Surgical Outcomes in Childhood Uveitic Glaucoma


To evaluate surgical outcomes and to describe a strategy for pediatric uveitic glaucoma.


Retrospective case series.


The clinical practice of a single surgeon identified 36 patients with juvenile uveitic glaucoma (diagnosed before 18 years of age) who underwent 1 or more intraocular pressure (IOP)-lowering surgical procedures. The first eye operated on was included in the analysis. The main indication for success was IOP of less than 21 mm Hg with controlled inflammation, without further IOP-lowering surgery or devastating complication.


Patients with uveitic glaucoma associated with juvenile idiopathic arthritis (n = 20), idiopathic uveitis (n = 9), other (n = 6) were included. Mean age ± standard deviation at initial glaucoma surgery was 11.1 ± 4.4 years with a follow-up of 5.6 ± 4.8 years (range, 0.2 to 16.4 years). Goniotomy was the initial surgical procedure in 31 (86%) eyes. Fifteen eyes did not require further IOP-lowering surgery. Sixteen eyes had additional IOP-lowering surgery: second goniotomy (n = 9), glaucoma drainage device (n = 6), and trabeculectomy (n = 1). By Kaplan-Meier survival analysis, the first versus the first or second goniotomy were successful at 10 years in 48% (95% confidence interval, 28% to 65%) versus 69% (95% confidence interval, 47% to 84%). Goniotomy failure was noncorrelative with phakic status or presence of preoperative synechiae. Five eyes (14%) had initial glaucoma drainage device implantation because of closed angles. Visual acuity was similar between initial and final examinations (0.37 ± 0.48 logarithm of the minimal angle of resolution units [Snellen 20/47] vs 0.28 ± 0.34 logarithm of the minimal angle of resolution units [Snellen 20/38]). IOP was reduced (33.0 ± 8.0 mm Hg vs 12.6 ± 4.5 mm Hg; P < .0001), as was number of glaucoma medications (3.2 ± 1.1 vs 1.2 ± 1.3; P < .0001). Eyes undergoing cataract removal after successful glaucoma surgery maintained IOP control. Glaucoma was controlled and vision was preserved in 33 (92%) eyes.


Refractory juvenile uveitic glaucoma was managed successfully by goniosurgery and glaucoma drainage device implantation. Cataract removal can be accomplished successfully, provided IOP and uveitis are well controlled before surgery.

Uveitis in children can engender significant visual impairment. Chronic inflammation can lead to glaucoma, cataracts, band keratopathy, synechiae, and macular edema. Fortunately, medical treatment with ocular and systemic corticosteroids and corticosteroid-sparing therapy has improved control of inflammation and visual prognosis. However, many eyes affected with uveitis still require surgical intervention to control intraocular pressure (IOP) and to maximize vision.

Elevation of IOP in uveitis has been attributed to a variety of inflammatory-related factors culminating in increased resistance through the outflow pathways, often exacerbated by necessary topical corticosteroid exposure. The complex balance between controlling both IOP and inflammation makes uveitic glaucoma a therapeutic challenge, and some cases become unresponsive to medical treatment.

The surgical management of glaucoma secondary to uveitis in children presents difficulties related both to the young age of the patients as well as the underlying uveitis, and reports specific to juvenile uveitic glaucoma are limited. Glaucoma drainage device surgery has been reported effective in a small series of pediatric uveitic glaucoma cases with short follow-up. Goniotomy has been proposed as an effective surgery for pediatric glaucoma associated with chronic anterior uveitis, with reasonable long-term success and low complication rates, but with lower success in eyes with anterior synechiae and after cataract removal. Despite the reported success and simplicity of goniosurgery for glaucoma related to uveitis in children, there remains no uniformly accepted strategy for addressing these difficult cases.

In the current study, we evaluated our surgical outcomes for cases of pediatric uveitic glaucoma that were uncontrolled by medical treatment and suggest a surgical algorithm for successfully treating these difficult cases.


Retrospective chart review identified all patients diagnosed with glaucoma secondary to childhood uveitis before age 18 years from July 1994 through July 2011 who underwent at least 1 surgical procedure to control IOP. The IOP and inflammatory response to a particular glaucoma surgery on the first eye largely governed the clinical decision regarding surgery on the second eye. Hence, in children requiring glaucoma surgery in both eyes, the first eye that underwent surgery was included for most analyses.

Surgical success was defined as a final IOP of less than 21 mm Hg (and considered clinically adequate for glaucoma control), controlled inflammation, and without serious complication or need for further surgical intervention to reduce IOP. Information including date of birth, self-reported race, sex, type of uveitis, and date of glaucoma diagnosis were obtained for each patient (and included eye). Preoperative and postoperative information collected for each glaucoma surgery included examination findings (visual acuity [VA], IOP, slit-lamp examination, inflammation status, gonioscopy, optic nerve appearance), glaucoma medications, and anti-inflammatory medications (topical and systemic). Inflammation was quantified uniformly in medical records by the number of cells per high-power field using a 1-mm beam on a Haag-Streit slit lamp (Haig-Streit International, Koeniz, Switzerland). IOPs were measured by Goldmann applanation when possible. Tono-Pen (Reichert, DePew, New York, USA) readings were used in the operating room under anesthesia, and Tono-Pen or Icare rebound tonometer (Icare, Espoo, Finland) readings were used if the child was unable to cooperate for Goldmann applanation tonometry. All intraocular surgical procedures were documented along with intraoperative and postoperative complications.

Glaucoma surgeries included in the analysis were performed by or under direct supervision of a single attending surgeon (S.F.F.). Goniotomies were performed using a standard, previously described technique with initial surgery involving the temporal 180 degrees and subsequent goniotomy performed on unoperated angle (usually nasal or inferior). Ahmed glaucoma valve (models S-2 or FP-7; New World Medical, Rancho Cucamonga, California, USA) or Baerveldt (250 or 350 mm 2 ; Advanced Medical Optics, Santa Ana, California, USA) glaucoma drainage devices were placed in the superotemporal (or rarely the inferotemporal) quadrant using a standard technique as previously described. The size of the Baerveldt plate was selected based on the age of the patient and size (axial length) of the eye. After surgery, topical antibiotics and corticosteroids were administered. Topical corticosteroids were tapered back to preoperative levels as dictated by inflammation. Oral corticosteroids were used briefly in most cases, and uveitis was well controlled before surgery in most cases.

For all statistical analyses, P values of less than .05 were considered to be statistically significant. No multiple comparisons were required. All data shown are mean ± standard deviation, unless otherwise stated. Comparisons were performed using a 2-tailed t test, analysis of variance test, Fisher exact test, Wilcoxon rank-sum test, and Kaplan-Meier survival analysis (Graph Pad Prism, LaJolla, California, USA) as appropriate. Survival (surgical success) was defined as IOP less than 21 mm Hg, clinically stable, and no additional IOP-lowering surgery.


Included were 36 patients diagnosed with glaucoma secondary to uveitis before 18 years of age. The mean age at time of the first glaucoma surgery was 11.1 ± 4.4 years (range, 0.3 to 21.2 years). Nineteen (53%) of the 36 patients required glaucoma surgery in both eyes. Complete analysis was conducted on the first eye operated. Twenty-three patients (64%) were female. There were 19 right eyes and 17 left eyes in the primary analysis. Racial demographics were as follows: 26 patients were white, 3 patients were Hispanic, 6 patients were black, and 1 patient was Asian. Twenty of the patients (56%) had uveitis associated with juvenile idiopathic arthritis, whereas 9 patients were diagnosed with idiopathic anterior uveitis. Two patients each had diagnoses of sarcoidosis, pars planitis, and tubular interstitial nephritis, respectively. One patient had toxocariasis.

Review of the first eye operated for each patient demonstrated that 56 goniotomy or glaucoma drainage device implantation surgeries were performed on 36 eyes ( Figure 1 ) with a mean follow-up time from the first glaucoma surgery of 5.6 ± 4.8 years (median, 3.5 years; range, 0.2 to 17.2 years). The logarithm of the minimal angle of resolution VA before the first glaucoma surgery was 0.37 ± 0.48 (Snellen equivalent, 20/47). The mean preoperative IOP was 33.0 ± 8.0 mm Hg with 3.2 ± 1.1 glaucoma medications and 2.0 ± 1.2 anti-inflammatory medications. Before initial glaucoma surgery, 26 of 36  patients (72%) were receiving systemic anti-inflammatory medications, including a combination of corticosteroid-sparing therapy (25 patients) and oral corticosteroids (5 patients). Twenty-seven of 36 patients (75%) were using topical corticosteroids before surgery in the eye included in the analysis. Five of 36 patients (14%) were not receiving any systemic or topical anti-inflammatory medications.

Figure 1

Diagram showing surgical strategy for treating pediatric uveitic glaucoma. Thirty-six eyes were included in this surgical strategy which represents surgeries performed by one authors (S.F.F.) for pediatric uveitic glaucoma. Shading denotes final outcome for given eyes. F/U = follow-up.

Thirty-one (86%) of 36 eyes underwent goniotomy surgery as the first glaucoma surgery, whereas 5 eyes (14%) underwent initial glaucoma drainage device placement ( Table 1 ). There were no significant differences in age, VA, IOP, number of glaucoma medications, or number of anti-inflammatory medications between eyes that underwent initial goniotomy versus glaucoma drainage device placement.

Table 1

Ocular Findings and Inflammatory Characteristics before Initial Glaucoma Surgery in Eyes with Glaucoma Secondary to Pediatric Uveitis

Initial Glaucoma Surgery Goniotomy (n = 31) Glaucoma Drainage Device (n = 5) All Eyes (n = 36)
Age (year) 10.9 ± 4.4 12.1 ± 4.3 11.1 ± 4.4
VA logMAR (Snellen equivalent) 0.36 ± 0.52 (20/46) 0.42 ± 0.11 (20/53) 0.37 ± 0.48 (20/47)
Intraocular pressure (mm Hg) 34.0 ± 7.8 27.2 ± 7.6 33.0 ± 8.0
No. of glaucoma medications a 3.3 ± 1.1 3.0 ± 1.2 3.2 ± 1.1
No. of anti-inflammatory medications b 2.1 ± 1.1 2.0 ±1.2 2.0 ± 1.2
Lens status Phakic (14 cataract), n = 21
Aphakic, n = 4
Pseudophakic, n = 6
Phakic, n = 0
Aphakic, n = 3
Pseudophakic, n = 2
Phakic (14 cataract), n = 21
Aphakic, n = 7
Pseudophakic, n = 8
Inflammation No cells/hpf, n = 13
1 to 4 cells/hpf, n = 17
≥5 cells/hpf, n = 0
Not reported, c n = 1
No cells/hpf, n = 0
1 to 4 cells/hpf, n = 3
≥5 cells/hpf, n = 2
Not reported, n = 0
No cells/hpf, n = 13
1 to 4 cells/hpf, n = 20
≥5 cells/hpf, n = 2
Not reported, c n = 1
Band keratopathy 11/31 2/5 13/36
Synechiae d 15/31 5/5 20/36

hpf = high-power field; logMAR = logarithm of the minimal angle of resolution; VA = visual acuity.

a Topical or systemic glaucoma medications.

b Topical or systemic anti-inflammatory medications.

c Patient was an infant and unable to assess anterior chamber inflammation at slit lamp.

d Denotes posterior or peripheral anterior synechiae or both.

Twenty-one (68%) of 31 eyes that underwent initial goniotomy were phakic, whereas all eyes that underwent initial glaucoma drainage device placement were aphakic or pseudophakic ( P = .008). Preoperative inflammation largely was well controlled in the goniotomy group, with 28 eyes (93%) showing fewer than 5 cells per high-power field. The median preoperative cup-to-disc ratio (recorded at examination by single clinician [S.F.F.]) was 0.7 (range, 0.2 to 0.95). Reliable automated visual field testing (less than one third fixation losses and less than 30% false-positive or false-negative responses) results were unavailable for most of all included eyes at the time of initial glaucoma surgery because of the age of the patient or poor VA. Before surgery, band keratopathy was documented in 11 (35%) of 31 eyes, and posterior or peripheral anterior synechiae, or both, was documented in 15 (54%) of 31 eyes.

All eyes that underwent glaucoma drainage device surgery had extensive peripheral anterior synechiae that contributed to the clinician’s decision to proceed with glaucoma drainage device placement as the initial glaucoma surgery. In addition, the 5 eyes that had initial glaucoma drainage device placement all had undergone prior cataract extraction (with or without intraocular lens placement) and 2 eyes had undergone prior Ahmed glaucoma drainage device placement by another ophthalmologist. These 2 eyes, on initial surgery, underwent explantation of the previous Ahmed shunt and replacement with a new Ahmed glaucoma drainage device. These 2 eyes are considered with the other 3 eyes that had initial glaucoma drainage device implantation at our institution, but analysis of this group was performed with and without these eyes.

Surgical Interventions


Kaplan-Meier survival analysis of the first goniotomy surgery demonstrated success at 1, 2, and 10 years of 76% (95% confidence interval [CI], 56% to 88%), 57% (95% CI, 37% to 73%), and 48% (95% CI, 28% to 65%), respectively ( Figure 2 , Top). In the 16 eyes in which the first goniotomy failed, 6 underwent glaucoma drainage device implantation and 1 underwent trabeculectomy with mitomycin C ( Figure 1 ). Nine eyes underwent a second goniotomy. Kaplan-Meier analysis showed cumulative surgical success of 1 or 2 goniotomy surgeries at 1, 2, and 10 years of 83% (95% CI, 63% to 92%), 79% (95% CI, 59% to 90%), and 69% (95% CI, 47% to 84%), respectively ( Figure 2 , Bottom). In 2 eyes, the second goniotomy failed and the patients required glaucoma drainage device surgery. One of these eyes had placement of an Ahmed glaucoma drainage device that controlled the IOP for 11.1 years, after which the first device was exchanged for a new Ahmed device ( Figure 1 ).

Figure 2

Kaplan-Meier analysis of surgical success of goniotomy for pediatric uveitic glaucoma. The probability of success versus follow-up time is shown (Top) after 1 or (Bottom) after 1 or 2 goniotomy surgeries. Dotted lines denote 95% confidence intervals (CIs). Mean cumulative probability of success for 1 versus 1 or 2 surgeries was 76% (95% CI, 56 to 88) versus 83% (95% CI, 63 to 92) at 1 year, 57% (95% CI, 37 to 73) versus 79% (95% CI, 59 to 90) at 2 years, and 48% (95% CI, 28 to 65) versus 69% (95% CI, 47 to 84) at 10 years. Vertical crosshatches indicate censored eyes (successful at last follow-up). For 1 versus 1 or 2 goniotomies, the number of surviving eyes was 20 versus 22 at 1 year, 14 versus 20 at 2 years, and 3 versus 4 at 10 years.

In this series of limited size, we did not find a statistically significant association between goniotomy success and preoperative lens status. Hence of the 22 eyes in which goniotomy surgery was successful, 15 were phakic, 3 were aphakic, and 4 were pseudophakic; the distribution of lens status was similar in the 9 eyes in which goniotomy surgery was not successful (6 phakic, 1 aphakic, and 2 pseudophakic). Goniotomy was not performed in eyes with extensive preoperative synechiae. Not surprisingly we failed to find a statistically significant association between goniotomy success and extent of preoperative synechiae. Hence, 8 (36%) of 22 eyes in which goniotomy was successful had synechiae, whereas 7 (78%) of the 9 eyes in which goniotomy failed had synechiae. Goniotomy was associated with small postoperative hyphemas in 32 (80%) of 40 total procedures, all of which resolved spontaneously.

Glaucoma Drainage Device

Five eyes underwent initial glaucoma drainage device placement. As previously mentioned, before surgery these eyes had extensive peripheral anterior synechiae formation and subsequent predominantly closed angles. Four of these eyes (80%) have not required further IOP-lowering surgery with a mean of 2.4 ± 1.7 years of follow-up (median, 2.11 years; range, 0.2 to 4.9 years). One eye had initial placement of an Ahmed shunt, which then was replaced with a Baerveldt device 3.8 years later, with IOP control at last follow-up 1 year later. Two eyes underwent additional glaucoma drainage device-related surgeries (not for IOP reduction) including revisions (i.e., tube shortening or lengthening). There were no intraoperative complications during glaucoma drainage device placement, and there were no cases of exposure of the tube or reservoir. Analysis of results excluding the 2 eyes treated after initial glaucoma implant surgery at an outside location did not alter the results above.

Second-Eye Analysis

Of the 36 patients whose first eyes were enrolled in this retrospective study, 19 patients also had uveitic glaucoma requiring surgery in their fellow eyes. The mean time between the initial glaucoma surgery on the first eye versus the second eye was 0.8 ± 1.0 years. Of these 19 eyes, 16 underwent goniotomy as the initial surgery, whereas 3 underwent placement of a glaucoma drainage device initially. Five of the eyes underwent a second goniotomy, whereas 5 eyes underwent glaucoma drainage device implantation after failed goniotomy. Kaplan-Meier analysis of the second eyes demonstrated a 1-, 2-, and 10-year survival of 88% (95% CI, 59% to 97%), 74% (95% CI, 45% to 90%), and 60% (95% CI, 24% to 83%), respectively, for up to 2 goniotomies. Kaplan-Meier analysis of both the first and second eyes (n = 55) for up to 2 goniotomies revealed a 1-, 2-, and 10-year survival of 84% (95% CI, 70% to 92%), 77% (95% CI, 62% to 87%), and 65% (95% CI, 45% to 79%), respectively.

Vision, Intraocular Pressure, and Uveitis Control After Glaucoma Surgery

The ocular characteristics and inflammatory status of all eyes was reviewed at last examination with consideration for surgical interventions chosen ( Table 2 ). At last examination, 34 (94%) of 36 eyes showed preservation of preoperative vision. The overall logarithm of the minimal angle of resolution VA at time of final examination (0.28 ± 0.34; Snellen equivalent, 20/38) was not significantly different from VA at the initial examination. Because our length of follow-up from initial glaucoma surgery ranged from 0.2 to 17.2 years, we also reported VA in eyes with follow-up of less than 1 year, 1 to 5 years, and more than 5 years ( Table 3 ). We found no statistically significant change of preoperative VA regardless of follow-up time.

Table 2

Ocular Findings and Inflammatory Characteristics at Final Examination in Eyes with Childhood Uveitic Glaucoma that Required Intraocular Pressure-Lowering Surgery

Surgery Strategy Initial Goniotomy (n = 31) a Initial Glaucoma Drainage Device b (n = 5) All Eyes (n = 36)
Goniotomy Only (n = 22) Goniotomy Then Glaucoma Drainage Device (n = 8)
Age (year) 16.4 ± 7.3 16.6 ± 7.3 14.4 ± 3.3 17.2 ± 8.1
Follow-up (year) 4.8 ± 4.4 8.5 ± 5.2 2.4 ± 1.7 5.6 ± 4.8
No. of glaucoma surgeries 1.4 ± 0.5 2.6 ± 0.9 2.0 ± 1.4 1.8 ± 0.9
VA logMAR (Snellen equivalent) 0.17 ± 0.17 (20/30) 0.52 ± 0.58 (20/66) 0.39 ± 0.24 (20/49) 0.28 ± 0.34 (20/38)
IOP (mm Hg) 13.5 ± 4.5 10.9 ± 3.0 13.6 ± 3.2 12.6 ± 4.5
No. of glaucoma medications 1.5 ± 1.4 0.9 ± 1.3 0.6 ± 0.9 1.2 ± 1.3
No. of anti-inflammatory medications 1.4 ± 1.3 1.9 ± 0.8 2.2 ± 1.3 1.7 ± 1.2
Lens status Phakic (7 cataract), n = 9
Aphakic, n = 3
Pseudophakic, n = 10
Phakic (1 cataract), n = 1
Aphakic, n = 0
Pseudophakic, n = 7
Phakic, n = 0
Aphakic, n = 3
Pseudophakic, n = 2
Phakic (8 cataract), n = 10
Aphakic, n = 7
Pseudophakic, n = 19
Inflammation No cells/hpf, n = 16
1 to 4 cells/hpf , n = 6
≥5 cells/hpf, n = 0
No cells/hpf, n = 4
1 to 4 cells/hpf, n = 2
≥5 cells/hpf, n = 2
No cells/hpf, n = 1
1 to 4 cells/hpf, n = 3
≥5 cells/hpf, n = 1
No cells/hpf, n = 21
1 to 4 cells/hpf, n = 11
≥5 cells/hpf, n = 3
Not reported, c n = 1
Band keratopathy 5/22 4/8 3/5 13/36
Synechiae 11/22 6/8 5/5 23/36

hpf = high-power field; IOP = intraocular pressure; logMAR = logarithm of the minimal angle of resolution; VA = visual acuity.

a Thirty-one eyes underwent goniotomy as initial glaucoma surgery performed by one of the authors (S.F.F.). Twenty-two of these eyes underwent 1 or 2 goniotomy surgeries as the only glaucoma surgeries. Eight eyes first underwent goniotomy surgery and then underwent placement of a glaucoma drainage device. One eye underwent goniotomy surgery followed by trabeculectomy. This eye, which had uncontrolled uveitis and became phthisical secondary to medication noncompliance, is not represented in this column, but is included in the last column, which displays data for all 36 eyes.

b Five eyes underwent glaucoma drainage device placement as initial glaucoma surgery performed by one of the authors (S.F.F.).

c Inflammation status was not reported for the phthisical eye that had undergone goniotomy and trabeculectomy.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jan 9, 2017 | Posted by in OPHTHALMOLOGY | Comments Off on Surgical Outcomes in Childhood Uveitic Glaucoma

Full access? Get Clinical Tree

Get Clinical Tree app for offline access