New Frontiers in Glaucoma Surgery
We now have several strategies to compare with trabeculectomy.
BY HEATHER A. NESTI, M.D., HERBERTO ALVIM, M.D., AND L. JAY KATZ, M.D.

For the past 30 years, trabeculectomy has been the primary surgical intervention for the lowering of intraocular pressure (IOP) in patients with glaucoma. In recent years, this procedure has been augmented with the introduction of anti-metabolite agents such as mitomycin-C and 5-flurouracil. Unfortunately, filtration failure due to excessive fibrosis still occurs, and late onset bleb-related complications such as hypotony, bleb leaks, infections and endophthalmitis have become commonplace.

As a result, many new techniques, devices and agents have recently been introduced. Here we'd like to provide a brief overview of some of the most promising.

In an effort to limit postoperative subconjunctival fibrosis, researchers have developed a monoclonal antibody (CAT-152) for use with standard trabeculectomy procedures. CAT-152 neutralizes TGF-ß2, a naturally occurring fibrosis stimulant. This may minimize the development of thin, avascular blebs that are prone to leaking and infection. (The potential advantage over anti-metabolites is its apparent lack of toxicity to the conjunctival tissue.)

Nonpenetrating Surgery

Nonpenetrating glaucoma surgery, which avoids entry into the anterior chamber, was first described more than 40 years ago. Today, surgeons are performing and studying two nonpenetrating procedures -- deep sclerectomy and viscocanalostomy. (Each is described below.)

The allure of these procedures is the safety of not entering the anterior chamber, thereby avoiding intraoperative and postoperative hypotony. Theoretically, no bleb (or a thick-walled, more diffuse bleb) forms after these procedures, minimizing bleb-related complications. However, the limbal anatomy may be disorganized in re-operations, which can make it difficult to identify Schlemm's canal -- an integral part of the procedure.

These procedures are either performed alone or in combination with cataract surgery.

Nonpenetrating deep sclerectomy (NPDS). To perform NPDS, the surgeon creates two separate scleral flaps. The first flap, about one-third of the scleral thickness, is dissected to clear cornea; a second flap is made in the remaining scleral bed. The second flap is dissected anteriorly to unroof Schlemm's canal, removing corneal stroma to Descemet's membrane. This thin remaining layer, termed the trabeculo-Descemet membrane (TDM), becomes a portal through which the aqueous can leave the anterior chamber.

To complete the procedure, the inner and outer walls of Schlemm's canal are stripped to reduce resistance to outflow. The superficial scleral flap is then sutured.

Modifications in the procedure include injection of viscoelastic in the remaining scleral bed, the use of collagen implants underneath the scleral flap, and concomitant use of anti-fibrosis agents (as documented by Galassi et al, Acta Ophthalmol Scand 2002, 63-5; Karlen et al., Br J Ophthalmol 1999, 83:6-11; and Kozobolis et al., J Glaucoma 2002, 11:287-93).

The effectiveness of this procedure is well documented:

• Chiou et al. have used ultrasound biomicroscopy to demonstrate aqueous flow through the TDM beneath the scleral flap (Ophthalmology 1998;105:746-50).
• After 36 months, Wishart et al. noted that 77% of patients maintained an IOP lower than 21 mm Hg, with no medication (Acta Ophthalmol Scand 2003;81:343-8).
• Kozobolis et al. reported increased long-term IOP reduction when placing mitomycin-C on the sclera prior to flap construction (Glaucoma 2002, 11:287-93).
• Mermoud et al. demonstrated that NPDS with collagen implant had a similar success rate to trabeculectomy after a 2-year follow-up (J Cataract Refract Surg 1999; 25:323-31).
• Similar studies comparing trabeculectomy with NPDS by Chiselita (Eye 2001:15:197-201) and Demailley et al. (Int Ophthalmol 1997;20:131-40), revealed lower long-term IOP reduction with trabeculectomy. (No anti-metabolite agents were used intraoperatively in either group in these studies.)

NPDS does have potential drawbacks. The procedure is technically challenging, with a steep learning curve. Studies such as Mermoud's, and one by Sayyad et al. (Ophthalmology 2000;107:1671-4) have described perforating the TDM, necessitating conversion to conventional trabeculectomy. Other reported complications have included postoperative steroid responsive rise in IOP, large cystic blebs, hyphema, postoperative iris prolapse, hemorrhagic Descemet detachment, and choroidal detachments. However, according to both Mermoud and Sayyad, the incidence of complications when performing NPDS is lower than with conventional trabeculectomy.

If IOP increases, and the surgeon suspects insufficient aqueous flow through the TDM, the surgeon can perform goniopuncture. An Nd:YAG laser is used to create small openings in the TDM, penetrating into the anterior chamber to increase outflow and lower IOP. (Wishart et al. noted that patients with previous ALT required goniopuncture of the TDM postoperatively. Vuori also reported iris prolapse following goniopuncture [Acta Ophthalmol Scand 2003; 81:573-6].)

Viscocanalostomy. In this procedure, as in deep sclerectomy, the surgeon creates two scleral flaps. The second scleral flap is deeper, and is used to unroof Schlemm's canal. After the canal has been opened, the surgeon inserts a cannula into the canal and injects high-viscosity viscoelastic.

The largest case series (214 patients) was reported by Stegmann, who first introduced the procedure (J Cataract Refract Surg 1999;25:316-22.). Results included:

• 82.7% (166 patients) achieved an IOP of less than 22 mm Hg with no additional medical therapy.
• Average pressure reduction was 64%.
• Bleb formation was rare (less than 5%).

Five-year results reported by Shaarawy et al. showed that 60% of patients maintained an IOP of less than 21 mm Hg with no medication. This increased to 90% with the addition of medication (Br J Ophthalmol 2003; 87:441-5). Also:

• Goniopuncture was performed in 33.8% of patients, with a 39.5% immediate reduction in IOP.
• Six of 57 eyes were considered failures.

Jonescu-Cuypers et al. conducted a prospective, randomized trial comparing viscocanalostomy and trabeculectomy. They reported that complete success (without re-operation or medication) was obtained in 5 of 10 cases of trabeculectomy, but no cases of viscocanalostomy (Ophthalmology 2001;108:254-8). This result differs from that reported by Carassa et al., whose data showed that 86.2% of their patients (25 out of 29) had an IOP between 2 mm Hg and 21 mm Hg, with no medication, after 12 months (Eur J Ophthalmol 1998;8:57-61).

Complications that have been reported with viscocanalostomy include perforation of the TDM window, hyphema, choroidal detachment and a postoperative steroid responsive increase in IOP.

The Ex-PRESS Shunt

The Ex-PRESS Mini Glaucoma Shunt (CIBA Vision), which was approved by the FDA in 2000, is an unvalved device made of stainless steel (316 L). It has a 3-mm-long tube (2.5 mm of which is intraocular), an outer flange to prevent deep implantation, and a spur-like inner projection to prevent extrusion. Unlike other shunts (e.g., Ahmed and Baerveldt) the Ex-PRESS shunt does not have a posteriorly positioned reservoir. As a result, filtration with the Ex-PRESS tube leads to limbal blebs.

Studies have shown that the Ex-PRESS shunt effectively achieves good intraocular pressure at various follow-up times (for example, see Kaplan-Messias et al., Invest Ophthalmol Vis Sci 2001;42:S552). Studies have also shown that the device has acceptable biocompatibility and may produce long-term functioning blebs (Nyska et al., J Glau-coma 2003;12:275-280).

Because there is no valve, all the complications associated with profound hypotony may occur. Some authors (such as Kaplan-Messias) describe an incidence of shunt-related complications as high as 17%, with almost 4% of the cases undergoing re-operation (Invest Ophthalmol Vis Sci 2001; 42:S552.) One strategy that may help protect against these risks is to incorporate the Ex-PRESS shunt under a sutured scleral flap.

While use of the device may be warranted in advanced cases of glaucoma, a larger, randomized study is needed to evaluate its risks and benefits for patients who have less severe glaucoma.

Bypassing the Trabecular Meshwork

The possibility of allowing aqueous to flow directly into Schlemm's canal has generated considerable excitement:

• In European clinical practice, ablation of the trabecular tissue has been elegantly done with an excimer laser using an ab interno approach.
• In early trials, surgeons have introduced stents. (The stent lumen connects the anterior chamber with Schlemm's canal.) These may be placed using ab externo (EyePass, GMP Companies, Inc., Drs. Brown and Lynch) or ab interno (Glaukos Inc., Dr. Hill) techniques.

These procedures are "blebless." They don't lead to hypotony because physiologic episcleral venous pressure is 9 to 10 mm Hg.

Only the Beginning

The search for the ideal glaucoma surgical procedure continues. The new techniques and devices described above are promising, but all of them need further study. They must be compared with trabeculectomy -- the gold standard for surgical intervention -- in randomized, prospective clinical trials, to determine their place in the management of glaucoma.

Dr. Nesti is a clinical glaucoma fellow at Wills Eye Hospital in Philadelphia, where Dr. Katz is the co-director of the Glaucoma Service. Dr. Alvim is a glaucoma research fellow from Brazil, also currently at Wills Eye Hospital.