Choriocapillaris Loss in Advanced Age-Related Macular Degeneration

Figure 1

A 76-year-old patient with geographic atrophy (GA). (a) Color fundus photograph. (b) En face projections of a 6 mm × 6 mm, unthresholded optical coherence tomography (OCT) angiography (OCTA) volume from Bruch’s membrane to 45 μm below. B.1 corresponds to a 1.5 ms interscan time OCTA volume, and B.2 corresponds to a 3.0 ms interscan time OCTA volume. Projection artifacts from large retinal vessels have been removed and colored black. The white contours trace the margin of atrophy, as determined by a subretinal pigment epithelium (RPE) slab of the OCT volume. Note that the 1.5 ms interscan OCTA image reveals substantially more choriocapillaris alteration than does the 3.0 ms interscan time image. In some regions, the OCTA signal is documented in the 3.0 ms interscan time but not the 1.5 ms interscan time, suggesting that these regions have flow impairment rather than complete choriocapillaris atrophy. (c) Binarized versions of the choriocapillaris OCTA images in B, where a constant threshold was used. C.1 corresponds to the 1.5 ms interscan time OCTA image, and C.2 corresponds to the 3.0 ms interscan time OCTA image. Again, note there are substantially more areas of low choriocapillaris flow (black) in the 1.5 ms interscan time OCTA image than in the 3.0 ms interscan time OCTA image. (d) OCT and OCTA B-scans extracted from the locations indicated by the dashed pink lines of B.1 and C.1. The OCT B-scan (D.1) shows RPE and photoreceptor loss, which causes increased light penetration into the choroid. The 1.5 ms OCTA B-scan is shown in D.2, and the 3.0 ms OCTA B-scan is shown in D.3. Note that both D.2 and D.3 are unthresholded OCTA images, which results in worse image quality. Unthresholded choriocapillaris OCTA images are useful for reducing the rate of false-positive flow impairment due to thresholding. (e-f) Enlargements of the dashed boxes in B-C. Red boxes correspond to 1.5 ms interscan time images, and orange boxes correspond to 3.0 ms interscan time images. The boxes have been rotated 90 degrees clockwise relative to their orientations in B and C. These regions of interest show that there is choriocapillaris flow impairment beyond the margin of RPE atrophy. Arrows point to an example area of flow impairment which changes as a function of interscan time. Note that in the 1.5 ms OCTA, there is less OCTA signal (more dark areas) than in the 3.0 ms OCTA, which makes the impairment more pronounced in the 1.5 ms OCTA (this is easiest seen in F.1 and F.2). This illustrates how shorter interscan time OCTA is more sensitive to flow alterations than is longer interscan time OCTA.

(a)

(b)

(c)

(d)

(e)

(f)