Uveitic Macular Edema

Macular edema (ME) is a leading cause of vision loss in both children and adults with uveitis.^1–7 Reported rates of uveitic ME (UME) from tertiary referral clinics have ranged from roughly 20% to 30%,^8–13 with the vast majority of cases occurring in the setting of intermediate, posterior, or panuveitis––although ME occurs occasionally in eyes with anterior uveitis, especially when inflammation spills over to involve the vitreous. Tomkins-Netzer et al. studied a large clinic-based uveitis cohort in the United Kingdom and identified ME as the primary cause of more than 25% of all long-term, moderate-to-severe vision loss––defined in their study as a best-corrected visual acuity (BCVA) of 20/50 or worse at two years.¹⁴ Similarly, the Multicenter Uveitis Steroid Treatment (MUST) Trial found that increased central retinal thickness (CRT) on optical coherence tomography (OCT) was the strongest predictor of decreased vision at presentation (median, 20/56) and continued to portend poor vision if increased CRT persisted at two years.¹⁵ More recently, the MUST Trial Research Group reported that among the 148 eyes of 117 subjects with UME in the study, CRT improved in 71% and resolved in 60% of eyes following two years of therapy. These improvements were observed whether subjects received systemic immunosuppression alone or an intravitreal fluocinolone acetonide implant therapy (Retisert®); although the implant eyes experienced higher rates of cataract and corticosteroid-induced ocular hypertension and glaucoma. Moreover, among eyes with ME at baseline in the MUST Trial, 72% were found to have cystoid and 28% diffuse ME––a distinction based on the presence or absence, respectively, of cystoid spaces on OCT. The authors found no statistically significant differences in increased vision or improvement or resolution of UME in eyes with cystoid verses diffuse UME.¹⁶

A number of treatment options are available for UME.^1–7,17,18 When inflammation is mild, a reasonable first approach might include the use a potent, topical corticosteroid––such as difluprednate 0.05%, prednisolone acetate 1%, or dexamethasone phosphate 0.1%, with or without a topical non-steroidal anti-inflammatory drug (NSAID). A series of monthly intravitreal (IVT) anti-vascular endothelial growth factor (VEGF) injections can also be used when the intraocular inflammation is minimal or largely controlled. For eyes that are incompletely responsive to topical or IVT therapy, or with moderately to severely active inflammation, systemic, periocular, or intravitreal (IVT) corticosteroids may be required. Biologic agents, such as tumor necrosis factor (TNF) inhibitors, interferon, and tocilizumab, can also be used. Finally, a clearing pars plana victrecomy may be considered for eyes that are unresponsive to medical therapy––especially when an epiretinal membrane is present and concomitant removal of that membrane may result in further restoration of vision. Six original articles^19-24 in this issue of Ocular Immunology & Inflammation address various aspects of the management of UME.

Lejoyeux et al.¹⁹ performed a retrospective analysis of patients with refractory UME treated with either adalimumab (ADA), 40 mg every other week (n = 12), or infliximab (IFX), 5 mg/kg every six to eight weeks (n = 13), at a university referral center in Paris, France, over a 10-year period beginning in 2006. Important study limitations acknowledged by the authors included the relatively small sample size; the retrospective patient identification and data collection; and notable baseline imbalances suggesting that subjects in the IFX arm may have had more severe inflammation. Identified imbalances in IFX-treated subjects included worse median baseline best-corrected visual acuity (BCVA; 20/56 vs. 20/28; p = 0.0078), and trends toward higher median daily corticosteroid dosage (12.5 vs. 10 mg/day; p = 0.035), longer median duration of treatment (18 vs. 12 months; p = 0.67), and higher median baseline central foveal thickness (CFT) as measured on optical coherence tomography (OCT; 469 vs. 382 microns; p = 0.43). These potential sources of bias and confounding acknowledged,²⁵ the authors reported that the proportion of eyes with a decrease in CFT of 50 microns or more following both 6 months––the primary outcome measure (ADA = 50% vs. IFX = 62%)––and 12 months (ADA = 50% vs. IFX = 56%) of treatment tended to be similar in the two subgroups. Changes from baseline to 6-month BCVA and corticosteroid-sparing effect in ADA and IFX arms were small and similarly not statistically significant. Treatment was discontinued in two subjects in each arm due to adverse events. The authors concluded that many, although not all, patients with refractory UME will improve on anti-TNF therapy. They cited several previously published reports in support of this conclusion.

Mackensen et al.²⁰ performed a randomized, controlled study to evaluate the efficacy and tolerability of adding ADA, 40 mg every other week, to existing low-dose corticosteroid plus conventional immunosuppressive therapy in 25 patients with refractory uveitis seen over a 10-year period at university centers in Heidelberg and Munster, Germany. All subjects were given 1 mg/kg of corticosteroid followed by a fixed tapering regimen after randomization. Outcome measures were assessed at 3 months. Important limitations in study design acknowledged by the authors included the relatively small sample size, the potential for referral bias, and modest imbalances in both the anatomic location of the inflammation and the underlying systemic conditions. Three-line improvement in BCVA, the primary outcome, was notably better in those who received ADA (60.0% vs. 13.3%; p = 0.022). Administration of ADA was also associated with statistically significant improvements in anterior chamber inflammation (p = 0.023), vitreous haze (p = 0.031), and CRT (p = 0.039). While no subjects discontinued ADA due to side effects, both infections (16.7% vs. 6.7%) and gastrointestinal events (14.8% vs. 6.6%) were more common in those who received ADA. The authors concluded that the addition of ADA to an existing regimen of low-dose corticosteroids plus a single, conventional immunosuppressive agent can offer added benefit to patients with refractory UME––a conclusion supported by the large, randomized VISUAL-1²⁶ and VISUAL-2²⁷ ADA registration trials. The VISUAL-1 study, in particular, demonstrated that the addition of ADA decreased both the occurrence and, when present, worsening of UME.

Eser Ozturk et al.²¹ described the use of tocilizumab (TCZ), a fully humanized monoclonal antibody directed against both soluble and membrane-bound interleukin-6 (IL-6) receptor, in five subjects with Behçet uveitis who failed two to five years of treatment with conventional immunosuppression, interferon alpha, and anti-TNF therapy seen at a university center in Istanbul, Turkey. Drug was administered at a dose of 8 mg/kg as monthly intravenous infusion. Clinical remission was achieved in 4 of the 5 subjects following a single infusion, and after the 5^th infusion in one subject. No subjects experienced recurrences while on TCZ, and none discontinued TCZ due to drug-related adverse events. Visual acuity remained stable or improved in all eyes and mean measures of anterior chamber cell and flare, vitreous haze, fluorescein angiography score, and CRT improved over the course of therapy. The authors noted specifically that CRT at last visit was less than 300 microns in all eyes except one, which had an epiretinal membrane that may have limited that particular eye’s ability to respond. Despite the small sample size and the retrospective nature and variable follow-up of the study, the authors concluded that TCZ may be an effective approach to the management of both uveitis and UME in patients who fail more conventional therapies. They cited several previously published reports that support such therapeutic benefit, especially for refractory UME.

Tugal-Tutkan et al.²² described the results of gevokizumab (GVK), a recombinant humanized monoclonal antibody directed against IL-1β, compared to placebo as added to standard of care in a prospective, randomized, double-masked, phase 3 study (EYEGUARD B) of 83 subjects with Behçet disease uveitis (BDU; 40 on GVK vs. 43 on placebo). Previous phase 1/2 studies of GVK in patients with active BDU demonstrated good tolerability and rapid and durable therapeutic responses.^28,29 Gevokizumab was administered subcutaneously every 4 weeks. The primary endpoint of the phase 3 study was an event-driven reduction in time to first ocular exacerbation across the treatment arms. While the GVK arm showed a trend toward decreased time to first recurrence as compared to placebo, this difference did not achieve statistical significance (hazard ratio 0.85; 95% CI 0.41–1.77). Similarly, the rate of two or more line BCVA loss was slightly lower in the GVK arm (10.3% vs. 26.8%; p = 0.035), and there were trends toward decreased emergent rates of retinal infiltrates (29.3% vs. 21.1%), retinal vasculitis (26.8% vs. 10.3%), ME (18.0% vs. 10.3%), hypopyon formation (9.8% vs. 2.6%), papillitis (2.4% vs. 0.0%), and neovascularization of the disk or retina (9.8% vs. 5.2%), but none of these differences achieved statistical significance. Adverse events were common, having occurred in 92.7% of GVK and 93.0% of placebo treated subjects, but were generally balanced across the two arms. Adverse events judged to be potentially drug-related were much less common, having occurred in 17.1% of GVK and 18.6% of placebo treated subjects. Five events of drug hypersensitivity that were considered drug-related occurred in the GVK arm. The authors concluded that while GVK failed to produce a statistically significant decrease in time to first ocular exacerbation, secondary outcome data suggested a modest benefit to BCVA and trends toward less frequent occurrences of specific signs of BDU reactivation, including UME, and improved corticosteroid sparing during the 6 months of treatment, supporting a role for IL-1β inhibition in the management of BDU. The development of GVK for non-infectious uveitis, initially pursued by Xoma, Inc, has been discontinued.

Steeples et al.²³ retrospectively studied the use of intraviteal (IVT) alcohol-free triamcinolone acetonide suspension, 40mg/ml (Triesence®), in 44 eyes of 40 subjects who received a total of 66 injections for UME at a university center in Manchester, United Kingdom, between February, 2013 and April, 2015. All subjects were followed for a minimum of 6 months. All eyes had UME and seven (16%) had both UME and active, severe posterior uveitis refractory to high-dose corticosteroid therapy. The cause of uveitis was either undetermined (47.5%) or due to sarcoidosis (17.5%) in approximately two-thirds of eyes. The most common anatomical classifications were panuveitis (35%) and intermediate uveitis (35%), followed by posterior (12.5%) and anterior (12.5%) uveitis. Five percent of eyes had sclerouveitis. Fifty-four percent of eyes were pseudophakic and 7% were aphakic. Eight of 44 eyes (18.2%) had a history of ocular hypertension and two (4.5%) had glaucoma. Over two-thirds of patients (67.5%) were on systemic prednisolone, one or more non-corticosteroid immunosuppressive agents, or both. Thirty eyes (68.2%) received a single IVT injection, eight (18.2%) received two, and six (13.6%) received three. The IVT dose was 2 mg in 31 eyes (70.5%), 4 mg in 12 (27.3%) and 5 mg in one (2.3%). Corticosteroid-induced ocular complications were common and included cataract in eight of 17 phakic eyes (47.1%), intraocular pressure (IOP) elevation to or above 21mmHg in 20 eyes (45.4%) and IOP above 30 mmHg in three eyes (6.8%). Elevated IOPs were more common with 4mg (50.0%) verses 2 mg (35.5%) doses. The highest mean IOPs tended to occur between six and eight weeks after injection. Drops were used to manage IOP in 10 eyes (22.7%), but no eyes required glaucoma surgery. CRT improved on average from 489 ± 150 microns at baseline, to 309 ± 111 microns at weeks 2–4, after which the mean CRT tended to state more or less flat through week 24. Mean BCVA tended to improve over eight weeks, after which it too tended to remain flat through week 24. No differences in mean responses were noted between eyes treated with 2 mg vs. 4 mg. The authors concluded that while cataract and medically manageable IOP elevations were common, IVT alcohol-free triamcinolone acetonide suspension could be used to effectively reduce CRT and improve vision in most eyes with non-infectious UME with or without active inflammation. The authors noted the small sample size, particular the 4 mg subgroup; the possible introduction of biases related to the retrospective and single center nature of their study; and added the potential for confounding in their outcomes analyses due to co-existing contributors to vision loss, such as cataract and epiretinal membrane formation.

Ratra et al.²⁴ retrospectively assessed the tolerability and efficacy of the dexamethasone intravitreal implant (OZURDEX®) for the retreatment of refractory UME in 42 eyes of 34 patients seen at a uveitis referral clinic in Chennai, India, between April, 2011, and December, 2015. A total of 56 implants were placed. Five subjects were children 16 years of age or less. The median duration of ME prior to implantation was 24 months (range 1–60 months). Follow-up varied from 6 to 53 months (mean 19.2 ± 2.2 months). Comparing presentation to last recorded visit, the duration of which varied by subject, mean vision improved from logMAR 0.48 ± 0.06 (~ 20/60) to logMAR 0.34 ± 0.1 (~ 20/44; p = 0.043) and mean CRT decreased from 472.2 ± 35 to 274.7 ± 60.6 microns (p = 0.003). Most improvement in vision and CRT occurred within 3 months. In addition, the rate of dependency on oral corticosteroids and systemic immunosuppression decreased from 97.6% to 59.5% and from 66.7% to 50.0%, respectively. A second implant was required to 11 eyes (26.2%) after a mean of 16.8 ± 2.1 months (range 12.6–20.9 months) and a third implant in three eyes (7.1%). Similar outcomes were observed in the small number of children who received implants. Implantation produced an increase in IOP equal to or greater than 21 mmHg in seven eyes (16.6%), one of which required trabeculectomy. Cataract developed in six eyes (14.3%). The authors concluded that patients with refractory UME can show rapid and durable therapeutic benefit from the dexamethasone intravitreal implant with acceptable rates of ocular hypertension and cataract, both of which were manageable. As above, they noted limitations related to the retrospective, single center nature of their study cohort.

Taken together, these studies highlight the broad range of treatment options available for the management of UME. While regional or systemic corticosteroids are generally required initially to treat ME in the setting of active inflammation, anti-VEGF and systemic immunosuppressive agents, both conventional and biologic, can be useful as adjunctive therapies for UME that is either refractory or incompletely responsive to corticosteroids. While many of these same approaches may be used to treat ME in eyes with infectious uveitis, long-term corticosteroid and non-corticosteroid immunosuppressive therapies need to be used with caution in such settings, should be considered only after the active infection has been controlled, and often require long-term co-administration of anti-microbial therapy to minimize the risk of recurrence.

Additional information

Funding

This work was supported by The Pacific Vision Foundation (ETC) and The San Francisco Retina Foundation (ETC).