Researchers have long sought ways to predict whether a patient will develop Alzheimer's or Parkinson's disease, but their current options are costly and impractical. Now, they're exploring whether an inexpensive picture of a patient's eye could accurately assess risk, the Washington Post's Elizabeth Brown reports.
A need for early detection
Experts and providers say that early detection of a patient's risk for neurodegenerative diseases could support advance planning—and potentially help researchers develop treatments.
Early detection "is sort of the holy grail," Ron Petersen, director of Mayo Clinic's Alzheimer's Disease Research Center and the Mayo Clinic Study of Aging, told Brown.
That's because patients typically already have experienced a significant amount of neurodegeneration by the time they start showing symptoms, Brown notes. And at that point, current treatments usually aren't very effective.
If providers are able to detect brain changes that could signal the onset of neurodegenerative diseases, the hope is they could slow the disease's progression, Petersen said. "[T]he sooner we intervene, the better we will be" at staving off cognitive impairment, he told Brown.
Why the eye?
As part of that quest, some researchers are examining whether retinal screenings could be a low-cost and noninvasive screening method, Brown reports.
According to Brown, an eye's "retina contains layers of neurons that seem to experience neurodegenerative disease along with" similar neurons located inside a person's skull. The neurons in the eye are located "right against the jellylike vitreous of the eyeball," meaning they "live and die where scientists can see them," Brown writes.
Some researchers believe examining those neurons could give hints at what's happening with neurons in patient's brains, according to Brown. She notes that "Alzheimer's and Parkinson's cause sweeping changes to the landscape of the brain before there are any behavioral shifts—blood vessels atrophy, neurons die prematurely, and snarls of misfolded proteins disrupt communication between surviving neurons."
But according to Brown, current methods for spotting those changes are costly and invasive. For instance, PET imaging can help providers "identify Parkinson's and Alzheimer's pathology in the brain before symptoms begin," but the scans often cost between $3,000 and $6,000, making them "too invasive and costly for widespread use," Brown writes.
In hopes of finding a cheaper alternative, Maya Koronyo-Hamaoui, a neuroscientist and professor of neurosurgery who studies early Alzheimer's intervention and treatment at Cedars-Sinai, has worked with colleagues to develop a technique to identify plaques linked with Alzheimer's disease in the retinal neurons of patients experiencing mild cognitive impairments. The scans cost about $285 each, and they require "modified ophthalmological equipment—and a lot of curry flavoring," Brown writes.
Before undergoing the scan, Koronyo-Hamaoui and her team instruct patients to drink protein shakes containing curcumin, which is "the naturally occurring compound that gives turmeric its color and flavor and is central in curry," Brown notes. Koronyo-Hamaoui and her team instruct patients to "load up" on curcumin because it binds to amyloid beta—the protein found in Alzheimer's plaques—and it "glows yellow when exposed to blue light," Brown reports, allowing the researchers to detect amyloid beta in a patient's retinal neurons.
However, Koronyo-Hamaoui acknowledges that detecting amyloid beta doesn't necessarily mean a patient will develop Alzheimer's disease, as the protein also accumulates in healthy brains as people age. In fact, many people with comparatively higher amyloid beta levels do not experience abnormal cognitive changes, Koronyo-Hamaoui said.
Elsewhere, Ruogu Fang, a biomedical engineer who researches applications of artificial intelligence in health and medicine, along with colleagues at the University of Florida, are using a different—yet comparably simple—method to measure changes in tiny blood vessels in the back of the eye, Brown reports.
Fang and colleagues "use a fundus camera," which is "a specialized iPhone attachment about the size of a book … to take high-resolution" pictures of the blood vessels, Brown writes. Certain changes in the vessels can signal oxygen deprivation, which "contributes to premature neuron death," Brown reports. And early research has shown that computer algorithms can "distinguish Parkinson's patients from healthy controls with an accuracy upward of 70%" from the fundus images, according to Brown.
But Fang similarly noted that identifying vascular changes in a patient's eyes may not always mean the patient is at risk of Parkinson's or Alzheimer’s disease, because vascular changes can occur with other medical conditions as well, such as diabetic retinopathy and traumatic brain injury.
Still, Sharon Fekrat, a vitreoretinal surgeon and co-director of the Duke Neurodegenerative Disease Retinal Imaging Repository, told Brown that the goal isn't necessarily to discover a single biomarker that absolutely indicates neurodegenerative disease risk. Fekrat is working with colleagues on a machine learning project that they're hoping can identify patterns in biomarkers of Alzheimer's disease and indicate which biomarkers are the most effective indicators of a patient's risk.
According to Brown, Fekrat's project so far has identified "thinning of the retina's ganglion cell layer [as] the most predictive power for an Alzheimer's diagnosis," which can be measured "using optical coherence tomography, a patient eye scanning system most eye doctors already use."
Hopes for the future
Fekrat told Brown her team's hope is that, one day, low-cost retinal scans that can detect warning signs for different neurodegenerative diseases will be widely accessible. "We could set up a machine next to a pharmacy," where patients could get their eyes imaged and "we'll have our deep learning algorithm analyze it and spit out a risk score," Fekrat said.
Laura Volpicelli-Daley, an assistant professor of neurology at the University of Alabama at Birmingham, told Brown that such information could inspire patients to take preventive measures to slow the potential onset or progression of neurodegenerative diseases, including exercising and improving their diets.
"Exercise is the only thing I've seen that really helps with Alzheimer's and Parkinson's," Volpicelli-Daley said, and getting patients information about their potential risk for those disease could motivate them to change their habits (Brown, Washington Post, 2/27).