Sustained-release Corticosteroids for Uveitis

Corticosteroids constitute both first-line and important adjunctive therapy in patients with non-infectious uveitis.1^,2 Three sustained-release corticosteroid dosage forms are currently available for the treatment of uveitis (Table 1),3–5 including a biodegradable triamcinolone acetonide (TA) injectable suspension (Triesence®, 40 mg/mL; Alcon Laboratories, Inc.); a biodegradable dexamethasone (DEX) intravitreal implant (Ozurdex®; 0.7 mg, Allergan, Inc.); and an anchored non-degradable fluocinolone acetonide (FA) intravitreal implant (Retisert®, 0.59 mg, Bausch and Lomb, Inc.). An unanchored non-degradable FA intravitreal implant (Medidur®, 0.19 mg; pSivida Corp.) is currently in phase 3 trials for the treatment of posterior uveitis and, with favorable results, will be submitted for review by the Food and Drug Administration (FDA) in the first half of 2017.6 All sustained-release corticosteroid products are contraindicated in eyes with active infection, or with glaucoma or a history of corticosteroid-induced ocular hypertension.

TABLE 1. Summary of commercially available sustained-release corticosteroids in uveitis.

Product	Delivery/placement	Durability	Cataract	Ocular hypertension/ filtering surgery	Other (uncommon)
Biodegradable triamcinolone acetonide injectable suspension (Triesence®, 40 mg/mL; Alcon Laboratories, Inc.) – 4 mg dose	27-gauge needle Pars plana injection Clinic or procedure room	2–4 months	∼33% at 12 months	∼one-third at 12 months 1–2% require filtering surgery	Retinal detachment Endophthalmitis Pseudoendophthalmitis
Biodegradable dexamethasone intravitreal implant (Ozurdex®; 0.7 mg, Allergan, Inc.)	22-gauge pre-filed auto-injector Pars plana injection Clinic or procedure room	4–6 months	2–3%	∼15–25% at 12 months Filtering surgery uncommon	Retinal detachment Endophthalmitis Anterior chamber migration
Anchored, non-degradable fluocinolone acetonide intravitreal implant (Retisert®, 0.59 mg, Bausch and Lomb, Inc.)	3.5 mm pars plana incision Surgically anchored to a suture strut Operating room	<30 months	>90% at 36 months	>two-thirds at 36 months Up to 45% require filtering surgery	Retinal detachment Endophthalmitis Hypotony Dislocation/dissociation (late, or at surgery)

All sustained-release corticosteroid products are contraindicated in eyes with active infection, or with glaucoma or a history of corticosteroid-induced ocular hypertension. An unanchored non-degradable FA intravitreal implant (Medidur®, 0.19 mg; pSivida Corp.) is currently in phase 3 trials for the treatment of posterior uveitis.

Each of the currently available sustained-release corticosteroid dosage forms has a unique product profile and, consequently, indicated and preferred usages. The most important differentiating features are ease of delivery, durability of effect, and rate of adverse events—most notably cataract and elevated intraocular pressure (IOP). In broad terms, the injectable TA is easily delivered through a standard 27-gauge needle, has the shortest duration of effect at 2–3 months, and is associated with a moderate risk of corticosteroid-induced cataract and ocular hypertension requiring IOP lowering medications—each of which occur in about one-third of eyes injected with 4 mg within 12 months of treatment. About 1–2% of injected eyes will require glaucoma surgery to control the IOP. The DEX implant, by comparison, is delivered with a pre-filled auto-injector through a 22-gauge needle and so requires slightly more skill, lasts twice as long as injectable TA—for approximately 4–6 months, and has the lowest risk of corticosteroid-induced cataract and ocular hypertension requiring IOP lowering medications following a single injection, which occur in 2–3% and about 15–25% of implanted eyes, respectively. The need for glaucoma surgery following placement of a DEX implant is uncommon. In contrast to both injectable TA and the DEX implant, the anchored FA implant is placed surgically, lasts up to 30 months, and has the highest rates of both implantation and drug-associated adverse events, with over 90% of patients requiring cataract surgery, over two-thirds requiring pressure lowering medications, and up to 45% requiring glaucoma surgery within 3 years of implantation. Hypotony or rhegmatogenous retinal detachment may occur in a minority of patients following placement. Four original articles1–4 and a letter5 in this issue of Ocular Immunology & Inflammation report findings related to the use of these three sustained-release corticosteroids in patients with non-infectious uveitis.

Shin and Yu7 enrolled 50 subjects in a single-center, randomized, controlled study of the tolerability and effectiveness of 4 mg/0.1 mL of intravitreal TA as adjunctive therapy to systemic anti-inflammatory treatment for patients with non-infectious macular edema (ME). All subjects had a thickness >300 µm in the central macula, defined as the 1 mm diameter region of the Early Treatment of Diabetic Retinopathy (ETDRS) grid, and/or a thickness >99th percentile normal distribution in any of the four quadrants of the paracentral area annulus, defined as the area of the ring between 1 mm and 3 mm from the foveal center. Subjects were followed up monthly for 6 months and, at each visit, were given a repeat TA or sham injection if the ME continued to meet either the central or paracentral thickness cut-off specified in the inclusion criteria. In addition, systemic anti-inflammatory medicines were increased if anterior chamber and/or vitreous inflammation increased by ≥1 grade, or if best-corrected visual acuity (BCVA) decreased due to inflammation. The criteria used for tapering systemic anti-inflammatory agents were not specified. Overall, demographic and clinical characteristics were well-balanced across the treatment arms at baseline, with a high proportion of patients on systemic prednisolone (88% and 96% in the TA and sham groups, respectively) and all patients on at least one non-corticosteroid immunosuppressive agent—most typically mycophenolate mofetil (56% vs 68% in the TA and sham groups, respectively) or cyclosporine (32% vs 48% in the TA and sham groups, respectively). The most common diagnoses in the TA and sham groups were Behçet uveitis (36% vs 40%, respectively), pars planitis (28% vs 16%, respectively), and idiopathic uveitis – which one is left to presume excludes ‘pars planitis’ (24% vs 28%). While central and paracentral retinal thickness improved in both the treatment and control arms over the course of the study, TA treated eyes showed modestly faster improvement in retina thickness, with approximately 20 µm more mean decrease in thickness during the first half of the study, and clearly quicker reversal of leakage on fluorescein angiography. Differences in the area of leakage and mean central and paracentral retinal thickness were not statistically different during the last 3 months of the study, however, and vision remained unchanged at all visits, both from baseline and across the two treatment arms, at about 70 ETDRS letters (Snellen equivalent 20/40). Moreover, cataract developed in four TA (22%) versus two sham (15%) subjects over the course of the study, whereas IOP was on average 3–6 mmHg higher in the TA treated group during the first 3 months of the study. While details regarding changes in systemic anti-inflammatory therapies over the course of the study were not provided, the authors noted that more patients in the TA arm were able to have their systemic prednisolone tapered to ≤10 mg/day (88% vs 56%). Of note, subjects who fulfilled the injection retreatment criteria for three consecutive visits were diagnosed with refractory ME and were discontinued from the study, and data from their last three visits was not included in the analysis. Given that refractory ME was diagnosed in 11 (44%) TA and 13 (52%) sham-treated subjects, it would appear that the sample size of the study was decreased by nearly 50% over the second half of the trial. The average number of injections in the TA group was 2.2. The authors suggested that adjunctive intravitreal TA may be used to achieve faster resolution of retinal vascular leakage and ME, but provided no comment on the cost–benefit of such an approach in patients with mild to moderate ME, as studied in their subjects.

Burkholder et al.8 reported the practice patterns of 45 respondents of approximately 176 surveyed members of the American Uveitis Society regarding the use of the DEX implant. All respondents completed a uveitis fellowship and committed at least one-third of their time to clinical care and/or research in uveitis. Approximately half had additional subspecialty training in retina. Two-thirds practiced in an academic setting. Most respondents performed the injection in either a procedure (41.0%) or clinic (23.1%) room, with just over one in five (20.5%) preferring an operating room. Over half (57.5%) used subconjunctival lidocaine as their primary anesthetic. The most frequent indication for injection was ME (76.5%), whereas the most commonly associated diagnosis was pars planitis (24.7%), followed by multifocal choroiditis (18.8%), birdshot chorioretinitis (18.8%), and sarcoid-associated uveitis (17.6%). The most commonly cited contraindications to injection were aphakia (62.2%), glaucoma requiring >2 pressure lowering medications (54.0%), a history of corticosteroid-induced ocular hypertension (32.4%), and the presence of an anterior chamber intraocular lens (32.4%). Over one-third (37.7%) of respondents were of the opinion that the DEX implant was effective for longer, whereas nearly three-quarters (71.4%) were of the opinion that the DEX implant was less likely to cause corticosteroid-induced ocular hypertension when compared with an anchored FA implant. Nearly half of all respondents (46.5%) preferred either a trial with at least one DEX implant prior to placing an anchored FA implant, or using multiple DEX implants in place of an FA implant (48.8%). Nearly two-thirds (64.3%) responded that they did fewer anchored FA implants since the approval of DEX implants. The authors noted the wide variability in practice patterns and the clearly perceived role for both the DEX and FA implants in treating patients with chronic non-infectious uveitis.

Coskun et al.9 investigated the use of adjunctive DEX implants in 17 eyes of 12 patients with active, refractory Behçet posterior uveitis seen over an 18-month period in their referral clinics in Turkey. Systemic corticosteroids were tapered and discontinued within 1 month of receiving the DEX implant, and patients were followed for a total of 12 months. By month 1, improvements occurred in BCVA, central macular thickness, and vitreous haze score. While each of these beneficial effects was durable through month 12, recurrence occurred in two eyes (11.8%) at month 3; in nine eyes (52.9%) at month 6; and in three eyes (17.6%) at month 12 (mean 6.9 months). While IOP rose above 20 mmHg in three eyes (17.6%) at the month 1 visit, the pressures were readily controlled with topical IOP, lowering medications in all patients. Posterior subcapsular opacities developed in four eyes (23.5%), but none required cataract surgery. The authors concluded that a single DEX implant can provide well-tolerated and effective adjunctive control of active, posterior uveitis in patients with Behçet disease that is otherwise incompletely controlled with systemic immunosuppressive therapy. The beneficial effects of these adjunctive DEX implants lasted for an average of approximately 6 months.

Freitas-Neto et al.10 reviewed the use and reasons for removal of anchored FA implants in 165 eyes of 116 patients over a 9-year period in a uveitis referral center in the USA. A total of 187 FA implants were placed. Seventeen eyes (10.3%) received multiple implants, at a median time of just over 4.5 years. The most common diagnosis was idiopathic panuveitis (35.2%), followed by birdshot chorioretinitis (11.2%), pars planitis (10.3%), and sarcoid associated uveitis (9.5%). All implanted eyes either had or developed cataract requiring surgery. Eight eyes (4.8%) required implant removal, reasons for which included spontaneous dislocation in five eyes (62.5%), and one eye each (12.5%) with dislocation during vitrectomy for epiretinal membrane removal, removal for uncontrolled glaucoma, and removal for uncontrolled hypotony. The authors concluded that the rates of both dislocation and required removal of FA implants were low.

Chang et al.11 described two cases of spontaneous dislocation into the anterior chamber of the medication reservoir of previously placed anchored FA implants. Time from implantation to dislocation was 6.5 years in one eye and 6 years in the second. Both eyes were aphakic and had mild to moderate anterior chamber inflammation with localized corneal edema in the area of reservoir-endothelial contact. In both instances, the reservoirs were retrieved using pars plana vitrectomy and removed from the eye through a corneal incision. The authors cited reports by others of similarly spontaneous dissociations 3–6 years following implantation.12^,13 Itty et al. recently reported the experience at two centers, Duke University and Texas Retina Associates, with 407 FA implants placed in 224 patients over a 14-year period.14 A total of 17 of 407 (4.2%) drug pellets spontaneously dislocated from the suture strut at an average post-implantation time of between 6 and 7 years (mean 77.4 months; range 33–132 months). The spontaneous dislocation rate was roughly 20-fold higher with implants that had been placed for ≥5 years. The authors identified an additional 11 cases (2.7%) during either exchange or removal in which there was pellet-strut dissociation, but the pellet had not separated or migrated posteriorly. Nicholson et al. identified dissociation without dislocation at the time of exchange or removal in two out of five cases (40.7%).15

Together, these studies support previously reported randomized clinic trials demonstrating that sustained-release corticosteroids provide effective adjunctive therapy for patients with active, non-infectious intermediate, posterior, and panuveitis.3–5 Side-effects vary, however, with the risk of subsequent cataract and ocular hypertension being lowest following placement of the DEX implant, moderate following injectable TA,16 and highest following surgical placement of an FA implant. A cost-effectiveness analysis of the FA implant versus systemic therapy for patients with non-infectious intermediate, posterior, and panuveitis from the Multicenter Uveitis Steroid Treatment (MUST) Trial Research Group found the incremental cost-effectiveness ratio for the FA implant to be most favorable in patients with unilateral disease.17

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Supported in part by The Pacific Vision Foundation (ETC) and The San Francisco Retina Foundation (ETC).