Shedding light on the retina to see healthy and pathological aging

With the advances in medicine and technology, life expectancy has increased significantly in recent decades. While this is positive news, it also brings major challenges to public health systems. As people live longer, the number of cases of age-related disorders, including neurodegenerative diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), has also grown. These debilitating conditions primarily affect the elderly and can have a devastating impact on the quality of life of patients and their families. Considering emerging treatments, one of the most challenging aspects of neurodegenerative diseases is their early diagnosis and continuous monitoring. Current diagnostic methods for most neurodegenerative diseases rely primarily on brain positron emission tomography (PET) scans, cerebrospinal fluid analysis, genetic testing in cases where pathogenic variants are implicated and, more recently, plasma biomarker analysis. In AD and other dementias, plasma biomarkers have been widely used in diagnosis in recent years, as a less invasive approach and with highly sensitive techniques such as single molecule array. In fact, a highly accurate blood test was developed and surpassed cognitive tests and CT scans in detecting AD. In the developed test, the researchers reached a high accuracy (over 88%) in diagnosing AD in both primary and secondary care (Palmqvist et al., 2024). In recent years, however, an increasing number of clinical findings and basic research studies have suggested that morphological or functional alterations in the retina may serve as an early indicator of neurodegenerative diseases. As part of the central nervous system, the retina is expected to exhibit morphological, cellular, and molecular alterations akin to those observed in the brain. In this perspective, we explore the potential role of the retina in the evaluation of healthy and pathological aging, including early diagnosis of neurodegenerative diseases. The retina is a neural tissue whose development is coordinated with the diencephalon’s. Structurally, the retina contains stratified layers with specific cell types that recognize and convert light stimuli into electrochemical signals, transmitting this information to the brain. Beyond its essential role in vision, the retina can mirror overall brain health. Several parallels can be drawn between brain and retina development, aging, disorders, and inflammation. Increasing evidence suggests that the retina may show early indications of neurodegenerative diseases (Casciano et al., 2024), even before clinical cognitive signs appear. Thus, advancements in our understanding of molecular markers and functional/morphological changes in the retina offer a promising avenue for the early diagnosis of neurodegenerative diseases. Animal models and post-mortem studies provide valuable resources for understanding the parallels between brain and retinal alterations in neurodegenerative diseases. Such studies are essential to understand how the retina may reflect disease progression. In a post-mortem study, Koronyo et al. (2023) observed the presence of amyloid-β peptide (Aβ) aggregates in the retina in both mild cognitive impairment and in the AD continuum, along with retinal gliosis and morphological alterations. The authors further found that intraneuronal accumulation of Aβ immunoreactivity predicted cognitive status, especially when the accumulation was observed in the peripheral region of the retina (Koronyo et al., 2023). Additionally, they reported that the retina thickens as pathology worsens, as does retinal gliosis. Taken together, retinal thickness and the presence of Aβ aggregates may represent a possible way to diagnose AD. Even though the underlying causes of sporadic and autosomal dominant neurodegenerative diseases are different, their symptoms and biological markers are often very similar (Almeida et al., 2024). Sporadic cases are the most common manifestation of neurodegenerative diseases; however, genetic cases provide the unique advantage of studying patients’ years before detectable biological markers or the onset of symptoms. These studies are instrumental in understanding disease progression, identifying early biomarkers, aiding in diagnosis, and developing targeted therapeutic strategies. Furthermore, insights gained from genetic cases can be used to compare different populations, enhance early diagnostic techniques, and tailor preventive interventions. By leveraging the genetic forms of diseases, researchers can elucidate pathophysiological mechanisms, ultimately contributing to improved clinical outcomes for all patients affected by neurodegenerative disorders. For example, Adhikari et al. (2023) showed that autosomal dominant AD presents with a remarkable retinal signature. Most common retinal changes of autosomal dominant AD include ganglion cell modifications leading to axonal projection alterations. Consequently, the optic tract was also altered in AD. The non-invasive nature of retinal clinical exams makes them particularly advantageous, offering the potential to detect a wide range of brain alterations associated with aging. Notably, PET scans and retinal optical coherence tomography (OCT) imaging have been utilized to detect amyloid plaques and measure retinal layer thickness, representing new possibilities for diagnosing neurodegenerative diseases and monitoring disease progression. Technological advancements have also brought up a range of new possibilities. One of these advancements includes OCT angiography, which enables detailed quantitative analysis of the retina and a high-definition evaluation of the retina’s microvascular network. Vascular analysis may include vessel size and diameter, blood flow, and real-time directionality. Sheriff et al. (2023) recently presented a meta-analysis of the potential relationship between retinal changes detected through OCT angiography and AD diagnosis. Although a high heterogeneity between published studies was found, they concluded that OCT angiography parameters such as the reduction in superficial and deep vessel density, and the alteration in the foveal avascular zone, may represent good indicators for diagnosing AD. On the other hand, it is crucial to assess additional and classical parameters, such as nerve fiber layer thickness and vascular density, as they could serve as additional indicators of AD. Although Sheriff et al. (2023) focused their meta-analysis on AD, it is worth mentioning that similar alterations may be observed in PD and other dementias. While retinal exams such as PET scan, OCT, and their variants offer non-invasive means of diagnosing neurodegenerative diseases, it is worth noting that these measures are typically conducted post-diagnosis of a neurodegenerative disease. Evidence suggests that, alongside morphological alterations, signs of atrophy of retinal ganglion cells are associated with AD. However, it is essential to recognize that morphological alterations in ganglion cells may only manifest after the onset of clinical AD. Nevertheless, the retina holds promise as a potential avenue for detecting AD markers years before the onset of classical symptoms such as cognitive impairment. In this scenario, Asanad et al. (2021) showed that retinal ganglion cells present electrophysiological alterations before clinical diagnosis of dementia. They showed that patients with preclinical AD, diagnosed based on the Aβ42/Tau ratio in the cerebrospinal fluid, presented alterations in the electroretinogram response. Combined with regression analysis, these changes could predict AD with high sensitivity before alterations in normal retinal morphology (Asanad et al., 2021). In clinical practice, diagnosis based on retinal alterations has traditionally relied on anecdotal experience. To avoid such personal bias, artificial intelligence (AI) and machine learning have been put into play. Feeding AI algorithms with retinal image data correlated with parallel diagnoses of neurodegenerative conditions – such as brain PET scans, cerebrospinal fluid analysis, and genetic testing – can assist in analyzing the complex data derived from retinal images and may standardize clinical conclusions. In other words, AI-driven approaches can improve diagnostic accuracy and predict disease progression using retinal exams that could be performed years before diagnosis (Ashayeri et al., 2024; Parmar et al., 2024). Indeed, a recent study showed that AI helped to diagnose PD patients based on eye scans up to 7 years before the onset of clinical signs (Wagner et al., 2023). A previous study has also shown that PD patients presented an accumulation of alpha-synuclein in the retinal inner cell layers, which correlates with a decrease in the morphometry of these layers (for review, see Movio et al., 2023). Thus, machine learning–guided comparison between healthy and PD retinas has consistently shown that thinning the retinal inner layer can predict PD risk with high efficacy (Wagner et al., 2023).Figure 1: General scheme of ideas proposed here.Created with BioRender.com. Aβ: Amyloid-beta; AI: artificial intelligence; αSyn: alpha-synuclein; ERG: electroretinogram; NFL: nerve fiber layer; OCT: optical coherence tomography; PET: positron emission tomography; pTau: phosphorylated tau.An exciting perspective that should be considered is the possibility of diagnosing neurodegenerative diseases long before symptoms, even in childhood, by replicating the patients’ retina in vitro. This approach may be feasible due to a novel stem cell-related technology known as retinal organoids (Movio et al., 2023). Stem cells per se hold potential in treating retinal diseases (Bovi Dos Santos et al., 2024), but they could also be used to generate retinal organoids, reflecting the phenotypes and genotypes of patients. Retinal organoids are 3D miniature structures derived from induced pluripotent stem cells that mimic the organization and function of the retina. They are mainly used to study retinal development, disease mechanisms, and potential therapies for retinal disorders. However, these 3D structures may have other powerful applications, such as modeling neurodegenerative diseases. Because several reports have shown that the retina may represent a window to brain health and pathologies, modeling the 3D structure of the retina in miniature could become a convenient system to facilitate early diagnosis of neurodegenerative diseases (Movio et al., 2023). This proposal is supported by recent evidence that retinal organoids exhibit tauopathy and that the pattern of phosphorylated α-synuclein accumulation enables the identification of retinal cell types displaying early or late-stage α-synuclein aggregation in PD. Indeed, retinal organoids may represent a novel resource for drug testing, biomarker identification, and pathophysiological research. In light of these recent advances, retinal examinations emerge as pivotal indicators of brain health. Beyond their non-invasive nature, retinal exams offer a more cost-effective alternative to brain scans. Exploring these parallels could redefine our approach to understanding neurodegenerative disease propensity and progression, representing a new treatment era. Furthermore, the combination of retinal exams and plasma biomarkers could provide a comprehensive diagnostic approach that is highly efficient, straightforward, and cost-effective. Finally, modeling neurodegenerative diseases using retinal organoids holds promise for early diagnosis, besides testing personalized treatments to reverse or delay disease progression. These advancements underscore the transformative impact of retinal research on both diagnostics and therapeutic interventions in Alzheimer’s disease and related dementias. C-Editors: Zhao M, Liu WJ, Qiu Y; T-Editor: Jia Y