Difluprednate achieves a similar clinical outcome after trabeculectomy compared to prednisolone acetate at a significantly lower drop frequency

Abstract

Background:

The success of trabeculectomy is highly dependent on postoperative control of inflammation and fibrosis. Prednisolone acetate is the most commonly used topical steroid after ophthalmic surgery. However, non-compliance and adherence problems are constantly thwarting this goal. A topical eye drop regimen that requires fewer drops per day and still achieves the same success rates regarding controlling fibrosis and inflammation is desirable.

Objective:

This study aimed to evaluate whether similar success rates can be achieved with topical difluprednate, comparable to prednisolone acetate, following trabeculectomy.

Design:

A single-center, single-surgeon retrospective chart review.

Methods:

Retrospectively, medical records were evaluated. Inclusion criteria were age ⩾18 years with no upper limit and a diagnosis of open-angle glaucoma. Exclusion criteria were follow-up of <3 months and a history of a concurrent surgery that lowers IOP other than laser trabeculoplasty, cataract surgery, or trabecular meshwork bypass procedures. Success was defined as IOP ⩽ 21 mmHg and ⩾20% reduction below baseline after 1 month, no hypotony (IOP > 5 mmHg), no subsequent glaucoma surgery, and no loss of light perception vision. Primary outcome measures were time to failure and Kaplan–Meier survival, and secondary outcome measures were change in IOP, number of anti-glaucoma medications (AGM), and postoperative interventions and complications.

Results:

In all, 115 eyes were analyzed: 75 eyes in the prednisolone acetate group, and 40 eyes in the difluprednate group. Baseline characteristics and demographics were similar between the groups. IOP and AGM were significantly lowered, with no difference between the two groups. Failure rates varied between 12% and 31% at 1 year in the prednisolone arm, and between 12% and 35% in the difluprednate arm. No differences between the two treatment arms were found regarding survival statistics.

Conclusion:

There was no difference in the success rate between the drug used to treat postoperative inflammation and prevent scarring after trabeculectomy. However, fewer drops per day were necessary in the difluprednate group. Both groups showed no difference in the amount by which IOP and AGM were reduced.

Introduction

Glaucoma is a progressive neurodegenerative disease that leads to visual field loss and irreversible blindness if left untreated or treated insufficiently. Lowering intraocular pressure (IOP) is the only evidence-based treatment.^{1 –4}

Lowering IOP by trabeculectomy (trab) remains the surgical gold standard with remarkable success rates between 57% and 88% for up to 20 years. ⁵ However, IOP control can be compromised by scarring due to inflammation and fibrosis. Two strategies have been routinely used to increase success rates. First, antimetabolites, that is, mitomycin C (MMC) or 5-fluorouracil (5-FU), are used routinely at the time of trabeculectomy surgery to inhibit fibroblast proliferation and to reduce their activity in the first weeks after the procedure. The intraoperative use of antimetabolites has improved success rates remarkably.^{6 –11} Second, topical corticosteroids, mainly prednisolone acetate, are used during the early postoperative period to further reduce inflammation and scarring. ¹²

Currently, conjunctival inflammation, fibrosis, and scarring are not directly measurable in a clinical setting. After trab surgery, all three of these factors result in an increase in IOP over time, which is considered failure. As a result of a trab, the aqueous humor is guided to the newly created outflow pathway from the anterior chamber through the scleral wall to the subconjunctival space. There is a high propensity of subconjunctival fibrous tissue formation during postsurgical healing, which may seal off the transscleral and delimit the area of subconjunctival area of filtration, that is, bleb size. ¹³ Platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF-A), and transforming growth factor beta-1 (TGF-β1) are involved in the early inflammatory phase after surgery. ¹³ TGF-β has a key role in wound healing as it induces the differentiation of fibroblasts to highly contractile myofibroblasts characterized by expression of alpha-smooth muscle actin (α-SMA) and enhanced expression of extracellular matrix proteins (fibronectin, tenascins, and collagens). ¹³ Because these factors of conjunctival inflammation, fibrosis, and scarring cannot be assed noninvasively in a clinical setting, the increase in IOP is commonly used as a surrogate maker for both inflammation and scarring after glaucoma filtration surgery.

Topical steroids are superior in controlling inflammation to non-steroid anti-inflammatory use after trab.^{14 –16} Following a repeat-trab, intensive hourly topical steroid use was not superior to QID use. ¹⁷ The addition of systemic prednisone to topical steroids had no additional benefit.^15,16

Difluprednate (Durezol^® 0.05% ophthalmic emulsion; Alcon Laboratories Inc., Fort Worth, TX, USA) was approved by the U.S. Food and Drug Administration on June 23rd, 2008, for the treatment of inflammation and pain following ocular surgery, and on June 13th, 2012, to treat endogenous anterior uveitis. ¹⁸ Since its approval, it has been proven to be a safe and effective topical anti-inflammatory drug. ¹⁹ In comparison to prednisolone acetate, difluprednate was fluorinated twice (position C-6 and C-9) to enhance the glucocorticoid receptor’s affinity. ²⁰ To further enhance glucocorticoid potency, a butyrate ester was added (position C-17). A third modification was designed to improve corneal penetration.^21,22 Therefore, an acetate group in position C-21 was added, resulting in an ester double bond (C-17 and C-21).^21,22 The ocular formulation was preserved with sorbic acid rather than the more toxic benzalkonium chloride.^23,24 Difluprednate is an emulsion and not a suspension. Thus, the concentration of the drugs in the eye drops is more uniform and less influenced by active shaking prior to usage. ²⁵ For the treatment of endogenous anterior uveitis, Durezol QID was shown to be equally effective to prednisolone acetate eight times per day.^18,26,27 The greater potency and different formulation may provide more predictable treatment outcomes compared to prednisolone acetate. At the moment, prednisolone acetate is still more commonly used, for example, 83% versus 12% after cataract surgery. ²⁸ A recent survey by the American Glaucoma Society shows that almost half of the participating members of the society were either unfamiliar with literature or that there was not enough evidence to support the superiority of difluprednate over prednisolone. ²⁹ There is one report comparing difluprednate with prednisolone acetate for the postoperative anti-inflammatory treatment after combined cataract surgery plus Kahook Dual Blade, but to our knowledge, not for after trab. ³⁰

We hypothesized that, due to higher glucocorticoid potency, higher glucocorticoid receptor affinity, and better corneal penetration, the success rate of difluprednate-treated eyes would be non-inferior even with a less frequent dosing regimen of difluprednate compared to prednisolone acetate. Therefore, retrospective data from subjects who have been treated with either difluprednate or prednisolone acetate after trab with intraoperative administration of MMC have been collected; success rates of both groups have been compared.

Materials and methods

The institutional review board of the Massachusetts Eye & Ear Infirmary (Protocol No. 12-049H) approved all study-related analyses. The analyses were conducted in adherence with the Declaration of Helsinki and the Health Insurance Portability and Accountability Act. A waiver of Consent/HIPAA authorization was granted in accordance with the regulations of the Common Rule and HIPAA Privacy Rule.

Retrospectively, the medical records of all patients who underwent trab with MMC between April 1, 2008, and April 30, 2011, by one surgeon (D.J.R.) were reviewed (retrospective, non-randomized, comparative case series). Billing records and the surgeon’s personal logs for the designated time period identified the cases. After its approval by the Food and Drug Administration in June 2008 to treat inflammation and pain following ocular surgery, there was a shift in the study’s surgeons’ practice pattern to treat more and more patients with difluprednate rather than with prednisolone acetate. Prior to June 2008, all patients were treated topically with prednisolone acetate 1% ophthalmic suspension in the early period after surgery. At the study institution, eyes received incisional filtration surgery (trab with MMC) if the IOP was outside patient’s individualized target range on maximally tolerated medical therapy or if the IOP was controlled, but required the use of three or more anti-glaucoma medications (AGM), or if the patient had such advanced glaucomatous visual field loss whereby it was deemed that the patient was at risk of loss of central fixation by the occurrence of a postoperative IOP spike.

The patient’s baseline data (age at surgery, gender, ethnicity, central corneal thickness, type of glaucoma, prior IOP-lowering surgery, refraction, pachymetry, and lens status) were collected. Patient’s baseline data were further collected once before surgery in addition to 1 ± 1 day, 1 week ±3 day, and 1, 3, 6, 9, 12 months ± 45 days after surgery, and subsequently every 6 months ± 45 days. During each study visit, best-corrected visual acuity (BCVA), IOP, and AGM were collected for statistical analysis. Baseline IOP was calculated as the average measurement value of the two visits immediately before surgery. If only one preoperative visit was available, this was used as the baseline IOP. Humphrey visual field (HVF; Carl Zeiss Meditec; Dublin, CA) data, such as mean deviation (M.D.) and pattern standard deviation (PSD), were collected preoperatively. The HVF was graded according to the Hodapp-Anderson-Parrish scale (HAP scale). ³¹

Inclusion and exclusion criteria

Inclusion criteria were age ⩾18 years with no upper limit and trab with MMC as the initial (i.e., primary) filtration surgery for the management of open-angle glaucoma (primary open-angle glaucoma, pseudoexfoliative glaucoma, pigmentary glaucoma, juvenile open-angle glaucoma, low tension glaucoma) between April 1, 2008 and April 30, 2011. The exclusion criteria were a history of a concurrent surgery that lowers IOP or any other type of glaucoma besides those mentioned above. Patients who had prior laser trabeculoplasty, cataract surgery, or trabecular meshwork bypass procedures were not excluded. In addition, all patients with postoperative data of <3 months were excluded.

Definition of failure and success

As recommended by the World Glaucoma Association, multiple definitions of success were used. ³² Failure was the opposite of the given definition of success. Our first definition of success was modified from the criteria used in the Tube-versus-Trabeculectomy (TVT) study and primary-Tube-versus-Trabeculectomy (pTVT) study, IOP ⩽ 21 mmHg and ⩾20% reduction below baseline.^{33 –38} All definitions of success evaluate success after 1 month rather than after 3 months based on our clinical experience that the IOP has then stabilized from the usual postoperative manipulations and vision has mostly returned to baseline around 1 month rather than at 3 months. The second definition of success was IOP ⩽ 18 mmHg and ⩾30% reduction below baseline; definition 3 was IOP ⩽ 21 mmHg; and definition 4 was IOP ⩽ 18 mmHg. All definitions of failure included the criteria of IOP ⩽ 5 mmHg (“hypotony”) on two consecutive follow-up visits after 1 month, subsequent glaucoma surgery, or loss of light perception vision. Additional glaucoma surgery refers to any surgical (laser or incisional) procedure required to lower the IOP with the exception of laser suture lysis or use of subconjunctival antimetabolites (5-fluorouracil), which are considered common postoperative interventions. The date of failure was the first date out of 2 consecutive examinations on which the definitions of success were no longer fulfilled, starting with month 1 after the procedure. Subsequently, incisional surgeries were considered failures and counted as those for Kaplan–Meier-Survival accordingly. The indication for subsequent glaucoma surgery was made by the surgeon and the patient and was based on one of the following justifications: (1) documented progression of retinal nerve fiber layer (RNFL) thinning on Optical Coherence Tomography scans, (2) progression of scotomas on visual field testing, or (3) an IOP outside a target range, which would lead to progression over time. Failure is the opposite of success.

Surgical techniques—Trabeculectomy

Surgery was performed after a retrobulbar block or under general anesthesia. A conjunctival Tenon’s peritomy was performed superiorly from 11:00 to 1:00 o’clock to create a limbal-based conjunctival flap. A Weck Cell sponge (O.R. Specialties, Milford, MI, USA) soaked in a 0.4 mg/ml solution of MMC was placed underneath the conjunctival and Tenon’s flap for a total of 1.5–3 min exposure time, at the intraoperative discretion of the surgeon after taking into consideration tissue quality. The pledgets were removed, and the area was then copiously irrigated using balanced salt solution (BSS; Alcon, Fort Worth, TX, USA). A rectangular-shaped scleral flap approximately 3.5 mm in length, extending 1.5–3.0 mm posterior from the limbus, was made using a supersharp blade. Interrupted 10-0 nylon sutures were placed and left untied. The flap was lifted, a limbal based corneal tunnel was created using a slit knife, and a Kelly’s Descemet’s punch (0.75 mm bite) was used to create an internal ostium at the clear cornea and carried posteriorly until the edge of the internal ostomy was between 0.75 and 1.0 mm from the edge of posterior border of the scleral flap. Vannas scissors and iris forceps were used to create an iridectomy. The previously prepared 10-0 nylon sutures were then tightened, with the tension of the sutures adjusted to allow for adequate drainage. In addition, a single 10-0 nylon suture was used as a releasable suture at the apex of the flap in some cases, at the surgeon’s discretion. The conjunctival Tenon’s flap was then closed using two interrupted 7-0 polyglactin (Vicryl; Ethicon, Somerville, NJ, USA) running wing sutures (fornix-based) or a running 8-0 polyglactin (limbus-based). An injection of subconjunctival dexamethasone and Ancef (or vancomycin if the patient had a prior allergy to penicillin or cephalosporins) was delivered to the inferior fornix. Tobramycin/dexamethasone (Tobradex; Alcon) ointment was placed on the eye. The eye was patched and shielded.

Medical treatment after surgical interventions

After trab, all eyes were treated either with prednisolone acetate 1% or difluprednate 0.05 % ophthalmic emulsion. Patients were advised to use prednisolone acetate every 1–2 h during the daytime for 2 weeks (i.e., approximately eight times per day), then four times per day for an additional 2 weeks, and then tapered out by one drop per week. Difluprednate was prescribed four times per day for 2 weeks, and then tapered out by one drop per week. In addition, moxifloxacin 0.5% (Vigamox^®; Alcon, Fort Worth, TX, USA) was prescribed four times per day until the Vicryl sutures dissolved, which was usually after week 5.

Postoperative interventions

Laser suture lysis was performed, or releasable sutures were removed in the postoperative period to titrate IOP at the discretion of the surgeon based on the target IOP for the individual patient.

Statistical analysis

All statistical analyses were performed using SPSS software version 21.0 for Microsoft Windows (SPSS Inc., IBM, Chicago, IL, USA). For continuous variables, Student’s t-tests and for categorical variables, Pearson’s Chi-Square tests were used to determine significant differences between the groups. Kaplan–Meier survival statistics of the two groups and the different definitions of failure were computed and compared using the log-rank test. A p value <0.05 was considered to be statistically significant. Continuous variables were expressed as mean ± standard deviation, and categorical variables as absolute numbers and percentages. All visual acuities were converted to the logarithm of the minimum angle of resolution visual acuity (logMAR) with counting fingers = 2.0, hand motions = 3.0, and light perception = 4.0 for statistical analysis.

Results

Billing records and surgeons’ logs identified 146 eyes. Four eyes were excluded for <3 months of follow-up, and 27 eyes were excluded due to a diagnosis that was not part of the inclusion criteria. 115 eyes had sufficient data, met the inclusion and exclusion criteria, and, thus, could be analyzed. Seventy-five eyes were identified in the prednisolone acetate group (control arm) and 40 eyes were identified in the difluprednate group (study arm). None of the eyes were required to be excluded due to prior glaucoma surgery in either of the two groups. Given the fact that difluprednate was approved in June 2008, it took some time for a change in clinical practice to treat patients after trab with difluprednate rather than the established treatment with prednisolone acetate. Thus, sufficient follow-up data in the difluprednate group were limited to 30 months.

Baseline characteristics, demographics, and visual field data

The baseline ophthalmologic characteristics and demographic data of the eyes in both groups were not different for age, gender, ethnicity, laterality, glaucoma diagnosis, best corrected visual acuity (BCVA), baseline refraction, central corneal thickness, baseline cup-to-disc ratio, lens status, previous laser trabeculoplasty, and previous glaucoma surgery (Table 1). The mean defect was higher in the prednisolone acetate arm (−16.1 vs. −11.5 dB, p = 0.023). However, there was no difference in pattern standard deviation or Hodapp-Anderson-Parish score between the two groups (see Table 1). Baseline IOP (22.8 ± 8.7 mmHg vs 21.4 ± 7.3 mmHg, p = 0.368) and AGM (3.3 ± 1.1 vs. 3.3 ± 1.1, p = 0.790) were not different between groups.

Table 1.

Demographic and baseline data of glaucoma patients who were treated with prednisolone or difluprednate after trabeculectomy.

Demographics	Prednisolone acetate group (n = 75)	Difluprednate group (n = 40)	p
Age (years)	64.6 ± 18.4 (60.4 to 68.8)	62.8 ± 11.8 (59.1 to 66.6)	0.584
Gender			0.077
Male	43 (57.3%)	16 (40.0%)
Female	32 (43.7%)	24 (60.0%)
Race			0.685
African-American	10 13.3%)	7 (17.5%)
Asian	5 (6.6%)	3 (7.5%)
Caucasian	53 (70.7%)	29 (72.5%)
Hispanic	5 (6.6%)	1 (2.5%)
Other	2 (2.7%)	0
Eye			0.810
OD	43 (57.3%)	22 (55.0%)
OS	32 (43.7%)	18 (45.0%)
Glaucoma diagnosis			0.205
POAG	41 (54.7%)	25 (62.5%)
PXFG	15 (20.0%)	4 (10.0%)
PDG	3 (4.0%)	0
JOAG	7 (9.3%)	2 (5.0%)
LTG	9 (12.0%)	9 (22.5%)
Baseline BCVA (logMAR)	0.5 ± 0.8 (0.3 to 0.7)	0.3 ± 0.5 (0.1 to 0.4)	0.053
Baseline IOP (mmHG)	22.8 ± 8.7 (20.8 to 24.8)	21.4 ± 7.3 (19.1 to 23.7)	0.368
Baseline AGM	3.3 ± 1.1 (3.0 to 3.5)	3.3 ± 1.1 (3.0 to 3.7)	0.790
Baseline MD (dB)	−16.1 ± 9.5 (−18.4 to −13.7)	−11.5 ± 9.4 (−14.7 to −8.3)	0.023
Baseline PSD (dB)	7.5 ± 3.6 (6.6 to 8.4)	6.8 ± 4.7 (5.2 to 8.5)	0.428
HAP score			0.215
Early glaucoma	15 (20.0%)	10 (25.0%)
Moderate glaucoma	14 (18.7%)	13 (32.5%)
Severe glaucoma	36 (48.0%)	12 (30.0%)
No data available	10 (13.3%)	5 (12.5%)
Baseline refraction (dpt.)
Sphere	–0.67 ± 2.99 (−1.42 to 0.08)	–0.69 ± 1.40 (−1.19 to −0.19)	0.973
Cylinder	−0.87 ± 1.12 (−1.15 to −0.58)	–0.61 ± 0.76 (−0.88 to −0.34)	0.252
CCT (µm)	539.9 ± 41.5 (529.4 to 550.4)	530.3 ± 40.4 (516.8 to 543.7)	0.262
Baseline c/d ratio	0.84 ± 0.17 (0.80 to 0.88)	0.97 ± 0.82 (0.71 to 1.23)	0.199
Lens state			0.287
Phakic	49 (65.3%)	30 (75.0%)
Pseudophakic	26 (34.7%)	10 (25.0%)
Previous LTP			0.603
ALT	7 (9.3%)	5 (12.5%)
SLT	24 (32.0%)	16 (40.0%)
ALT and SLT	5 (6.7%)	1 (2.5%)
None	39 (52.0%)	18 (45.0%)
Previous IOP-lowering surgery			0.724
Trabectome	13 (17.3%)	8 (20.0%)
None	62 (82.7%)	32 (80.0%)
Mean defect (dB)	−16.1 ± 9.5 (n = 65/75)	−11.5 ± 9.7 (n = 35/40)	0.023
Pattern standard deviation (dB)	7.5 ± 3.6 (n = 65/75)	6.8 ± 4.7 (n = 35/40)	0.428

AGM, anti-glaucoma medication; ALT, argon-laser-trabeculoplasty; BCVA, best corrected visual acuity; CACG, chronic angle-closure glaucoma; CCT, central corneal thickness; dB, decibel; HAP score, Hodapp-Anderson-Parrish scale score; IOP, intraocular pressure; LPI, laser peripheral iridotomy; MD, mean defect; NVG, neovascular glaucoma; PDG, pigment dispersion glaucoma; POAG, primary open-angle glaucoma; PSD, pattern standard deviation; PXFG, pseudoexfoliative glaucoma; SLT, selective-laser-trabeculoplasty.

Loss of sample size over time

Both groups had a loss of sample size over time. The loss of sample size was similar for up to 2 years with a loss of sample size at 1, 1.5, 2, and 2.5 year(s) in the prednisolone acetate group of 31%, 20%, 21%, and 32% and the difluprednate group of 28%, 13%, 35%, and 50%, respectively. The loss of sample size can be partially explained by the end of follow-up at the department where the surgery was performed. Though there was a loss of sample size during follow-up in both groups, it affected both groups equally for an extent of approximately 2 years (Table 2). Due to the loss of sample size after 2.5 years, only data up to this time were analyzed.

Table 2.

Comparison of intraocular pressure, anti-glaucoma medication numbers, and best-corrected visual acuity between the prednisolone and the difluprednate groups.

Time-point	Prednisolone acetate group	Difluprednate group	p
	IOP (mmHg)
Baseline	22.8 ± 8.7 (n = 75/75)	21.4 ± 7.3 (n = 40/40)	0.368
Postoperative 1 day	18.0 ± 10.4 (n = 74/74; p = 0.002)	16.6 ± 8.6 (n = 40/40; p = 0.002)	0.452
Postoperative 1 week	13.5 ± 9.4 (n = 75/75; p < 0.001)	13.5 ± 9.3 (n = 40/40; p < 0.001)	0.969
Postoperative 1 month	10.7 ± 5.0 (n = 75/75; p < 0.001)	11.1 ± 7.2 (n = 37/37; p < 0.001)	0.714
Postoperative 3 months	11.1 ± 4.6 (n = 70/70; p < 0.001)	10.0 ± 5.0 (n = 37/38; p < 0.001)	0.248
Postoperative 6 months	10.6 ± 4.1 (n = 65/67; p < 0.001)	9.6 ± 5.1 (n = 35/36; p < 0.001)	0.256
Postoperative 9 months	10.8 ± 4.4 (n = 46/52; p < 0.001)	10.6 ± 5.8 (n = 28/29; p < 0.001)	0.921
Postoperative 1 year	11.0 ± 4.1 (n = 54/60; p < 0.001)	10.7 ± 5.5 (n = 33/35; p < 0.001)	0.764
Postoperative 1.5 years	12.3 ± 6.5 (n = 53/59; p < 0.001)	12.0 ± 6.3 (n = 23/26; p < 0.001)	0.881
Postoperative 2 years	11.4 ± 4.6 (n = 43/51; p < 0.001)	10.6 ± 3.3 (n = 17/20; p < 0.001)	0.515
Postoperative 2.5 years	11.8 ± 5.3 (n = 39/51; p < 0.001)	10.4 ± 4.0 (n = 10/14; p < 0.001)	0.441
	AGM
Baseline	3.3 ± 1.1 (n = 75/75)	3.3 ± 1.1 (n = 40/40)	0.790
Postoperative 1 day	0.0 ± 0.0 (n = 74/74; p < 0.001)	0.0 ± 0.0 (n = 40/40; p < 0.001)	a
Postoperative 1 week	0.0 ± 0.2 (n = 75/75; p < 0.001)	0.1 ± 0.6 (n = 40/40; p < 0.001)	0.370
Postoperative 1 month	0.0± 0.0 (n = 75/75; p < 0.001)	0.1 ± 0.5 (n = 37/37; p < 0.001)	0.043
Postoperative 3 months	0.4 ± 1.0 (n = 70/70; p < 0.001)	0.2 ± 0.8 (n = 37/38; p < 0.001)	0.321
Postoperative 6 months	0.7 ± 1.2 (n = 65/67; p < 0.001)	0.7 ± 1.2 (n = 35/36; p < 0.001)	0.979
Postoperative 9 months	0.6 ± 1.1 (n = 46/52; p < 0.001)	0.6 ± 1.2 (n = 28/29; p < 0.001)	0.891
Postoperative 1 year	0.8 ± 1.2 (n = 54/60; p < 0.001)	0.8 ± 1.6 (n = 33/35; p < 0.001)	0.943
Postoperative 1.5 years	1.0 ± 1.4 (n = 53/59; p < 0.001)	0.8 ± 1.4 (n = 23/26; p < 0.001)	0.463
Postoperative 2 years	0.9 ± 1.3 (n = 43/51; p < 0.001)	0.4 ± 1.1 (n = 17/20; p < 0.001)	0.175
Postoperative 2.5 years	0.8 ± 1.3 (n = 39/51; p < 0.001)	1.0 ± 1.3 (n = 10/14; p = 0.012)	0.699
	BCVA (logMAR)
Baseline	0.5 ± 0.8 (n = 75/N)	0.3 ± 0.5 (n = 40/40)	0.053
Postoperative 1 day	0.8 ± 0.7 (n = 74/74; p < 0.001)	0.7 ± 0.8 (n = 40/40; p < 0.001 )	0.375
Postoperative 1 week	0.8 ± 0.7 (n = 75/75; p < 0.001)	0.6 ± 0.8 (n = 40/40; p = 0.002)	0.433
Postoperative 1 month	0.7 ± 0.9 (n = 75/75; p = 003)	0.4 ± 0.5 (n = 37/37; p < 0.001)	0.016
Postoperative 3 months	0.6 ± 0.8 (n = 70/70; p = 0.015)	0.4 ± 0.7 (n = 37/38; P = 0.002)	0.243
Postoperative 6 months	0.5 ± 0.8 (n = 65/67; p = 0.080)	0.4 ± 0.8 (n = 35/36; p = 0.003)	0.525
Postoperative 9 months	0.4 ± 0.7 (n = 46/52; p = 0.368)	0.5 ± 0.8 (n = 28/29; p = 0.002)	0.713
Postoperative 1 year	0.6 ± 0.8 (n = 54/60; p = 0.094)	0.5 ± 0.8 (n = 33/35; p = 0.001)	0.968
Postoperative 1.5 years	0.5 ± 0.7 (n = 53/59; p = 0.777)	0.5 ± 0.9 (n = 23/26; p = 0.024)	0.895
Postoperative 2 years	0.6 ± 0.8 (n = 43/51; p = 0.051)	0.5 ± 0.9 (n = 17/20; p = 0.205)	0.811
Postoperative 2.5 years	0.7 ± 0.9 (n = 39/51; p = 0.054)	0.1 ± 0.1 (n = 10/14; p = 0.170)	0.064

The numerator of the sample size refers to the number of patients who had not yet failed with data at that specific time point, due to the definition of success 1 (IOP ⩽ 21 mmHg and ⩾20% reduction below baseline). The denominator of the sample size refers to the total number of patients with data at that specific time point. p Values in parentheses compare the values at the different time points with baseline, while the p values in the last column reflect statistical significance between the two groups (i.e., prednisolone acetate group vs difluprednate group).

^aA student’s t-test cannot be computed because the standard deviations of both groups are 0.

AGM, anti-glaucoma medications; BCVA, best corrected visual acuity; IOP, intraocular pressure; LogMAR, logarithm of the minimal angle of resolution.

IOP, anti-glaucoma medications, and visual acuity

In both groups, IOP and AGM were lowered from the first day after the procedure. This change was significant for up to 6 years in the prednisolone acetate group and for up to 3 years for IOP and up to 2.5 years for AGM in the difluprednate group, with no difference between the two groups (with the exception of the 1-month postop time point for AGM).

Early and late postoperative complications

Early (i.e., within 3 months after surgery) and late complications occurred similarly in both groups (p = 0.437 and p = 0.649, see Table 3). 28% of eyes in the prednisolone acetate group and 35% of eyes in the difluprednate group had complications in the early postoperative period. Bleb leakage, hypotony with or without bleb leakage, choroidal effusions, or macula folds have been the most common early complications in both groups. There was only one eye in each group having late complications (Table 3).

Table 3.

Comparison of postoperative complications in the prednisolone acetate and difluprednate group.

Complications, postoperative interventions, and subsequent surgeries	Prednisolone acetate group	Difluprednate group	p
Early complications (within 3 months after surgery)
Bleb leakage	7	6
Bleb leakage and anterior chamber hemorrhage	0	1
Hyphema	0	1
Hypotony	7	2
Hypotony with bleb leakage	2	1
Hypotony with choroidals	3	2
Hypotony with macula folds	1	0
Suprachoroidal hemorrhage	0	1
Cystoid macula edema	1	0
Total no. of patients with early complications	21	14
Late complications (>3 months after surgery)
Pseudophakic bullous keratopathy	1	0
Retinal detachment	0	1
Total no. of patients with late complications	1	1	0.649
Postoperative interventions
Laser suture lysis	23	16
Laser suture lysis and bleb revision	2	0
Bleb revision	3	0
Releasable suture removed	1	0
Releasable suture removed and laser suture lysis	1	0
Total no. of patients with postoperative interventions	30	16	1.000
Surgery
Tube implant (Baervaeldt)	13	3
Transscleral cyclo-photocoagulation	0	0
. . . transscleral-cyclophotocoagulation, endo-cyclophotocoagulation	1	0
. . .combined cataract surgery and trabeculectomy ab interno with the Trabectome	1	0
. . .combined cataract surgery and trabeculectomy ab interno with the Trabectome, tube implant (Baervaeldt)	0	2
. . . combined cataract surgery and trabeculectomy ab interno with the Trabectome and needling	1	0
. . . trabeculectomy ab interno with the Trabectome	2	0
. . . trabeculectomy ab interno with the Trabectome, re-trabeculectomy, endo-cyclophotocoagulation	1	0
Total no. of patients with subsequent surgeries	19	5	0.107

Postoperative interventions

In both groups, laser suture lysis was commonly performed (in 32% of the cases in the prednisolone acetate group and 40% in the difluprednate group, see Table 3). In addition, a releasable suture was removed in two eyes and the bleb was revised in three eyes in the prednisolone acetate group. Laser suture lysis was counted once for each eye, even if done more than once on a single patient.

Subsequent surgery

Subsequent IOP-lowering surgery was performed in 19 eyes (25%) in the prednisolone acetate group and 5 eyes (13%) in the difluprednate group (p = 0.107, see Table 3). A glaucoma drainage device (Baerveldt) was implanted in 13 eyes (17%) in the prednisolone acetate group and in 5 eyes (13%) in the difluprednate group (p > 0.05). Cataract surgery was performed in 2 eyes in both groups during the follow-up (p > 0.05).

Kaplan–Meier survival statistics

Kaplan–Meier survival analysis was used to compare failure rates between the difluprednate (intervention) and the prednisolone (control) arm (Figure 1). The cumulative probability of failure is in Table 4. In addition, various definitions of success have been used. Patients with persistent hypotony, that is, IOP ⩽ 5 mmHg, subsequent surgery to lower IOP, or loss of light perception vision were still classified as treatment failures. However, the upper limit of IOP (⩽21 vs ⩽18 mmHg) and the relative IOP reduction (⩾20% vs ⩾30%) distinguishing treatment success from treatment failure were changed. For definitions 2, 3, and 4, the results are presented in Table 4 and Figure 1(b)–(d). No differences between the two treatment arms were found for any definition of success. Failure rates varied between 12% and 31% at 1 year, between 55% and 70% at 2.5 years in the prednisolone arm, and between 12% and 35% and between 90% and 95% in the difluprednate arm. A censored event in Kaplan–Meier statistics refers to an event (in this study, failure per definition 1–4) that was not observed. This can be that the participant dropped out of the study was lost to follow-up, or simply reached the end of the study without experiencing the event. As a result, the exact survival time is unknown; however, the data are still considered in the analysis by marking them as censored at the time they left the study. Hence, the Kaplan–Meier statistics estimates survival probabilities even in the event of incomplete information.

Figure 1.

Kaplan–Meier survival analysis of the cumulative probability of failure. Log-rank test was performed for failure definitions 1 ((a) p = 0.855), 2 ((b) p = 0.789), 3 ((c) p = 0.667), and 4 ((d) p = 0.885) for trabeculectomy plus MMC treated with prednisolone acetate (solid line) or difluprednate (dashed line).

MMC, mitomycin C.

Table 4.

Kaplan–Meier survival analysis and relative treatment failures at 1, 1.5, 2, and 2.5 years revealed no difference between study arm (difluprednate) and the control arm (prednisolone) for the different definition of failure.

Time after surgery	1 Year		1.5 Years		2 Years		2.5 Years		p Value e
	Prednisolone arm (%)	Difluprednate arm (%)	Prednisolone arm (%)	Difluprednate arm (%)	Prednisolone arm (%)	Difluprednate arm (%)	Prednisolone arm (%)	Difluprednate arm (%)
Failure Definition 1 a	27	25	37	43	40	58	64	90	0.855
Failure Definition 2 b	31	35	45	53	51	68	71	95	0.789
Failure Definition 3 c	11	18	24	43	28	58	55	90	0.667
Failure Definition 4 d	12	20	26	45	31	60	58	90	0.885

^aIOP>21 mmHg, IOP reduction of <20% from baseline, IOP ⩽5 mmHg (hypotony), subsequent glaucoma surgery, or loss of light perception vision.

^bIOP>18 mmHg, IOP reduction of <30% from baseline, IOP ⩽5 mmHg (hypotony), subsequent glaucoma surgery, or loss of light perception vision.

^cIOP>21 mmHg, IOP ⩽5 mmHg (hypotony), subsequent glaucoma surgery, or loss of light perception vision.

^dIOP>18 mmHg, IOP ⩽5 mmHg (hypotony), subsequent glaucoma surgery, or loss of light perception vision.

^eLog-rank test.

Discussion

Glaucoma is a progressive optic neuropathy characterized by structural and functional abnormalities of the optic nerve and visual field. IOP is the only evidence-based modifiable risk factor for disease onset and progression.^{39 –42} Trab remains the surgical gold standard. To achieve high success rates, inflammation and fibrosis must be minimized. Both lead to scarring, which can compromise the efficacy of a trab. Thus, antimetabolites (MMC or 5-FU) are routinely used intraoperatively and postoperatively and topical anti-inflammatory steroids are prescribed in the first 6–8 weeks after a trab.^{6 –12} However, the effect of topical anti-inflammatory steroids is highly dependent on the patient’s adherence to the prescribed drop regimen. Hence, a simple topical drop regimen is desirable, that is, using as few drops per day as possible with satisfying effectiveness.

In this retrospective, comparative case review, failure rates after trabeculectomy were comparable with topical anti-inflammatory prednisolone and difluprednate, with no statistical difference (p > 0.05 in Kaplan–Meier survival statistics for all definitions of failure). The required drop frequency per day in the difluprednate group was approximately half the dosing of the prednisolone acetate group and, thus, significantly lower. In addition, treatment duration was relatively shorter. Such should increase adherence and, hence, could ultimately result in equivalent postoperative anti-inflammatory control after trab. ⁴³ Furthermore, a postoperative anti-inflammatory drop regimen with fewer drops per day could result in potential improved quality of life and better adherence. ⁴⁴ Low adherence of glaucoma patients to their drop regimen is a known problem. Eye drops can result in dry eye and/or pain, which can negatively influence adherence. Thus, a drop regimen that consists of fewer drops per day but with similar efficacy is always welcomed from both the patient’s and treating physician’s point of view. The different drop frequencies of prednisolone and difluprednate have been studied previously after phacoemulsification cataract surgery and after Ahmed valve glaucoma drainage device implantation.^45,46

There is consensus that a glaucoma intervention’s success shall be assessed as a combination of a relative reduction in IOP (percentage) plus an absolute upper limit. ⁴⁷ The most common criteria is based on the TVT study with a 20% reduction and an absolute IOP of ⩽21 mmHg.^{48 –50} Clinically, an IOP of 21 mmHg is usually too high for a glaucoma patient. We added the additional success criteria of a 30% reduction in IOP plus an absolute IOP of ⩽18 mmHg, which is in accordance with Kass et al. ⁴² and Mathew et al. ⁵¹ Thus, we used different definitions of success. The first is adapted from the TVT study and is currently the most widely used in reporting success in glaucoma studies.^{48 –50} However, an IOP of 21 mmHg for glaucoma would not be classified as a clinical success. Hence, we used definition 2 with an upper limit of 18 mmHg, which is probably sufficient for most early and moderate glaucoma cases. The relative IOP reductions, which are often reported and part of the success criteria, are useful for comparing different procedures independent of baseline IOP in these different studies. On the other hand, there is the disadvantage that with a lower baseline IOP, some cases can be classified as failure because a 20% or 30% reduction in IOP was not achieved, regardless of a sufficient absolute IOP value. Thus, we used definitions 3 and 4, which do not account for relative IOP reduction.

The IOP and AGM lowering efficacy of trab of this study are comparable to previously published studies, for example, the pTVT study ⁴⁸ In addition, the TVT study showed similar results in regard to IOP and AGM.^{36 –38} Failure rates of this study at 1 year are comparable to the TVT and pTVT reports. But with loss of sample size at 2.5 years in this study (data aftyer 2.5 years was not reported due to loss of sample size), higher than the reported data from the TVT and pTVT study at 3 years.^37,49

This study’s main limitation includes the flaws inherent in a retrospective case review, including but not limited to non-randomization, loss of sample size, and the lack of regular follow-up data for all study participants. All of such limitations can cause bias. Despite these limitations, retrospective data can elucidate a directional change. The decision to treat eyes after trab with difluprednate was based on the time frame, as the surgeon at one point switched from prednisolone to difluprednate in the treatment of postoperative inflammation after trab. Careful interpretation is necessary due to the loss of sample size over time. The loss of sample size was comparable in both groups up to 2 years. Loss of sample size is a problem even in prospective randomized controlled trials, for example, the Tube Versus Trabeculectomy Study had a loss of sample size of 25% after 3 years.^37,49 And higher numbers in retrospective glaucoma studies are not uncommon, for example, a 50% loss of sample size after 3 years. ⁵² The higher age of glaucoma patients contributes to the loss of sample size. Another potential explanation is that patients with a favorable postoperative outcome tend to be seen by non-specialized general or comprehensive ophthalmologists, while patients with a less favorable outcome tend to stay with the operating glaucoma specialist during the postoperative course. Furthermore, interpretation of the data should be taken carefully, as there is a difference in the number of eyes treated in each group. A potential bias regarding failure and postoperative interventions is possible. The prednisolone acetate group trended toward worse baseline glaucoma, that is, borderline worse baseline vision, significantly worse MD, and more severe glaucoma. This should be considered as a potential confounder. Last, the Kaplan–Meier curves do not meet the assumptions of proportionality. The KM curves of both study groups do not vary by a lot. However, they cross each other at least twice. Therefore, we advise to cautiously interpret the KM statistics.

Similar potency to control inflammation of post-phacoemulsification has been reported for difluprednate and prednisolone.^{53 –55} Furthermore, good results for graft rejection treatment after penetrating keratoplasty with difluprednate have been reported as well. ⁵⁶ For endogenous anterior uveitis, treatment with difluprednate showed superiority in clearing of anterior chamber cells compared to prednisolone.^26,27 However, to the best of our knowledge, this is the first study to compare the outcome of trab treated postoperatively with difluprednate and prednisolone.

In conclusion, our results show that both drugs sufficiently control postoperative inflammation, and, thus, achieve sufficient IOP control and prevent scarring after trab in two-thirds of eyes. However, difluprednate had to be applied significantly less frequently to achieve comparable outcomes; patients received on average 98 drops of difluprednate versus 210 drops of prednisolone acetate during the first weeks after trabeculectomy, which may be an advantage of this drug, especially concerning adherence and quality of life for our glaucoma patients.