To correlate features of indocyanine green (ICG) videoangiography with stereo film fluorescein angiography (FA) in the analysis of age-related macular degeneration (AMD).

A retrospective study with concurrent comparison of two methods of assessment on one set of patients with AMD undergoing sequential FA-ICG videoangiography.

One hundred four patients with AMD who had undergone immediately sequential FA-ICG videoangiography were selected in a consecutive fashion from the photographic files of the Casey Eye Institute. Three interpreters independently graded in an unbiased fashion the FA features of AMD. These were compared with ICG videoangiography features of hypocyanescence (ICG videoangiography hypofluorescence), moderate or intense hypercyanescence (ICG videoangiography hyperfluorescence), or absence of fluorescence over background (ICG videoangiography isocyanescence). We also assessed ICG videoangiography features in the opposite eye of those with choroidal neovascularization (CNV) by FA in one eye by examining an additional 96 FA-ICG videoangiography scans.

Outcomes were the ICG videoangiography characteristics of classic and occult CNV, serous pigment epithelial detachments (SPEDs) both with and without CNV, macular hemorrhage, and the comparative size of these features.

There were 25 eyes that had significant macular hemorrhage by FA. Of these, the FA revealed CNV in its entirety under or near the hemorrhage in 11. The ICG examination revealed all 11 plus an additional 12 hypercyanescent features that were not visible by FA. Although 26 of 32 SPEDs had CNV by FA analysis, 31 of 32 had well-defined hypercyanescence under or at the SPED edge by ICG videoangiography. Eighty-seven percent of eyes with classic choroidal neovascular membranes (CCNV; 20 of 23) and 93% of eyes with fibrovascular pigment epithelial detachments (FVPED; 66 of 71) were hypercyanescent with distinct edges. Fifty percent of eyes with only late leakage of undetermined source (LLUS) were hypercyanescent, whereas 50% were isocyanescent on ICG videoangiography. There were four of 104 eyes in which ICG videoangiography revealed poorly defined hypercyanescent areas when, by FA, there were only drusen or elevated blocked fluorescence. Indocyanine green videoangiography revealed only three eyes with poorly defined hypercyanescence of 200 fellow eyes in patients with unilateral AMD.

Indocyanine green videoangiography correlated fairly well with stereoscopic FA for the presence of CCNV or FVPED, but correlated poorly when LLUS was present. In certain cases where CNV was associated with SPED or macular hemorrhage, ICG videoangiography demonstrated features not apparent on FA.