While topical non-steroidal anti-inflammatory drugs (NSAIDs) have proven very useful in pseudophakic ME, studies have shown that there is limited to no role for topical NSAID monotherapy in the treatment of inflammatory ME. This class of medications may have a weak synergistic effect when used in conjunction with other local therapies such as intravitreal triamcinolone acetate (IVTA) or intravitreal anti-VEGF, prolonging the effect of CSMT reduction [13]. There is a higher concentration NSAID, indomethacin 0.5 %, which when administered 4 times daily, has shown some early promise in uveitic ME [14].

Topical prednisolone 1 % has shown more success than topical NSAIDs in decreasing vascular permeability and inflammatory-mediated reactions. However, its poor penetration to the posterior segment still limits its utility and the general principle is that topical corticosteroids may be used primarily for mild ME. There is a higher chance of achieving sufficient drug levels around the macula in pseudophakic or aphakic patients

More recently, difluprednate, a more potent topical corticosteroid, has become available. In the pediatric population, its use has been associated with an improvement in uveitic ME in 78 % (7/9) of studied eyes and a mean decrease in CSMT of 192 µm [15]. Because difluprednate is more potent, cataract and intraocular pressure (IOP) rise can occur earlier and be more severe than with topical prednisolone. Close monitoring for these side effects is important.

Periocular corticosteroid is often effective for treating uveitic ME and is a common first-line therapy. Triamcinolone acetate 40 mg injected into the posterior sub-Tenon’s space or transeptally into the inferior orbital floor is one commonly used injectable steroid. The goal of periocular steroid injection is to bring the medication as immediately adjacent to the site of inflammation as possible. Hence, some have suggested that delivering the medication in the sub-Tenon’s location may be more effective than orbital floor injection, but this has not been definitively proven. The sub-Tenon’s injection is usually delivered using the Smith and Nozik method with lid retraction and drug delivery with a 16 mm 25-gauge needle into the posterior superotemporal space with the bevel facing the globe. A trans-septal injection is delivered at the lateral third of the inferior orbital rim with a 25- or 27-gauge needle [16]. A study looked at the location of the deposited steroid injection using ultrasound scanning immediately before and after sub-Tenon’s corticosteroid injection [17]. It found that a superotemporal depot is more successful than an inferotemporal depot in delivering the medication to the macular area.

Leder et al. [18] reviewed the results of their experience over a decade for 156 eyes and found that 53 % of eyes treated with a single periocular corticosteroid injection had complete clinical resolution of ME within one month of injection. An additional 22 % of eyes needed repeated periocular injections in one month intervals to achieve resolution. In most depot cases, removal after instillation is very difficult. Hence, close follow-up for IOP control and cataract development is essential. Single injections have limited rates of complications, but with repeated applications these side effects will increase. There are also side effects specific to periocular delivery methods. Trans-septal injections may cause adipose tissue prolapse through the septum, particularly if multiple injections are given, while superior sub-Tenon’s injections may lead to ptosis.

IVTA has been reported to be non-inferior if not superior to periocular corticosteroid use [19]. It is commonly used in cases of ME that are incompletely responsive to periocular corticosteroids. As with periocular corticosteroid injections, it can used as a fast-acting treatment for ME while waiting for longer acting systemic medication to take effect. A single injection of 4 mg/0.1 ml IVTA has clinical efficacy for approximately 2–6 months in non-vitrectomized eyes [20]. In some cases, just one IVTA can result in ME resolution without recurrence, but control of the underlying uveitis inflammation is imperative for this result. One large series in the literature followed 65 eyes over 8 months and detected a 83 % response rate to 4 mg of IVTA with significant visual improvement with 51 % of patients gaining >2 Snellen lines [11]. Half-dose IVTA (2 mg/0.05 ml) produced an 80 % resolution of ME in a 29-patient cohort, with 39 % of eyes gaining >2 Snellen lines [21]. Studies in diabetic patients have shown 4 mg IVTA to cause no additional retinotoxic effect [22]. However, benzyl alcohol at concentrations slightly higher than that which is present in some commercially available triamcinolone preparations caused histological damage to outer retinal structures in rabbits [23]. Therefore, removal of preservatives from a preservative-containing corticosteroid solution is recommendable. Currently, the preservative-free formulation of triamcinolone acetate (Triescence) is the agent of choice due to its design specifically for intraocular use [2]. Intravitreal injections have the associated risks of endophthalmitis (0.5–0.87 %), cataracts leading to cataract surgery (15–20 % of elderly patients within 1 year of injection), IOP spikes in about 30–40 % of eyes leading to surgical glaucoma intervention in about 1–2 % of cases and other retinal complications including retinal detachment [24, 25].

Fluocinolone acetonide (Retisert) and dexamethasone (Ozurdex) intravitreal implants are both sustained-release corticosteroid devices. Both were designed in an attempt to create longer lasting local treatments for uveitis with a possibility of reducing dependency on systemic corticosteroids and immunosuppressive agents. They both also show efficacy in the treatment of uveitic ME. The 0.59 mg fluocinolone implant has been approved since 2005 by the FDA for the treatment of noninfectious intermediate, posterior and panuveitis. It releases fluocinolone acetonide at an initial rate of 0.6 µg/day in the first month, then decreasing to a steady state of 0.3–0.4 µg/day for approximately 3 years thereafter. This implant requires surgical placement in the operating room. While the main indication for fluocinolone implantation is to control uveitic inflammation, the implant is also efficacious in controlling the associated ME.

The implant was investigated for the treatment of uveitis and ME in a prospective randomized study of 278 patients. In the study, not only did the implants significantly reduce uveitis recurrence and decrease need for adjunctive treatment, of the 100 patients with uveitic ME, a quarter of eyes had a >3 line of sustained visual acuity improvement at 34 weeks [26]. The great efficacy of the fluocinolone implant is accompanied by significant rates of corticosteroid-related side effects. At three years after implantation, greater than 90 % of implanted phakic eyes will require cataract surgery [27] and a >10 mmHg IOP rise from baseline occurs in more than 70 % of eyes [28]. IOP-lowering surgery will be necessary in at least of third of eyes. In addition, there is also a small risk of retinal complications including detachment with the placement of the fluocinolone implant [29, 30]. In summary, in cases of uveitic ME where there is ongoing underlying uveitic inflammation, the fluocinolone implant is very likely to result in concurrent improvement or complete resolution of the ME but with a high likelihood of IOP rise and cataract formation.

The dexamethasone 0.7 mg implant is also approved for the treatment of noninfectious intermediate, posterior and panuveitis. Its duration of effect is 4–6 months. It can be injected intravitreally in the clinic, obviating the need to go to the operating room. With regards to ME, the dexamethasone intravitreal implant has been shown to significantly decrease CSMT in patients with uveitic ME, including pediatric patients and those previously refractory to treatment with IVTA, periocular corticosteroids or anti-VEGF [31, 32]. The improvement in visual acuity, which is due in large part to resolution of ME in these patients, has been shown to take effect with one single injection and to manifest in a best-corrected visual acuity (BCVA) gain of >2 lines in more than half of patients at 3 months and in more than a quarter of eyes at 2 years [32, 33]. With repeated dexamethasone intravitreal implant injections, the rates of cataract development and ocular hypertension were approximately 10 and 20 %, respectively, in one series [34]. The dexamethasone implant is particularly useful in treating ME in vitrectomized eyes where the duration of directly injected corticosteroid can be very short.

Bevacizumab (Avastin) and ranibizumab (Lucentis), recombinant humanized monoclonal antibodies directed against VEGF, have both been used in the treatment of uveitic ME. It is believed that anti-VEGF drugs may inhibit the breakdown of the blood–retinal barrier and hence decrease vascular leakage and fluid accumulation leading to ME [2]. Ranibizumab 0.5 mg intravitreal injections can improve vision and reduce ME. In a small group of seven patients with refractory ME who failed corticosteroid treatment, all patients receiving intravitreal ranibizumab demonstrated improvement in vision by one month [35]. The mean gain of vision was 13 letters at the 6-month time point with monthly injections administered on an as needed basis. Similarly, many centers have studied the use of bevacizumab (1.25 mg or 2.5 mg) for the same purposes due to its wide availability and better cost competitiveness [36]. Single or repeated injections of intravitreal bevacizumab lead to a reduction in CSMT and improvement in vision of >2 Snellen lines in approximately 40–50 % of patients [36, 37]. Intravitreal anti-VEGF agents have the advantage over corticosteroids of causing less cataract and IOP rise. However, they are also more transient in effect and do not have any significant anti-inflammatory effect to treat concomitant uveitis [38–41]. In a randomized clinical trial of 31 eyes comparing use of 1–3 intravitreal bevacizumab injections to 1–3 injections of IVTA for refractory uveitic ME, the IVTA group did manifest better control of leakage and CSMT reduction [42].

Diclofenac 500 µg/0.1 ml has been previously administered safely intravitreally in the eyes of some patients with uveitic ME with no signs of ocular toxicity at 8-week follow-up. However, when comparing IVTA and intravitreal diclofenac (IVD) head-to-head, it was found that while IVTA significantly reduced CSMT and increased mean BCVA at 6 and 24 months of study, the IVD group did not reach any statistically significant changes in these parameters [43]. Further study on intravitreal NSAIDs are needed to establish efficacy for uveitic ME.