| Objective: To investigate the correlation between the retinal structure and function of the central retina in diabetic macular edema(DME)after 3consecutive intravitreal injections of Conbercept through multifocal-electroretinogram(mf-ERG)and optical coherence tomography(OCT),to find out the correlation between the central retinal function and anatomy,to explore the evaluation value of mf-ERG in anti-VEGF therapy.Methods:Fifty-two eyes of 52 patients with non-proliferative diabetic retinopathy and DME were collected from the fudus department in the First Affiliated Hospital of Chongqing Medical University during June 2020 to December 2020.They were diagnosed by fundus examination,OCT and fundus fluorescein angiography(FFA)and were given an intravitreal injection of Conbercept 0.05 m L/0.5mg per month for 3 consecutive sessions.We recorded the best corrected visual acuity(BCVA),length of inner retinal disturbance,the thickness of central retinal macular(CRT),external limiting membrane/ellipsoid zone,and P1 and N1 wave amplitude density and latency data for each month before and after mf-ERG R1,R2,and R3 ring.In addition,we analyzed the correlation between changes in mf-ERG parameters(from the objective function)and visual acuity(from the subjective function)and the morphological characteristics of OCT(from the structural aspect).The mf-ERG parameters of the corresponding points in the mf-ERG field map were evaluated.The SD-OCT foveal B-scan was used to assess any correlation between the external limiting membrane(ELM)status,the ellipsoid zone region(EZ)status,the presence of a cyst with any correlation between the retinal inner layer(DRIL)or disordered,and the central corresponding region of mf-ERG parameters.All data were statistically analyzed using SPSS23.0 software.K-S test was used to assess the normality of the parameter distribution.Intergroup comparisons were normalized by repeated measures of variance analysis,while nonnormality was tested by Friedman’s nonparametric test.We evaluated the correlation between the data using Pearson or Spearman coefficients.BCVA was used as the dependent variable,and the impact of each factor was further evaluated by multiple regression analysis.A P value less than 0.05 is considered significant.Result: After three intravitreal injections,the mean BCVA in all 52 eyes was significantly improved compared to the first treatment(0.21±0.17 VS 0.28±0.13 VS 0.34±0.12 VS 0.37±0.19;P <0.001),the last time was not significantly improved compared with the third time(P> 0.05).The mean CRT decreased(485.00±149.76μm VS 366.62±80.25μm VS 323.09±58.65μm VS 298.44±56.86μm;P<0.001),but there was no significant difference between the third time and the second time,and no thinner than the last time(P>0.05).The length of DRIL of the last test(per100μm)was significantly different from that of the first test(P<0.001).According to mf-ERG parameters,the length of the last ELM/EZ defect(per100μm)was statistically different from that of the third defect(P<0.001).Each point increase in the amplitude density of P1 and N1 waves in the R1 ring was associated with a decrease during the waiting time(P<0.001).Each point increase in the amplitude density of P1 waves in the R2 ring was associated with a statistically significant difference(P<0.001).There was no significant correlation between BCVA and CRT at any subsequent time point(P>0.05).BCVA was significantly positively correlated with the amplitude density of P1 and N1 waves of the R1 ring,while the latency of P1 and N1 was significantly negatively correlated with BCVA.At 1Mo,the amplitude density of N1 wave in the R1 ring was negatively correlated with CRT(r=-0.326,P=0.019).The length of DRIL is significantly negatively correlated with the amplitude density of R1’s P1 wave.The increase of amplitude density of P1 wave was correlated with the decrease of DRIL length(r=0.437,P=0.001).The length ofELM/EZ defect is significantly positively correlated with the P1 wave latency of R1 ring.The decrease of P1 wave latency was related to the decrease ofELM/EZ defect length(r=0.283,P=0.042).Multiple regression analysis showed that the best corrected visual acuity was related to the amplitude density of the fundus.P1 wave in R1 ring(P=0.046),N1 wave amplitude density in R1 ring(P=0.013),length ofEZ/ELM defect(P=0.036)and length of DRIL defect(P=0.009).Conclusion: After three intravitreal injections of Conbercept,the 52 DME eyes showed a significant increase in mean BCVA after treatment.The improvement was not significant after the third dose.The CRT was significantly reduced.However,the time of reduction after the second dose was not significant.By the end of treatment,the recovery of DRIL and ELM/EZ defect length was significant.Compared with the prior time point,the amplitude density of P1 and N1 waves of the R1 ring of mf-ERG was significantly increased.At the same time the latency period was shortened.They suggested that Conbercept could improve retinal function and sensitivity in the fovea of macular area.Multifocal electroretinogram and OCT showed that there was a correlation between changes in retinal structure and function.Changes in some parameters of the central ring of mf-ERG(such as P1 wave amplitude density and latency)correlate with structural abnormalities of the central retina revealed by OCT(such as DRIL and ELM/EZ rupture).Compared with the changes of DRIL and ELM/EZ length,the changes of mf-ERG parameters are more sensitive,intuitive,objective and easy to obtain.It can be used as a supplement for OCT monitoring the anti-VEGF efficacy of DME.In addition,multiple regression analysis showed that the findings of mf-ERG and OCT could complement each other and even have predictive value in predicting visual acuity in patients with DME. |