As retina specialists, we have become remarkably adept at managing the late-stage complications of age-related macular degeneration (AMD). The anti-VEGF era has undeniably transformed our ability to protect patients from the legal blindness associated with neovascular AMD (nAMD). However, a sobering longitudinal reality remains we are treating the sequelae, not the pathogenesis. High-level evidence, including the work of Khanani et al. (2024), underscores that even with optimal suppression of choroidal neovascularization (CNV), vision loss in our patients remains progressive. We must confront the fact that CNV is a complication of AMD, much like a myocardial infarction is a complication of atherosclerosis, but it is not the disease itself.
The recent introduction of anti-complement therapies for geographic atrophy (GA) has provided another piece of the puzzle, yet it remains incomplete. Phase 3 trials for anti-C3 (pegcetacoplan) and anti-C5 (avacincaptad pegol) agents demonstrated only a “nominal” slowing of GA growth rates. The fact that two different molecules targeting disparate steps in the complement cascade yielded nearly identical, partial responses suggests that while the complement pathway is a participant, it is likely not the primary driver. The data compels us to move beyond these narrow confines and seek a unifying hypothesis—one that explains why this disease is inextricably linked to aging and why it remains so uniquely localized to the macula.
Clinical Realities
- The Anti-VEGF Plateau: Suppression of exudative complications does not halt the underlying degenerative process. Vision loss remains a steady, progressive inevitability.
- The Complement Ceiling: Anti-complement therapies offer modest gains, indicating that complement overactivation is a downstream event rather than the primordial insult.
- The Unmet Mandate: We require a hypothesis that integrates clinical observations, genetic risk, and imaging biomarkers into a cohesive model of end-organ vascular failure.
Ischemia and Hypoxia: The Initial Insults
A perfusion-based model shifts our clinical focus from primary neurodegeneration to chronic vascular insufficiency. In this framework, we define hypoperfusion not as an acute event, but as a chronic, low-grade reduction in blood flow—an “ischemia-lite” that persists for decades. This state induces chronic hypoxia at the level of the neurovascular unit.
Mechanism of Action: Ischemia vs. Hypoxia
- Ischemia: A reduction in tissue-level blood flow, resulting in compromised delivery of oxygen and glucose, alongside impaired clearance of metabolic waste.
- Hypoxia: A cellular state of oxygen deficiency. In the outer retina, ischemia is the primary driver of hypoxia, which subsequently impairs RPE metabolism, lipid handling, and mitochondrial efficiency.
The Kurihara mouse model provides the mechanistic blueprint for this transition. By genetically triggering hypoxia specifically in the RPE, researchers observed metabolic shifts that mirror human AMD: lipid droplet accumulation, progressive thickening of Bruch’s membrane, and eventual photoreceptor atrophy. This suggests that metabolic stress alone, driven by oxygen deprivation, is sufficient to trigger the AMD phenotype.
The “chicken or the egg” debate regarding cell death in GA has been largely settled by AI-driven longitudinal analysis. Schmidt-Erfurth et al. (2025) utilized machine learning to analyze the Ellipsoid Zone (EZ) and RPE in GA eyes. Their data demonstrates that photoreceptor degeneration (EZ loss) occurs first, driving the expansion of GA lesions, with RPE loss following as a secondary consequence. This finding is highly consistent with a perfusion model: the photoreceptors, being the cells furthest from the choriocapillaris blood supply, are the first to succumb as oxygen diffusion becomes limited.
The Macula’s “Achilles Heel”: Unique Vascular Anatomy
The macula’s susceptibility to AMD is a direct consequence of its “exceptional” metabolic and anatomic profile. The outer retina possesses the highest metabolic rate in the human body; photoreceptors consume more oxygen per gram than any other cell type. To support this, the choroid receives the highest perfusion per unit mass of any tissue. However, the macular choriocapillaris is uniquely vulnerable due to its planar, non-anastomosing “end-artery” architecture.
Anatomic Vulnerabilities
| Feature | Retinal Circulation | Choriocapillaris Circulation (Macula) |
| Pressure System | High Pressure | Low Pressure |
| Vessel Architecture | Anastomosing | Planar, Non-anastomosing “End-Arteries” |
| Autoregulation | Highly Autoregulated | Highly Perfusion-Pressure Dependent |
| Metabolic Demand | High | Exceptional (Highest in the body) |
| Vulnerability | Resistant to systemic flow drops | Susceptible to “Watershed Zone” hypoxia |
In the macula, these end-arteries create “watershed zones”—regions that rely entirely on perfusion pressure and diffusion. Any drop in systemic perfusion or increase in resistance renders these zones hypoxic. This anatomic vulnerability is not unique to the eye; it mirrors other “perfusion-deficient cerebrovascular neurodegenerations.” The substantia nigra and the hippocampus share similar end-artery structures, making them susceptible to the chronic hypoperfusion observed in Parkinson’s and Alzheimer’s diseases. AMD may well be the ocular manifestation of a system-wide cerebral small-vessel vulnerability.
Macula-Wide Evidence of Vascular Loss
Recent evidence compels us to recognize AMD not as a focal disease limited to lesions, but as a global macular vascular dysfunction. Histological evidence from Ramrattan, Curcio, and Mullins has identified choriocapillaris dropout and “ghost vessels” even in the earliest stages of dry AMD.
The Three Pillars of Macula-Wide Evidence
- Global Flow Reduction: Linton et al. (2025) used Laser Speckle Flowgraphy to demonstrate a 33% reduction in choroidal blood flow across all stages of nonexudative AMD. Crucially, these observations could not be accounted for by blood pressure or intraocular pressure, suggesting an intrinsic vascular pathology.
- Choriocapillaris Flow Deficits (CCFD): Swept-source OCTA reveals that CCFD throughout the entire macular region—even in areas remote from GA—predicts future lesion growth.
- Metabolic Stress Markers: AI-based measurements show that the mean thickness of the EZ layer anywhere in the macula is qualitatively predictive of growth rates, suggesting a state of macula-wide metabolic “simmering.”
Reinterpreting Genetic Risk: The Vascular Lens
When viewed through a vascular lens, the dominant genetic risk factors for AMD transition from retinal markers to vascular drivers.
The Complement/Lipid Axis (CFH)
Complement Factor H (CFH) is a fluid-phase immune regulator in the blood. A critical finding by Demirs et al. (2021) shows that 99% of complement activity occurs in the choroid and choriocapillaris, not the retina. This settles the “where to treat” debate: the pathology is vascular. Furthermore, the “Oil Spill” hypothesis (Curcio et al.) links AMD to systemic atherosclerosis. While esterified cholesterol in Bruch’s membrane is RPE-derived, the source of unesterified cholesterol is likely the systemic circulation. Polymorphisms in CFH may fail to prevent the inflammatory cascade triggered by these retained systemic lipoproteins, mirroring atherogenesis in the vessel wall.
The Small Vessel/Structural Axis (HTRA1/ARMS2)
HTRA1 is famously associated with CARASIL, a cerebral small vessel disease. In the eye, the HTRA1/ARMS2 gene product localizes to the intercapillary pillars of the choriocapillaris. These pillars maintain the stromal integrity and precise microanatomical alignment between the choriocapillaris and Bruch’s membrane. If this support structure fails, the multipolar vascular geometry becomes distorted, creating diffusion-limited regions that starve the outer retina of oxygen.
A Tale of Two Phenotypes: Soft Drusen vs. SDD
We must move toward a binary classification of AMD pathways. Subretinal drusenoid deposits (SDD), or reticular pseudodrusen, represent a distinct ischemic pathway.
Phenotypic Differentiation
| Feature | Complement-Associated AMD (Soft Drusen) | Non-Complement Associated AMD (SDD) |
| Genetic Driver | CFH (Chromosome 1) | HTRA1/ARMS2 (Chromosome 10) |
| Systemic Link | Metabolic Syndrome / Atherosclerosis | High-Risk Vascular Disease (HRVD) |
| Mechanism | Lipid/Complement Axis (“Oil Spill”) | Small vessel failure / Systemic hypoperfusion |
| Association | Systemic Lipid Retention | Carotid Stenosis / Myocardial Infarction |
The link between SDD and High-Risk Vascular Disease (HRVD)—including carotid stenosis—is the “smoking gun” for the perfusion model. Basloe and Smith identified a “gravity effect” in HRVD-associated SDD, where deposits are significantly more prevalent in the superior macula. This suggests that in patients with compromised systemic flow, gravity further reduces perfusion to the superior fundus, tipping it into ischemia. In contrast, HTRA1-driven SDD shows a global distribution, consistent with an intrinsic genetic defect in stromal support.
AMD as a Cardiometabolic Syndrome
Aging is, in many respects, “inevitable atherosclerosis” driven by the accumulation of progerin in senescent cells. AMD should be viewed as an end-organ manifestation of cardiometabolic syndrome, which now affects 60% of individuals over age 60.
Drug Repurposing Watchlist
- Statins: Simvastatin has demonstrated a 12-fold reduction in AMD progression risk for patients homozygous for the CFH Y402H allele.
- PCSK9 Inhibitors: Large database studies show reduced AMD risk; because these monoclonal antibodies are too large to cross the Blood-Retinal Barrier (BRB) and work exclusively in the liver, their protection is purely systemic.
- Metformin: Associated with reduced GA and nAMD incidence, potentially acting via the gut-retina axis.
- SGLT2 Inhibitors: Emerging data suggests protective effects via hyperglucosuria and improved metabolic homeostasis.
- GLP-1RAs: These agents provide systemic cardiovascular protection; retrospective data shows a significantly lower hazard risk for AMD compared to statins or insulin.
Conclusion: Moving Toward a Unifying Hypothesis
The evidence is clear: it is time to redefine AMD as an end-organ vascular disease. Whether the driver is the complement-lipid axis of soft drusen or the small-vessel/systemic hypoperfusion axis of SDD, the final common pathway is a failure of the neurovascular unit to maintain oxygen exchange.
As specialists, we have a clinical mandate to look beyond the macula. The management of AMD must evolve into a multidisciplinary effort, co-managing our patients’ systemic vascular health with cardiology and neurology. By addressing blood flow and metabolic homeostasis, we may finally move from treating the complications of AMD to halting the disease itself.
Clinical Pearl AMD is a state of chronic, low-grade hypoperfusion. Photoreceptor loss precedes RPE loss because the furthest cells succumb first to oxygen debt. Treatment must target the vascular supply, where 99% of complement activity resides.
Reference Guide (Abridged)
- Demirs et al. (2021): Localization of 99% of ocular complement activity to the choroid/choriocapillaris.
- Linton et al. (2025): 33% choroidal flow reduction in AMD, independent of BP and IOP.
- Basloe & Smith (2024): Superior macular “gravity effect” in HRVD-associated SDD ischemia.
- Schmidt-Erfurth et al. (2025): AI analysis confirming EZ loss precedes RPE loss in GA progression.
- Curcio et al. (2011): The “Oil Spill” hypothesis and the systemic source of unesterified cholesterol.
- Hayreh (1990): Foundational anatomy of macular end-arteries and watershed zones.
- Kortvely et al. (2016): Localization of HTRA1 to the intercapillary pillars.
