Abstract
OBJECTIVE: To compare the efficacy and safety of different anti-vascular endothelial growth factor (anti-VEGF) agents in combination with pars plana vitrectomy (PPV) for the treatment of proliferative diabetic retinopathy (PDR), in order to guide clinical decision-making. METHODS: A systematic search was conducted in PubMed, EMBASE, the Cochrane Library, and Web of Science for relevant randomized controlled trials (RCTs) up to November 1, 2025. The outcomes assessed included intraoperative bleeding rate, postoperative vitreous hemorrhage rate, changes in best-corrected visual acuity (BCVA), incidence of iatrogenic retinal breaks, reoperation rate, and the duration of surgery. A network meta-analysis was performed to evaluate the efficacy and safety of bevacizumab (IVB), ranibizumab (IVR), conbercept (IVC), and aflibercept (IVA) in combination with PPV. The risk of bias was assessed using the RoB 2 tool, and the quality of evidence was graded using the GRADE framework. This study was registered with the Prospective Register of Systematic Reviews (CRD420251251306). RESULTS: A total of 22 RCTs involving 1, 388 patients (1, 416 eyes) and five treatment regimens were included. Compared with PPV alone, all anti-VEGF agents significantly reduced the risk of intraoperative bleeding, as indicated by 95% confidence intervals (CIs) not crossing the null value of 1. IVR demonstrated the most favorable outcome (OR = 0.03, 95% CI: 0-0.44; SUCRA = 86.1%). For preventing postoperative vitreous hemorrhage, IVA was the most effective (OR = 0.19, 95% CI: 0.05-0.66; SUCRA = 73%), followed by IVC (OR = 0.25, 95% CI: 0.13-0.49) and IVB (OR = 0.30, 95% CI: 0.17-0.53). Regarding postoperative visual improvement, IVC showed the most favorable trend (MD=-0.72, 95% CI: -1.64 to 0.20; SUCRA = 85.2%). IVB significantly shortened the duration of surgery (MD=-19.96, 95% CI: -29.09 to -10.82; SUCRA = 93.5%) and reduced reoperation rate (OR = 0.36, 95% CI: 0.21-0.63; SUCRA = 90.8%). IVR was the most effective in reducing iatrogenic retinal breaks (OR = 0.08, 95% CI: 0.01-0.52; SUCRA = 83.9%). CONCLUSIONS: Different anti-VEGF agents demonstrate distinct advantages during the perioperative period of PDR. IVR and IVC excel in controlling intraoperative and postoperative bleeding, IVB is the most efficient in improving surgical efficiency, and IVC shows the greatest potential for vision improvement. Clinical decisions should be based on individualized treatment goals. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251251306, identifier CRD420251251306.