1. Introduction: The Long-Term Vision Gap
In the two decades since the advent of anti-vascular endothelial growth factor (anti-VEGF) therapy, the management of neovascular age-related macular degeneration (nAMD) has been defined by a striking paradox. While we have achieved unprecedented control over exudative activity, long-term visual outcomes continue to plateau and eventually decline. This “long-term vision gap” suggests that successfully drying the retina is merely the first step in a much longer battle against structural decay.
Real-world data and meta-analyses of 10-year outcomes reveal a sobering reality: despite treatment, patients experience an average loss of 8 letters from baseline over a decade. This progressive decline is rarely the result of uncontrolled exudation. The Age-Related Eye Disease Study 2 (AREDS2) identified macular atrophy (MA) and subretinal fibrosis as the principal causes of poor visual outcomes (20/200 or worse) in 60% and 40% of cases, respectively. As clinicians, we must recognize that managing these structural sequelae—specifically the transition from an exudative to an atrophic phenotype—is the critical challenge remaining in modern nAMD care.
2. Standardizing the Terminology: CAM Definitions and the nAMD Context
To facilitate precise clinical communication, the Classification of Atrophy Meetings (CAM) consortium established an OCT-based framework that classifies the spectrum of degeneration. While these stages are not always strictly linear, they provide a standardized roadmap based on histological correlates.
The CAM Spectrum of Atrophy
- Outer Retinal Atrophy (Intact RPE):
- iORA (Incomplete Outer Retinal Atrophy): Thinning of the outer retina with a disrupted but detectable ellipsoid zone (EZ) and an intact RPE. The external limiting membrane (ELM) remains intact.
- cORA (Complete Outer Retinal Atrophy): Characterized by the complete loss of the EZ and interdigitation zone (IZ), with severe outer retinal thinning and an intact RPE.
- RPE and Outer Retinal Atrophy:
- iRORA (Incomplete RPE and Outer Retinal Atrophy): Defined by an interrupted RPE band, subsidence of the inner nuclear layer (INL) and outer plexiform layer (OPL), and discontinuous choroidal hypertransmission (< 250 µm).
- cRORA (Complete RPE and Outer Retinal Atrophy): The “OCT correlate of geographic atrophy (GA).” Diagnosis requires four stringent criteria:
- A region of choroidal hypertransmission ≥ 250 µm.
- Corresponding RPE attenuation or disruption ≥ 250 µm.
- Definitive overlying photoreceptor loss (EZ, IZ, and ONL thinning).
- Absence of an RPE tear.
The Proposed Category of Fibro-atrophy
The neovascular environment poses a unique challenge: overlying subretinal hyperreflective material (SHRM), fibrosis, or the neovascular complex itself often attenuates or obscures choroidal hypertransmission. In these cases, the classic cRORA criteria are unassessable.
To address this, we use the term fibro-atrophy. While the term “atrosis” has been proposed, fibro-atrophy is preferred as it more descriptively conveys both the fibrotic/HRM component and the underlying atrophic loss. This category describes regions where B-scan OCT demonstrates clear RPE and outer retinal loss spanning ≥ 250 µm, but where the signature hypertransmission is masked by overlying fibrotic tissue. This distinction is vital for clinical trials and prognostic counseling, as the fibro-atrophic endpoint represents a complex failure of the RPE-choriocapillaris complex compared to “clean” cRORA.
3. The Epidemiological Burden: From Incidence to 10-Year Prevalence
The incidence of MA in treated nAMD is far higher than early anti-VEGF trials suggested. Data indicates that MA becomes a nearly universal feature of the disease given a long enough timeline.
- Short-term Cumulative Incidence: Pivotal trials provided the baseline. The CATT study reported an 18.3% incidence at 2 years, while the IVAN trial documented 24.4%. A meta-analysis of treatment-naïve eyes estimated a risk of 11% at year one, rising to 29% by year two.
- Long-term Prevalence: The burden increases relentlessly over time. In the VIEW 2 extension cohort, MA was present in 96% of eyes after a mean of 7 years. Other studies (6–10 years) report prevalence ranging from 78.3% to 89.5%. A 10-year retrospective analysis documented a cumulative incidence of approximately 80%.
- Progression Rates: Once established, MA enlargement varies from 0.20 mm²/year to 1.2 mm²/year. This variability is largely dictated by the underlying macular neovascularization (MNV) subtype and baseline atrophic burden.
4. Risk Stratification: Identifying the Atrophy-Prone Eye
A proactive management strategy requires identifying patients at high risk for structural failure at the time of diagnosis.
| Category | Key Biomarkers / Predictors |
| Demographics | Older age; presence of geographic atrophy or MA in the fellow eye. |
| Clinical Status | Poor baseline visual acuity (≤ 20/200); smoking history (most consistent modifiable risk). |
| MNV Morphology | Type 3 MNV (highest risk); large lesion area; presence of Reticular Pseudodrusen (RPD) (linked to 75% risk of atrophy within 3 years). |
| OCT Biomarkers | iRORA (20-fold increased risk of cRORA within 1 year); Nascent GA (subsidence of OPL/INL, hyporeflective wedge-shaped bands); Discrete Hyperreflective Foci (HRF); subfoveal choroidal thinning. |
| Genetics | ARMS2/HTRA1 locus variants; genetic risk scores for the complement pathway. |
| Fluid Dynamics | Intraretinal Fluid (IRF) (highly associated with atrophy and SHRM); significant fluctuations in central subfield thickness (CST). |
5. The MNV Subtype Divergence: Trophic Support vs. Aggressive Decay
The pathological trajectory of an eye is profoundly dictated by the MNV subtype.
Type 3 MNV: The Aggressive Phenotype
Type 3 MNV (retinal angiomatous proliferation) is the most atrophy-prone subtype. These lesions demonstrate rapid expansion (averaging 1.21 mm²/year), with nearly 50% of eyes developing new MA by month 24. The pathways to decay include RPE cell migration, direct mechanical damage from sub-RPE neovessels, and the collapse of serous pigment epithelial detachments (PEDs) following fluid reabsorption. This subtype is characterized by extensive choriocapillaris dysfunction and a failure of metabolic support to the outer retina.
Type 1 MNV: The “Protective Hypothesis”
In contrast, Type 1 MNV (located beneath the RPE) is associated with significantly slower atrophy growth (approx. 0.18 mm²/year). This has led to the hypothesis that Type 1 MNV may act as a “physiological repair mechanism.” This support is mediated by the neo-CC, a choriocapillaris-like vascular structure that develops on the anterior surface of the MNV. Extensive neo-CC coverage is associated with a significantly lower likelihood of cRORA development.
Academic Caveat: While Type 1 MNV may offer localized trophic benefits, it remains a pathologic complex. It is a persistent source of potential hemorrhage, SHRM, and eventual fibrotic scarring. The “protection” is relative, not absolute.
6. Multimodal Imaging: The Modern Diagnostic Toolkit
Monitoring MA requires a shift toward multimodal analysis, with structural OCT serving as the reference standard (92.4% sensitivity).
- Fundus Autofluorescence (FAF): Sensitive for RPE loss (hypoautofluorescence). However, it is often confounded by macular pigment (absorbing blue light), SHRM, and hemorrhage.
- Near-Infrared Reflectance (NIR): Essential for en face tracking. Hyperreflective zones demark atrophic areas, but reflectance is influenced by choroidal thickness and hyperpigmentation.
- En face OCT: Highly efficient for quantifying hypertransmission defects (hyperTDs). However, clinicians must be wary of “masking” effects where hyperpigmentation or calcified drusen hide underlying hypertransmission.
- OCTA: Best used as an exploratory tool for vascular context (e.g., neo-CC). It is currently limited by projection and motion artifacts, as well as signal attenuation from drusen and hemorrhage.
Functional Metrics
Clinicians must look beyond Best-Corrected Visual Acuity (BCVA), which often fails to capture the patient’s lived experience. Contrast sensitivity has a stronger correlation with the anatomical extent of MA than BCVA. Furthermore, reading speed is specifically disrupted by atrophy in the inner-right ETDRS subfield, which breaks the left-to-right flow of reading.
7. Clinical Management: The Anti-VEGF Paradox and Fluid Tolerability
A significant controversy persists regarding whether cumulative anti-VEGF exposure causes atrophy. While early trials (MARINA) showed higher MA incidence in treated vs. sham eyes, the relationship is likely confounded by disease severity.
Injection Frequency and the IRF Trigger
Eyes with aggressive disease—specifically those with persistent intraretinal fluid (IRF)—require more frequent injections. Because IRF is a primary risk factor for atrophy, the injection frequency is likely a marker of a severe phenotype rather than a causal agent of MA. Therefore, “drug holidays” in the presence of active IRF are not supported; IRF remains a mandatory treatment trigger.
The SRF Debate
Evidence from FLUID and HARBOR suggests that stable subretinal fluid (SRF) may be tolerated in selected eyes. SRF is often associated with lower rates of MA, likely reflecting the biology of Type 1 MNV and a more preserved RPE barrier. Conversely, fluctuations in fluid are highly detrimental, promoting neuronal stress and atrophy.
Newer Agents and Durability
Faricimab and Aflibercept 8 mg offer extended durability and more stable fluid control. While these agents reduce the injection burden, current evidence does not yet prove they reduce the long-term incidence or progression of MA.
8. Future Perspectives: Personalizing Care
The roadmap for nAMD management involves moving toward a personalized strategy that balances exudation control with structural preservation.
- Standardization: Validating the cRORA/fibro-atrophy framework for consistent cross-study endpoints.
- AI Integration: Moving toward automated quantification of SHRM burden, RPE loss, and hyperTD expansion to track progression objectively.
- Complement Inhibitors: While agents like pegcetacoplan slow growth in GA, their role in nAMD is uncertain. Safety remains the primary concern, as GA trials showed a dose-dependent signal for new-onset exudation (MNV) following complement inhibition.
9. Conclusion: Summary of Key Takeaways
- Acknowledge the Structural Ceiling: Exudation control is necessary but insufficient; macular atrophy and fibrosis are the true determinants of late-stage visual failure.
- MNV Subtype Informs Prognosis: Type 3 MNV requires aggressive monitoring for rapid decay, while Type 1 MNV may provide localized trophic support via the neo-CC.
- Prioritize Multimodal OCT: Use structural OCT B-scans to differentiate between cRORA and fibro-atrophy, as fundus-based imaging is frequently obscured by the neovascular environment.
Reference:
Zhuang, Xuenan, et al. “Macular atrophy in neovascular age-related macular degeneration: Prevalence, risk factors, imaging features, and clinical implications.” Surv. Ophthalmol., vol. 0, no. 0, 6 Aug. 2026, doi:10.1016/j.survophthal.2026.08.002.
