Clinical Outcomes of Ahmed Glaucoma Valve Implantation With or Without Intravitreal Bevacizumab Pretreatment for Neovascular Glaucoma: a Systematic Review and Meta-analysis
Zhou M, Xu X, Zhang X, Sun
The objective of this meta-analysis was collating studies comparing placement of the Ahmed Glaucoma Valve (AGV; New World Medical) in the treatment of neovascular glaucoma (NVG) with and without pretreatment using intravitreal bevacizumab (IVB). Zhou and colleagues conducted a literature search using several pertinent keywords in four electronic databases as well as reference lists of retrieved trials. Seven studies met the inclusion criteria for comparative studies of patients who had NVG treated with the AGV, with and without antivascular endothelial growth factor pretreatment, and observed for at least 6 months postoperatively for IOP, medication use, and adverse outcomes.
At A Glance
• A meta-analysis collated studies comparing placement of the Ahmed Glaucoma Valve in the treatment of neovascular glaucoma with and without pretreatment using intravitreal bevacizumab. The investigators concluded that the use of intravitreal bevacizumab can improve outcomes without increasing the risk of adverse effects when used in conjunction with the device for this form of the disease.
• In a follow-up study to the Early Manifest Glaucoma Trial (EMGT), researchers reported that perimetry found progression more frequently than optic nerve photos in patients with preexisting visual field loss.
From these seven articles published between 2009 and 2014, the investigators extracted data regarding the reduction in IOP as a primary outcome. Secondary outcomes were reduction and elimination of glaucoma medications. Tolerability was assessed with a comparison of reported adverse events. Pooled data comparing subjects treated with AGV alone to those with both IVB and AGV were stratified by study design, use of panretinal photocoagulation, and length of follow-up greater or less than 15 months. Finding a significantly greater rate of elimination of postoperative IOP medications in the treated group, the investigators concluded that the use of IVB can improve outcomes without increasing the risk of adverse effects when used in conjunction with the AGV for NVG. The rate of partial medication reduction, however, was not significantly different between the two groups.
What is the evidence in favor of the use of IVB prior to the placement of the AGV in NVG?
The pooled data demonstrated numerically greater but statistically insignificant comparative lowering of postoperative IOP with bevacizumab pretreatment compared to the surgical control group. These results were statistically significant, however, in data pooled from the two randomized controlled trials (RCTs), with an IOP reduction of 8.52 mm Hg (0.85-16.19) when IVB was used. The data revealed an equivalent reduction in the number of glaucoma medications between groups (qualified success), but the pooled data were not statistically significant. Four studies found increased complete success rates—defined as achieving target IOP without glaucoma medications—in the IVB group, with a pooled odds ratio of 3.18 (1.41-7.19).
Is the use of bevacizumab associated with adverse outcomes?
Reported adverse reactions included hyphema, a flat anterior chamber, hypotony, choroidal detachment, and tube obstruction or exposure. Use of bevacizumab was associated with an odds ratio of 0.15 (0.07-0.32) for hyphema, and no significant effect was observed with respect to other adverse outcomes. Thus, this meta-analysis found a statistically significant reduction in hyphema and no increase in other complications, supporting the conclusion that IVB is a safe adjunct to the AGV.
What are the limitations of this meta-analysis?
Noteworthy weaknesses of this analysis include the limited number of studies and the small sample sizes, with particularly few RCTs meeting the inclusion criteria. With the exception of the measurements discussed earlier, statistical significance was not observed in most comparisons. Limited follow-up data and a nonuniform interval between bevacizumab injection and AGV placement may contribute to the heterogeneity of these data. A statistical analysis of heterogeneity was conducted, and it identified high heterogeneity among IOP measurements, suggesting validity limitations of these pooled data for meta-analysis. A 2013 Cochrane review of literature drawn from the same period as this analysis found no RCTs that met inclusion criteria to evaluate the IOP-lowering effects of antivascular endothelial growth factor agents as an adjunctive treatment modality for NVG; those researchers noted insufficient available evidence to evaluate the effectiveness of either IVB or intravitreal ranibizumab.2 Well-designed, longer-term RCTs are needed.
Structural and Functional Progression in the Early Manifest Glaucoma Trial
Ohnell HM, Heijl A, Brenner L, et al 3
In this follow-up study to the Early Manifest Glaucoma Trial (EMGT), the investigators set out to determine whether or not disease progression would manifest initially as changes in the optic nerve structure or visual field. They evaluated optic nerve photos and glaucoma probability change maps for progression in a total of 249 subjects that included 306 study eyes meeting EMGT visual field criteria for manifest glaucoma and 192 fellow eyes without field defects. Changes in visual field progression were detected before alterations in the optic nerve’s appearance at four times the frequency in eyes with manifest glaucoma and with similar frequency for either test in fellow eyes. From these data, the investigators concluded that perimetry finds progression more frequently than optic nerve photos in patients with preexisting visual field loss.
How were optic nerve photos and visual fields assessed for this study?
Pairs of monoscopic optic nerve photos that included the baseline photo—but were masked to temporal order—were examined independently by a panel of three investigators. These images were assessed for changes in the course of small blood vessels as the finding indicating progression. Pallor, peripapillary atrophy, and optic disc hemorrhages were not considered indications of progression, because these findings can be confounded by cataract or photographic exposure. Automated optic disc imaging techniques were not used, because the technology became available later in the study and would have unmasked the temporal order of the photos. Visual fields were determined to progress if three consecutive studies showed three or more locations with statistically significant progression based on glaucoma change probability maps.
Was one test able to detect glaucomatous changes with greater sensitivity?
Progression was detected in 203 study eyes, 81% of which were identified with visual field testing, whereas 19% were identified with optic nerve photos. Confidence intervals did not overlap and demonstrated with statistical significance that visual field testing was the more sensitive technique for detecting progression. This relationship was not observed in fellow eyes not meeting the EMGT criteria for manifest glaucoma at baseline. The investigators did not stratify these data for intervention and observation arms of study eyes, and doing so might have provided additional insight.
What can this study tell us about progression with one method after change has been observed with the other?
According to the EMGT protocol, study eyes randomized to either treatment with a laser and betaxolol or observation were eligible for a change in treatment when disease progression was detected, thus confounding any data that would be collected subsequent to reaching the study endpoint. No conclusions can be drawn about the time from detection of progression with one method to positivity of the other.
Do risk factors affect how progression is initially identified in fellow eyes?
One hundred ninety-two fellow eyes without manifest glaucoma were stratified by risk factors for glaucomatous progression to eyes with an IOP greater than 21 mm Hg, eyes with preperimetric glaucoma (abnormal rim appearance, notch or disc hemorrhage), and otherwise normal eyes. Overall, progression was observed in 62 eyes. Progression overall and within these subgroups was detected by visual field and nerve appearance at similar frequency with no identification of specific risk factors for progression. n
1. Zhou M, Xu X, Zhang X, Sun X. Clinical outcomes of Ahmed Glaucoma Valve implantation with or without intravitreal bevacizumab pretreatment for neovascular glaucoma: a systematic review and meta-analysis. J Glaucoma. 2016;25:551-557.
2. Simha A, Braganza A, Abraham L, et al. Anti-vascular endothelial growth factor for neovascular glaucoma. Cochrane Database Syst Rev. 2013;10:CD007920.
3. Ohnell HM, Heijl A, Brenner L, et al. Structural and functional progression in the Early Manifest Glaucoma Trial. Ophthalmology. 2016;123:1173-1179.
Section Editor James C. Tsai, MD, MBA
• president of New York Eye and Ear Infirmary of Mount Sinai and
chair of ophthalmology for the Mount Sinai Health System in
• New York
Jonathan Paul, MD
• glaucoma fellow, University of Kentucky, Lexington
• (859) 323-5875; email@example.com
• financial interest: none acknowledged
Sheila Sanders, MD
• associate professor of ophthalmology, University of Kentucky, Lexington
• (859) 323-5875; firstname.lastname@example.org
• financial disclosure: on the speaker’s bureau of and has been a research collaborator with Allergan