Combined pars plana vitrectomy and pars plana baerveldt tube placement in eyes with neovascular glaucoma

Purpose: To describe characteristics and outcomes of combined pars plana vitrectomy and Baerveldt tube insertion procedure from 2005 to 2010 in eyes with neovascular glaucoma.

Method: Seventy-nine patients (89 eyes) with ≥2 months of follow-up were included. Outcome measures were visual acuity, intraocular pressure (IOP), number of glaucoma medications, and complications. Changes in mean logMAR visual acuity, IOP, and glaucoma medications were compared by a two-tailed t-test.

Results: Mean patient age was 69.0 years. Forty-three (54%) were male. Mean follow-up time was 19.9 months. Most common causes of neovascular glaucoma was diabetes (n = 63 [71%]) and central retinal vein occlusion (n = 21 [24%]). Eighty-six eyes (97%) underwent a 250 mm Baerveldt drainage device and 3 (3.4%) a 350 mm Baerveldt. Forty-five (51%) 20-gauge, 12 (13%) 23-gauge, and 32 (36%) 25-gauge pars plana vitrectomies were performed. Fifty-two eyes (58%) preoperatively and 23 (33%) postoperatively received intraocular injections for rubeosis and macular edema. Mean ± standard deviation logMAR visual acuity at 18-, 24-, 36-, and 48-month follow-up time points was significantly better than preoperative vision (P < 0.05). Preoperative versus final IOP and number of glaucoma medications were significantly decreased (P < 0.05). Fourteen eyes (16%) had a final visual acuity of no light perception. Most common complications included transient ocular hypertension (n = 82 [92%]), transient hypotony (n = 20 [22%]), hyphema (n = 19 [21%]), corneal edema (n = 17 [19%]), and vitreous hemorrhage (n = 14 [16%]). The frequency of transient hypotony, vitreous hemorrhage, and rubeosis was significantly (P < 0.05) higher in 20-gauge versus 23-/25-gauge pars plana vitrectomy eyes. Nine eyes (10%) required return to the operating room after combined procedure, including 4 eyes (4.5%) for retinal detachment and 3 (3.4%) for high IOP due to tube occlusion. Three eyes (3.4%) developed endophthalmitis and 2 (2.2%) progressed to being pre/phthisical (none were enucleated).

Conclusion: Combined pars plana vitrectomy and Baerveldt glaucoma shunt may be a useful procedure in reducing IOP and number of glaucoma medications in eyes with neovascular glaucoma along with stabilizing visual acuity in a majority of these eyes. Further studies are warranted to verify and expand on these findings.