Optical coherence tomography angiography measured area of retinal neovascularization is predictive of treatment response and progression of disease in patients with proliferative diabetic retinopathy

BACKGROUND: The purpose of this study was to evaluate the area of retinal neovascularization in patients with treatment-naïve proliferative diabetic retinopathy (PDR) as measured by optical coherence tomography angiography (OCT-A) as a marker of subsequent treatment response after panretinal photocoagulation (PRP), and to examine if this area correlated with area of retinal neovascularization as measured by fluorescein angiography (FA). METHODS: En face OCT-A scans (4.5 × 4.5 mm) of neovascularizations were obtained at baseline (BL) before PRP and at month (M) 3 and M6 after treatment. Progression of PDR were defined as lesion growth (assessed by ophthalmoscopy and wide-field fundus photo) or increasing leakage by Optos ultra-widefield FA, and patients were divided into two groups; progression or non-progression. Mann–Whitney U test and Wilcoxon signed-rank test were used to analyse differences between groups and between time points. Areas of retinal neovascularizations (OCT-A and FA) were calculated by algorithms developed in Python (version 3.6.8, The Python Software Foundation, USA). RESULTS: Of 21 eyes included, 14 had progression of disease. Median OCT-A area did not differ between the two groups (progression vs. non-progression) at BL (76.40 ± 162.03 vs. 72.62 ± 94.15, p = 0.43) but were statistically significantly larger in the progression group at M6 (276.69 ± 168.78 vs. 61.30 ± 70.90, p = 0.025). Median FA area did not differ in the progression vs. the non-progression group at BL (111.42 ± 143.08 vs. 60.80 ± 54.83, p = 0.05) or at M6 (200.12 ± 91.81 vs. 123.86 ± 162.16, p = 0.62). Intraclass correlation between area by OCT-A and FA was −5.99 (95% CI: −35.28–0.993), p = 0.71. CONCLUSIONS: In this study of patients with treatment-naïve PDR, we showed that increasing area of retinal neovascularizations measured by OCT-A at M6 indicated progression of disease after PRP treatment. Our results suggest that area by OCT-A reflects disease activity and that it can be used as an indicator to monitor the progression of PDR over time, and to evaluate treatment response six months after PRP. Trial registration https://clinicaltrials.gov (identifier: NCT03113006). Registered April 13, 2017.