Associations Between Baseline Biomarkers and Functional Outcomes in Patients with Diabetic Macular Edema after Anti-VEGF Treatment.

Objective:To investigate associations between baseline structural and functional biomarkers and 6-month post-treatment best-corrected visual acuity (BCVA) among patients with diabetic macular edema (DME) receiving intravitreal anti-VEGF therapy. Methods:Data from patients treated at Hebei Eye Hospital between December 2022 and January 2024 were retrospectively analyzed. Baseline biomarkers, including the disruption of the external limiting membrane (ELM) and ellipsoid zone (EZ), disorganization of retinal inner layers (DRIL), and epiretinal membrane (ERM) on spectral domain optical coherence tomography (SD-OCT), as well as the amplitude and implicit time of N1 and P1 waves in multifocal electroretinography (mfERG), were collected. The associations between these biomarkers and BCVA after anti-VEGF treatment were then assessed. Linear regression analysis was applied to quantify the correlations between baseline SD-OCT morphological biomarkers and 6-month anti-VEGF visual therapeutic efficacy. Multivariate logistic regression analysis was conducted to assess the efficacy of anti-VEGF treatment and baseline mfERG biomarkers. Results:A total of 87 eligible DME patients completed three monthly anti-VEGF loading injections and 6-month follow-up, including 48 patients in the response group and 39 in the nonresponse group. Linear regression assessed whether baseline retinal layer disruption predicted 6-month visual response and revealed that extensive ELM (t = 2.914, p = 0.004) and EZ (t = 2.112, p = 0.036) disruption at baseline and extensive DRIL (t = 2.922, p = 0.004) were associated with nonresponse defined as < 3 lines of BCVA improvement at 6 months after the final anti-VEGF injection. A multivariate logistic regression analysis indicated that, for an N1 implicit time of R1, each 1-ms reduction in N1 implicit time for Ring 1 corresponded to an adjusted odds ratio of 1.252 for favorable visual response (95% CI [1.101, 1.422], p = 0.001) after anti-VEGF treatment. The results were the same for an N1 implicit time of R2 (1.111, 95% CI [1.028, 1.201], p = 0.008), a P1 implicit time of R1 (1.049, 95% CI [0.998, 1.102], p = 0.029), and a P1 implicit time of R2 (1.105, 95% CI [1.012, 1.206], p = 0.025). Conclusions:Baseline SD-OCT and mfERG biomarkers were significantly associated with anti-VEGF treatment outcomes in this DME cohort. Shorter N1 and P1 implicit times in the first and second mfERG rings were correlated with favorable treatment responses, while extensive disruption of ELM and EZ, severe DRIL, and the presence of tractional ERM were linked to poorer anti-VEGF responses. These structural and functional biomarkers may offer potential prognostic reference value for evaluating treatment prognosis. Further large-sample and prospective studies are still required to validate their clinical applicability and routine use in guiding therapeutic decision-making.