Age-Related Macular Degeneration

Sodium iodate has become the most widely used chemical model for acute RPE ablation because its toxicity is dose-titratable, reproducible, and mechanistically selective for the RPE. However, the utility of the model depends entirely on the availability of validated functional endpoints to confirm that the intended RPE ablation has occurred and to quantify the downstream visual deficit. Histological confirmation (RPE flat mounts, OCT layer thickness) is terminal or technically demanding; ERG requires anaesthesia, trained operators, and is difficult to perform repeatedly in longitudinal studies. Without a reliable functional readout at each time point, researchers cannot determine the optimal window for therapeutic intervention or distinguish partial from complete functional rescue.

In AMD, rod photoreceptors are selectively vulnerable because they are concentrated at the rod-rich parafoveal ring and have the highest metabolic demands from the RPE. RPE ablation by NaIO3 therefore impairs scotopic (rod-mediated) function before photopic acuity deteriorates, mirroring the clinical observation that delayed dark adaptation precedes visual acuity loss in early AMD. An optomotor-based assay capable of distinguishing scotopic from photopic thresholds is therefore essential for faithfully recapitulating the AMD functional phenotype in the NaIO3 model. For the ocular toxicity context of this model, see Ocular and CNS Toxicity Models. For the retinal degeneration broader context, see also Retinal Degeneration.

Complement dysregulation is one of the strongest genetic and mechanistic contributors to dry AMD: variants in CFH, C3, and related regulators promote sustained complement imbalance that drives RPE injury and extracellular deposit formation. In parallel, microglia – the resident immune cells of the retina – play a context-dependent, dual role in AMD pathogenesis. Early-stage microglia provide neuroprotective support, clearing photoreceptor debris and trophic factor support for the RPE, while chronically activated late-stage microglia become neurotoxic and contribute to RPE and photoreceptor loss. Understanding this duality is critical for evaluating therapeutic microglial depletion strategies, which risk eliminating protective microglial function.

An additional mechanistic link connects AMD to Alzheimer’s disease pathology: amyloid-beta (Aβ) deposits have been identified within drusen in AMD patients, and impaired amyloid clearance from the aging retina may exacerbate RPE stress and complement activation. This molecular overlap positions the AMD retina as a window into systemic amyloid burden and neuroinflammatory processes shared with Alzheimer’s disease. For the broader neuroinflammatory perspective, see Neuroinflammation and Autoimmune CNS Disease and the neuroinflammation cluster. For the aging context, see Systemic Aging and CNS Decline and the aging cluster.

The fundamental challenge is that most endpoints used to assess neuroinflammatory state (immunohistochemistry for microglial markers, complement protein levels) are terminal or require tissue collection, precluding longitudinal tracking within the same animal. ERG captures aggregate photoreceptor function but does not directly reflect the microglial or complement contribution to the visual phenotype. A non-invasive functional endpoint that can be repeated at multiple time points across the treatment course is essential for dissecting the temporal relationship between neuroinflammatory intervention and visual outcome.

In human AMD, delayed rod-mediated dark adaptation (RMDA) – reflecting slowed retinoid re-supply across the RPE and Bruch’s membrane – is the earliest detectable functional biomarker of incident early AMD, preceding the appearance of structural changes on OCT (Owsley et al. (2016). Delayed Rod-Mediated Dark Adaptation Is a Functional Biomarker for Incident Early AMD. Ophthalmology.). Rod loss occurs predominantly at the parafoveal ring – the region of highest rod density – and is the largest aging effect in the outer retina, occurring well before cone-mediated central visual acuity declines (Curcio et al. (2024). Age-Related Macular Degeneration, a Mathematically Tractable Disease. Invest. Ophthalmol. Vis. Sci.). This rod-first vulnerability means that preclinical AMD models assessed only with photopic endpoints miss the earliest and most sensitive functional signal of disease onset or therapeutic rescue.

Translating the RMDA concept to preclinical models requires an assay that can separately quantify rod (scotopic) and cone (photopic) visual function in the same animal at multiple time points. Conventional ERG approaches can do this but require anaesthesia, trained operators, and impose procedural stress that limits longitudinal frequency. The optomotor reflex platform with scotopic extension provides an automated, anaesthesia-free alternative that achieves the same photoreceptor subclass distinction in freely moving awake mice.

The first approved therapies for geographic atrophy – pegcetacoplan (Syfovre, a C3 inhibitor) and avacincaptad pegol (Izervay, a C5 inhibitor) – slow GA lesion growth by 14-21% at 12 months in clinical trials (Comparative Study of Pegacetacoplan versus Avacincaptad Pegol. Clin Ophthalmol. 2025.), but their effects on functional visual endpoints are modest. Translating these and more potent complement-targeting strategies requires preclinical models with measurable functional visual deficits that are sensitive enough to detect partial rescue. Similarly, cell therapy approaches – RPE stem cell transplants, iPSC-derived RPE monolayers, co-transplantation of photoreceptor precursors – require functional readouts that go beyond structural engraftment to demonstrate that transplanted cells are restoring visual circuit function.

A key obstacle is that the NaIO3 model, the most widely used AMD surrogate for cell therapy and neuroprotection studies, lacked a validated functional characterisation until Carido et al. (2014) established the OptoDrum-based timeline. Without this baseline, researchers could not confirm that the model had produced the intended deficit, nor demonstrate that an intervention had achieved functional rescue above the degeneration trajectory. The optomotor-based baseline makes both claims possible in a non-invasive, longitudinal assay. For cell therapy and vision restoration studies more broadly, see Maintaining and Restoring Vision and the blindness cluster.

A common design question in AMD research is whether a model or phenotype reflects AMD-specific outer retinal degeneration or a genetically-driven inherited retinal dystrophy (IRD). In AMD, degeneration begins in the parafoveal rod-rich zone and progresses centripetally as geographic atrophy, with cone-mediated foveal acuity preserved until late disease. In IRDs (retinitis pigmentosa, Stargardt disease, Best disease), the genetic defect typically determines which photoreceptor class is initially affected and the spatial pattern of degeneration. Functional distinction matters because therapeutic strategies differ: complement inhibitors, anti-VEGF, and RPE cell replacement are AMD-specific approaches, while gene therapy and antisense oligonucleotides target specific inherited mutations.

OptoDrum-based longitudinal profiling can contribute to this distinction: the relative decline rate of scotopic versus photopic acuity, the age of onset of functional deficit, and the dose-response to RPE-selective toxins (NaIO3) versus genetic models provide a functional phenotypic signature that supplements structural imaging. For the comprehensive view of inherited retinal disease models and their genetic foundations, see Retinal Degeneration and Inherited Retinal Disease and the retinal dystrophy cluster, which covers genetically-defined IRDs in mechanistic contrast to the age-related degeneration described here.