AIM:To identify the reasons for not improving visual acuity of patients with diabetic macular edema(DME)and its correlation with optical coherent tomography(OCT)imaging characteristics. They were treated with anti-vascular endothelial growth factor(VEGF)again after an interruption of treatment due to the COVID-19 pandemic.

METHODS: Retrospective cases study. A total of 31 patients(40 eyes)with DME and treated with anti-VEGF drugs in the Department of Ophthalmology of the First Affiliated Hospital of Xi'an Jiaotong University between December 2019 and July 2020 were selected. All patients had an treatment history of interruption due to COVID-19. Based on the behavioral standard of best corrected visual acuity(BCVA)at 1mo after re-treatment, they were divided into improved visual acuity group(14 eyes)and non-improved visual acuity group(26 eyes). All patients received comprehensive ophthalmic examinations, including BCVA, intraocular pressure, OCT, fundus photography, and fundus fluorescein angiography(FFA). The central retinal thickness(CRT), hyper-reflective retinal dots(HRD), subretinal fluid(SRF), intraretinal fluid(IRF), disorganization of the inner retinal layers(DRIL)and ellipsoid zone disruption(EZD)were quantified by Image J. The correlation between BCVA and CRT, the number of HRD, the diameter of SRF and IRF, and the height of DRIL and EDZ width at 1mo after treatment were analyzed by Spearman correlation.

RESULTS: The proportion of HRD in OCT was significantly higher in the non-improved visual acuity group than that in the improved visual acuity group(χ²=5.43, P=0.0072); There was a positive correlation between the difference of BCVA(LogMAR)and CRT change in DME patients after treatment(r_s=0.406, P=0.009); There was a positive correlation between BCVA(LogMAR)and EZD width after treatment(r_s=0.358, P=0.023). There was no correlation between BCVA(LogMAR)after treatment and the number of CRT(r_s=-0.086,P=0.600)and HRD(r_s=-0.115,P=0.482), the diameter and height of SRF(r_s=0.102,P=0.530; r_s=0.100,P=0.538)and IRF(r_s=0.134,P=0.410; r_s=-0.014,P=0.932), and the width of DRIL(r_s=0.089,P=0.587).

CONCLUSION:The probability of retinal HRD was significantly higher in patients with no visual improvement than that in patients with visual improvement after the re-treatment with anti-VEGF due to the treatment interruption during COVID-19. EZD width was also closely correlated to the prognosis of visual acuity. As for these patients, HRD manifestations in OCT should be comprehensively considered, taking intravitreal injection of glucocorticoid as an auxiliary strategy if necessary.