by David S. Greenfield, M.D.

Clinical Assistant Professor of

Ophthalmology and Neurology

The New York Eye and Ear Infirmary

New York Medical College

Unquestionably, the use of antifibrotic agents such as mitomycin C and 5-fluorouracil has dramatically improved the outcome of glaucoma filtering surgery in eyes at high risk for failure. As such, we have witnessed the prevailing use of these agents in eyes with uncontrolled intraocular pressure (IOP) associated with disorders such as neovascularization, uveitis, and previous intraocular surgery including failed glaucoma filtration procedures. Reports of late complications associated with any surgical technique, however, represent the fundamental basis of what many have long referred to as "the curse of long-term follow-up". Clearly, the augmentation in IOP reduction provided by mitomycin C has been no stranger to this dictum as we have noticed an increase in the incidence of delayed-onset postoperative endophthalmitis.

Bleb-associated endophthalmitis is among the most dreaded and sight-threatening complications following glaucoma filtering surgery. In contrast to acute postoperative endophthalmitis, delayed-onset is a more common presentation and often develops after the first postoperative month. Patients typically present with pain, diffuse ocular injection, acute visual loss, and marked intraocular inflammation. Often there is a fibrino-purulent precipitate within the filtering bleb associated with a conjunctival epithelial defect on the surface. Although hypopyon and vitritis are the hallmark features of advanced infection, varying degrees of intraocular inflammation may make the diagnosis more difficult to establish.

The true incidence of endophthalmitis after glaucoma filtering surgery has traditionally been very difficult to establish. The reported rates have ranged between 0.2% and nearly 10%. Some of the variables that influence the calculated incidence include the total number of patients studied (often referred to as the "denominator"), how one defines endophthalmitis (often referred to as the "numerator"), the length of follow-up, the type of glaucoma surgery performed (e.g. full-thickness or guarded approach), the location of surgery (e.g. at the superior or inferior limbus), and the adjunctive use of antifibrotic agents.

Recently, we have reported the increased incidence of bleb-associated endophthalmitis after trabeculectomy with mitomycin C, particularly when performed at the inferior limbus. We undertook a retrospective analysis of 773 consecutive eyes that underwent glaucoma filtering surgery by a single surgeon. The course of 609 eyes from 485 patients with a minimum of 3 months of follow-up were reviewed. Mean follow-up was 16.0±11.5 months. Overall, 13/609 eyes (2.1%) developed endophthalmitis an average of 18.5±13.2 months after surgery. However, Kaplan-Meier estimates demonstrated that the rate of endophthalmitis was significantly greater after inferior trabeculectomy (7.8% per patient-year) than after superior trabeculectomy (1.3% per patient-year). In addition, a trend was noted with a greater overall rate of infection after trabeculectomy alone (2.5% per patient-year) as compared to trabeculectomy combined with cataract surgery and intraocular lens implantation (1% per patient-year).

In addition to trabeculectomy performed at the inferior limbus, other risk factors for the development of infection were identified. One patient developed endophthalmitis shortly after the onset of bilateral bacterial conjunctivitis. More importantly, bleb trauma preceded the development of endophthalmitis in 4/13 (31%) of patients. Three of these patients underwent argon laser suture lysis or bleb needle revision less than three weeks before infection and 1 patient had recent blunt ocular trauma. As previously reported, the most common organisms responsible for infection were Streptococcus viridans and Haemophilus influenza.

Although the visual outcomes in this cohort were poor with almost 40% of patients achieving less than 20/400 acuity after infection, it is important to note that nearly all the filtration capacity of these blebs survived the episode of infection. The mean IOP after recovery from endophthalmitis was 13±6.8 and more than 50% of patients were controlled without medication. Thus, an important consideration relates to whether or not these individuals remain at high risk for re-infection.

In summary, it appears that the overall incidence of endophthalmitis after trabeculectomy with mitomycin C exceeds the reported rate of infection for trabeculectomy without antifibrotic therapy. Excluding trabeculectomy at the inferior limbus, the incidence is similar to the reported rate of infection following full-thickness procedures. As such, it is likely that the thickness of the conjunctival bleb, rather than the nature of the surgical procedure itself, is responsible for the increased incidence. Inferior filters are unique in that they are frequently exposed and poorly covered by the eyelid, and are subject to the effects of repeated epithelial trauma as the lower eyelid rubs the bleb with each blink. As a result, one should strongly consider a superiorly placed glaucoma drainage device when the superior conjunctiva demonstrates significant cicatrization.

[ About NYGRI ][ Education ][ Glaucoma Network ]
[ How You Can Help ][ Question Corner ][ Research ]