eAppendix. Diabetic Retinopathy Clinical Research Network
Customize your JAMA Network experience by selecting one or more topics from the list below.
Jampol LM, Glassman AR, Bressler NM, Wells JA, Ayala AR, for the Diabetic Retinopathy Clinical Research Network. Anti–Vascular Endothelial Growth Factor Comparative Effectiveness Trial for Diabetic Macular Edema: Additional Efficacy Post Hoc Analyses of a Randomized Clinical Trial. JAMA Ophthalmol. 2016;134(12):1429–1434. doi:10.1001/jamaophthalmol.2016.3698
Are there additional analyses from the randomized clinical trial comparing aflibercept, bevacizumab, and ranibizumab for diabetic macular edema that might aid in the interpretation of the results?
Mean letter change in visual acuity over 2 years among eyes with initial visual acuity 20/50 or worse was greater in the aflibercept group than in the bevacizumab or ranibizumab groups.
Although post hoc analyses should be viewed with caution, in eyes with visual acuity of 20/50 or worse, aflibercept produces the greatest mean improvement over 2 years.
Post hoc analyses from the Diabetic Retinopathy Clinical Research Network randomized clinical trial comparing aflibercept, bevacizumab, and ranibizumab for diabetic macular edema (DME) might influence interpretation of study results.
To provide additional outcomes comparing 3 anti–vascular endothelial growth factor (VEGF) agents for DME.
Design, Setting, and Participants
Post hoc analyses performed from May 3, 2016, to June 21, 2016, of a randomized clinical trial performed from August 22, 2012, to September 23, 2015, of 660 participants comparing 3 anti-VEGF treatments in eyes with center-involved DME causing vision impairment.
Randomization to intravitreous aflibercept (2.0 mg), bevacizumab (1.25 mg), or ranibizumab (0.3 mg) administered up to monthly based on a structured retreatment regimen. Focal/grid laser treatment was added after 6 months for the treatment of persistent DME.
Main Outcomes and Measures
Change in visual acuity (VA) area under the curve and change in central subfield thickness (CST) within subgroups based on whether an eye received laser treatment for DME during the study.
Post hoc analyses were performed for 660 participants (mean [SD] age, 61  years; 47% female, 65% white, 16% black or African American, 16% Hispanic, and 3% other). For eyes with an initial VA of 20/50 or worse, VA improvement was greater with aflibercept than the other agents at 1 year but superior only to bevacizumab at 2 years. Mean (SD) letter change in VA over 2 years (area under curve) was greater with aflibercept (+17.1 [9.7]) than with bevacizumab (+12.1 [9.4]; 95% CI, +1.6 to +7.3; P < .001) or ranibizumab (+13.6 [8.5]; 95% CI, +0.7 to +6.0; P = .009). When VA was 20/50 or worse at baseline, bevacizumab reduced CST less than the other agents at 1 year, but at 2 years the differences had diminished. In subgroups stratified by baseline VA, anti-VEGF agent, and whether focal/grid laser treatment was performed for DME, the only participants to have a substantial reduction in mean CST between 1 and 2 years were those with a baseline VA of 20/50 or worse receiving bevacizumab and laser treatment (mean [SD], −55  µm; 95% CI, −82 to −28 µm; P < .001).
Conclusions and Relevance
Although post hoc analyses should be viewed with caution given the potential for bias, in eyes with a VA of 20/50 or worse, aflibercept has the greatest improvement in VA over 2 years. Focal/grid laser treatment, ceiling and floor effects, or both may account for mean thickness reductions noted only in bevacizumab-treated eyes between 1 and 2 years.
clinicaltrials.gov Identifier NCT01627249
The Diabetic Retinopathy Clinical Research Network (DRCR.net) Protocol T, which is a comparison of the anti–vascular endothelial growth factor (VEGF) agents aflibercept, bevacizumab, and ranibizumab for treatment of center-involving diabetic macular edema (DME) with vision loss,1-4 has prompted discussions at conferences and requests by peer-reviewed and non–peer-reviewed journals for post hoc analyses. At 1 and 2 years,1,4 the mean visual acuity (VA) improvement with all 3 agents was similar for eyes with VAs of 20/32 to 20/40 at randomization. For a baseline VA of 20/50 or worse, aflibercept produced a greater improvement in mean VA than the other agents at 1 year. Improvement in VA at the 2-year time point with aflibercept remained statistically superior to bevacizumab but not ranibizumab.4 Mean letter changes in VA from 1 to 2 years in each group were small (aflibercept: letter score change, −1.2; 95% CI, −3.2 to 0.8; P = .24; bevacizumab: letter score change, +0.3; 95% CI, −2.3 to 2.8; P = .85; and ranibizumab: letter score change, +1.1; 95% CI, −1.4 to 3.6; P = .40; P values testing whether within-group change differed from 0 [2-tailed t test]). Despite these small differences, some might interpret the 2-year results as revealing convergence in the subgroup with worse baseline VA. This article examines VA outcomes over time, explores the potential effect of focal/grid laser treatment on retinal thickness, and hypothesizes potential reasons for the decreasing number of injections over 2 years.
Post hoc analyses were performed from May 3, 2016, to June 21, 2016, of a randomized clinical trial performed from August 22, 2012, to September 23, 2015, of participants comparing 3 anti-VEGF treatments in eyes with center-involved DME causing vision impairment. Participants were randomized to intravitreous aflibercept (2.0 mg), bevacizumab (1.25 mg), or ranibizumab (0.3 mg) administered up to monthly based on a structured retreatment regimen. Focal/grid laser was added after 6 months for persistent DME. Outcomes measured were change in VA area under the curve (AUC) and change in central subfield thickness (CST) within subgroups based on whether an eye received laser treatment for DME during the study. The study adhered to the tenets of the Declaration of Helsinki and was approved by local institutional review boards or a central institutional review board if the site did not have a local board. Study participants provided written informed consent, and data were deidentified. Data analysis was performed with SAS statistical software, version 9.4 (SAS Institute Inc). P < .05 was considered statistically significant. P values for treatment group differences were adjusted for multiple comparisons.
Post hoc analyses were performed for 660 participants (mean [SD] age, 61  years; 47% female, 65% white, 16% black or African American, 16% Hispanic, and 3% other). For eyes with a VA of 20/50 or worse, mean (SD) changes in the VA letter score from baseline averaged over each of the follow-up visits through 2 years (AUC) were as follows: aflibercept, +17.1 (9.7) letters; bevacizumab, +12.1 (9.4) letters; and ranibizumab, +13.6 (8.5) letters. The letter difference was +4.5 for aflibercept vs bevacizumab (95% CI, +1.6 to +7.3; P < .001), +3.4 (95% CI, +0.7 to +6.0; P = .009) for aflibercept vs ranibizumab, and +1.1 (95% CI, −1.2 to +3.5; P = .35) for ranibizumab vs bevacizumab. The mean percentage of time per eye during the 2 years that the eye was improved by 15 letters or more from baseline (Table 1) was greater with aflibercept (58%) than with bevacizumab (43%; 95% CI for aflibercept vs bevacizumab, 1%-25%; P = .02) or ranibizumab (46%; 95% CI for aflibercept vs ranibizumab, 0%-22%; P = .05).
Compared with aflibercept and ranibizumab, bevacizumab reduced retinal thickness less in both baseline VA subgroups at 6 months, the earliest time when focal/grid laser could be administered, and at 1 year.1 However, differences in CST on optical coherence tomography between bevacizumab and the other 2 agents, observed at 1 year, diminished at 2 years in the subgroup with worse baseline VA.1 Among eyes with a baseline VA of 20/50 or worse, changes in CST between 1 and 2 years were small, with a mean (SD) of −2 (73) µm (P = .82) for aflibercept and +4 (81) µm (P = .66) for ranibizumab. However, additional reduction occurred with bevacizumab between 1 and 2 years (mean [SD] change, −42  µm; P < .001, 2-tailed t test). Because treatment with laser was based on CST, it is not possible to separate the effect of laser on thinning from the course of DME over time independent of laser or the effect of subsequent anti-VEGF injections, which were given more frequently in the bevacizumab group in year 2 (Table 2).
Because laser treatment for DME was more frequently used with bevacizumab than with aflibercept or ranibizumab (Table 2), a post hoc analysis was performed stratified by whether an eye did or did not receive focal/grid laser treatment during the study (Figure). Among eyes with baseline VAs of 20/50 or worse, the only decrease in CST from 1 to 2 years occurred in eyes in the bevacizumab group that received focal/grid laser treatment during the study (mean [SD], −55  µm; P < .001, 2-tailed t test). In this subgroup with worse baseline VA receiving focal/grid laser treatment, at 2 years there were no treatment group differences in the mean (SD) change of CST from baseline (aflibercept, −213  µm; bevacizumab, −204  µm; and ranibizumab, −191  µm). However, in eyes with worse baseline VA never receiving focal/grid laser treatment during the study, the mean reduction from baseline to 2 years was greater with aflibercept (−209 µm) than with bevacizumab (−144 µm), a difference of −46 µm (95% CI, −99 to +8) (adjusted for baseline VA and CST, Hochberg-adjusted P = .14). The difference between ranibizumab (−153 µm) and bevacizumab (−144 µm) was −8 µm (95% CI, −57 to +41) (adjusted for baseline VA and CST, Hochberg-adjusted P = .75). Within the bevacizumab eyes with VAs of 20/50 or worse at baseline, the VA change between 1 and 2 years was similar for eyes receiving focal/grid laser treatment (+0.03 letter score) and eyes not receiving focal/grid laser treatment (+0.75 letter score). Differences in thickness suggest a possible role of the laser treatment reducing the CST of eyes in the bevacizumab group in year 2. Similar findings were not noted among eyes with baseline VAs of 20/32 to 20/40 (Figure). Postrandomization subgroup analyses have the potential for selection biases, and laser treatment, continued anti-VEGF therapy, and natural history all remain potential explanations for thinning of the central subfield in the bevacizumab group. It is also possible that the mean CST with aflibercept and ranibizumab had reached its maximal amount of resolution by 1 year and additional substantial improvement was therefore not possible because of floor and ceiling effects.
The DRCR.net treatment algorithm involves deferring injections after the initial 6 months of treatment if VA and CST are stable for 2 consecutive visits (ie, CST changing <10% or <5 letters of VA). Stability of optical coherence tomography CST and VA after 6 months, not necessarily resolution of DME or 20/20 VA, is required to defer injections. Implementation of this strategy was associated with a decreased injection frequency between years 1 and 2 (Table 2). Further decreases were noted over 5 years in a prior DRCR.net protocol that evaluated ranibizumab for DME (Protocol I).6 The substantial decrease in the number of injections from years 3 to 5 may be partially attributable to DME stabilization over time or to change in retinopathy influenced by treatment decreasing VEGF production over time. This hypothesis is supported by other studies7-9 that have found less likelihood of worsening and greater likelihood of improving diabetic retinopathy severity levels with anti-VEGF treatments compared with no anti-VEGF treatment for DME.
Some interpret small decreases in mean VA with aflibercept and small increases with bevacizumab or ranibizumab from 1 year to 2 years as a convergence of drug effects on VA. Additional changes beyond 2 years cannot be determined. Convergence of effects over time, if real, may represent a benefit of each anti-VEGF drug’s effect on retinopathy severity as reported in prior studies1,6,8 and possibly associated with a decreased VEGF drive. This mechanism might allow agents eventually to have similar efficacy over time. Another explanation for the possible decreasing VA differences by treatment group might be more injections in the bevacizumab and ranibizumab groups compared with aflibercept. Yet another explanation may be decreases in VEGF production over time as part of the natural evolution of diabetic retinopathy.
For eyes with DME and worse initial VA, aflibercept continues to exhibit greater VA benefit compared with bevacizumab at 2 years, but the difference in mean VA between participants receiving ranibizumab and aflibercept was no longer statistically significant. Nevertheless, over 2 years, the mean VA improvement was greater for aflibercept compared with bevacizumab or ranibizumab. When initial VA loss is mild, there was little difference in VA outcomes, although bevacizumab thins the retina less than the other 2 agents. Although post hoc analyses should be viewed with caution given the potential for bias, these results, coupled with safety10 and cost-effectiveness3 outcomes, provide further information that is potentially helpful when interpreting DRCR.net Protocol T results through 2 years.
Corresponding Author: Adam R. Glassman, MS, Jaeb Center for Health Research, 15310 Amberly Dr, Ste 350, Tampa, FL (email@example.com).
Accepted for Publication: August 15, 2016.
Published Online: October 6, 2016. doi:10.1001/jamaophthalmol.2016.3698
Author Contributions: Dr Glassman had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Jampol, Glassman, Bressler, Wells.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Jampol, Glassman, Bressler, Wells.
Critical revision of the manuscript for important intellectual content: Jampol, Bressler, Wells, Ayala.
Statistical analysis: Glassman, Ayala.
Administrative, technical, or material support: Jampol, Glassman.
Study supervision: Jampol, Bressler, Wells.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Jampol reported performing data monitoring for Janssen/QLT. Mr Glassman reported receiving grants from the National Institutes of Health, Genentech/Roche, and Regeneron. Dr Bressler reported receiving a subcontract through Northwestern University from the National Eye Institute and clinical or laboratory research grants from Genentech/Roche, Lumenis, Bayer, Novartis, and Regeneron. Dr Wells reported receiving clinical or laboratory research grants from Regeneron and Genentech. Ms Ayala reported receiving grants from the National Institutes of Health, Genentech/Roche, and Regeneron. No other disclosures were reported.
Funding/Support: This study was supported by grants EY14231, EY23207, and EY18817 through a cooperative agreement from the National Eye Institute and the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, US Department of Health and Human Services. Regeneron provided the aflibercept and Genentech provided the ranibizumab for the study. Genentech also provided funding for blood pressure cuffs and the collection of plasma and urine that are not part of the main study results reported in this article.
Role of the Funder/Sponsor: As per the DRCR.net Industry Collaboration Guidelines (available at http://www.drcr.net), the DRCR.net had complete control over the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Regeneron and Genentech had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Group Information: The most recently published list of the Diabetic Retinopathy Clinical Research Network investigators and staff who participated in this study can be found in the eAppendix in the Supplement.
Disclaimer: Dr Bressler is the editor of JAMA Ophthalmology but was not involved in the editorial review or the decision to accept the manuscript for publication.
Create a personal account or sign in to: