Short-term Safety and Efficacy of Intravitreal Triamcinolone Acetonide for Uveitic Macular Edema in Children

Objectives  To evaluate the short-term safety and efficacy of intravitreal (IV) triamcinolone acetonide (TA) for treating pediatric cystoid macular edema (CME) secondary to noninfectious uveitis.

Methods  A retrospective noncomparative interventional case series. The medical records of 15 consecutive children (16 eyes) with uveitic CME treated with IVTA (2 or 4 mg) were reviewed. Data collected included details of uveitis, CME, visual acuity, intraocular pressure, and cataract development. The median follow-up time was 16 months (range, 9-36 months).

Results  Resolution of CME was achieved in all of the treated eyes. The median time taken for CME to resolve was 3 weeks (range, 1-24 weeks). The mean improvement of visual acuity after IVTA was 0.6 logarithm of the minimum angle of resolution. Following initial response to IVTA, CME relapsed in 5 eyes (31%) after a median time of 7 months (range, 3-13 months). The most common adverse effect was increased intraocular pressure, with an increase of more than 15 mm Hg in 5 eyes (31%). Steroid-induced cataract was observed in 6 of 11 phakic eyes (55%).

Conclusions  We found that IVTA is efficacious in the treatment of uveitic CME in children and results in CME resolution and visual acuity improvement. As in adults, treatment in children may be associated with elevated intraocular pressure and cataract.

Intraocular inflammation is uncommon in children and accounts for between 5% and 10% of all uveitis cases seen in tertiary referral hospitals.^1,2 However, pediatric uveitis presents a distinct challenge for ophthalmologists because of the associated sight-threatening complications and age-related treatment. Cystoid macular edema (CME) is one of the major causes of ocular morbidity in these children,² and although it generally responds to systemic corticosteroids and steroid-sparing drugs, long-term treatment in children is associated with marked adverse effects such as growth retardation.^2-4 Intravitreal (IV) triamcinolone acetonide (TA) has recently been advocated in the treatment of adult uveitis and other intraocular exudative and proliferative disorders.^5-13 Higher vitreous concentration of TA can be achieved through the IV route as compared with delivery to the periocular space.¹⁴ In this study, we describe our experience with the use of IVTA in a series of children with CME secondary to noninfectious uveitis.

A case-note review was conducted with consecutive children (aged ≤ 16 years) treated with IVTA for CME secondary to noninfectious uveitis who were seen at tertiary referral uveitis clinics of 3 of us (J.R.G., P.J.M., and S.L.). Details of the uveitis, CME, IVTA injection, visual acuity (VA), macula, intraocular pressure (IOP), and lens status were collected. There were 15 children (16 eyes) in this study, 10 girls and 5 boys, with a mean age of 10 years (range, 4-16 years). The median follow-up time was 16 months (range, 9-36 months). All of the eyes had VA of 0.3 logarithm of the minimum angle of resolution (logMAR) or worse and clinically diagnosed CME. Macular edema was confirmed by optical coherence tomography (OCT) and/or fluorescein angiography in all but 1 child (case 3) who was too young to cooperate with the imaging procedure but in whom there were obvious clinical signs of CME. One eye (case 12) had a preexisting inferior temporal branch retinal vein occlusion at the time of referral. Ten eyes (9 patients) treated with IVTA had shown inadequate response to periocular corticosteroids and/or systemic immunosuppression. Two eyes (left and right eyes in case 11) had recurrent CME after initially responding to an orbital floor steroid injection. In 4 further cases, primary IVTA therapy was considered for marked unilateral CME. None of these eyes had a history of previous steroid-induced ocular hypertension, but 1 eye (case 10) was receiving topical ocular hypotensive agents for hypertensive uveitis. Five patients (5 eyes) were receiving systemic immunosuppressives at the time of IVTA injection. This was to control the systemic features of the disease in 2 patients, for uncontrolled posterior uveitis in 2 patients, and for CME in association with systemic features of juvenile idiopathic arthritis in 1 patient. Pretreatment characteristics of the cases are shown in Table 1. All of the eyes had a complete ophthalmic examination including refraction. Visual acuity was measured using a logMAR chart or appropriate VA tests with VA converted to logMAR retrospectively for the younger children.

Informed written consent was obtained from children's parents after thorough discussion of the benefits and risks of the procedure. All of the injections were performed under general anesthesia and using a standard aseptic technique. A dose of 4 mg of IVTA in 0.1 mL (Kenalog; Bristol-Myers Squibb, Middlesex, England) was injected in all but 2 eyes (cases 4 and 7) that were treated with only 2 mg of IVTA (in 0.05 mL). We used commercially available TA and injected it along with its alcohol vehicle. The injection was placed 4 mm posterior to the limbus (3.5 mm if pseudophakic and 3 mm if aphakic) using a 27-gauge needle. Fundus indirect ophthalmoscopy was performed immediately after the procedure to confirm central retinal artery patency. In 4 eyes, IVTA was injected in combination with other procedures including lensectomy, pars plana vitrectomy, and panretinal photocoagulation as indicated. Four (25%) of the studied eyes were vitrectomized at the time of IVTA injection. Postoperatively, all of the eyes were examined at weekly intervals for the first month, then every 1 to 4 months according to clinical need.

The mean data values were compared using paired and independent samples t tests. All of the analyses were conducted using SPSS 10.0 for Windows statistical software (SPSS, Inc, Chicago, Illinois).

Resolution of CME was achieved in all of the eyes following treatment with IVTA, and this was documented by reduction of the central retinal thickness on OCT and/or by fluorescein angiography in most eyes. Macular edema completely resolved in all but case 15, which showed a partial reduction of macular thickness on OCT from 516 μm to a minimum thickness of 397 μm 24 weeks after IVTA. The time taken for CME to resolve after IVTA ranged from 1 to 24 weeks, with a median of 3 weeks. After initial resolution of CME with IVTA treatment, CME recurred in 5 of 16 eyes (31%), with a median time to relapse of 7 months (range, 3-13 months). Figure 1 shows CME relapse with time. Further treatment was considered in only 4 cases that showed improvement of vision with the CME resolution (Table 2).

Treatment with IVTA resulted in a statistically significant improvement in VA, from a mean (SD) of 1.0 (0.5) logMAR to 0.5 (0.3) logMAR (P < .001, paired samples t test). The mean (SD) VA improvement following IVTA was 0.6 (0.5) logMAR. There was no difference in outcome between nonvitrectomized eyes (mean [SD] VA gain of 0.6 [0.5] logMAR) and vitrectomized eyes (mean [SD] VA gain of 0.6 [0.5] logMAR) (P = .92, independent samples t test).

Twelve eyes (75%) achieved VA of 0.2 logMAR or better following IVTA, whereas VA remained unchanged in 4 cases despite CME resolution. Of eyes with improved VA, 11 (92%) achieved clinically significant (at least 2-fold) improvement in the visual angle (0.4 logMAR) (Table 2). The median time to best VA following IVTA was 4 weeks (range, 1-136 weeks). In the 12 eyes that had demonstrated initial visual improvement following IVTA, 4 eyes experienced CME recurrence that was associated with a statistically significant decrease in VA from a mean (SD) of 0.6 (0.4) logMAR to 0.8 (0.4) logMAR (P = .005, paired samples t test).

Of 5 cases with unilateral uveitis in patients younger than 8 years, 2 eyes with persistent poor VA despite CME resolution appeared to be amblyopic and underwent occlusion therapy. In 1 case (case 7), CME resolved completely 6 weeks after IVTA, with improvement of VA from 1.0 to 0.6 logMAR and a final VA of 0.2 logMAR. Amblyopia therapy was also undertaken in case 3 (VA, 1.0 logMAR), but vision did not improve any further.

By the end of the 15 months of follow-up, 9 eyes (56%) showed VA improvement of 0.4 logMAR or more, 6 eyes (38%) had the same level of VA or showed improvement of less than 0.4 logMAR, and 1 eye (6%) had worse VA compared with the baseline VA before the initial IVTA injection. Table 2 shows the VA of each case at the time of final review.

The mean (SD) IOP increased from the baseline value of 12.56 (3.24) mm Hg (range, 5-19 mm Hg) to a maximal value of 22.62 (12.21) mm Hg (range, 10-44 mm Hg) following TA treatment. Figure 2 shows each eye's baseline IOP and maximal IOP after injection. The median time to elevation in IOP was 3 weeks (range, 1-22 weeks), which coincided with the time to maximal IOP reached after IVTA. There was a TA-induced IOP increase greater than 15 mm Hg in 5 eyes (31%), between 6 and 15 mm Hg in 3 eyes (19%), and less than 6 mm Hg or no IOP elevation in 8 eyes (50%). All of the eyes with IOP elevated more than 6 mm Hg were satisfactorily managed with topical ocular hypotensive agents alone except for 4 eyes with IOP increases greater than 15 mm Hg that also needed short-term treatment with oral acetazolamide. One of these eyes (case 14) required a mitomycin C–enhanced trabeculectomy 39 weeks after TA treatment. Eight eyes (50%) were treated with ocular hypotensive therapy, with a median treatment duration of 33 weeks (range, 1-69 weeks). At a median follow-up time of 15 months, no eye developed a change in optic disc appearance by fundus examination and IOP had normalized without medical treatment in all but 4 eyes (median treatment duration in these eyes was 48 weeks).

At the time of IVTA injection, 11 eyes (69%) were phakic, 1 (6%) was pseudophakic, and 4 (25%) were aphakic (of which 1 had undergone lensectomy and IVTA injection). All of the phakic eyes had a clear lens at the time of IVTA injection except for 2 eyes that had some posterior subcapsular cataract. Four eyes with a clear lens developed posterior subcapsular cataract 5 to 16 months after IVTA injection, and it rapidly progressed to dense intumescent cataract with significant reduction of vision in 2 eyes. Increased lens opacification was also seen in eyes that had preexisting posterior subcapsular cataract.

Phacoemulsification and intraocular lens implantation surgery were indicated in 4 eyes (cases 1, 5, 10, and 14) with visually significant TA-related cataract. Posterior capsule rupture was not encountered in any of the cases during surgery. The intraocular lens was placed in the capsular bag in all but 1 eye (case 1) that developed a radial tear in the anterior capsule and had the lens implanted in the ciliary sulcus.

No IVTA-treated eyes developed vitreous hemorrhage, retinal tear or detachment, or endophthalmitis during the study period.

A change in the dose of immunosuppressives was made in 1 of the 6 eyes receiving systemic therapy after IVTA treatment. In this eye (case 15), reduction in the oral prednisolone dose and eventual cessation of cyclosporine were possible with no relapse of CME or uveitis during the follow-up period.

One eye (case 3) not previously receiving systemic treatment was started on oral prednisolone in conjunction with vitrectomy surgery for recurrent vitritis and CME after IVTA but went on to develop phthisis bulbi after the surgery.

Four eyes received a second IVTA injection. Two eyes (cases 1 and 15) had repeat treatment for recurrent CME, resulting once again in CME resolution and visual improvement within 4 weeks. In the remaining 2 eyes (cases 10 and 14), repeat IVTA was undertaken in the absence of macular edema 14 to 16 months after the initial injection. In case 10, repeat IVTA combined with pars plana vitrectomy and panretinal photocoagulation were undertaken for active Eales disease. In the second case (case 14), TA was used at the time of phacoemulsification surgery to prevent postoperative reactivation of intraocular inflammation. Only 1 of the retreated eyes (case 1) experienced an IOP increase of more than 6 mm Hg after retreatment. This eye also had a short-lived increase in IOP after the first injection. Ocular hypertension induced by TA was also seen in another eye (case 14) after the first treatment, but the IOP was surgically controlled several months before repeat IVTA.

To our knowledge, this is the first study investigating IVTA treatment in children. Results suggest that IVTA is effective in children with sight-threatening uveitic CME and can avoid adding or increasing the dose of systemic immunosuppressives. In all of the treated eyes, resolution of CME was achieved with IVTA and was documented by OCT and/or fluorescein angiography in most eyes. This was associated with improvement in VA in 12 eyes (75%), in keeping with findings of a similar study of adult uveitic CME treated with IVTA.⁵ In most patients, the time to best VA was associated with the time to CME resolution.

In eyes with long-standing macular edema, the resolution of uveitic CME may not be necessarily accompanied by functional improvement in VA as degenerative changes to the retina may ensue.^5,7,8,15 In such cases, other causes of loss of vision such as cataract, epiretinal membrane, or optic nerve damage need to be excluded. Furthermore, in the context of pediatric uveitis, it is also important to consider the presence of refractive errors and amblyopia in young children with unilateral or asymmetrical disease as demonstrated in our study. Damage of the neurosensory retina was presumably the cause of poor vision after recovery of CME in 4 eyes in which we failed to find any clinically apparent cause. However, no foveal thinning was shown on OCT and electrophysiological studies were not done in these eyes (Table 2). In this series, we were unable to investigate any correlation between the duration of CME and the responsiveness to IVTA. Most of the eyes had CME at the time of referral to the clinics, and it was not possible to precisely determine the duration of CME prior to IVTA treatment. In adults with uveitic macular edema, 2 recent retrospective studies have provided mixed results as to the effect of the duration of macular edema on the response to IVTA. Androudi et al⁷ found no direct relationship between CME duration and CME resolution or VA improvement. However, Kok et al⁵ reported that most visual improvement was seen in eyes with CME of less than 1 year's duration.

The tendency for macular edema to recur following IVTA is well known, has previously been documented in uveitic patients,^5-10 and is related to the pharmacokinetics of TA.^16,17 In this series, only 5 (31%) of the treated eyes had recurrent CME, which is consistent with the results of Kok et al⁵ but less than the rate that has been reported in other series of IVTA for uveitic CME in adults.^5,7,8

The response to IVTA was previously reported as reduced in adult vitrectomized eyes as compared with nonvitrectomized eyes and has been attributed to the rapid elimination of TA from the vitreous cavity following vitrectomy.^16,17 However, although 4 (25%) of our eyes were vitrectomized, no statistical difference in the mean gain of VA was found between vitrectomized and nonvitrectomized eyes and only 1 of the 4 eyes with recurrent CME was vitrectomized. It is possible that the pharmacokinetics of TA clearance following vitrectomy are different in adults and children, which may explain the greater efficacy of TA seen here in vitrectomized eyes as compared with that in previous adult studies. However, the numbers in this series are small, making it difficult to draw any firm conclusion in this regard.

There is now a growing body of evidence that younger age is a prominent risk factor for corticosteroid-induced elevation of IOP.^5,18-20 In children, the ocular hypertensive response to topical corticosteroids is more rapid, is more severe, and occurs more frequently.²¹ In this study, TA-induced ocular hypertension greater than 15 mm Hg was seen in nearly one-third of the cases, a finding in accordance with previous work on adults with uveitis treated by IVTA.^5-8 The onset of a TA-induced increase in IOP occurred earlier in children as compared with adults.²² In most children, an initial increase in IOP was observed at 3 weeks. Intraocular pressure was controlled by topical therapy alone in 5 of 8 eyes that required IOP-lowering treatment. It is obviously important to monitor IOP closely in these patients as the induced IOP elevation may occur at any time.^20,22,23

As previously reported in adults,^5,18,24 we found that children without increased IOP after IVTA treatment are less likely to experience increased IOP with repeat injections, but it is still difficult to be sure given the number of cases that received reinjection (4 eyes) in the study and the fact that no eye received more than 2 treatments. The use of IVTA should be avoided in children who have previously exhibited a corticosteroid-induced elevation of IOP unless the IOP has been surgically controlled.

In this series, corticosteroid-related cataract was seen in 6 of the 11 phakic eyes (55%), an incidence higher than that reported for adults with uveitic macular edema.^5,7 However, 4 of the study eyes were already aphakic and a further eye had a posterior chamber lens implant prior to IVTA treatment. We hypothesize that this high incidence of TA-related cataract may relate to slow dispersion of the TA through the nonsyneretic pediatric vitreous with high levels remaining near the posterior pole of the lens.

The results of our study need to be interpreted with caution because of the bias associated with nonrandomization and its retrospective design. In addition, the number of eyes reviewed was small and the follow-up period was variable. We also feel that in eyes with CME and active inflammation, visual loss cannot be attributed to only macular edema because severe vitritis can itself affect vision. However, we remain encouraged by the fact that IVTA can improve both active inflammation and CME. Variation in the dose of IVTA used is another drawback of our work and is due to a change in protocol with children younger than 5 years receiving half the dose of TA. Although this was used in only 2 eyes, it did not seem to affect the outcome because these eyes had no recurrence of their macular edema by the last visit. Finally, our study was not quantitative in objective terms because we performed OCT and/or fluorescein angiography evaluation for the study eyes.

In summary, these data suggest that IVTA is worth considering in children with uveitic CME either as a first-line treatment in severe unilateral disease or as a second-line agent in refractory cases. The most important adverse effects of IVTA in children appear to be elevated IOP and rapid progression of cataract. Further prospective trials are needed to ascertain the safety and efficacy of IVTA in children.

Correspondence: Susan Lightman, PhD, FRCOphth, Department of Clinical Ophthalmology, Institute of Ophthalmology, Moorfields Eye Hospital, City Road, London EC1V 2PD, England (s.lightman@ucl.ac.uk).

Submitted for Publication: April 1, 2007; final revision received July 22, 2007; accepted July 30, 2007.

Author Contributions: Dr Sallam had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: None reported.

Funding/Support: This work was supported by grants 22928 and 351030 from the National Health and Medical Research Council of Australia (Dr Chang).

Previous Presentation: This article was presented in part as a poster at the 2006 Annual Meeting of the American Academy of Ophthalmology; November 13, 2006; Las Vegas, Nevada.