Retinal Vascular Permeability in Diabetic Subjects without Retinopathy Compared with Mild Diabetic Retinopathy and Healthy Controls

Abstract

Objective

To investigate retinal vascular permeability mapping as a potential biomarker for diabetic retinopathy in subjects with diabetes with no signs of retinopathy and with mild nonproliferative retinopathy.

Design

This is a case-control study.

Subjects

Participants included 7 healthy controls, 22 subjects with diabetes mellitus and no clinical signs of retinopathy (DMnoDR), and 7 subjects with mild nonproliferative diabetic retinopathy (NPDR).

Methods

All participants underwent routine retinal fluorescein videoangiography (FVA). Each FVA dataset was analyzed with the dynamic tracer kinetic model (DTKM) method to estimate 5 parameters: extraction fraction (E), blood flow, arrival time, transit time, and rate constant defined via adiabatic solution. The DTKM method was based on indicator dilution theory, including sequential use of 2 prominent kinetic models: the plug flow model and the adiabatic approximation to the tissue homogeneity model.

Main Outcome Measures

Extraction fraction, i.e., the fluorescein dye leakage measured during 1 pass through surrounding retinal tissue, is extracted via DTKM method and directly relates to retinal vascular permeability. Thus, E represents the preclinical biomarker, retinal vascular permeability.

Results

The 3 diagnostic groups were found to have significantly different permeability (P = 0.003). Despite having no clinical signs of retinopathy, the mean rank of average vascular E was significantly higher in DMnoDR subjects compared with healthy controls (P = 0.04), as was the mean rank of E for mild NPDR subjects (P = 0.002). The average E for mild NPDR, DMnoDR, and control subjects was 0.10 ± 0.04, 0.07 ± 0.04, and 0.04 ± 0.01, respectively.

Conclusions

The vascular permeability extracted from FVA datasets using the DTKM method is a promising biomarker for detecting preclinical retinal pathology in patients with diabetes. Longitudinal studies are ongoing to explore the ability of this biomarker to distinguish those subjects with diabetes who will progress to clinically apparent retinopathy from those who will not.

Financial Disclosure(s)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.