Using optical coherence tomography to distinguish intraretinal microvascular abnormalities from neovascular complexes in diabetic retinopathy

Abstract

 

Purpose

 

To identify novel optical coherence tomographic (OCT) derived parameters, which distinguish between intraretinal microvascular abnormalities (IRMA) and neovascular complexes (NVC).

 

Methods

 

Data was collected retrospectively from 12 patients with diabetic retinopathy who underwent concurrent fluorescein angiography (FA) and OCT scanning over areas of IRMA or NVE. Classification of IRMAs and NVEs were based on colour and red-free photographs according to Early Treatment Diabetic Retinopathy Study (ETDRS) criteria. Two independent graders assessed FA and OCT images. OCT images were graded according to the following parameters: (1) Breach of the internal limiting membrane (ILM); (2) presence of vitreous dots; (3) breach of the posterior hyaloid interface.

 

Results

 

31 lesions from 18 eyes (12 patients) were examined. Reproducibility of grading of all images was good with a weighted Kappa=0.87 (SE=0.09). According to ETDRS criteria, 16/31 (51.6%) were IRMAs and 15/31 (48.4%) NVE. FA and OCT features which significantly distinguished IRMA and NVE were (1) FA leak [3/16 (18.8%) IRMA vs 15/15 (100%) NVE, p=0.0001], (2) internal limiting membrane breach [9/16 (56.3%) IRMA vs 15/15 (100%) NVE, p=0.007), (3) posterior hyaloid breach [2/16 (12.5%) vs 11/15 (73.3%), p=0.001], and (4) adjacent vitreous hyper-reflective dots [5/16 (31.3%) IRMA vs 12/15 (80.0%) NVE, p=0.01]. Stepwise multiple logistic regression analysis revealed that the “breach of the posterior hyaloid” alone on OCT images, could independently distinguish NVE from IRMAs (OR=19.3, CI=2.96-125.2, p=0.002).

 

Conclusions

 

In this study, we identified novel OCT-derived parameters, which distinguish IRMAs and NVCs. If these parameters are successfully validated, they may not only serve as end-points to clinical trials, but also enable clinicians to closely monitor patients with less reliance on regular invasive FA investigation.

 

Keywords: 499 diabetic retinopathy • 552 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • 550 imaging/image analysis: clinical