Averaging optical coherence tomography angiography (OCT-A) images, rather than evaluating single images, improves vessel visualization in patients with diabetic retinopathy (DR), allowing for earlier disease identification and treatment, according to research published in Ophthalmology Retina.
In a retrospective, cross-sectional cohort study, researchers averaged multiple OCT-A images to quantitatively compare retinal vascular features and correlate these metrics to patient best-corrected visual acuity (BCVA) and DR severity. Ultimately, investigators sought to evaluate the efficacy of this averaging method in terms of both quality and outcome severity as compared with the use of single OCT-A images.
The study cohort included 84 eyes from 55 patients, including: 28 eyes from 19 healthy, age-matched controls, 11 eyes from 8 people with diabetes without DR, 9 eyes from 7 patients with mild DR, 10 eyes from 7 patients with moderateDR, 5 eyes from 3 patients with severe DR, and 21 eyes from 15 patients with proliferative DR. All OCT-A images were consecutively obtained from a single sitting and were analyzed using spectral-domain OCT-A with an angiography 3×3 mm scan pattern.
At baseline, laterality, gender, and lens status were not significantly different between groups, but mean BCVA was better in healthy eyes.
Lower vessel length density (VLD) and perfusion density (PD) parameters were “almost uniformly” associated significantly with worse DR and best-corrected visual acuity. A univariate regression analysis comparing quantitative metrics for both average and single images indicated that foveal avascular zone size in the averaged images was significantly larger compared with single images in eyes with nonproliferative or proliferative DR (0.39 ± 0.17 mm2 vs 0.23 ± 0.14 mm2; P <.034 and 0.46 ± 0.16 mm2 vs 0.27 ± 0.14 mm2; P <.001). Conversely, VLD and PD demonstrated nearly uniform significantly lower values in the average images.
In a univariate linear regression analysis comparing quantitative metrics for best-corrected visual acuity, foveal avascular zone size was not associated with best-corrected visual acuity in single images, compared with a significant association found in averaged images (coefficient, 0.21; P =.004). A significant association was also noted for deep retinal layer temporal VLD among averaged images (coefficient, -0.013, P =.009), though this was not statistically significant in single images (P =.058).
A final univariate linear regression analysis compared quantitative metrics with DR severity and found significant associations in all superficial and deep retinal layer parameters, excepting the superficial retinal layer inferior PD.
Investigators then used a multivariate linear regression model, including all OCT-A parameters significantly associated with DR severity, and found that averaging did not result in an increased R2 value (0.82 and 0.77 for single and averaged groups, respectively). After including all potential explanatory variables, none remained associated significantly with DR severity in multivariate analysis with single images. Among averaged images, increased superficial retinal layer superior Early Treatment of Diabetic Retinopathy Study perfusion density was a significant predictor of worse DR severity (coefficient 52.7; P =.026), and a similar outcome for BCVA showed comparably low R2 values (0.42 and 0.47, respectively).
A study limitation was the limited sample size overall as well as a small sample of eyes for each stage of DR severity. Other limitations include the cross-sectional design of the study and its use of only 1 OCT-A system, which limits the potential reproducibility of the investigation.
“Our study is the first to demonstrate the qualitative and quantitative improvement with registered image averaging in eyes with varying degrees of DR severity,” the researchers concluded. “Earlier identification of eyes at significant risk because of lower vascular density and macular ischemia may provide for better individualized management of diabetic retinal disease.”
Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.
Jung JJ, Yu DJG, Zeng A, et al. Correlation of quantitative measurements with diabetic disease severity using multiple en face OCT angiography image averaging. Ophthalmol Retina. 2020;4(11):1069-1082.