Prognosis and risk factors for early postoperative wound leaks after trabeculectomy with and without 5-fluorouracil

To evaluate the prognosis of patients who develop early postoperative wound leaks after trabeculectomy, to determine risk factors for these leaks, and to document adverse outcomes.

Multicenter randomized clinical trial.

Two-hundred thirteen patients with previous cataract surgery or failed filtering surgery were randomized to either trabeculectomy (standard group) or trabeculectomy with postoperative subconjunctival 5-fluorouracil injections (5-FU group). Masked measurements of intraocular pressure were performed throughout 5 years of follow-up. Failure was defined as a reoperation to control intraocular pressure or an intraocular pressure greater than 21 mm Hg with or without medication at or after the 1-year examination. An early postoperative wound leak was defined as one that developed within 2 weeks postoperatively, determined by daily topical fluorescein testing.

At 1 year (5 years), the success rate for the 5-FU group was 80% (54%) in eyes without a leak, and 60% (28%) in those with a leak. The 1-year (5-year) success rate in the standard group was 50% (24%) in those without a leak and 44% (15%) with a leak. (P =.018, log-rank test, adjusted for treatment). Leaks were more common in the 5-FU group, 34/105 (32%), than in the standard treatment group, 22/108 (20%), (P =.066, chi(2)). More eyes with one-layer conjunctiva-Tenon capsule closure developed leaks 40/115 (35%) than those with two-layer closure 16/95 (17%) (P =.006, chi(2)). More eyes with a trabeculectomy located inferiorly 29/76 (38%) developed leaks than those done superiorly 27/137 (20%) (P =.006, chi(2)). Patients with leaks were older (66 years) than those without leaks (60 years) (P =.011, t test).

An early postoperative wound leak was a risk factor for trabeculectomy failure. We recommend that trabeculectomy be performed in a superior location with a two-layer closure.