Choroidal neovascularization (CNV) is the growth of abnormal blood vessels from the choroid into or beneath the retina. These vessels leak fluid and blood, damaging photoreceptors and causing rapid central vision loss. CNV is the defining lesion of neovascular age-related macular degeneration (nAMD) and a complication of pathologic myopia, angioid streaks, and inflammatory eye disease. Untreated, it can destroy reading vision within weeks to months.
The anatomical fault line: why the choroid is vulnerable
The choroid is a vascular layer sandwiched between the sclera and the retina. It supplies the outer retina, including the photoreceptors, through a capillary bed called the choriocapillaris. Separating choroid from retina is Bruch's membrane, a five-layer extracellular matrix. That membrane is the structural gatekeeper of the macula.
In a healthy eye, Bruch's membrane blocks vessel invasion while permitting metabolic exchange. In CNV, this barrier fails. The triggers are cumulative: age-related lipid deposition, oxidative stress, chronic inflammation, and genetic susceptibility. Each degrades membrane integrity and shifts the local balance of pro-angiogenic and anti-angiogenic signals.
The decisive molecular switch is vascular endothelial growth factor (VEGF). Hypoxia and inflammatory cytokines upregulate VEGF, which binds receptors on choroidal endothelial cells. Those cells proliferate, degrade matrix with enzymes called matrix metalloproteinases, and migrate through defects in Bruch's membrane. This is not random leakage. It is organized, VEGF-driven angiogenesis.
A plain-language way to hold this: CNV is a wound-healing response that misfires inside the eye, building plumbing where none belongs.
How the abnormal vessels damage the macula
Once through Bruch's membrane, new vessels occupy one of two compartments. Subretinal CNV grows between the retina and retinal pigment epithelium (RPE). Sub-RPE CNV, also called type 1, grows between the RPE and Bruch's membrane. A third pattern, type 3 or retinal angiomatous proliferation, begins within the retina and anastomoses with choroidal vessels.
The damage is mechanical and biochemical. Leaky neovessels exude plasma and lipids, producing subretinal fluid and hard exudates. Hemorrhage can occur beneath the RPE or retina. The resulting fibrin scaffold contracts, forming a fibrotic disc that permanently replaces photoreceptors. Fluid alone disrupts photoreceptor alignment; fibrosis destroys them outright.
Location determines prognosis more than size. A lesion outside the foveal avascular zone may spare central acuity. A lesion under the fovea threatens it immediately. This is why clinicians describe CNV by its relation to the fovea, not just its area.
Type matters too. Type 1 lesions often remain occult and progress slowly. Type 2 lesions are more often classic on angiography and tend to declare themselves with sudden visual distortion. Type 3 lesions carry a high risk of bilateral involvement.
Quantified context: who develops CNV and how often
Age-related macular degeneration is the leading cause of irreversible central vision loss in older adults in high-income countries. Neovascular disease accounts for a minority of all AMD but the majority of severe vision loss from it. Roughly 10 to 15 percent of AMD patients develop neovascular complications over their lifetime.
Progression risk is measurable. Fellow eyes of patients with CNV in one eye face a substantial annual risk of conversion, estimated in the range of 10 to 20 percent per year in untreated historical cohorts. That figure is central to surveillance strategy.
Pathologic myopia is the second major driver. CNV complicates a meaningful minority of highly myopic eyes, and the risk rises with axial length and with lacquer cracks in the posterior pole. Angioid streaks, seen in pseudoxanthoma elasticum and related disorders, carry a high lifetime rate of CNV. Inflammatory conditions such as punctate inner choroidopathy are less common but disproportionately affect younger women.
Prevalence figures shift with population aging and with diagnostic coding changes. Treat incidence estimates in older sources as historical context, not current measurement.
Edge cases that change management
Not every CNV behaves alike. The pattern and the underlying disease alter both urgency and prognosis.
Type 1 CNV may persist for years with minimal exudation and can be observed in selected cases; not every lesion requires immediate injection. Type 3 lesions are often bilateral over time and can present with retinal hemorrhage before obvious fluid. Subretinal hemorrhage can be massive and vision-threatening within days, requiring urgent retinal evaluation. Peripapillary CNV threatens the optic nerve and may produce field loss before acuity drops. CNV in pregnancy can be aggressive; hormonal and hemodynamic changes may accelerate exudation. Recurrence after successful treatment is common and does not imply treatment failure. Anti-VEGF therapy does not eliminate the underlying neovascular stimulus; it suppresses activity.
The counterintuitive professional caveat: a quiet, dry-looking scar is not proof the disease is inactive. Fibrovascular tissue can persist beneath the retina with no fluid, and treating an inactive lesion adds risk without benefit. Clinicians therefore treat activity, not the mere presence of a vessel.
Misconceptions worth correcting
The first misconception is that CNV is purely a leakage problem. Leakage is the visible symptom; angiogenesis is the disease process. This distinction explains why anti-VEGF drugs, which block vessel growth signaling, outperform therapies aimed only at drying tissue.
The second is that OCT alone defines treatment need. Optical coherence tomography is the workhorse for detecting fluid, but fluorescein angiography and indocyanine green angiography still characterize lesion type and extent in ambiguous cases. A dry OCT can coexist with active neovascularization elsewhere.
The third is that anti-VEGF injections restore lost vision in every patient. They often stabilize and sometimes improve acuity, but photoreceptors destroyed by fibrosis do not regenerate. Delay is the strongest modifiable predictor of poor outcome.
The fourth is that a normal eye exam excludes CNV. Early lesions can hide beneath the RPE and produce subtle distortion that only a home Amsler grid or OCT detects. Self-monitoring between visits catches conversion earlier than annual examinations.
Practical actions: detection, thresholds, and escalation
Detection rests on three tools: symptoms, home monitoring, and imaging. Metamorphopsia, a new central scotoma, or sudden blurred vision in one eye warrants prompt dilated retinal examination. An Amsler grid checked daily per eye is cheap and sensitive to distortion.
Imaging thresholds guide decisions. OCT showing intraretinal or subretinal fluid in a patient with known AMD risk is a red flag for active CNV and prompts angiography. Fluorescein angiography showing leakage confirms the diagnosis. Indocyanine green angiography helps classify type and identify feeder vessels in select cases.
Sudden new distortion or central blind spot: seek retinal evaluation within days, not weeks. Subretinal hemorrhage: same-day or next-day evaluation. Sudden painless vision loss in a previously stable AMD patient: treat as possible conversion. Fellow eye conversion: escalate surveillance frequency. No improvement after repeated injections: reassess diagnosis and lesion type.
Reconciling thresholds: conversion risk of roughly 10 to 20 percent annually in fellow eyes justifies frequent self-monitoring, while a single OCT with fluid justifies angiography and treatment discussion. These are different thresholds serving different decisions. Do not conflate population risk with individual diagnostic criteria.
Treatment is specialist territory. Anti-VEGF agents are injected into the vitreous on schedules determined by lesion activity, not by a fixed public algorithm. Dosing intervals, agent selection, and switching decisions belong to the treating retina specialist and are outside the scope of this explainer. What patients and referring clinicians control is time to diagnosis.
Limits of current knowledge
Mechanistic understanding is strong; prediction is not. We know VEGF drives neovascularization, yet patients with identical risk profiles diverge sharply in outcome. Genetic variants in complement pathway genes associate with AMD, but no polygenic score yet guides CNV surveillance in routine practice.
Treatment durability remains unsolved. Some eyes require frequent injections for years; others quiet after a few. The biological reasons for this heterogeneity are incompletely characterized. Fibrosis prevention has no proven therapy.
Imaging advances, including OCT angiography, detect flow without dye and improve lesion characterization. Their role in routine treatment decisions is still being defined. Long-term data on progression under sustained therapy continue to accumulate.
What can be concluded with confidence: CNV is a VEGF-mediated invasion across Bruch's membrane, its location and type determine visual threat, early detection changes outcomes, and treatment suppresses activity without erasing the underlying lesion. Everything else is an open question worth answering carefully.
Timeline: from trigger to clinical consequence
| Stage | Approximate time course | What happens |
|---|---|---|
| Barrier degradation | Years to decades | Lipid and oxidative injury weaken Bruch's membrane |
| VEGF upregulation | Months to years before symptoms | Angiogenic signaling rises; vessels begin to migrate |
| Breach of Bruch's membrane | Weeks to months | Neovessels enter sub-RPE or subretinal space |
| Leakage and fluid | Days to weeks | Exudation produces subretinal fluid and distortion |
| Hemorrhage or fibrosis | Days to months | Vision loss accelerates; scarring may begin |
| Treatment response | Days to weeks after injection | Fluid often resolves; acuity variable |
| Recurrence | Months to years | Activity returns without ongoing suppression |
Comparison: CNV subtypes and their clinical character
| Subtype | Location | Typical angiographic appearance | Clinical behavior |
|---|---|---|---|
| Type 1 (sub-RPE) | Beneath RPE | Occult, late leakage | Slower, often less exudative |
| Type 2 (subretinal) | Between retina and RPE | Classic, well-defined | Sudden distortion, distinct leakage |
| Type 3 (RAP) | Intraretinal origin | Anastomotic complex | Bilateral risk, hemorrhage-prone |
| Peripapillary | Near optic disc | Localized leakage | Field loss risk |
| Myopic CNV | Posterior pole in high myopia | Small, well-defined | Younger patients, recurrence common |
Type classification matters because it predicts behavior, not because it dictates a single pathway. Many lesions are mixed.
The practical takeaway for anyone reading this with concern about their own vision: this article describes disease mechanisms, not individual diagnosis or treatment. Sudden visual distortion, a new blind spot, or painless vision loss deserves a dilated eye examination promptly. The mechanisms are well understood; the outcome depends heavily on how quickly the lesion is found.