Choriocapillaris perfusion assessed using swept source optical coherence tomographic angiography and the severity of diabetic retinopathy

Aims To investigate the relationship between choriocapillaris (CC) blood perfusion and the severity of diabetic retinopathy (DR) using swept-source optical coherence tomography angiography (SS-OCTA). Methods Type 2 diabetes mellitus in the Guangzhou was recruited. DR was graded according to the standardised seven-field colour retinal images adhering to the ETDRS scale. CC perfusion was quantified by evaluating the flow deficit (FD) density, FD number and FD size using SS-OCTA. Multivariable linear regressions were used to evaluate the correlation between CC FD metrics and DR severity. Results A total of 1692 patients (eyes) were included in the final analysis. Participants with DR had a significantly higher FD density than the NDR group, with the differences of 1.61% (95% CI 1.04 to 2.18; p<0.001) among mild non-proliferative DR (NPDR) patients, 2.23% (95% CI 1.76 to 2.70; p<0.001) among moderate NPDR patients and 3.31% (95% CI 2.27 to 4.36; p<0.001) among severe DR than NDR patients after adjusting for confounding factors. Similarly, the higher FD number and size were also correlated with severer degree of DR (all p<0.05). Furthermore, the addition of FD density to conventional risk factors significantly improved the performance for discriminating DR from NDR patients (AUC 0. 829, 95% CI 0. 804 to 0. 855; p<0.001). Conclusion Quantitative FD density, number and size assessed by using SS-OCTA were independently correlated with the severity of DR. The assessment of the CC perfusion metrics in the macula may be a sensitive biomarker of DR changes.

Automated summary

This study investigated the relationship between choriocapillaris blood perfusion and the severity of diabetic retinopathy (DR) using swept-source optical coherence tomography angiography (SS-OCTA). The results showed that FD density, number, and size were independently correlated with the severity of DR. CC perfusion metrics may serve as a sensitive biomarker for evaluating DR changes.