Outcomes of Valved and Nonvalved Tube Shunts in Neovascular Glaucoma.

Purpose: To determine the outcomes of Ahmed glaucoma valve (AGV; New World Medical Inc) and Baerveldt glaucoma implant (BGI; Advanced Medical Optics) surgery in the setting of neovascular glaucoma (NVG).

Design: Single-center, retrospective study.

Participants: Consecutive patients who underwent AGV or BGI surgery for the treatment of NVG and had ≥6 months of follow-up.

Methods: Chart review of AGV and BGI surgical outcomes in patients with NVG.

Main Outcome Measures: Progression to no light perception (NLP) vision and 6-month surgical failure, which was defined as intraocular pressure (IOP) >21 mmHg with medications or <5 mmHg at 2 consecutive visits, or glaucoma reoperation.

Results: A total of 152 eyes (91 AGV, 61 BGI) were included with an average follow-up of 29.6 ± 25.8 months. Baseline demographics and clinical characteristics were comparable between groups. At month 6, failure was similar between AGV and BGI eyes (21.6% vs. 25.9%; P = 0.552), but glaucoma medication use was lower in BGI eyes (P < 0.001). At the final visit, 18.7% of AGV and 14.8% of BGI eyes progressed to NLP vision (P = 0.530), and medication use was lower in BGI eyes (P < 0.0001). Multivariate analysis identified lower preoperative visual acuity (VA) (P = 0.001), failure to receive panretinal photocoagulation within 2 weeks of surgery (P = 0.003), and bilaterality of the underlying ischemic retinal pathology (P = 0.026) as the strongest predictors of NLP outcome. Age, sex, race, NVG etiology, tube type, preoperative IOP, extent of synechial angle closure preoperatively, preoperative hyphema, IOP at the first NLP visit, and final IOP were not significant predictors of NLP vision.

Conclusions: Eyes with AGV and BGI had comparable outcomes in NVG, although fewer medications were required in BGI eyes to control IOP. Progression to NLP vision was associated with poor baseline VA, delayed retinal treatment, and bilaterality of the underlying ischemic retinal pathology.