Different Anti-Vascular Endothelial Growth Factor for Patients With Diabetic Macular Edema: A Network Meta-Analysis

Abstract

Background: Antiangiogenic therapy with anti-vascular endothelial growth factor (VEGF) is commonly used to treat diabetic macular edema (DME), which can reduce edema, improve vision, and prevent further visual loss. There is little head-to-head trial data to guide the selection of an individual VEGF inhibitor. Therefore, we aimed to investigate the efficacy and safety of anti-VEGF for patients with DME and to assess the differences between clinically relevant options by using network meta-analysis (NMA). Methods: MEDLINE, Embase, the Cochrane Library, Web of Science, Chinese Biomedical Literature Database, Wanfang, China National Knowledge Infrastructure, and VIP databases were searched for published randomized controlled trials (RCTs) from their inception to November 2020. We included RCTs of anti-VEGF drugs (intravitreal aflibercept (IVT-AFL), intravitreal ranibizumab (IVR), and intravitreal conbercept (IVC)) treating adult patients who were diagnosed with DME, regardless of stage or duration of the disease. We estimated summary odds ratios (ORs) and mean differences (MDs) with 95% credible intervals (CrIs) using a Bayesian NMA. This study's registration number is CRD42021259335. Results: We identified 43 RCTs comprising 8,234 patients. Beneficial effects were observed in patients who used IVT-AFL compared with those who used other anti-VEGF therapies at 1-year follow-up on corrected visual acuity (BCVA) improvements (all patients: versus IVR: MD 2.83, 95% CrIs 1.64, 4.01, versus IVC: MD 2.41, 95% CrIs -0.52, 5.32; patients with worse baseline visual acuity (VA): versus IVR: MD 3.39, 95% CrIs 1.89, 4.9, versus IVC: MD 3.49, 95% CrIs 0.49, 6.44) and the proportion of patients with a gain of at least 15 Early Treatment Diabetic Retinopathy Study (ETDRS) letters (all patients: versus IVR: OR 1.55, 95% CrIs 1.11, 2.17, versus IVC: OR 2.78, 95% CrIs 1.23, 6.04; patients with worse baseline VA: versus IVR: OR 2.05, 95% CrIs 1.18, 3.58, versus IVC: OR 2.85, 95% CrIs 1.24, 6.41). The effect of improvement in BCVA was identified for IVT-AFL compared to intravitreal bevacizumab. Based on the surface under the cumulative ranking curve (SUCRA), IVT-AFL had the highest probability of being the most effective option (99.9% and 99.5% in terms of the two primary outcomes, respectively). At the 2-year follow-up, numerical differences were identified favoring IVT-AFL; however, they did not reach statistical significance when comparing IVT-AFL to IVR. In the analysis of adverse events, IVT-AFL showed a lower risk of incidence of ocular adverse events compared to other anti-VEGF therapies at 1-year follow-up (versus IVR: OR 0.45, 95% CrIs 0.28, 0.7; versus IVC: OR 0.36, 95% CrIs 0.21, 0.63). Conclusion: IVT-AFL resulted in greater beneficial effects on BCVA and a higher proportion of patients with a gain of at least 15 ETDRS letters compared to IVR or IVC one year after treatment (especially in DME patients with worse baseline VA). In addition, fewer ocular adverse events occurred in the IVT-AFL group compared to the IVR or IVC groups. After two years, there was insufficient evidence to identify which anti-VEGF has superior efficacy or safety. Clinical Trial Registration: https://www.crd.york.ac.uk/prospero/, PROSPERO; https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42021259335, CRD42021259335.

Keywords: aflibercept; conbercept; diabetic macular edema; network meta-analysis; ranibizumab.

FIGURE 1

PRISMA flow diagram.

FIGURE 2

Assessment of risk of bias for included RCTs [(A) Risk of bias graph; (B) Risk of bias summary]. Notes: (A) Risk of bias graph: reviewers’ judgements about each risk of bias item are presented as percentages across all randomized controlled trials (RCTs). (B) Risk of bias summary: reviewers’ judgements about each risk of bias item for each included RCT. “Low” risk of bias in green, “Unclear” in yellow, and “High” risk of bias in red.

FIGURE 3

Network geometry for BCVA (ETDRS letters) mean change from baseline. All populations at 1-year follow-up [(A) 18 trials] and 2-year follow-up [(C) 5 trials]. Population with worse baseline VA at 1-year follow-up [(B) 12 trials] and 2-year follow-up [(D) 5 trials]. BCVA, best-corrected visual acuity; ETDRS, Early Treatment Diabetic Retinopathy Study; IVB, intravitreal bevacizumab; IVC, intravitreal conbercept; IVR, intravitreal ranibizumab; IVT-AFL, intravitreal aflibercept. Notes: Direct comparisons are represented by the black lines connecting the different interventions. Line width is proportional to the number of trials including every pair of interventions, whereas circle size is proportional to the total number of trials for each intervention in the network.

FIGURE 4

GRADE for the primary outcomes (A-1, B-1, C-1) are for the outcome BCVA (ETDRS letters) mean change from baseline; (A-2, B-2, C-2) are for the proportion of patients with a gain of at least 15 ETDRS letters. Notes: (A) Summary of study limitations of the included randomized controlled trials (RCTs). The colors in the circles indicate the percentage of low risk of bias RCTs (green), moderate risk of bias RCTs (yellow), and high risk of bias RCTs (red) involving each intervention. The colors of the line then indicate the summative risk of bias assessment of each comparison based on the above information–low risk of bias comparison (green), moderate risk of bias comparison (yellow), high risk of bias comparison (red). (B) Contribution of risk of bias comparisons to focus comparison estimates (between IVT-AFL, IVR, and IVC). (C) Table of domains for downgrading. BCVA, best-corrected visual acuity; ETDRS, Early Treatment Diabetic Retinopathy Study; GRADE, Grading of Recommendations Assessment, Development and Evaluation; IVC, intravitreal conbercept; IVR, intravitreal ranibizumab; IVT-AFL, intravitreal aflibercept.

FIGURE 5

Network geometry for the proportion of patients with at least 15 ETDRS letters. All populations at 1-year follow-up [(A) 11 trials] and 2-year follow-up [(C) 7 trials]. Population with worse baseline VA at 1-year follow-up [(B) 8 trials] and 2-year follow-up [(D) 7 trials]. ETDRS, Early Treatment Diabetic Retinopathy Study; IVB, intravitreal bevacizumab; IVC, intravitreal conbercept; IVR, intravitreal ranibizumab; IVT-AFL, intravitreal aflibercept. Notes: Direct comparisons are represented by the black lines connecting the different interventions. Line width is proportional to the number of trials including every pair of interventions, whereas circle size is proportional to the total number of trials for each intervention in the network.