, defined by the reference database population. For any layer and retinal location, the reference population will demonstrate a range of thickness measurements. They form a bell-shaped normal distribution. The majority of patients have thickness measurements that correspond to the central body of the bell curve (middle and either side), while some have thicker or thinner values that fall into the curve´s two tails. Thickness values falling within the main central body of the normal distribution curve are colour-coded as green on OCT reports. Around 4% of the reference population will show thinner measurements in the tail of the distribution and well outside the main central body of the distribution. Thickness values in this region are colour-coded yellow or orange. A further 1% of the population will show very thin values at the tail end that are colour-coded red.

OCT reports present the measured retinal nerve fibre layer thickness around the optic disc, colour-coded to compare with the reference data. The thickness data is averaged and displayed over four quadrants and also 12 clock segments. If one or more segments have yellow or red thickness measurements, it indicates that the retinal nerve fibre layer thickness corresponds to the thin tail of the reference database distribution. Consequently, there is a higher suspicion of glaucoma. Since glaucoma rarely develops symmetrically, yellow and red segments for one eye but not the other eye increase the level of suspicion further. However, interpret the colour-coded comparison with the reference data with caution. The presence of yellow or red segments does not definitively indicate glaucoma. They merely increase the chance of glaucoma being present. Some “normal” patients without glaucoma have a naturally thin retinal nerve fibre layer all their lives, corresponding to the yellow or red tails. It may be because they have a longer eye and hence a thinner retina.

There are also variations on normal anatomy, including tilted discs, large optic discs, eyes with a large angle between the major retinal arcades. All these normal variations can lead to yellow or red segments on the OCT report, without glaucoma being present. Conversely, some patients with glaucoma will have retinal nerve fibre layer thickness measurements well within the green range of the reference database and yet be diagnosed with glaucoma. That can occur if the patient started life with a particularly thick retinal nerve fibre layer. The 3D OCT-1 Maestro2 has the advantage of simultaneously capturing a colour fundus image, which is pinpoint registered to the OCT scan. The fundus image provides information about wedge defects or confounding retinal disease along with the disc appearance, including variations on normal anatomy.

OCT is already providing valuable insight into the detection and management of glaucoma and retinal disease. Capturing a wide field OCT scan covering a 12x9mm area provides an analysis of the retinal layers affected by glaucoma over both the disc and macular regions. The layers of interest are the retinal nerve fibre layer (around the optic disc and generally across the posterior pole) and the ganglion cell layer (around the macula). You can compare thickness data for both layers with a reference population but be careful to ensure the quality of scans captured since it impacts the accuracy of the thickness measurements.

Suspicious OCT findings indicated by yellow or red colour-coding of thickness measurements – compared to the reference database – do not necessarily indicate glaucoma. It is important to rule out other possible causes, such as tilted discs or high myopia. You should always interpret OCT findings in the context of the complete clinical picture, including the disc and fundus appearance, history, risk factors, and other clinical findings. It is rare to find a suspicious OCT report without there also being risk factors for glaucoma or some other clinical findings that raise suspicion of glaucoma.

With the rapid spread of OCT use in primary eye care, it won’t be long before we can reap additional benefits based on repeat scans on the same patient. Historical data allows for progression analysis, providing an extra tool for the detection of glaucoma but also other neurodegenerative conditions that can cause RNFL loss such as Alzheimer’s disease and multiple sclerosis. In the future, OCT Angiography (OCTA) is likely to become more widely used in glaucoma detection and management. It can support clinical decision making in tricky cases by imaging the Radial Peripapillary Capillaries (RPC) to look for vascular defects that may pre-empt visible retinal nerve fibre layer damage. Read more.

When screening visual fields for signs of glaucoma, you must choose the right test. As optometrists, we are interested in monitoring the progression of visual field loss, and therefore complete threshold testing is valuable. The metrics allow progression analysis to support treatment decisions. But complete threshold tests are time-consuming and stressful for patients, and therefore not practical for glaucoma screening. When it comes to visual field testing, the reliability of the data is dependent on a fine balance between the thoroughness of the test and the duration of the test. Testing more points will increase sensitivity to early glaucomatous field loss, but longer-lasting tests result in fatigue, variations in fixation and less reliable patient responses (ultimately increasing false positives).

An efficient visual field screening test is required to avoid a significant impact on clinical workflow. Conventional screening tests present a suprathreshold target at each particular visual field location. This stimulus is slightly brighter than the one a patient of that age would expect to see based on reference data.

Modern screening tests such as the Henson Smart Supra take screening to a new level. Test points are located in the visual field areas most susceptible to early glaucomatous damage rather than a regular grid pattern. Also, the test offers the flexibility to retest points, investigate additional ones in the same area or extend with an additional test pattern. You can thus avoid performing a whole additional test of, for example, the central 10 degrees. It takes less than 1 minute to test one eye of a healthy patient while offering high sensitivity and specificity – the Goldilocks test.

Visual field tests generally have a working distance of 25-30cm. Visual field thresholds are affected by the presence of an uncorrected refractive error, so it is important to correct it. Uncorrected refractive error causes generalised depression of the visual field, which shows up on the mean deviation map and can mask early visual field loss.

The refractive correction worn for near vision is the most suitable, but patients who wear bifocals or multifocals will need to view the test through an auxiliary lens to avoid suppressing the upper portion of the visual field. Make sure the auxiliary lens is well-positioned to avoid generating rim artefacts. If delegating visual field testing to your support staff, choose a visual field screener that auto-calculates the correction needed based on the patient’s spectacle prescription. That will help minimise measurement errors due to refractive blur.

Although test algorithms like Smart Supra make wide-spread visual field testing a reality, visual fields are usually performed when there is a reason to be suspicious. For example, in case of a suspicious optic disc appearance, thinning of the RNFL or GCL on the OCT report, or the presence of risk factors such as raised IOP, thin cornea, family history of glaucoma etc.

Some OCTs can generate a Hood report from a 3D wide scan. The Hood report is designed by U.S. professor Don Hood from Columbia University to support early glaucoma detection. It incorporates OCT probability maps and highlights areas of the retinal nerve fibre layer or ganglion cell layer with a high probability of being abnormal. Helpfully, the 24° and 10° visual field test locations are superimposed over the probability maps. The location of suspicious areas on the probability maps can help you decide which visual field test to perform. If the suspicious areas coincide with some of the 24-2 test points, then a visual field test over 24° is a natural choice. If, however, the suspicious areas on the OCT probability map are limited to the central area, a 10° visual field test is recommended, because a test over 24° is not sensitive to early central visual field loss.

Remember that structural (OCT) and functional (visual field) losses do not necessarily occur in parallel. OCT is very sensitive, picking up thinning of retinal layers of just a few microns. Perimetry, in comparison, is a very subjective and crude measure of target detection over a grid pattern that may or may not coincide with the area of damage. In addition, an estimation is that between 40-50% of retinal nerve fibres need to be lost before a visual field defect becomes apparent. Hence, you would expect early structural glaucomatous damage to show on an OCT before a functional loss can be detected using visual fields.

If you are going to assess visual fields, do it right!  Choose a test optimised for glaucoma screening that will not significantly extend your examination time. Make sure that the patient is fresh and not tired when starting a test, and bring them back another day if necessary. Keep them engaged throughout the test. Choose a smart screening test that keeps the test time short but has good sensitivity and specificity, such as the extendable Smart Supra test. Implementing short test times makes it more realistic to offer visual field testing to all patients with a higher risk of glaucoma.  Remember that functional damage detectable by subjective visual field tests tends to occur after the structural glaucoma damage that is revealed through an OCT scanning. The Hood report provides a method to link structural and functional damage, and it supports the choice of the visual field test area based on the location of areas of suspicion on OCT.

Viewing clinical data on one screen enhances clinical decision making and improves efficiency. Software such as Harmony RS offers a hanging protocol function, allowing you to decide how you want clinical information presented to you. Alternatively, you can select one of the default hanging protocols, such as colour fundus image/OCT/visual field report, which will be presented side-by-side on the screen for any visit you select from the timeline. The platform is vendor-neutral, so you can freely choose devices to connect. The morph function provides an easy way to compare visual field or OCT reports between visits.

Do not be tempted to refer based on one apparent abnormal finding such as raised pressure or an OCT scan with circumpapillary RNFL sectors that are red (thin) compared to the reference database. In such cases, additional tests or a second opinion from a colleague or an ophthalmologist can help you decide whether to refer or not. Look out for normal anatomical variations, particularly in higher myopes. Be suspicious if similar disc or OCT findings occur for both eyes – glaucoma is rarely – if ever – symmetrical in its development. Large optic discs, tilted disc, a large angle between the disc and fovea, or a large angle between the retinal arcades are all anatomical reasons why the OCT or disc metrics for a patient may not be reflected within the reference population, resulting in a false positive.

Familiarise yourself with your national or local referral criteria for suspected glaucoma. Which patients do the ophthalmologists want to see, and what information do they want to receive from you? OCT and visual field reports along with pressure measurement can be shared through the referral mechanism in software such as Harmony RS. You may have the option of seeking a second opinion through Harmony RS, before deciding whether or not to refer. Sharing information to avoid the patient having to repeat tests is generally appreciated by all involved.

Going forward, communication between primary and secondary eye care needs to increase, if we want to provide an efficient system for managing the growing and ageing population. We already see technologies designed to support shared care and delegation to primary care, the prime example being Artificial Intelligence (AI) technology. For example, you can plug Thirona AI technology into Harmony RS to provide automated analysis of the fundus image for age-related macular degeneration, diabetic retinopathy or glaucoma. We are likely to see more AI technology in the future.