Lecture: How to Diagnose and Monitor Glaucoma

This lecture covers the basic and advanced diagnostic framework for detection and monitoring glaucoma, illustrating with examples how to evaluate the retinal nerve fiber layer and neuroretinal rim in clinical practice. Interpretation of visual field and optical coherence tomography findings is also discussed.

Lecturer: Dr. Chris Leung, Head of Dept., Clinical Professor, Dept. of Ophthalmology, The University of Hong Kong

Transcript

[Chris] Hello, everyone. In this webinar we’ll talk about how to diagnose and monitor glaucoma. This webinar has 45 minutes of presentation, followed by 15 minutes of question and answer. If you have any questions you want to ask during the presentation, please feel free to type a question in the Q&A box and I can answer that at the end of the presentation.

In this talk we have two sections. The first is about how to diagnose glaucoma and the second would be on how to monitor glaucoma. And then each section, before each section we have some poll questions.
Let’s go to the poll questions, which is about which of the following parameters are important to make a diagnosis of glaucoma? I will list five different parameters and I would like to vote using your Zoom poll tool and so we can begin the polling. Which of the following parameters are important to make a diagnosis of glaucoma? The first is visual field sensitivity. Is it important to have visual field sensitivity to make a diagnosis of glaucoma? Yes or no? All right, so we have overwhelmingly 92% of individuals believe visual field sensitivity is important. So let’s go to the next question, I will come back to that in the presentation.

Is intraocular pressure important to make a diagnosis of glaucoma? Yes or no? All right, so we have 75% of our participants believe intraocular pressure is important to make a diagnosis of glaucoma.

The third parameter here is cup-to-disc ratio. Is it important? Yes or no? Right, again, 90% of our participants believe the cup-to-disc ratio is important to make a diagnosis of glaucoma. Okay.

The fourth parameter, retinal nerve fiber layer thickness. Is it important? Do you use retinal nerve fiber layer thickness measurements to make a diagnosis of glaucoma? Yes or no? All right. 93% think it’s important.
The last one is neuroretinal rim width? Is it important? Do you use neuroretinal rim width to make a diagnosis of glaucoma? A or B, yes or no? Choice one is yes, choice two is no. Right. 85%. Good.

Now, these are the take home messages on how to diagnose glaucoma. We need to know glaucoma is an optic neuropathy and the diagnosis, clinical diagnosis actually is predicted on two key parameters which is the retinal nerve fiber layer and neuroretinal rim. And then we also need to be aware of the fact that wide-field assessment is important. Wide field means we need to include both the region around the optic disc or parapapillary region and the macula. We need to look at the macula and the parapapillary region to assess the retinal nerve fiber layer and ganglion cell-inner plexiform layer for diagnosis or evaluation of glaucoma. We will come back to these points during the lecture.

During the 10th World Glaucoma Association Consensus Meeting, we had a discussion among a group of glaucoma specialists, clinicians, and scientists from different parts of the world to talk about how to make a diagnosis of glaucoma. It was five years ago, and this is one key consensus statement we made during that meeting, which said, clinical diagnosis of glaucoma is predicted on the detection of a thin retinal nerve fiber layer and narrow neuroretinal rim.

Notice that this statement contains no cup-to-disc ratio, contains no intraocular pressure, contains no visual field measurements. It’s surprising, perhaps, to many of you because many of you believe intraocular pressure is important, visual field is important. But diagnosis of glaucoma, actually, we didn’t list any of the IOP or visual field or even cup-to-disc ratio parameters in this consensus. We basically focused on two key parameters which is RNFL and neuroretinal rim.

We also, here in this consensus, mentioned that diagnosis of glaucoma does not always require the detection of visual field defects with perimetry. The very reason here is that glaucoma is a disease of the optic nerve, it’s about chronic progressive degeneration of the optic nerve. To detect the pathology we need to look at the optic nerve. And the way that we look at the optic nerve is typically with structural assessment of the optic nerve head and the retinal nerve fiber layer.

Retinal nerve fiber layer assessment is important because the retinal nerve fiber layer is composed of the exons, retinal ganglion cells. Glaucoma, as I said, is a disease about progression degeneration of the optic nerve. What is the optic nerve? The optic nerve comprises the exons of retinal ganglion cells and glaucoma is about degeneration of these exons, glaucoma is about the degeneration of retinal ganglion cells. So that’s why to study the pathology to make a diagnosis of glaucoma, we really need to evaluate the retinal nerve fiber layer, the exons of retinal ganglion cells.

Conventionally, we apply two techniques to examine the retinal nerve fiber layer. We have red-free photography and also OCT. OCT is a more commonly adopted approach for evaluation of retinal nerve fiber layer nowadays. But in the old days before we had OCT, we relied more on red-free photography.

Photography is easy, is straightforward, but it has its own limitations. It’s been used a long time ago and this is a paper by Harry Quigley in 1980, when he used red-free photography to detect retinal nerve fiber defects in glaucoma. We can see the striations, you can see the trajectories of these axonal fiber bundles. You see loss of these axonal fiber bundles as a down looking area in this red-free photograph. The challenge here is, in myopic eyes, in eyes with hypopic mendafundus, just as what we see on the right hand side, actually very difficult to appreciate the retinal nerve fiber layer refractivities. It is so hard to identify retinal nerve fiber defects in eyes with hypopic mendafundus in eyes with myopia. It’s actually difficult in that regard.

Another challenge is that when we have diffuse loss of retinal nerve fiber layer. When the glaucoma is at the advanced stages, it would be difficult for us to pinpoint where exactly the retinal nerve fiber layer loss is using red-free photography. Detecting early retinal nerve fiber layer defects would be okay with red-free photography. But when the loss is diffuse, when glaucoma is advanced, it would be difficult to rely on this tool to delineate the presence or absence of retinal nerve fiber layer thinning. That’s why we have now OCT, which provides a cross sectional analysis of the retina. Retinal nerve fiber layer typically is imaged over the parapapillary region or the region around the optic nerve head. It is because all retinal ganglion cells sitting on the retina, they all sense exons converge at the optic nerve head. And that’s why around the optic nerve head, we have relatively thick retinal nerve fiber layer so it is easier to study, measure retinal nerve fiber layer over the parapapillary region.

With OCT this is what we typically obtain. I encourage you to look at the retinal nerve fiber layer thickness map. In the old days, not long ago, actually, 10-20 years ago, when we used OCT to image a retinal nerve fiber layer, we typically used a circle scan with a diameter of about 3.4 millimeter, just as one I show you in arrow. By convention, many OCT devices they display this circumpapillary retinal nerve fiber thickness profile and this is what we typically rely on. We look at where the quadrants and clock hours being labeled as red. Red here indicates the retinal nerve fiber thickness is below the normal reference range, below the 99th percentile of the normal value, so it’s indicated in red.

If we’re limiting ourselves to this circle, we would be missing a lot of information. Instead of looking at the panel below, instead of looking at the circumpapillary retinal nerve fiber layer profile, I would encourage you all to study the retinal nerve fiber thickness map, the one in the top right and top left corner. To understand the retinal nerve fiber thickness map, we need to understand the fact that the retinal nerve fiber layer is thickest over the supero temporal and inferotemporal sector of the optic disc. And in this map, or in this display, thick retinal nerve fiber layer is indicated in red. Thin retinal nerve fiber layer is indicated in blue. Compared to the left eye, you can clearly now appreciate there is less red, or in general, blue, over the inferotemporal sector of the optic disc. This indicates loss of the retinal nerve fiber layer.

Retinal nerve fiber thickness map provides you a qualitative assessment but for quantitative evaluation you can always refer to the deviation map, which shows whether the retinal nerve fiber layer thickness on a particular region in this map is below the 99th percentile, which would be indicted in red in the deviation map, or below the 95th percentile which would be indicated in yellow. You can see the correspondence between the thickness map and the deviation map, in this case. Giving you high confidence that there is genuine loss of retinal nerve fiber layer over the inferotemporal region.

When we look at the left eye, you see the thickness of the retinal nerve fiber layer in the thickness map. Basically it follow what we expect to see being thick over the superotemporal and inferotemporal sector of the optic disc. And then in a deviation map, you don’t see any abnormal pixels highlighted in red. You may then think or you may say the left eye, you don’t see any abnormal retinal nerve fiber layer thickness. Before we interpret the retinal nerve fiber thickness map and deviation map, we often need to look at the signal strength. By convention, we need to have our signal strength of at least six or seven because if a OCT printout has a low signal strength, the retinal nerve fiber layer would be falsely reduced. Giving you a false positive assessment of retinal nerve fiber layer abnormality. We need to have a careful check on whether the signal strength is good enough for you to interpret the thickness measurement. Low signal strength typically is related to mediopexy like cataract.

We image the retinal nerve fiber layer because, as I mentioned, retinal nerve fiber layer is the exons of retinal ganglion cells. But we also need to study the macula. The reason here is that a lot of time some early defects can be missed over the parapapillary region, and I will show that to you in the next slide. But I would like to demonstrate here the anatomy of the retinal layers. The top layer is the retinal nerve fiber layer and then we have the ganglion cell layer which represents the cell body, retinal ganglion cells. And below the retinal ganglion cell layer we have the inner plexiform layer, which is composed of dendrites of retina ganglion cells. Retina ganglion cell is a central nervous system neurons so it has cell body has exon and dendrites. OCT provides us an opportunity to measure different layers on the retina.

We notice that this section corresponds to the macula because we see many layers of the ganglion cell layer. We don’t typically image the retinal nerve fiber layer alone when we evaluate the macula, because the RNFL is relatively thin over the macula. It is relatively thick over the optic nerve head region. But it is relatively thin over the macula. By contrast, the ganglion cell layer is relatively thick over the macula, because we have five to six layers of these cell bodies locating at the macula. By contrast, the parapapillary region, the GCL, is relatively thin. That’s why of the macula, we would image the GCL and IPL. Sometimes we also image the RNFL together as well.

In the size instrument, the Cirrus OCT, it measures the GCIPL without measuring the RNFL. That’s okay. There’s some instruments like the Optovue outer view, they also measured RNFL together with the GCL and IPL and they label at ganglion cell complex. The interpretation, basically, is more or less the same as the RNFL thing especially. You look at the thickness map. Red indicates thick, blue indicates thin. And then you look at the deviation map, do you see whether there is a correspondence. You always look for symmetry between eyes and symmetry between the upper and lower retina. What you can appreciate here in these eyes, you have loss of this on the shape thickness, over here on the right eye and here on the left eye. And then corresponding changes can be found over the deviation map. Again, look at the signal strength, it’s nine, which is good, about seven. You have loss of the GCIPL. Now, this is a case where we have loss of the retinal nerve fiber layer with right eye and loss of the GCIPL over both eyes.
How about the neuroretinal rim assessment? I mentioned, which in the World Glaucoma Association Consensus Meeting, the diagnosis of glaucoma requires the assessment of the retinal nerve fiber layer and also the neuroretinal rim. The reason is that loss of neuroretinal rim is relatively specific for glaucoma. But there are also kinds, other types of non-glaucomatous optic neuropathy. All optic neuropathies would have loss of the retinal nerve fiber layer. All optic neuropathies, glaucomatous or non-glaucomatous, would have thin retinal nerve fiber layer because all optic neuropathy basically means loss of retinal ganglion cells, degeneration of retinal ganglion cells. And that’s why we have loss of RNFL in all kinds of optic neuropathies. But loss of the neuroretinal rim is relatively specific for glaucoma. That’s why the evaluation of the neuroretinal rim is important in assessment in the diagnosis of glaucoma.

We have two approaches to evaluate the neuroretinal rim. We have optic disc assessment and also we can use OCT to quantify the width of the neuroretinal rim. And this is an OCT, a child OCT which provides this measurement. Because this BMO, Bruch’s membrane opening, which is over here, BMO. And it extends from the BMO to the nearest distance to the internal limiting membrane. This is the definition of minimum rim width. By scanning here, 24 radar scans, around optic nerve head, the OCT automatically identified the BMO and then measures the neuroretinal rim width.

Sometimes the BMO location will be wrong and then you have to check carefully each scan the OCT is able to identify the BMO correctly. Because if the BMO location’s wrong, you would get a wrong or incorrect BMO minimum rim width measurements. And then based on this, again, color coded plot in red indicates the rim width is below the 99th percentile. And then if it’s in yellow it’s below the 95th percentile. What we can appreciate here in this picture, is that you have loss of the inferotemporal neuroretinal rim, both line loss over the inferonasal rim. We have OCT providing us the opportunity to quantify the measurements of neuroretinal rim.

Thinking that we can rely on OCT without looking at the optic disc using our slit lamp biomicroscopy, the answer is no, we still need to examine the optic disc with color examination with our slit lamp biomicroscopy. Because assessment of the color is actually very important to differentiate glaucomatous from non-glaucomatous optic neuropathies. This is also one of our consensus meetings during the 10th World Glaucoma Association Consensus Meeting, it’s one of the key consensus points. A pale rim suggests non-glaucomatous optic neuropathy. We have to look for the color. Pink rim indicates normal glaucomatous optic neuropathy if the rim is also thin. But a pale neuroretinal rim would suggest the presence of non-glaucomatous optic neuropathies.

I had the opportunity to attend this program organized by Orbis in China, Setran. We had a team there doing operations and doing some assessment for glaucoma patients. And here is an OCT given to me which they had been treating this patient as glaucoma for many years. And this is the OCT printouts. Over the right eye you can see the retinal nerve fiber layer thickness is relatively normal. Red over the superotemporal, inferotemporal sector and no abnormalities highlighted in red in the deviation plot.

Over the left hand side, obviously, we see loss of the retinal nerve fiber layer, more on the superior side than in the inferior side. But when you examine the color of the rim carefully, what you notice here is that the rim, actually, is pale, particularly when you look at the right eye together. We can appreciate the difference in color. The rim is pink over the right and pale over the left. In fact, this is not a case of glaucoma, this is a case of non-glaucomatous optic neuropathy, but being treated as glaucoma because of abnormal retinal nerve fiber layer thickness in the OCT. We have to be careful when we see abnormalities in the retinal nerve fiber layer in OCT, we need to look at the neuroretinal rim to check whether the green dimension is normal, if it is abnormal. Then if it is thin then it is likely to be glaucoma. If the rim looks pale, then it is likely to be non-glaucomatous optic neuropathy. There are many different courses of non-glaucomatous neuropathies. Hereditary, inflammatory, compressive, then you have to work out to do additional investigation to find out the pathology of non-glaucomatous, or etiology of non-glaucomatous optic neuropathy.

Here is a study done many years ago demonstrating that rim color was found to be 94% specific for non-glaucomatous optic neuropathy. That’s why examination of the color of the rim is critical.

We often see the scenario in patients with glaucoma and sometimes we then say hey, the disc looked pale. I would like to highlight to you the point we need to study here is that glaucomatous optic neuropathy, although the disc here may appear pale, but the residual neuroretinal rim remains pink. When we comment on the color of the optic disc, pay attention not the cup, but to the rim. Because it is the rim color that matters. Cup always look pale, glaucomatous optic disc look pale because most rim, especially in advanced glaucoma case, would be gone. Leaving you only the disc when you examine the optic disk. That’s why it looks pale. When you study the residual rim, it is pink.

Here is another case to demonstrate the importance of examination of the retinal nerve fiber layer and the rim as well to evaluate glaucoma. Over the right hand side, you can see even with the OCT, loss of the rim, you have enlarged optic cup. But the left eye you may then say it is roughly normal, the rim looks pink for both eyes and the cup-to-disc ratio is not too bad. There is no apparent loss of the rim tissue evidenced by clinical examination of the optic disc photograph.

When we use OCT to quantify the rim width, we’re using a more objective approach to discern whether the rim measurement is normal or abnormal. Again, the right eye, you have loss of the inferotemporal rim. And then on the left side, OCT is not able to show any abnormalities. Looking back again, on the photograph, we have loss of the rim, more on the inferotemporal side than inferonasal side. And then the left eye, the neuroretinal rim is normal.

When we look at the OCT, the same OCT picture you all just saw, and then the right hand side loss of the retinal nerve fiber layer over the inferotemporal side. And then on the left side, which I mentioned early, there is low loss of retinal nerve fiber layer. But it’s only when we look at the GCIPL for the macula, you can see a wedge-shaped defect of the GCIPL over the left eye as well. That’s why, in the beginning, I emphasized the importance of wide-field imaging. I think it’s a key take home message when we evaluate glaucoma, we often we’re just focused on the optic disc and in the old days, before we had OCT, we go by the cup-to-disc ratio. But we know this is not a very sensitive way to detect glaucomatous damage. We need to examine wide-field, we need to examine both the parapapillary region in the macula, like in this case. If we don’t examine the macula, we would miss a diagnosis of glaucoma.

This is particularly important in early cases of glaucoma. The reason why we don’t see any rim loss, the reason why we don’t see any retinal nerve fiber layer loss close to the optic disc, because the damage is very focal. Remember, retinal nerve fiber defects, retinal nerve fiber abnormalities, they are often wedge shape. Meaning that the abnormality expands from the optic disc margin to where is the macula. In other words, by the chance you detect abnormality of the macula because the fact that the damage expands away from the optic disc so you see bigger damage of the macula. This is the wedge you see over the macula. When you trace this wedge towards the optic disc, the loss can be so small that it is difficult to detect when you study the rim over the optic disc and also even for retinal nerve fiber layer, the loss is so little that it’s difficult to detect even with OCT. Remember, wide-field assessment is critical for evaluation of glaucoma.

Again, I would also like to say we need to be careful in the interpretation of those we’ve printed out because sometimes we do see false positive, particularly in eyes with myopia. Here is a case demonstrating that in the thickness map you see red, or thick retinal nerve fiber layer over this inferotemporal side and superotemporal side. However, when you look at the deviation map you see hey, this is a normal retinal nerve fiber layer thickness highlighted in those printouts. In fact, it is false positive. The reason why it is false positive is because the arrangement or distribution of the retinal nerve fiber layer in eyes with myopia is slightly different from eyes without myopia.

Typically, the superotemporal and inferotemporal retinal nerve fiber bundles, they tend to move closer to the macula, they converge to the macula. Which, indirectly, leave the superior and inferior retina around the optic nerve head relatively normal compared with the normative database distribution. The OCT normative data was largely collective from eyes without myopia. If we are using myopia, non-myopic eyes as reference to gauge abnormalities in myopic eyes, this is not going to be good. This is always to be non-specific detection of retinal nerve fiber layer abnormalities. And in fact, this is what we show in a previous study demonstrating that alarmingly the specificity can be actually very low. Lots of false positive can be detected in eyes with myopia.
And then when we look at the finding, when we apply a normative database collected from eyes with myopia, this false positive would be gone. We need to be careful in the interpretation of the retinal nerve fiber layer thickness, particularly in eyes with high myopia.
One useful trick you all can apply, is that you always look for correspondence between the thickness map and the deviation map. Here, you don’t see correspondence, you have relatively normal looking retinal nerve fiber layer appearance with the thickness map, but abnormal annotation in the deviation map. It is not in agreement with what we expect. Always look in pair in this OCT analysis.

To conclude the first part, we need to be careful when we evaluate glaucoma. We don’t need IOP to diagnose glaucoma, we don’t need visual field to diagnose glaucoma. Cup-to-disc ratio is not very sensitive because it varies with optic disc morphology and optic disc size. So that’s why the current standard for diagnosis of glaucoma is relying on the retinal nerve fiber layer assessment and neuroretinal rim assessment. And in that regard, we need to understand rim assessment is important because we need to differentiate glaucomatous for non-glaucomatous optic neuropathy.

For non-glaucomatous optic neuropathy, we’ll see pale neuroretinal rim. And for assessment of the retinal nerve fiber layer, we need to see wide-field. Of the macula, more often, we rely on the GCIPL because the ganglion cell in the plexiform layer is thicker at the macula. That gives us a higher reliability in the measurements of layer thickness. So you can study GCIPL over the macula or you can study both RNFL combined with GCIPL of the macula. The lesson here is we need wide-field. Because retinal nerve fiber defects are often wedge-shaped in glaucoma. You get a better chance to detecting abnormality of the macula.

The second poll question, let’s go to the second part, how to detect more glaucoma progression? We have this eye with visual field collected at three different time points spanning 10 years. Do you think this eye has glaucoma progression? Let’s have a go.
Yes, or no? Is there evidence of glaucoma progression? All right, over half, 77% of our participants we’re 444 this evening, so 77% of them will say there is evidence of progression.
These are the key points. For evaluation of glaucoma, we need structural assessment. And typically we rely on the OCT imaging. And again, we look at RNFL and GCIPL and again, we need wide-field, including both the parapapillary region and the macula. It is the structural assessment of the optic nerve.

For functional assessment we rely on standard automated perimetry. A lot of us are familiar with using the 24-2 or that’s for examination of visual field loss. But in advanced glaucoma, when the peripheral field is almost gone we then often will refer to 10-2, giving us more test points of the centralized rim.

We typically test our patients with OCT and visual field every four months. There’s no standard, there’s no consensus on how often we should prescribe OCT and visual field tests. For eyes at risk of change, we need to test more often, even two to three months. But we can test less often in eyes with low risk of progression. And to evaluate the risk there are many factors we need to consider. Intraocular pressures, obviously, is important in these regards. Intraocular pressure is not relevant to the diagnosis of glaucoma but it is important to the assessment of risk of progression. Higher the IOP, higher the risk of progression. And then you also need to evaluate the stage of disease and the prior history. How fast is the disc of views progressing? All this you have to take into consideration when you determine how often this patient needs to be tested.

I think this is a very important point, we need to use available software support to determine the progression. What do I mean available software support? We need statistical analysis, we just can’t trust our eyeballs to determine where the change has occurred or not in visual field and optic disc. We need to rely on statistical analysis and there are two approaches. One is event analysis and another is trend analysis.

What is event analysis? The reason here is visual field tests, in particular visual field tests, and also actually OCT measurements, all these investigations have test-retest reliability. Visual field because it is a subjective test. This type of variability would be typically higher compared with objective tests like OCT. We can’t just say when we see a change this is progression because this change can be variability. For event analysis we define genuine progression where this change is greater than a threshold. When the change is smaller than the threshold we don’t define progression. When the change is greater than the threshold, we label it as genuine change. This is event analysis.

For trend analysis we make use of all the data points we collected throughout a period of time and then we perform a linear regression analysis to see whether there is a significant negative trend in this analysis. And if there is a significant negative trend we call this progression. In visual field printouts, the available software support we currently have is the GPA or Guided Progression Analysis, which provides trend analysis of VFI. VFI is Visual Field Index, very similar to MD min/deviation. VFI also takes consideration of general reduction in sensitivity. And VFI of 100% means a normal view, VFI of 0% means a total dark field. And this is a linear regression analysis between VFI and age or years. In this case, we see the rate of progression is -2.9% and it is statistically significant with P less than .1%. We see a significant negative trend so this is progression.

In the GPA we also have event analysis. It is based on the EMGT criteria which is the Early Manifest Glaucoma Trial, a trial that was done many, many years ago. And based on this trial criteria, we compare the follow up to the baseline assessment point by point comparison. When a location of visual field sensitivity dropped more than the test-retest reliability the first time it would be highlighted in an empty triangle. At the same location, if the same loss is detected for a second time, it is labeled as a half-filled triangle. At the same location, if the same loss is detected three times for three consecutive tests, it would be labeled as a black triangle. When we see a list of three black triangle it would be called or diagnosed or defined as likely progression. When you have three half-filled triangles it would be labeled as a possible progression. The reason we need this strange written criteria is because of the highly subjected nature of visual field testing.

Going back to the poll question, 77% of you would say this is progression because you see this pattern deviation plot. From 2009 to 2019, you see, actually, an increase in abnormalities in these abnormal points in the visual field. But we just don’t yet have enough evidence because eyeballs, as I said, are not trustworthy. We need to look at statistical analysis. When we look at trend analysis the rate is .2% but the slope is not statistically significant. When we look at the event analysis, there’s actually no progression detected. There are only two half-filled triangles, not fulfilling the criteria of visual field loss.

Next time, when patient come back for follow up the patient may well be having a higher VFI, may well be having a bad visual field sensitivity, and resulting in the fact that apparent increase in this abnormal points in the pattern deviation plot may be gone next time when the patient come back or because test-retest reliability.

Likeway, for evaluation of structural change. Again, I don’t know how many of you will see there is a change or there’s a loss of neuroretinal rim, it’s so difficult to see. And this is why OCT Is particularly important in these regards which allow us to perform an event analysis. Again, this is also called Guided Progression Analysis or GPA. And in this eye what we see is a series of retinal nerve fiber thickness maps on the top and then the change map, we call this RNFL thickness change map, which is a statistical comparison, pixel by pixel comparison. When this pixel RNFL thickness is lower than when this pixel retinal nerve fiber layer thickness in a follow up exam is smaller than the baseline, and the difference is greater than the test-retest reliability of that pixel, that pixel will be highlighted in red. You see additional RNFL thinning in this eye detected since 2009 August, illustrating that there is progressive thinning of the retinal nerve fiber layer. The importance of using OCT to evaluate retinal nerve fiber thinning is based on the fact that a lot of these changes can be predictive of future loss of the visual field.

Here, in this example, demonstrating that you have progressive retinal nerve fiber layer thinning detected 33 months before you see visual field progression detected by GPA. We did a study some years back and the risk is very high when you see eyes having progressive retinal nerve fiber layer thinning detected by OCT, these eyes actually have a more than threefold increase in risk of development of visual field progression.

As I mentioned earlier, it is important also to look at the macula. In the macula you reevaluate the GCIPL. In this example, we see progressive GCIPL thinning can be detected 21 months before we see progressive retinal nerve fiber layer thinning detected by GPA. Again, it is the same principle, it is the event analysis comparing pixel by pixel GCIPL thickness. When the change is greater than the variability, it will be highlighted in red. And in this case the change map you see this additional patch of thinning of the GCIPL over this area. One may wonder then can we just rely on the macula because we see macula change can be detected earlier than even parapapillary retinal nerve fiber layer assessment? But sometimes we do see the other way around. In this example, we see progressive retinal nerve fiber layer thinning detected before we see progressive GCIPL thinning. The time gap here is 37 months.

Why will some cases have retinal nerve fiber layer thinning detected first before we see progressive GCIPL thinning? Whereas in some cases we see progressive GCIPL thinning detected first before we see retinal nerve fiber layer thinning? As I mentioned it is important to understand retinal nerve fiber layer defects, they’re often in worse shape. When these abnormalities progress over time, it would be actually easier to detect change of the macula, so it’s not surprising to see change to be detected of the macula before we see change to be detected around the optic nerve head.

But there are cases we see change over the RNFL near the optic nerve head detected before we see change of the macula. It is because the limitation of the current OCT, we scanned the macula using six by six millimeter trip scan. But when the OCT analyze the GCIPL, it only focused on this elliptical annulus. Imagine if you have retinal nerve fiber layer defects developing outside this scanned zone, you would not be able to detect abnormality, you would not be able to detect change. And that’s why wide view assessment is important to maximize the opportunity to detect change.

With the model here, where we evaluate progression using RNFL and GCIPL, we need to be aware of the fact that we need to focus both at the macula and the parapapillary region. And we have a model here indicating that progressive retinal nerve fiber layer and progressive GCIPL thinning, they are mutually predictive. Importantly, they are both predictive of visual field loss. A question that some of you may ask is shall we treat glaucoma patients or shall we increase our IOP lowering medications when we see progressive retinal nerve fiber layer thinning when we see progressive GCIPL thinning? The answer here is that we need to understand why we need to lower intraocular pressure.

This is the EMGT, Early Manifest Glaucoma Trial, and in this study this is the idea how we get 21 mmHg. We typically say when the patient has an intraocular pressure more than 21, he has high intraocular pressure and then we need to provide IOP lowering medication. The rationale here is that in this study, what it shows is when eyes having an intraocular pressure more than 21, these eyes will have a risk of development of visual field progression and the risk is 1.77 fold higher compared with eyes with intraocular pressure below 21. This is the rationale of lowering the intraocular pressure because we want to lower the risk of visual field progression.

Now, with this rationale, imagine if we see progressive retinal nerve fiber thinning, these eyes have a 3.6 fold higher in risk of development of visual field progression. If the rationale of lowering the intraocular pressure is to reduce risk of visual field loss, then we need to pay very careful attention when we see progressive retinal nerve fiber thinning, when we see progressive GCIPL thinning. Because these eyes would be at high risk for development of future loss in the visual field. But thinking that we need to treat everyone with progressive retinal nerve fiber layer thinning, the question is similar to the one I’m asking here. Do you need to treat everyone with an intraocular pressure more than 21? The answer, probably no. You need to really understand the whole risk profile. You need to determine the stage, how bad the glaucoma is, how bad the optic nerve is damaged in the first place. How fast is the optic nerve is progressing over time, you need to also consider how high it’s intraocular pressure, you need to evaluate individual risk effect one by one to get an overall idea how high the risk is to determine how lower the pressure to be targeted.

Again, this is a summary. The take home message is I strongly advise you when you evaluate visual field loss, you really need to look at GPA, Guided Progression Analysis. This is actually one consensus statement we made in another consensus meeting about glaucoma progression from the World Glaucoma Association Consensus Meetings. And then we also need to evaluate both the parapapillary region and the macula for structural evaluation of optic nerve progression. This is what I want to share with you and thank you very much for your participation.
I guess there are some questions in the Q&A. There are a lot of them, I apologize, I don’t think I can answer every one of your questions but let me pick some of these questions for you.
What’s BMO again? BMO is Bruch’s membrane opening. It is a landmark in the retina, a reference point for us to measure minimum rim width.

Another question, is assessment of neuroretinal rim the best screening test for priority? The answer probably I would say is no. What I demonstrated to you earlier is that glaucoma damage, glaucoma optic nerve damage, it expands away from the rim to the parapapillary region to the macula. It’s a triangular wedge shape. For early changes, for mild, local damage, it’s actually useful to examine the macula instead of the rim because you see tiny loss of the rim tissue for early damage. But you may be able to see much bigger abnormality of the GCIPL over the macula. For detection of glaucomatous damage, I would say still you need a wide-field imaging, you need OCT.

And then another question is, is there any image-guided AI-based assessment for RNFL loss? Yes, there is. But it’s not yet commercially available, it’s not routinely used in our clinic because it hasn’t been fully validated. I believe this is something that we’re going to have in the near future.

Another question, is there any road for multifocal ERG in determining glaucoma progression? I don’t think it has an important role at this point because we still, very much, rely on visual field and OCT for evaluation of the functional and structural integrity of the optic nerve to determine progression. I’m not saying that multifocal ERG is not useful, but it’s not something that we typically rely on because the evidence there in the literature is sparse and the information multifocal ERG provides would be relatively limited compared with visual field.
How important is assessment of parapapillary atrophy in the diagnosis of glaucoma? We don’t rely on parapapillary atrophy to make a diagnosis of glaucoma because it is not specific for glaucomatous damage, although we do see a lot of glaucomatous optical changes to be associated with parapapillary atrophy. We also see in myopic eyes, we also see that in aged eyes, so it’s not a very specific sign for glaucoma. We don’t look at PPA to make a diagnosis of glaucoma.
BMO and minimum rim width reliable in high myopia cases? Very good question. I don’t think it is very reliable to use BMO/MRW in high myopic eyes because high myopic optic disc, the configuration of these discs are very different from normal optic disc. BMO is often difficult to be located in high myopic eyes. Making it actually BMO/MRW assessment less reliable compared with normal eyes. I would not recommend using BMO/MRW in highly myopic eyes.
And then, let me see if there’s any question…

Another question that I found many of you asked before you joined the talk is NTG, Normal Tension Glaucoma. As I said, in Asia, in Japan, in Korea, in China, we have a lot of patients with intraocular pressure less than 21. That’s why, again, I emphasize the fact that we don’t make a diagnosis of glaucoma based on intraocular pressure. A lot of patients with glaucoma, their pressure is less than 21. The question here is how do we treat, how do we manage our patients with NTG.

The key here is you need to first make the correct diagnosis, you need to identify retinal nerve fiber loss, thinning of the rim. And then once a diagnosis is made, you look at the visual field. Sometimes the visual field can be normal, some of you may call this preperimetric glaucoma. A lot of time if the IOL risk profile is not high, so for example, the IOP is at the low teens and then the patient is able to be followed up, optic nerve nerve damage is mild, I can’t serve. Meaning that I would follow up the patient, at say three month or two months from now, to get a series of visual field and OCT tests done. The reason here is I want to know how fast the optic nerve is progressing over time. Because at a one time point, you will not be able to get the speed or the rate of change. But with a more frequent follow up over a half year or a year, you can get an idea how fast the visual field is progressing, how fast the retinal nerve fiber layer is thinning. And then if the nerve is thinning, if the field is progressing, obviously you would then prescribe IOP lowering treatments. Typically with medication or sometimes now we use SLT.

But if you see the nerve and the view are relatively stable, I think if the overall risk profile is not very high, you can keep observed, provided that frequent follow up can be provided. Provided that the patient can come back for follow up, otherwise if you don’t follow up the patient, the patient can actually develop higher IOP, patient can develop progressive retinal nerve fiber layer loss if you don’t measure it. It is important for you to follow glaucoma patients on a regular basis.

I’m really sorry, I just don’t have time to answer all questions. It’s 8pm. I hope you all enjoyed the lecture and I hope I can share with you more about how to treat, how to manage glaucoma in the future webinars. Thank you very much, indeed.

Download Slides

PDF

August 27, 2021

Last Updated: October 31, 2022

6 thoughts on “Lecture: How to Diagnose and Monitor Glaucoma”

  1. I learned a lot from this webinar! Lecture is very clear and concise. Easy to follow and understand! Thank you Dr. Chris.

    Reply

Leave a Comment