comparative effectiveness of ranibizumab and bevacizumab in the treatment of diabetic retinopathy: A Systematic Review and Meta-Analysis

 Corresponding author: James Success ODUBIA Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution License 4.0. comparative effectiveness of ranibizumab and bevacizumab in the treatment of diabetic retinopathy: A Systematic Review and Meta-Analysis James Success ODUBIA 1, *, Samson Shina AKINRINADE 2, Samuel Olaniyi AGBESEYI 2, Adekoyejo Ayomide SOWUNMI 2, Daniel Ataije NKAROIJO 2 and Sunday Ezekiel AJILEYE 2 1 Department of Health Promotion and Education, Faculty of Public Health, University of Ibadan, Ibadan, Nigeria. 2 Department of Epidemiology and Medical Statistics, Faculty of Public Health, University of Ibadan, Ibadan, Nigeria. World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 Publication history: Received on 22 July 2025; revised on 25 August 2025; accepted on 29 August 2025 Article DOI: https://doi.org/10.30574/wjbphs.2025.23.2.0790 Abstract Background: Diabetic retinopathy (DR) is a leading cause of vision loss globally. While panretinal photocoagulation has been a long-standing treatment, intravitreal anti-VEGF agents such as ranibizumab have demonstrated superior efficacy, establishing it as a first-line therapy. However, bevacizumab, though off-label, is widely used due to its lower cost and comparable pharmacological profile. This study systematically reviewed and meta-analysed randomized controlled trials (RCTs) published between 2012 and 2021 to compare the therapeutic effectiveness of ranibizumab and bevacizumab in improving visual outcomes among patients with DR. Specific objectives included: (1) rigorous selection and appraisal of eligible RCTs; (2) systematic extraction of trial data; (3) pooled statistical analysis using odds ratios and relative risks; and (4) critical interpretation of findings for clinical practice. Methods: A quantitative approach was employed, using systematic review and meta-analysis in line with PRISMA guidelines. Data were analysed with RevMan 5.4 software, and statistical comparisons were made using odds ratios and relative risk analysis. Results: Pooled analysis of eligible trials demonstrated no heterogeneity (I² = 0%) and revealed a p-value of 0.74, indicating bevacizumab was superior to ranibizumab in improving visual outcomes. Conclusion: Bevacizumab demonstrates greater therapeutic effectiveness than ranibizumab in the treatment of DR. Given its lower cost and accessibility, bevacizumab should be prioritized in clinical practice, particularly in resourcelimited settings. Increased awareness and adoption by healthcare professionals are recommended to optimize DR management. Keywords: Diabetic Retinopathy; Ranibizumab; Bevacizumab; Meta-Analysis; Anti-VEGF 1. Introduction Diabetes mellitus (DM) is a complex metabolic disorder marked by chronic hyperglycemia caused either by insufficient insulin secretion or ineffective utilization of insulin in peripheral tissues. The American Diabetes Association (2010) defines diabetes when plasma glucose levels reach or exceed 200 mg/dl (11.1 mmol/L) two hours post–glucose challenge, while values between 140 and 199 mg/dl (7.8–11.0 mmol/L) indicate prediabetes. DM exists primarily in two forms: type 1 (T1DM), an autoimmune-mediated destruction of pancreatic β-cells, and type 2 (T2DM), which arises from insulin resistance coupled with relative insulin deficiency (Kumar, 2018; Jiang and Dutta, 2017). Globally, the rising prevalence of T2DM, fueled by obesity and sedentary behavior, has intensified the public health burden (Nguyen World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 420 et al., 2012). Chronic hyperglycemia is not only linked to metabolic instability but also to systemic complications such as cardiovascular disease, nephropathy, and neuropathy (Petrie et al., 2018). Among these, diabetic retinopathy (DR) stands out as the leading cause of visual impairment in working-age populations. Its pathogenesis is multifactorial, involving capillary basement membrane thickening, endothelial dysfunction, and ischemia-induced expression of vascular endothelial growth factor (VEGF). VEGF upregulation promotes neovascularization, vascular leakage, and retinal edema, eventually leading to irreversible blindness if untreated (Nguyen et al., 2012). For decades, the treatment of proliferative diabetic retinopathy (PDR) relied on panretinal photocoagulation (PRP), which reduces the risk of severe vision loss but at the cost of peripheral vision and night vision (Madan et al., 2017). The introduction of intravitreal anti-VEGF therapy has transformed DR care by directly targeting angiogenesis and vascular leakage. Ranibizumab, a humanized monoclonal antibody fragment specific to VEGF-A, has been shown to stabilize or improve vision in patients with diabetic macular edema (DME) and proliferative disease (Chong, 2016). Results from the DRCR.net Protocol S trial demonstrated that ranibizumab provides similar or superior visual outcomes compared to PRP, with fewer vitrectomies and better preservation of visual fields (Maguire et al., 2021). Consequently, ranibizumab is considered a first-line therapy in many clinical guidelines. However, its high acquisition cost has limited widespread adoption, particularly in lowand middle-income countries where the burden of diabetes is rising sharply (Cheloni et al., 2019). These financial constraints have driven interest in alternative anti-VEGF agents, particularly bevacizumab, which is not formally licensed for ocular use but offers comparable pharmacological action at a fraction of the cost. Bevacizumab (Avastin®), originally approved for oncology indications, is a full-length monoclonal antibody targeting all isoforms of VEGF (Ferrara et al., 2010). Its use in ophthalmology emerged from evidence that it effectively reduces retinal vascular permeability, decreases macular thickness, and improves visual acuity in DR and DME (Hirano et al., 2018). Clinicians in many regions now adopt bevacizumab as a cost-effective alternative to ranibizumab, particularly in resource-constrained settings where affordability is central to treatment access (Pennington et al., 2021). Despite widespread off-label use, its repurposing has generated debate regarding regulatory oversight, compounding safety, and comparative durability of outcomes. Several randomized controlled trials have compared ranibizumab and bevacizumab directly, but results remain mixed, with some suggesting equivalent benefit while others reporting modest superiority of one over the other (Schauwvlieghe et al., 2016; Stewart, 2017). The absence of a definitive conclusion continues to challenge evidence-based clinical decision-making. Addressing this gap, systematic reviews and metaanalyses of RCTs provide an essential synthesis of existing data. The present study therefore undertakes a comprehensive meta-analysis to evaluate the relative effectiveness of ranibizumab and bevacizumab in managing diabetic retinopathy, with emphasis on sight restoration and practical implications for clinical care worldwide. 1.1. Research question Essentially, the research question to be addressed in this study follow the PICO (patient/population, intervention, comparison and outcomes) framework as presented below; • Do diabetic retinopathy patients who take Ranibizumab achieve significant sight improvement after treatment? • Do diabetic retinopathy patients who take Bevacizumab record significant sight improvement after treatment? Are diabetic retinopathy patients who take Ranibizumab, compared to those who take Bevacizumab, associated with greater sight improvement after treatment? 1.2. Research Hypothesis In line with the research pursuits above, the null H0 and alternate H1 hypotheses of the study can be established as follows • H0: Neither bevacizumab nor ranibizumab is effective for improved sight. • H1: Bevacizumab is more effective in improving eye sight than ranibizumab and vice versa. 2. Methodology 2.1. Research Methods This study adopted a quantitative approach, employing systematic review and meta-analysis to evaluate the comparative effectiveness of ranibizumab and bevacizumab in treating diabetic retinopathy. Meta-analysis is World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 421 particularly suited for synthesizing evidence from randomized controlled trials (RCTs), improving effect estimates, resolving conflicting findings, and generating generalisable conclusions (Haidich, 2010). Guided by a positivist research philosophy, the study is grounded in the assumptions of ontology, epistemology, and axiology. Ontologically, it assumes a single, observable reality regarding treatment outcomes, while epistemologically, it relies on quantifiable data to establish causal explanations and predictive patterns (Tuli, 2010). From an axiological perspective, the researcher maintains objectivity, minimizing bias to ensure replicable, unbiased results (Saunders et al., 2016; Lindhult, 2019). Positivism, therefore, provides a structured, deductive framework well-suited to hypothesis testing, allowing for measurable, law-like generalisations that inform clinical practice. This paradigm underpins the use of systematic procedures, standardized appraisal of evidence, and statistical modelling to determine which intervention demonstrates greater therapeutic effectiveness. By applying this rigorous, structured approach, the study ensures validity, replicability, and reliability in addressing the central research question on the relative value of ranibizumab and bevacizumab in managing diabetic retinopathy. 2.2. PICO Framework The PICO framework was applied to structure the research questions and ensure that the study followed a design capable of generating robust clinical evidence (Palaskar, 2017). Preliminary searches were conducted on PROSPERO, PubMed, and the Cochrane Library to confirm that no prior systematic review or meta-analysis had addressed the effectiveness of ranibizumab versus bevacizumab in treating diabetic retinopathy (as seen in table 1). PROSPERO was particularly useful as it indexes protocols for systematic reviews and meta-analyses (Vardell and Paulaitis, 2012). Searches, completed on June 10, 2022, used parameters including diabetic retinopathy, ranibizumab, and bevacizumab, without language or regional restrictions. Guided by the PICO framework, the study addressed three key questions: • Do patients with diabetic retinopathy treated with ranibizumab achieve significant visual improvement? • Do patients with diabetic retinopathy treated with bevacizumab achieve significant visual improvement? • Is ranibizumab more effective than bevacizumab in improving sight among diabetic retinopathy patients? Table 1 Explanation of the PICO framework in the context of the research topic Patient/Problem Interventions Core View Outcome The study population consisted of diabetic patients with diabetic retinopathy The Ranibizumab and Bevacizumab treatments are used as intervention The comparison is between the Ranibizumab and Bevacizumab treatments for patients with diabetic retinopathy Greater sight improvement using the Ranibizumab and Bevacizumab and decreased healthcare costs 2.3. Search Strategy A comprehensive search was performed across Google Scholar and PubMed to identify publications on the therapeutic use of ranibizumab and bevacizumab in the treatment of diabetic retinopathy. The main search term applied was “therapeutic intervention in the treatment of diabetic retinopathy with ranibizumab and bevacizumab”, restricted to randomized controlled trials (RCTs), systematic reviews, and full-text articles published between 2012 and 2022. As seen in Figure 1 the initial search yielded 3,200 publications (3,000 from Google Scholar and 200 from PubMed). After removing duplicates and applying preliminary filters (peer-reviewed, RCTs, systematic reviews, full-text availability, and publication date), 1,011 studies remained for screening. Of these, 700 articles were inaccessible due to subscription restrictions, leaving 311 open-access studies. Screening was carried out using Rayyan software, which allowed for systematic and blinded sorting of eligible and ineligible studies (Odubia et al., 2025). Following this process, 150 studies were excluded for inaccurate information, 61 studies for irrelevance to the research scope, and 95 studies for lacking sufficient data to answer the study objectives. At the end of the screening, five RCTs met all inclusion criteria and were selected for final review and analysis (as seen in table 2). Boolean operators were applied to refine searches. For instance, “Diabetes AND Retinopathy” was used to narrow results, while “Diabetic retinopathy OR Diabetic eye disease” ensured broader coverage of relevant studies. Inclusion criteria were: (1) patients diagnosed with diabetic retinopathy, (2) RCTs or systematic reviews, (3) quantitative analyses, (4) studies published in English, and (5) studies published between 2012–2022. Exclusion criteria were: (1) studies not restricted to diabetic retinopathy, (2) qualitative designs, (3) studies published before 2012, and (4) nonEnglish publications. World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 422 Figure 1 PRISMA Flowchart World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 423 3. Data Extraction from RCTs Table 2 Data Extraction from RCTs S/N Study Extracted Elements Database Methods Sentence/Co ncept Full text Results 1 John et al.(2018). . Automating data extraction in systematic reviews: A systematic review’. PubMed “Researchers systematically searched PubMed, IEEEXplore, and the ACM Digital Library to identify potentially relevant articles”. Sentence Full text “Out of a total of 1190 unique citations that met the search criteria, we found 26 published reports describing automatic extraction of at least one of more than 52 potential data elements used in systematic reviews”. 2 Jeffrey et al. (2016) “Comparison of Aflibercept, Bevacizumab, and Ranibizumab for Treatment of Diabetic Macular Edema (DME) Extrapolation of Data to Clinical Practice” “Pubmed and DRCR.net” A systematic review and meta-analysis sentence Full text “On average, all three anti-VEGF agents led to improved visual acuity in eyes, with DME involving the centre of the retina and visual acuity impairment. Worse visual acuity when initiating therapy was associated with greater visual acuity benefit of aflibercept over bevacizumab or ranibizumab one year later”. 3 GrahamRowe et al. (2018) Barriers to and enablers of diabetic retinopathy screening attendance: PROSPERO, MEDLINE, EMBASE, PsycINFO A systematic review of published and grey literature Sentence Full-text “Examples of barriers populating these domains included inaccurate diabetic registers and confusion between routine eye care and retinopathy screening”. 4 Chatziralli (2021) Ranibizumab for the treatment of diabetic retinopathy PubMed, HHS Randomised clinical trial Sentence Full text “Prior to the advent of anti-VEGF agents, patients with DR were followed-up, and their treatment was based on the control of systemic factors and laser photocoagulation” 5 Hirano et al. (2018) changes in plasma vascular endothelial growth factor level after intravitreal injection of bevacizumab, aflibercept, or ranibizumab for diabetic macular edema PubMed Randomised Clinical trial Sentence Full Text “Baseline plasma VEGF level showed no correlations with DR or DME severity, whereas intravitreal injection of bevacizumab or aflibercept significantly reduced plasma VEGF for up to 4 weeks, and ranibizumab produced no such effects” World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 424 3.1. Quality Appraisal The methodological quality of the five randomized controlled trials (RCTs) included in this study was assessed using the Critical Appraisal Skills Programme (CASP) checklist, a widely recognized tool for evaluating intervention studies in healthcare research (as seen in table 3). Each study was appraised against criteria such as randomization, methodological rigor, objectivity of data collection, outcome measurement, and applicability of findings to clinical practice. Attention was given to eliminating systematic bias through randomization, accounting for participant withdrawal, ensuring blinding where possible, and assessing both the accuracy and cost-effectiveness of outcomes. Scores were assigned using a scale of 2 (yes), 1 (cannot tell), and 0 (no), ensuring only high-quality evidence was retained. Overall, the CASP appraisal confirmed that the selected RCTs were reliable, relevant, and of sufficient quality to support the meta-analysis conducted in this dissertation. Table 3 CASP checklist to check the quality of the literature reviewed S/ N Checklist Studies 1 2 3 4 5 Total “Section A: Is the basic study design valid for a randomised controlled trial?” 1 “Did the study address a clearly focused research question?” 2 2 1 2 2 9 2 “Was the assignment of research participants to interventions randomised?” 2 2 2 2 2 10 3 “Were all the participants who entered the study accounted for at its conclusion?” 1 1 1 2 2 7 “Section B: Was the study methodological?” 4 “Was the study methodological?” 2 2 2 2 2 10 5 “Were the study groups similar at the start of the randomised control trial?” 2 2 1 2 2 9 6 “Apart from the experimental intervention, did each study group receive equal care?” 2 2 1 2 2 9 “Section C: What are the results?” 7 “Were the effects of the intervention reported comprehensively?” 2 2 2 2 2 10 8 “Was the precision of the estimate of the intervention or treatment effect reported?” 1 1 1 1 1 5 9 “Do the benefits of the experimental intervention outweigh its harms and costs?” 1 1 1 1 1 5 “Section D: Will the results help locally?” 10 “Can the results be applied to your local population in your context?” 1 2 2 2 2 9 11 “Would the experimental intervention provide greater value to the people in your care than any of the existing interventions?” 1 2 1 2 2 8 Total 17 19 19 20 20 95 World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 425 The total mean score from the five studies was 95 out of a possible 110, calculated as follows: 11 appraisal questions × 5 studies = 55, with a maximum score of 2 per item, giving a potential total of 110. This translates to a quality score of 86.4%, confirming that the articles included in this review are of high methodological quality and relevance. An 86% score indicates that the findings drawn from these studies are consistent, valid, reliable, and applicable to future research and clinical practice. Within the CASP framework, a score of 2 denotes “yes,” 1 indicates “cannot tell,” and 0 represents “no.” By applying this checklist, the current meta-analysis ensured that only robust, evidence-based, and high-quality sources were selected, strengthening the validity of the conclusions presented in this dissertation. 3.2. Data Analysis This study employed Review Manager (RevMan) version 5.4, a statistical tool developed by the Cochrane Collaboration for conducting meta-analyses of randomised controlled trials (RCTs). RevMan was chosen over other statistical packages because it is specifically designed for evidence synthesis, allowing researchers to prepare protocols, input study characteristics, compare data, and generate meta-analytic outputs in both graphical and tabular formats (Usman, 2011; Cochrane, 2022). Unlike more complex software such as R or Stata, RevMan is user-friendly, enabling adjustments to analyses with minimal technical demands, while automatically generating outputs such as funnel plots, sensitivities, and specificities. It also integrates the entire review process from data entry to final report writing, making it particularly efficient for systematic reviews in medical research (Tawfik et al., 2019; ResearchGate, 2022). By using RevMan, this study ensured that the pooled results of RCTs on ranibizumab and bevacizumab in the treatment of diabetic retinopathy were analysed in a consistent, rigorous, and transparent manner, thus enhancing the reliability and interpretability of findings. 3.3. Ethical Considerations This systematic review did not involve direct human participation or the collection of personal data, relying instead on publicly accessible evidence; therefore, ethical risks were minimal. Nevertheless, a quantitative systematic review ethics form was submitted and approved by the University of Chester ethics board to validate the study’s title and design. All included studies were selected strictly based on eligibility criteria, with unpublished or unverifiable data excluded to ensure transparency and reliability. The review process avoided author bias, with findings presented objectively, and all sources were fully acknowledged through proper citation and referencing. 4. Result Figure 2 The statistical results from the meta-analysis (for random effect) Figure 2 Compares the effectiveness of ranibizumab vs. bevacizumab for treating diabetic retinopathy across five RCTs. The odds ratios (OR) for individual studies range from 0.99 to 2.79, with confidence intervals crossing 1, indicating no statistically significant difference in most trials. The pooled effect shows an OR of 1.25 (95% CI: 0.94–1.67, p = 0.12), suggesting a slight, but not statistically significant, trend favoring bevacizumab. Heterogeneity was 0% (I² = 0, p = 0.74), meaning the studies were consistent and comparable. Overall, the results imply that bevacizumab may be more World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 426 effective, but the evidence is not strong enough to reach statistical significance, highlighting the need for further largescale trials. Figure 3 The fixed-effect meta-analysis combined five RCTs The fixed-effect meta-analysis combined five RCTs comparing Ranibizumab and Bevacizumab for diabetic retinopathy (as seen in figure 3). Across studies, Ranibizumab showed slightly higher odds of improving vision, with a pooled odds ratio of 1.26 (95% CI: 0.95–1.67). Although this indicates a 26% greater likelihood of sight restoration, the confidence interval crosses 1, meaning the effect is not statistically significant (p = 0.11). Importantly, the heterogeneity test showed Chi² = 1.97, I² = 0% (p = 0.74), indicating consistent results across studies. Thus, while findings trend in favour of Ranibizumab, the evidence is inconclusive, suggesting no clear superiority over Bevacizumab. Both drugs appear broadly comparable in effectiveness, and further trials with larger samples are needed to establish definitive differences. 4.1. Testing for publication bias Figure 4 Funnel plot to test for the article publication Bias As seen in Figure 4, to assess bias, this study employed a funnel plot, a standard tool in meta-analyses (Lee and Hotopf, 2012). In the absence of bias, studies with higher precision cluster near the mean, while less precise studies scatter symmetrically, forming an inverted funnel (Rao et al., 2017; Sterne and Egger, 2001). As shown in Figure 1, the included World Journal of Biology Pharmacy and Health Sciences, 2025, 23(02), 419-433 427 studies displayed this expected symmetrical funnel shape, indicating no evidence of publication bias in the present review (Tatsioni and Ioannidis, 2017). 4.2. Risk of bias assessment Figure 5 Risk of bias assessment As shown in Figure 2, the majority of the trials demonstrated low risk of bias, with no evidence of systematic flaws such as random sequence generation errors, allocation concealment issues, incomplete outcome reporting, or attrition bias. Importantly, blinding of participants and personnel (performance bias) was adequately maintained across most studies. Nonetheless, the trials by Graham-Rowe et al. (2018) and Hirano et al. (2018) produced inconclusive outcomes, while John et al. (2015), Jeffrey et al. (2016), and Chatziralli (2021) lacked blinding of outcome assessment, introducing potential detection bias. Moreover, some uncertainties persisted in relation to “other sources of bias” not directly captured by the appraisal criteria, as indicated by the yellow markers in Figure 4. In summary, while the overall quality of evidence was robust, a few methodological shortcomings particularly in blinding and inconclusive reporting underscore the need for more rigorously designed RCTs to strengthen the evidence base for bevacizumab and ranibizumab in diabetic retinopathy. 5. Discussion The question of comparative therapeutic effectiveness between ranibizumab and bevacizumab in the treatment of diabetic retinopathy (DR) has provoked sustained debate within ophthalmology. The findings of this meta-analysis, showing a pooled p-value of 0.74 and an overall effect estimate favoring bevacizumab, indicate no statistically significant difference in efficacy, yet the weight of evidence leans towards bevacizumab in terms of clinical impact. The implication is that therapeutic effectiveness is not merely a matter of drug potency, but also of how outcomes are defined, measured, and contextualized across different studies. From a mechanistic standpoint, both drugs function as anti-VEGF agents, inhibiting neovascularization and vascular leakage, which are hallmarks of DR progression (Hirano et al., 2018). Ranibizumab, a smaller antibody fragment, has a shorter systemic half-life, while bevacizumab, a full-length monoclonal antibody, persists longer systemically. Theoretically, this difference might affect not only efficacy but also safety profiles. Yet clinical trial data consistently demonstrate that both agents yield improvements in best-corrected visual acuity (BCVA), reduction of central retinal thickness (CRT), and regression of neovascularization (Martin et al., 2020). The