Association of OCT Angiography-Detected Intraretinal Microvascular Abnormalities with Diabetic Retinopathy Severity.

PURPOSE: To assess the severity and clinical significance of intraretinal microvascular abnormalities (IRMAs) using expanded field swept-source OCT angiography (SS-OCTA) in eyes with nonproliferative diabetic retinopathy (NPDR).

DESIGN

Cross-sectional, observational study.

PARTICIPANTS

One hundred thirty-nine eyes from 101 subjects with NPDR.

METHODS

The montage of 12 × 12-mm angiography centered on the macula and optic nerve was evaluated by 2 masked graders for (1) the presence of IRMA in each 6 × 6-mm field, including center, superotemporal, inferotemporal, superonasal (SN), and inferonasal (IN) to the macula, and SN and IN to the optic nerve and (2) subtypes of IRMA (dilated trunk, net shape, loop, sea fan, and tufted IRMA). Nonperfusion areas (NPA) were quantified using FIJI. Nonproliferative diabetic retinopathy grading was initially collected from chart diagnoses and subsequently verified using ultra-widefield color fundus photos. Logistic and linear regression models were used to evaluate the relationships between IRMA features, diabetic retinopathy (DR) severity, and NPA.

MAIN OUTCOME MEASURES: Intraretinal microvascular abnormality features associated with severe NPDR.

RESULTS

Intraretinal microvascular abnormalities, observed with SS-OCTA, were present in 58.3% of all NPDR eyes and more prevalent in severe (96.6%) than mild (28.8%) to moderate (70.6%) NPDR. The number of affected fields and IRMA subtypes increased with DR severity (P < 0.01). The most common subtype of IRMA is the dilated trunk, comprising 58.3%, followed by the net shape subtype at 35.3%, loop at 11.5%, sea fan at 6.5%, and tufts at 5.8%. Significant predictors of severe NPDR included the presence of IRMA in the central field (odds ratio [OR]: 8.7; P = 0.01), more widely distributed IRMA (OR: 2.2; P < 0.01), a greater variety of IRMA subtypes (OR: 4.2, P < 0.01), and the presence of specific subtypes such as net shape (OR: 16.1; P = 0.02), sea fan (OR: 26.0; P < 0.01), and tufted IRMA (OR: 13.4; P = 0.03). Center-involving IRMA (β = 5.8; P = 0.046) and IRMA with loops (β = 7.0; P = 0.043) were found to be associated with increased NPA.

CONCLUSIONS

Our study demonstrates that IRMA lesions identified on OCT angiography, particularly their distribution and morphology, are associated with DR severity and provide complementary information that may facilitate the integration of SS-OCTA into clinical evaluation of DR.

FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.