Alzheimer disease (AD) is a progressive neurodegenerative disorder that is estimated to currently affect more than 6 million Americans. Its prevalence is expected to increase to over 13 million by 2060 as the US population ages.1 Alzheimer disease poses a significant financial challenge to both individual caregivers (unpaid dementia caregiving was valued at $339.5 billion) and health care systems, with long-term care estimated at $345 billion in 2023.1
Alzheimer disease is characterized by the accumulation of abnormal protein aggregates in the brain. It involves the formation of hyperphosphorylated tau protein leading to neurofibrillary tangles and the accumulation of amyloid-𝛽 plaques which result from the processing of amyloid precursor protein. These pathologic changes disrupt neuronal function, leading to neuronal loss, brain atrophy, and cognitive decline.
Early identification of AD is critical for making lifestyle changes and weighing the potential benefits of available therapies to possibly slow the progression of AD, particularly at the early stages. It also empowers families to plan for the future and make informed decisions in the face of a life-altering diagnosis. Recently, retinal imaging metrics have emerged into the biomarker landscape for the possible early detection of AD. These metrics promise a chance at early diagnosis and an enhanced understanding of the pathophysiology of AD.
Current Diagnostic Challenges
Established methods to diagnose AD include mental and cognitive status testing (eg, mini mental state examination, Montreal cognitive assessment), cerebrospinal fluid examination for amyloid-𝛽 and tau, and brain imaging such as volumetric magnetic resonance imaging and positron emission tomography. Many of these diagnostic tools are costly, invasive, and not readily available to all patients or considered standard of care for reimbursement. Additionally, these methods typically diagnose AD at later stages when amyloid-𝛽 deposition is significant and the disease has become symptomatic.3
AD is known to have a long preclinical asymptomatic stage typically preceding onset of symptoms by 15 to 20 years in which certain neuropathologic features are evident.4 This raises the possibility for detection of AD at this presymptomatic stage. Recently, advances in retinal research have demonstrated the possibility of retinal imaging as a noninvasive, efficient, and economical way to screen for earlier findings attributed to AD.
The Retina as an Extension of the Central Nervous System
Developmentally, the retina arises from brain tissue and is considered an extension of the central nervous system.5 Some diseases of the central nervous system have readily detectable intraocular manifestations, making the retina a helpful window to monitor concurrent abnormalities in the brain. For example, retinal venular dilation has been observed in cerebral small vessel disease,6 thinning of the retinal nerve fiber layer (RNFL) has been found in multiple sclerosis,7 and foveal thinning and foveal avascular zone (FAZ) remodeling has been identified in Parkinson disease.8 Retinal biomarkers in AD research are a relatively recent and promising area of investigation. These biomarkers are specific features identified in the retina that can provide valuable insights into the presence, progression, response to intervention, and potentially some of the underlying mechanisms of AD.
Types of Retinal Biomarkers
Optical coherence tomography (OCT) and OCT angiography (OCTA) have emerged as potential imaging tools to quantify differences in retinal architecture and microvasculature at all stages of AD, from the preclinical phase to the symptomatic stage.9 Investigations of retinal microvasculature changes in AD involve a comprehensive array of imaging metrics extracted from these multimodal retinal images. These imaging metrics include macular ganglion cell-inner plexiform layer (GC-IPL) thickness, central subfield thickness (CST), peripapillary RNFL thickness, choroidal vascularity index (CVI), fractal dimension (FD), vessel width gradient (WG), FAZ area, vessel density, perfusion density, and peripapillary capillary perfusion density (CPD), among others.
Retinal Biomarker Associations With Alzheimer Disease
Investigations of retinal microvascular findings in AD have identified several key imaging metrics that have been able to distinguish AD from controls. Several studies have found FAZ area to be an important metric that is increased in AD compared to controls.10-12 Other studies have also found a significant increase in FAZ area in mild cognitive impairment (MCI) as well.11,13,14 Vessel density in the superficial capillary plexus (SCP) is another potentially promising retinal biomarker for AD. Multiple investigations have demonstrated decreased vessel density in AD compared to controls,10,15-18 a decrease that is also evident in AD compared to MCI.15 Perfusion density in the SCP is reduced as well in AD compared to MCI and to controls.15
In a pair of identical twins discordant for AD, OCTA-based metrics were used to compare the retinal vasculature and showed that the twin with AD had significantly lower vessel density and a significantly larger FAZ area in the SCP.19 This demonstrates that distinct changes in retinal vasculature occur in AD even when other genetic and nongenetic factors are not markedly different. It also supports the evidence of the utility of retinal microvascular changes on OCTA as a potential biomarker for AD. It is worth noting, however, that while these differences have been seen between an AD cohort and controls, no significant distinctions in vessel density and perfusion density have been observed between early onset (starting before age 65) and late onset (starting after age 65) AD.18
In peripapillary OCTA images, CPD — representing the percentage of perfused capillaries in a unit scan area — also differs significantly between AD and controls. Capillary perfusion density is significantly greater in AD vs controls and in MCI vs controls. This makes CPD another potentially useful biomarker for AD.20
Using macular OCT, GC-IPL thickness was found to be decreased in AD and MCI compared to controls.15 However, no difference was found in CST and RNFL thickness in AD compared to both MCI and controls.15
Choroidal vascularity index, an OCT-based metric for choroidal structural analysis defined as the ratio of vascular luminal area to total choroidal area,21,22 was found to have a more mixed pattern of results. CVI was found to be significantly lower in patients with MCI compared to control subjects,23 but a difference in CVI was not observed between subjects with AD and control subjects. This finding suggests complex choroidal vascular changes in MCI and AD, and this relationship needs to be further studied.
Several parameters derived from ultrawidefield (UWF) imaging may also help distinguish AD from MCI and controls. Midperipheral FD quantifies the level of vascular branching in the midperipheral retina and is increased in AD compared to controls. In MCI, midperipheral FD is decreased compared to controls.24 Vessel WG measures the rate of vessel thinning as vessels move farther from the optic nerve. Retinal arterial WG was found to be increased in AD and MCI compared to controls, indicating significant arteriolar thinning in both AD and MCI. Retinal venous WG was increased in MCI compared to controls, indicating venular thinning, but no significant difference was found in venular WG between AD and controls.24
Clinical Applications
Multimodal retinal imaging, such as OCT and OCTA, is more affordable and much less invasive than lumbar puncture procedures. It also has fewer medical contraindications than magnetic resonance imaging (MRI) with far shorter acquisition times. These considerations, along with strong evidence for retinal microvascular changes in AD, make OCTA a promising tool that can be widely used for screening purposes.
These findings also highlight the complexity of retinal vascular changes in relation to different stages of cognitive impairment, with variations in findings among studies. Several significant biomarkers associated with AD in some studies were found to be insignificant in others.9 Differences among study results collectively underscore the intricate nuances involved in retinal image assessment, encompassing factors such as the choice of imaging technology, segmentation and measurement protocols, and participant characteristics, all of which can significantly influence the outcomes across different stages of cognitive impairment.
Currently, published studies use different OCT and OCTA imaging devices. Each imaging device has its own segmentation algorithm and proprietary software for quantitative analysis, and thus data between such devices are not directly comparable.25-27 To draw meaningful conclusions about optimal biomarkers for AD screening, it will be necessary to standardize both the segmentation algorithms and microvascular metrics being reported and this will in turn improve efficiency and collaboration.9 Metrics with good intrasession repeatability are critical for studies measuring retinal microvasculature over time. Peripapillary average CPD, FAZ area, vessel density, and perfusion density have all shown excellent intrasession repeatability and they also have good interocular symmetry.28-30 These metrics may be especially suitable for further use in studies of AD-associated retinal biomarkers.
Future Directions: Artificial Intelligence
Machine learning approaches have been applied to the task of identifying AD using OCTA and other types of retinal images. Such approaches further highlight the pressing challenges surrounding the need for standardizing imaging platforms and analysis metrics. Machine learning may facilitate high-throughput screening of populations at risk for AD. Many image-based models use a convolutional neural network (CNN) architecture, which is a type of deep learning model designed for image processing that iteratively learns patterns in image data. A CNN has been developed to differentiate MCI from normal cognition.31 Another CNN model is able to identify AD from multimodal images and performs well when combining GC-IPL, quantitative data, and patient data as model inputs.32 These and other models can be paired with deep learning‒based image quality assessment models to further automate the workflow of using retinal images to screen for neurodegenerative diseases.33
Conclusion
The growing burden of AD underscores the urgent need for accurate diagnostic tools and the ability to monitor pathophysiologic disease progression. Retinal biomarkers have emerged as a promising avenue, offering a less invasive, more cost-effective, and more widely accessible modality to achieve these goals. These biomarkers hold the potential to provide valuable insights across the spectrum of AD stages and are increasingly important as novel therapeutics enter the previously barren treatment landscape.
As we strive to better understand the existing heterogeneity among some study results, standardizing measurement techniques and protocols to facilitate meaningful comparisons across studies may offer insights. The integration of artificial intelligence, particularly CNNs, is showing promise in automating probability of AD and streamlining the analysis of retinal images. In combination with artificial intelligence, retinal images have demonstrated impressive performance as we attempt to distinguish AD disease stages, offering hope for not only more precise and consistent results, but also faster and more cost-effective screening to help us better manage the expected future increase in the burden and prevalence of AD. RP
Editor’s note: This article is discussed on the Retina Podcast at www.retinapodcast.com.
References
1. 2023 Alzheimer’s disease facts and figures. Alzheimers Dement. 2023;19(4):1598-1695. doi:10.1002/alz.13016
2. Skaria A. The economic and societal burden of Alzheimer disease: managed care considerations. 2022;28. Am J Manag Care. 2022;28(10 Suppl):S188-S196. doi:10.37765/ajmc.2022.89236
3. Jack CR, Knopman DS, Jagust WJ, et al. Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol. 2010;9(1):119-128. doi:10.1016/S1474-4422(09)70299-6
4. Sperling RA, Aisen PS, Beckett LA, et al. Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7(3):280-292. doi:10.1016/j.jalz.2011.03.003
5. London A, Benhar I, Schwartz M. The retina as a window to the brain-from eye research to CNS disorders. Nat Rev Neurol. 2013;9(1):44-53. doi:10.1038/nrneurol.2012.227
6. Ikram MK, De Jong FJ, Van Dijk EJ, et al. Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan Study. Brain. 2006;129(Pt 1):182-188. doi:10.1093/brain/awh688
7. Petzold A, Boer JF de, Schippling S, et al. Optical coherence tomography in multiple sclerosis: a systematic review and meta-analysis. Lancet Neurol. 2010;9(9):921-932. doi:10.1016/S1474-4422(10)70168-X
8. Kundu A, Ma JP, Robbins CB, et al. Longitudinal Analysis of retinal microvascular and choroidal imaging parameters in parkinson disease compared to controls. Ophthalmol Sci. 2023:100393. doi:10.1016/j.xops.2023.100393
9. Rifai OM, McGrory S, Robbins CB, et al. The application of optical coherence tomography angiography in Alzheimer’s disease: a systematic review. Alzheimers Dement. 2021;13(1):e12149. doi:10.1002/dad2.12149
10. Bulut M, Kurtuluş F, Gözkaya O, et al. Evaluation of optical coherence tomography angiographic findings in Alzheimer’s type dementia. Br J Ophthalmol. 2018;102(2):233-237. doi:10.1136/bjophthalmol-2017-310476
11. Wu J, Zhang X, Azhati G, Li T, Xu G, Liu F. Retinal microvascular attenuation in mental cognitive impairment and Alzheimer’s disease by optical coherence tomography angiography. Acta Ophthalmol. 2020;98(6):e781-e787. doi:10.1111/aos.14381
12. Zabel P, Kaluzny JJ, Wilkosc-Debczynska M, et al. Comparison of retinal microvasculature in patients with Alzheimer’s disease and primary open-angle glaucoma by optical coherence tomography angiography. Invest Ophthalmol Vis Sci. 2019;60(10):3447-3455. doi:10.1167/iovs.19-27028
13. Criscuolo C, Cennamo G, Montorio D, et al. Assessment of retinal vascular network in amnestic mild cognitive impairment by optical coherence tomography angiography. PLoS One. 2020;15(6):e0233975. doi:10.1371/journal.pone.0233975
14. O’Bryhim BE, Apte RS, Kung N, Coble D, Van Stavern GP. Association of preclinical Alzheimer disease with optical coherence tomographic angiography findings. JAMA Ophthalmol. 2018;136(11):1242-1248. doi:10.1001/jamaophthalmol.2018.3556
15. Yoon SP, Grewal DS, Thompson AC, et al. Retinal microvascular and neurodegenerative changes in Alzheimer’s disease and mild cognitive impairment compared with control participants. Ophthalmol Retina. 2019;3(6):489-499. doi:10.1016/j.oret.2019.02.002
16. Lahme L, Esser EL, Mihailovic N, et al. Evaluation of ocular perfusion in Alzheimer’s disease using optical coherence tomography angiography. J Alzheimers Dis. 2018;66(4):1745-1752. doi:10.3233/JAD-180738
17. Jiang H, Wei Y, Shi Y, et al. Altered macular microvasculature in mild cognitive impairment and Alzheimer disease. J Neuroophthalmol. 2018;38(3):292-298. doi:10.1097/WNO.0000000000000580
18. Robbins CB, Grewal DS, Stinnett SS, et al. Assessing the retinal microvasculature in individuals with early and late-onset Alzheimer’s disease. Ophthalmic Surg Lasers Imaging Retina. 2021;52(6):336-344. doi:10.3928/23258160-20210528-06
19. Grewal DS, Polascik BW, Hoffmeyer GC, Fekrat S. Assessment of differences in retinal microvasculature using oct angiography in Alzheimer’s disease: a twin discordance report. Ophthalmic Surg Lasers Imaging Retina. 2018;49(6):440-444. doi:10.3928/23258160-20180601-09
20. Robbins CB, Grewal DS, Soundararajan S, et al. Radial peripapillary capillary network microvascular changes in Alzheimer’s disease, mild cognitive impairment, and cognitively healthy controls. Alzheimers Dement. 2021;17(S4):e050969. doi:10.1002/alz.050969
21. Betzler BK, Ding J, Wei X, et al. Choroidal vascularity index: a step towards software as a medical device. Br J Ophthalmol. 2022;106(2):149-155. doi:10.1136/bjophthalmol-2021-318782
22. Agrawal R, Gupta P, Tan KA, Cheung CMG, Wong TY, Cheng CY. Choroidal vascularity index as a measure of vascular status of the choroid: measurements in healthy eyes from a population-based study. Sci Rep. 2016;6(1):21090. doi:10.1038/srep21090
23. Robbins CB, Grewal DS, Thompson AC, et al. Choroidal structural analysis in Alzheimer disease, mild cognitive impairment, and cognitively healthy controls. Am J Ophthalmol. 2021;223:359-367. doi:10.1016/j.ajo.2020.09.049
24. Pead E, Thompson AC, Grewal DS, et al. Retinal vascular changes in Alzheimer’s dementia and mild cognitive impairment: a pilot study using ultra-widefield imaging. Transl Vis Sci Technol. 2023;12(1):13. doi:10.1167/tvst.12.1.13
25. Lu Y, Wang JC, Zeng R, et al. Quantitative comparison of microvascular metrics on three optical coherence tomography angiography devices in chorioretinal disease. Clin Ophthalmol. 2019;13:2063-2069. doi:10.2147/OPTH.S215322
26. Munk MR, Giannakaki-Zimmermann H, Berger L, et al. OCT-angiography: a qualitative and quantitative comparison of 4 OCT-A devices. PLOS ONE. 2017;12(5):e0177059. doi:10.1371/journal.pone.0177059
27. Lei J, Pei C, Wen C, Abdelfattah NS. Repeatability and reproducibility of quantification of superficial peri-papillary capillaries by four different optical coherence tomography angiography devices. Sci Rep. 2018;8(1):17866. doi:10.1038/s41598-018-36279-2
28. Robbins CB, Grewal DS, Thompson AC, Yoon SP, Plassman BL, Fekrat S. Repeatability of peripapillary optical coherence tomography angiography parameters in older adults. J Vitreoretin Dis. 2020;5(3):239-246. doi:10.1177/2474126420953968
29. Lee WH, Lee MW, Kim MS, Ryu CK, Kim JY. Long-term repeatability of peripapillary optical coherence tomography angiography measurements in healthy eyes. Sci Rep. 2021;11(1):23832. doi:10.1038/s41598-021-03469-4
30. Akrobetu DY, Robbins CB, Ma JP, et al. Intrasession repeatability of OCT angiography parameters in neurodegenerative disease. Ophthalmol Sci. 2023;3(2). doi:10.1016/j.xops.2023.100275
31. Wisely CE, Richardson A, Henao R, et al. A convolutional neural network using multimodal retinal imaging for differentiation of mild cognitive impairment from normal cognition. Ophthalmol Sci. 2024;4(1). doi:10.1016/j.xops.2023.100355
32. Wisely CE, Wang D, Henao R, et al. Convolutional neural network to identify symptomatic Alzheimer’s disease using multimodal retinal imaging. Br J Ophthalmol. 2022;106(3):388-395. doi:10.1136/bjophthalmol-2020-317659
33. Lee T, Rivera A, Brune M, et al. Convolutional neural network-based automated quality assessment of OCT and OCT angiography image maps in individuals with neurodegenerative disease. Transl Vis Sci Technol. 2023;12(6):30. doi:10.1167/tvst.12.6.30
