Neovascular Morphology and Persistent Disease Activity Among Patients With NV AMD

Neovascular Age-Related Macular Degeneration (NV AMD) remains the leading cause of vision loss among people over 65. Intravitreal injections with drugs that block VEGF, a major protein mediator of angiogenesis and vascular leakage, have revolutionized treatment of NV AMD. However, less than 40% of treated patients have clinically significant improvement in vision. Further, in spite of continuous monthly anti-VEGF therapy, up to 40-50% of patients demonstrate sustained persistent disease activity (PDA), defined as (1) unresolved intraretinal, subretinal, or sub-retinal pigment epithelium fluid; (2) progressive lesion enlargement and fibrosis; and/or (3) persistent or new hemorrhage, assessed after either loading dose therapy or after sustained treatment with anti-VEGF. Since affected patients are at increased risk for long-term vision loss, PDA remains a vital clinical unmet need.

We are interested in the relationship between NV lesion morphology and response to therapy. Specifically, we hypothesize that specific NV morphologic subtypes are more frequently associated with PDA, based on preliminary retrospective analyses of indocyanine green (ICG) imaging data from NV AMD patients in our Duke Medical Retina practice. We have observed that eyes with Capillary pattern, seen as a discrete homogenous focus of microvessels, are highly responsive to anti-VEGF therapy and rarely exhibit PDA (<20% of cases). In contrast, eyes with Arteriolar pattern (large-caliber feeding artery, many branching arterioles, and minimal capillary component) and eyes with polypoidal choroidal vasculopathy (variably sized and numbered, discrete saccular dilations of choroidal vasculature), demonstrate PDA in up to 70% of cases. A third subtype, choroidal leak syndrome, visible as choroidal hyperpermeability and leakage, manifest PDA in over 60% of cases. These data suggest that complex NV lesion morphology is the primary cause of PDA, and that anti-VEGF therapy alone is insufficient for these patients. However, the relative frequency of these subtypes and the association of PDA and NV lesion morphology, in a treatment-naive population free of selection bias, are unknown.

In this study, we will determine the relative frequency of NV subtypes in two groups: (i) a representative, treatment-naïve NV AMD patient population, and (ii) a population of patients who develop recurrent NVAMD activity while off treatment and assess the frequency of PDA according to specific NV morphologic subtypes. This information will clarify the scope of the PDA problem, and will identify patients with PDA who may benefit from additional therapeutic strategies.