Glaucoma Drainage Implant Surgery – An Evidence-Based Update with Relevance to Sub-Saharan Africa

Abstract

Glaucoma represents a leading cause of preventable vision loss in Sub-Saharan Africa. Recent studies evaluating outcomes of glaucoma drainage implant (GDI) surgery suggest an important role for this approach in the African patient population. The Tube Versus Trabeculectomy study demonstrated a higher success rate with non-valved GDI surgery compared to trabeculectomy with mitomycin C after five years. The Ahmed Baerveldt Comparison study showed no difference in surgical failure rates between the Ahmed Glaucoma Valve and the Baerveldt Glaucoma Implant (BGI) but better intraocular pressure outcomes with the BGI at one year. The Ahmed Versus Baerveldt study demonstrated a lower failure rate for the BGI, but also a requirement for more post-operative interventions. Further study of GDI surgery in the Sub-Saharan Africa is necessary to determine its optimal place in the treatment paradigm for glaucoma patients in the region.

INTRODUCTION

Glaucoma is the second leading cause of blindness worldwide, and the Sub-Saharan African population shares a disproportionate amount of this disease burden.¹ Intraocular pressure (IOP) lowering is the only proven method to prevent the development² and/or slow the progression of glaucomatous optic neuropathy.³ The use of glaucoma drainage implants (GDIs) has increased considerably in recent years.⁴,⁵ This shift is likely related to a growing number of recently published studies documenting the efficacy of GDIs earlier in the glaucoma treatment paradigm.⁶ This article aims to provide insight into current evidence-based approaches to GDI surgery with specific relevance to the Sub-Saharan African population.

GLAUCOMA IN SUB-SAHARAN AFRICA: SCOPE OF THE PROBLEM

In a recent population-based, cross-sectional study, Budenz and colleagues⁷ reported on the prevalence and causes of blindness and visual impairment among adults in Tema, Ghana. In that study, glaucoma accounted for 21.9% of all cases of non-refractive blindness (defined as a best-corrected visual acuity < 20/400 in the better-seeing eye or visual field constriction to < 10% from fixation). In another study, Entekume et al.⁸ reported on the prevalence and causes for functional low vision in Nigeria. In that study, glaucoma was the most common cause of functional low vision, accounting for 26.5% of all cases. Preventable vision loss due to glaucoma in Sub-Saharan Africa is indeed an important clinical problem, and further study is necessary to elucidate the optimal treatment paradigm in this population.

HISTORY OF GLAUCOMA DRAINAGE IMPLANT SURGERY

GDIs are intended to create a pathway for aqueous humor outflow from the anterior chamber and into a potential space within an overlying fibrous capsule that develops postoperatively. The first attempt at GDI surgery, in which a horsehair was implanted into the anterior chamber, was performed by Rollet and Moreau in 1906.⁹ In 1969, Molteno described the use of a silicone tube attached to a polyethylene plate, which acted as a shunt reservoir and was fixated to the equatorial sclera. This approach was reported to decrease the risk of subsequent fibrosis.¹⁰ Since Molteno's introduction, the field of GDI surgery has evolved to include several types of implants which differ in surface area, shape, and presence or absence of a flow-restricting valve.¹¹ Non-valved devices include the Baerveldt Glaucoma Implant (BGI; Abbott Medical Optics, Abbott Park, IL, USA) and the Molteno3 implant (IOP Ophthalmics, Costa Mesa, CA, USA). Valved devices include the Ahmed Glaucoma Valve (AGV; New World Medical, Rancho Cucamonga, CA, USA) and the Krupin Valve Implant. Over the past decade, three important clinical trials were undertaken in order to evaluate the role of GDI devices in glaucoma management.

THE TUBE VERSUS TRABECULECTOMY (TVT) STUDY

In order to compare the safety and efficacy of GDI surgery with standard trabeculectomy with mitomycin C, Gedde and colleagues performed a multicenter randomized clinical trial comparing the outcomes of each procedure.¹² The TVT study included 17 clinical centers and a total of 212 patients with uncontrolled glaucoma and an IOP ≥ 18 mm Hg and ≤ 40 mm Hg on tolerated medical therapy. Baseline IOP was 25.1 ± 5.3 mm Hg in the GDI group and 25.6 ± 5.3 mm Hg in the trabeculectomy group (P = 0.56). Patients enrolled in the study had undergone previous trabeculectomy, cataract extraction with intraocular lens implantation, or both. Each patient was randomized to GDI surgery, using the 350-mm² BGI, or to standard trabeculectomy with mitomycin C (0.4 mg/mL applied for 4 minutes). The primary study outcome was post-operative IOP. Secondary outcomes included rates of surgical failure (defined as IOP > 21 mm Hg or not reduced by 20%, IOP < 5 mm Hg, need for additional glaucoma surgery, or loss of light perception vision), need for supplemental glaucoma medications, post-operative visual acuity, and complication rates.

After five years of follow-up in the TVT study,¹³ IOP outcomes between the GDI and trabeculectomy study groups were found to be similar (14.4 ± 6.9 mm Hg versus 12.6 ± 5.9, respectively; P = 0.12) and represented a 41.4% and 49.5% IOP reduction from baseline (P = 0.097), respectively. The proportion of patients who experienced a post-operative IOP < 14 mm Hg was also similar in each of the study groups (63.9% in the GDI group versus 63.5% in the trabeculectomy group; P > 0.99). Although patients randomized to trabeculectomy required less supplemental medications through the first two years of follow-up, this difference did not persist at five years (1.4 ± 1.3 medicines required in the GDI group versus 1.2 ±1.5 in the trabeculectomy group; P = 0.23). Patients randomized to trabeculectomy were more likely to experience surgical failure (29.8% failure rate in the GDI group versus 46.9% in the trabeculectomy group; P = 0.002) and were more likely to require a re-operation for glaucoma (9% in the GDI group versus 29% in the trabeculectomy group; P = 0.025) during five years of follow-up. The proportion of patients who lost > two lines of Snellen visual acuity was similar in the GDI and trabeculectomy groups at five years (46% versus 43%, respectively; P = 0.93), and the most frequent cause of loss of vision was progression of glaucomatous disease (11% and 13%, respectively).

Although early post-operative complications were more frequent in TVT patients randomized to trabeculectomy (37% versus 21% in the GDI group; P = 0.012), the rate of late post-operative complications was similar (36% in the trabeculectomy group versus 34% in the GDI group; P = 0.81) at five years.¹⁴ The rate of serious complications (defined as events producing a loss of ≥ 2 lines of Snellen visual acuity and/or requiring reoperation) was similar in each treatment group (20% in the trabeculectomy group versus 22% in the GDI group; P = 0.79).

In sum, the TVT study demonstrated that both trabeculectomy and GDI surgery were capable of producing sustained IOP reduction after five years of follow-up. Although GDI surgery was associated with the use of more supplemental glaucoma medications during the first two years, this difference did not persist at five years. Patients who were treated with trabeculectomy experienced a higher surgical failure rate and were more likely to require re-operation for glaucoma compared with patients treated with GDI surgery. Vision loss occurred at a similar rate in the two groups, and although the rate of early post-operative complications was greater in the trabeculectomy group, there were similar rates of late and serious post-operative complications in the two groups.⁶

THE AHMED BAERVELDT COMPARISON (ABC) STUDY

The ABC study¹⁵ is an ongoing multicenter, randomized controlled clinical trial comparing long-term outcomes and complications of two GDIs: The AGV and the BGI. This is the first prospective trial to compare two different GDI types. In this trial, 276 glaucoma patients with an IOP ≥ 18 mm Hg (mean IOP 31.2 ± 11.2 mm Hg in the AGV group versus 31.8 ± 12.8 mm Hg in the BGI group; P = 0.071) at 16 clinical centers worldwide were randomized to surgical treatment with one of the aforementioned devices. The primary outcome was the rate of surgical failure, which was defined as IOP > 21 mm Hg or not reduced ≤ 20%, IOP ≤ 5 mm Hg, need for additional glaucoma surgery and/or removal of the implant, or loss of light perception vision. Secondary outcomes included post-operative IOP, visual acuity, and complication rates.

At one year, surgical failure rates between the AGV and BGI groups were similar (16.4% versus 14.0%, respectively; P = 0.52).¹⁶ However, a lower post-operative IOP was achieved in patients randomized to the BGI (IOP was reduced to 15.4 ± 5.5 mm Hg in the AGV group versus 13.2 ± 6.8 mm Hgin the BGI group; P = 0.007). There was a decrease of ≥ 2 lines of Snellen visual acuity in 30% of patients in the AGV group and 34% of patients in the BGI group (P = 0.57). The overall intra-operative complication rate was 10% and similar between treatment groups (P = 0.31); intra-operative hyphema accounted for 79% of these complications. However, a greater proportion of patients randomized to the BGI experienced early post-operative complications (58% in the BGI group versus 43% in the AGV group; P = 0.016), and this was attributed to a greater number of tube occlusions and episodes of corneal edema in patients randomized to BGI.

Recent three year results¹⁷ of the ABC study also demonstrated a similar risk of surgical failure in the AGV and BGI groups (P = 0.88). However, patients randomized to the BGI achieved a lower post-operative IOP (12.9 ± 4.4) compared to patients randomized to AGV (14.3 ± 4.9; P = 0.049). Results of the ABC study suggest that although GDI surgery with the BGI may lead to better IOP outcomes, there is an increased risk of post-operative complications. The ABC study is ongoing and is designed to continue follow-up of participants to five years.

THE AHMED VERSUS BAERVELDT (AVB) STUDY

In another study comparing the AGV and BGI devices,¹⁸ Christakis and colleagues performed a randomized clinical trial including 238 glaucoma patients at seven clinical sites. Eligible patients with inadequate baseline IOP (defined as an IOP greater than the clinical target, which had not responded to conventional medicinal, laser, or surgical therapy; mean IOP 31.1 ± 10.5 mm Hg in the AGV group versus 31.7 ± 11.1 mm Hg in the BGI group; P = 0.71) were randomized to surgical therapy with either of the two devices. The primary outcome was the rate of post-operative surgical failure, which was defined as an IOP > 18 mm Hg, < 5 mm Hg, < 20% IOP reduction, vision-threatening complications, need for additional glaucoma procedures, or loss of light perception vision. Secondary outcomes included post-operative IOP, visual acuity, and complication rates.

At one year, a significantly lower proportion of patients randomized to BGI experienced surgical failure compared to patients randomized to AGV (28% versus 43%, respectively; P = 0.02).¹⁹ A lower post-operative IOP was achieved in patients randomized to BGI versus the AGV (16.5 ± 5.3 mm Hg versus 13.6 ± 4.8 mm Hg, respectively; P < 0.001). Visual acuity outcomes were similar in the two treatment groups, with no difference in mean logMAR acuity change from baseline between the two devices at one year (P = 0.66). There was no difference in the overall rate of post-operative complications in each treatment group at one year (44% in the AGV group versus 54% in the BGI group; P = 0.19). However, a greater proportion of patients randomized to BGI required post-operative interventions (26% in the AGV group versus 42% in the BGI group; P = 0.009).

Recent two-year results of the AVB study²⁰ continued to demonstrate a higher surgical failure rate in the AGV group compared to the BGI group (48% versus 31%, respectively; P = 0.02). By two years, mean IOP was no longer significantly lower in the BGI group (14.5 ± 6.6 mm Hg in the BGI group versus 16.1 ± 6.4 mm Hg in the AGV group; P = 0.12). There was a continued increased likelihood of post-operative interventions in patients randomized to BGI versus AGV at two years (47% versus 32%; P = 0.02). Results of the AVB study demonstrate a lower post-operative failure rate in patients undergoing GDI surgery with the BGI, but with an increased need for post-operative interventions. The AVB study is ongoing and is designed to continue follow-up of participants to five years.

GLAUCOMA DRAINAGE IMPLANT SURGERY IN SUB-SAHARAN AFRICA

Prior studies have suggested that medical therapy may be less effective than primary surgical therapy in the Sub-Saharan African population.²¹,²² Furthermore, there is an increased risk of trabeculectomy failure in eyes of African individuals compared to Caucasian eyes.²¹,^23–25 Kim et al.²⁶ studied outcomes of primary trabeculectomy with mitomycin C in 30 Black West African individuals and reported a mean post-operative IOP of 18.3 mm Hg after a mean 6.5 years of follow-up, with only 27.6% of eyes experiencing a post-operative IOP < 15 mm Hg. There have been no studies to suggest decreased efficacy of GDI surgery in an African compared to a Caucasian population, but data are limited. Forty-five percent of subjects in the TVT study¹² were White and 39% were Black; racial status did not predict risk of surgical failure and treatment outcome in both univariate (P = 0.64) and multivariate (P = 0.67) risk factor analyses.¹³ Extrapolation of the TVT results suggests that GDI surgery would be worth considering as an effective option for long-term IOP control and preservation of visual field in Sub-Saharan African glaucoma patients.

Neither the ABC¹⁶ nor the AVB¹⁹ studies demonstrated an increased risk of GDI surgical failure based on racial status. Forty-nine percent of patients enrolled in the ABC study were White and 25% were Black. The relative risk for GDI surgical failure among Black subjects was 0.9 (95% confidence interval = 0.5-1.9). In the AVB study, 71% of subjects were White and 12% were Black. In a univariate risk factor analysis, racial status did not predict risk of GDI surgical failure (P = 0.71). These results suggest that the efficacy of GDI surgery may not differ between these study populations and the Sub-Saharan African population.

Kiage et al.²⁷ performed a retrospective case series investigating outcomes of GDI surgery with the AGV in 25 East African patients with glaucoma. Six of these patients had undergone prior trabeculectomy and six had undergone prior angle surgery (trabeculotomy or goniotomy). Mean pre-operative IOP in this patient group was 36.4 mm Hg and was reduced to 16.7 mm Hg, for a mean percentage IOP-lowering of 53.2% after a median follow-up period of two months. Results of this study support the use of GDI surgery in the Sub-Saharan African population. Encouraging results with GDI surgery in an Ethiopian population with refractory glaucoma were recently reported by Giorgis.²⁸ In this retrospective analysis of 13 eyes of 12 Ethiopian patients that underwent AGV implantation, mean IOP was reduced from a pre-operative level of 31.38 ± 8.67 mm Hg to post-operative levels of 15.52 ± 8.80 mm Hg at one week, 14.77 ± 6.39 mm Hg at one month, 16.62 ± 4.48 mm Hg at three months, and 17.15 ± 3.87 mm Hg at six months of follow-up. Complications were encountered in 46% of eyes and included, hypotony, tube-corneal touch, acceleration of cataract, choroidal effusion, flat anterior chamber, reduction in visual acuity, and tube exposure. Interestingly, subjects in this series did not tend to experience a “hypertensive phase” previously associated with AGV implantation²⁹ and postulated to be secondary to gradual, transient congestion and edema of the fibrous capsule surrounding the implant plate. Future studies prospectively documenting the occurrence and timing of the hypertensive phase will prove valuable to evaluation of the expected time course of IOP reduction and required post-operative care in glaucomatous patients undergoing GDI surgery.

As mentioned, there is an increased risk of trabeculectomy failure in African compared to Caucasian eyes.²¹,^23–25 This raises the issue of primary GDI surgery for long-term IOP control in African eyes afflicted with glaucoma. GDI surgery is typically associated with less need for post-operative follow-up compared with trabeculectomy, and the TVT study revealed a decreased requirement for re-operation among eyes treated with GDI compared to eyes treated with trabeculectomy.¹⁴ These advantages are of particular importance in Sub-Saharan Africa, where problems with poor compliance are well documented.²¹,²² Further, AGV rather than BGI surgery may be preferable in this population as both the ABC¹⁶ and AVB¹⁹ trials demonstrated a higher rate of early post-operative complications and greater likelihood of need for post-operative interventions, respectively, in patients randomized to BGI. However, if post-operative compliance with office visits is judged to be satisfactory, BGI surgery may be a better choice as the ABC and AVB studies both demonstrate a lower risk of surgical failure.

As studies on GDI are planned in Sub-Saharan Africa, cost-effectiveness should also be considered.

CONCLUSIONS

Glaucoma represents a leading cause of blindness in Sub-Saharan Africa. Glaucoma management with medical therapy and/or trabeculectomy may not be optimal due to a strict need for compliance and post-operative follow-up. Several advances in the field of GDI surgery have led to an improved safety and efficacy profile, as demonstrated in three recent randomized controlled trials. GDI surgery offers a potentially safe and effective option for IOP control and the long-term preservation of vision in the Sub-Saharan African population; however, further research is needed before widespread use can be recommended.