Implant technology improving, becoming ‘primary surgical option’ in some conditions

Surgeon shares his techniques for avoiding and correcting complications with glaucoma drainage implants.

ByPaul Palmberg, MD, PhD

Paul Palmberg, MD, PhD [photo]
Paul Palmberg

Glaucoma drainage implants are assuming a steadily expanding role in the glaucoma surgeon’s armamentarium. The devices work by shunting aqueous from the anterior chamber through a long tube to an equatorial plate that maintains a space for fluid absorption.

More and more, we are turning to glaucoma drainage implants (GDI) after a single failed trabeculectomy, rather than performing a second, and in certain conditions, such as neovascular glaucoma, iridocorneal endothelial syndrome, penetrating keratoplasty with glaucoma and glaucoma following retinal detachment surgery, GDIs are coming to be the primary surgical option.

As with all surgical procedures, the implantation of GDIs is not without risk. However, after 2 decades in the evolution of devices and techniques, we are now able to markedly reduce the risk of complications. This article illustrates solutions we have found useful in avoiding and fixing problems in GDI surgery and in postoperative management with the Advanced Medical Optics Baerveldt glaucoma implant.

Avoiding diplopia

Most cases of diplopia with GDIs are caused by the formation of a high bleb that displaces the globe or interferes with ocular rotations, rather than by the formation of restrictive adhesions to the recti muscles. This can be demonstrated by simply decompressing the bleb cavity, which will temporarily eliminate the motility disturbance. A simple and safe way to indirectly decompress the bleb cavity is to perform an anterior chamber paracentesis through the cornea with a 30-gauge needle in eyes with nonvalved GDIs.

Fortunately, motility problems are now rare for the Baerveldt glaucoma implant since the manufacturer modified the device by placing four fenestrations in the reservoir plate that allow the formation of “rivets” of tissue between the scleral surface and the bleb wall. With a resultant lower profile of the bleb cavity, globe displacement is avoided.

A technical advance came from the realization that one could reduce the chance of injury to muscles and of causing bleeding in the region of the future bleb cavity by simply clearing space for the device in the superior temporal (or inferior nasal) quadrant. This is accomplished by entering the sub-Tenon’s space in that quadrant by blunt dissection and then passing two muscle hooks into the quadrant and separating them until they reach the adjacent recti muscles. Then one muscle hook is placed behind the insertion of the lateral rectus and the second behind it, and run posteriorly, stripping any check ligaments. One then repeats the process for the superior rectus. That is all the cleaning of muscles that is required, as one need not separate Tenon’s capsule from the anterior surface of the muscles. Care is also taken to be sure that the device is placed with each wing of the plate at least 1 mm behind the muscle insertion, so it does not alter the direction of the muscle fibers.

In those rare cases when diplopia occurs from a bleb that is too large, the tube may be ligated, allowing the bleb to collapse. A second device may then be placed in another quadrant for pressure control.

Avoiding corneal problems

Corneal endothelial injury leading to edema may occur when the tube entry is anterior to the trabecular meshwork, allowing the destruction of endothelial cells at the site of tube entry by the slight movement of the tube with blinks, followed by migration of cells from the adjacent cornea. This may occur more rapidly by contact of the tip of the tube to the corneal endothelium.

As further protection against the tendency of the tube to gradually straighten over time and migrate through the limbal tissue, eventually allowing the tube tip to touch the corneal endothelium, I recommend the technique of Dr. Oscar Albis, of Barranquilla, Columbia, which employs a longer entry path than the commonly used short puncture at the limbus. I make a 23-gauge needle track that first tunnels in sclera parallel to the surface for 2 to 3 mm to reach the limbus. The track then gradually turns to become parallel to and just above the iris surface for the actual entry into the anterior chamber. This provides sufficient resistance to prevent a gradual straightening of the tube.

When the anterior chamber is shallow, the tube may need to be placed through the peripheral iris into the posterior chamber and out into the pupil over an IOL to avoid allowing the tube to be slowly pulled forward by the formation of peripheral anterior synechiae. Alternatively, the tube may be passed through the pars plana into the vitreous cavity if a pars plana vitrectomy has already been performed or is also needed.

Avoiding exposure of the tube

The tube must be covered with a preserved patch of cornea or sclera to prevent erosion to the surface. An alternative is to pass the tube through a scleral tunnel of several millimeters before entering the anterior chamber.

When exposure does occur, the tube must be covered with corneal or scleral tissue, not just conjunctiva-Tenon’s, or re-erosion will occur.

Placement in phakic eyes

In phakic eyes I prefer to route the tube to enter at the 12 o’clock position and to angle the tube nasally in the far periphery of the chamber angle to avoid lens touch and to keep the tube out of the way for future clear corneal cataract extraction.

Tip obstruction

Tube tip obstruction by the iris is usually avoided by having a bevel-up tip for the tube. When it does occur, a 30-gauge needle may be used at the slit lamp to lift the tube above the iris surface, freeing the tip. In some cases of rubeosis, the tube tip may become wrapped by iris contraction. When this occurs, argon laser application can be applied through a goniomirror to obliterate the iris at the tube tip, freeing it.

Obstruction of the tube tip by a blood clot occurs in cases of rubeosis or in which the entry has to be through the iris. The risk of obstruction of the tube tip by blood can be reduced by placing a small bubble of Healon (sodium hyaluronate, Advanced Medical Optics) at the tube tip whenever considerable anterior chamber bleeding occurs intraoperatively. This keeps a clot from forming at the tube tip. When a clot obstructs the tip, 10 units of tissue plasminogen activator (tPA) may be injected into the anterior chamber to dissolve it.

Early pressure control

When fenestrations do not adequately control the IOP in the early postoperative period, the tube may be refenestrated at the slit lamp. This is easier to do when a clear half-thickness corneal patch has been used. At 2 or more weeks postop, the Vicryl ligature on the implant may be lasered at 0.5 W for a 0.5-second duration with a 50-µm spot.

Hypotony

Hypotony of marked degree is rare with a ligated nonvalved implant or valved implant. It may occur with the use of a large, nonvalved implant in cases with marked uveitis or after previous cyclodestruction procedures, in which aqueous production is reduced. When it occurs, it may be necessary to drain choroidal detachments and to temporarily re-ligate the tube, or to replace a nonvalved implant with a valved one.

Conclusion

The art of glaucoma treatment is just that, an art. Different cases require individual screening and treatment plans to assure the most efficacious outcomes. The Baerveldt glaucoma implant is a useful tool for managing IOP and can greatly benefit a wide spectrum of glaucoma patients. Careful management of pre-, intra- and postoperative procedures will minimize surgical complications and garner more positive patient results.

For Your Information:

Paul Palmberg, MD, PhD, is a professor of ophthalmology at the Bascom Palmer Eye Institute, University of Miami, Miller School of Medicine, Miami, Fla. He can be reached at the Bascom Palmer Eye Institute, Box 016880, Miami, FL 33101 U.S.A.; +1-305-326-6386; fax: +1-305-326-6474; e-mail: ppalmberg@med.miami.edu. Dr. Palmberg has no direct financial interest in the products mentioned in this article, nor is he a paid consultant for any companies mentioned.