Surgeons should weigh efficacy, long-term complications of trabeculectomy

George L. Spaeth, MD

Glaucoma surgery has changed remarkably little since Holth, LaGrange and Elliot developed different forms of filtration procedures more than 100 years ago. Important, clinically significant modifications have been made from those early forms of sclerectomy, but the basic goal, to allow aqueous to percolate the surface of the globe, is the same. Seidel developed his test to help ascertain whether such operations were likely to work; specifically, when aqueous seeped through the conjunctiva — Seidel positive — the IOP was likely to be well-controlled. But that thin conjunctival covering sometimes developed full-thickness leaks, allowing bacteria to enter the anterior chamber; sometimes even blebs that were intact became infected. Some infected blebs developed endophthalmitis.

These issues are mentioned because complications continue to be the bane of the glaucoma patient and surgeon. The Collaborative Initial Glaucoma Treatment Study (CIGTS) provided a great opportunity to get a handle on the frequency and nature of complications associated with trabeculectomy performed by glaucoma specialists and community ophthalmologists. What this showed was that a definite infection inside the eye occurred in one of 285 trabeculectomies. Two other cases developed an anterior chamber reaction in association with an inflammation of the bleb, but neither had a decrease in vision, and neither probably had real endophthalmitis. Endophthalmitis is “life-threatening” to the eyeball, and while the patient in CIGTS ended up with relatively good vision, this is not always the case.

Is an incidence of 0.3% really “low,” to use the authors’ adjective? When an operation results in a potentially “vision-threatening“ complication in around 1% of the cases having that operation, but is effective in preventing an inevitable loss of vision in about 80%, is that “good” or “bad?” Are there other approaches with a better risk/benefit ratio?