Anti-VEGF drugs compared with laser photocoagulation for the treatment of diabetic retinopathy: a systematic review and meta-analysis

Article history

The contractual start date for this research was in August 2021. This article began editorial review in November 2023 and was accepted for publication in August 2024. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright statement

Copyright © 2024 Simmonds et al. This work was produced by Simmonds et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2024 Simmonds et al.

Inclusion criteria

All RCTs that recruited people with diabetic retinopathy (proliferative and non-proliferative); patients with a principal indication for treatment of DMO or vitreous haemorrhage were excluded. The technologies of interest were any anti-VEGF therapy, anti-VEGF combined with PRP, PRP alone and sham injection.

A full list of outcomes of interest were reported in the review protocol. This paper focuses particularly on best corrected visual acuity (BCVA), as this was the only outcome reported in all trials. The appendices to this paper report evidence on all protocol-specified outcomes reported in the trials.

Review methods

An Information Specialist (HF) designed a preliminary search strategy in Ovid MEDLINE which consisted of terms for the condition (diabetic retinopathy), that were combined with terms for the intervention (anti-VEGF, angiogenesis inhibitors, or specific drugs used for the treatment of diabetic retinopathy). A RCT study filter was applied. No date or language limits were applied. The final MEDLINE strategy was adapted for use in all resources searched. All search strategies are presented in full in Appendix 1.

The searches were performed on 27 August 2021 and were updated on 13 July 2022 and again on 26 May 2023. The following databases were searched: Ovid MEDLINE(R) ALL, EMBASE (Ovid), Science Citation Index Expanded (Web of Science), Conference Proceedings Citation Index Science (Web of Science), Cochrane Central Register of Controlled Trials [CENTRAL (Wiley)], Cochrane Database of Systematic Reviews (Wiley), Database of Abstracts of Reviews of Effects {DARE [Centre for Reviews and Dissemination (CRD)]}, PROSPERO (CRD) and Epistemonikos. The following trial registries were searched: World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), ClinicalTrials.gov and the EU Clinical Trials Registry. Search results were imported into EndNote 20 (Clarivate Analytics, Philadelphia, PA, USA) and deduplicated.

Two researchers (RW, AL) independently screened all titles and abstracts retrieved for consideration of the full text. The reviewers then screened full texts of potentially eligible studies to determine inclusion. Disagreements were resolved through discussion or with a third reviewer (MS).

A data extraction form was developed and piloted. Data on interventions used, patient characteristics, outcomes reported, and all outcome data were extracted for all included publications. Data extraction was completed by one reviewer and checked by a second (RW, AL). Risk of bias in all included trials was assessed using the Cochrane Risk of Bias 2 tool, focusing on the BCVA outcome, given limited reporting of other outcomes. 18

Statistical analysis

Effect estimates were pooled across trials using standard DerSimonian–Laird random-effect pairwise meta-analyses, according to the duration of follow-up. Heterogeneity was assessed in terms of I²19 and by inspecting the between-study heterogeneity standard deviations (SDs; τ), relative to the treatment effect size.

Network meta-analyses were performed using standard Bayesian methods of NMA in R (version 4.3.1, The R Foundation for Statistical Computing, Vienna, Austria) using the R package multinma (version 0.5.1). 11,20 This extends the standard NMA modelling approach to investigate the potential impact of patient factors (e.g. type of retinopathy) and timing of assessments on the effectiveness of anti-VEGF therapy. 20 Network consistency was checked by comparing the model fit and between-study heterogeneity from the NMA models to an unrelated mean effects model (similar to a model performing direct meta-analysis for each treatment comparison, but with a shared heterogeneity parameter). 21

Visual acuity (BCVA) in diabetic retinopathy is commonly measured using the logarithm of the minimum angle of resolution (log-MAR) and Early Treatment Diabetic Retinopathy Study (ETDRS) scales. As both are widely used, NMAs were performed for both scales. Published data were transformed from one scale to the other, as required. This paper presents results on the log-MAR scale, with ETDRS results reported in the appendices.

The potential impact of unpublished or ongoing trials on the NMAs was investigated using threshold analysis. Threshold analysis investigates where in a NMA results might not be robust to changes in the observed evidence. 22

All R code and data used for this paper are available on GitHub (github.com/marksimmondsyork/AVID).

Patient and public involvement

Patient and clinical representatives were involved in all stages of this project as part of our advisory group including: the funding application, protocol development, discussing the review and its findings, and writing this paper. Further patient and stakeholder involvement was engaged through the NICE committee currently developing guidance on diabetic retinopathy management.

Equality, diversity and inclusion

As this was a review project of existing trial data, we could not account for equality issues in this field beyond what was reported in included publications or data. We note that reporting on potential equality areas such as ethnicity or socioeconomics was limited.

General results

Key findings for BCVA, DMO, vitrectomy, vitreous haemorrhage and adverse events are presented here. A full presentation of all analyses performed for all outcomes is provided in the appendices.

Figure 1 shows the PRISMA flow chart for this review. Studies excluded from the review are listed in Appendix 1. Overall, 14 RCTs were included in the meta-analyses. The searches also identified 21 other RCTs, which were unsuitable for meta-analyses. These included trials reported only as conference abstracts, not in English, published before 2010 (and therefore judged to be out-of-date), that used types of anti-VEGF not in widespread use, or did not include a PRP arm. Those trials therefore could not be reasonably included in the NMAs. These are summarised in Appendix 1.

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram. DMO, diabetic macular edema; VH; vitreous haemorrhage; MA, meta-analysis.

The included RCTs are summarised in Table 1. Trials varied substantially in sample size from only 40 eyes up to just over 400 persons. There were six trials of ranibizumab, five of bevacizumab and three trials of aflibercept. Five trials used anti-VEGF as the intervention, while nine used anti-VEGF combined with PRP. Twelve trials were of patients with proliferative retinopathy. Two trials recruited patients with non-proliferative retinopathy; both evaluated aflibercept. 23,24 Trials of aflibercept and ranibizumab were conducted in Europe, North America or Brazil. All trials of bevacizumab were conducted in the Middle East or South Asia. BCVA was the only outcome reported consistently in all trials.

Risk of bias

For the risk-of-bias assessment of the included trials, see Table 2 and Appendix 1. Overall, four trials were classed at low risk of bias, three moderate and seven at high risk of bias. Risk of bias across individual domains was predominately of ‘some concerns’, primarily due to poor reporting, although larger trials tended to be better reported. Concerns were most common for the outcome measurement domain, due to the lack of masking of participants and outcome assessors. Other concerns included limited description of randomisation and allocation concealment processes, and missing patients and outcome data. The direction of bias was generally unpredictable. Overall, all the trials of bevacizumab were judged to be at high risk of bias. Only the larger trials of ranibizumab and aflibercept were at low risk of bias.

Impact on vision (best corrected visual acuity)

Figure 2 summarises all the data on BCVA for anti-VEGF compared to PRP, as reported across all trials. Results are shown as difference in ETDRS letters between anti-VEGF and control arms. This plot highlights significant variation in the design of the included studies, which precludes combining them all in a standard meta-analysis and demonstrates the need for NMA and meta-regression. First, some trials compare anti-VEGF to PRP directly, while others combine anti-VEGF with PRP, therefore motivating the need for NMA. Second, the time at which BCVA is measured varied enormously across trials, from 1 month to five years. Shorter trials were generally smaller in size, more likely to use bevacizumab and possibly showed larger effect sizes.

FIGURE 2.

All BCVA data (ETDRS letters) from all trials of anti-VEGF.

Network meta-analyses of best corrected visual acuity in proliferative retinopathy

Given the variations in timing at which BCVA results were reported, two NMAs were performed:

1. Analysis up to and including 1 year of follow-up, using the longest follow-up in each trial

2. Analysis only of trials with 1 or 2 years’ follow-up

Note that trials reporting at exactly 1 year were included in both analyses. Given the clinical differences between proliferative and non-proliferative disease, the two trials of non-proliferative disease were not included in the NMA. The network diagrams for both analyses are shown in Figure 3. The size of the circles indicates the number of participants, and the width of the lines and the number of trials. Note that all the trials of bevacizumab combined with PRP had follow-up durations of < 1 year, so are not included in the analyses at 1–2 years. In both networks, there is only one trial of aflibercept and one of bevacizumab (without PRP).

FIGURE 3.

Network diagrams at (a) up to 1 year and (b) 1–2 years.

Figure 4 shows the results of all treatment comparisons from the NMA for data up to 1 year, and Figure 5 for data from 1 to 2 years. Full results of these NMAs are given in Appendix 2. In both figures, the point estimates are shown by the dots, with the horizontal lines being 95% credible intervals (CrIs). Negative relative effects (to the left of the vertical line) indicate favouring the first-named intervention.

FIGURE 4.

Comparison of interventions from NMA of BCVA up to 1 year.

FIGURE 5.

Comparison of interventions from NMA of BCVA from 1 to 2 years.

For the primary comparisons with PRP at up to 1 year, all trials favoured anti-VEGF over PRP and improved vision (reduced log-MAR scores). Changes in log-MAR scores when compared to PRP ranged from −0.078 (or 3.8 ETDRS letters) for ranibizumab with PRP to −0.198 (or 6.8 ETDRS letters) for bevacizumab. Results for aflibercept and bevacizumab (without PRP) were inconclusive because there was only one trial of each. Indirect comparisons between anti-VEGFs found no conclusive evidence that any one anti-VEGF was superior to the others. Heterogeneity across the network appeared to be modest, with an estimated heterogeneity standard error (τ) of 0.04 (95% CrI 0 to 0.14). Results for trials with a follow-up duration of 1–2 years (see Figure 5) were similar to those at up to 1 year, suggesting no obvious trend in treatment effects at up to 2 years.

Given the similarity in magnitude of effect for the various anti-VEGF agents compared to PRP, it is not surprising that the indirect comparisons between agents show no conclusive evidence of difference between any of them. There appears to be no difference between using ranibizumab alone versus ranibizumab combined with PRP, particularly at 2 years.

Treatment rankings are shown in Appendix 2 (Figures 23 and 26). Given the similarity in effect sizes across the different types of anti-VEGF, it is difficult to draw conclusions from the ranking diagrams beyond the fact that PRP alone is likely to be the least effective treatment. The limited data on bevacizumab mean its ranking is very uncertain.

Impact of follow-up time and vision at randomisation

To further examine the impact of follow-up time on the effectiveness of anti-VEGFs, we fitted a range of NMA models including time as a covariate. This meant that all trials could be combined in a single NMA, and whether the effectiveness of anti-VEGFs varied with time could be assessed. Models were also fitted including BCVA at randomisation, to account for possible variation in the effectiveness of the anti-VEGFs with initial vision (see Appendix 2).

Overall, results were very similar to the NMAs at up to 1 year and 1–2 years. Figure 6 shows the effect estimates for anti-VEGFs compared to PRP alone from a model with a linear association between anti-VEGF effect and both follow-up time and BCVA at randomisation. Estimates are presented for 1 year of follow-up and the mean BCVA at baseline across all trials (log-MAR 0.3). The pattern of effect sizes is very similar to that seen in Figures 4 and 5, but CrIs are wider, suggesting that adjusting for follow-up time and baseline BCVA leads to greater uncertainty.

FIGURE 6.

Network meta-analysis of log-MAR with adjustment for follow-up time and BCVA at baseline.

There was no clear evidence that the relative effectiveness of anti-VEGFs varied with time or with vision at randomisation. However, it should be noted that almost all the data were for follow-up times of 2 years or less. Only one trial followed up patients for 5 years, and that found no evidence of difference between anti-VEGF (ranibizumab) and PRP after 5 years. 31

Further network meta-analyses

To further compare the anti-VEGFs to each other, simplified NMAs were performed by combining treatment arms. Two NMAs were performed:

1. Comparing anti-VEGF (of any type), anti-VEGF (any type) combined with PRP and PRP alone

2. Comparing aflibercept, ranibizumab (with or without PRP), bevacizumab (with or without PRP) and PRP alone

In both cases, NMAs included adjustment for follow-up time and BCVA at randomisation. Full results for these NMAs are presented in Appendix 2. In summary, there was good evidence that, when all types of anti-VEGF were combined, anti-VEGF in general improved BCVA when compared to PRP (mean difference −0.089, 95% CrI −0.180 to −0.019), as did anti-VEGF combined with PRP compared to PRP alone (mean difference −0.108, 95% CrI −0.192 to −0.048).

When comparing the three anti-VEGFs (with or without concomitant PRP), there was no clear evidence of any difference in effectiveness between the three types of anti-VEGF; for example, there was no difference between aflibercept and ranibizumab (mean difference −0.003, 95% CI −0.166 to 0.163).

Other outcomes

Results on outcomes other than BCVA were inconsistently reported, with most being reported in no more than three trials. Complete results for these outcomes are presented in Appendix 3. The limited data meant that NMAs were not feasible for these outcomes. A meta-analysis was performed for outcomes reported in two or more trials by assuming that the impact of anti-VEGFs is the same for all types of anti-VEGF, for anti-VEGF alone or in combination with PRP, and at all times up to 2 years. While these are strong assumptions, they may be reasonable given the results observed for BCVA, and the apparent lack of heterogeneity in the data.

Forest plots of neovascularisation of the disc (NVD) and neovascularisation elsewhere (NVE) are shown in Appendix 3. These suggest that neovascularisation was reduced while using anti-VEGF. The results of meta-analyses for other non-vision outcomes are shown in Figure 7. Although data were limited, the results suggest that anti-VEGF treatment substantially reduces the rate of macular oedema (DMO), the need for vitrectomy and reduces the rate of vitreous haemorrhage. No data on progression of diabetic retinopathy were reported.

FIGURE 7.

Meta-analysis of non-vision outcomes. RR, relative risk.

Adverse events

As with non-BCVA outcomes, adverse events were not widely reported, with little consistency across trials as to which adverse events were reported. A meta-analysis was performed for adverse event types reported in two or more trials by assuming that the impact of anti-VEGFs is the same for all types of anti-VEGF, for anti-VEGF alone or in combination with PRP, and at all times up to 2 years.

The meta-analysis results are shown in Figure 8. Due to the small numbers of events, and limited numbers of trials reported each adverse event, most results are inconclusive. Anti-VEGF appeared to reduce the incidence of retinal detachment. It appeared to increase the rate of ocular pain, but it was unclear whether this was procedure-related or post-intervention pain. Full results are presented in Appendix 3.

FIGURE 8.

Meta-analyses of adverse event outcomes.

Non-proliferative retinopathy

Two trials compared aflibercept to sham injection in patients with non-proliferative retinopathy with a follow-up of 2 years (PANORAMA and Protocol W). Meta-analysis of their BCVA results found no clear evidence of any benefit of aflibercept over sham injection [mean difference (log-MAR) −0.02, 95% CI −0.05 to 0.01]. Progression to macular oedema was the only other outcome reported by both trials, with strong evidence to suggest that aflibercept reduces the risk of macular oedema [relative risk (RR) 0.283, 95% CI 0.18 to 0.44]. Protocol W reported that aflibercept reduced the rate of vitrectomy compared to sham injection (RR 0.38, 95% CI 0.24 to 0.60). Full results are presented in Appendix 4.

Protocol W found that aflibercept slowed the rate of progression to proliferative retinopathy when compared to sham injection (hazard ratio 0.40, 97.5% CI 0.28 to 0.57). PANORAMA found that more patients on aflibercept experienced a 2 point or more improvement in Diabetic Retinopathy Severity Scale (DRSS) (RR 4.41, 95% CI 2.81 to 6.94).

Patient and public perspectives

Patient representatives noted several key areas of continued concern. Most critically was that most trials of anti-VEGF used BCVA as their primary outcome, without any consideration of how that impacted on quality of life, ability to work, drive or care for family. The lack of long-term evidence also raised concerns because there is substantial uncertainty about how PDR will be managed and treated long term.

CRediT contribution statement

Mark Simmonds (https://orcid.org/0000-0002-1999-8515): Conceptualisation (lead), Data curation (lead), Formal analysis (lead), Funding acquisition (lead), Investigation (lead), Methodology (lead), Project administration, Writing (lead).

Alexis Llewellyn (https://orcid.org/0000-0003-4569-5136): Conceptualisation, Data curation, Formal analysis, Funding acquisition, Investigation (co-lead), Methodology, Writing.

Ruth Walker (https://orcid.org/0000-0003-2765-7363): Conceptualisation, Data curation, Formal analysis, Funding acquisition, Investigation, Writing.

Helen Fulbright (https://orcid.org/0000-0002-1073-1099): Investigation, Methodology.

Matthew Walton (https://orcid.org/0000-0003-1932-3689): Conceptualisation, Funding acquisition, Writing.

Rob Hodgson (https://orcid.org/0000-0001-6962-2893): Conceptualisation, Funding acquisition, Writing.

Laura Bojke (https://orcid.org/0000-0001-7921-9109): Conceptualisation, Funding acquisition, Writing.

Lesley Stewart (https://orcid.org/0000-0003-0287-4724): Conceptualisation, Funding acquisition, Writing.

Sofia Dias (https://orcid.org/0000-0002-2172-0221): Conceptualisation, Funding acquisition, Methodology, Writing.

Thomas Rush: Conceptualisation, Funding acquisition, Writing (patient and public involvement advisor).

John G Lawrenson: Conceptualisation, Funding acquisition, Writing.

Tunde Peto (https://orcid.org/0000-0001-6265-0381): Conceptualisation, Funding acquisition, Writing.

David Steel (https://orcid.org/0000-0001-8734-3089): Conceptualisation, Funding acquisition, Writing.

Acknowledgements

We acknowledge the help and advice given by all persons involved in the NICE diabetic retinopathy guidance development process.

Data-sharing statement

Data and code to reproduce the meta-analyses are available on GitHub (https://github.com/marksimmondsyork/AVID). For all other data requests please contact the corresponding author.

Ethics statement

As this was a systematic review of existing published data, no ethics approval was required.

Information governance statement

All data used in this paper were taken from published sources: no personal data were included.

Disclosure of interests

Full disclosure of interests: Completed ICMJE forms for all authors, including all related interests, are available in the toolkit on the NIHR Journals Library report publication page at https://doi.org/10.3310/PCGV5709.

Primary conflicts of interest: Laura Bojke declares that she was on the HS&DR Researcher-Led awards panel (December 2019–December 2022). All other authors have no conflicts of interest to declare.

Department of Health and Social Care disclaimer

This publication presents independent research commissioned by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, NIHR Coordinating Centre, the Health Technology Assessment programme or the Department of Health and Social Care.

This article was published based on current knowledge at the time and date of publication. NIHR is committed to being inclusive and will continually monitor best practice and guidance in relation to terminology and language to ensure that we remain relevant to our stakeholders.

Study registration

This study is registered as PROSPERO (CRD42021272642).

Funding

This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR132948.

About this article

The contractual start date for this research was in August 2021. This article began editorial review in November 2023 and was accepted for publication in August 2024. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright

Copyright © 2024 Simmonds et al. This work was produced by Simmonds et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

Database search strategies

The aim of the literature search was to identify RCTs on anti-VEGFs, angiogenesis inhibitors and other specific drugs used for the treatment of diabetic retinopathy.

An Information Specialist (HF) designed a preliminary search strategy in Ovid MEDLINE in consultation with the research team. The strategy consisted of terms for the condition (diabetic retinopathy), which were combined with terms for the intervention (anti-VEGF, angiogenesis inhibitors, or specific drugs used for the treatment of diabetic retinopathy) using the Boolean operator AND. Text word searches for terms appearing in the title and abstracts of database records were included in the strategy alongside searches of relevant subject headings. A RCT study filter was applied using the Boolean operator AND. No date or language limits were applied. The final MEDLINE strategy was adapted for use in all resources searched.

The searches were performed on 27 August 2021. The following databases were searched: Ovid MEDLINE(R) ALL, EMBASE (Ovid), Science Citation Index Expanded (Web of Science), Conference Proceedings Citation Index Science (Web of Science), Cochrane CENTRAL (Wiley), Cochrane Database of Systematic Reviews (Wiley), DARE (CRD), PROSPERO (CRD) and Epistemonikos. The following trial registries were searched: WHO ICTRP, ClinicalTrials.gov, and the EU Clinical Trials Registry.

Search results were imported into EndNote 20 and deduplicated. All search strategies are presented in full below.

The searches were updated on 13 July 2022 and again on 26 May 2023 using all the databases and strategies as used previously, except for DARE as this database is no longer updated. For each update search, the results of the databases were deduplicated against each other in a separate EndNote 20 Library before being merged with the results of the original EndNote Library and deduplicated for a second time.

Ovid MEDLINE(R) ALL

(Includes Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE)

via Ovid http://ovidsp.ovid.com/

Date range searched: <1946–25 May 2023>

Date searched: 26 May 2023

Records retrieved: 3172

The MEDLINE strategy below includes a search filter to limit retrieval to RCTs using the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity-maximising version (2008 revision); Ovid format.

Lefebvre C, Glanville J, Briscoe S, Littlewood A, Marshall C, Metzendorf MI, et al. Technical Supplement to Chapter 4: Searching for and Selecting Studies. In Higgins JPT, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, Welch VA, editors. Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from: www.training.cochrane.org/handbook.

1. (*Diabetes Mellitus/ or *Diabetes Complications/) and exp *Retinal Diseases/ (3199)

2. Diabetic Retinopathy/ (29304)

3. ((diabet* or DM) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath* or maculopath*)).ti,ab,kw. (30685)

4. (((proliferat* or PDR or pre-proliferat* or preproliferat* or non-proliferat* or nonproliferat* or NPDR or background) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath*)) and (diabet* or DM)).ti,ab,kw. (7895)

5. (new blood vessel* and diabet*).ti,ab,kw. (273)

6. (((retin* or subretina* or sub-retina* or interretina* or inter-retina* or vitreoretin* or vitreo-retin* or chorioretin* or chorio-retin* or choroid* or macula* or intraocular or intra-ocular or intravitreal or intra-vitreal) adj4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*).ti,ab,kw. (13654)

7. ((retinal vein* adj3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*).ti,ab,kw. (1473)

8. or/1‐7 (44519)

9. exp Vascular Endothelial Growth Factors/ai (9366)

10. exp Receptors, Vascular Endothelial Growth Factor/ai (3393)

11. (anti adj2 VEGF*).ti,ab,kw. (9210)

12. (anti-VEGF* or antiVEGF*).ti,ab,kw. (9455)

13. ((anti vascular or anti-vascular or antivascular) adj2 endothelial growth factor*).ti,ab,kw. (5745)

14. (((vascular endothelial adj2 growth factor*) or vasculotropin or VEGF* or vascular permeability factor* or VPF) adj2 (trap* or inhibit* or antagonist*)).ti,ab,kw. (11005)

15. (vascular proliferation adj4 inhibit*).ti,ab,kw. (38)

16. or/9‐15 (28125)

17. Angiogenesis Inhibitors/ (28876)

18. exp Angiogenesis Inducing Agents/ai (118)

19. (angiogen* adj2 (antagonist* or inhibit*)).ti,ab,kw. (14831)

20. ((antiangiogen* or anti angiogen* or anti-angiogen*) adj2 (agent* or drug* or effect*)).ti,ab,kw. (10949)

21. (angiostatic adj2 (agent* or drug*)).ti,ab,kw. (103)

22. ((neovasculari?ation or vasculari?ation) adj2 inhibit*).ti,ab,kw. (1243)

23. or/17‐22 (45139)

24. Aflibercept*.ti,ab,kw,rn. (3315)

25. (Eylea or Zaltrap or Ziv-Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005).ti,ab,kw. (316)

26. Bevacizumab/ (14139)

27. Bevacizumab*.ti,ab,kw,rn. (22533)

28. (Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb-VEGF or rhuMAb VEGF or “NSC 704865” or NSC704865).ti,ab,kw. (1675)

29. (IVB adj2 inject*).ti,ab,kw. (316)

30. Ranibizumab/ (4684)

31. Ranibizumab*.ti,ab,kw,rn. (6307)

32. (Lucentis or “rhuFab V2”).ti,ab,kw. (456)

33. (IVR adj2 inject*).ti,ab,kw. (139)

34. Pegaptanib*.ti,ab,kw,rn. (671)

35. (“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838).ti,ab,kw. (140)

36. or/24‐35 (28353)

37. 8 and (16 or 23 or 36) (4979)

38. randomized controlled trial.pt. (593242)

39. controlled clinical trial.pt. (95314)

40. randomized.ab. (604126)

41. placebo.ab. (238387)

42. drug therapy.fs. (2592996)

43. randomly.ab. (408822)

44. trial.ab. (649200)

45. groups.ab. (2520111)

46. or/38‐45 (5663345)

47. 37 and 46 (3308)

48. exp animals/ not humans.sh. (5123796)

49. 47 not 48 (3190)

50. limit 49 to yr=“2000-Current” (3182)

51. remove duplicates from 50 (3172)

Key:

/ or.sh. = indexing term (Medical Subject Heading: MeSH)

/ai = indexing term with subheading for antagonists & inhibitors

exp = exploded indexing term (MeSH)

* or $ = truncation

? = adds up to 1 additional character

ti,ab,kw = terms in either title, abstract or keyword fields

rn = registry number/name of substance

adj3 = terms within three words of each other (any order).

pt = publication type

fs = floating sub-heading

EMBASE

via Ovid http://ovidsp.ovid.com/

Date range searched: <1974–25 May 2023>

Date searched: 26 May 2023

Records retrieved: 2558

The EMBASE strategy below includes the Cochrane EMBASE RCT filter (Ovid format).

Glanville J, Foxlee R, Wisniewski S, Noel-Storr A, Edwards M, Dooley G. Translating the Cochrane EMBASE RCT filter from the Ovid interface to EMBASE.com: a case study. Health Info Libr J. 2019. doi:10.1111/hir.12269

1. *diabetes mellitus/ and exp *retina disease/ (4826)

2. exp diabetic retinopathy/ (53891)

3. ((diabet* or DM) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath* or maculopath*)).ti,ab,kw. (43573)

4. (((proliferat* or PDR or pre-proliferat* or preproliferat* or non-proliferat* or nonproliferat* or NPDR or background) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath*)) and (diabet* or DM)).ti,ab,kw. (11148)

5. (new blood vessel* and diabet*).ti,ab,kw. (391)

6. (((retin* or subretina* or sub-retina* or interretina* or inter-retina* or vitreoretin* or vitreo-retin* or chorioretin* or chorio-retin* or choroid* or macula* or intraocular or intra-ocular or intravitreal or intra-vitreal) adj4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*).ti,ab,kw. (20734)

7. ((retinal vein* adj3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*).ti,ab,kw. (2199)

8. or/1‐7 (70501)

9. vasculotropin inhibitor/ (7663)

10. (anti adj2 VEGF*).ti,ab,kw. (15751)

11. (anti-VEGF* or antiVEGF*).ti,ab,kw. (16291)

12. ((anti vascular or anti-vascular or antivascular) adj2 endothelial growth factor*).ti,ab,kw. (7400)

13. (((vascular endothelial adj2 growth factor*) or vasculotropin or VEGF* or vascular permeability factor* or VPF) adj2 (trap* or inhibit* or antagonist*)).ti,ab,kw. (17346)

14. (vascular proliferation adj4 inhibit*).ti,ab,kw. (50)

15. or/9‐14 (38838)

16. angiogenesis inhibitor/ (20415)

17. (angiogen* adj2 (antagonist* or inhibit*)).ti,ab,kw. (20444)

18. ((antiangiogen* or anti angiogen* or anti-angiogen*) adj2 (agent* or drug* or effect*)).ti,ab,kw. (15734)

19. (angiostatic adj2 (agent* or drug*)).ti,ab,kw. (125)

20. ((neovasculari?ation or vasculari?ation) adj2 inhibit*).ti,ab,kw. (1718)

21. or/16‐20 (45260)

22. aflibercept/ (8877)

23. Aflibercept*.ti,ab,kw,dy,tn. (9141)

24. (Eylea or Zaltrap or Ziv-Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005).ti,ab,dy,tn. (1741)

25. bevacizumab/ (72890)

26. Bevacizumab*.ti,ab,kw,dy,tn. (75152)

27. (Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb-VEGF or rhuMAb VEGF or “NSC 704865” or NSC704865).ti,ab,kw,dy,tn. (11007)

28. (IVB adj2 inject*).ti,ab,kw. (395)

29. ranibizumab/ (12442)

30. Ranibizumab*.ti,ab,kw,dy,tn. (12826)

31. (Lucentis or “rhuFab V2”).ti,ab,kw,dy,tn. (3216)

32. (IVR adj2 inject*).ti,ab,kw. (197)

33. pegaptanib.dy,tn. (2470)

34. Pegaptanib*.ti,ab,kw,dy,tn. (2544)

35. (“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838).ti,ab,kw,dy,tn. (1266)

36. or/22‐35 (85594)

37. 8 and (15 or 21 or 36) (8778)

38. randomized controlled trial/ (785964)

39. controlled clinical trial/ (469252)

40. Random$.ti,ab,ot. (1968994)

41. randomization/ (99178)

42. intermethod comparison/ (297283)

43. placebo.ti,ab,ot. (366311)

44. (compare or compared or comparison).ti,ot. (604093)

45. ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. (2766233)

46. (open adj label).ti,ab,ot. (109016)

47. ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab,ot. (274477)

48. double blind procedure/ (210575)

49. parallel group$1.ti,ab,ot. (32223)

50. (crossover or cross over).ti,ab,ot. (124540)

51. ((assign$ or match or matched or allocation) adj5 (alternate or group or groups or intervention or interventions or patient or patients or subject or subjects or participant or participants)).ti,ab,ot. (415063)

52. (assigned or allocated).ti,ab,ot. (489023)

53. (controlled adj7 (study or design or trial)).ti,ab,ot. (450984)

54. (volunteer or volunteers).ti,ab,ot. (282270)

55. human experiment/ (650911)

56. trial.ti,ot. (403295)

57. or/38‐56 (6311902)

58. 37 and 57 (2810)

59. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$).ti,ot. and animal experiment/ (1227092)

60. animal experiment/ not (human experiment/ or human/) (2577203)

61. 59 or 60 (2645661)

62. 58 not 61 (2689)

63. limit 62 to yr=“2000-Current” (2686)

64. remove duplicates from 63 (2558)

Key:

/ or.sh. = indexing term (Emtree Subject Heading)

exp = exploded indexing term (Emtree)

* or $ = truncation

? = adds up to 1 additional character

ti,ab,kw = terms in either title or abstract fields

dy,tn = drug index terms word or drug trade name fields

adj3 = terms within three words of each other (any order).

pt = publication type

ot = original title

Cochrane Central Register of Controlled Trials

via Wiley http://onlinelibrary.wiley.com/

Date range searched: Issue 5 of 12, May 2023

Date searched: 26 May 2023

Records retrieved: 1825

1. ([mh ^“Diabetes Mellitus”] or [mh ^“Diabetes Complications”]) and [mh “Retinal Diseases”] 250

2. [mh ^“Diabetic Retinopathy”] 1934

3. ((diabet* or DM) NEAR/3 (retinopath* or vitreoretinopath* or chorioretinopath* or maculopath*)):ti,ab,kw 4547

4. (((proliferat* or PDR or preproliferat* or nonproliferat* or NPDR or background) NEAR/3 (retinopath* or vitreoretinopath* or chorioretinopath*)) and (diabet* or DM)):ti,ab,kw 1326

5. (“new blood” NEXT vessel* and diabet*):ti,ab,kw 32

6. (((retin* or subretina* or interretina* or vitreoretin* or chorioretin* or choroid* or macula* or intraocular or intravitreal) NEAR/4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*):ti,ab,kw 3457

7. ((retinal NEXT vein* NEAR/3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*):ti,ab,kw 254

8. {OR #1-#7} 5751

9. [mh “Vascular Endothelial Growth Factors”/ai] 758

10. [mh “Receptors, Vascular Endothelial Growth Factor”/ai] 154

11. (anti NEAR/2 VEGF*):ti,ab,kw 1610

12. (antiVEGF*):ti,ab,kw 1523

13. ((anti NEXT vascular or antivascular) NEAR/2 “endothelial growth” NEXT factor*):ti,ab,kw 699

14. (((“vascular endothelial” NEAR/2 growth NEXT factor*) or vasculotropin or VEGF* or “vascular permeability” NEXT factor* or VPF) NEAR/2 (trap* or inhibit* or antagonist*)):ti,ab,kw 2048

15. (“vascular proliferation” NEAR/4 inhibit*):ti,ab,kw 1

16. {OR #9-#15} 3671

17. [mh ^“Angiogenesis Inhibitors”] 1681

18. [mh “Angiogenesis Inducing Agents”/ai] 0

19. (angiogen* NEAR/2 (antagonist* or inhibit*)):ti,ab,kw 2126

20. ((antiangiogen* or anti NEXT angiogen*) NEAR/2 (agent* or drug* or effect*)):ti,ab,kw 717

21. (angiostatic NEAR/2 (agent* or drug*)):ti,ab,kw 10

22. ((neovasculari?ation or vasculari?ation) NEAR/2 inhibit*):ti,ab,kw 37

23. {OR #17-#22}2691

24. Aflibercept*:ti,ab,kw 1081

25. (Eylea or Zaltrap or Ziv NEXT Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005):ti,ab,kw 252

26. [mh ^Bevacizumab] 2633

27. Bevacizumab*:ti,ab,kw 7386

28. (Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb NEXT VEGF or “NSC 704865” or NSC704865):ti,ab,kw 941

29. (IVB NEAR/2 inject*):ti,ab,kw 89

30. [mh ^Ranibizumab] 1049

31. Ranibizumab*:ti,ab,kw 2266

32. (Lucentis or “rhuFab V2”):ti,ab,kw 451

33. (IVR NEAR/2 inject*):ti,ab,kw 32

34. Pegaptanib*:ti,ab,kw 166

35. (“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838):ti,ab,kw 82

36. {OR #24-#35}10087

37. #8 and (#16 or #23 or #36) 1847

38. (rat or rats or rodent* or mouse or mice or “mus musculus” or “mus domesticus” or murine or murinae or bovine or sheep or ovine or “ovis aries” or porcine):ti,ab,kw 17188

39. #37 not #38 with Publication Year from 2000 to 2023, in Trials 1825

Science Citation Index Expanded

via Web of Science, Clarivate Analytics https://clarivate.com/

Date range searched: 1900–26 May 2023

Date searched: 26 May 2023

Records retrieved: 2394

1. 32 #29 NOT #30 2,394 Limited by 2000-01-01 to 2023-05-26

2. 31 #29 NOT #30 2,410

3. 30 TI=(animal or animals or rat or rats or rodent* or mouse or mice or “mus musculus” or “mus domesticus” or murine or murinae or porcine or pig or pigs or piglet or piglets or sow or sows or minipig or minipigs or sheep or ovine or “ovis aries” or lamb or lambs or ewe or ewes or rabbit or rabbits or leporide or leporidae or kitten or kittens or dog or dogs or puppy or puppies or monkey or monkeys or horse or horses or foal or foals or equine or bovine or calf or calves or cattle or heifer or heifers or hamster or hamsters or chicken or chickens or livestock or alpaca* or llama*) 3,259,653

4. 29 #27 AND #28 2,524

5. 28 TS=(random* or control* or trial* or “single blind” or “double blind” or “triple blind” or placebo)8,083,064

6. 27 #6 AND #26 6,121

7. 26 #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 83,065

8. 25 TS=(“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838) 142

9. 24 TS=(Pegaptanib*) 716

10. 23 TS=(IVR NEAR/2 inject*) 177

11. 22 TS=(Lucentis or “rhuFab V2”) 564

12. 21 TS=(Ranibizumab*) 9,347

13. 20 TS=(IVB NEAR/2 inject*) 307

14. 19 TS=(Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb-VEGF or “rhuMAb VEGF” or “NSC 704865” or NSC704865) 3,355

15. 18 TS=(Bevacizumab*) 36,279

16. 17 TS=(Eylea or Zaltrap or Ziv-Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005) 320

17. 16 TS=(Aflibercept*) 4,076

18. 15 TS=((neovascularisation or neovascularization or vascularisation or vascularization) NEAR/2 inhibit*) 1,858

19. 14 TS=(angiostatic NEAR/2 (agent* or drug*)) 105

20. 13 TS=((antiangiogen* or “anti angiogen*” or anti-angiogen*) NEAR/2 (agent* or drug* or effect*)) 11,802

21. 12 TS=(angiogen* NEAR/2 (antagonist* or inhibit*)) 19,846

22. 11 TS=(“vascular proliferation” NEAR/4 inhibit*) 44

23. 10 TS=(((“vascular endothelial” NEAR/2 “growth factor*”) or vasculotropin or VEGF* or “vascular permeability factor*” or VPF) NEAR/2 (trap* or inhibit* or antagonist*)) 14,540

24. 9 TS=((“anti vascular” or anti-vascular or antivascular) NEAR/2 “endothelial growth factor*”) 5,018

25. 8 TS=(anti-VEGF* or antiVEGF*) 10,111

26. 7 TS=(anti NEAR/2 VEGF*) 10,549

27. 6 #1 OR #2 OR #3 OR #4 OR #5 43,073

28. 5 TS=((“retinal vein*” NEAR/3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*) 1,546

29. 4 TS=(((retin* or subretina* or sub-retina* or interretina* or inter-retina* or vitreoretin* or vitreo-retin* or chorioretin* or chorio-retin* or choroid* or macula* or intraocular or intra-ocular or intravitreal or intra-vitreal) NEAR/4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*) 16,980

30. 3 TS=(“new blood vessel*” and diabet*) 288

31. 2 TS=(((proliferat* or PDR or pre-proliferat* or preproliferat* or non-proliferat* or nonproliferat* or NPDR or background) NEAR/3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath*)) and (diabet* or DM)) 7,763

32. 1 TS=((diabet* or DM) NEAR/3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath* or maculopath*)) 36,053

Key:

TS= terms in either title, abstract, author keywords, and keywords plus fields

TI= search in title field

NEAR/3 = terms within three words of each other (any order).

* = truncation

Conference Proceedings Citation Index – Science

via Web of Science, Clarivate Analytics https://clarivate.com/

Date range searched: 1990–26 May 2023

Date searched: 26 May 2023

Records retrieved: 86

1. 32 #29 NOT #30 86 Limited by 2000-01-01 to 2023-05-26

2. 31 #29 NOT #30 89

3. 30 TI=(animal or animals or rat or rats or rodent* or mouse or mice or “mus musculus” or “mus domesticus” or murine or murinae or porcine or pig or pigs or piglet or piglets or sow or sows or minipig or minipigs or sheep or ovine or “ovis aries” or lamb or lambs or ewe or ewes or rabbit or rabbits or leporide or leporidae or kitten or kittens or dog or dogs or puppy or puppies or monkey or monkeys or horse or horses or foal or foals or equine or bovine or calf or calves or cattle or heifer or heifers or hamster or hamsters or chicken or chickens or livestock or alpaca* or llama*) 295,290

4. 29 #27 AND #28 92

5. 28 TS=(random* or control* or trial* or “single blind” or “double blind” or “triple blind” or placebo) 1,616,551

6. 27 #6 AND #26 458

7. 26 #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 8,998

8. 25 TS=(“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838) 14

9. 24 TS=(Pegaptanib*) 39

10. 23 TS=(IVR NEAR/2 inject*) 1

11. 22 TS=(Lucentis or “rhuFab V2”) 29

12. 21 TS=(Ranibizumab*) 564

13. 20 TS=(IVB NEAR/2 inject*) 7

14. 19 TS=(Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb-VEGF or “rhuMAb VEGF” or “NSC 704865” or NSC704865) 196

15. 18 TS=(Bevacizumab*) 4,659

16. 17 TS=(Eylea or Zaltrap or Ziv-Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005) 60

17. 16 TS=(Aflibercept*) 577

18. 15 TS=((neovascularisation or neovascularization or vascularisation or vascularization) NEAR/2 inhibit*) 177

19. 14 TS=(angiostatic NEAR/2 (agent* or drug*)) 6

20. 13 TS=((antiangiogen* or “anti angiogen*” or anti-angiogen*) NEAR/2 (agent* or drug* or effect*)) 634

21. 12 TS=(angiogen* NEAR/2 (antagonist* or inhibit*)) 1,209

22. 11 TS=(“vascular proliferation” NEAR/4 inhibit*) 6

23. 10 TS=(((“vascular endothelial” NEAR/2 “growth factor*”) or vasculotropin or VEGF* or “vascular permeability factor*” or VPF) NEAR/2 (trap* or inhibit* or antagonist*)) 1,025

24. 9 TS=((“anti vascular” or anti-vascular or antivascular) NEAR/2 “endothelial growth factor*”) 224

25. 8 TS=(anti-VEGF* or antiVEGF*) 836

26. 7 TS=(anti NEAR/2 VEGF*) 869

27. 6 #1 OR #2 OR #3 OR #4 OR #5 5,826

28. 5 TS=((“retinal vein*” NEAR/3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*) 74

29. 4 TS=(((retin* or subretina* or sub-retina* or interretina* or inter-retina* or vitreoretin* or vitreo-retin* or chorioretin* or chorio-retin* or choroid* or macula* or intraocular or intra-ocular or intravitreal or intra-vitreal) NEAR/4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*) 2,140

30. 3 TS=(“new blood vessel*” and diabet*) 29

31. 2 TS=(((proliferat* or PDR or pre-proliferat* or preproliferat* or non-proliferat* or nonproliferat* or NPDR or background) NEAR/3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath*)) and (diabet* or DM)) 642

32. 1 TS=((diabet* or DM) NEAR/3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath* or maculopath*)) 4,723

Key:

TS= terms in either title, abstract, author keywords, and keywords plus fields

TI= search in title field

NEAR/3 = terms within three words of each other (any order).

* = truncation

Cochrane Database of Systematic Reviews

via Wiley http://onlinelibrary.wiley.com/

Date range searched: Issue 5 of 12, May 2023

Date searched: 26 May 2023

Records retrieved: 14

1. ([mh ^“Diabetes Mellitus”] or [mh ^“Diabetes Complications”]) and [mh “Retinal Diseases”] 250

2. [mh ^“Diabetic Retinopathy”] 1934

3. ((diabet* or DM) NEAR/3 (retinopath* or vitreoretinopath* or chorioretinopath* or maculopath*)):ti,ab,kw 4547

4. (((proliferat* or PDR or preproliferat* or nonproliferat* or NPDR or background) NEAR/3 (retinopath* or vitreoretinopath* or chorioretinopath*)) and (diabet* or DM)):ti,ab,kw 1326

5. (“new blood” NEXT vessel* and diabet*):ti,ab,kw 32

6. (((retin* or subretina* or interretina* or vitreoretin* or chorioretin* or choroid* or macula* or intraocular or intravitreal) NEAR/4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovasculari?ation*)) and diabet*):ti,ab,kw 3457

7. ((retinal NEXT vein* NEAR/3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*):ti,ab,kw 254

8. {OR #1-#7} 5751

9. [mh “Vascular Endothelial Growth Factors”/ai] 758

10. [mh “Receptors, Vascular Endothelial Growth Factor”/ai] 154

11. (anti NEAR/2 VEGF*):ti,ab,kw 1610

12. (antiVEGF*):ti,ab,kw 1523

13. ((anti NEXT vascular or antivascular) NEAR/2 “endothelial growth” NEXT factor*):ti,ab,kw 699

14. (((“vascular endothelial” NEAR/2 growth NEXT factor*) or vasculotropin or VEGF* or “vascular permeability” NEXT factor* or VPF) NEAR/2 (trap* or inhibit* or antagonist*)):ti,ab,kw 2048

15. (“vascular proliferation” NEAR/4 inhibit*):ti,ab,kw 1

16. {OR #9-#15} 3671

17. [mh ^“Angiogenesis Inhibitors”] 1681

18. [mh “Angiogenesis Inducing Agents”/ai] 0

19. (angiogen* NEAR/2 (antagonist* or inhibit*)):ti,ab,kw 2126

20. ((antiangiogen* or anti NEXT angiogen*) NEAR/2 (agent* or drug* or effect*)):ti,ab,kw 717

21. (angiostatic NEAR/2 (agent* or drug*)):ti,ab,kw 10

22. ((neovasculari?ation or vasculari?ation) NEAR/2 inhibit*):ti,ab,kw 37

23. {OR #17-#22} 2691

24. Aflibercept*:ti,ab,kw 1081

25. (Eylea or Zaltrap or Ziv NEXT Aflibercept or “AVE 0005” or AVE0005 or “AVE 005” or AVE005):ti,ab,kw 252

26. [mh ^Bevacizumab] 2633

27. Bevacizumab*:ti,ab,kw 7386

28. (Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb NEXT VEGF or “NSC 704865” or NSC704865):ti,ab,kw 941

29. (IVB NEAR/2 inject*):ti,ab,kw 89

30. [mh ^Ranibizumab] 1049

31. Ranibizumab*:ti,ab,kw 2266

32. (Lucentis or “rhuFab V2”):ti,ab,kw 451

33. (IVR NEAR/2 inject*):ti,ab,kw 32

34. Pegaptanib*:ti,ab,kw 166

35. (“EYE 001” or EYE001 or Macugen or “NX 1838” or NX1838):ti,ab,kw 82

36. {OR #24-#35} 10087

37. #8 and (#16 or #23 or #36) 1847

38. (rat or rats or rodent* or mouse or mice or “mus musculus” or “mus domesticus” or murine or murinae or bovine or sheep or ovine or “ovis aries” or porcine):ti,ab,kw 17188

#39 #37 not #38 with Cochrane Library publication date between January 2000 and May 2023, in Cochrane Reviews 14

Key:

mh = exploded indexing term (MeSH)

mh ^ = unexploded indexing term (MeSH)

/ai = indexing term with subheading for antagonists & inhibitors

* = truncation or additional characters within a word

? = adds up to 1 additional character

ti,ab,kw = terms in either title or abstract or keyword fields

near/3 = terms within three words of each other (any order)

next = terms are next to each other

Epistemonikos

via https://www.epistemonikos.org/

Date range searched: Inception – 26 May 2023

Date searched: 26 May 2023

Records retrieved: 1026

((title:((title:(((diabet* OR proliferat* OR PDR OR pre-proliferat* OR preproliferat* OR non-proliferat* OR nonproliferat* OR NPDR OR background) AND retinopath*)) OR abstract:(((diabet* OR proliferat* OR PDR OR pre-proliferat* OR preproliferat* OR non-proliferat* OR nonproliferat* OR NPDR OR background) AND retinopath*))) OR (title:((new blood vessel* AND diabet*)) OR abstract:((new blood vessel* AND diabet*)))) OR abstract:((title:(((diabet* OR proliferat* OR PDR OR pre-proliferat* OR preproliferat* OR non-proliferat* OR nonproliferat* OR NPDR OR background) AND retinopath*)) OR abstract:(((diabet* OR proliferat* OR PDR OR pre-proliferat* OR preproliferat* OR non-proliferat* OR nonproliferat* OR NPDR OR background) AND retinopath*))) OR (title:((new blood vessel* AND diabet*)) OR abstract:((new blood vessel* AND diabet*))))) AND (title:((anti AND VEGF*)) OR abstract:((anti AND VEGF*))) OR (title:((anti-VEGF* OR antiVEGF*)) OR abstract:((anti-VEGF* OR antiVEGF*))) OR (title:(((“anti vascular” OR anti-vascular OR antivascular) AND “endothelial growth factor”)) OR abstract:(((“anti vascular” OR anti-vascular OR antivascular) AND “endothelial growth factor”))) OR (title:(((“vascular endothelial growth factor” OR vasculotropin OR VEGF* OR “vascular permeability factor” OR VPF) AND (trap* OR inhibit* OR antagonist*))) OR abstract:(((“vascular endothelial growth factor” OR vasculotropin OR VEGF* OR “vascular permeability factor” OR VPF) AND (trap* OR inhibit* OR antagonist*)))) OR (title:((angiogen* AND (antagonist* OR inhibit*))) OR abstract:((angiogen* AND (antagonist* OR inhibit*)))) OR (title:(((antiangiogen* OR “anti angiogen” OR anti-angiogen* OR angiostatic) AND (agent* OR drug* OR effect*))) OR abstract:(((antiangiogen* OR “anti angiogen” OR anti-angiogen* OR angiostatic) AND (agent* OR drug* OR effect*)))) OR (title:((Aflibercept* OR Eylea OR Zaltrap OR Ziv-Aflibercept OR “AVE 0005” OR AVE0005 OR “AVE 005” OR AVE005 OR Bevacizumab* OR Avastin OR Mvasi OR Alymsys OR Aybintio OR Equidacent OR Onbevzi OR Oyavas OR Zirabev OR rhuMAbVEGF OR rhuMAb-VEGF OR “rhuMAb VEGF” OR “NSC 704865” OR NSC704865 OR Ranibizumab* OR Lucentis OR “rhuFab V2” OR Pegaptanib* OR “EYE 001” OR EYE001 OR Macugen OR “NX 1838” OR NX1838)) OR abstract:((Aflibercept* OR Eylea OR Zaltrap OR Ziv-Aflibercept OR “AVE 0005” OR AVE0005 OR “AVE 005” OR AVE005 OR Bevacizumab* OR Avastin OR Mvasi OR Alymsys OR Aybintio OR Equidacent OR Onbevzi OR Oyavas OR Zirabev OR rhuMAbVEGF OR rhuMAb-VEGF OR “rhuMAb VEGF” OR “NSC 704865” OR NSC704865 OR Ranibizumab* OR Lucentis OR “rhuFab V2” OR Pegaptanib* OR “EYE 001” OR EYE001 OR Macugen OR “NX 1838” OR NX1838))) OR (title:(((IVB OR IVR) AND inject*)) OR abstract:(((IVB OR IVR) AND inject*))))

Filter: Publication year 2000–2023

Publication type: Systematic Reviews

= 1026

Key:

* = truncation

title: = searches in title field

abstract: = searches in abstract field

PROSPERO

via https://www.crd.york.ac.uk/prospero/

Date range: Inception – 26 May 2023

Date searched: 26 May 2023

Records retrieved: 159

1. MeSH DESCRIPTOR Diabetic Retinopathy 107

2. ((diabet* or DM) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath* or maculopath*)) 609

3. (((proliferat* or PDR or pre-proliferat* or preproliferat* or non-proliferat* or nonproliferat* or NPDR or background) adj3 (retinopath* or vitreoretinopath* or vitreo-retinopath* or chorioretinopath* or chorio-retinopath*)) and (diabet* or DM)) 110

4. (new blood vessel* and diabet*) 9

5. (((retin* or subretina* or sub-retina* or interretina* or inter-retina* or vitreoretin* or vitreo-retin* or chorioretin* or chorio-retin* or choroid* or macula* or intraocular or intra-ocular or intravitreal or intra-vitreal) adj4 (damage* or deteriorat* or degnerat* or disease* or edema or oedema or neovascularisation* or neovascularization*)) AND diabet*) 373

6. ((retinal vein* adj3 (occlu* or obstruct* or clos* or stricture* or steno* or block* or emboli*)) and diabet*) 64

7. #1 OR #2 OR #3 OR #4 OR #5 OR #6 740

8. MeSH DESCRIPTOR Vascular Endothelial Growth Factors EXPLODE ALL TREES WITH QUALIFIER AI 0

9. MeSH DESCRIPTOR Receptors, Vascular Endothelial Growth Factor EXPLODE ALL TREES WITH QUALIFIER AI 0

10. (anti adj2 VEGF*) 327

11. (anti-VEGF* or antiVEGF*) 327

12. ((anti vascular or anti-vascular or antivascular) adj2 endothelial growth factor*) 153

13. (((vascular endothelial adj2 growth factor*) or vasculotropin or VEGF* or vascular permeability factor* or VPF) adj2 (trap* or inhibit* or antagonist*)) 96

14. (vascular proliferation adj4 inhibit*) 0

15. #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 412

16. MeSH DESCRIPTOR Angiogenesis Inhibitors 40

17. MeSH DESCRIPTOR Angiogenesis Inducing Agents EXPLODE ALL TREES WITH QUALIFIER A I0

18. (angiogen* adj2 (antagonist* or inhibit*)) 74

19. ((antiangiogen* or anti angiogen* or anti-angiogen*) adj2 (agent* or drug* or effect*)) 145

20. (angiostatic adj2 (agent* or drug*)) 0

21. ((neovascularisation* or neovascularization* or vascularisation* or vascularization*) adj2 inhibit*) 0

22. #16 OR #17 OR #18 OR #19 OR #20 OR #21 224

23. (Aflibercept*) 141

24. (Eylea or Zaltrap or Ziv-Aflibercept or AVE 0005 or AVE0005 or AVE 005 or AVE005) 22

25. MeSH DESCRIPTOR Bevacizumab 46

26. (Bevacizumab*) 445

27. (Avastin or Mvasi or Alymsys or Aybintio or Equidacent or Onbevzi or Oyavas or Zirabev or rhuMAbVEGF or rhuMAb-VEGF or rhuMAb VEGF or NSC 704865 or NSC704865) 59

28. (IVB adj2 inject*) 0

29. MeSH DESCRIPTOR Ranibizumab 7

30. (Ranibizumab*) 142

31. (Lucentis or rhuFab V2) 23

32. (IVR adj2 inject*) 0

33. (Pegaptanib*) 30

34. (EYE 001 or EYE001 or Macugen or NX 1838 or NX1838) 5

35. #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 500

36. #15 OR #22 OR #35 839

37. #7 AND #36 159

Key:

MeSH DESCRIPTOR = indexing term: Medical Subject Heading (MeSH)

QUALIFIER AI = indexing term subheading for antagonists & inhibitors

EXPLODE ALL TREES = exploded indexing term (MeSH)

* = truncation

adj3 = terms within three words of each other (order specified).

:TI,KW = terms in either title or keyword fields

ClinicalTrials.gov

via https://clinicaltrials.gov/

Date searched: 26 May 2023

Records retrieved: 286

Two separate searches were used in Advanced Search, retrieving 286 records in total, which were imported into EndNote 20 and deduplicated.

1. Condition or Disease: (diabetic retinopathy)

Other Terms: (Aflibercept OR Eylea OR Zaltrap OR Bevacizumab OR Avastin OR Mvasi OR Alymsys OR Aybintio OR Equidacent OR Onbevzi OR Oyavas OR Zirabev OR rhuMAb VEGF OR Ranibizumab OR Lucentis OR rhuFab OR Pegaptanib OR Macugen) = 190 hits

2. Condition or Disease: (diabetic retinopathy)

Other Terms: ((VEGF OR vascular endothelial growth factor OR vasculotropin OR vascular permeability factor or VPF) AND (anti OR trap or inhibitor or antagonist)) = 96 hits

European Union Clinical Trials Register

via www.clinicaltrialsregister.eu/ctr-search/search

Date searched: 26 May 2023

Records retrieved: 163

Two separate searches were used, retrieving 163 records in total, which were imported into EndNote 20 and deduplicated.

1. ((“diabetic retinopathy”) AND (Aflibercept OR Eylea OR Zaltrap OR Bevacizumab OR Avastin OR Mvasi OR Alymsys OR Aybintio OR Equidacent OR Onbevzi OR Oyavas OR Zirabev OR “rhuMAb VEGF” OR Ranibizumab OR Lucentis OR rhuFab OR Pegaptanib OR Macugen)) = 113 hits

2. ((“diabetic retinopathy”) AND ((anti OR trap or inhibitor OR antagonist) AND (VEGF OR “vascular endothelial growth factor” OR vasculotropin OR “vascular permeability factor” OR VPF))) = 50 hits

WHO International Clinical Trials Registry Platform

via https://trialsearch.who.int/

Date searched: 26 May 2023

Records retrieved: 198

Two separate searches were used in Advanced Search, retrieving 198 records in total, which were imported into EndNote 20 and deduplicated.

1. Advanced Search

Condition: (diabetic retinopathy)

Intervention: (Aflibercept OR Eylea OR Zaltrap OR Bevacizumab OR Avastin OR Mvasi OR Alymsys OR Aybintio OR Equidacent OR Onbevzi OR Oyavas OR Zirabev OR rhuMAb VEGF OR Ranibizumab OR Lucentis OR rhuFab OR Pegaptanib OR Macugen)

Recruitment Status: ALL = 194 records for 180 trials

2. Advanced Search

Condition: (diabetic retinopathy)

Intervention: ((VEGF OR vascular endothelial growth factor OR vasculotropin OR vascular permeability factor or VPF) AND (anti OR trap or inhibitor or antagonist))

Recruitment Status: ALL = 23 records for 18 trials

List of excluded studies

Trials not included in meta-analyses

Risk-of-bias assessment

Figures and forest plots summarising best corrected visual acuity data

Note from these figures that there appears to be a possible decline in benefit to vison over time, and that the benefit of ant-VEGF may be greater in people with poorer initial vision, but these differences may be confounded by differences between types of anti-VEGF.

FIGURE 13.

Forest plot of all mean differences in ETDRS between anti-VEGF and control (right side favours anti-VEGF).

FIGURE 14.

Forest plot of all mean differences in log-MAR between anti-VEGF and control (left side favours anti-VEGF).

Forest plots of meta-analyses of best corrected visual acuity

Up to 1 year

FIGURE 15.

Meta-analysis of mean differences in ETDRS between anti-VEGF and control up to 1 year of follow-up (right side favours anti-VEGF).

FIGURE 16.

Meta-analysis of mean differences in log-MAR between anti-VEGF and control up to 1 year of follow-up (left side favours anti-VEGF).

One to 2 years’ follow-up

FIGURE 17.

Meta-analysis of mean differences in ETDRS between anti-VEGF and control with 1–2 years’ of follow-up (right side favours anti-VEGF).

FIGURE 18.

Meta-analysis of mean differences in log-MAR between anti-VEGF and control with 1–2 years’ of follow-up (left side favours anti-VEGF).

Maximum follow-up in trial (up to 2 years)

FIGURE 19.

Meta-analysis of mean differences in ETDRS between anti-VEGF and control at end of trial (right side favours anti-VEGF).

FIGURE 20.

Meta-analysis of mean differences in log-MAR between anti-VEGF and control at end of trial (left side favours anti-VEGF).

Network meta-analyses of best corrected visual acuity (using logarithm of the minimum angle of resolution)

Note: From this point forward on, meta-analyses of BCVA measured using log-MAR are presented. Some analyses using ETDRS were performed but are not included here. Similarly, only random-effects analyses are presented for simplicity, as differences between random- and fixed-effect analyses were minimal.

Analyses at up to 1 year of follow-up

FIGURE 21.

Network diagram of BCVA at up to 1 year of follow-up.

FIGURE 22.

All treatment comparisons for 1-year random-effects NMA of log-MAR.

FIGURE 23.

Probability of treatments for 1-year random-effects NMA of log-MAR.

TABLE 6 - Results of NMA of log-MAR up to 1 year – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.084	−0.222	0.056
Bevacizumab	PRP	−0.198	−1.213	0.785
Bevacizumab + PRP	PRP	−0.172	−0.279	−0.069
Ranibizumab	PRP	−0.121	−0.233	−0.006
Ranibizumab + PRP	PRP	−0.078	−0.165	0.013
Bevacizumab	Aflibercept	−0.115	−1.142	0.853
Bevacizumab + PRP	Aflibercept	−0.088	−0.273	0.082
Ranibizumab	Aflibercept	−0.037	−0.213	0.130
Ranibizumab + PRP	Aflibercept	0.006	−0.151	0.173
Bevacizumab + PRP	Bevacizumab	0.026	−0.947	1.027
Ranibizumab	Bevacizumab	0.077	-0.913	1.098
Ranibizumab + PRP	Bevacizumab	0.121	−0.867	1.151
Ranibizumab	Bevacizumab + PRP	0.051	−0.095	0.208
Ranibizumab + PRP	Bevacizumab + PRP	0.094	−0.040	0.236
Ranibizumab + PRP	Ranibizumab	0.043	−0.067	0.160

TABLE 7 - Results of NMA of log-MAR up to 1 year – ranking probabilities

Treatment arm	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)	5th (%)	6th (%)
PRP	0.00	0.13	0.65	4.93	37.65	56.65
Aflibercept	4.33	12.65	24.65	30.48	23.10	4.80
Bevacizumab	50.73	5.05	2.38	2.63	3.83	35.40
Bevacizumab + PRP	33.60	44.23	14.73	5.68	1.73	0.05
Ranibizumab	10.25	30.30	33.60	17.40	7.25	1.20
Ranibizumab + PRP	1.10	7.65	24.00	38.90	26.45	1.90

Analyses at 1–2 years’ follow-up

FIGURE 24.

Network diagram of BCVA at up to 1–2 years of follow-up.

FIGURE 25.

All treatment comparisons for 1–2 year random-effects NMA of log-MAR.

FIGURE 26.

Probability of treatments for 1–2-year random-effects NMA of log-MAR.

TABLE 8 - Results of NMA of log-MAR 1–2 years – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.080	−0.225	0.100
Bevacizumab	PRP	−0.182	−1.181	0.816
Ranibizumab	PRP	−0.072	−0.171	0.017
Ranibizumab + PRP	PRP	−0.068	−0.152	0.020
Bevacizumab	Aflibercept	−0.102	−1.095	0.899
Ranibizumab	Aflibercept	0.008	−0.200	0.187
Ranibizumab + PRP	Aflibercept	0.012	−0.174	0.189
Ranibizumab	Bevacizumab	0.110	−0.887	1.104
Ranibizumab + PRP	Bevacizumab	0.114	−0.885	1.112
Ranibizumab + PRP	Ranibizumab	0.004	−0.100	0.114

TABLE 9 - Results of NMA of log-MAR 1–2 years – ranking probabilities

Treatment arm	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)	5th (%)
PRP	0.05	1.20	6.25	37.93	54.58
Aflibercept	21.38	34.58	20.98	17.00	6.08
Bevacizumab	56.20	3.15	2.88	3.35	34.43
Ranibizumab	13.13	30.28	34.13	20.15	2.33
Ranibizumab + PRP	9.25	30.80	35.78	21.58	2.60

Analysis at maximum follow-up time (up to 2 years)

FIGURE 27.

All treatment comparisons for end-of-trial random-effects NMA of log-MAR.

FIGURE 28.

Probability of treatments for end-of-trial random-effects NMA of log-MAR.

TABLE 10 - Results of NMA of log-MAR at end of trial – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.087	−0.228	0.049
Bevacizumab	PRP	−0.209	−1.176	0.782
Bevacizumab + PRP	PRP	−0.171	−0.284	−0.064
Ranibizumab	PRP	−0.085	−0.177	−0.004
Ranibizumab + PRP	PRP	−0.069	−0.151	0.016
Bevacizumab	Aflibercept	−0.122	−1.117	0.881
Bevacizumab + PRP	Aflibercept	−0.085	−0.265	0.093
Ranibizumab	Aflibercept	0.002	−0.167	0.151
Ranibizumab + PRP	Aflibercept	0.017	−0.139	0.180
Bevacizumab + PRP	Bevacizumab	0.038	−0.956	1.030
Ranibizumab	Bevacizumab	0.124	−0.868	1.100
Ranibizumab + PRP	Bevacizumab	0.140	−0.830	1.116
Ranibizumab	Bevacizumab + PRP	0.087	−0.054	0.225
Ranibizumab + PRP	Bevacizumab + PRP	0.102	−0.032	0.243
Ranibizumab + PRP	Ranibizumab	0.015	−0.086	0.128

TABLE 11 - Results of NMA of log-MAR at end of trial – ranking probabilities

Treatment arm	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)	5th (%)	6th (%)
PRP	0.00	0.63	4.80	35.88	58.70	0.00
Aflibercept	20.58	29.20	23.05	17.80	4.35	20.58
Bevacizumab	5.90	2.70	1.85	3.65	33.43	5.90
Bevacizumab + PRP	46.80	9.78	3.88	1.23	0.10	46.80
Ranibizumab	17.03	31.83	31.10	16.20	0.93	17.03
Ranibizumab + PRP	9.70	25.88	35.33	25.25	2.50	9.70

Network meta-analyses allowing for interaction with follow-up time and best corrected visual acuity at randomisation

Allowing for variation with follow-up time

Network meta-analyses incorporating all follow-up times (longest in each trial), allowing varying effect of anti-VEGF with follow-up time. Time variation is assumed to be the same for all types of anti-VEGF. A selection of output plots is presented. Results are presented for the predicted effects after 1 year of follow-up.

FIGURE 29.

All treatment comparisons for time-adjusted random-effects NMA of log-MAR.

TABLE 12 - Results of NMA of log-MAR adjusting for time – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.086	−0.221	0.045
Bevacizumab	PRP	−0.199	−1.218	0.858
Bevacizumab + PRP	PRP	−0.112	−0.245	0.027
Ranibizumab	PRP	−0.119	−0.214	−0.023
Ranibizumab + PRP	PRP	−0.075	−0.153	0.001
Bevacizumab	Aflibercept	−0.112	−1.135	0.945
Bevacizumab + PRP	Aflibercept	−0.026	−0.216	0.163
Ranibizumab	Aflibercept	−0.033	−0.200	0.130
Ranibizumab + PRP	Aflibercept	0.011	−0.138	0.163
Bevacizumab + PRP	Bevacizumab	0.086	−0.957	1.127
Ranibizumab	Bevacizumab	0.080	−0.973	1.105
Ranibizumab + PRP	Bevacizumab	0.123	−0.919	1.138
Ranibizumab	Bevacizumab + PRP	−0.007	−0.197	0.184
Ranibizumab + PRP	Bevacizumab + PRP	0.037	−0.123	0.195
Ranibizumab + PRP	Ranibizumab	0.044	−0.064	0.155

TABLE 13 - Results of NMA of log-MAR adjusting for time – ranking probabilities

Treatment arm	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)	5th (%)	6th (%)
PRP	0.00	0.05	0.65	5.13	38.10	56.08
Aflibercept	7.60	19.98	25.30	25.10	17.48	4.55
Bevacizumab	51.98	3.88	2.58	2.78	4.35	34.45
Bevacizumab + PRP	18.73	30.70	17.90	17.35	12.35	2.98
Ranibizumab	19.40	34.13	25.88	14.88	5.25	0.48
Ranibizumab + PRP	2.30	11.28	27.70	34.78	22.48	1.48

Allowing for variation over time and by logarithm of the minimum angle of resolution at randomisation

Network meta-analyses incorporating all follow-up times (longest in each trial), allowing for varying effect of anti-VEGF by follow-up duration and varying effect by trial mean log-MAR at randomisation. Time and log-MAR variation are assumed to be the same for all types of anti-VEGF. A selection of output plots is presented. Results are presented for the predicted effects after 1 year of follow-up and at mean baseline BCVA across trials.

FIGURE 30.

All treatment comparisons for time-adjusted and baseline BCVA adjusted random-effects NMA of log-MAR.

TABLE 14 - Results of NMA of log-MAR adjusting for time and baseline BCVA – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.085	−0.310	0.119
Bevacizumab + PRP	PRP	−0.116	−0.281	0.050
Ranibizumab	PRP	−0.117	−0.247	0.011
Ranibizumab + PRP	PRP	−0.073	−0.187	0.041
Bevacizumab + PRP	Aflibercept	−0.031	−0.287	0.228
Ranibizumab	Aflibercept	−0.032	−0.288	0.233
Ranibizumab + PRP	Aflibercept	0.012	−0.218	0.251
Ranibizumab	Bevacizumab + PRP	−0.001	−0.239	0.225
Ranibizumab + PRP	Bevacizumab + PRP	0.043	−0.151	0.234
Ranibizumab + PRP	Ranibizumab	0.044	−0.112	0.203

TABLE 15 - Results of NMA of log-MAR adjusting for time and baseline BCVA – ranking probabilities

Treatment arm	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)	5th (%)
PRP	0.08	0.70	4.23	19.50	75.50
Aflibercept	19.25	23.45	21.43	22.70	13.18
Bevacizumab + PRP	38.40	23.30	18.10	15.68	4.53
Ranibizumab	36.93	30.48	19.63	10.75	2.23
Ranibizumab + PRP	5.35	22.08	36.63	31.38	4.58

Network meta-analyses of reduced networks

Assuming anti-vascular endothelial growth factor and anti-vascular endothelial growth factor + panretinal photocoagulation are equivalent

This analysis assumes that anti-VEGF only arms and anti-VEGF + PRP arms have equal effect. To be used to assess differences between anti-VEGF types. A model allowing effect to vary with time and baseline log-MAR was used. Results are presented for the predicted effects after 1 year of follow-up and at mean baseline BCVA across trials.

FIGURE 31.

Results from a reduced network to compare anti-VEGFs.

TABLE 16 - Results of reduced network to compare anti-VEGFs – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Aflibercept	PRP	−0.091	−0.245	0.063
Bevacizumab	PRP	−0.126	−0.261	0.007
Ranibizumab	PRP	−0.094	−0.173	−0.023
Bevacizumab	Aflibercept	−0.035	−0.238	0.174
Ranibizumab	Aflibercept	−0.003	−0.166	0.163
Ranibizumab	Bevacizumab	0.032	−0.142	0.200

TABLE 17 - Results of reduced network to compare anti-VEGFs – ranking probabilities

Treatment	Probability of ranking
	1st (%)	2nd (%)	3rd (%)	4th (%)
PRP	0.00	0.80	11.48	87.73
Aflibercept	25.08	32.93	33.35	8.65
Bevacizumab	53.60	23.33	20.18	2.90
Ranibizumab	21.33	42.95	35.00	0.73

Assuming all types of anti-vascular endothelial growth factor are equivalent

This analysis assumes that all three anti-VEGF drugs have equal effect. To be used to assess the overall effect of anti-VEGF. A model allowing effect to vary with time and baseline log-MAR was used.

FIGURE 32.

Results from a reduced network to compare treatment classes.

TABLE 18 - Results of reduced network to compare treatment classes – comparisons between treatments

Intervention	Comparator	Mean difference	95% CI
Anti-VEGF	PRP	−0.089	−0.180	−0.019
Anti-VEGF + PRP	PRP	−0.108	−0.192	−0.039
Anti-VEGF + PRP	Anti-VEGF	−0.019	−0.126	0.083

TABLE 19 - Results of reduced network to compare treatment classes – ranking probabilities

Treatment	Probability of ranking
	1st (%)	2nd (%)	3rd (%)
PRP	0.03	1.20	98.78
Anti-VEGF	33.05	65.88	1.08
Anti-VEGF + PRP	66.93	32.93	0.15

Threshold analyses

Up to 1 year

FIGURE 33.

Threshold analyses of data up to 1 year of follow-up.

One to 2 years

FIGURE 34.

Threshold analyses of data with 1–2 years of follow-up.

Maximum follow-up (up to 2 years)

FIGURE 35.

Threshold analyses of data at end of trial (up to 2 years).

Allowing for effect variation with time and baseline logarithm of the minimum angle of resolution

FIGURE 36.

Threshold analyses of model adjusting for effect of time and baseline log-MAR.

Reduced network (for comparing anti-vascular endothelial growth factors)

Adjusted for follow-up time and BCVA at baseline.

FIGURE 37.

Threshold analysis of simplified network to compare anti-VEGF types, with time and baseline BCVA adjustment.

Reduced network (comparing anti-vascular endothelial growth factor to panretinal photocoagulation)

Adjusted for follow-up time and BCVA at baseline.

FIGURE 38.

Threshold analyses of simplified network to compare anti-VEGF to PRP, adjusted for follow-up time and baseline BCVA.

Forest plots of outcomes without meta-analysis

These forest plots show results for all anti-VEGF types, and at all follow-up times. Note that this means some trials appear more than once in a forest plot.

Neovascularisation of the disc

FIGURE 39.

Forest plot of all NVD data (left side favours anti-VEGF).

Neovascularisation elsewhere

FIGURE 40.

Forest plot of all NVE data (left side favours anti-VEGF).

Diabetic macular oedema

FIGURE 41.

Forest plot of DME incidence (left side favours anti-VEGF).

Improvement in diabetic retinopathy severity score (Diabetic Retinopathy Severity Scale)

FIGURE 42.

Forest plot of improvement in DRSS severity (right side favours anti-VEGF).

Proliferative retinopathy incidence

FIGURE 43.

Forest plot of proliferative DR (left side favours anti-VEGF).

Regression of neovascularisation

FIGURE 44.

Forest plot of regressive neovascularisation (left side favours anti-VEGF).

Use of other treatments

FIGURE 45.

Forest plot of use of other treatments (left side favours anti-VEGF).

Vitrectomy

FIGURE 46.

Forest plot of vitrectomy incidence (left side favours anti-VEGF).

Vitreous haemorrhage

FIGURE 47.

Forest plot of vitreous haemorrhage incidence (left side favours anti-VEGF).

Adverse event outcomes

These forest plots show results for all anti-VEGF types, and at all follow-up times. Note that this means some trials appear more than once in a forest plot. For simplicity, only adverse event outcomes reported in two or more studies are presented.

Cataracts

FIGURE 48.

Forest plot of cataracts data (left side favours anti-VEGF).

Conjunctival haemorrhage

FIGURE 49.

Forest plot of conjunctival haemorrhage data (left side favours anti-VEGF).

Cardiovascular mortality

FIGURE 50.

Forest plot of cardiovascular mortality data (left side favours anti-VEGF).

Death (all-cause mortality)

FIGURE 51.

Forest plot of death data (left side favours anti-VEGF).

Myocardial infarction

FIGURE 52.

Forest plot of myocardial infarction data (left side favours anti-VEGF).

Ocular pain

FIGURE 53.

Forest plot of ocular pain data (left side favours anti-VEGF).

Raised intraocular pressure

FIGURE 54.

Forest plot of raised intraocular pressure data (left side favours anti-VEGF).

Retinal detachment

FIGURE 55.

Forest plot of retinal detachment data (left side favours anti-VEGF).

Retinal tear

FIGURE 56.

Forest plot of retinal data (left side favours anti-VEGF).

Serious adverse event (however defined)

FIGURE 57.

Forest plot of serious adverse event data (left side favours anti-VEGF).

Stroke

FIGURE 58.

Forest plot of stroke data (left side favours anti-VEGF).

Meta-analyses of other outcomes and adverse events

All meta-analyses presented assumed that the impact of anti-VEGF on outcome (or adverse event) is the same for all types of anti-VEGF (in isolation or combined with PRP), and at all follow-up times. For trials with multiple time points, the longest follow-up was used. For trial with multiple arms, only one anti-VEGF arm was used; arms using anti-VEGF alone were preferred.

Neovascularisation elsewhere

FIGURE 59.

Meta-analysis of NVE (left side favours anti-VEGF).

Neovascularisation of the disc

FIGURE 60.

Meta-analysis of NVD (left side favours anti-VEGF).

Other non-vison outcomes

This forest plot shows the summary results of each meta-analysis (each bar is a meta-analysis result). Meta-analyses are restricted to trials of proliferative retinopathy. Full forest plots for each outcome are not presented.

FIGURE 61.

Meta-analysis summary for non-vision outcomes in PDR trials (left side favours anti-VEGF).

Adverse events

This forest plot shows the summary results of each meta-analysis (each bar is a meta-analysis result). Meta-analyses are restricted to trials of proliferative retinopathy. Full forest plots for each outcome are not presented.

FIGURE 62.

Meta-analysis summary for adverse events (left side favours anti-VEGF).

Best corrected visual acuity

FIGURE 63.

Mean difference in ETDRS after 2 years in NPDR trials.

FIGURE 64.

Mean difference in log-MAR after 2 years in NPDR trials.

Diabetic macular oedema in non-proliferative retinopathy

Diabetic macular oedema was the only outcome other than BCVA reported in both trials of NPDR.

FIGURE 65.

Diabetic macular oedema incidence in NPDR trials.

Other outcomes

FIGURE 66.

Non-BCVA outcomes in NPDR trials.

Adverse events

FIGURE 67.

Adverse event outcomes in NPDR trials.