Fornix-Based vs. Limbal-Based Conjunctival Trabeculectomy Flaps
American Journal of Ophthalmology, February 2017
Opinions differ on surgical techniques for trabeculectomy and, in particular, whether the conjunctival incision should be made in the fornix or the limbus. In a Cochrane systematic review of randomized controlled trials (RCTs) involving adults with glaucoma, Al-Haddad et al. found similar efficacy for the 2 conjunctival flap locations.
Six RCTs (N = 361 patients) met the authors’ inclusion criteria for comparing the benefits and complication rates of these methods. All trials had at least 12 months of follow-up, and 3 had at least 24 months. The main outcome measures were the proportion of failed trabeculectomies at 24 months and the mean intraocular pressure (IOP) at 24 months. Trabeculectomy failure was defined as the need for repeat surgery or uncontrolled IOP (>22 mm Hg) despite additional topical/systemic drugs.
None of the trials reported trabeculectomy failure at 24 months. Of the 3 studies that evaluated failure at 12 months, 1 reported a 2% rate of failure in the fornix-based group (1 of 50 eyes) and 6% in the limbal-based group (3 of 50 eyes). Another trial reported that 2 of 43 (4.6%) eyes in each group required reoperation during 3 years of follow-up. Of the 3 RCTs that included 24-month IOP data, no significant differences were found in mean postoperative IOP between the surgical groups.
With respect to complications, postoperative shallow anterior chamber was twice as common for limbal- versus fornix-based procedures. There were no other significant differences in complications between the 2 procedures.
The authors concluded that fornix-based and limbal-based flap incisions appear similar in terms of IOP control and bleb failure, but the risk of shallow anterior chamber was significantly greater after limbal-based surgery. However, the authors acknowledged that their low event rates and wide confidence intervals indicate a high degree of uncertainty about the findings.
The original article can be found here.