Predictors of Selective Laser Trabeculoplasty Efficacy: Results from the Swedish Optimal Selective Laser...

To investigate the impact of potential predictor variables on selective laser trabeculoplasty (SLT) efficacy in the Swedish Optimal SLT (OSLT) trial.

Post hoc analysis of a multicenter, masked, randomized controlled trial.

Five hundred twelve eyes from 399 patients enrolled in the OSLT trial.

Patients were randomized to one of 4 SLT variants, differing in treatment extent (180° or 360°), and laser power (standard or high). Analyses were performed with linear mixed models, which allowed optimal use of all the data and applied comprehensive adjustment to the results. A wide selection of potential predictors for SLT efficacy were first analyzed separately regarding their association with intraocular pressure reduction (IOPR). Predictors meeting a threshold of P < 0.1 were then included in a multivariable candidate model, which was refined through manual stepwise backwards selection until only significant variables (P < 0.05) remained. Thereafter, other variables of interest were evaluated in this model.

The mean IOPR 1 to 6 months after SLT.

Selective laser trabeculoplasty performed in 360 degrees, at an energy level giving a 50% to 75% cavitation bubble frequency (the 360/high protocol) remained the most efficacious SLT variant after adjustment for predictors (P < 0.001). The coefficient estimate (CE) of 360/high SLT was 2.0; SLT performed in 360 degrees, at an energy level titrated just below the cavitation bubble threshold (360/standard) CE was 0.9; and SLT performed in 180 degrees, at an energy level giving a 50% to 75% cavitation bubble frequency (180/high) CE was -0.1, with SLT performed in 180 degrees, at an energy level titrated just below the cavitation bubble threshold (180/low) as the reference. Higher baseline IOP was a positive predictor for both absolute IOPR (CE 0.35; P < 0.001) and relative IOPR (CE 0.72; P < 0.001). A larger IOPR from the prior SLT (CE 0.18; P = 0.004), as well as IOPR in the contralateral eye (CE 0.57; P < 0.001), were also highly significant positive predictors. Conversely, the present analyses suggest that SLT efficacy is diminished by pseudoexfoliations (CE -0.69; P = 0.02), the number of prior SLTs (CE -0.54; P = 0.004), and increased corneal thickness (CE -0.01; P < 0.002). Consultants/specialists and residents achieved similar SLT results, but individual surgeon performance mattered (P = 0.003), as well as if the surgeon used the dominant hand (CE 0.35; P = 0.050). However, age, glaucoma medication, cataract surgery, anterior chamber angle pigmentation, inflammatory symptoms, or postoperative anterior chamber flare had no correlation with SLT efficacy.

The 360/high SLT protocol remained the most effective and reliable technique after adjustment for multiple predictors. No evidence was found to advise against 360/high SLT for any particular group of patients.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.