Oral Ibuprofen is Associated With Reduced Likelihood of Early Bleb Failure After Trabeculectomy in High-Risk Glaucoma Patients

Abstract

Précis: 

Individuals prescribed ibuprofen after trabeculectomy have better postoperative intraocular pressure (IOP) control and a higher chance of bleb survival despite being at a higher risk of scarring.

Purpose: 

To investigate the effects of early adjunctive oral ibuprofen treatment on IOP and bleb failure in eyes at high risk of scarring.

Methods: 

In these retrospective analyses, 288 eyes of 273 patients (mean ± SD age: 68.56 ± 10.47 y; 32.60% females) with primary glaucoma who underwent trabeculectomy/phacotrabeculectomy at the Singapore National Eye Centre between April 2020 and April 2021 with a follow-up duration ≥1 year were included. Of these, 77 (26.7%) eyes deemed to be at high risk of scarring were administered oral ibuprofen ≥3 months postoperatively (mean ± SD ibuprofen administration duration: 4.08 ± 2.28 wk). Participant’s IOPs at baseline and at postoperative weeks 1, 2–3; and months 1, 2, 3, 6, and 12 were recorded. Bleb failure was defined as 2 consecutive IOP readings of >21, >18, and >15 mm Hg, and/or requiring remedial postoperative laser or surgery.

Results: 

The ibuprofen group experienced significantly greater postoperative IOP reductions at week 1 [mean difference, 95%CI: −2.89 (−5.22, −0.56) mm Hg] and month 1 [−2.29 (−4.53, −0.05) mm Hg]; and substantially lower odds of bleb failure at the >18 mm Hg [odds ratio, 95% CI: 0.39 (0.20–0.79)] and >15 mm Hg [0.52 (0.29–0.94)] thresholds, compared with the non-ibuprofen group. No differences in adverse ocular hypotony events were observed.

Conclusion: 

Early adjunctive oral ibuprofen administered to individuals at high risk of posttrabeculectomy scarring is associated with greater IOP reductions and reduced likelihood of bleb failure. Our results suggest that oral nonsteroidal anti-inflammatory drugs may be a safe way of improving trabeculectomy survival in high-risk eyes.

Trabeculectomy surgery, the most commonly performed incisional surgical procedure for glaucoma,¹ is often regarded as the “gold standard” surgical treatment for this disease.² It aims to reduce intraocular pressure (IOP), which remains the major modifiable risk factor,³ by creating a fistula between the anterior chamber and subconjunctival space through a sclerostomy.⁴ A major determinant of long-term surgical success lies in the wound healing response, as excessive inflammation can lead to scarring and eventual bleb failure.⁵ As such, intraoperative antimetabolites such as mitomycin C (MMC) are used to inhibit fibroblasts, in conjunction with postoperative topical corticosteroids to reduce the production of inflammatory mediators, thereby increasing surgical success rates.^6,7

Despite these prophylactic interventions, Asian, Afro-Caribbean and Hispanic eyes have been observed to be at significantly higher risk of an aggressive wound-healing response.^8,9 Adjunctive treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) is increasingly being viewed as an attractive option in such cases as it offers an alternative pathway to dampen the inflammatory response by inhibiting cyclooxygenase enzymes, with minimal side effects.¹⁰

However, there remains sparse and conflicting literature relating to the effectiveness of postoperative administration of NSAIDs in further reducing the risk of bleb failure after trabeculectomy. Early Molteno implant trials were one of the first studies to demonstrate the efficacy of a 6-week adjunctive dose of postoperative oral diclofenac in controlling IOP, a finding that was later replicated in a randomized controlled trial (RCT) by Petrov and colleagues, who administered a 2-week regime of topical nepafenac as an adjunct to topical dexamethasone pretrabeculectomy.^11,12 In contrast, an RCT by Cantor et al¹³ observed no significant difference in postoperative outcomes despite the administration of adjunctive topical flurbiprofen posttrabeculectomy. To address this knowledge gap, we evaluated the independent associations of the early adjunctive use of oral ibuprofen in posttrabeculectomy patients with IOP reduction and bleb failure over 1 year in a retrospective clinical cohort. We hypothesized that the use of early adjunctive oral ibuprofen is associated with a reduced likelihood of bleb failure in eyes at high risk of failure, compared with eyes not administered ibuprofen.

METHODS

Study Design and Population

Our study was a retrospective, clinical review of patients with glaucoma receiving care at the Singapore National Eye Centre. In brief, we reviewed the medical, pharmaceutical, and intraoperative records of all patients who underwent trabeculectomy or phacotrabeculectomy at Singapore National Eye Centre from April 2020 to April 2021, with a follow-up duration of at least 1 year. A standard Cairn’s type trabeculectomy with sclerostomy creation and scleral flap closure was performed for all patients, either independently by a glaucoma specialist or by a glaucoma fellow under the supervision of a glaucoma specialist. We included patients who were diagnosed with primary open angle glaucoma, primary angle closure glaucoma, and normal tension glaucoma. MMC was routinely used intraoperatively for all surgeries, either through injection of 0.2 mL of 0.2 mg/mL MMC or application of 0.4 mg/mL for 1–3 minutes using a sponge. All patients have prescribed a combination of topical corticosteroids and antibiotics postoperatively as per standard in the institution. During each subsequent clinic visit, the blebs were assessed and IOP was measured by the attending glaucoma specialist for any signs of imminent bleb failure. Clinical signs of early bleb failure included at least one of the following features: increased conjunctival vascularization, low-lying flat blebs, encapsulation of the blebs, presence of tenon cysts, or raised IOP.¹⁴ Those at-risk subjects would be started on oral ibuprofen for a minimum of 1 week at a dose of 400 mg Three times a day with concomitant prescription of gastric protective medications such as omeprazole or famotidine, barring any contraindications such as drug allergies or severe renal impairment. Patients diagnosed with secondary glaucoma, or who were administered ibuprofen >3 months posttrabeculectomy for any reason, were excluded from the analyses. The latter was done as the postsurgical inflammatory response should have largely abated after 3 months, and anti-inflammatory drugs are not usually indicated to manage fibrosis after this point. A waiver of informed consent was obtained from the SingHealth Centralized Institutional Review Board (CIRB Reference #2022/2550) due to the retrospective nature of this study, and all data were anonymized before analyses were carried out.

Data Collection

We collected sociodemographic data from the patient’s electronic medical records, including age, sex, and ethnicity. Preoperative details and medical histories such as type of glaucoma, preoperative visual acuity and IOP, type, number and duration of baseline topical glaucoma medications, use of systemic glaucoma medications, pre-existing ocular surface conditions (scarring and ocular surface disease), and history of any previous ocular surgeries or lasers were included. Postoperative data included the use of ibuprofen, duration between surgery and starting ibuprofen, total duration of ibuprofen use, postoperative IOP at week 1, week 2–3, month 1, month 2, month 3, month 6, and year 1, presence and a number of bleb needlings, the type of adjunctive antifibrotic injected [5-fluorouracil (5-FU), MMC, or dexamethasone], duration between surgery and bleb needlings, restarting of topical glaucoma medications, and any laser procedures (ie, trans-scleral cyclophotocoagulation) or repeat surgeries performed. We also tabulated the number of bleb failures, defined as 2 consecutive IOP measurements over 3 stated thresholds (>15 mm Hg, >18 mm Hg, and >21 mm Hg), and/or having had surgery trans-scleral cyclophotocoagulation to resolve uncontrolled IOP after trabeculectomy, based on previously published trial reporting guidelines.

Statistical Analyses

All analyses were performed using STATA 16.1 (StataCorp LP, College Station). Differences in baseline patient characteristics and postoperative variables between individuals who were prescribed oral ibuprofen after trabeculectomy (“ibuprofen group”) versus those who were not (“non-ibuprofen group”) were tested using the independent t test for continuous variables, and the χ² test and Fisher exact tests for categorical variables. Generalized linear models were then constructed to determine the association between ibuprofen use with change in IOP from baseline across the 7 time points. We also utilized generalized linear models, assuming a log-link function for binary data, to elicit the associations between ibuprofen use and bleb failure at the 3 IOP thresholds. In both cases, we adjusted for baseline and postoperative variables that are known confounders or were significantly different between the ibuprofen and non-ibuprofen groups. This included age, sex, ethnicity, total duration of baseline topical glaucoma medications and oral acetazolamide use, type and a number of antifibrotic agents used for bleb needling and subconjunctival injections, need for postoperative glaucoma medications, and postoperative laser procedures. Because of the small number of individuals prescribed ibuprofen in this study, the backward stepwise selection was utilized for both models, with the threshold for variable removal set at 0.1. Robust standard errors were used to account for any potential clustering effects resulting from high between-eye correlations. Based on previous work and guidelines,^15,16 we also tested for potential interactions (ie, effect-modification) between patient sociodemographic variables (age, sex, and ethnicity) and the outcomes. No significant interactions were found; hence no stratified analyses were conducted. A 2-sided P value of <0.05 was considered to be statistically significant.

Sample Size Calculations

Assuming a margin of equivalence of 2 mm Hg between the ibuprofen and the non-ibuprofen groups and based on the current proportion of ibuprofen cases within the study database (~30%), we estimate that we would need 370 total individuals, comprising 111 persons on ibuprofen and 259 not prescribed this oral NSAID, to achieve results at 80% power and the 5% significance level.

RESULTS

A total of 344 eyes of 324 patients with glaucoma who underwent phacotrabeculectomy or trabeculectomy with at least 1 year of follow-up were identified. After excluding 34 patients who were diagnosed with secondary glaucoma, and 16 patients who were administered ibuprofen more than 3 months after trabeculectomy was performed, a total of 288 eyes from 273 eligible patients were eventually included in our final analysis (Table 1).

TABLE 1 - Comparison of Baseline Participant Characteristics Between Eyes Treated With and Without Ibuprofen

	Mean (SD)
Parameter	All patients, 288 eyes 273 patients; n (%)	Without ibuprofen, 211 eyes 200 patients; n (%)	With ibuprofen, 77 eyes 73 patients; n (%)	P *
Patient demographics
Age (y)	68.56 (10.47)	70.29 (9.50)	63.82 (11.56)	<0.001
Sex
Male	184 (67.40)	131 (65.50)	53 (72.60)	0.268
Female	89 (32.60)	69 (34.50)	20 (27.40)	—
Ethnicity
Chinese	229 (83.88)	166 (83.00)	63 (86.30)	0.427
Malay	13 (4.76)	8 (4.00)	5 (6.85)	—
Indian	24 (8.79)	20 (10.00)	4 (5.48)	—
Others	7 (2.56)	6 (3.00)	1 (1.37)	—
Type of glaucoma
POAG	159 (55.21)	120 (56.87)	39 (50.65)	0.322
PACG	81 (28.12)	60 (28.44)	21 (27.27)	—
NTG	48 (16.67)	31 (14.69)	17 (22.08)	—
Preopertaive VA	0.42 (0.34)	0.40 (0.31)	0.46 (0.41)	0.253
Preopertaive IOP	18.44 (6.52)	17.90 (5.78)	19.92 (8.08)	0.045
Type of surgery
Trabeculectomy only	76 (26.39)	61 (28.77)	17 (21.79)	0.242
Phacotrabeculectomy	212 (73.61)	151 (71.23)	61 (78.21)	—
Alpha antagonist
No	81 (28.12)	61 (28.77)	20 (25.64)	0.616
Yes	207 (71.88)	151 (71.23)	58 (74.36)	—
Beta blocker
No	35 (12.15)	26 (12.26)	9 (11.54)	0.869
Yes	253 (87.85)	186 (87.74)	69 (88.46)	—
CAI
No	85 (29.51)	59 (27.83)	26 (33.33)	0.379
Yes	203 (70.49)	153 (72.17)	52 (66.67)	—
PGA
No	7 (2.43)	3 (1.42)	4 (5.13)	0.088
Yes	281 (97.57)	209 (98.58)	74 (94.87)	—
No. eye drops utilized by each patient	3.28 (0.89)	3.29 (0.86)	3.23 (0.97)	0.645
Duration (mo) of use of each type of topical glaucoma eyedrop
Alpha antagonist	25.34 (30.42)	25.91 (32.08)	23.84 (25.77)	0.649
Beta blocker	62.18 (58.75)	68.12 (60.15)	46.01 (51.80)	0.005
Carbonic Anhydrase Inhibitor	38.23 (41.08)	40.88 (41.98)	30.33 (37.54)	0.101
Prostaglandin analogs	66.21 (54.89)	71.89 (56.79)	50.02 (45.64)	0.002
Total duration (mo) of use of any topical glaucoma eyedrop	68.28 (58.05)	74.54 (59.30)	52.62 (51.48)	0.003
Acetazolamide
No	254 (88.19)	189 (89.15)	65 (83.33)	0.185
Yes	34 (11.81)	23 (10.85)	13 (16.67)	—
Pre-existing conjunctival scarring
No	287 (99.65)	211 (99.53)	78 (100.00)	NA
Yes	1 (0.35)	1 (0.47)	0 (0.00)	—
Pre-existing ocular surface disease
No	278 (96.53)	205 (96.70)	75 (96.15)	0.829
Yes	10 (3.47)	7 (3.30)	3 (3.85)	—
Previous eye operations on the operated eye
No	233 (80.90)	169 (79.72)	64 (82.05)	0.657
Yes	55 (19.10)	43 (20.28)	14 (17.95)	—
Previous eye lasers on the operated eye†
No	193 (67.01)	138 (65.09)	57 (73.08)	0.219
Yes	95 (32.99)	74 (34.91)	21 (26.92)	—

^*Eye-level characteristics: generalized linear models adjusted for possible clustering effect by eye.

^†Includes selective laser trabeculoplasty, pan retinal photocoagulation, laser peripheral iridotomy, laser-assisted in situ keratomileusis, photodynamic therapy, and micropulse laser trans-scleral cyclophotocoagulation.

Patient-level characteristics (age, sex, ethnicity, and diagnosis): t test or χ² test at patient-level.

NA indicates not possible to estimate differences due to null (“0”) value in comparison group.

IOP indicates intraocular pressure; NTG, normal tension glaucoma; PACG, primary angle closure glaucoma; PGA, Prostaglandin analogue; POAG, primary open angle glaucoma; VA, visual acuity.

The mean age of all patients was 68.56 ± 10.47 years, with those on ibuprofen (N = 77 eyes of 73 patients) being significantly younger (63.82 ± 11.56 y) than those not prescribed ibuprofen (N = 211 eyes of 200 patients; 70.29 ± 9.50 y; P < 0.001). Compared with those not on ibuprofen, those on ibuprofen had a significantly higher preoperative IOP (19.92 vs 17.90 mm Hg, P = 0.045) and a significantly shorter duration of use of topical glaucoma eyedrops (52.62 vs 74.54 mo, P = 0.003), especially for beta-blockers and prostaglandin analogs (Table 1). No other significant differences were noted in terms of preoperative characteristics.

Those on oral ibuprofen were started on average at 2.2 ± 2.3 weeks after trabeculectomy for an average total duration of 4.08 ± 2.28 weeks (Table 2). Supplemental Table (Supplemental Digital Content 1, https://links.lww.com/IJG/A727) further shows the breakdown of patients stratified by the number of postoperative days before ibuprofen administration. A greater proportion of patients prescribed ibuprofen underwent early (<3 mo) bleb needling postoperatively compared with those not prescribed ibuprofen (32.47% vs 12.80%, P < 0.001; Table 2); however, the average number of bleb needlings per person did not differ between the 2 groups at any timepoint [all P <0.05 for early, intermediate (3–6 mo) and late (>6 mo); Table 2]. Similarly, individuals in the ibuprofen group were more likely to have undergone ≥1 subconjunctival injection with 5-FU and dexamethasone (9.09% vs 1.42% for 5-FU; 7.79% vs 3.79% for dexamethasone, P=0.007; Table 2), compared with those not on ibuprofen, although the mean number of injections (1.28 ± 0.61 vs 1.50±0.85, P = 0.452) per person again did not differ between the 2 groups. There was also no significant increase in low IOP events requiring remedial surgery, for example, ocular hypotony, between the 2 groups (1.28% vs 2.83%, P=0.446; Table 2).

TABLE 2 - Comparison of Participant Postintervention Characteristics

	Mean (SD)
Parameter	Without ibuprofen, 211 eyes of 200 patients; n (%)	With ibuprofen, 77 eyes of 73 patients; n (%)	P *
Total duration (wk) of postoperative ibuprofen given	—	4.08 (2.28)	—
No. postoperative months before starting ibuprofen (d)	—	0.56 (0.58)	—
Postoperative IOP
POW1	14.50 (7.10)	14.06 (8.45)	0.687
POW2–3	15.23 (7.11)	15.72 (6.57)	0.601
POM1	14.20 (6.49)	14.10 (6.16)	0.901
POM2	12.37 (6.17)	12.89 (6.03)	0.528
POM3	11.55 (4.37)	12.40 (5.27)	0.236
POM6	11.54 (3.92)	11.61 (3.78)	0.898
POY1	12.27 (3.57)	12.19 (3.53)	0.879
Change in postoperative IOP from preoperative IOP
POW1	−3.46 (8.51)	−5.65 (10.57)	0.108
POW2–3	−2.46 (8.37)	−4.23 (8.86)	0.139
POM1	−3.75 (8.05)	−5.85 (8.30)	0.058
POM2	−5.64 (8.65)	−7.16 (9.63)	0.231
POM3	−6.45 (6.89)	−7.60 (9.16)	0.337
POM6	−6.39 (7.20)	−8.41 (8.37)	0.065
POY1	−6.03 (6.59)	−7.70 (8.29)	0.130
Bleb failure† (>21 mm Hg on 2 consecutive visits)
No	166 (78.67)	62 (80.52)	0.733
Yes	45 (21.33)	15 (19.48)	—
Bleb failure† (>18 mm Hg on 2 consecutive visits)
No	142 (67.30)	57 (74.03)	0.274
Yes	69 (32.70)	20 (25.97)	—
Bleb failure† (>15 mm Hg on 2 consecutive visits)
No	94 (44.55)	38 (49.35)	0.469
Yes	117 (55.45)	39 (50.65)	—
POD1 VA	0.67 (0.67)	0.58 (0.47)	0.170
Change in POD1 VA compared with preoperative VA	0.27 (0.64)	0.11 (0.38)	0.011
Intraoperative complications
No	210 (99.06)	78 (100.00)	NA
Yes	2 (0.94)	0	—
Bleb needling
Total	43 (20.75)	28 (37.18)	<0.001
Fluorouracil	12 (5.69)	0	—
MMC	22 (10.43)	21 (27.27)	—
Dexamethasone	9 (4.27)	7 (9.09)	—
No. bleb needlings (per person)	1.40 (0.65)	1.72 (0.84)	0.077
Early bleb needling (0–<3 mo)
Total	28 (12.80)	25 (32.47)	<0.001
Fluorouracil	7 (3.32)	0	—
MMC	16 (7.58)	19 (24.68)	—
Dexamethasone	4 (1.90)	6 (7.79)	—
No. bleb needlings (per person)	1.21 (0.49)	1.45 (0.73)	0.189
Intermediate bleb needling (3–6 mo), yes
Total	13 (6.16)	7 (9.09)	0.221
Fluorouracil	2 (0.95)	0	—
MMC	6 (2.84)	6 (7.79)	—
Dexamethasone	5 (2.37)	1 (1.30)	—
No. bleb needlings (per person)	1.00 (0.00)	1.14 (0.37)	0.292
Late bleb needling (>6 mo)
Total	14 (6.64)	1 (2.60)	0.308
Fluorouracil	5 (2.37)	0	—
MMC	8 (3.79)	1 (1.30)	—
Dexamethasone	1 (0.47)	1 (1.30)	—
No. bleb needlings (per person)	1.07 (0.26)	1.00 (0.00)	0.315
Subconjunctival injection of antifibrotics
Total	14 (6.64)	14 (18.18)	0.007
Fluorouracil	3 (1.42)	7 (9.09)	—
MMC	3 (1.42)	1 (1.30)	—
Dexamethasone	8 (3.79)	6 (7.79)	—
No. subconjunctival injections of antifibrotics (per person)	1.5 (0.85)	1.28 (0.61)	0.452
Use of glaucoma medications postoperative
No	157 (74.41)	55 (71.43)	0.612
Yes	54 (25.59)	22 (28.57)	—
Uncontrolled IOP requiring repeat laser‡
No	210 (99.53)	77 (100.00)	0.545
Yes	1 (0.47)	0	—
Uncontrolled IOP requiring repeat surgery§
No	210 (99.53)	77 (100.00)	NA
Yes	1 (0.47)	0	—
Low IOP requiring surgery
No	206 (97.17)	77 (98.72)	0.446
Yes	6 (2.83)	1 (1.28)	—
Any intervention∥ done postoperative to lower IOP
No	127 (60.19)	36 (46.75)	0.042
Yes	84 (39.81)	41 (53.25)	—

Bold values indicate P <0.05.

^*Eye-level characteristics: generalized linear models adjusted for possible clustering effect by eye.

^†Excluding postoperative day 1 visit, also includes need for additional surgery and/or laser (trans-scleral cyclophotocoagulation) for uncontrolled intraocular pressure.

^‡Trans-scleral cyclophotocoagulation.

^§Repeat trabeculectomy.

^∥Includes glaucoma medication, repeat laser or repeat surgery.

NA indicates not possible to estimate differences due to null (“0”) value in comparison group.

IOP indicates intraocular pressure; MMC, mitomycin C; POD1, Post-operative day 1; POM1, Post-operative month 1; POW1, Post-operative week 1; POY1, Post-operative year 1; VA, visual acuity.

In multivariable analyses, all patients experienced significant reductions in IOP across all postoperative time points charted. However, the ibuprofen group consistently registered a greater IOP reduction from baseline compared with those not on ibuprofen, reaching significance at postoperative week 1 (mean difference: −2.89 mm Hg, 95% CI: −5.22 to −0.56, P = 0.015) and month 1 (−2.29 mm Hg, 95% CI: −4.53 to −0.05, P = 0.045) timepoints as compared with those not on ibuprofen (Table 3 and depicted graphically in Fig. 1). We further observed that ibuprofen use was significantly associated with lower odds of bleb failure when failure threshold was defined as 2 consecutive IOP readings >18 mm Hg (OR: 0.39, 95% CI: 0.20–0.79, P = 0.009) and >15 mm Hg (odds ratio: 0.52, 95% CI: 0.29–0.94, P = 0.032), but not at the >21 mm Hg threshold (odds ratio: 0.53, 95% CI: 0.25–1.10, P = 0.088), although a nonsignificant reduction in the likelihood of failure was observed (Table 4).

TABLE 3 - Association Between Postsurgical Change in Intraocular Pressure and Postoperative Adjuvant Treatment With Ibuprofen

	Mean (SE)†
IOP change from baseline (mm Hg)	No ibuprofen*	Percentage reduction from baseline‡	Ibuprofen	Percentage reduction from baseline§	Between group difference (95% CI)	P
1 wk	−3.46 (0.62)	19.33	−5.65 (1.23)	28.36	−2.89 (−5.22, −0.56)	0.015
2–3 wk	−2.46 (0.65)	13.74	−4.23 (1.02)	21.23	−1.13 (−3.52, 1.26)	0.355
1 mo	−3.75 (0.58)	20.95	−5.85 (0.94)	29.37	−2.29 (−4.53, −0.05)	0.045
2 mo	−5.64 (0.65)	31.51	−7.16 (1.09)	35.94	−0.97 (−3.55, 1.60)	0.459
3 mo	−6.45 (0.49)	36.03	−7.60 (1.10)	38.15	−1.29 (−3.54, 0.96)	0.261
6 mo	−6.39 (0.51)	35.70	−8.41 (0.98)	42.22	−1.98 (−4.14, 0.17)	0.072
1 y	−6.03 (0.50)	33.69	−7.70 (0.98)	38.65	−1.20 (−3.33, 0.92)	0.266

Bold values indicates P <0.05.

^*Adjusted for age, sex, race, total duration of medication use at baseline, type and number of early bleb needling (<3 mo) with antifibrotic agent used, type of subconjunctival antifibrotic injection agent used, postoperative glaucoma medication prescribed (yes/no), postoperative laser done (yes/no), using stepwise selection.

^†Predicted means from generalized linear mixed models adjusted for potential clustering effect by study eye.

^‡Calculated from mean preoperative IOP of 17.95 mm Hg.

^§Calculated from mean preoperative IOP of 20.00 mm Hg.

All changes from baseline were statistically significant using paired t tests.

Tests for interaction were done for sex and race.

IOP indicates intraocular pressure.

FIGURE 1: Postoperative intraocular pressure reduction of non-ibuprofen group versus ibuprofen group. Adjusted for age, sex, race, total duration of medication use at baseline, bleb needling done (yes/no), subconjunctival antifibrotic injection done (yes/no), postoperative glaucoma medication prescribed (yes/no), postoperative laser done (yes/no), using stepwise selection.

TABLE 4 - Multivariable Adjusted^* Association Between Ibuprofen and Bleb Failure^†

OR (95% CI)
	IOP >21 mm Hg	P	IOP >18 mm Hg	P	IOP >15 mm Hg	P
No ibuprofen	Reference	—	Reference	—	Reference	—
Ibuprofen	0.53 (0.25, 1.10)	0.088	0.39 (0.20, 0.79)	0.009	0.52 (0.29, 0.94)	0.032

Bold values indicates P <0.05.

^*Adjusted for age, sex, race, total duration of preoperative medication use, type of early bleb needling with antifibrotic agent used (<3 mo), type of subconjunctival antifibrotic injection agent used, postoperative glaucoma medication prescribed (yes/no), using stepwise selection.

^†Defined as any 2 consecutive readings above the stated IOP thresholds and/or need for corrective surgery/laser (ie, trans-scleral cyclophotocoagulation) for uncontrolled IOP.

Tests for interaction was done for sex and race.

IOP indicates intraocular pressure; OR, odds ratio.

DISCUSSION

Our study found that the use of adjunctive oral ibuprofen in patients at high risk of early bleb failure posttrabeculectomy resulted in greater levels of IOP reduction and a reduced likelihood of experiencing bleb failure up to a year posttrabeculectomy, compared with patients who were not prescribed ibuprofen. In addition, we did not observe any significant increase in severe adverse events related to low IOP between the 2 groups. Despite the lack of statistical power, our results suggest that starting adjunctive oral ibuprofen early may be a potential low-risk management strategy to decrease the likelihood of bleb failure in high-risk individuals.

Our findings are consistent with early trials of Molteno and Dempster¹¹ implants, which showed that the addition of an oral NSAID (diclofenac) on top of oral prednisone, prescribed for a duration of about 6 weeks to all patients regardless of the risk of bleb failure, produced a synergistic anti-inflammatory effect that greatly reduced bleb inflammation and provided better postoperative IOP control than using oral steroids alone. Similarly, an RCT by Petrov et al¹² showed that patients who have prescribed a combination of topical nepafenac and dexamethasone for 2 weeks before trabeculectomy had a less frequent need for postoperative needling, remedial hypotensive therapy, and a greater likelihood of complete trabeculectomy success in the first postoperative year, compared with patients on topical dexamethasone only. In contrast, Cantor et al¹³ did not find any additional benefit to using topical flurbiprofen on top of topical prednisolone after trabeculectomy in terms of postoperative IOP and visual acuity. We suspect that the differences in outcomes may be because of the differences in type, dosage, and method and timing of administration of the NSAIDs used. Moreover, Cantor’s study was limited by a high attrition rate (~28.6%) over its follow-up period. Larger-scale RCTs evaluating the efficacy of adjunctive oral NSAID administration in improving posttrabeculectomy outcomes are therefore warranted to confirm the validity of our findings.

Patients who have prescribed a course of ibuprofen in our study were deemed to have a higher risk of bleb failure and a greater proinflammatory and scarring profile. Despite the higher risk of bleb failure in the ibuprofen group, which would typically require multiple bleb needling interventions to rescue, the total number of bleb needlings that the 2 groups underwent remained similar over the 1-year follow-up duration, even during the early postoperative period (<3 mo), suggesting that the adjunctive treatment with ibuprofen may play a role in reducing the need for multiple instances of needling in eyes at high risk of early subconjunctival scarring. Moreover, the lack of serious adverse events related to low IOP, such as ocular hypotony consequent to bleb leaks, supports the relatively low risk of ibuprofen administration. Of note, the significantly greater IOP reductions observed in the ibuprofen group at week 1 may be misleading as only ~30% of the ibuprofen group had been prescribed the medication at this time; the greater decrease observed may simply reflect these patients’ higher preoperative IOP levels and, therefore, represent a regression to the mean. In contrast, the reduced risk of bleb failure by absolute IOP criteria suggests that the effect may be real. These results warrant more in-depth RCT studies to validate the efficacy of early adjunctive oral ibuprofen to rescue failing blebs.

The observed synergistic effect of ibuprofen in combination with topical steroids observed in our study may be explained by the different mechanisms of action that steroids and NSAIDs use to block the synthesis of inflammatory mediators.¹⁷ Steroids inhibit the action of phospholipase-A₂ and the subsequent release of arachidonic acid, preventing its metabolism, whereas ibuprofen inhibits the conversion of available arachidonic acid to leukotrienes and prostaglandins as explained.¹⁸ Moreover, although oral NSAIDs are almost always prescribed with gastro-protection (eg, omeprazole), there is no evidence that their anti-inflammatory effects are modulated by this class of gastro-protective drugs.¹⁹ All things considered, the synergistic effect of oral NSAIDs with routinely prescribed posttrabeculectomy topical steroid regimes, together with their relatively benign side-effect profile, further supports their use as adjunctive treatment to safely manage posttrabeculectomy outcomes in routine clinical care.

Strengths of our study include it being an audit of trabeculectomies performed in a real-world clinical setting, with a comprehensive list of patient sociodemographic and clinical characteristics, enabling detailed multivariable statistical models to be constructed. Therefore, not only does our study sample resemble that of real-world clinical populations, but it also reflects the decision-making process that clinicians go through on a day-to-day basis, ensuring the generalizability of results. However, our study also had some limitations. Firstly, our study did not reach sufficient statistical power as the number of recruited patients in our study fell short of the optimal number derived from sample size calculations, which prevented us from conducting further stratified analyses, for example, by the type of glaucoma. Secondly, our study design did not allow us to establish a definitive cause-and-effect relationship between the use of oral ibuprofen and the reduced likelihood of bleb failure, nor could we eliminate the effects of selection bias.²⁰ Thirdly, even though multivariable analyses were conducted in an attempt to adjust for the effects of potential confounders, for example, the number and type of antifibrotic agents used in needling and subconjunctival injections, not all information were readily available, for example, visual fields data, hence resulting in residual confounding that may be unaccounted for. Fourthly, the retrospective nature of this analysis did not permit us to include any photos or utilize a formal bleb grading system, that would have allowed us to assess the effects of oral ibuprofen on bleb morphology. Finally, we were unable to ascertain if patients self-administered ibuprofen or other NSAIDs postoperatively for other conditions given that ibuprofen is readily available as an over-the-counter medication, or whether patients prematurely discontinued the ibuprofen themselves without informing the clinician due to the side effects, which could potentially result in spurious associations. Our findings should hence be interpreted with caution.

CONCLUSION

Early adjunctive oral ibuprofen use in patients at risk of posttrabeculectomy bleb failure is associated with greater postoperative IOP reduction and a lower likelihood of bleb failure, as compared with patients not prescribed ibuprofen. We suggest that starting oral NSAIDs early is a potentially low-risk intervention for clinicians, for managing blebs at high risk of scarring and subsequent failure. Future studies using an RCT design are warranted to explore and confirm our findings.