Relationship between central retinal sensitivity, thickness, perfusion density and visual acuity in patients with branch retinal vein occlusion.

Abstract

Editor, Patients with branch retinal vein occlusion (BRVO) show reduced perfusion density (PD) in the area affected by the vein occlusion on optical coherence tomography angiography (OCTA). Although PD positively correlates with retinal sensitivity (RS) in the affected area (Rachima et al. 2020), patients may have decreased PD with preserved visual acuity (VA) posttreatment. Therefore, we hypothesized that the condition of the unaffected side is more important for VA than the PD or RS of the affected side. We examined how the central foveal retinal thickness (CFT), RS and PD at four points in the central region affect VA. Patients with major BRVO who received ranibizumab 0.5 mg/0.05 ml (Lucentis kit for intravitreal injection, 10 mg/ml; Novartis, Basel, Switzerland) 1+ pro re nata regimen at Shinshu University Hospital during 2017–2018 were assessed at baseline, 6 and 12 months postinitial treatment. CFT (CIRRUS HD-OCT; Carl Zeiss Meditec, Inc., Jena, Germany), RS (Microperimeter MP-3; Nidek Co., Ltd.) in quadrants around the central 2° (affected and unaffected, temporal and nasal), PD of the whole retinal segmentation (OCTA PLEX Elite 9000; Carl Zeiss Meditec, Inc.) and VA were measured (Fig. 1A–D). The association between VA and each variable was evaluated. To minimize OCTA artefacts, an eye-tracking function was used. Baseline PD was 1 month postinjection (Hasegawa et al. 2019), and all images had a signal intensity of 7 or higher. All procedures followed the tenets of the Declaration of Helsinki. Fifty-seven test results with 19 eyes of 19 patients (mean age 70.0 ± 8.2 years; 8 men, 11 women) were analysed. VA correlated negatively with CFT (logMAR equivalent, r = 0.50, p < 0.0001, Pearson's correlation coefficient [Fig. 1E]) and positively with central RS in all quadrants (affected nasal: r = −0.58, p < 0.0001; affected temporal: r = −0.37, p = 0.004; unaffected nasal: r = −0.58, p < 0.0001; unaffected temporal, r = −0.67, p < 0.0001 [Fig. 1F]), but not with PD (affected nasal: r = −0.12, p = 0.391; affected temporal: r = 0.13, p = 0.339; unaffected nasal: r = 0.08, p = 0.546; unaffected temporal: r = −0.07, p = 0.603). Multiple regression analysis using CFT and 4-quadrant RS showed that the unaffected temporal RS was a significant determinant of VA (estimate: −0.0162, p = 0.0125). Baseline central RS positively correlated with VA, except for that in the affected temporal side (affected nasal: r = −0.27, p = 0.0430; affected temporal: r = −0.21, p = 0.1089; unaffected nasal: r = −0.32, p = 0. 0147; unaffected temporal: r = −0.41, p = 0.0015 [Fig. 1G]). In patients with major BRVO, VA negatively correlated with CFT, and RS of the unaffected temporal side was a significant determinant of VA. However, PD did not correlate with VA. This suggests that decreased PD does not affect VA, but macular oedema and decreased RS in unaffected areas have a greater effect. Because blood flow in the macula is divided into upper and lower arcade vessels, macular PD reduction caused by BRVO should not extend to the unaffected side. Therefore, the decreased RS on the unaffected side may be due to macular oedema and photoreceptor cell damage with serous retinal detachment (SRD) (Ota et al. 2007). Patients with a shorter time between onset and treatment had a better visual prognosis (Scott et al. 2011). In this study, RS at baseline positively correlated with VA. Because natural recovery of VA beyond 20/40 occurs in few patients (Rogers et al. 2010), early treatment initiation is strongly recommended before macular oedema or SRD reduces RS on the unaffected side.