On Friday the 6th of February, I attended the Glaucoma 360 New Horizons forum, a meeting bringing together clinicians, scientists, venture capitalists, as well as representatives from device manufacturers, pharmaceutical companies and the FDA.
Typically, the Friday meeting’s focus – at least in the three years that I have attended – has been on surgical devices, methods of monitoring intra-ocular pressures (IOP), new drug development, and, necessarily, the reporting of results from different phases of various clinical trials. Far less time is given to diagnostics, although the key note speeches have in the past touched on this. This year was no exception, so the following reports on that and also a talk on biomarkers, both, therefore, relating to my own interests in diagnostics.
Opportunities and Challenges for Innovation in Glaucoma – Paul P. Lee, MD, JD
This year’s keynote was presented by Professor Paul Lee, Chair of the Department of Ophthalmology at the University of Michigan. This talk was an interesting, high-level overview of the state of glaucoma management and treatment. Professor Lee started with a forewarning of the forecasted growth of glaucoma cases in the US given in part due to the shift and growth in demographic. The cost of treatment will therefore only rise, yet early intervention is a known way to keep these costs down; the flipside being that, at least initially, glaucoma is an asymptomatic disease, so diagnostics and screening will have to improve. In terms of early intervention, Professor Lee stated that, “the focus of the health technology sector needs to be on community health metrics”, where non-professionals could employ simple to use technologies to look for early warnings of disease. Thereby: “continued technological innovation in our therapies and diagnostics [could be attained via] greater use of data.”
Professor Lee reinforced this notion with an example: just 15 years ago, 40% to 80% of glaucoma specialists were not using optic nerve head (ONH) exams routinely. Furthermore, around half of their primary open angle glaucoma patients were only being seen every 15 months. Imaging has improved and resolved the prior lack of ONH examinations, but frequency of visit can still be an issue given the workload of today’s ophthalmologist. Options such as telemedicine are certainly compelling, yet, despite effectiveness, this can also lead to more ophthalmologist visits.
A clear solution is not there, but improved diagnostics, in the hands of the clinician when in the presence of the patient, offer to lighten the load. This would need: “[the] automatic provision of decision support as a part of the workflow.” Indeed, Professor Lee cited study from Kawamoto et al. showed that in such cases, the effective treatment was significantly improved. If we take imaging as an example, looking at OCT in particular, this all means that structural quantification needs to be not only accurate, but fast.
Professor Lee book-ended his entire keynote with reference to a Robert Wood Johnson foundation study that, based on the how healthcare is applied, characterized outcomes to be in the range of “Healthcare if you can get it”, to “Big Data; Big Gains”. “Slow reform, better health”, was the middle ground, which he suggested, we should rise above, aiming instead for the big gains applying appropriate and innovative technologies to get there.
Early Neuronal Changes: Markers for Glaucoma – Andrew D. Huberman, PhD
The second talk I will report on was that from Dr. Andrew Huberman, Assistant Professor at the Department of Neurosciences at the University of California, San Diego. Dr. Huberman is also part of the “Catalyst for a Cure” program, funded by the Glaucoma Research Foundation, the organizers of this meeting.
Very clearly stated, the goal of Dr. Huberman’s collaborative work is the discovery of novel biomarkers to monitor glaucoma disease progression, and to ultimately support diagnosis as, for example, “[an] eye exam is not a direct indicator of retinal ganglion cell [GC] health.” Indeed, Dr. Huberman continued, you cannot resolve the ganglion cells using OCT as the resolution is simply not there. Instead, clinically, one needs to “see and count individual cells”, as “an indirect measure is not going to work”.
To step back a little here and give some context to the talk, it is clear that there are biomarkers for glaucoma. These include IOP, Visual Fields (VFs) and also OCT-based measures; primarily of the retinal nerve fiber layer [RNFL] thickness measured peripapillary around the ONH. But, we do know from clinical studies that the GCs, measured using OCT as a complex that typically includes the inner plexiform layer (IPL – the GCs’ dendrites), are showing efficacy as a biomarker. It is true, however, that the interface between GC layer (GCL) and the IPL is not easy to resolve in today’s OCT devices but as a surrogate, this particular complex (GCL + IPL) is showing clinical efficacy that, I think, was somewhat unfairly overlooked in this talk.
Dr. Huberman is instead looking at the “different types of ganglion cells” in mice, using for example, genetic tools, to see what is affected when IOP is artificially increased simulating the onset of glaucoma. The different types of ganglion cells they have seen were broadly described as large and small, with some responding to light-on stimuli and others responding to light-off events. Early results seem to indicate that it is the light-off RGCs that are affected first with glaucoma. The clinical consequence of these findings, if confirmed, would be to configure VF tests to probe the “off-cell” visual pathway.
Their collaborative effort is looking too to see if the retinal microvasculature, that sits along the GCL, to see if that is somewhat deficient. Currently, this is only via visualization techniques, the most exciting looking being depth-resolved OCT spectroscopy that can image function in the form of flow speed.
The “ultimate readout”, however, was concluded to be via the use of adaptive optics where you might count the number of GCs within a given area. Unfortunately, although this has been done effectively for imaging rods and cones at the back of the retina, this has not yet been done for the GCs which will likely not show the necessary contrast for such imaging as they are basically transparent to light. Furthermore, exciting though adaptive optics is, any opacities in the vitreous render it very difficult to capture high-resolution images. More optimistically, we were shown “individual nerve fiber axons”, although, at the scale shown, they were actually bundles of fibers, and not individual axons. Nonetheless, the use of adaptive optics to image the RNFL seemingly showed appearance change with disease progression. This has yet to be confirmed in larger populations and is likely difficult to quantify, but in regard to the use of adaptive optics for glaucoma disease management, that’s all we were offered.
Clearly, these are exciting areas of research, but this is work very much embedded in the lab for the near future. The structural changes we saw using imaging were still only hypothesized to relate to glaucoma progression as studies that would confirm this will be very hard to perform given the complexity of the acquisition and the lack of objective analysis tools. Quantification, even in animal studies, will take some effort. That said, clearly this is important work as such methods, probing both function and structure, will certainly embellish future diagnostics, potentially adding to the clinician’s armamentarium. The transition from the lab to the clinic is the stated and very worthy objective. I very much look forward to future updates from this effort.