Association Between Clinical Biomarkers and Optical Coherence Tomography Angiography Parameters in Type 2 Diabetes Mellitus

Abstract

Purpose: To investigate the clinical significance of the changes in the macular microvasculature in patients with diabetes mellitus type 2 without diabetic retinopathy.

Methods: Fifty-five patients with diabetes mellitus type 2 without diabetic retinopathy and 48 healthy individuals were enrolled in a prospective cross-sectional study. We identified the changes of optical coherence tomography angiography parameters (foveal avascular zone [FAZ] area and circularity, vessel density, and perfusion index) of the 6 × 6-mm macular scan. Correlation and multiple regression analyses were performed between optical coherence tomography angiography parameters and previously known diabetes mellitus type 2-related demographic and systemic characteristics, and serum biochemical markers.

Results: FAZ parameters and perfusion index of the superficial and deep vascular plexus showed significant correlation with serum insulin level, and homeostasis model assessment indices. In multiple linear regression analysis, low insulin levels predicted increased FAZ areas in both the superficial (β = -0.007; P = 0.030) and deep layers (β = -0.010; P = 0.018) and a decreased perfusion index in the deep layer (β = 0.003; P = 0.001).

Conclusions: The expansion and loss of circularity of the FAZ and the decrease in the perfusion index may be affected by insulin resistance and secretory capacity in patients with diabetes mellitus type 2 with no diabetic retinopathy.

Figure 1.

Optical coherence tomographic angiography (OCT-A) images of the right eye of a 57-year-old man with type 2 diabetes with no DR on clinical examination. En-face OCT-A images (a, b) and corresponding cross-sectional B-scan with flow (c, d) within the 6 × 6-mm macular cubes were captured at the level of the superficial capillary plexus (SCP; a, c) and the deep capillary plexus (DCP; b, d), respectively. The dotted purple lines on the B-scan image represent the boundaries of the segmentation slabs with each vascular network from which the OCT-A image was generated. The projection artifacts of the SCP layer were removed in the DCP during en-face OCT-A.

Figure 2.

Image processing steps for microvascular analysis of an en face optical coherence tomographic angiography projection image within the 6 × 6-mm macular cubes at the level of the superficial capillary plexus. (a) FAZ area is shown in yellow. (b) A binarized image was used to measure the perfusion index, and (c) skeletonization was performed to calculate the vessel density using the tools provided in the ImageJ software. The scan-logo in the bottom right of the images was masked for analysis.