Comparison of Outcomes of Circumferential Versus Hemi-Gonioscopy-Assisted Transluminal Trabeculotomy in Primary Open-Angle Glaucoma: 24-Month Results

Abstract

Purpose: To compare the clinical outcomes of circumferential (360°) versus hemi (180°) gonioscopy-assisted transluminal trabeculotomy (GATT) in patients with primary open-angle glaucoma (POAG) over a 24-month follow-up. Methods: This retrospective, comparative study included 90 eyes of 90 patients with POAG who underwent either standalone circumferential GATT (n = 46) or hemi-GATT (n = 44). Inclusion criteria required uncontrolled intraocular pressure (IOP) (≥ 20 mmHg) despite maximal medical therapy or intolerance to medications. Main outcomes included surgical success (complete and qualified), IOP, number of anti-glaucoma medications, and postoperative complications. Complete success was defined as achieving a target IOP of 6–18 mmHg with ≥ 20% reduction from baseline without medications or further surgery. Qualified success allowed medication use. Results: At 24 months, mean IOP reduction was greater in the circumferential GATT group (− 10.1 ± 5.3 mmHg; 39.7%) than the hemi-GATT group (− 8.0 ± 3.8 mmHg; 32%) (p < 0.05). Complete success was achieved in 58.7% and 34.1% of eyes in the circumferential and hemi-GATT groups, respectively (p = 0.006). Qualified success rates were 78.2% (circumferential GATT) vs. 61.3% (hemi-GATT) (p = 0.051). The most common complication was transient hyphema, observed more frequently after circumferential GATT (54.6%) than hemi-GATT (31.8%) (p = 0.03). IOP spikes (> 30 mmHg) occurred in 10.8% of circumferential and 9% of hemi-GATT eyes (p = 0.71), all managed conservatively. Conclusions: Both circumferential and hemi-GATT effectively lowered IOP in patients with POAG. Circumferential GATT achieved greater IOP reduction and higher complete success, while hemi-GATT was associated with fewer hyphema events. Hemi-GATT may be preferable in patients at higher bleeding risk, whereas circumferential GATT may provide superior IOP control.