May 31, 2019 Conference
Dr. Mohamed Elmasry, Joslin Diabetes Center
Dr. Paolo Silva, Joslin Diabetes Center
Dr. Lloyd Paul Aiello, Joslin Diabetes Center
Dr. Jennifer K. Sun, Joslin Diabetes Center
Methods: Eyes of diabetic patients were imaged with optical coherence tomography angiography (OCTA) 3x3 mm macular scans. Projection resolved automated software calculated SCP and DCP VD and central retinal thickness (CRT). ETDRS DR severity was graded on ultrawide field images. Clinical data, including treatment history and demographics were reviewed and recorded on standardized forms.
Results: Of 503 sequentially imaged eyes with DR (N=151 [30.0%] proliferative DR) from 282 patients [mean±SD: age 53.8 ± 15.0 yrs, diabetes duration 31.8 ± 16.8 yrs, A1c 8.3% ± 1.5, type 1 diabetes 69.7% (193), female 44.0%(124)], 51 eyes (10.1%) had anti-VEGF therapy within the prior year (total injections: 3.0±1.8). Untreated vs treated eyes had higher SCP (42.0±5.0 vs 39.1±4.3, p<.0001) and DCP VD (46.5±5.2 vs 42.7±5.2, p<.0001) and thinner CRT (277.3±38.1 vs 306.0±54.9 ?m, p<.0001). In bivariate analyses, number of anti-VEGF injections within the past year was inversely correlated with SCP (point estimate [95% CL]: -0.03 [-0.05, -0.02], p=0.003) and DCP VD (-0.05 [-0.07, -0.03], p<.0001). SCP (p=0.03) and DCP VD (p<.0001) both remained significantly associated with past anti-VEGF injection number after adjusting for CRT and DR severity. Time to last injection was not related to SCP or DCP VD. In 54 eyes imaged a mean of 134±33 days post baseline, 22 eyes (9 treatment naïve) had anti-VEGF therapy (2.3±1.2 injections) before follow-up imaging. There was an inverse correlation between baseline DCP, but not SCP VD and subsequent injection number (-0.08 [-0.14, -0.01], p=0.02) even after adjusting for baseline CRT and DR severity. No relationship was found between change in SCP or DCP VD and need for anti-VEGF therapy or number of anti-VEGF injections over a mean period of 124±42 days of anti-VEGF exposure.
Conclusions: These findings suggest that decreased macular VD is potentially associated with past and future need for anti-VEGF treatment in eyes with DR. This is consistent with the hypothesis that lower VD reflects greater severity of diabetic pathology. Further prospective studies may determine whether macular VD is a reliable biomarker of future treatment burden in eyes with DR.