Glaucoma drainage implants (also known as tube shunts, aqueous shunts, and glaucoma drainage devices) have been increasingly utilized in the surgical management of glaucoma. Recent Medicare claims data and surveys of the American Glaucoma Society membership have demonstrated a decrease in the popularity of trabeculectomy and a concurrent increase in glaucoma drainage implant surgery. These devices share a common design consisting of a silicone tube that is inserted into the eye through a scleral fistula and shunts aqueous humor to an end plate located in the equatorial region of the globe. Fibrous encapsulation of the end plate produces a reservoir into which aqueous pools. The major resistance to aqueous flow through drainage implants occurs across the fibrous capsule around the plate. Therefore, the final intraocular pressure (IOP) that is achieved following glaucoma drainage implant surgery is determined by capsule thickness and surface area. In particular, thinner and larger surface area capsules are associated with lower postoperative pressures.

Commercially available glaucoma drainage implants differ with respect to the size, shape, and material of the end plate. They may be further subdivided into valved and nonvalved implants, depending on whether a valve mechanism is present that limits aqueous flow if the IOP becomes too low. Nonvalved implants require a temporary restriction of flow by tube ligation or occlusion at the time of surgical implantation. This maneuver allows a capsule to develop around the end plate before the device is functional, and minimizes the risk of hypotony in the early postoperative period. Valved implants have flow restrictors that are designed to prevent hypotony, even prior to encapsulation of the plate. Valved implants do not require temporary flow restriction and offer the advantage over nonvalved implants of providing immediate IOP reduction postoperatively. The end plates of valved implants receive an earlier exposure to aqueous humor after surgery with a higher concentration of inflammatory mediators, and this has been postulated to induce a higher rate of bleb encapsulation and poorer long-term IOP control. Implants with larger end plates have been shown to produce greater pressure reduction, presumably by promoting a larger surface area capsule around the end plate for filtration. However, there appears to be an upper limit beyond which an increase in end plate size does not allow greater efficacy.

Glaucoma surgeons have questioned whether one drainage implant is superior to another. Several retrospective studies have compared different implant types, but they were limited by biases that are inherent to retrospective studies. A lesser number of randomized prospective trials have compared implants of the same type. Nassiri and associates offer the first randomized prospective study comparing drainage implants of different types. Patients with refractory glaucoma were randomly assigned to treatment with an Ahmed glaucoma valve implant or single-plate Molteno implant, and results were reported during the first 2 years of follow-up in this single-surgeon study.

No significant differences in surgical success rates were observed between the Ahmed group and the Molteno group using Kaplan-Meier survival analysis. Failure was defined as IOP >21 mm Hg or <6 mm Hg on 2 consecutive visits, phthisis bulbi, loss of light perception vision, removal of the implant, reoperation for glaucoma, or any devastating intraoperative or postoperative complication. The success and failure criteria used in this study are similar to those in other glaucoma surgical studies, thereby facilitating comparison with prior publications. However, the results of several recent multicenter randomized clinical trials have suggested that IOP of 21 mm Hg or less may be inadequate to prevent glaucomatous progression in many patients. This is especially true in patients with severe glaucoma, such as those enrolled in this study with an average mean deviation on Humphrey visual field testing of approximately −19 decibels.

The Ahmed group had significantly greater IOP reduction in the early postoperative period than the Molteno group, and this is expected given the design differences between the 2 implants (ie, valved vs nonvalved). However, mean IOP was reported as significantly lower in the Molteno group compared with the Ahmed group in the later postoperative period. It is noteworthy that the authors censored patients from analysis at the time of failure, and all patients in this study failed because of inadequate IOP reduction. This is analogous to a school reporting the average grade point average of its student body, but excluding all children with a D average or less. It would have been more informative to the reader to provide the mean pressure of all patients, rather than only those who had a successful outcome. The tendency toward greater use of supplemental medical therapy in the Molteno group likely contributed to the greater pressure reduction (among successful patients) in this group relative to the Ahmed group.

The approach to adjunctive medical therapy is different between the 2 implants under investigation, and this should be considered when interpreting the study results. Valved implants like the Ahmed provide immediate pressure lowering, and medical therapy is generally added in the later postoperative period as needed. The Molteno implant and other nonvalved implants are not functioning initially because of the temporary restriction of flow. Therefore, use of glaucoma medications is expected in the early postoperative period, and they must be purposefully weaned after the tube opens. These differences in the use of supplemental medical therapy may explain the higher mean number of medications and lower mean IOP in the Molteno group compared with the Ahmed group.

A hypertensive phase was observed in 17.4% of patients in the Molteno group and 28.3% of patients in the Ahmed group. The hypertensive phase that is frequently seen following glaucoma drainage implant surgery is presumed secondary to bleb encapsulation. A hypertensive phase was defined in this study as IOP >21 mm Hg during the first 3 postoperative months (with or without medications) after reduction of IOP <22 mm Hg during the first postoperative week and not caused by tube obstruction, retraction, or valve malfunction. While this definition seems reasonable for valved implants and has been applied in a prior study of the Ahmed implant, it may not be appropriate for nonvalved implants. Tube fenestrations were performed in all patients in the Molteno group, and this has been shown to be an effective method for immediate IOP reduction postoperatively. However, fenestrations commonly fail early in the postoperative period, even prior to the lysis of the vicryl ligature around the tube (the method of temporary tube occlusion in this study). Some of the cases of fenestration failure in the Molteno group were likely misclassified as a hypertensive phase, and the author’s definition was met even prior to the formation of a bleb over the end plate (and, therefore, not related to bleb encapsulation).

The rate of postoperative complications was comparable between the Molteno group and the Ahmed group, and no devastating complications were observed in either treatment group. All complications were transient and self-limited, resolving with medical or minor surgical interventions (ie, anterior chamber reformation, suturing of a wound dehiscence). These data support a favorable safety profile for both implants. However, additional late complications associated with drainage implants may occur with longer follow-up (ie, tube erosion, corneal edema).

The mean visual field (VF) score did not significantly change in the Ahmed group during the course of the study, but the Molteno group experienced a 1-decibel loss from baseline. This VF deterioriation occurred despite a lower mean IOP in the Molteno group compared with the Ahmed group, and the authors are left trying to explain this counterintuitive finding. However, only a single VF was used as a baseline, and it is well known that VF results fluctuate. The fact that 10% of patients in the Ahmed group and 18% of patients in the Molteno group were excluded from analysis because of a “learning effect” with VF testing underscores the need to establish a good baseline to provide meaningful VF data.

Nassiri and associates are to be congratulated for delivering an important comparative prospective study. Despite the above limitations and a relatively small study population, the authors have provided valuable information about the outcomes of glaucoma drainage implant surgery. Additional studies comparing different implants are needed, and randomized clinical trials provide the highest level of evidence based medicine. The Ahmed Baerveldt Comparison (ABC) Study is another randomized clinical trial that compares 2 different drainage implants, and it promises to also provide important data to guide surgical decision making.