Abstract
Purpose
To compare the five year outcomes of the Ahmed FP7 Glaucoma Valve (AGV) and the Baerveldt 101-350 Glaucoma Implant (BGI) for the treatment of refractory glaucoma.
Design
Multicenter randomized controlled clinical trial.
Participants
276 patients, including 143 in the AGV group and 133 in the BGI group.
Methods
Patients 18 to 85 years of age with previous intraocular surgery or refractory glaucoma and intraocular pressure (IOP) of ≥ 18 mmHg in whom glaucoma drainage implant surgery was planned were randomized to implantation of either an AGV or BGI.
Main Outcome Measures
IOP, visual acuity, use of glaucoma medications, complications, and failure (IOP > 21 mmHg or not reduced by 20% from baseline, IOP ≤ 5 mmHg, reoperation for glaucoma, removal of implant, or loss of light perception).
Results
At 5 years, IOP (mean ± SD) was 14.7 ± 4.4 mmHg in the AGV group and 12.7 ± 4.5 mmHg in the BGI group (p = 0.012). The number of glaucoma medications in use at 5 years (mean ± SD) was 2.2 ± 1.4 in the AGV group and 1.8 ± 1.5 in the BGI group (p = 0.28). The cumulative probability of failure during 5 years of follow-up was 44.7% in the AGV group and 39.4% in the BGI group (p = 0.65). The number of subjects failing due to inadequately controlled IOP or reoperation for glaucoma was 46 in the AGV group (80% of AGV failures) and 25 in the BGI group (53% of BGI failures, p=0.003). Eleven AGV eyes (20% of AGV failures) experienced persistent hypotony, explantation of implant, or loss of light perception compared to 22 (47% of failures) in the BGI group. The 5-year cumulative reoperation rate for glaucoma was 20.8% in the AGV group compared to 8.6% in the BGI group (p=0.010). Change in logMAR Snellen visual acuity (mean ± SD) at 5 years was 0.42 ± 0.99 in the AGV group and 0.43 ± 0.84 in the BGI group (p=0.97).
Conclusions
Similar rates of surgical success were observed with both implants at 5 years. BGI implantation produced greater IOP reduction and a lower rate of glaucoma reoperation than AGV implantation but BGI implantation was associated with twice as many failures due to safety issues such as persistent hypotony, loss of light perception, or explantation.
Glaucoma drainage implants (GDI) have been used with increasing frequency in the management of glaucoma refractory to trabeculectomy, even in the era of antifibrotic agent use. Medicare data reveals a marked increase in the use of GDIs, from just over 2,000 in 1994 to almost 12,000 in 2012 (William Rich, III, personal communication). In addition, surveys of the membership of the American Glaucoma Society performed in 1996, 2002, and 2008 show a significant increase in the use of GDIs in patients who had undergone prior surgery or who had neovascular or uveitic glaucoma compared to trabeculectomy with mitomycin-C.1–3 This shift in practice pattern has been validated by the results of the Tube Versus Trabeculetomy (TVT) Study,4 which found that patients with prior trabeculectomy and/or prior cataract surgery had a higher success rate with GDI surgery compared with trabeculectomy with mitomycin-C.
GDIs share a common design consisting of a tube that is inserted into the eye through a scleral fistula and shunts aqueous humor to an end plate placed in the equatorial region. They differ with respect to the size and material composition of the end plate, as well as the presence or absence of a valve that restricts aqueous flow if the intraocular pressure (IOP) becomes too low. A limited number of studies exist comparing different implant designs, and most of these are retrospective case series.5 A recent Ophthalmic Technology Assessment of GDIs performed by the American Academy of Ophthalmology’s Technology Assessment Committee concluded that “Too few high-quality direct comparisons of various available shunts have been published to assess the relative efficacy or complication rates of specific devices….”6 The Ahmed Baerveldt Comparison (ABC) and Ahmed Versus Baerveldt (AVB) Studies were initiated to compare the safety and efficacy of the Ahmed Glaucoma Valve (AGV FP7, New World Medical, Cucamonga, CA) and the Baerveldt Glaucoma Implant (BG 101-350, Abbott Medical Optics, Abbott Park, IL), the two most commonly used GDIs in the United States. These randomized prospective clinical trials have shown similar results through 3 years of follow-up.7,8 Specifically, both studies showed a small difference in IOP (1.2 – 1.3 mmHg lower in the BGI group) on slightly fewer medications (0.5 – 0.7 in the BGI group) with more subjects failing due to elevated IOP in the AGV group. The purpose of the current study is to report the 5-year treatment outcomes in the ABC Study.
METHODS
The Institutional Review Board at each of 16 Clinical Centers approved the study protocol before recruitment was started, and each patient gave informed consent. The study was registered at www.clinicaltrials.gov (accessed February 16, 2014). The design and methods of the ABC Study are described in detail in a baseline methodology paper,9 and are summarized as follows.
Randomization and Treatment
Patients age 18–85 years with refractory glaucoma and IOPs ≥to 18 mmHg in whom GDI surgery was planned were enrolled in the study. Patients with primary glaucomas with a previous failed trabeculectomy or other intraocular surgery were included. Also, patients without previous intraocular surgery were eligible if they had secondary glaucomas known to have a higher risk of trabeculectomy failure such as neovascular glaucoma (NVG), uveitic glaucoma, or glaucoma associated with iridocorneal endothelial (ICE) syndrome. Exclusion criteria included no light perception at baseline, uveitic glaucoma secondary to Juvenile Rheumatoid Arthritis, prior glaucoma drainage implant or cyclodestructive procedure, need for concurrent or anticipated (within 6 months) non-glaucoma surgery (cataract, corneal, vitreoretinal, superotemporal scleral buckle or retinal sponge precluding superotemporal placement of an implant, or inability to provide informed consent.
Eligibility was independently confirmed at the Statistical Coordinating Center (SCC) at the Bascom Palmer Eye Institute. Individuals enrolled in the study were randomized to placement of an AGV FP7 or BG 101-350 according to a permuted variable block randomization scheme, stratified by surgeon within Clinical Center and type of glaucoma. Patients were allocated to one of 4 strata according to their type of glaucoma, as follows: (1) Primary glaucomas with previous intraocular surgery; (2) Secondary glaucomas (excluding uveitic glaucoma and NVG); (3) NVG; and (4) Uveitic glaucoma. Neither the subject nor investigator was masked to the randomization assignment. Only one eye of each patient was eligible for enrollment. Details of the inclusion and exclusion criteria, recruitment method, and surgical procedures for AGV and BGI implantation used in this study are described in the baseline paper.9
Patient Visits
Follow-up visits were scheduled one day, one week, one month, three months, six months, one year, 18 months, two years, three years, four years, and five years postoperatively. Information about data obtained at baseline and follow-up visits is contained in the baseline paper.9
Primary and Secondary Outcome Measures
The primary outcome measure was failure, based on consensus definitions contained in the World Glaucoma Association Guidelines on Design and Reporting of Surgical Trials.11 These criteria for failure were defined prospectively as IOP > 21mmHg or less than a 20% reduction below baseline on 2 consecutive study visits after three months, IOP ≤ 5 mmHg on 2 consecutive study visits after three months, reoperation for glaucoma, loss of light perception, or removal of the implant for any reason. Reoperation for glaucoma was defined as additional glaucoma surgery requiring a return to the operating room. Cyclodestruction was counted as a reoperation for glaucoma, irrespective of whether or not the procedure was performed in the operating room. Interventions performed at the slit lamp, such as needling procedures, removal of occluding stents, or laser suture lysis, were not considered glaucoma reoperations. IOP, use of glaucoma medications, visual acuity, visual fields, and rates of surgical complications were secondary outcome measures in the ABC Study. Eyes that had not failed by the above criteria and were not on glaucoma medical therapy were considered complete successes, and those requiring adjunctive medical therapy were defined as qualified successes.
Statistical Analysis
Snellen VA measurements were converted to logarithm of the minimum angle of resolution visual acuity (logMAR VA) equivalents for the purpose of data analysis, as reported previously.12 The time to failure was defined as the time from GDI placement to either reoperation for glaucoma, loss of acuity to no light perception (NLP) in the study eye, or the first of two consecutive follow- up visits after three months in which the patient had persistent hypotony (IOP ≤ 5 mmHg) or inadequately controlled IOP (IOP > 21 mmHg or not reduced by 20%). Data on IOP and numbers of glaucoma medications were censored once a patient underwent a reoperation for glaucoma, explantation of the implant, or loss of light perception, but not after failure due to high IOP, hypotony, or reoperation for a complication. There was no censoring of visual acuity results. Univariate comparisons between treatment groups were performed with the two-sided Student t-test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. Risk factors for treatment failure were assessed for statistical significance with the Kaplan-Meier survival analysis log-rank test. Multivariate analysis was performed with Cox proportional hazard regression analysis with forward stepwise elimination. Patients’ data were analyzed in the group to which they were assigned during randomization (intent-to-treat analysis). A p–value of 0.05 or less was considered statistically significant in our analyses.
RESULTS
Recruitment and Retention
A total of 276 patients were enrolled between October 2006 and April 2008, including 143 patients (52%) who were randomized to placement of an AGV and 133 patients (48%) to a BGI. Protocol violations are described in the baseline paper.9
Retention of patients in the study through 5 years of follow-up is shown in Figure 1. In the overall study group, 174 (63%) patients completed their 5-year visit. This included 87 (61%) patients in the AGV group and 87 (65%) in the BGI group. We compared the numbers of patients who did not complete a five year visit (n=81) by treatment group, excluding from analysis those who had died prior to the end of the 5 year visit window (n=21). No significant difference was observed in the proportion of patients who did not complete 5-year visits in the AGV group, 44 (34%), and the BGI group, 37 (30%) (p = 0.59, Fisher exact test). There was no difference between randomized treatment groups in either mean IOP or mean numbers of IOP lowering medicines between those followed until the next annual visit and those lost to follow up (all p > 0.2, two way analysis of variance with test of interaction between treatment group and loss to follow up during the next year).
Figure 1. Study Flow Diagram.
Recruitment and retention in the ABC Study at five years.
Baseline Characteristics
There were no differences in baseline demographic or clinical characteristics between the two groups, as detailed in the baseline paper.9
IOP Reduction
The baseline and follow-up IOPs for the two groups are reported in Table 1 and Figure 2. Patients who underwent additional glaucoma surgery, removal of the implant, or loss of light perception were censored from analysis after these events. Both study groups experienced a significant reduction in IOP postoperatively. Among patients with 5 year follow-up in the AGV group, IOP (mean ± standard deviation; SD) was reduced from 29.6 ± 10.1 mmHg at baseline to 14.7 ± 4.4 mmHg at the five-year follow-up visit (p < 0.001, paired t-test). In the BGI group, IOP (mean ± SD) was reduced from 28.3 ± 9.3 mmHg at baseline to 12.7 ± 4.5 mmHg at the five-year follow-up visit (p < 0.001, paired t-test). The IOP difference between the two treatment arms at five years was statistically significant (p=0.015) using analysis of covariance, which takes into account preoperative IOP differences. The AGV group had a significantly lower mean IOP than the BGI group at the one day and one week follow-up visits. However, the mean IOP in the BGI group was approximately 1 to 2 mmHg lower than the AGV group thereafter, except at the two-year visit. The 1.3mmHg difference in baseline IOP between AGV and BGI patients who returned for 5 year follow up was not statistically significant (p=0.37). Furthermore, accounting for preoperative IOP with analysis of covariance did not alter the statistical significance of any of the comparisons of post-operative IOPs between the AGV and BGI groups. Mean IOP in the AGV and BGI groups did not vary significantly between the four study strata at any of the annual follow up visits (all p > 0.2, two way analysis of variance with test of interaction).
Table 1.
Intraocular Pressure and Medical Therapy at Baseline and Follow-up in the Ahmed Baerveldt Comparison Study
| Ahmed Group | Baerveldt Group | P-value* | |
|---|---|---|---|
| Baseline IOP (mm Hg) Glaucoma medications N |
31.2 ± 11.2 3.4 ± 1.1 143 |
31.8 ± 12.5 3.5 ± 1.1 133 |
0.71 0.34 |
| 1 year IOP (mm Hg) Glaucoma medications N followed (% of baseline) |
15.4 ± 5.5 1.8 ± 1.3 133 (93%) |
13.4 ± 6.9 1.5 ± 1.4 117 (88%) |
0.018 0.078 |
| 2 years IOP (mm Hg) Glaucoma medications N followed (% of baseline) |
14.5 ± 5.5 1.9 ± 1.3 122 (85%) |
14.2 ± 6.0 1.4 ± 1.5 110 (83%) |
0.76 0.020 |
| 3 years IOP (mm Hg) Glaucoma medications N followed (% of baseline) |
14.4 ± 4.7 2.0 ± 1.4 106 (74%) |
13.1 ± 4.5 1.5 ± 1.4 100 (75%) |
0.078 0.018 |
| 4 years IOP (mm Hg) Glaucoma medications N followed (% of baseline) |
15.5 ± 6.2 2.2 ± 1.7 102 (74%) |
13.4 ± 4.4 1.7 ± 1.4 99 (74%) |
0.017 0.025 |
| 5 years IOP (mm Hg) Glaucoma medications N followed (% of baseline) |
14.7 ± 4.4 2.2 ± 1.4 87 (61%) |
12.7 ± 4.5 1.8 ± 1.5 87 (65%) |
0.015 0.28 |
Data are presented as mean ± standard deviation.
IOP and number of medications are censored after treatment Failure by no light perception, reoperation for glaucoma, or explantation for complication.
IOP = intraocular pressure
Student t-test
Figure 2. IOPs by randomized treatment group and follow-up visit.
Graph of IOP (mmHg) in the ABC Study by study group from preoperative level through five year follow-up visit (mean ± SD).
Medical Therapy
Table 1 also shows the number of glaucoma medications in both groups at baseline and follow-up. Patients who underwent surgery glaucoma reoperation, removal of the implant, or loss of light perception were censored from analysis. There was a significant reduction in the need for medical therapy in both treatment groups (Figure 3). The number of glaucoma medications (mean ± SD) in the AGV group decreased from 3.5 ± 1.0 at baseline to 2.2 ± 1.4 at the five-year follow-up visit (p < 0.001, paired t-test). The number of glaucoma medications (mean ± SD) in the BGI group was reduced from 3.5 ± 1.1 at baseline to 1.8 ± 1.5 at the five-year follow-up visit (p < 0.001, paired t-test). Patients in the AGV group were using significantly more medications at years two through four as compared to the BGI group. There was no statistical difference between treatment groups with regard to the number of medications in use at 5 years or the reduction in medications from baseline to 5 years. Mean number of medications in the AGV and BGI groups did not vary significantly between the four study strata at any of the annual follow up visits (all p > 0.2, two way analysis of variance with test of interaction).
Figure 3. Medication use by randomized treatment group and follow-up visit.
Histogram of the number of classes of ocular hypotensive medication used from before surgery through five year follow-up visit (mean ± SD).
Treatment Outcomes
Table 2 compares the outcomes and reasons for failure of randomized patients, unadjusted for follow-up time. All patients who were seen at the five-year follow-up visit and/or failed during the first five years of the study were included in this analysis. While the total numbers of failures were similar in the two groups, the reasons for treatment failure were different between the AGV and BGI groups (p=0.012, exact chi-square test). The number failing due to inadequately controlled IOP or reoperation for glaucoma was 46 in the AGV group (representing 80% of AGV failures) compared to 25 in the BGI group (53% of BGI failures), a statistically significant difference (p=0.003). Only 11 AGV eyes (20% of AGV failures) experienced persistent hypotony, complications for which explantation was performed, or loss of light perception in the study eye compared to 22 (47% of failures) in the BGI group.
Table 2.
Reasons for Treatment Failure in the Ahmed Baerveldt Comparison Study
| Ahmed Group | Baerveldt Group | |
|---|---|---|
| Inadequate IOP control without additional glaucoma surgery# | 23 (40%) | 17 (36%) |
| Reoperation to lower IOP | 23 (40%) | 8 (17%) |
| Explantation for complication | 3 (5%) | 4 (8%) |
| Persistent hypotony* | 1 (2%) | 6 (13%) |
| Loss of light perception | 7 (12%) | 12 (26%) |
| Total | 57 | 47 |
Data are presented as number (percentage of the total number of Failures in each respective treatment group). There was a statistically significant difference in the distribution of types of Failures between the AGV and BGI implants (p=0.012, exact chi-square test).
IOP = intraocular pressure
IOP > 21 mmHg at 2 consecutive visits after 3 months
IOP ≤ 5 mm Hg at 2 consecutive visits after 3 months
Kaplan-Meier survival analysis was used to compare failure rates between the two treatment groups (Figure 4). The cumulative probability of failure (standard error, SE) was 44.7% (4.6%) in the AGV group and 39.4% (4.6%) in the BGI group at five years (p=0.65, log-rank test). The relative risk of treatment failure in the AGV group was 1.1 times that in the BGI group (95% Confidence Interval (CI): 0.8–1.7, p=0.52 Cox proportional hazards regression). There was no suggestion of different treatment effects in the four randomized strata (p = 0.52, three degree of freedom test of treatment group by stratum interaction). To investigate the timing of failures over follow-up, we calculated annual hazard rates during each of the five study years. The hazard rate (SE) of failure was highest in the first two years of follow-up, 1.5% (0.02%) and 1.3% (0.2%), respectively, and decreased in years three and four, 0.5% (0.2%) and 0.4% (0.8%), with a modest increase in the last year of the study, 0.8% (0.2%). There was no difference in the pattern of hazard rates over time between the study strata.
Figure 4. Kaplan-Meier cumulative surgical failure rates by randomized treatment group.
Kaplan-Meier survival curve of cumulative surgical failures through five years of follow-up by treatment group.
The cumulative proportion of patients undergoing reoperation for glaucoma during 5 years of follow-up was 20.8% in the AGV group compared to 8.6% in the BGI group (p = 0.010, log-rank test; Figure 5). The relative risk of reoperation for glaucoma in the AGV group was 2.6 times that of the BGI group (95% CI:1.2–5.3, p = 0.012; Cox proportional hazards regression). Table 3 presents the specific reoperations for glaucoma performed in the two treatment groups.
Figure 5. Kaplan-Meier cumulative reoperation rates by randomized treatment group.
Kaplan-Meier survival curve of cumulative reoperation rates in through five years of follow-up by treatment group.
Table 3.
Reoperations for Glaucoma in the Ahmed Versus Baerveldt Study
| Ahmed Group (n = 143) |
Baerveldt Group (n = 133) |
|
|---|---|---|
| Additional tube shunt | 13 | 8 |
| Cyclodestructive procedure | 12 | 2 |
| Tube revision followed by cyclodestructive procedure | 1 | 0 |
| Total number of patients (5 year cumulative Kaplan-Meier percentage ± SE) with reoperation for glaucoma* | 26 (20.8 ± 3.7%) | 10 (8.6 ± 2.6%) |
Data are presented as number of patients.
P = 0.010 for the difference in 5-year cumulative reoperation rates for glaucoma between treatment groups from Kaplan-Meier analysis (log rank test adjusted for stratum)
The failure rates for the AGV and BGI treatment groups were examined using alternative outcome criteria. Patients with persistent hypotony, reoperation for glaucoma, or loss of light perception were still classified as treatment failures; however, the upper IOP limit defining success and failure was changed. When inadequate IOP control (with or without medications) was defined as IOP greater than 17 mmHg or not reduced by 20% on two consecutive follow-up visits after three months, the cumulative probability of failure at five years (SE) was 60.4% (4.5%) in the AGV group and 46.1% (4.7%) in the BGI group (p = 0.048, stratified log rank test). When inadequate IOP control was defined as IOP greater than 14 mmHg or not reduced by 20% on two consecutive follow-up visits after three months, the cumulative probability of failure was 77.6% (3.9%) in the AGV group and 64.4% (4.5%) in the BGI group at five years (p = 0.003, stratified log rank test).
Patients with a five-year follow up visit who were still successful through that visit were divided into complete and qualified success based on the requirement for IOP-lowering medical therapy at five years. The number (%) of complete successes at five years was 9 (8%) in the AGV group compared to 14 (14%) in the BGI group (Table 4, p = 0.27). Table 4 also reviews in detail the percent of treatment failures, complete, and qualified success in the two arms of the study by stratum. The study was not adequately powered to reach conclusions about the differences between treatment arms in these subgroups. That being said, there does not appear to be much difference in outcomes within the diagnostic strata.
Table 4.
Treatment Outcomes after 5 Years of Follow-up in the Ahmed Versus Baerveldt Comparison Study
| Ahmed Group | Baerveldt Group | |
|---|---|---|
| Stratum 1—primary glaucomas with previous intraocular surgery Failure Qualified success Complete success |
26 (47%) 25 (46%) 4 (7%) |
18 (35%) 23 (44%) 11 (21%) |
| Stratum 2—secondary glaucomas (excluding neovascular and uveitic glaucomas) Failure Qualified success Complete success |
7 (50%) 5 (36%) 2 (14%) |
7 (58%) 4 (33%) 1 (8%) |
| Stratum 3—neovascular glaucoma Failure Qualified success Complete success |
19 (66%) 9 (31%) 1 (3%) |
20 (71%) 6 (21%) 2 (7%) |
| Stratum 4— uveitic glaucoma Failure Qualified success Complete success |
5 (56%) 2 (22%) 2 (22%) |
2 (33%) 4 (67%) 0 (0%) |
| Overall group Failure Qualified success Complete success* |
57 (53%) 41 (38%) 9 (8%) |
47 (48%) 37 (38%) 14 (14%) |
Data presented as number of patients (percentage).
P = 0.27 for the difference in Complete success rates between treatment groups (binomial logistic regression model including both randomized treatment group and stratum as independent variables)
Because the surgeon was not masked to the treatment assignment, a potential bias existed in the decision to reoperate for IOP control. To evaluate for reoperation bias, the IOP levels were compared between treatment groups among patients who failed because of inadequate IOP control. For cases failing by high IOP at two consecutive study visits without reoperation, the average of the failing IOPs was calculated and compared between the two treatment groups. The failing IOP (mean ± SD) in the AGV group was 20.0 ± 4.4 mmHg compared to 23.0 ± 6.4 mm Hg in the BGI group (p=0.089, two-sample t237 test). The IOP immediately prior to glaucoma reoperation was also compared between treatment groups. Among AGV cases reoperated for glaucoma, the preoperative IOP (mean ± SD) immediately prior to reoperation was 28.9 ± 9.0 mm Hg compared to 29.4 ± 6.3 mm Hg in the BGI group (p=0.90).
Visual Acuity
Visual acuity results are shown in Table 5. There was a significant decrease in Snellen VA in both treatment groups during the five years of follow-up. In the AGV group, logMAR Snellen VA (mean ± SD) decreased from 1.07 ± 1.01 at baseline to 1.42 ± 1.15 at the five-year follow-up visit (p <0.001, paired t-test). In the BGI group, logMAR Snellen VA (mean ± SD) decreased from 1.04 ± 1.00 at baseline to 1.43 ± 1.40 at the five-year follow-up visit (p < .001, paired t-test). There was no significant difference in logMAR Snellen VA between the two groups at five years (p = 0.97, student t-test).
Table 5.
Visual Acuity Results in the Ahmed Baerveldt Comparison Study
| Ahmed Group N=86 |
Baerveldt Group N=87 |
P-value† | |
|---|---|---|---|
| Snellen VA, logMAR mean ± SD Baseline (n=276) 5 years (n=174) Change at 5 years (n=174) |
1.07 ± 1.01 1.42 ± 1.15 0.42 ± 0.99 |
1.04 ± 1.00 1.43 ± 1.40 0.43 ± 0.84 |
0.80 0.94 0.97 |
| Loss of ≥ 2 Snellen lines at 5 years, n (%)* Glaucoma Retinal disease Corneal opacity, edema, graft Failure Cataract Other‡‡ Unknown |
36 (42%) 14 (39%) 10 (28%) 3 (8%) 3 (8%) 1 (3%) 5 (14%) |
38 (44%) 17 (45%) 5 (13%) 10 (26%) 3 (8%) 5 (13%) 2 (5%) |
0.88‡ |
logMAR = Logarithm of the Minimum Angle of Resolution
Patients may have more than one reason for decreased vision.
Two sided Student t-test
Fisher exact test
Other reasons for vision loss included phthisis bulbi (n=3), posterior capsule opacification (n=2), inability to perform acuity test (Alzheimer’s Disease, n=1)
Snellen VA was decreased by two or more lines from baseline in 36 (42%) patients in the AGV group and 38 (44%) patients in the BGI group at five years, and this difference was not significantly different (table 5, p = 0.88, Fisher’s exact test). The most frequent causes of vision loss during five years of follow-up were glaucoma, retinal disease, and anterior segment pathology. The reason for decreased vision was unknown in 5 (14%) patients in the AGV group and 2 (5%) patients in the BGI group. The other miscellaneous cause for reduced vision in the AGV group was due to a patient with Alzheimer’s who did not perform the acuity test well. Other causes of vision loss in 5 patients in the BGI group included phthisis bulbi and posterior capsular opacification. There were no significant differences in the reasons for visual acuity loss between the two treatment groups. Of 161 patients with visual acuities measured at both the three and five year visits, 32 patients (20%) lost 2 or more Snellen lines of acuity between their 3 and 5 year visits. Reasons for acuity loss were glaucoma alone or in combination with another cause in 14 patients (44%), retinal disease in the absence of glaucoma in 7 patients (22%), corneal disease in 5 patients (16%), cataract alone in 1 patient (3%), and in 4 patients (13%) the reason was not recorded.
Twenty-five patients (9%) progressed to NLP vision, 6 of which had previously failed by one of the other criteria, and all but one of these (96%) of these were in the neovascular glaucoma (NVG) stratum. We compared the incidence of NLP between randomized treatment groups among the 80 NVG patients. At 5 years, the cumulative proportion of NVG patients who progressed to NLP in the AGV group was 28.3% (SE=8.9%) compared to 51.1% (SE=9.2%) in the BGI group, a difference which was statistically significant (p=0.030, log-rank test). In the judgment of the surgeons, neovascular eyes which lost light perception in the Ahmed group did so for the following reasons: glaucoma (4), progressive diabetic retinopathy (3), no reason provided (1). In the Baerveldt group reasons included: macular disease (2), phthisis bulbi (3, 1 following retinal detachment), vitreous hemorrhage/hyphema (1), glaucoma (2), enucleation of painful eye (1), unable to determine due to anterior segment pathology (1), progressive diabetic retinopathy (1), ischemia (1), no reason provided (4).
DISCUSSION
The ABC Study is a multicenter prospective clinical trial comparing the two most popular GDIs. Patients with previous intraocular surgery or refractory glaucoma were enrolled in the study and randomly assigned to surgical treatment with the Ahmed FP7 and Baerveldt BG 101-350. Baerveldt implantation was more effective in providing long-term IOP control than Ahmed implantation. The BGI produced greater IOP reduction with fewer adjunctive medications and required fewer glaucoma reoperations compared with the AGV during 5 years of follow-up.
We recognize that the goal of glaucoma therapy is the prevention of further glaucomatous optic nerve damage and visual field loss with preservation of visual function. The degree of IOP reduction is a surrogate for successful glaucoma therapy, primarily because IOP is easily measurable and the only known treatable risk factor for glaucoma progression. As such, it serves as an important measure of surgical success. Both the AGV and BGI produced profound reductions in IOP, from baseline averages of 31 – 32 mmHg to final average IOPs of 14.7 mmHg in the AGV group and 12.7 mmHg in the BGI group. The total IOP reduction was greater than 50% in both treatment groups, which is comparable to previous studies of GDIs.5 The BGI group had a mean IOP approximately 2 mmHg lower IOP than the AGV group at most of the annual study visits, including at 5 years, and this represents a statistically significant difference. The lower IOPs in the BGI group were achieved with fewer glaucoma medications compared with the AGV group at most time intervals.
There are two reasons that may be offered to explain the superior IOP control observed with the BGI relative to the AGV. First, studies have shown that glaucoma drainage devices with larger end plates result in lower IOPs.5 Heuer and colleagues13 found higher success rates and lower long-term IOPs using the double-plate compared to the single-plate Molteno implant. However, there may be an upper limit of benefit of end plate size, as Britt et al14 subsequently noted similar outcomes when comparing the Baerveldt 500 mm2 implant to the 350 mm2 in a prospective clinical trial. A recent retrospective study by Seah et al15 comparing the Baerveldt 250 mm2 versus 350 mm2 implant found no difference in final IOP between the two implant sizes. A prospective randomized trial comparing these two end plate sizes is underway (clinicaltrials.gov, registered July 8, 2010). The second possible explanation for lower long300 term IOPs with the BGI relates to exposure of the filtering bleb to postoperative inflammatory material. In the valved AGV, there is immediate flow of aqueous to the bleb, exposing it to inflammatory cells, cytokines, and proteins resulting from the surgery, which may produce more vigorous scarring of the fibrous capsule surrounding the end plate.5,16 In the non-valved BGI, complete occlusion of the tube for the first four to six weeks is critical to prevent early hypotony and hypotony-related complications such as flat anterior chambers, choroidal effusions, and suprachoroidal hemorrhages.17 By occluding the BGI for a period of several weeks, the bleb is exposed to much less inflammatory material. Whatever the explanation, the larger, non-valved BGI tends to produce better long-term IOP control, which may make it the preferred implant in patients in whom one is trying to achieve the lowest possible IOP postoperatively.
The primary outcome in the ABC Study was cumulative failure rate at 5 years. Approximately 40% of subjects in both groups failed by criteria defined a priori, based on failure criteria recommended by a consensus group of the World Glaucoma Association.11 It is interesting to note that the two treatment groups failed at approximately the same rate, but they did so for different reasons. The AGV group failed due to high IOP endpoints, while the BGI group failed due to safety endpoints. Higher IOPs in the AGV group resulting in failure or reoperation for glaucoma may be related to the smaller end plate or immediate release of inflammatory factors to the sub-Tenon’s space, as discussed above. The higher rate of hypotony in the BGI group is likely related to the larger size of the end plate and the lack of a flow restrictor, the same design features that resulted in fewer failures due to lack of IOP control. A higher rate of surgical success was seen with the BGI compared with the AGV in post hoc analyses when IOP failure was stringently defined as IOP greater than 14 mm Hg.
Only 8% of subjects who underwent AGV implantation and 14% of those undergoing BGI implantation had controlled IOP without medications at 5 years (complete success). In the TVT study, the tube (BGI) group had a complete success rate of 25% but the subjects in the TVT study were at lower risk of surgical failure than the current study since the TVT study excluded patients with secondary glaucomas such as ICE syndrome, uveitis, and neovascular glaucoma. Table 4 shows the complete success rate by stratum in the ABC study at 5 years. In stratum 1, which is identical to the subjects addressed by the TVT study, the complete success rate is 21%, very similar to the 25% complete success rate in the TVT BGI group.
The rate of reoperation for glaucoma was higher in the AGV group relative to the BGI group. Patients who required additional glaucoma surgery underwent placement of a second GDI or cyclodestruction in both treatment groups. Because investigators were not masked to the treatment assignment and the decision to reoperate was left to the surgeon’s discretion, a potential for bias existed in the decision to reoperate for glaucoma. No significant difference in mean IOP at the time of failure was seen between treatment groups in patients who had a reoperation for glaucoma, or in patients who failed because of inadequate IOP reduction but did not have additional glaucoma surgery. These observations suggest that no selection bias was present for glaucoma reoperation.
Visual acuity decreased in both treatment groups during the five years of follow-up. Approximately 43% of subjects lost 2 or more lines of Snellen visual acuity. Snellen acuity was the same in the treatment groups at year 5 and no significant differences in the rates and reasons for vision loss were present in the AGV and BGI groups. Many of the causes of vision loss, such as progression of diabetic retinopathy or age related macular degeneration, were not directly attributable to the surgical procedures under study. Compared to the 3 year study results,7 there were no additional subjects who lost 2 or more lines of vision but there were eight additional subjects in the BGI group who lost 2 or more lines of vision. The proportion of subjects who lost 2 or more lines of vision in the current study and the magnitude of vision lost between the preoperative and 5 year visit was very similar to that seen in the 5-year results of the TVT Study.4
Several retrospective case series have compared the AGV and BGI.18–22 Unfortunately, the surgeon’s GDI selection in these studies may have been influenced by the patient’s presumed risk of failure and could bias the results. Randomized clinical trials are designed to produce comparison groups that differ only by the treatment provided, and they offer the highest level of evidence-based medicine. The Ahmed Versus Baerveldt (AVB) Study is another multicenter randomized prospective clinical trial comparing the safety and efficacy of the AGV and BGI. Both the ABC and AVB Studies observed significantly greater long-term IOP reduction and less need for glaucoma medical therapy with the BGI compared to the AGV with similar success rates after 3 years of follow-up. The similarity in results between these clinical trials has allowed each study to validate the other.
There are several limitations to the ABC Study. Neither the patient nor the surgeon was masked to the implant used. The study only evaluated the AGV and BGI, and the results cannot be extrapolated to other GDIs or different models of the AVG or BGI. Patients were excluded if other ocular procedures were required in conjunction with glaucoma surgery, so the study does not provide information about the preferred implant when concurrent ocular surgery is needed. While aspects of both surgical procedures were standardized, some variation in surgical technique occurred between surgeons. We felt that it was important to provide latitude for the surgeon to perform the procedures under study in a manner in which he/she was proficient. Also, the results apply only to the diagnostic groups included in the study. Specifically, these results cannot be generalized to patients without prior incisional surgery who are low-risk for failure of standard surgery (such as trabeculectomy).
The ABC Study does not demonstrate clear superiority of one implant over the other. In addition to efficacy and safety data, there are other important considerations. The individual patient characteristics and surgeon’s comfort and experience with each implant are critical in device selection. The benefits of each implant in reducing IOP must be interpreted in light of its surgical complications (manuscript forthcoming). The valve mechanism of the AGV allows the implant to function immediately postoperatively, and this may be particularly advantageous in patients with markedly elevated IOP preoperatively. For instance, patients with neovascular glaucoma with completely closed anterior chamber angles typically have markedly elevated IOP that is unresponsive to medical therapy and need immediate IOP lowering; one would prefer a valved implant in this instance since one would not want to wait the typical 5 – 7 weeks for a suture ligature to dissolve in an non-valved implant such as the BGI. In addition, NVG patients typically do not have significant glaucomatous cupping at presentation since their IOP has been elevated for a relatively short period of time. For these reasons, perhaps one would prefer the smaller-plated valved AGV implant and be willing to sacrifice the modestly lower average IOP achievable with the larger, non-valved BGI implant. The AGV implant may also be preferred in patients at greater risk for postoperative hypotony, such as those with uveitic glaucoma or prior cyclodestruction. In these patients, decreased aqueous humor production may induce hypotony if there is excess outflow in a large non-valved implant such as the BGI. In either group of patients, NVG or inflammatory glaucomas, if the IOP is too high in the long-term, a larger, non-valved implant can usually be placed in a second quadrant. The above clinical suggestions should be backed up, however, with future properly powered randomized clinical trials since the current study did not have enough subjects in these subgroups to come to definitive conclusions on which implant is best used in which subgroup.
It is interesting to compare the results of the ABC Study at three and five years. From zero to three years, the failure rates in the two groups were approximately 10% per year, with a cumulative failure rate of 30% in both groups at year three. From three to five years, the failure rate seems to flatten such that an additional 10% of subjects failed in the last two years of follow up for a rate of 5% failure per year. It seems that, once patients make it through the first 3 years there is a lower rate of failure going forward, although longer follow-up would be helpful to confirm this. Also, it is interesting to note that the IOP and number of medications remained stable between years three to five as the had been in years one three. Similar to the five year results of BGI 101-350 in the Tube vs. Trabeculectomy Study,4 IOP was, on average, between 13 and 15 mmHg on an average of two medications.
In summary, BGI implantation produced greater IOP reduction and a lower incidence of glaucoma reoperation than AGV implantation after 5 years of follow-up. The AGV decreased IOP to a greater degree in the early postoperative period compared with the BGI. Similar rates of surgical success were observed with both implants during 5 years of follow-up, but the reasons for treatment failure were different. Failure after AGV was usually due to high IOP endpoints, while failure with the BGI was most commonly related to safety endpoints (hypotony, implant explantation, and loss of light perception). The approximate 2 mmHg additional IOP lowering obtained with the BGI must be weighed against the higher number of safety endpoints in the BGI group compared to the AGV group. A detailed account of the complications following 5 years of follow-up from this study is forthcoming.
Acknowledgments
Supported by: NIH P30 EY014801 and unrestricted grants from Research to Prevent Blindness and New World Medical, Inc.
The Ahmed versus Baerveldt Comparison Study Group
Clinical Centers:
Duke University Eye Center, Leon Herndon, MD;
Glaucoma Associates of Texas, David Godfrey, MD, Ronald Fellman, MD;
Medical College of Wisconsin, The Eye Institute, James Robinson, MD, David Dueker, MD;
Minnesota Eye Consultants, Patrick Riedel, MD, Thomas Samuelson, MD;
NIHR Biomedical Research Centre for Ophthalmology, Moorfields Eye Hospital, Keith Barton, MD, Renata Puertas, MD;
National University Hospital, Department of Ophthalmology, Paul Chew, MD, Cecilia Aquino, MD;
Southern California Glaucoma Consultants, Alfred M Solish, MD;
Toronto Western Hospital Eye Clinic, Yvonne Buys, MD, Graham Trope, MD;
University of California Davis Medical Center, Department of Ophthalmology, James Brandt, MD, Michele Lim, MD;
University of California Los Angeles, Jules Stein Eye Institute, Simon Law, MD;
University of Campinas, Ophthalmology, Vital Costa, MD;
University of Miami, Bascom Palmer Eye Institute, Donald Budenz, MDi, Steven Gedde, MD, Fouad El Sayyad, MD;
University of Oklahoma, Dean A. McGee Eye Institute, Steve Sarkisian, MD;
University of Southern California, Doheny Eye Institute, Vikas Chopra, MD, Brian Francis, MD, Mario Meallet, MD, Rohit Varma, MD;
University of Tennessee at Memphis, Department of Ophthalmology, Peter Netland, MD,ii PhD, Sarwat Salim, MDiii;
University of Texas Houston, Cizik Eye Clinic, Robert Feldman, MD, Nicholas Bell, MD;
Safety and Data Monitoring Committee, Voting Members: Philip Chen, MD, University of Washington, Department of Ophthalmology; Dale Heuer, MD, Medical College of Wisconsin, The Eye Institute; Kuldev Singh, MD, MPH, Stanford University, Department of Ophthalmology; Martha Wright, MD, University of Minnesota, Department of Ophthalmology. Non-voting members: Donald L Budenz, MD, MPH, William J. Feuer, MS, Joyce C. Schiffman, MS
Steering Committee: Keith Barton, MD, Donald L Budenz, MD, MPH, William J. Feuer, MS.
Statistical Coordinating Center: William J. Feuer, MS, Joyce C. Schiffman, MS, Wei Shi, MS; Coordinator, Luz Londono; Database Manager, Yolanda Silva, BS, University of Miami, Bascom Palmer Eye Institute.
iCurrently affiliated with University of North Carolina at Chapel Hill
iiCurrently affiliated with University of Virginia
iiiCurrently affiliated with Medical College of Wisconsin
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest: No conflicting relationship exists for any author
REFERENCES
- 1.Chen PP, Yamamoto T, Sawada A, et al. Use of antifibrosis agents and glaucoma drainage devices in the American and Japanese Glaucoma Societies. J Glaucoma. 1997;6:192–196. [PubMed] [Google Scholar]
- 2.Joshi AB, Parrish RK, II, Feuer WF. 2002 survey of the American Glaucoma Society: practice preferences for glaucoma surgery and antifibrotic use. J Glaucoma. 2005;14:172–174. doi: 10.1097/01.ijg.0000151684.12033.4d. [DOI] [PubMed] [Google Scholar]
- 3.Desai MA, Gedde SJ, Feuer WJ, et al. Practice preferences for glaucoma surgery: a survey of the American Glaucoma Society in 2008. Ophthalmic Surg Lasers Imaging. 2011;42:202–208. doi: 10.3928/15428877-20110224-04. [DOI] [PubMed] [Google Scholar]
- 4.Gedde SJ, Schiffman JC, Feuer WJ, et al. Tube Versus Trabeculectomy Study Group. Treatment outcomes in the Tube Versus Trabeculectomy (TVT) Study after five years of follow-up. Am J Ophthalmol. 2012;153:789–803. doi: 10.1016/j.ajo.2011.10.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gedde SJ, Panarelli JF, Banitt MR, Lee RK. Evidenced-based comparison of aqueous shunts. Curr Opin Ophthalmol. 2013;24:87–95. doi: 10.1097/ICU.0b013e32835cf0f5. [DOI] [PubMed] [Google Scholar]
- 6.Minckler DS, Francis BA, Hodapp EA, et al. Aqueous shunts in glaucoma: a report by the American Academy of Ophthalmology. Ophthalmology. 2008;115:1089–1098. doi: 10.1016/j.ophtha.2008.03.031. [DOI] [PubMed] [Google Scholar]
- 7.Barton K, Feuer WJ, Budenz DL, et al. Ahmed Baerveldt Comparison Study Group. Three-year treatment outcomes in the Ahmed Baerveldt Comparison Study. Ophthalmology. 2014;121:1547–1557. doi: 10.1016/j.ophtha.2014.01.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Christakis PG, Tsai JC, Kalenak JW, et al. The Ahmed Versus Baerveldt Study: three-year treatment outcomes. Ophthalmology. 2013;120:2232–2240. doi: 10.1016/j.ophtha.2013.04.018. [DOI] [PubMed] [Google Scholar]
- 9.Barton K, Gedde SJ, Budenz DL, et al. Ahmed Baerveldt Comparison Study Group. The Ahmed Baerveldt Comparison Study: methodology, baseline patient characteristics, and intraoperative complications. Ophthalmology. 2011;118:435–442. doi: 10.1016/j.ophtha.2010.07.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Budenz DL, Barton K, Feuer WJ, et al. Ahmed Baerveldt Comparison Study Group. Treatment outcomes in the Ahmed Baerveldt Comparison Study after 1 year of follow-up. Ophthalmology. 2011;118:443–452. doi: 10.1016/j.ophtha.2010.07.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Heuer DK, Barton K, Grehn F, et al. Consensus on definitions of success. In: Shaarawy TM, Sherwood MB, Grehn F, editors. Guidelines on Design and Reporting of Surgical Trials. Amsterdam, the Netherlands: Kugler: World Glaucoma Association; 2008. pp. 15–24. [Google Scholar]
- 12.Fluorouracil Filtering Surgery Study Group. Five-year follow-up of the Fluorouracil Filtering Surgery Study. Am J Ophthalmol. 1996;121:349–366. doi: 10.1016/s0002-9394(14)70431-3. [DOI] [PubMed] [Google Scholar]
- 13.Heuer DK, Lloyd MA, Abrams DA, et al. Which is better? One or two? A randomized clinical trial of single-plate versus double-plate Molteno implantation for glaucomas in aphakia and pseudophakia. Ophthalmology. 1992;99:1512–1519. doi: 10.1016/s0161-6420(92)31772-5. [DOI] [PubMed] [Google Scholar]
- 14.Britt MT, LaBree LD, Lloyd MA, et al. Randomized clinical trial of the 350-mm2 versus the 500-mm2 Baerveldt implant: longer term results: is bigger better? Ophthalmology. 1999;106:2312–2318. doi: 10.1016/S0161-6420(99)90532-8. [DOI] [PubMed] [Google Scholar]
- 15.Seah SKL, Gazzard G, Aung T. Intermediate-term outcome of Baerveldt glaucoma implants in Asian eyes. Ophthalmology. 2003;110:888–894. doi: 10.1016/S0161-6420(03)00088-5. [DOI] [PubMed] [Google Scholar]
- 16.Freedman J, Iserovich P. Pro-inflammatory cytokines in glaucomatous aqueous and encysted Molteno implant blebs and their relationship to pressure. Invest Ophthalmol Vis Sci. 2013;54:4851–4855. doi: 10.1167/iovs.13-12274. [DOI] [PubMed] [Google Scholar]
- 17.Nguyen Q, Budenz DL, Parrish RK., II Complications of Baerveldt glaucoma implants. Arch Ophthalmol. 1998;116:571–575. doi: 10.1001/archopht.116.5.571. [DOI] [PubMed] [Google Scholar]
- 18.Tsai JC, Johnson CC, Dietrich MS. The Ahmed shunt versus the Baerveldt shunt for refractory glaucoma: a single-surgeon comparison of outcome. Ophthalmology. 2003;110:1814–1821. doi: 10.1016/S0161-6420(03)00574-8. [DOI] [PubMed] [Google Scholar]
- 19.Syed HM, Law SK, Nam SH, et al. Baerveldt-350 implant versus Ahmed valve for refractory glaucoma: a case-controlled comparison. J Glaucoma. 2004;13:38–45. doi: 10.1097/00061198-200402000-00008. [DOI] [PubMed] [Google Scholar]
- 20.Wang JC, See JL, Chew PT. Experience with the use of Baerveldt and Ahmed glaucoma drainage implants in an Asian population. Ophthalmology. 2004;111:1383–1388. doi: 10.1016/j.ophtha.2003.11.005. [DOI] [PubMed] [Google Scholar]
- 21.Tsai JC, Johnson CC, Kammer JA, Dietrich MS. The Ahmed shunt versus the Baerveldt shunt for refractory glaucoma II: longer-term outcomes from a single surgeon. Ophthalmology. 2006;113:913–917. doi: 10.1016/j.ophtha.2006.02.029. [DOI] [PubMed] [Google Scholar]
- 22.Goulet RJ, III, Phan AD, Cantor LB, WuDunn D. Efficacy of the Ahmed S2 glaucoma valve compared with the Baerveldt 250-mm2 glaucoma implant. Ophthalmology. 2008;115:1141–1147. doi: 10.1016/j.ophtha.2007.10.034. [DOI] [PubMed] [Google Scholar]