CNV Detection on OCTA
Type 1 vs Type 2 CNV flow signatures. Treatment-naive vs treated appearance. How OCTA compares to fluorescein angiography for sensitivity and specificity.
CNV Detection on OCTA
Choroidal neovascularization is the single finding where OCTA delivers the most dramatic advantage over structural OCT. A B-scan tells you there is fluid — OCTA tells you there is flow. That distinction changes management: fluid can be traced to multiple sources, but active CNV flow is unambiguous. Understanding the OCTA appearance of different CNV subtypes is fundamental to using OCTA for AMD monitoring.
OCTA detects CNV by identifying decorrelation signal in the outer retina and choriocapillaris slabs — areas that should be avascular in health. Any flow in these slabs requires explanation. Not all outer retina flow is CNV (projection artifact must be excluded), but all active CNV produces detectable flow.
Type 1 CNV: Sub-RPE Flow
Type 1 CNV (also called occult CNV) grows beneath the RPE — within Bruch's membrane or in the sub-RPE space. On structural OCT, it appears as a flat or irregular RPE elevation, sometimes with mild subretinal fluid, often with no fluid at all. It is notoriously difficult to detect on B-scan alone.
On OCTA, Type 1 CNV appears as flow signal beneath the RPE boundary, visible on the outer retina or choriocapillaris slab. The characteristic pattern is a dense, irregular vascular network — sometimes described as a "sea fan" or "medusa head" shape — confined to the sub-RPE space. The flow is often lower velocity and the vessel caliber smaller than Type 2, making the network appear less bright and more irregular.
In AMD, Type 1 CNV is often quiescent — present for months to years with minimal exudative activity. The OCTA finding of sub-RPE flow does not automatically indicate treatment, but it establishes the presence of neovascularization, changes your monitoring interval, and defines the lesion that future fluid would be arising from.
| Feature | Type 1 CNV (Sub-RPE) |
|---|---|
| Location | Beneath RPE (sub-RPE space / Bruch's membrane) |
| B-scan appearance | Flat or irregular RPE elevation; ± mild SRF; often no fluid |
| OCTA outer retina slab | Dense, irregular vascular network (sea fan / medusa pattern) |
| Flow characteristics | Lower velocity; small caliber vessels; irregular branching |
| Natural history | Often quiescent; may activate with fluid if Bruch's integrity lost |
| Treatment trigger | Fluid development on B-scan, not OCTA flow detection alone |
Type 2 CNV: Sub-Retinal Neovascularization
Type 2 CNV (classic CNV) breaks through the RPE and grows into the subretinal space above it. On B-scan, it appears as a hyperreflective membrane above the RPE with associated subretinal fluid and intraretinal fluid — the classic exudative AMD picture that demands urgent management.
On OCTA, Type 2 CNV has a dramatically different appearance from Type 1. The neovascular membrane sits above the RPE, making it visible on the outer retina slab as a bright, well-defined vascular network. The classic description is a "flower" or "glomerulus" pattern — a central feeding vessel with peripheral branching loops — surrounded by a relatively avascular halo. This bright, elevated appearance reflects the higher flow velocity and larger caliber vessels compared to the more tortuous Type 1 network.
In active disease, the flower pattern is dense and bright. After anti-VEGF treatment, vessel density decreases and the pattern becomes progressively less distinct — a key objective monitoring marker for treatment response.
Mixed-Type CNV and Treatment Response
Mixed-type CNV (combined Type 1 and Type 2, or retinal angiomatous proliferation) accounts for a significant portion of neovascular AMD. On OCTA, mixed lesions show flow in both sub-RPE and subretinal compartments with complex patterns that do not fit cleanly into either template. The key clinical skill is recognizing the presence of any neovascular flow and tracking its behavior over time.
Treatment-naive vs. treated CNV on OCTA:
| State | OCTA Appearance | B-scan Correlation |
|---|---|---|
| Treatment-naive (active) | Dense, bright network; Type 2 shows glomerulus; Type 1 shows full-extent sea fan | Active IRF and/or SRF present |
| After anti-VEGF (treated) | Decreased vessel density; outer branches regress first; central core persists | Fluid resolved or reduced |
| Quiescent (long-term treated) | Ghost vessels — faint network outline, very low density; irregular hyporeflective area | RPE changes persist; no active fluid |
| Reactivating CNV | Increasing vessel density in previously quiescent lesion; new peripheral branching | New fluid appearing — injection trigger |
OCTA vs Fluorescein Angiography for CNV
For detection, OCTA is non-inferior to FA and adds depth-resolved anatomical information FA cannot provide. For characterizing leakage pattern and guiding initial treatment decisions in complex lesions, FA retains value. The practical question for optometrists is not which replaces which — it is understanding where OCTA changes your clinical workflow.
| Parameter | OCTA | Fluorescein Angiography |
|---|---|---|
| CNV detection sensitivity | ~82–90% overall (higher for Type 2; lower for quiescent Type 1) | ~70–85% (occult CNV often poorly characterized on FA) |
| Invasiveness | None — no dye, no systemic risk | IV dye injection; rare anaphylaxis risk |
| Depth localization | Yes — sub-RPE vs. subretinal differentiation | No — cannot distinguish Type 1 from Type 2 by depth |
| Leakage detection | No — flow only, not dye leakage | Yes — dye extravasation shows active leakage |
| Monitoring frequency | Every visit — repeatable, non-invasive | Not practical for routine monitoring visits |
| Treatment response quantification | Excellent — vessel density changes are objective and reproducible | Qualitative; requires repeat dye injection each assessment |
For optometrists managing AMD and co-managing with retina, OCTA occupies the primary monitoring role at every visit — non-invasive, giving objective CNV activity data. FA remains the retina specialist's tool for initial treatment planning in complex cases and for confirming suspected CNV when OCTA findings are equivocal.
Sensitivity, Specificity, and Clinical Thresholds
OCTA CNV detection performance varies significantly by subtype. Type 2 (classic) CNV achieves ~90–95% sensitivity because the network sits above the RPE, generates strong decorrelation signal, and has large caliber vessels. Type 1 (occult) CNV achieves ~70–85% sensitivity because the network is hidden beneath the RPE, vessel caliber is smaller, and flow velocity is lower.
Practical clinical thresholds:
- Negative OCTA does not exclude CNV: If B-scan shows active fluid and clinical suspicion is high, a negative outer retina slab does not rule out occult CNV. Refer for FA or specialist-performed imaging in these cases.
- Outer retina flow without fluid: Can represent early, quiescent Type 1 CNV. Requires closer monitoring — not necessarily immediate treatment.
- Signal strength threshold: Reliable CNV detection requires signal strength ≥7. Below this, outer retina noise may mimic flow signal.
- Projection artifact mimics CNV: If outer retina "flow" exactly follows the topography of the superficial vascular arcades, it is projection artifact. Apply projection removal and cross-reference with B-scan before concluding CNV is present.
- Type 1 CNV: sub-RPE, sea fan pattern, often quiescent — monitoring finding without fluid
- Type 2 CNV: subretinal, glomerulus/flower pattern, exudative — treatment-urgent finding
- Treated CNV: decreasing vessel density; quiescent = ghost vessels, no fluid
- OCTA sensitivity ~82–90%; negative scan does not exclude occult CNV in high-suspicion cases
- Always exclude projection artifact before diagnosing outer retina flow as CNV
Sign up to track your progress and access all lessons.
Create Free Account