Fig 1.
En-face OCTA angiograms of a healthy subject (left columns, A to H) and a T1D patient with severe NPDR (right columns, A’ to H’).
Black and white scans (A, C, E, and G) represent OCTA angiograms. Color map scans (B, D, F, and H) represent color-coded VD in the corresponding OCTA angiograms. Warmer colors represent higher VD. A-A’ and B-B’: Superficial vascular plexus (SVP). C-C’ and D-D’: Deep vascular complex (DVC). E-E’ and F-F’: Intermediate capillary plexus (ICP). G-G’ and H-H’: deep capillary plexus (DCP). A rarefaction of the capillary meshwork on all OCTA angiograms was observed in the eye with severe NPDR compared to the healthy eye.
Table 1.
Parafoveal vessel density (VD), foveal vascular zone (FAZ) and parafoveal retinal thickness (RT) mean values for the two groups.
Values are expressed as mean ± standard deviation.
Fig 2.
Parafoveal VD box-plots of the retinal vascular plexuses in the two groups.
Healthy subjects: blank box-plots; diabetic retinopathy (DR) patients: box-plots with vertical stripes. VD (%) is represented on the y axis. Superficial vascular plexus (SVP), deep vascular complex (DVC), intermediate capillary plexus (ICP) and deep capillary plexus (DCP) are presented on the x axis. Asterisks over braces indicate a significant difference (p <0.05) between healthy subjects and DR patients in the corresponding plexus. Squares represent outside values. Upper and lower whiskers, respectively, represent the upper and lower adjacent values. Upper and lower box margins represent the 25th and 75th percentiles. The black horizontal line inside the box is the median value.
Fig 3.
Correlations between vessel density in the intermediate and deep capillary plexuses.
A and B: Scatter-plots with solid line showing the relationships between the parafoveal VD (%) in the intermediate capillary plexus (ICP, y axis) and in the deep capillary plexus (DCP, x axis) in healthy subjects (A) and diabetic retinopathy patients (B). Significant correlations (all p <0.01) were observed between the VD in the ICP and in the DCP in both healthy subjects and T1D patients. r = Pearson correlation coefficient.
Fig 4.
En-face OCTA angiograms (black and white) and retinal thickness maps (color-coded maps) of a healthy subject (left columns, A to D) and a T1D patient (right columns, A’ to D’).
A-A’: OCTA angiograms of the superficial vascular plexus (SVP); B-B’: color-coded maps of the inner retinal thickness (from the inner limiting membrane, ILM, to the border between the inner plexiform layer, IPL, and the inner nuclear layer, INL). C-C’: OCTA angiograms of the deep vascular complex (DVC); D-D’: color-coded maps of the intermediate retinal thickness (including the INL and the outer plexiform layer). Thickness values in micron are reported from each quadrant of the parafoveal area within the color-coded maps.
Fig 5.
Correlations between vessel density (VD) and retinal thickness.
Scatter-plots with solid line showing the relationships between the parafoveal vessel density (VD, %; y axis) and the retinal thickness (micron; x axis) in healthy subjects (A,C) and in diabetic retinopathy (DR) patients (B,D). A and B: significant correlations (all p <0.01) were observed between the VD in the superficial vascular plexus (SVP) and the inner retinal thickness (from the inner limiting membrane, ILM, to the border between the inner plexiform layer, IPL, and the inner nuclear layer, INL) in both healthy subjects (A) and DR (B) patients. C and D: no correlations were observed between the VD in the deep vascular complex (DVC) and the intermediate retinal thickness (including the INL and the outer plexiform layer) in both healthy subjects (C) and DR (D) patients. r = Pearson correlation coefficient.
Fig 6.
En-face OCTA angiograms and structural B-scans of two T1D patients.
En-face OCTA angiograms of the superficial vascular plexus (SVP, A and D) and deep vascular complex (DVC, C and F) of two T1D patients with the corresponding structural B-scan with angio-overlay (B and E), both passing at the green line shown on the en-face OCTA angiograms. Different patterns of capillary drop-out are seen among the plexuses. In box A, an area of capillary drop-out in the SVP is delimited by the yellow bracket and corresponds to an area of inner retinal thinning and no flow signal in B. In the DVC, the flow signal is visible in the same area as in A, with visible capillaries in both boxes C and B (white arrows indicate red spots on the angio-overlay B-scan). In box D, two arrows from the SVP indicate two bigger vessels with no flow signal between them and a corresponding retinal thinning in E. In the DVC, the flow signal is visible in the same area as in D, with visible capillaries in both boxes F and E (white arrows indicate red spots on the angio-overlay B-scan). The inner nuclear layer and outer plexiform layer show an irregular wavy pattern on both boxes B and E.
Fig 7.
En-face OCTA angiograms of vascular plexuses in a T1D patient.
En-face OCTA angiograms of the superficial vascular plexus (SVP, A) and deep vascular complex (DVC, B) in severe NPDR. A’ and B’ represent the same areas magnified in boxes 1,2,3,4. Boxes 1,2,3,4 represent consecutive color-coded OCTA slabs with the same thickness acquired at increasing depth from the SVP to the DVC. White vessels arise from the SVP while purple vessels arise from the DVC. Arrows (white and yellow) and the white arrow-head indicate residual capillaries in an area of capillary drop-out in the SVP (A) connecting with a denser capillary network in the DVC.