The Ahmed versus Baerveldt Study at King Khaled Eye Specialist Hospital: Three-Year Treatment Outcomes.

PURPOSE: The purpose of this study was to analyze the outcomes of two frequently used surgical valves in treating refractory glaucoma.
METHODS: This was a retrospective and nonrandomized study comparing patients aged 18 years or older who underwent implantation using standardized surgical techniques.
RESULTS: A total of 86 patients were included in the study, 48 in the Ahmed group and 38 in the Baerveldt group. The overall success rate was 63.1% in both the groups. At the 3-year follow-up, the Ahmed group had a mean intraocular pressure (IOP) of 14.0 ± 4.8 mmHg (60% reduction) compared with 15.8 ± 6.2 mmHg (53.3% reduction) in the Baerveldt group (0.536). The Ahmed group required an average of 1.6 ± 1.3 medications (59% reduction) compared with 2.1 ± 1.7 (40% reduction) in the Baerveldt group (P < 0.001).
CONCLUSION: Despite a high failure rate, both devices were effective in lowering IOP and the need for medications. Lower IOP and medications were needed in the Ahmed group. Copyright:

Introduction

Glaucoma is one of the leading causes of irreversible vision loss globally.[1] Worldwide in 2013, it was predicted that 64.3 million people (aged 40–80 years) had glaucoma, and this total is predicted to reach 76 million by 2020 and 112 million by 2040.[2] The prevalence and burden of glaucoma are prone to growth in future due to a rise in glaucoma risk factors, such as aging population and diabetes.[34] Measured among global determent causes of blindness, it is estimated that glaucoma represents 8% of all causes, adjoining cataract at 51%.[1] Provided that, the sum of years of life lost through premature death and years lived with disability (disability-adjusted life years) for glaucoma is 4.7 million.[5] These years represent the gap between the ideal health situation and the current health status.

Treating glaucoma can be established by reducing the intraocular pressure (IOP) by 20%–50%, which is achieved by drug treatment, laser therapy, or surgery.[6] In moderate-to-severe glaucoma, trabeculectomy or aqueous drainage shunts are used commonly as the first line.[7] In the recent years from 1996 to 2008, an increase in the use of the glaucoma drainage device from 17.5% to 50.8% and a decrease in the usage of trabeculectomy from 80.8% to 45.5% among the American Glaucoma Society was observed.[8] Furthermore, in general practice, a 184% increase in aqueous shunting devices procured was measured from 2728 cases in 1995–7744 in 2004.[9] In the 5-year follow-up, supplemental medical therapy and efficacy of lowering the IOP were almost the same, with a 49.5% reduction in IOP with trabeculectomy and mitomycin C, and 41.4% with Baerveldt tube.[9] Nevertheless, the failure rate and revision surgeries were lower using the tube.[9]

Two of the most popular shunts are usually used, Ahmed-FP7 valve (New World Medical, Inc., Rancho Cucamonga, CA, USA) or Baerveldt-350 implant (Abbott Medical Optics, Inc., Santa Ana, CA, USA). In a meta-analysis done by Wang et al. that included 1048 eyes from 10 clinical trials from 2003 to 2015, comparing the two valves suggested that the success rate for short-term complications (6–18 m) during follow-up was the same.[10] On the other hand, long-term (>18 m) complications at follow-up concluded that the daring system of Ahmed glaucoma valve (AGV) had a lower success rate.[10] The effectiveness of Baerveldt glaucoma implant (BGI) in controlling IOP and requiring lesser medications postoperatively was higher than AGVs.[10] However, AGV had a lower incidence of severe and total complications.[10] In this study, we intended to compare the results of a 3-year follow-up regarding the efficacy and safety of two drainage systems.

Methods

After the approval from the Ethical Committee in King Khaled Eye Specialist Hospital (KKESH) for a retrospective, nonrandomized, comparative study, the review chart of glaucoma patients that underwent implantation of AGV-FP7 or Baerveldt-350 implant was collected. This study enrolled patients who underwent implantation using standardized surgical technique from October 2015 to June 2018 by several physicians in KKESH. Study Manual created by Christakis PG. et al.[11] was used in this study and will be briefly summarized here. Data collected included baseline characteristics, preoperative assessment, and follow-up visits that evaluated visual acuity, IOP, number of glaucoma medication, ocular examination, and postoperative complications or interventions at 1 day, 1 week, 2 weeks, 1 month, 2 months, 3 months, 6 months, 12 months, 18 months, 2 years, and 3 years postoperatively. Consequently, outcomes were classified as primary or secondary. Furthermore, all of our methods obligated to the Declaration of Helsinki. Patients were not directly involved in the design of this study.

Inclusion criteria

Retrospective nonrandomized control trial included inadequately controlled glaucoma patients (IOP not reaching clinical target despite using conventional medical interventions, such as laser or surgical therapy) aged 18 years or older that underwent Ahmed-FP7 valve or Baerveldt-350 implantation using standardized surgical technique. One eye was included for each patient.

Exclusion criteria

Any patient that had a missed follow-up was excluded. Eyes that had two or more glaucoma drainage implants (GDIs), had a GDI replacement surgery, or eyes that had interventions added to the GDI surgery, were also excluded.

Baseline data collection

The following data were collected: age, sex, glaucoma diagnosis and cause, ocular history (interventions and laser therapy), best-corrected visual acuity with Snellen chart, Goldmann applanation tonometry calculated IOP, glaucoma medications, and abnormal slit-lamp or fundoscopy findings.

Surgical procedures

The surgical procedures used were standardized according to the Ahmed Versus Baerveldt Study Manual,[11] and complications occurred during surgery were recorded.

Follow-up schedule and study measurements

Patients were scheduled to be seen at 1 day, 1 week, 1 month (±2 weeks), 3 months (±2 weeks), 6 months (±2 weeks), 12 months (±2 weeks), 18 months (±2 weeks), 2 years (±2 weeks), and 3 years (±2 weeks) after surgery. At each visit, the following data were collected:

Visual Acuity: best-corrected visual acuity through Snellen chart and low vision chart were recruited

IOP: Goldmann, Tono-Pen, and pneumotonometer were recruited

Glaucoma medications

Ocular examination: documented systematic and comprehensive slit-lamp biomicroscopy and dilated fundoscopy ocular examination were obtained

Complications: during any visit, a record of complications resulting from the surgery or glaucoma progression was obtained along with the intervention to these complications.

Outcome measures

Figure 1 depicts the outcome measures in both the primary and secondary outcomes; the primary result held three categories.

Figure 1

Outcome measures

Complete success

When all of the following criteria are met: (1) IOP of 5–18 mmHg or more than 20% reduction in all visits after 3 months, (2) no glaucoma medications, (3) no significant vision loss (>2 Snellen lines), (4) no vision-threatening complications (endophthalmitis, choroidal effusion, or suprachoroidal hemorrhage), and (5) no surgical intervention.

Qualified success

When all of the following criteria are met: (1) IOP of 5–18 mmHg or more than 20% reduction in nonconsecutive visits, (2) allow glaucoma medications, and (3) surgical intervention for nonvision-threatening complication.

Failure

If any of the following criteria are met: (1) IOP outside the 5–18 mmHg range or <20% reduction for 2 consecutive visits after 3 months, (2) vision-threatening complications, (3) de novo glaucoma procedure, and (4) no light perception (NLP) vision.

Secondary outcome measures

Success was analyzed using 2 alternative intraocular pressures (IOPs), 14 mmHg and 21 mmHg, as recommended by the World Glaucoma Association's Guidelines on Design and Reporting of Surgical Trials.[12]

Statistical analysis

Data were collected and stored in a spreadsheet using Microsoft Excel 2010® software. Data were analyzed using SPSS® version 21.0 (IBM Inc., Chicago, Illinois, USA).

Descriptive analysis was done, where categorical variables were presented in the form of frequencies and percentages and continuous variables in the form of mean (± standard deviation), range (minimum to maximum) for the normally distributed variables, and median (interquartile range). Independent t-test or Mann–Whitney U-test was used to compare the means between the two groups for the normally and nonnormally distributed variables. Chi-squared test was used to compare proportions between the groups. Kaplan–Meier survival curves were plotted to show the percentage of survival/failure, and log rank (Mantel–Cox) was used to test the differences between the survival rates. Any output with P < 0.05 was interpreted as an indicator of statistical significance.

Results

Baseline characteristics

Forty-eight (55.8%) eyes receiving AGV and 38 (44.2%) eyes receiving BGI were collected, and a comparative chart review was carried out. The baseline characteristics, such as age, sex, proportion of right eyes, diagnosis, and preoperative ocular and surgical history, are plotted in Table 1. The mean age of patients was 59, with many of them having secondary complicated glaucoma, including neovascular glaucoma (26.7%), or having undergone previous glaucoma surgery, such as trabeculectomy (14%), trans-scleral cyclodestructive procedure (6%), deep sclerectomy (3%), and endocyclophotocoagulation (1.2%). The median preoperative IOP was similar in both the groups (P = 0.66, Mann–Whitney U-test), where it was 35 mmHg in the AGV group and 34 mmHg for the BGI group, and 85% of them were on 4 classes of glaucoma medications (P = 0.002).

Table 1

Baseline demographic and ocular characteristics

Characteristics	Overall (n=86 eyes)	Ahmed-FP7 (n=48 eyes)	Baerveldt-350 (n=38 eyes)	P
Age, median (IQR)	59.5 (33.0-66.0)	59.5 (45.5-72.3)	58.0 (23.8-63.3)	0.052±
Sex: female	37 (43.0)	19 (39.6)	18 (47.4)	0.469
Glaucoma diagnosis
Primary open angle glaucoma	12 (14.0)	7 (14.6)	5 (13.2)	0.853
Secondary open angle glaucoma	10 (11.8)	7 (14.6)	3 (7.9)	0.339
Primary closed angle glaucoma	10 (11.6)	6 (12.5)	4 (10.5)	0.775
Secondary angle closure glaucoma	4 (4.7)	2 (4.2)	2 (5.3)	0.812
Congenital glaucoma	9 (10.5)	4 (8.3)	5 (13.2)	0.464
Juvenile open angle glaucoma	3 (3.5)	0 (0.0)	3 (7.9)	0.049
Neovascular glaucoma	23 (26.7)	16 (33.3)	7 (18.4)	0.123
Combined mechanism glaucoma	12 (14.0)	4 (8.3)	8 (21.1)	0.091
Pseudoexfoliative glaucoma	3 (3.5)	2 (4.2)	1 (2.6)	0.690
IOP (mmHg), mean±SD	34.5±8.3	35.0±8.6	34.0±8.0	0.663±
Glaucoma medications, mean±SD (median)	3.8±0.7 (4.0)	3.9±0.3 (4.0)	3.5±0.9 (4.0)	0.002
Previous surgery
Deep sclerectomy	1 (1.2)	1 (2.1)	0 (0.0)	0.371
Phaco	34 (39.5)	19 (39.6)	15 (39.5)	0.992
Trab	12 (14.0)	7 (14.6)	5 (13.2)	0.850
ECCE	5 (5.8)	5 (10.4)	0 (0.0)	0.040*
ECP	1 (1.2)	1 (2.1)	0 (0.0)	0.371
CPC	5 (5.8)	3 (6.3)	2 (5.3)	0.846
PPV	5 (5.8)	3 (6.3)	2 (5.3)	0.846
Cornea graft	9 (10.5)	3 (6.3)	6 (15.8)	0.151
Previous laser	12 (14.0)	8 (16.7)	4 (10.5)	0.414
Previsual acuity
Median (Snellen)	20/200	20/200	20/200

IQR (Q1-Q3) *Statistically significant at 5% level of significance, ±Mann-Whitney U-test. Values are n (%) unless otherwise indicated and compared using Chi-squared test. IOP: Intraocular pressure, SD: Standard deviation, IQR: Interquartile range, CPC: Cyclophotocoagulation, ECCE: Extracapsular cataract extraction, ECP: Endocyclophotocoagulation, PPV: Pars plana vitrectomy

Treatment outcomes

The cumulative probability of failure was 33.3% in the Ahmed group and 44.7% in the Baerveldt group (P = 0.283). When using an alternate IOP criterion of 21 mmHg, the cumulative probability of failure was 14.6% in the Ahmed group and 15.8% in the Baerveldt group (P = 0.878). When using an IOP criterion of 14 mmHg, the cumulative probability of failure was 37.5% in the Ahmed group and 42.1% in the Baerveldt group (P = 0.666). The most common cause of failure in both the groups was high IOP [Table 2 and Figure 2].

Table 2

Reasons for failure during 3 years of follow-up

	Ahmed-FP7 (n=48 eyes)	Baerveldt-350 (n=38 eyes)
Success	32 (66.7)	21 (55.3)
Complete	8 (16.7)	4 (10.5)
Failurey	16 (33.3)	17 (44.7)
High IOP (>18 mmHg for consecutive visits)	12	17
Requiring additional glaucoma surgery	3	3
Progression to NLP	1	2
Devastating complications	1	3

IOP: Intraocular pressure, NLP: No light perception

Figure 2

Kaplan–Meier analysis using the primary outcome criteria

Intraocular pressure

Postoperative data of 36 months were collected from both the groups [Figure 3], and the mean IOP reduction rate at 3 years postoperatively was 60% and 53% for the AGV and BGI groups, respectively (P = 0.536). The mean IOP in the Ahmed group decreased from 35.0 ± 8.6 mmHg preoperatively to 14.0 ± 4.8 mmHg at 3 years postoperatively. The mean IOP in the Baerveldt group decreased from 33.9 ± 8.0 mmHg preoperatively to 15.8 ± 6.2 mmHg at 3 years postoperatively. The rate of complete success was low in both the groups, with only 16.7% in the AGV group and 10.5% in the BGI group meeting the criteria.

Figure 3

Mean intraocular pressure during the 3 years after surgery. Error bars represent standard deviation. Corresponds to a statistically significant difference between groups

Glaucoma medications use

The mean glaucoma medication use in the Ahmed group decreased from 3.9 ± 0.3 preoperatively to 1.6 ± 1.3 at 3 years postoperatively (59% reduction, P < 0.001). The mean glaucoma medication use in the Baerveldt group decreased from 3.5 ± 0.9 preoperatively to 2.1 ± 1.7 at 3 years postoperatively (40% reduction, P < 0.001), as shown in Figure 4.

Figure 4

Mean number of glaucoma medications during the 3 years after surgery

Visual outcomes

Both the groups had a moderate but similar reduction in visual acuity over 3 years of follow-up (P = 0.809). The mean logMAR visual acuity in the Ahmed group decreased from 1.1 ± 0.6 at baseline to 1.3 ± 0.9 at 3 years postoperatively (P = 0.144). The mean logMAR visual acuity in the Baerveldt group decreased from 1.0 ± 0.7 at baseline to 1.3 ± 1.0 at 5 years postoperatively (P = 0.007), as illustrated in Figure 5.

Figure 5

Distribution of change in visual acuity from baseline to 3 years

De novo glaucoma procedures

Subsequent glaucoma surgery was required in two patients of each group. All of them underwent cyclodestructive procedure (4% in AGV and 5% in BGI).

Postoperative complications

The rate of postoperative complications was similar in both the groups [Table 3]; the most frequent complications in the AGV group were hypotony maculopathy and hyphema (10% each) and in the BGI group were hypotony maculopathy and vitreous hemorrhage (13% each).

Table 3

Postoperative complications during 3 years of follow-up

	Ahmed-FP7 (n=48 eyes), n (%)	Baerveldt-350 (n=38 eyes), n (%)	P
Tube occlusion	1 (2.1)	0	0.372
Suprachoroidal hemorrhage	0	0	-
Endophthalmitis	1 (2.1)	3 (7.9)	0.208
Cystoid macular edema	2 (4.2)	0	0.204
Shallow AC	0	3 (6.3)	0.117
Hypotony maculopathy	5 (10.4)	5 (13.2)	0.689
Tube cornea touch	2 (4.2)	0	0.204
Tube erosion	0	0	-
Encysted bleb	3 (7.9)	0	0.049
Retinal detachment	3 (6.3)	0	0.117
Vitreous hemorrhage	0	5 (13.2)	0.010
Hyphema	5 (10.4)	1 (2.6)	0.160
Total complications	25	17	0.985

AC: Anterior chamber

Postoperative intervention

The most common interventions were cataract surgery (12.5% Ahmed and 8% Baerveldt, P = 0.49), cyclophotocoagulation (4% Ahmed and 5% Baerveldt, P = 0.81), and tube-related interventions (4% Ahmed and 8% Baerveldt, P = 0.46), as shown in Table 4.

Table 4

Postoperative interventions during 3 years of follow-up

Subsequent surgery	Ahmed-FP7 (n=48 eyes), n (%)	Baerveldt-350 (n=38 eyes), n (%)	P
AGV	0	0	-
BGI	0	0	-
CPC	2 (4.2)	2 (5.3)	0.810
Tube revision	2 (4.2)	3 (7.9)	0.463
Cataract	6 (12.5)	3 (7.9)	0.488
Tube removal	2 (4.2)	1 (2.6)	0.700
PKP	2 (4.2)	0	0.203

AGV: Ahmed glaucoma valve, BGI: Baerveldt glaucoma implant, CPC: Cyclophotocoagulation, PKP: Penetrating keratoplasty

Discussion

In our study, 86 retrospective, nonrandomized charts of glaucoma patients who underwent implantation of AGV-FP7 or a Baerveldt-350 were reviewed. Patients had uncontrolled IOP preoperatively despite maximum medical treatment, and several patients had previous failed laser treatment, deep sclerectomy, and trabeculectomy. Preoperative diagnosis was mostly due to secondary glaucoma, then primary open-angle glaucoma, and combined mechanism of glaucoma.

A review of the 3-year follow-up showed that the failure rate was 33% in the Ahmed group and 45% in the Baerveldt group (P = 0.283). The main reason of failure in both the groups was high IOP (more than 18 mmHg). The Ahmed failure rate was similar to previous studies, ranging between 33% and 51%, and the Baerveldt failure rate was higher, ranging between 15% and 36% in previous studies.[1314151617]

When using an alternate IOP criterion of 21 mmHg, the failure rate was 15% in the Ahmed group and 16% in the Baerveldt group (P = 0.878). In comparison to previous studies, our results showed less failure rate.[1819] When using an IOP criterion of 14 mmHg, the cumulative probability of failure was 37.5% in the Ahmed group and 42.1% in the Baerveldt group (P = 0.666). Our results are comparable to previous studies.[1118]

The overall treatment success rate was 63.1% in both the groups. Further subdivision of treatment success into complete and qualified was applied. Complete success required IOP to be in range at all visits after 3 months without the use of glaucoma medications and without any significant loss of vision or additional surgical procedures required. Although at 3-year follow-up, 66.7% of the Ahmed group and 55.3% of the Baerveldt group were considered successes, with only 16.7% of the patients in the Ahmed group and 10.5% of the patients in the Baerveldt group as complete successes (P = 0.413). Low complete success rates were comparable to previous studies and indicate complicated postoperative course, including IOP outside the range, the need for glaucoma medications, and complications requiring any surgical interventions. In addition, there is poor prognosis of our population with advanced glaucoma.[182021]

Although both the groups showed a high rate of failure, both devices were effective in lowering IOP and the need for medications. At 3 years after surgery, the Ahmed group had a mean IOP of 14.0 ± 4.8 mmHg (60% reduction from baseline) compared with 15.8 ± 6.2 mmHg (53.3% reduction) in the Baerveldt group (0.536). The Ahmed group required an average of 1.6 ± 1.3 medications (59% reduction) compared with 2.1 ± 1.7 medications (40% reduction) in the Baerveldt group (P < 0.001). In this study, we reported lower IOP and medications needed in the Ahmed group. Opposite to our study, after 3 years of follow-up, failure had occurred in 51% and 39% of the patients in the Ahmed and Baerveldt groups, respectively, and the Baerveldt group showed lower IOP and less medications needed.[18] In a recent prospect randomized study by Moschos et al., 552 patients with refractory glaucoma were recruited to compare between Ahmed-FP7 and Baerveldt-350 implants. They noted similar results to our study, with the mean IOP found to be significantly lower in the Ahmed group at 1- and 5-year follow-up. The IOP was postoperatively controlled using a lower number of medications in both the groups, although this number was significantly lower in patients with an Ahmed implant at the 3-year follow-up.[22] Another retrospective study performed by Tsai et al. noted a comparable result to our study, wherein during a follow-up of 12 months, the success rate (defined as 6 mmHg ≤ IOP ≤ 21 mmHg without additional glaucoma surgery or devastating complication) in the Ahmed group (48 patients) was higher than in the Baerveldt one (70 patients). In addition, lower IOP was found in patients with the Ahmed implant at 1 day and 1 week postoperatively, and these patients required fewer medications at 1 week and 1 month (P < 0.001) after surgery.[14]

The total rate of postoperative complications was similar in both the groups. Transient postoperative hypotony was observed in both the groups, with 10.4% of the Ahmed group and 13.2% of the Baerveldt group having choroidal effusions. Our results were similar to the Ahmed versus Baerveldt study, in which, after the 3-year follow-up, they noticed 13% of the Ahmed group and 14% of the Baerveldt group experienced choroidal effusions.[18] It has been postulated that hypotony in Ahmed implant is the result of a defective valve, damage from priming, or excessive peritubular filtration at the anterior chamber insertion site.[2324] In Baerveldt implant, the absence of flow restriction system leads to severe hypotony if there is ligation malfunction or excessive filtration through fenestration.[2324] In this study, none of the patients in both the groups experienced vision-threatening complications secondary to hypotony.

Most complications occurred in the early postoperative period. Similar to previous retrospective studies, our study showed that Ahmed implant showed a higher rate of encapsulation in comparison to Baerveldt implant (8% vs. 0%, P = 0.04).[1418] Several studies proposed that differences in early flow rates and exposure to inflammatory mediators, as well as differences in end plate material and topography, are the causes of encapsulation.[14252627] Others suggested that the cause of encapsulation is early exposure of the Ahmed bleb to mechanical compression from aqueous flow and exposure to surgery-induced inflammatory cytokines.[2526] In addition, electron microscopy comparing the Ahmed-FP7 end plate with the Baerveldt-350 implant found that it had a root-mean-square roughness 10-fold more, causing increased in vitro tenon fibroblast adhesion.[27] Antimetabolite use showed a lower rate of encapsulation, but failed to lower IOP, and showed a higher rate of hypotony and graft melt.[192528]

Five patients underwent tube revision (4% in AGV and 8% in BGI). De novo glaucoma procedures were needed in patients with uncontrolled IOP despite medications or tube revision. Ciliary body destructive procedure was performed in four patients who needed de novo glaucoma procedure (4% in AGV and 5% in BGI).

In both the groups, there was a moderate but similar reduction in visual acuity over 3 years of follow-up. However, as reported previously, it is difficult to know whether the definitive cause is the vision loss of glaucomatous progression, complications of the surgery, or associated ocular pathology. One patient (2%) in the Ahmed group and two patients (5%) in the Baerveldt group progressed to NLP, the majority of whom had neovascular glaucoma.

Several limitations were encountered in this study. Besides the fact that we had the difficulties of a retrospective study, different physicians with different experiences may contribute to different outcomes. Most of our recruited patients had a high risk of surgical failure because they were diagnosed with advanced glaucoma that was refractory to the maximum tolerated medical therapy. The outcomes of this study concern only Ahmed-FP7 model and Baerveldt-350 model and cannot be applied to different models. Furthermore, these outcomes cannot be applied if other concurrent surgeries are needed.

Conclusion

This study found that the overall treatment success rate was 63.1% in both the groups. Despite a high rate of failure, both devices were effective in lowering IOP and the need for medications. At the 3-year follow-up, the Ahmed group had a mean IOP of 14.0 ± 4.8 mmHg (60% reduction) compared with 15.8 ± 6.2 mmHg (53.3% reduction) in the Baerveldt group (0.536). The Ahmed group required an average of 1.6 ± 1.3 medications (59% reduction) compared with 2.1 ± 1.7 (40% reduction) in the Baerveldt group (P < 0.001).

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.