NeuroVoices: Sharon Cohen, MD, FRCPC, on the Advantages Retinal Imaging Brings to Alzheimer Disease Care

Sharon Cohen, MD, FRCPC

This is a 2-part interview. Click here for Part 1.

Biomarkers found in blood, body fluids, organs, and tissues are measurable indicators of what’s happening in the body. When in combination with other tests, these biomarkers can help doctors determine whether a person may be at risk of developing Alzheimer disease (AD) or a related dementia. Although the boom of biomarkers has become a critical component to the community, there is no consensus on which are the most valuable or necessary to accurately show disease progression.

There are several different types of biomarkers and tests, including brain imaging scans such as CT, MRI, PET, and more specifically, amyloid PET, tau PET, and fluorodeoxyglucose PET. Cerebrospinal fluid, a clear fluid that surrounds the brain and spinal cord, has also been a valuable tool for early detection of neurodegenerative diseases and to assess the impact of experimental medications. The retina, a protrusion of the central nervous system, is a promising candidate for the noninvasive identification of various AD biomarkers. One such technology, RetiSpec’s hyperspectral retinal imaging, has shown to predict brain amyloid-ß status in high-risk individuals better than standard approaches.

RetiSpec and retinal imaging in general, could have an extremely positive impact on the AD community, Sharon Cohen, MD, FRCPC, believes. Cohen, a neurologist and assistant professor at the University of Toronto, led the validation study of RetiSpec and presented its results at the 14th Clinical Trials in Alzheimer Disease (CTAD), November 9-12, 2021.As part of a new iteration of NeuroVoices, Cohen provided insight on the advantages of retinal imaging, the ways in which it optimizes drug development, and whether this approach changes the clinician views of other biomarkers.

NeurologyLive^®: What ways can RetiSpec positively impact the drug development stage?

Sharon Cohen, MD, FRCPC: Alzheimer trials have traditionally been very long, expensive, and tended to disappoint in terms of their results. The trial designs are getting more sophisticated, and biomarkers are being embedded to assess the biology of the disease, not just the clinical symptoms. If we go by clinical symptoms, what’s memory loss for me vs you vs someone down the street, is not as crisp of a measure when we measure something biologically. We can’t develop effective treatments if we’re not diagnosing properly. We know that on clinical grounds, when we say someone has mild cognitive impairment and we think it’s due to Alzheimer disease, we are wrong 50% of the time when we look for biomarker confirmation using PET, let’s say.

In the best hands, clinical diagnosis is not good enough. If we’re going to have a therapeutic that targets the mild cognitive impairment stage of Alzheimer disease, and we’re wrong 50% of the time, we’re going to be giving drugs to the wrong people and exposing them to risk. We need accurate diagnosis. For drug development, you can’t expect to see a treatment effect if you have people enrolled who don’t actually have the disease. That has been the flaw of many past trials. The field is moving forward with biomarker confirmation of cohorts being exposed to experimental therapies; however, this increases the cost. That includes PET amyloid scans, or the burden to patients at sites of spinal fluid analysis, and lumbar punctures. If we could use RetiSpec to prescreen or perhaps in the future to screen from a biomarker standpoint, that would be awesome. It would be cost-saving, and make trials more effective and more efficient.

Is there any potential for this tool to change how we value certain biomarkers?

Biomarkers have exploded onto the scene in the last few years. We’re learning a great deal about the disease from them. I think we have to think about biomarkers in various categories. There are biomarkers of Alzheimer disease that are important for predicting diagnosis, ones that are confirmatory of diagnosis, and then others that will help us stage the disease. This allows us to not have to use just clinical criteria, how much memory has changed or how much language is lost, or how well someone’s functioning. If I’m retired and I’m not working, the demands on me are less than if I have an intense job, and it may be fuzzy what my true level of functional decline is. Having biomarkers that can stage the disease just the same way we stage and diagnose in cancer, is important. We stage is by how much it spreads to local areas, to lymph nodes, to other organs. We need to be able to stage the extent of Alzheimer disease and then we need biomarkers that track the impact of interventions and predict whether an intervention is effective.

There’s a lot of discussion about, is amyloid lowering enough for us to say that will predict clinical benefit? For example, do we need to see a tau biomarker decline either on a PET tau scan or one of the isoforms in plasma or spinal fluid that measure tau. It’s very complicated. Then there are markers of neurodegeneration that are not specific to Alzheimer but give us a take on how much brain cell loss there is. Then there are biomarkers that have been used effectively in other neurologic diseases like multiple sclerosis or spinal muscular atrophy. There’s a biomarker not specific to Alzheimer called neurofilament light that can be measured in the blood, so that’s awesome. It can track not just disease progression, but the impact of an intervention of an approved therapy. In Alzheimer disease, we’ve got this whole spectrum and more biomarkers coming onto the scene. Some are imaging biomarkers, some are spinal fluid, some are plasma, and now there’s this retinal biomarker. I think we’re going to need several. It’s not going to be a matter of having one biomarker that will tell the whole story, but a select few biomarkers for the right purpose. Is it for prescreening? Is it for confirmation of diagnosis? Is it to predict results to therapy? Is it to track or stage disease?

Is there anything else notable you’d like to mention?

While we do need a full range of biomarkers, what we do in research and what’s possible in the real world are two different things. Having the technology like the RetiSpec biomarker or plasma biomarkers really facilitate real-world diagnosis and care of patients so that the cost is low, you’re not doing invasive procedures, you’re not needing to ship fluids across the country to specialized labs or exposing patients to radiation. With the RetiSpec scan, it also allows other players into the arena of early diagnosis. We know we don’t have enough family doctors or enough specialists, but ophthalmologists, optometrists who regularly see seniors and take pictures of their eye to look for other retinal diseases, could potentially be partners in identifying and screening for early changes in Alzheimer. It opens another opportunity for us to have greater skilled practitioners involved in the early diagnosis where we need that. That’s where the field is moving in therapeutics, to try to intervene early in the disease. How are we going to do that if we’re not identifying enough people in the real world?

Transcript edited for clarity. Click here for more installments of NeuroVoices.

REFERENCE
1. Cohen S, Plante A, Hazan Al, et al. RetiSpec’s hyperspectral imaging system: results of a validation study in preclinical AD and MCI. Presented at CTAD 2021; November 9-12. Poster LP16