Dr. Weinreb: I'm gonna begin this evening talking about imaging in clinical practice with an emphasis on glaucoma and how we use it--the OCT in glaucoma
I have a conflict of interest to disclose, which is I'm a consultant for Optovue. I'm also a consultant for other companies, all other companies who work with OCT
So, glaucoma's a continuum in which there's an acceleration of retinal ganglion cell apoptosis. And we go from normal to abnormal, which undetectable disease to asymptomatic disease to functional impairment
At the current time, we stage the disease, early, moderate, advanced. And many of us predict loss of function. We do this by assessing risk
Some of the risk factors that are emerging are intraocular pressure, thin central cornea, the condition of the optic nerve, the visual field, age, presence of exfoliation or an optic disc hemorrhage
As we move forward, we're beginning now to detect progression with our OCT instruments, and we're also beginning to measure the rate of change to estimate the likelihood and expected severity of functional loss as it moves ahead by itself. So, we--we're beginning to have trend based analysis where we could do linear aggression on our results
There are a number of advantages to imaging for glaucoma. One of the things that we can do with the digital information is we can measure features that are otherwise not measure--we're not able to measure
Some of the features include, for example, the depth of the optic cup. We can measure optic cup volume. We can measure the thickness of the retinal nerve fiber layer. More recently, we've been able to look at the pores in the lamida corosa [sp] and measure the shape of the lamida corosa and how it responds to intraocular pressure
One particular aspect that we cannot otherwise measure is the thickness of the retinal ganglion cell layers within the macula. The retinal ganglion cells are lost throughout the retina in glaucoma, but they're also lost in the macula
And the macula is particularly important because a large number of retinal ganglion cells are found in the macula. In a healthy eye, as you know, we have about 700,000 to 1.5 million retinal ganglion cells. The density of retinal ganglion cells is greatest in the macula where they can be as many as six cells deep
About one-half of retinal ganglion cells are located in the central 4.5 millimeters or 16 degrees. So, you can imagine that there's a propensity of retinal ganglion cells within the macula. And if we're monitoring a disease like glaucoma, if we can detect the loss of retinal ganglion cells in the macula, it would be very helpful to us clinically
Unfortunately, the macula represents only about 7 percent of the total retinal area. But, the retinal ganglion cells in the macula are lost in glaucoma just as they're lost everywhere throughout the retina
Originally introduced with the RTView several years ago, the ganglion cell complex has become a very important parameter for those of us who take care of glaucoma. The ganglion cell complex is something that you can't measure at your slip lamp and you can't measure with other technologies at this time
Well, what is the ganglion cell complex? Well, the ganglion cell complex consists of three major constituents - first, the retinal nerve fiber layer, and of course, the retinal nerve fibers are the axons of the retinal ganglion cells. It also includes the retinal ganglion cell soma or cell bodies, and then it also includes the inner plexiform layer. The inner plexiform layer are the dendrites of the retinal ganglion cell
So, when we're measuring GCC with an OCT, we are now measuring everything relating to the retinal ganglion cell, including its dendrites, the cell body and the nerve fibers. The retinal ganglion cells extend through these three layers
Many of you have probably seen the GCC. On the top, we're looking here at the standard OCT image of the retinal nerve fiber layer thickness around the optic disc. And on the left, we're looking in the right eye here at the GCC. In the center, we don't have the optic disc, but we have the fovea
In this particular eye, there's loss of GCC, which means there's loss of retinal ganglion cells in the foveal area. You can see the same thing in the cross-section here of the peripapillary retinal nerve fiber layer. This is the cross sectional image, the double hump, which of course is thickest in the superior temporal and inferior temporal regions
On the bottom, we're looking at the left eye. And here again, you see in the center, the optic disc peripapillary area here. And again, this looks--doesn't look too bad
But, you look here at the GCC map and you can see there's extensive loss of retinal ganglion cells
So, GCC - very, very helpful for monitoring glaucoma. There are some--some colleagues think that the GCC might be more sensitive than standard retinal nerve fiber layer thickness measurements taken around the optic nerve head
In fact, there are many glaucoma specialists now who believe that the GCC analysis may detect damage before standard retinal nerve fiber layer damage, the types of damage that we're used to seeing in the peripapillary region
As another example, here you can see, on the right, we're looking at the GCC, and over here, you're looking at right eye and left eye, the optic nerve head analysis with the optic disc in the center and the cross-sectional analysis of the retinal nerve fiber layer thickness
In this particular patient, we look at the thickness map, we can look at the deviation map, which is the difference from normal, and then we can look at the significance map and see that there are major losses of retinal ganglion cells
So, one advantage of imaging, current imaging is that we can detect features like the GCC that we can't do with the slip lamp. Another advantage is that the data collection is digital. It's objectively processed and analyzed, and you can export it to an electronic medical record for analysis and storage
Digital data is increasingly important as we all move to electronic medical records. Optovue is releasing very soon a new digital imaging device called the EyeCam Fundus Camera, the EyeCam Camera. This is non-mydreatic camera, digital camera. It can give you a 45 degree color fundus image, or you could digitally zoom and magnify the optic disc for analysis in glaucoma, or you can use it for looking at the retina, as well. And you can also have red free fundus images
What's unique about this camera is the cost is gonna be something that's very affordable to all clinicians
Another advantage of imaging is something that I call the structure-structure advantage. All of you have heard about structure and function comparisons
What you can do, now that we have digital images of the optic disc, the macula and the entire fundus, is you can take the structure of the optic disc from your photographic image and you can superimpose on it the structural information that you've acquired from your OCT. And this has been done here with the EyeCam and the RTView in which you're looking at the peripapillary nerve fiber layer superimposed on the fundus image. And here, you can look at the GCC superimposed on the macula
Glaucoma is heading towards the analysis of structure and function. Both structure and function are necessary for detecting and monitoring change in glaucoma
Most patients don't change in both structure and function at the same time. Very early in the course of disease, we typically see changes in structure proceeding changes in function, and very late in the course of the disease, it's more common to see changes in function without changes in structure
There are benefits from combining structure and function. It's not clear, though, what the best ways for doing this are at the current time. There have been linear aggression models, there have been machine learning classifiers, and there's been very sophisticated statistical analyses
One approach employed by Ted Garway-Heath and his colleagues at Moorefield's is to map the visual field to the optic disc and red nerve--retinal nerve fiber layer defects in red free photographs are used to map the visual field test points to the corresponding location on the optic disc, and this creates a map where the visual field locations can be compared to the retinal nerve fiber layer thickness at the optic disc margin
Another approach is using something called neural networks where--in particular, something called a relevance vector machine. And with these types of analyses, you can more specifically and more sensitively detect glaucoma and glaucoma change than with structure or function alone
And then, the third type of approach is to use fairly sophisticated bio statistical means such as Bayesian Analysis in order to combine structure and function. With this type of approach, information derived from one test is allowed to influence the inferences obtained from the other test
As an example, a visual field that appears to be normal or unchanged might with such an analysis be shown to actually have changed. Here's an example in which we're looking here at a series of visual fields obtained over a five year period
And in this particular situation, if you look at the one in 2004 and you look at the one obtained in 2008, they don't appear different. The rate of progression appears very slow. And in fact, the slope calculated from conventional analysis is not significant
When we begin to use some of the structural information, whether from photographs, whether from an instrument like a GDX or from an OCT, what we find is that the same information now shows that there are significant changes. The structure and functional information are used together, and it allows us to reanalyze the visual field data and determine that, in fact, there is progression
This type of method allows us to detect more patients who are progressing. For example, if we use only standard automated perimetry in a group of subjects, very few patients progress. If we look at progression with the GDX instrument, an imaging instrument, there's a larger group of patients that progress, but we can begin to refine our data when we use our Bayesian Analysis approach
Structure and function is the future of glaucoma diagnosis and monitoring. One approach to being able to use this in your clinical practice is to take visual field information that you've obtained from your patients and present it simultaneously on a computer monitor or on a printout. It's done here with the RTView data on the bottom and information from a perimeter - in this particular case, an oculus perimeter - on the top. And it allows for ready comparison between the two
One of the biggest challenges with progression is being able to reliably detect change. And one new advance that is coming to the Optovue technology with the introduction of a new faster OCT will be specialized motion correction technology that was developed by Dr. James Fujimoto at MIT and his colleagues
And this particular instrument now has been made much faster, it has 70 kilohertz speed, and it has this proprietary motion correction algorithm, which generates distortion-free 3D images. In brief, what's done is, instead of scanning in one direction, scanning is now done very quickly in two directions, horizontal and vertical. And the information then is combined and registered to get a much better image and much more accurate data
Here's an example of what happens when you don't have a fast OCT. As you know, when you require an OCT image, whether you're doing it for glaucoma or for retina, it takes a couple of seconds. And during that couple of seconds, the eye can move
And so, what happens--and these are a series of serial B scans here. You can see, the blood vessels, which are shown here by the shadows, from image to image, they're aligned here, but they're out of alignment as you move through the different times from here down to here
With the new imaging software, which acquires everything much quicker, you can see that the blood vessels, not only are they aligned here, but they're aligned throughout the entire scan
So, this new software offers accurate image registration based on distortion free images, reproducible test locations, and it provides a very good foundation for progression detection with reduced test retest variability
Both the World Glaucoma Association Consensus Initiative and the American Academy of Ophthalmology preferred practice patterns now recommend imaging as part of routine clinical care. And most of us are using it as part of routine clinical care
There are some caveats, however, that are worth mentioning. Imaging is not 100 percent accurate. So, you need--still need good clinical judgment and consideration of all clinical information
Second, assessment of quality is very important as poor quality is susceptible to artifacts, the type of artifacts that you get from slow image acquisition, slow scanning. With the faster instruments now and the new correction software, this becomes less of a problem
Third, you know, when there are atypical patients who are problematic--the ones in my practice that are the most challenging are the high myopes or the individuals with tilted discs. Normative databases may not be representative of all populations
If you're looking, for example, at a patient who is of Asian ancestry - as an example, Chinese ancestry - the normative database would be very different from one that has been acquired for patients of European ancestry. So, you need specific types of normative databases that might be ethnic dependent
And then, last, normative databases are also based on statistics. So, there's always a low probability that a patient outside normal limits is just a normal at the low end of the distribution
So, is imaging ready for clinical practice? I think all the evidence suggests that we should be employing OCT in our clinical practice. I think the technology now has sufficiently improved that it provides meaningful data for diagnosis and for progression
And as the speed of these instruments has increased, and we begin to have specialized software that allows us to correct some of the distortions that are artifacts of the imaging process, the types of information that we'll obtain are gonna be even better than we've had before
Thank you very much